key: cord-023055-ntbvmssh
authors: nan
title: Immunogenicity
date: 2004-02-19
journal: J Cell Biochem
DOI: 10.1002/jcb.240410506
sha: 
doc_id: 23055
cord_uid: ntbvmssh

nan

Ia moyecules with respect to their roles as peptide receptors and target structures for TCR interaction. Particular attention has been paid to distinguishing between local and distant effects of amino acid substitutions on Ia function and to determining which residues interact with peptide antigen and which (if any) with the TCR. This ex erimental approach has led to the identification of several regions of the pofvorphic amino-terminal domains of the a and p chains as playing critical roles in chain-chain association and quaternary Ia conformation. The a1 and p l putative helical regions have been found to have distinct degrees of structural lability, with the a1 helix showing much greater susceptibility to conformational change due to allelic variation in other re ions of the molecule. Allelically polymor hic residues in the a1 and p 1 domainstave been shown to play important roles in &e activity of the assembly/folding control regions, and hence, analysis of local binding roles of specific residues in Ia molecules must take this additional effect of substitutions at these positions into account. By controlling for large scale conformational effects, individual residues in the p chain have been assigned to desetopic ( eptide interaction) and histotopic (TCR interaction) roles. In the cytochrome c molel, a putative peptide bindin "pocket" involvin residues from both the postulated p l a helix and also the p-stran% floor has been defined, residues controlling both the extent of binding and the orientation of the bound peptide have been identified, and at least one residue with TCR interaction potential without obvious peptide binding properties has been localized. Combining these data with those of other investigators leads us to propose a general model of class I1 MHC structure-function relationships. We have shown previously that memory B cells transferred into K-allotype distinct congenic rats in the absence of any priming antigen are deleted from the adoptive host within a matter of weeks (half-life of 1-2 weeks). In contrast co-injection of antigen with the cells facilitates their survival and the maintenance of a donor response for periods in excess of one year. In the experiments reported here we ask if the persistence of T cell memory is also dependent on antigen. 10. carrier (KLH) primed T cells were transferred in the presence or absence of antigen into irradiated, K-allotype distinct adoptive host. A t various times after transfer these rats were injected with 2x10' hapten-carrier (DNP-KLH) primed B cells together with 50 Ig of soluble DNP-KLH. This limiting number of B cells makes a secondary type response only if carrier-specific memory T cells survive in the adoptive host. We found that already at 6 weeks following transfer without antigen, no memory T cell help was available for these B cells. In contrast T cells transferred together with 10 ~g KLH provided help for secondary type donor responses at 6 and 12 weeks after transfer. We conclude that longterm memory at both the T and B cell levels does not reside with small, very long-lived, resting cells but. with active clones that are maintained by small amounts of antigen that may persist for long periods. Once antigen is lost from lymphoid tissues both T and B cell memory wanes within a relatively short time. T cells recognize antigen in the form of short peptides associated to class I or class I1 MHC molecules. Each MHC molecule has the ability to bind a large number of peptides and peptides with unrelated sequences can compete for binding to the same MHC molecule, as well as in vitro. In vivo competition strictly correlates with the capacity of the competitor peptide to bind to the MHC molecule presenting the antigenic peptide and its extent dependes on the molar ratio between antigen and competitor.

competition among different peptides derived by processing of hen egg-white lysozyme

(HEL) appears to exert a major influence on the immunodominance of antigenic determinants recognized by T cells. Thus, the H L peptides 1-18 and 25-43 are both generated by HEL processing and are both able to bind to the I-E molecule but only 1-18 becomes immunodominant because it has th ability to compete in vivQ with other HEL peptides, such as 25-43, for the available sites on the I-E molecule. However, two immunodominant T cell epitopes, such as those in HEL peptides 51-66 and 112-129, both interacting with I-Ak molecules, do not compete with each other when injected together at equimolar concentrations. Such a coexistance is anticipated between peptides that bind with relatively high affinity to the presenting molecule and thus have both the chance to occupy a number of binding sites sufficient for T cell activation. r v s e III xen ic tr lantation. In v i m lnvesti$ation uslng mocloml a n t M i e s r e 4 E -t E X e y skin gracs-val on m i c e w a s significantl pr:lrd l g anti-antihdy trea-t but n o t a b anti-antibody: W i d e r the saue animals but a n t i C D 4 antibody did prolong minor a n t w -d i allqrdts. In v i m studies r m that primary proliferation a n F Z . 2 prcdwtion & -F I cells in response to mmkey stimulators was weak conpared to allogeneic reqonses. secondary responses t o xenogeneic stinulation were strong after in v i m priming but required the presence of responder NC's. Assays for c y G t s T cell effectors in m i c e which had rejected monkey skin revealed few such cells. zhese results est that widely d i a t e xencgeneic processing and presentation, Since xenogeneic antigens require that such presentation be in association with the.= antigens of regponder Apc's, the xenogeneic r a f t s have a functional similarity t o AFf2 leted allografts. shoved t h a t f e t a l r e n a l and f e t a l and p o s t n a t a l testis a l l o g r a f t s survived longer than corresponding a d u l t t i s s u e i n non-immunosuppressed outbred r a t hosts. The c u r r e n t study a s k s v h e t h e r t h e d i f f e r e n c e i n s u r v i v a l betveen r e n a l and t e s t i c u l a r g r a f t s and between g r a f t s of d i f f e r e n t ages is r e l a t e d t o d i f f e r e n t i a l t i s s u e expression of Class I and Class I1 mRNA t r a n s c r i p t s or s u r f a c e antigens. and i f t h e s e p a t t e r n s change w i t h t r a n s p l a n t a t i o n . congeneic mice w e found t h a t prolonged s u r v i v a l of C57BL/6 f e t a l r e n a l (n=42; p<0.008) and f e t a l (n=14; ~( 0 . 0 5 ) and p o s t n a t a l (n= 8; ~( 0 . 0 5 ) t e s t i s mouse a l l o g r a f t s t r a n s p l a n t e d beneath t h e r e n a l c a p s u l e of a d u l t r e c i p i e n t BIO.A mice and t h i s s u r v i v a l c o r r e l a t e s i n v e r s e l y w i t h t h e expression of Class I and Class I1 mRNA (northern a n a l y s i s ) and p r o t e i n s (immunohistochemistryy) and t h a t both p r o t e i n and mRNA increased throughout ontogeny f o r both t h e testis and kidney. After t r a n s p l a n t a t i o n t h e r e vas a marked i n d u c t i o n of MHC mRNA t r a n s c r i p t s f o r both testis (n=207) and kidney (n=320). Implanted f e t a l kidney t i s s u e t h a t survives. however. f a i l e d t o express d e t e c t a b l e MHC p r o t e i n , i n d i c a t i n g t h a t some p o s t -t r a n s c r i p t i o n a l modification i n t h i s t i s s u e occurs. t o a f f o r d it p r o t e c t i o n from r e j e c t i o n . Implanted testis shoved i n d u c t i o n of both mRNA and p r o t e i n v e l l above i t s much lower baseline. i n d i c a t i n g t h a t i t s r e g u l a t i o n , i n c o n t r a s t t o t h e kidney may be t r a n s c r i p t i o n a l . Thus t h e f e t u s may lower t h e MHC burden as a s t r a t e g y t o escape r e j e c t i o n e i t h e r by p o s t t r a n s c r i p t i o n a l modification of p r o t e i n expression a s i n t h e kidney or by t r a n s c r i p t i o n a l modification of mRNA as i n t h e testis. Culture of thymus tissue in 2-deoxyguanosine (2dGua) is thought to reduce tissue immunogenicity by selectively depleting highly immunogenic, thymic immigrants of bone marrow origin. In the mouse 2dGua treated thymus tissue survival is markedly enhanced compared to untreated tissue when transplanted under the kidney capsule of allogeneic recipients. These experiments were repeated in the rat. As expected, strain DA neonatal thymus tissue was rejected when transplanted under the kidney capsule of normal allogeneic strain PVG rats.

Surprisingly. acute rejection occurred even when the tissue was cultured for 14 days in 4 mM PdGua (3x the effective dose in mice). By in vitro criteria this dose was very effective in destroying thymocytes. To test whether residual marrow derived cells that escaped PdGua treatment were responsible for inducing rejection we "parked" the 2dGua treated DA tissue in T cell depleted PVG rats. Our working hypothesis was that the few remaining donor derived cells of marrow origin would be overgrown by host type cells. When PdGua-treated DA thymus tissue was transplanted into T cell depleted PVG recipients graft rejection did not occur. However DA PdGua treated thymus tissue, parked for as long as 200 days in T cell depleted PVG rats, was acutely rejected when retransplanted into normal PVG recipients. We interpret these results to suggest that rat thymic epithelium devoid of marrow derived cells is innately immunogenic. c 104

Corinne Amiel, Violaine Gugrin, Thierry May, Philippe Canton, Gilbert C Faure, Laboratoire d'Immunologie and Maladies Infectieuses, CHU de Nancy, Facult6 de Mgdecine, 54500 Vandoeuvre les Nancy, France. LFAl is a dimeric membrane molecule composed of a specific alpha chain (CDlla) and a beta chain (CD18) common to three members of the LFA family. LFAl is physiologically expressed on all white blood cells, while other molecules of the LFA family (with CDllb and CDllc alpha chains) are restricted to cells of myeloid lineage. A defective expression of LFAl has been described in some congenital immune deficiency and in AIDS. We investigated the LFAl defect on peripheral blood lymphocytes from 100 HIV+ patients. Three different monoclonal antibodies were used, respectively directed to chain-specific epitopes of CDlla (SPVL7, Sanbio) and CD18 (IOT18, Immunotech) and to a conformational epitope involving both chains (IOT16, Immunotech). Cell suspensions were stained in indirect immunofluorescence and a flow cytometer (Epics Profile, Coultronics) was used to assess the percentages of stained cell, the fluorescence intensity and the shape of fluorescent peaks. Our data suggest that LFAl expression is impaired in HIV+ patients both through the quantitative expression of each chain and through conformational alterations. The adhesion molecule LFA-1 is known to be important in antigen presentation. We have previously shown that both monocyte and T cell LFA-1 play a role in the interaction between these two cells (EJI D; 943, 1987) . Antibody to ICAM-1 (known to act as a ligand for LFA-1) also inhibits antigen presentation, although ICAM-1 is not thought to be expressed on resting T cells (EJI 18: 35, 1988) . We have looked at the expression of ICAM-1 on T cells after incubation with 12 cytokines and found that only IL-2 consistently effects an increase in both the percentage of ICAM-1 positive cells and in the level of expression. In addition we have found that a proportion of resting T cells express very low levels of ICAM-1. Double labelling experiments have shown that these cells are part of the memory T cell population as defined by antibodies to UCHLI, LFA-3 and LFA-1, and furthermore that ICAM-1 negative cells are unable to respond to to antigens such as PPD and Flu but are able to respond to PHA. This suggests that ICAM-1 represents an additional marker on the memory T cell population which more precisely defines the subset able to respond to recall antigens ICAM-1 EXPRESSION ON T CELLS, Anne-Marie Buckle and Nancy Hogg, Macrophage Lab. ICRF, Lincolns Inn Fields, London, WC2A 3PX, U.K.

IMMUNIZATION, Francis R. Carbone and Michael J. Bevan, Department of Immunology, Research Institute of Scripps Clinic, La Jolla, CA 92037. CTL recognize peptide forms of processed, foreign antigens in association with class I molecules of the MHC and are usually directed against endogenously synthesized "cellular antigens" such as those expressed by virusinfected cells. In vifro studies have shown that small exogenous peptides can directly associate with class I molecules on the cell surface and mimic the target complex derived by intracellular processing and presentation. We have recently generated OVA-specific, H-2Kb-restricted CTL by immunizing C57BL/6 mice with a syngeneic tumor line transfected with the OVA cDNA. The CTL recognize the OVA transfectant EG7-OVA and the synthetic peptide OVA but fail to recognize the native protein. We reasoned that given the potential for direct peptide/class ?'&$%ation observed in vifro, OVA2s,-2ra may induce CTL after in vivo priming. However, we found that this is not the case. OVA,,,-,,, and peptides of increasing lengths up to which are all able to form the target complex in vitro, are inefficient at priming ~% -%~~ specific ?h%sponses following intravenous injection. This is also true for both native and denatured OVA. In contrast to these results, the synthetic peptide OVA22g:z76 corresponding to a peptide in a partial tryptic digestion of OVA can efficiently prime C57BL/6 mice in vrvo following intravenous injection.

This peptide elicits CTL which appear identical to those derived from animals immunized with syngeneic cells producing OVA endogenously. It is now well established that human T lymphocytes can be activated via the T cell specific CD2 antigen. in order to determine if a factor@) other than the single CD2 polypeptide is involved In CD2 mediated signal transduction, we have staMy transfected murine L cells with the human CD2 cDNA. We report that such transfectants expressed hah levels of CD2 at the cell surface. formed sheep erythrocyte rosettes and expressed the three CD2 epitqm previously defined on human T lymphocytes, including the "activation associated' T I 13 epitope. The latter observation unequivocally demonstrating that expression of the TI13 epitope, in contrast to a previous report, is entirely independent of T cell specific factors. Combinations of CD2 mAbs that are potent stimulators of human T cells. however, failed to elicit either an increase In the concentration of intracellular free calcium or augment [3H]-thymidine inmrporatbn in the transfectants. These results provide both formal identification of the CD2 cDNA and dearly demonstrate that the single CD2 polypeptide expressed in an heterokgous cell system devoid of T cell specific factors, cannot alone transduce intracellular signals in response to stimulatory combinations of CD2 mAbs. The results are therefore consistent with the notion. that the functional CD2 antigen expressed in human T lymphocytes, requlres the association of another, as yet, undefined factor@). This conclusion was based on several lines of evid nce incl ding the observation that mAbs specific for the class I a3 domain of either H-ZL8 or L 2 D B interfered with t e generation of CD8-dependent (low substitution at position 227 in the a3 domain are not lysed by CD8-dependent primary CTL but are lysed by secondary CD8-independent (high affinity) CTL generated in the presence of antibody to the a3 domain. populations of CTL. We have isolated and characterized a C D W , CD4-Da-specific CPL line. This line is CD8-independent and is capable of lysing the addition, we are currently generating clones from primary $-specific CTL cultures to obtain CD8-dependent (low affinity) CTL. directed rnutagenesis are being tested with the CD8-dependent and CD8-independent clones to define additional residues important for CD8 recognition. The comparison of CTL clones with different CD8 dependencies will allow us to more precisely define the role of CD8 in T cell recognition. 

Percolle from the buffy coat of one unit ot blood. These cells (=400x10 ) are introduced into a Curame 5000 elutriation centrifuge (rotor speed of 3000 rpm; loading flow of 10 ml/min). Nine fractions can be obtained. The first three containing >90% lymphocytes; fraction 4 (3000 rpm-18 ml/min) and fraction 5 (2900 rpm-18 ml/min! contain both lymphocytes and monocytes and the next three fractions contain >90% monocytes; the finat fraction (rotor off) contains monocytes + granulocytes. Cells from each fraction (5x10 /well) are incubatee for five days with tetanus toxoid (1.5LF/well) and an enriched population of T cells (5x10 /well). Quadriplicate samples are then pulsed for 16 hours with 'H methyl thymidine. Maximum APC activity is found in fractions 4 and 5 representing 4 to 7% of the mononuclear cells. APC activity for these two fractions can be further purified by selective absorption of the cells onto gelatin coated surfaces that have been preincubated with plasma. The non adherent lymphocytes are rgmoved after two hours. After overnight incubation spontaneously released cells (1-5x10 ) can be harvested which have a higher APC activity than cells rotated by elutriation alone. These methods are now highly reproducible in our laboratory, so we can now begin to characterize and study these cells. The male s p e c i f i c H-Y a n t i g e n h a s been shown t o behave as a minor Histocompatibility a n t i g e n in man and mouse. I n t r a n s p l a n t a t i o n , male t i s s u e may t r i g g e r t h e c l o n a l expansion of H-Y reactive HHC r e s t r i c t e d effector cells of female o r i g i n .

Although male epidermal cells (EC) can induce an anti-H-Y T cell response in female mice, so far in v i t r o techniques have f a i l e d t o i d e n t i f y t h e cell-defined H-Y a n t i g e n on murine EC (1). Here w e developed a 51Cr release assay t o use human c u l t u r e d k e r a t i n o c y t e s (K) as t a r g e t cells for HLA-A2 specific and MA-A2 r e s t r i c t e d H-Y s p e c i f i c T cell clones. HLA-A2+ but n o t H L A -A T K were l y s e d by anti-MA-A2 CTLs i n a dose dependent manner. Low but d e t e c t a b l e l e v e l s of anti-H-Y k i l l i n g were found a g a i n s t MA-A2+ male K b u t n o t a g a i n s t H L A -a T male or HLA-A2+ female K. Both l e v e l s of a l l o r e a c t i v e and H-Y s p e c i f i c l y s i s were d r a m a t i c a l l y enhanced after exposure of K t o IFN gamma. These r e s u l t s s t r o n g l y suggest t h a t h m a n male s k i n cells are d i r e c t l y s u s c e p t i b l e for H-Y d i r e c t e d T c e l l k i l l i n g through t h e expression of f u n c t i o n a l H-Y/HLA complexes on t h e i r cell s u r f a c e . I n view of t h e s e f i n d i n g s , t o g e t h e r w i t h our r e c e n t s t u d i e s on t h e expression of H-Y CTL determinants on h m a n hematopoietic p r o g e n i t o r c e l l s (21, t h e r o l e of H-Y a s a t a r g e t s t r u c t u r e f o r c e l l mediated immunity i n o l i n i c a l t r a n s p l a n t a t i o n should be s e r i o u s l y taken i n t o account.

1. Steinmuller D. and Burlingham W.J. T r a n s p l a n t a t i o n 19@4,37,1,22.

2. Voogt P.J., Goulmy E., Fibbe W.E., e t a l . J . Clin. Invest. sept.1988. c 116 DIPHTERIA TOXOID (DT) PRESENTATION BY HLA DR7 TRANSFECTED MURINE FIBROBLASTS Bismuth, Laboratory of C e l l u l a r and T i s s u l a r Immunology, CHU P i t i e S a l p B t r i B r e , P a r i s , France and Veterans Medical C e n t e r , Iowa c i t y , USA. L t r a n s f e c t a n t s e x p r e s s i n g s i n g l e type of human MHC c l a s s I1 molecules produced by DNA Conjugate formation has been studied with cloned T cell lines and a B cell hybridoma and with T cells and B cells from normal mice. Resting T cells and B cells do not form appreciable numbers of conjugates but conjugates are formed between T cells stimulated with alloantigen for four days and B cells activated by 24 hour culture with LPS. Irrelevant lymphocytes do not affect the rate of specific conjugate formation in suspensions of cells agitated by gentle rocking but impair conjugate formation when cells are allowed to settle in round bottom tubes. In further experiments, it was shown that the conditions for the induction of lymphokine secretion by the T cell were not indentical to the conditions for conjugate formation.The significance of these and other observations for the interaction of T cells and B cells in vivo will be discussed. of the primary mixed leukocyte reaction (MLR) and that this reaction occurs in multicellular dendritic cell-CD4+ T cell clusters [Cellular Immunology 111, 183-195(1988) 

Dendritic cells are able to Contact, cluster, and retain allogeneic T cells and induce these alloreactive cells to proliferate and divide. tions labeled with a vital flvorescent dye, we show that only dendritic cells efficiently form stable clusters. Labeled monocytes and B cells do not form clusters with T cells. When labeled monocytes and unlabeled dendritic cells are used to stimulate T cells, unlabeled clusters form. Labeled monocytes do not move into the clusters until the third day of the MLR. Significant levels of IL-2 and a-IFN appear in the culture supernatant by the first or second day. blast transformation by the second day of the MLR as demonstrated by Giemsa staining of cluster cytopreps. also been studied by immunoperoxidase staining.

It is known that human peripheral blood dendritic cells are potent stimulators Using purified dendritic cell popula-

The distribution of certain adhesion molecules within clusters has c1m Microbiology and Immunology, Emory University, Atlanta, GA 30322. Immunization of SJL/J mice with myelin basic protein (MBP) induces the T cell-mediated autoimmune central nervous system disease, experimental allergic encephalomyelitis. Response against a dominant epitope (residues 89-101) leads to disease. Lymph node T cells from MBP-immune mice react against several epitopes in addition to 89-101 indicating that the I-As molecule is able to form immunogenic complexes with several MBP peptides. The question asked in these studies was whether subdominant epitopes from the same molecule would compete with the dominant epitope for binding sites on the I-AS molecule. To address this question two T cell clones, one specific for 89-101 (SP4.2) and a second specific for a second epitope present in peptide 89-170 (SP4.7) were tested for responsiveness when cultured with the dominant epitope alone or with mixtures of peptides containing dominant and subdominant epitopes. Reactivity of SP4.2 against peptide 89-101 was inhibited by peptides 1-37 and 43-88. Reactivity of SP4.7 against peptide 89-170 was not inhibited by peptide 89-101 although peptides 1-37 and 43-88 were inhibitory. Controls indicated that inhibitory reactivity was not due to toxicity at high concentrations of peptides. These findings imply that subdominant epitopes are able to compete with dominant epitopes of MBP for binding sites on I-As molecules. Linda R. Gooding, Frances C . Rawle, David I . Kusher, W i l l i a m S . M. Wold+ and Barbara Knowles*. Department of Microbiology and Immunology, Emory University School of Medicine, Atlanta, GA 30322, 'Institute f o r Molecular Virology, S t . Louis University School of Medicine, S t . Louis, MO 63110 and *The Wistar I n s t i t u t e , Philadelphia, PA 19104. I n several v i r u s systems e a r l y non-structural proteins localized predominantly i n the nucleus of infected cells are major t a r g e t antigens f o r cytotoxic T lymphocytes (CTL). Whether early synthesis o r nuclear l o c a l i z a t i o n are important factors i n immunodominance is not known. W e have recently developed a system f o r studying the CTL response t o human group C adenoviruses i n mice. By us ng both transfected t a r g e t s and virus deletion mutants w e have shown t h a t , response t o wild type AD5. There are two E1A t r a n s c r i p t s , 12s and 1 3 s . which both encode major e a r l y nuclear antigens d i f f e r i n g by a 46 amino a c i d insertion: both antigens are recognized equally w e l l by CTL. The E3 encoded 19K glycoprotein (gpl9K) of Ad5 binds t o MHC c l a s s I antigens i n the endoplasmic reticulum preventing t h e i r translocation t o the c e l l surface and strongly inhibiting l y s i s by Ad5 specific CTL. However, the presence of gpl9K i n the priming v i r u s does not a f f e c t the s p e c i f i c i t y of the CTL generated f o r E l A , so the immunodominance of t h i s protein cannot be due t o the fact t h a t i t is the only major protein synthesized before gpl9K i n the course of infection. Using virus deletion mutants we are investigating whether CTL s p e c i f i c f o r other Ad5 antigens can be induced i n the absence of ElA, and whether E1A is also the dominant antigen recognized i n mice of other MHC haplotypes. respond to antigens present on non-replicative virions. In contrast, we have obtained BALC/c I-Erestricted T hybridomas specific for the neuraminidase (NA) glycoprotein of A/PR8 influenza which recognize infectious, but not non-replicative virus, closely resembling recognition requirements observed for most class I MHC-restricted responses to influenza. Recognition correlated with the rte nova synthesis of viral NA within antigen-presenting cells, but did not depend strictly upon the amount of NA present in cultures, since high NA concentrations could be achieved by addition of non-replicative virus without being stimulatory for NA-specific T cells. Recognition of a neo-antigen was ruled out, since, in high concentration, NA isolated from purified virions, even if reduced and alkylated, was recognized by the T hybridoma clone. Isolated NA was recognized when added to pre-fixed APC, suggesting that this form of antigen was able to bypass the usual processing pathway of exogenous proteins. This suggests that endogenously-synthesized antigen may use different pathways to achieve class 11-associated presentation. T lymphocyte activation is a complex event which is influenced by a variety of distinct cell surface molecules. In order to determine the role of individual molecules in the activation process, we have developed an efficient methodology for generating cell variants in which expression of molecules is selectively inhibited by expression of anti-sense RNA from an Epstein-Barr virus episomal replicon. In a previous study, we reported that marked inhibition of CD8 cell surface expression could be achieved in a human T cell clone using this approach. We have now extended this strategy to another T cell surface molecule, CD2, as a first ste towards ascertaining its role in T cell activation. To this end, we s nthesized a &-mer oli onucleotide corresponding to a sequence in. the 5: end of the c d i n g re ion of human CD'i and inserted it in an anti-sense orientation into this replicon. This a-C%2/REP3 construct was electroporated into Jurkat cells. Analysis of stable a-CD2IREP3 transfectants by immunofluorescence staining and flow cytometry demonstrated complete and selective inhibition of CD2 expression. In contrast to the nontransfected arent, this CD2-variant demonstrated a partial loss in its ability to form conjugates a n 8 to secrete interleukin 2 when stimulated with anti-CD2 monoclonal antibodies. However, stimulation of the CD2-variant with A23187 and PMA did result in interleukin 2 secretion. Several observations suggest that CD8 functions not only as an adhesion molecule recognising MHC class I on the adjacent cells but also potentiate the transducting capacity of the TCR/CDg complex. Comparison of the mouse Ly 2 protein sequence with the homolog rat OX8 and human T8 sequences revealed most highly conserved regions in the membrane and cytoplasmic part of the molecule. The conservation of the transmembrane and cytoplasmic sequences in different species may be significant for the function of the CD8 molecule. In order to initiate the functional dissection of the CD8-molecule we constructed mutations in different parts of the molecule. By transfecting the a and B chain genes donated by a CD8 dependent cytotoxic T cell clone(KB5 C20) into the MHC class I1 restricted agd CD4 T cell hybridoma DO-11.10 we were able to reconstitute the ability to respond to K only if the transfer was done with the Ly 2 molecules (Gabert et al., 1987. Cell, 50. 545-554) .

In this system surface expression of mutated and non mutated Ly-2 molecules were checked by FACS-analysis and the molecuar size of the proteins were analysed by immunoprecipitation with the anti-Ly-2 monoclonal antibody 19/lJ8. Finally functional effects of the mutations were investigated in response towards the K alloantigen. We have simulated graft versus host and host versus graft reactivity in vitro by studying primary anti-minor H responses in a limiting dilution culture system. The ability of BMM and peripheral blood mononuclear cells (PBM) to stimulate and respond in this system were compared by estimating the number of proliferating cells. In GvH-direction the combination of donor-BMM (d-BEIM) and host-PBM (h-PBM) was 2 to 15 times more effective in stimulating proliferation than any other combination; the same applied to the combination h-PBM/d-BMM in HvG-direction.-Using these combinations the median frequency of proliferating cells in GvH-direction was 1/5300 (range 1/59C-1/18100) in 10 pairs, in HvG-direction (7 pairs) 1/2330 (range 1/115-1/6100). 85% of the proliferating cells had the phenotype of mature T-cells.-Using the same combination of responder/stimulator cells we have also estimated the number of cytotoxic cells specific for the HLA-identical target cell. In GvH-direction the median estimate (n=8) was 1/10300 (range 1/66O-1/45700), in HvG-direction (n=5) 1/2250 (range 1/400-1/6500). By split well analysis similar or higher frequencies of cytotoxic cells with specificity for NK-targets were detected (GvHR: 1/5750, HvGR: 1/3620). It was however possible to identify a significant number of minor H-specific clones by segregation analysis; their specificity could be confirmed after clonal expansion. The clones had the phenotype of typical cytotoxic T-cells.-The relevance of the two cytotoxic subpopulations described above to clinical events such as GvHD, graft rejection and relapse needs to be clarified.-

MOLECULAR CLONING OF MURINE ICAH-1, K.J. Horley, B. Baker, and F. Takei, Terry

Pathology, University of British Columbia, Vancouver, B.C., Canada.

We have previously reported a novel cell surface antigen expressed on activated and proliferating murine lymphocytes. The antigen, termed HALA-2, is absent or present at low densities on thymocytes, lymph node cells, and fibroblast cell lines, indicating it is not a universal proliferation antigen. Some cells of the spleen and bone marrow express MALA-2 at a high density possibly representing in vivo proliferation in these tissues. apparent molecular weight of 95-100 kD under both reducing and nonreducing conditions, and is susceptible to Endo F digestion. The monoclonal antibody YN1/1.7 that reacts with this antigen, profoundly inhibits MLR. A XgtlO cDNA library was constructed from NS-1 cells that express a high level of MALA-2, and screened with synthetic oligonucleotides resulting in the isolation of a full length cDNA clone (-3.2 Kb). The cDNA sequence has high homology with the human ICAU-1 sequence, indicating that HALA-2 may be the murine homologue of this characterized protein. Hines, Trudeau I n s t i t u t e , Inc., P.O. Box 59, Saranac Lake, NY 12983 A tumor c e l l l i n e , ET-5, has been derived from an apparent fibrosarcoma t h a t arose i n a C57BL/6 male mouse. antigens. Mice t h a t have r e j e c t e d ET-5 become imnune t o these minor H antigens, judged by accelerated s k i n g r a f t r e j e c t i o n , and t h i s imnunity can be t r a n s f e r r e d t o imnunod e f i c i e n t mice w i t h lymphoid c e l l s . However, spleen c e l l s from mice t h a t have r e j e c t e d According to the widely accepted view, CD2 (T11, sheep erythrocyte receptor) is the first T cell-specific antigen to appear on differentiating thymocytes during ontogeny. It follows that CD2 should be expressed on all immature and mature T cells. Using two-color cytofluorometry I have here identified subsets of CD2-CD3+ T cells both in fetal human thymus or spleen and in adult peripheral blood. CD2-CD3+ T cells constitute 1-25% of fetal thymocytes and 0.1-0.8% of peripheral blood T cells. IL-2-dependent longterm clones of CDI-CDJ+ cells do not react with a panel of monoclonal antibodies (mab) directed against the T1ll, Tlll or T113 epitopes of CD2 and do not transcribe CD2 mRNA. Fetal tissue-derived clones react with the TigammaA mab and thus express a functional TCR gamma chain, while CD2-CD3+ clones from peripheral blood are BHA031+ and express a full-length 1.3 kB TCR C S .RIA.

The clones established here are currently being characterized with respect to functional capacities. I conclude that expression of CD2 is not an absolute prerequisite for the expression of the CD3/TCR molecular complex on human T cells. if they are added after 24 hours. These interactions are bidirectional. since both CDlla and CD18. and t h e i r ligand I-Cam 1. are expressed on the presenting c e l l s as well as the T cells. However, a l l such e a r l y adhesion related events are not bidirectional since anti-CDZ and anti-LFA-3. which are expressed d i f f e r e n t i a l l y on T c e l l s and presenting c e l l s respectively are also effective as inhibitors. antibodies, a n t i CD4 and a n t i CD25 antibodies do not i n h i b i t clustering but do i n h i b i t p r o l i f e r a t i o n , and t h i s i s seen irrespective o f when the antibodies are added i n t o the assay. Our findings suggest t h a t there are two mechanisms involved i n dendritic c e l l -T c e l l interaction, f i r s t l y an inrnediate Cell-Cell adhesion step and l a t e r A secondary signal transduction process possibly mediated v i a cytokines. The q u a l i t a t i v e d i fferences between dendritic c e l l and B c e l l induced i m n o g e n l c i t y may thus l i e i n e i t h e r o f these two steps.

King, Department o f Eathology, The Bland-Sutton I n s t i t u t e . University College and Middlesex School I n contrast, a n t i class I1 M C lmmunogenicity C130 CULTVRED TISSUE IS CAPABLE OF STIMULATING AN RggwwsE WHEN l " S P L M l E D ~E N E I C W Y , Robert J. Ketchum and Orion D. Hegre, Dept. of Cell Biology and Neumanatoay, University of Uinnesota, Minneapolis HN 55455. Neonatal rat islets derived by culture-isolation have teen shown to k free of MlC class 11+ cells, and are immunologically silent when transplanted to either syngeneic or allogeneic hosts. Allogeneic transplantation of cultured neonatal non-islet pancreatic tissue, which is known to contain class 11+ cells, results in rapid allograft rejection. Unexpectedly, m i c transplantation of cultured non-islet ductal tissue also resulted in lononuclear lm,me cell (HNC) infiltration of the graft in 86% of grafts examined. Highly purified syngeneic islets and ductal elements grafted syngeneically at remote sites display an i u n e response in the ductal element graft, while the islet graft is free of any imnme cell infiltrate. This syngeneic imune response does not result from the use of xenogeneic serum in the medium, since cultures carried out using syngeneic rat serum supplemented medium yielded identical results. Uncultured neonatal pancreatic tissue grafted syngeneically does not result in IQK:

infiltrate, Thi6 i.rmne response to a syngeneic stimulus correlates with the presence of class 11+ (antigen presenting) cells. In grafts free of class 11+ cells (culture-isolated islet grafts) no i.rmne responrrc to syngeneic stimulus was observed, while a response was present when syngeneic ductal elements, known to include class 11+ cells, were grafted. This indicates a need for cells capable of antigen presentation to stimulate this syngeneic rerrponne, and suggests that either a modified self antigen or a nomally sequestered antigen is being presented. This syngeneic imune response demonstrates many of the same characteristics of, and may be analagous to, the In vitro syngeneic, or autologous, mixed lymphocyte reactions.

indicating this response is not to developnental antigen. c 132 THE PRESENCE OF "SELF" MHC CLASS I1 (MA-DR) ANTIGENS DETERMINES WHETHER BLOOD TRANSFUSIONS IHMUNISE OR SUPPRESS. EL Lagaaij, A Termijtelen. E Goulmy, & JJ Van Rood, Leiden University Hospital, The Netherlands. Blood transfusions can immunise the recipient, as well as induce prolonged allograft survival. It is not known what makes that some transfusions inrrmnise the recipient whereas others induce immune suppression. We investigated if certain MHC compatibilities or differences between recipient and transfusion donor and organ donor are required to induce the beneficial "transfusion effect" in man. We studied graft survival and blood transfusion induced changes in cellular and humoral immunity in 4 different patient groups. The patients received a single blood transfusion of a randomly choosen donor. We found in all 4 groups that to induce a beneficial "transfusion effect" compatibility for at least 1 HLA-DR antigen between recipient and transfusion donor is required. If the transfusion and recipient are mismatched for both MA-DR antigens, the recipient is immunised, resulting in an increased antibody production (P=O.OOl), an increased cytoxicity (CML) (P=O.OOS), an increased mixed lymfocyte reaction (MLR) (P=O.OOS) and a decreased graft survival (P-0.003). After a beneficial (MA-DR sharing) transfusion. the in vitro test remain unchanged or decrease. Graft survival increases with the number of shared antigens between transfusion donor and organ donor (P=O.O2), suggesting that a donor specific suppression is induced. Recent experiments have revealed a direct interaction between the CD4 molecule and HLA-DR antigen.

To address the nature of this interaction we have used a xenogeneic system in which a human CD4 cDNA was expressed in the murine CD4-and CD8-negative hybridoma 3DT52.5.8. The TcR of 3D3752.5.8 recognizes the murine class I molecule od. A class I 1 expressing Dd-positive cell line was obtained by cotransfection of the human class I1 cDNAs together with the murine od gene int.0 the murine fibroblast line DAP3. Coculture of 3DT52.5.8 and DAP3 expressing DP-Dd resulted in a 20 fold increase in IL-2 production and in rosette formation only when both CD4 and DP were present on the responding T hybridoma and the presenting cell, respectively. We are using this system to map regions of the CD4 molecule that interact with the class I1 MHC Ag.

The CD4 molecule has also been shown to be the receptor for the human immunodeficiency virus (HIV) via the gp120 molecule. Since gp120 and class I1 both interact Kith CD4, we have used our functional assay to verify if gp120 exerts an inhibitory function on CD4 class I 1 interaction. Recombinant gp120 inhibits the functional interaction and rosette formation in a concentration dependent fashion with maximal inhibition at about 10 pg/ml.. This inhibition is specific since it can be reversed by recombinant soluble CD4. The fact that recombinant. gp120 can inhibit the functional interaction between CD4 and its physiological ligand (class I1 Ags) suggests that the use of gp120 on a vaccine against HIV infection could alter the immune response of such individuals. This work was supported by SRC, MRC and NCI. T lymphocytes discern self from non-self molecules through the interaction of their antigen-specific receptors and proteins encoded by the MHC. Although the nature of this association is not well-defined, a model has been proposed whereby the v-segments of the T cell receptor interact with residues along the 2 alpha helices of the class I antigen (Davis et al.; 334:395, 1988) . We have recently shown that CTL generated against the class I molecule QIOd crossreact on several unrelated murine class I antigens containing the shared QIOd residues at amino acid positions 152, 155, and 156 (Mann et al.; J X 168:307, 1988) . These residues contributed by the a-2 domain occur in the alpha helical portion of the class I molecule and amino acids 152 and 155 could interact directly with the T cell receptor. To further characterize the role of these amino acids, we are in the process of determining whether insertion of these 3 residues by site-directed mutagenesis into a human class I molecule will allow for the antigen's recognition by anti-910 CTL. Here the T cell repertoire becomes restricted, so that foreign antigen can be recognized only when associated with the MHC products of the host, and mature T cells are tolerized to self antigens, a process which also seems to be MHC-restricted. Thus, T cells should be non-reactive to self antigens when they are associated with MHC products present on the tolerance-inducing thymic cells, whereas they may still react to the same self antigens when associated with different MHC products.

To examine MHCrestricted tolerance in vivo, a model system must have: a) self antigen in the context of one MHC haplotype. and b) tolerance to both that and a second MHC haplotype. Chimeras were prepared by aggregation of preimplantation embryos of two strains of mice, C57BL/6 (B6) and BALB/c.

The thymus of such chimeras should be composed of two distinct and completely intermixed populations of cells, one from each parental strain (isozyme analysis indicates no detectable fusion of cells).

Thus, T cells maturing in the chimeric thymus should be exposed to and tolerized to minor histocompatibility antigens (mHAs) of one parental strain only in association with the MHC of that strain. For example, mice might be expected to express B6 mHAs only with H-2b (the 8 6 MHC).

However, our chimeras were fully tolerant to F1 skin grafts, which have "hybrid" combinations of mHAs and MHC (e.g., B6 mHAs with H-2d). These results are most consistent with either, a) "wholesale" antigen processing and presentation of all mHAs by the tolerizing thymic cells, and/or, b) functional sharing of MHC products between the parental thymic cell populations.

Many of the events critical to the maturation of T lymphocytes occur in the thymus. In the case of T-cell responses against viruses such response defects are associated with a marked increase in disease susceptibility as illustrated by class I MHC controlled susceptibility to lethal pneumonia induction by Sendai virus. Certain class I or class I1 MHC determined TC response defects (four out of six tested by us) can be restored by imunization in vivo and/or restimulation in vitro with DC. DC are the most effective APC. Their superior APC capacity is due to 1) a very high absolute number of class I and class I1 MHC molecules, and 2 ) a low degree of sialylation of MHC and other surface molecules, reducing negative charge and facilitating access of the T-cell receptor to the MHC groove presenting the antigenic peptide and/or improved clustering with T cells. The more effective antigen presentation by DC allows a more prominent role for a CD4+ Q cell independent pathway of CD8+ TC activation. It is postulated that the more effective direct triggering of CD8+ Tc precursors lowers the threshold for IL-2 production by CD8+ cells, reducing the requirement for IL-2 production by the CD4+ cells. Failure of DC to overcome certain MHC-linked specific TC response defects probably reflects complete failure of any foreign peptide derived from the processed antigen to interact efficiently with the MHC or a true Tc repertoire defect.

Donald Pious, Department of Pediatrics, University of Washington, Seattle, W 98195 MHC class I1 molecules bind inmunogenic peptides derivsd fro13 soluble antigen and the complex is recognized by specific T cells. Ue have isolated eight independent mutant B -L U clones which are altered in their ability to present antigen. In standard proliferation assays using four different soluble protein antigens, the mutants are unable to stimulate the majority of T cell clones restricted to HLFl DR or DP. Cllthough unable to present *hole Hepatitis B surface antigen (HBdg), they effcctively present a HBdg peptide to a DPrestricted T cell clone.

The fact that both DR and DP restricted antigen presentation is abnormal in these mutants made it likely that the class I 1 structurual genes are unaltered. This hypothesis is supported by the finding that DNO sequences from the DR genes of one mutant are norml. However, two observations indicate that the uture class I1 di-expmsccd by the mutants are structurally altered.

Binding to the mutants with tuo polymorphic anti-DR antibodies and one anti-DP antibody is reduced, although the level of cell surface class I1 expression is normal.

Second, the Class I1 diners from the mutants dissociate into NO-rs under in vitro conditions (SDS-PWE) which preserve dimers in the progenitor line. Together, these functional and structural data suggest that the mutants are defective in a molecule that either associates with or post-translationally modifies class I1 molecules and is required for the physiologic formation of an tWC/antigen complex. they do not function as restriction elements presenting foreign antigens to T-cells. To investigate the nature of this functional defect we have constructed 3 different recombinant class I genes using DNA segments from the B10 Q9 (Qa-2) or H-2Db genes. structural protein encoded by the recombinant genes is derived entirely from the Q9 gene whereas the cis-acting transcriptional regulatory elements or the DNA segment encoding the membrane anchoring domain is derived from the H-2Db gene. into fertilised CBA/Ca embryos by microinjection and transgenic lines were established. To date we have established 13 transgenic lines. We have shown that the Q9 (Qa-2) antigen encoded by the recombinant genes behaves as a major transplantation antigen in skin grafts and provokes strong secondary cytotoxic T-cell responses in grafted animals irrespective of the tissue distribution or mode of membrane anchorage of the Q9 antigen. At present, we are investigating whether the Q9 antigens encoded by the recombinant genes are able to present influenza virus or mouse minor histocompatibility antigens to T-cells during immune responses and hence whether they can function as restriction elements. is a distinctive system because it permits an analysis of the activation requirements for antigen specific, resting T cells.

been isolated following culture with anti-Ig-sepharose and compared to dendritic cells as stimulators of CD4' T cells in the MIX. I1 MHC products and independently stimulated the 1 ' MLR and the production of several T derived lymphokines, including IL-2 and 11-4. However, the relative potencies of dendritic cells and anti-Ig blasts as 1'MIR stimulators varied in a strain dependent fashion. times more active in stimulating Hls-mismatched. MHC-matched T cells, relative to syngeneic T cells. anti-Ig blasts when stimulating acroas an MHC barrier and were likewise more effective in binding IQIC-disparate T cells to form the clusters in which the MIX was generated. Dendritic cell-T cell clustering was resistant to anti-LFA-1 mAb, while B blast-T cell clustering was totally blocked. Thus, anti-Ig B lymphoblasts and dendritic cells, two cell types which differ markedly in phenotype, also differ in efficiency and mechanism for initiating responses in allogeneic T cells.

Only anti-Ig blasts could stimulate across an Mls barrier, being at least 100

In contrast, dendritic cells were 10-30 times more potent than the as the lymphocyte f u n c t i o n antigen 3 (LFA-3) and the i n t r a c e l l u l a r adhesion molecule (ICAM). We i n t e r p r e t t h i s as an increase i n the membrane expression o f these s t r u c t u r e s f o l l o w i n g incubation. The increase i s blocked by the t r a n s l a t i o n i n h i b i t o r , cycloheximide, implying t h a t p r o t e i n synthesis i s involved. Helper T cell responses to soluble globular proteins require processing of the protein by Ia-expressing antigen presenting cells (APC). Antigen is internalized into acidic vesicles, proteolyzed, and peptides containing T ceU antigenic determinants are transported to the APC surface where they are recognized by the antigen-specific T cell in conjunction with Ia. Most Ia-"pressing cells are competent APC, however, only B cells have antigen-specilic receptors on their surface aUowing bound antigen to be processed and presented at 1/lW the antigen concentration required by nonspecific APC Little is known about B cell antigen processing function during differentiation, or if Ig-mediated APC function is altered at different maturational stages, thus allowing regulation of B cell-helper T cell interactions. Neonatal acquisition of APC function was examined in mice ages day 1 to day 15. Splenic cells from d l to d10 mice process and present pigeon qochrome $, PG at 0-20% of adult levels. By d15 neonatal spleen cells acquire the ability to process and present soluble PG at 3 5 4 of adult levels. The ability to internalize antigen through Ig rwptors was determined using an antigen-antibody conjugate, P$-&(ab')z. Neonatal spleen cells acquire the ability to process antigen through Ig simultaneously with the ability to process soluble antigen. Lack of prowsing by neonatal spleen cells prior to d10 is not attributable to insufficient levels of surface Ia. since d3 neonate spleen cells are able to activate T cell hybrids to 50% of adult levels when provided with P$ 81-104, containing the T cell determinant. DlOd15 neonate presentation of P@1-104 is indistinguishable from adult levels. B cell maturation into memory B cells was identified by the loss of the Jlld differentiation marker. Splenic Jlld'O B cells increase from 540% following immunization and return to nonimmune levels after 4 weeks. During antigen-induced B cell maturation, Jllb B cells are indistinguishable from splenic B cells in the ability to present antigen introduced into the processing pathway either pinocytotically or via surface Ig. P p antibody conjugates specific for mouse F(ab')Z I N , IgD, or IgG are presented equally well by both splenic and Jll@ B cells.

Thus, acquisition of B cell processing function appears to be developmentally regulated and may play an important role in B cell tolerance meshanisms. Once B cells have acquired the ability to process antigen this function is maintained and is not regulated during maturation into memory B cells. We are currently investigating the role of Ig isotype during neonatal acquisition of antigen processing. (Supponed by NIH grants AI-18939, AI-12001, and AI-2317) 

As the preliminary studies suggested that carrier SC (APC) for TS vs. Tcs activation might be distinct, studies were done to directly address this possibility by assessing tha ability of S3 coupled to various cell populations to activate TS and Tcs. The results indicated that TS activation required that S3 be coupled to plastic adherent cells which bear both I-A and I-J determinants. These cells are nonadberent to anti-Ig and nonfunctional in cyclophosphamide (Cy) treated mice.

I n contrast activation of Tcs required coupling of S3 to plastic non-adherent and anti-Ig adherent cells. These cells are functional in Cy treated mice and bear the B cell markers Jlld and I-A but not 1-3. Thus S3-specific TS are activated by I-A+ I-J+ adherent cells (presumably macrophages) whereas Tcs are activated when antigen is presented by B cells. (NIH Grant CA25054.) Interleukin-2 activated killer (IAK) lymphocytes (also known as LAK cells) which destroy a broader spectrum of tumors invitro than NK cells have been used sucessfully in an adoptive immuno-therapy protocol for the treatment of patients with a variety of advanced cancers. The cell suface molecule(s) on tumor cells that a r e involved in specific binding to IAK cells and in programming IAK cells for cytolysis (IAK acceptor molecules) have not been characterized. Inorder to identify such acceptor molecules a crude membrane digest of the lung carcinoma cell line A549 was biotinylated and adsorbed to IAK cells or to unstimulated human peripheral blood lymphocytes (UPBL) (each from the same person). Proteins from the washed solubilized cells were separated by PAGE, Western Blotted and probed with streptavidin-alkaline phosphatase. Several experiments demonstrated that different tumor membrane proteins bound to IAK cells compared to UPBL. The unstimulated cells bound one tumor membrance protein (about 40KD) not found on the IAK-adsorded blot. The IAK cells bound three tumor proteins (approximately 30,46 & 50KD) not found on the UPBL-adsorded blot. Three other proteins (about 35,44 & 55KD) were found to adhere equally well t o IAK cells and UPBL. Utilizing a streptavidin affinity column, solubilized tumor membrane proteins that bound to IAK cells could be separated from solubilized IAK membrane proteins. The isolated tumor membrane proteins that adsorded to IAK cells inhibited IAK mediated lysis of A549 tumor cells by >85%. These studies suggest that specific cellular adsorption techniques may be useful in isolating and characterizing tumor membrane proteins involved in interactions unique to cytolytic lymphocyte-tumor cell target binding and lysis. Activation of human T lymphocytes occurs via the T cell receptor-CD3 complex but can also be induced through the non-antigen-specific CD2 molecule. Selected combinations of mAbs or the soluble CD2 ligand, namely LFA-3 and a unique anti-CDZ mAb (CD2.1) induce human T cell activatlon. CD4 is an accessory molecule implicated in t h e activation of human T lymphocytes. This molecule may exert this function by increasing intercellular avidity through binding to MHC class I1 molecules and/or by transmitting intracellular signals. We have investigated the action of mAbs directed against different epitopes on the CD4 molecules in the activation of human T cells via the CDZ pathway. We show that anti-CD4 mAbs inhibit CD2 induced T cell proliferation in an epitope-depe dent fashion. This inhibition does not appear to be linked to the lower CD2 mediated [ C a ' + ] response induced by anti-CD4 mAbs, since [Ca"] response is equally affected by anti-CD4 mAbs whether or not they inhibi ed T cell proliferation. In conclusion, the partial inhibition of the CD2 induced [Cb'] response of T cells by various anti-CD4 mAbs suggest that : 1) this inhibition does not totally account for the inhibitory effect of anti-CD4 mAbs, 2) and the proliferation induced by anti-CDZ mAbs may not be completely ascribed to the [Ca2+] response of T cells. Rosenberg. and Alfred Singer, Experimental Immunology Branch, NCI, NIH, Bethesda, MD 20892 . We have devised a model to study the in vivo generation of suppressor cells by using mice congenic at Qa-1, a class I-like molecule encoded to the right of H-2D. disparate tail skin grafts (TSG) unless a second graft with additional helper determinants was also present. without any source of additional help, failed to reject their Qa-1 graft, and were unable to reject them even upon the subsequent addition of exogenous help. Thus, exposure to Qa-1 disparate grafts, in the absence of additional help, either led to Qa-1 specific tolerance or suppression. mice failing to reject Qa-1 allografts revealed the presence of Qa-1 specific suppressor cells that inhibited the in vivo activation of antigen specific effector cells capable of rejecting Qa-1 bearing allografts. experiments using T cell subpopulations should allow for further characterization of these Qa-1 specific suppressor cells.

We found that B6 mice did not reject Qa-1 However, animals engrafted with a Qa-1 graft alone, In the thymus, the T-cell receptor genes are rearranged, the T cells learn to recognize their own major histoconpatibilitp complex "HC). and they learn to respond to foreing HEC. These events seem to be linked to the interaction between T-cell precursors and the stromal cells of the thyms. Thus increasing evidence points to an essential role for the thymus epithelial cells (TE cells) in development of at least HEC class I1 recognition by the T cells. To be able to study the importance of TE cells in T cell maturation. we have developed a method for growing murine T cells in serum-free Pediun with well defined constituents. The m d i u a allows far growth of TE cells without concomitant growth of bone marrow derived cells as macrophages and fibroblasts. Data obtained by EN and lmmunocytochenistry showing the epithelial nature of the cultured cells, as well as autoradiographic data on the growth pattern, and characterization of TB cell supernatants will be provided in addition to results obtained from co-culture of TE cells and T-cell precursors (CM-CDS-thymytes). lmmunogenicity c 152 Branch, National Cancer Institute, Bethesda, MD 20892 The effector limb mediating skin allograft rejection is highly antigen specific, rejecting cells that express allogeneic MHC antigens while sparing those which fail to express allogeneic MHC determinants. disparate skin grafts are completely rejected in spite of the fact that only a small percentage of the cells within the graft express Ia antigens. Thus, it is possible that MHC class I1 disparate grafts are rejected by a mechanism that does not assess the expression of MHC determinants on each cell. We assessed the specificity of the rejection of Ia disparate grafts by using allophenic skin grafts in an adoptive transfer system and concluded that skin grift rejection across an MHC class I1 disparity required recognition of allo-Ia determinants expressed by every cell in the graft. Therefore, we reasoned that MHC class I1 antigens must be induced on these Ia negative populations. Indeed, injection of mice with gamma interferon dramatically induced Ia antigens on previously negative keratinocytes. We next tested whether the induction of allogeneic Ia determinants on keratinocytes was necessary for graft rejection by engrafting parental strain mice with skin from F >parent bone marrow chimeras. Such grafts failed to be rejected, in spite of the speciic rejection of the allogeneic Langerhans cells, indicating that the failure of keratinocytes to express allogeneic class I1 determinants leads to graft preservation. conclusion, MHC class I1 disparate skin allografts are rejected in a highly antigen specific fashion, secondary to the induction of MHC class I1 antigens on skin cells that fail to constitutively express them. To address whether the extensive polymorphism characteristic of class I molecules influences CD8 binding, we have screened a panel of transfectants expressing individual class I MHC alleles. Of 18 alleles tested, only Aw68.1 did not bind. All other molecules dld bind, including A2.1 and Aw69, which differ by 13 and 7 amino acids respectively from Aw68.1. Position 245 in the alpha 3 domain was identified by sitedirected mutagenesls as the critical residue differing between A2.1 and Aw68.1 which determines binding. A mutant Aw68.1 molecule containin alanine at position 245 bound CD8, while a mutant of A2.1 with valine at 245 did not. Alanine is found at position 245 of all human and murine class I molecules sequenced to date except Aw68.1 and Aw68.2, which have valine at that position. Bulk cultures of A2-allospecific CTL were also sensitive to this substitution, and preferenhally recognized both molecules with alanine at 245. This study shows that Aw68.1 differs from other class I molecules in its capacity to bind CD8, and raises the possibility that Aw68.1 may not function as a restriction element as effectively as other class I alleles. T cell hybridomas derived from H-Zd LNC recognked S and preS2 antigens in a I-Ad restricted way, while T cell hybridmas from H-2k LNC manifested a specificity for either preS2 in association w i t h I-Ak or for S in association with I-!& The activation of lhese hybridomas by antigen and antigen presenting cells (AF'C), as measured by IL-2 secretion, was found to be sensitive to prostaglandines and could be completely inhibited by anti-LFA-1 moncclonal antihxlies. Different AFT populations were tested for their capacity to present SpreSZ particles to these T cell hybridomas. Various macmphage like populations such as resident, Con A induced, Thiiglycolate induced peritoneal exudate cells as well as splenic adherent cells were found to present efficiently the SpreS2 antigen. In conhast B cells and la+ B cell lines (TA3, M12.4) could not function as accessory cells in the SpreSZ specific stimulation of these T cell hybridomas. The inability of these cells to present this antigen was not due t o inhibitory effects since these cells did not inhibit the presentation capacity of other potent APC's. Funhcnnore addition of APC's of a different haplotype could not complement for the defective presentatiW of SpreSZ by B cells and B ceU lines indicating that MHC independent accessory factors are not implicated in this process. Hence it is clear that maaophage-like APC's and B cells d i f f a in their capacity to process and present SpdZ antigens. Since SpreSZ is a very stable particle composed of lipids and proteins it is conceivable that such antigen quires a smng degradation and swh plocessing might occur in Cenain -phage-like APC's but not in B cells. Recombinant human insulin biosynthetically labeled with k n d " S at several amino acids was used as an antigen and was exposed for varyillg lengths of time to TA3 mouse B cell APC. Subcellular fractionation and HPLC chromatography permitted several of the processed peptides distributed throughout the insulin molecule to be monitored. Many insulin peptides localized to both the extracellular (8 peptides) and intracellular (4 peptides) compartments of TA3 cells were detected. membrane-associated peptides is in progress. Many of the peptides processed by TA3 APC in situ co-elute with those obtained upon digestion i n vitro by the insulin-specific insulin degrading enzylg5 (IDE). f o r the processing of 51-labelled human insulin suggest that insulin may be processed in B cell APC into immunogenic peptides by an enzyme(s) present on the plasma-membrane, intracellularly and extracellularly. (Supported by MRC and CDA) .

We investigated the pathway of antigen processing 5 itu in cell APC. Hurine CTL clones specific for HIA-A2 were generatef with the human cell line JY. Four of five CTL clones were found to lyse A2K transfected murine cells more effectively than A2 transfectants. Anti-CD8 specific mAb inhibited t 3 lysis by these four clones, and this inhibition was more pronouncgd for A2K One clone, which lysed A2 and A2K transfectants equivalently, was shown to be insensitive to anti-CD8 antibody inhibition.

These findings indicate that A2-specific T r i n e CTL clones possess greater avidity for murine target cells expressing the A2K hybrid molecule relative to those expressing the A2 molecule.

This implies that a CD8 interaction with the same molecule seen by the T cell receptor is important for target cell recognition. HLA class I1 antigens are highly polymorphic cell surface proteins involved in initiation and regulation of the immune response.

Allelic sequence variation primarily affects the structure of the first external domains of the a and j3 component chains. Here we provide evidence for other types of allelic polymorphism for these genes.

The sequences of two cDNA clones corresponding to the HLA-DQp mRNAs from an HLA homozygous cell line exhibit both alternative splicing and readthrough of polyadenylation. Furthermore, the alternative splicing event is associated with only a subset of HLA-DQB alleles, while the polyadenylation site readthrough is found in a larger subset. This suggests that polymorphic & acting elements within the HLA-DQp gene control both processing steps. Proteins, presumably encoded by the alternatively spliced mRNAs lacking transmembrane exons, are immunoprecipitated with a monomorphic monoclonal antibody directed against HLA-DQ. These proteins are found in supernatants of cultured cell lines for which secretion is predicted, but not in those of cell lines which do not contain the alternatively spliced mRNAs.

Such secretion class of I1 allelic products could profoundly affect interactions between effector and target cells in an immune response. Departments of Biology and Chemistry and the Cancer Center, Q-063, University of California at San Diego, La Jolla, CA 92093. We are studying the murine ICAM-1 gene and the effects of ICAM-1 on antigen recognition. Using the human ICAM-I cDNA, we have isolated cDNA and genomic clones encoding the murine homologue. The murine ICAM-1 gene is a single-copy gene that consists of multiple exom spanning 25kb of DNA and encodes a 2.5kb mRNA that is expressed at high levels in a wide variety of different cell types. Sequence analysis indicates that murine ICAM-1 is 50% and 6096 homologous to human ICAM-1 on the protein and DNA levels, respectively, and is a member of the immunoglobulin gene superfamily, consisting of several V-like domains linked tanddy. We are also studying the effects of ICAM-1 on antigen recognition. The T-cell clone 14D m s f 1988). This response is blocked with the anti LFA-1 antibody FD441.8 when antigen is presented by BlO.A(2R) spleen cells but not when antigen is presented by DCEK, a fibroblast transfected with I-Ek. We are currently aansfecting the murine ICAM-1 cDNA into DCEK in order to determine if we can enhance the 14D response and to determine if the enhancement is LFA-1 dependent.

wth an alp a beta TCR that recognizes moth The T cell differentiation anti en, CD4, is expressed by MHC class I1 restricted T lymphocytes.

MHC class I1 products. The association between CD4 expression and restriction by MHC class I1 RfiOducts has led to the hypothesis that CD4 may interact with monomorphic determinants of A large body of experimental evidence suggests that CD4 interaction with MHC class I1 molecules leads to an increase in the binding avidity of T cell-stimulator cell interactions. A direct test for a functional CD4-MHC class I1 interaction in T cell activation requires a separate evaluation of CD4-Ia interactions from TcR-A /Ia recognition. However, a separate evaluation proves difficult since the T cell receptor and Cb4 may interact with the same MHC class I1 molecule.

In this report, we use a T cell activation protocol, where TcR-Ag/Ia recognition is replaced by TcR complex-antLCD3 antibodies interactions. Using this activation protocol, we pave analyzed the effects of monoclonal anti-MHC class I1 antibodies on the activation o$a CD4 hybridoma in the absence of its TcR restr$tinJ MHC class I1 molecule (IE ) but in the presence of unrelated MHC class I1 molecules (IE ,IA ). The data obtained clearly indicate a functional role for CD4-MHC class I1 interactions in T cell triggering. We have targeted hen egg lysozyme (HEL) to murine B cells using heterocrosslinked antibodies which specifically bind to surface IgD or different MHC molecules. occurred more quickly with targeting to IgD than to MHC structures as assessed by fixation and pronase stripping experiments. HEL was internalized quickly into acidic compartments when targeted to IgD but was detected much later when targeted to MHC molecules, as assessed by shifts of fluorescent signal of internalized FITC-HEL. However, the data indicate that not all endocytosed HEL entered low pH (<5.5) compartments. Degraded HEL was released from B cells following endocytosis of 125-I-HEL. This release was detected earlier with targeting to IgD than to MHC structures. Interestingly, the total amount of internal 125-I-HEL decreased with time after endocytosis via IgD, but the internal 125-I-HEL was almost entirely whole undegraded HEL at all times following endocytosis. These data and those of chloroquine and leupeptin inhibition studies indicate differences in the fate of antigen entering B cells via IgD or MHC structures, and support the notion of a neutral pH storage compartment for antigen endocytosed via surface IgD on normal splenic B cells.

Internalization and presentation of HEL to hybridoma T cells

Laboratories, Department of Surgery, University o f Iowa, Iowa City, IA 52242 Target cell lysis by CD8' CTL is a highly specific phenomenon in vitro, as we have confirmed repeatedly in reverse labelling tests by showing that admixed "third party" target cells are not lysed in the presence of specific CTL-mediated cytolysis. However, when mixtures of CTL and their specific targets are inoculated into the skin of hosts syngeneic to the CTL, host cells at the site of inoculation are destroyed, often to an extent that results in grossly observable, full -thickness necrotic lesions. We have evoked these "innocent bystander" reactions in mice with CTL directed against single and multiple non-H-2 antigens and TNPhapten and influenza A virus-specific antigens. Thus, the ability to trigger bystander tissue destruction appears to be a general characteristic of CTL-target cell interaction in vivo. Our current evidence suggests that host inflamatory cells recruited and activated by factors stimulated by CTL-target cell recognition actually mediate the tissue destruction. These CTL-initiated bystander reactions may be the basis of the non-specific tissue destruction that contributes to allograft rejection and that is observed in many serious virus infections and in intense DTH reactions and contact dermatitis.

Rong h a Lin. Baael I n s t i t u t e f o r I u m l o g y , Baael, Snitzerland. W e have investigated the b a d e for i n u n i t y or tolerance t o a m a e aezw proteinthe f i f t h caponent of c o l p l a c n t (a). I n C5 deficient nice t h i s protein is absent from s e r m and as a cnrwquence they are not tolerized t o CS. C5 deficient dca generate UY bearing T c e l l s which recognize C5 i n the c m t e x t of clasa 11. I n contrast, C5 s u f f i c i e n t mice i n which C5 protein is continuously probced d m t mtmt T c e l l reaponsea againmt U. We have tested i f t h i s self protein i s proceseed and presented with clase I1 i n n o m l mice and can be recognized by C5 specific T c e l l s i n the absence of exogenewsly added antigen. A l l clasa I1 bearing c e l l s fm C5 s u f f i c i e n t l i c e activated C5 specific T c e l l clonea without additional antigen. Presentation was mt a cansaquence o f C5 secretion by macrophages i n culture but was a h t o be derived Prom endqlenewaly generated CS/cl.ss I1 caplexea. Thus t h i s self protein is e f f i c i e n t l y preaented ,inin and available f o r tolerance in&xtion. Although C5 deficient dca cannot secrete C5 they s t i l l synthesize a precursor mlecule, pro-C5, i n Accumulating evidence from a number of models suggests that unique subsets of antigenpresenting cells a r e responsible for the induction of specific T cell-mediated responses. W e have previously described an age-dependent maturational defect in the ability of the SJL strain of mice to activate DTH-inducer T cells to a wide variety of antigenic stimuli. None of the other 14 strains tested exhibited a similar defect and all other accessory cell dependent responses were unaffected in the DTH unresponsive SJL. W e have also shown that the adoptive transfer a macrophage from older DTH responsive SJL or other DTH responsive LAS strains can overcome this defect in DTH responsiveness. W e have recently found that a subpopulation with the Mac-I+, Mac-3' and Mac-2-surface phenotype a r e able to transfer responsiveness. FACS analysis indicate that the Mac-3 phenotype is expressed on less than 20% of macrophages. Titrations of the Mac-3' cells isolated by FACS indicate that adoptive transfer of o ly 100 Mac-3' cells can overcome the defect in DTH responsiveness. By contrast, transfer of 10 Mac 3-or Mac 2' cells were unable to overcome the defect. Our data suggest that the induction of CD4' antigen specific cells DTH-inducer T cells is mediated by a phenotypically unique small subset of macrophage accessory cells. In our studies, we have examined the effect of administering Fab' fragments of anti-L3T4 MoAb (Fabl-GK1.5) on the inhibition of humoral immunity. Treatment of KLH-primed mice with 0.5 mg Fabl-GK1.5 depleted L3T4' cells from lymph node tissue while leaving other lymphocyte subpopulations intact. After injection of KLH in complete Freund's adjuvant, these T,depleted mice were unable to produce anti-KLH antibodies. Long-lasting unresponsiveness against KLH (12 weeks) was observed despite the apparent regeneration of the T, population of the lymph node. The results obtained using either Fab' or intact GK1.5 antibody were comparable and suggest that a transient depletion of T, does not account entirely for the long-term humoral unresponsiveness. The aim of this study was to gain a more detailed insight into the molecular aspects of antigen processing during the imune response. As a first approach, endosomal vesicles were isolated from bovine alveolar macrophages and their proteolytic activity with respect to a model protein antigen, sperm whale myoglobin (Mb), was characterized. During the first stage of digestion of Mb by the endosomes, a limited number of fragments were preferentially released from the antigen. We have isolated and identified these fragments.

The digestion of myoglobin is completely prevented by pepstatin, a specific inhibitor of aspartic proteinases, and only marginally by other proteinase inhibitors. When Mb fragments preferentially released upon digestion with purified bovine cathepsin D, an aspartic proteinase abundant in macrophages, were identified, almost all coincided with the fragments released by the endosomes. To define in more detail the selectivity of cathepsin D under the mild conditions applied, other protein antigens were similarly treated with the enzyme and the peptides released were identified.

The location of the preferential cleavage siteswhen related to known T-cell epitopessuggests a dominant role for cathepsin D in the processing of protein antigens to yield fragments for presentation to T-cells. Possibly, the observed selectivity of the enzyme may account for the structural similarities among T-cell epitopes, noted by others. Actively acquired tolerance in mice to the antigens of the MHC (H-2) is induced by exposure of the animals to allogeneic lymphocytes within 24 hours of birth. Actively acquired tolerance to the MHC in humans (HLA) cannot be studied in the same way. However, we have evidence for the existence of actively acquired tolerance in humans in a study of 26 highly sensitized patients waiting for a renal allograft. They had developed complement dependent antibodies to the HLA antigens of almost all unrelated Caucasoid donors. The sera of these highly sensitized patients were tested against a panel of lymphocytes that were mismatched for only one HLA class I antigen. We found for these 73 patients HLA class I antigens that, although different from those present in the recipient, did not lead to a positive crossmatch. We called such antigens "permissible mismatches" and show that they often included those HLA antigens of the patient's mother that the patient had not inherited (noninherited maternal antigens; NIMA). In 15 of the 26 patients, the permissible class I mismatches included the NIMAa. The noninherited paternal antigens (NIPAS) were analyzed as a control; only two of the 25 NIPAs tested were acceptable mismatches, which emphasized the preferential nonresponsiveness to NIMA. Recent experiments indicate that what holds true for antibody formation also holds true for T cell activation. of HLA class I and HIA class I1 allospecific CD8-positive CTL clones. Monoclonal antibodies (McAb) directed against the CD8 structure were only found to inhibit antigen-specific cytotoxicity of a series of class I allospecific CD8-positive CTL clones and not of a class I1 allospecific CD8-positive CTL clone. However cytotoxicity induced by CD3 McAb (used at suboptimal concentrations) or CD2 McAbs in both types of CTL clone was blocked by CD8 McAbs. The absence of CD8 McAb blocking of antigen-specific cytotoxicity of the class-11-specific CD8positive CTL clone may be explained by assuming that it results from a triggering signal which is to strong to be overcome by the down-regulatory signal of the CD8 antigen. These combined findings clearly suggest a functional involvement of CD8 not only in TCR/CD3 activation, but also in TCR/CD3 controlled alternative activation routes, such as the CD2 activation pathway. Moreover it shows that even an HLA class I1 allospecific CD8-positive CTL clone expresses a functional active CD8 antigen. The absence of HIA class I expression on the target cells (DAUDI cells) used in the experiments described indicate that the CD8 antigens not act solely in an adhesion-like fashion, but exhibit also a more general regulatory function in T-cell activation. This regulatory role of CD8 may be explained by assuming the induction of a threshold for activation, which is triggered after binding of CD8 McAb or binding to its natural ligand, HLA class I. In our view, CDE-mediated regulation of T-cell activation could therefore prevent non-specific triggering of cytotoxicity by interactions of insufficient affinity. *This study was supported by a grant from the Dutch Kidney Foundation. The CD4 T-cell surface antigen 1s felt to have the dual functlon of stabilizlng the interaction of the T-lymphocyte with the antigen presenting cell (APC) a s well as transduclng an Independent signal that can potentiate the actlvatlon related alteratlons generated through the T-cell receptor. We have found that upon antibody-mediated cross-linklng of the CD4 molecules of cloned murlne T-lymphocytes there is a time and temperature dependent decrease In the abundance of the lymphocyte-speciflc tyroslne klnase p66lok. This co-modulation is speclflc for CD4 and p66lCk slnce cross-linking of other T-cell surface antlgens (CD3. T200, Thyl.2) does not result In detectable alteratlons in the abundance of the lck protein and slnce CD4 cross-llnklng does not induce any alteratlon In the abundance of p60*.. another srcrelated tyrosine klnase highly expressed In T-cells. Such data suggest that CD4 and the Internal membrane lck protein are in close proximity within the cell. Further analysls has revealed that slgnlflcant amounts of lck can be immunopreclpitated by antl-CD4 antibodies. In addltlon. CD4 can be speciflcally preclpltated by anti-lck antibodies. Our data imply that CD4 and p661Ok are physically associated in CD4+ T-lymphocytes. The flndlngs that CD4 is Msoclated to the lck proteln in either murlne or human T-cells and that CD8 is also complexed to p66lck ln CD8* T-cells suggest that the lck tyroslne klnase is involved In the functlon of the CD4 and CD8 accessory molecules. these APC do not appear to present processed Klsa determinants. In light of these findings, of apparent interest is the issue of which cells types are responsible for Hlsa-specific T cell tolerance induction. Studies in mice treated from birth with anti-p antibodies suggest an important but perhaps not exclusive role for B cells in this process. We are currently pursuing the identity of other cell types which may be involved.

In addition, lmmunogenlclty c173 University of Texas Southwestern Medical Center at Dallas, Dallas, Texas75235 QlO is a soluble Class I-like major histocompatibility antigen produced specifically by the liver. Previously, it has been shown that mice possessing soluble QlO can generate anti-Q10 cytotoxic T lymphocytes (CTL), suggesting that this soluble molecule does not function as a tolerogen. We have recently constructed C3H transgenic animals which express an exon shuffled Q10 (al, P2)/Ld (03, TM) molecule. This QIO/Ld molecule is expressed specifically in the liver on hepatocytes but not on nonparenchymal liver cells, spleen, thymus, kidney or brain. The expression of QIO/Ld in the transgenic hepatocytes is equivalent to La expression on BALB/c hepatocytes, suggesting the animals are expressing physiologic levels of the transgene. The presence of membrane bound QIO/Ld in C3H animals has not caused anti-910 CTL precursors to be deleted, however, because primary in vitro CTL assays show these transgenic animals can specifically lyse QIO/Ld targets. Histopathologic examination of the livers of these animals does not show extensive lymphocytic infiltration or inflammation.

In addition, serum levels of alanine aminotransferase. aspartate aminotransferase, and alkaline phosphatase are also normal, confirming that these animals do not show overt signs of liver rejection. results are ampatable with a t least two different pathways of antigen hardling, a pathway for degradaticm of antigen, and a "pmcessiq" pathway for antigen presenbtim. my+ L monocytcq enes appear t o i n t e r f e r e with t h e processing pathway, either by i n h i b i t i n g production of antigenic material t h a t can associate with Ia o r by i n h i b i t i n g putative intracellular event@) imr0lvb-q the binding of Ia to processed antigen and tmnsprt of annplexes to the cell surface The immunogenicity and antigenicity of synthetic peptides (SP) derived from the sequences of a streptococcal antigen were investigated in macaque monkeys. Immunization with the free peptides of 17 and 21 residues failed to elicit serum antibodies or T cell responses. However, both serum antibodies and lymphocyte responses were elicited by immunization with the SP linked to tetanus toxoid (T) as a carrier. Indeed, SPl7-lT and SP21-TT elicited serum antibodies and proliferative responses of lymphocytes, not only to the SP but also to the native strtptococcal antigen.

recall of SP17-TT or SP21-1T immunized monkeys w i t h suboptimal doses of the native srreptococcal antigen resulted in a significant increase in antibodies, both to the SP and native antigen, confirming that the two SP share antigenic epitopes with the native antigen. The B and T cell epitopes were then determined and the B cell epitopes resides in residue 8-13, whereas the T cell epitope overlaps and consists of residue 7-15. The T cell epitope has an amino-terminal leucine and carboxy-terminal glycine and alanine added to residue 8-13 of the B cell epitope. In spite of the B and T cell epitopes being expressed in SP17 (residues 1-15), the monomer failed to induce serum antibodies without a carrier. However, immunization with dimers of peptide-linked or disulphide-linked residues 1-15, without a canier, elicited both serum antibodies and proliferative responses of lymphocytes. The results suggest that the monomeric SP17 is not immunogenic, whereas the dimeric peptide elicits both antibodies and T cell responses. The minimal T cell-B cell structure required for immunogenicity is now being determined.

lmmunogenlclty Section A

Departments of Immunology and Rheumatology, Mayo Clinic, Rochester, PIN 55905.

Susceptibility to collagen induced arthritis (CIA) in mice maps to the I-A loci in H-29 mice. However, SWR (H-29) mice are CIA resistant, suggesting a role of non-MHC genes. We have recently shown gene complementation between H-29 from SWR and TCR V genes from several non-susceptible strains. CIA has been induced in B10, C3H.A and A fackcrosses with SWR with similar high incidences; 63, 71 and 67% respectively. C57L shares a similar background with B10 and is H-2b, but has the same V TCR mutation as SWR. backcrosses showed a very low incidence (17%) of CIA, and tte arthritis observed was of a much milder and transient nature. The invariant chain associated with HLA class I1 molecules is a 31-33 kd glycoprotein implicated in antigen processing and assembly and intracellular transport of class I1 molecules. Class I1 molecules and invariant chain are expressed primarily by B lymphocytes and antigen-presenting cells such as macrophages and can be induced by interferon-y in a variety of cell types.

TO define sequences involved in the human invariant chain gene regulation, 790 bp 5 ' to the initiation of transcription were subcloned upstream of the CAT gene. Transfection into invariant chain-producing cell lines and non-producing cell lines demonstrated that this 5 ' region displayed tissue specificity and responsiveness to interferon-y. Deletion mutants were constructed to ascertain the functional properties of specific regions of the invariant chain upstream regulatory regions.

These deletion mutants have led to the identification of 3 putative regulatory regions: 394 to 239, 2 3 to 216, and 216 to 165 bp 5 ' to the cap site of the invariant chain gene. Deletion of any one of these 3 regions results in decreased CAT activity. Protein-DNA interactions of these sequences have been characterized by mobility gel shift assay and DNAse I footprinting. Two regions have been identified that exhibit cell type dependent binding of nuclear proteins.

Two color flow cytometry was used to characterize the surface phenotypes of human bronchoalveolar lymphocytes (n=32). The CD4/CD8 ratio was highly variable (0.3-6.6, mean-2.1). A high proportion of the T cells expressed HLA-DR (9-38%, mean=2l%) indicative of T cell activation. However, detectable levels of the I+-2 receptor were expressed on <3% of the cells. CD45R was absent from CD4 cells in most preparations (0-10% mean=3%) suggesting that the cells are inducers of Ig synthesis. UCHL1, a marker of memory cells was present on 68-100 % of lung T cells. UCHLl+ CD45R-lung lymphocytes responded poorly to PHA and ConA but did respond to IL-2 in the presence of accessory cells. Together these data suggest that lung lymphocytes are recently activated memory cells. IL-2 induced lung T cell lines were also characterized for antigen expression and w ( activity. High LAK activity was obtained in preparations containing a high proportion of CD8 cells. These cultures appeared to be suicidal. In contrast, lines with a high proportion of CD4+ had low or absent LAK activity but proliferated in the presence of IL-2 for at least 3 months expressing a CD2 CD45R-phenotype. This abstract is a proposed presentation and does not necessarily reflect EPA policy. The polymorphic second exons of the HLA-DP, and DPD genes have been specifically amplified in vitro by the polymerase chain reaction (PCR) method, using the thermostable DNA polymerase of aauaticus. Sequence analysis of MI3 clones containing the amplified DP sequences from a panel of thirty-four Df typed cell lines revealed only the two previously characterized alleles for DP, . Fourteen allelic variants were defined for DPw Eight of these are associated with

the T-celldefined DPwl-6 types; two subtypes were found for both DPw2 and DPw4. Six additional DP alleles which were previously typed in the T cell assay as blanks were also idenlfied. Based on this sequence information, non-isotopic sequence specific oligonucleotide probes have been developed and used to type a Margarita Betz, Dominic Dordai, Brian E. Lacy, and Barbara S. Fox. Department of Medicine, University of Maryland School of Medicine, Baltimore MD 21201. Murine type 2 helper T cells (Th2) secrete interleukin 4 (IL 4) in response to antigen. Despite the likely importance of these cells, little is known about their priming and expansion in vivo. We have demonstrated IL 4 production in response to a cytochrome p peptide following T cell expansion in vitro. This antigen has not previously been shown to induce Th2 cells. BIO.A mice were primed with a peptide fragment of pigeon cytochrome c in CFA. Lymph node cells were restimulated in vitro with antigen for 5-7 days, Ficolled and rested for 3 days without antigen. Cells were then tested by limiting dilution for the presence of antigen-specific IL 4 producing cells. IL 4 was detected using the IL 4 sensitive cell line CT4S (provided by Dr. W. E. Paul, NIH). The specificity of the response was confirmed by blocking with the anti-lL 4 antibody 11 B1 1. Following in vitro restimulation of the primed lymphocytes with antigen, IL 4 production was detectable from as few as1 03 cells per well. IL 4 secretion was antigen dependent and required both in vivo priming and restimulation in order to be detected. It is not clear why primed lymph node cells, placed in limiting dilution culture directly after removal from the animal, failed to secrete detectable amounts of IL 4 in response to antigen. Suppression is an unlikely mechanism as fresh primed lymph node cells were unable to inhibit IL 4 production by restimulated cells. We are now investigating the factors that may regulate the development of IL 4 producing T cells.

Mark R. Boothby, Ellen Gravallese, Hsiou-Chi Liou. and Laurie H. Glimcher, Department o f Cancer Biology, Harvard School o f Public Health, Boston, MA 02115. regulated pattern, and normally expression is limited to certain cell types such as 6 cells and macrophages. cells is accompanied by the loss o f class I 1 MHC expression. These genes also respond to external stimuli such as the cytokine IL-4, which increases B cell Ia. A region o f the Aa MHC gene activated expression of a CAT reporter gene in a B lymphoma cell line but not in a myeloma cell line. A nuclear protein that bound to two sites within this region was found. This binding activity was present in spleeiis that lack T cells and in B cell lines, but it was absent from all three myeloma cell lines tested. IL-4 treatment of normal and athymic mouse spleen cells greatly increased the binding of this nuclear protein to its A a target sites, concomitant with increased Aa transcription; "Thus, 6 cells contain a sequence-specific binding activity regulated both by IL-4 and by differentiation.

The differentiation o f B cells to plasma lmmunogenicity c210 Peter van den Elsen3. lDepartment of Immunology, The Netherlands Cancer Institute, Amsterdam, ZDepartment of Immunology, Erasmus University, Rotterdam, 3Department of Immunohaematology, Academic Hospital, Leiden, The Netherlands. Human TCR y6 occurs in disulphide-linked (Type 1) or non-disulphide-linked (Type 2) forms, dependent on the use of the Cyl or Cy2 gene segment. The CyZ gene segment can contain a duplication or triplication of exon 2 , which gives rise to different protein forms (Types 2bc or Zabc). It is not known whether functional differences exist between these receptor types. Protein chemical analysis of Type 1 and Type 2bc receptors on functional human T cell clones derived from peripheral blood (PB) has indicated that not only the y chains, but also the 6 chains have a molecular mass and charge which set apart Type 1 and q p e 2bc receptors. Two sets of fifteen clones were randomly generated from PB of two normal donors after selection with the anti-TCR y6-1 mAb, which recognizes all receptor types. DNA rearrangement and mRNA expression analysis of y and 6 genes allowed us to map the specificity of the anti-TCR y6 mAbs 6TCS-1 and TiyA to the V6l and Vy9 gene segments respectively. Subsequently it could be concluded from the analysis of random clones that the majority of Type 1 receptors use vy9, while this preference seems absent in Type 2 receptors. The great majority of Type 1 receptors do not use V6l. while the majority of Type 2 receptors do. This was confirmed by fluorescence analysis of PBL of a large panel of normal donors. We conclude that Vy and V6 gene segments in functional TCR y6 in PB are used in non random combination and that their expression is correlated with rearrangement of the y gene to Cyl or Cy2. We have previously shown that polymorphic residues in the NH2-tenninal half of the p1 domain (amino acids 1-48; hypervariable regions 1 and 2 [PHVl and 21) determine with which allelic or isotypic a chain a particular b chain can achieve efficient cell surface beterodimer expression. This result might be understood in terms of the current model for Ia smcture which predicts that W V l would lie adjacent to region. Therefore, to examine the role of aHVl residues in conmlling hetcroduncr expression. a mutant A d cDNA was created in which the codon for amino acid 11 was mutated to code for the Auk residue at this position. In addition, recombinant A d and A& cDNAs, in which the segments encoding the three a hypervariable regions were exchanged between the two alleles, were used to study the connibutions of other a chain polymorphisms to this process. Interestingly, the polymorphic residues in aHV2 are predicted to lie in a region of the A a chain a-helix which is adjacent to the pHV4 region of the p chain a-helix. Allelic substitutions in this latter region of AD have been shown to similarly affect surface Ia h e t e r o d i i expression.

Taken together, these results suggest that there are at least two spatially separate areas in which the a and p chains interact and that these interactions are affected by polymorphic rtsidues in both areas, conmbuting to the efficiency of heteroditner expression and, most likely, Ia quartcmary conformation. The aim of this project is to identify contact residues of the T cell receptor (TCR) with antigen and/or MHC class I1 molecules. As a model system, a Vp17-containing TCR has been chosen since the majority of VB17' T cell hybrids react with IE molecules of the k,s,d, and b haplotype. T cell hybrids have been made which have a dual reactivity: they are Vp17+ and recognize IE molecules but also show reactivity towards a known antigen, namely chicken ovalbumin (Ova). One such hybrid has been mutagenized with ethyl methane sulfonate (EMS). Mutants were selected on the basis of their survival after stimulation by either antigen or IE. It was expected that mutations in all different kind of genes involved in T cell recognition and T cell activation would be found. Mutants obtained fall into two major groups: 1) loss variants of TCR a or fl chains,T3 or L3T4: 2) mutants with point mutations in one of these genes. We are currently analyzing the mutants biochemically and functionally in order to identify the particular gene affected. Point mutations in the a and p genes of TCR mutants will be localized using the polymerase chain reaction in combination with dideoxy sequencing. Activation of T lymphocytes requires the intracellular fragmentation of foreign antigens and their presentation by class I or class I1 Major Histocompatibility Complex glycoproteins. The direct binding of peptides to class I1 molecules has been shown in a number of experimental systems and its specificity compared to that of T cell activation. In contrast, direct binding of peptides to class I molecules has been difficult to detect; although peptide sensitization experiments and the crystallographic structure of HLA-A2 persuasively argue for its occurence and importance.

In this study, we demonstrate specific binding to HLA-A2 of an influenza matrix peptide (FLU-M1 residues 56-68) that has previously been shown to act as a target for certain HLA-A2 restricted influenza-specific cytotoxic T lymphocytes. We estimate that less than 0.3% of the purified HLA-A2 molecules were able to bind the added peptide. We and others have shown that allorecognition by cytolytic T lymphocytes (CTL) is analogous to T cell recognition of foreign antigens in that both can occur via presentation of antigenic peptides by products of the major histocompatibility complex. We have used peptides corresponding to the alphal helix of selected HLA molecules to analyze T cell recognition of this polymorphic region. The alpha alpha helix of HLA-B44 and -B13 are identical, and show a high degree oh homology with those of HLA-Bw58, -B47, and -B27. Peripheral blood lymphocytes from 5 normal donors were stimulated in vitro with targets expressing HLA-B44 to derive allospecific CTL lines and clones. In some individuals, the allospecific response was almost totally directed against the alphal helix. The ability of peptides corresponding to the alphal helix of these HLA molecules to inhibit and induce lysis as well as to modify other assays of T cell activation will be discussed. Diseases and *National Institute of Child Health and Human Development, d Bethesda, MD 20892 . Classical transplantation antigens are constitutively expressed on cells of all tissues exce t brain. Transaiption is regulated by the interaction of nuclear factors with 5' flanking regions tiat include the class I re latory element (CRE). Previously, the CRE has been divided mto 3 re om on the basis of nuclear t%or blndin . Several studies have implicated the nuclear protein (rI) wfich binds to the inverted repeat GEGGATTCCCCA) of re 'on I as necessary for gene transai tion Although region I is identicarin all se uencedouse $D and L enes, it is not conservefin Qa r y genes. A comparison of the C& from H-2Ld with that of 610, a Qa region gene expressed o y in the liver and fetal olk sac, shows that there are two changes within the inverted r eat se uence (TGaGGAcTCC$A). These differences disru t the dyad symmetry. Another nugotide diierence between H-2L and QlO falls within re 'on Ifof the CRE. However, QlO can bind to the nuclear factor (rII) that binds to region II of the fit2Ld CRE, whereas QlO region I can not bind to the nuclear factor (rI) that binds to the region I inverted re at. To test whether the differences m region I contribute to the restricted tissue e ression of Q l r w e have used site-directed in vitro mutagenesis to make the inverted repeat of]cglO region I like that of the classical class I genes. A change at either base enhances transcription as measured in a transient transfection system. Either change also allows binding of the nuclear factor that binds to the classical r alterations in the CRE regon I contribute to the limited tissue expression of310. The presence of disrupted CRE region I in other region genes likely contributes to their tissue restricted expression.

on I sequence. Thus,

Molecular analysis of T cell receptor structure/function in sperm whale myoglobin specific T cell clones. Jayne S.Danska, Alexandra M. Livingstone, Toshi Isihara and C. Garrison Fathman, Stanford University Medical School, Stanford, CA. We have undertaken structural characterization of the T cell receptors (TCR) utilized by a well defined panel of murine DBA/2 T cell clones that recognize epitopes within the 110-120 peptide of sperm whale myoglobin (Sp WMb) presented by I-Ad or I E . Only 2 of 14 independent clones show alloreactivity for 10 WHC haplotypes. Using the polymerase chain reaction (PCR) and DNA sequencing of the TCR a and fc chains from matched sets of clones bearing either WHC restriction or epitope specificity in common, we are addressing structural relationship between TCR and MHC/antigen for this model system. Among 6 I-Ed restricted T cell clones reactive with SpWMb 110-120, all have highly homologous TCR 6 chains associated with a minimum of three different TCR a chains, some of which are derived from novel V gene families. To further characterized the specificity of these clones we are generating substituted peptides to identify residues within the epitope important for interaction with TCR or restricting MHC molecule. Functional verification of the relationship between given TCR primary sequences, and recognition capability will be addressed by transfer of the a and/or 8 chafns cDNAs created by PCR amplification into T-cell hybridomas expressing endogenous TCR genes of known sequence and specificity. with MHC fine specificity. The differential impact of substitutions with the N-terminal and C-terminal portions of the Apl domain is consistent with models of AuAp structure in which the N-terminus interacts with peptide while the C-terminus interacts with both peptide and the TCR.

or AaUApu-resmcted T cell clones. The antigens tested were L-tymsine-p-lmmunogenicity C 218 EXPRESSION OF THE Q4P GENE, Patricia M. Day, Katherine E. LaPan and Jeffrey A. Frelinger, Department of Microbiology and lmmunolog University of North Carolina at Chapel Hill, Chapel Hill, NC 27599. Generally, the transcription of Class I genes tom the Q a a region is limited to tissues of hematopoietic origin. Previous work in our lab demonstrated widespread transcription of the Q4 gene in the 81 0.P mouse, with high levels of mRNA found in liver, lung, lymph node, spleen, testes and thymus. Less RNA was present in muscle and brain tissues. However,,it is not known whch individual cell types within these tissues are responsible for the transcription of the Q4 gene. We raised polyclonal antisera against a synthetic peptide, derived from the predicted amino acid sequence of the Q4P transmembrane region. We selected this region since it is the most locus specific. This antisera immunoprecipitates a Class lsized protein.

A monoclonal antibody, directed against the same peptide, has also been produced. SV40-transformed H-2P fibroblasts show an abundance of Q4 message. Suprisingly, indirect immunoluorescent staining with the monoclonal antibody reveals a cytoplasmic localization of the protein with a perinuclear concentration. Different patterns have been observed in examination of the H-2b em onal carcinoma cell lines 402AX and PCC4. QGspecific antibodies allow us to identify the cell ty es which express the24 gene product. The application of in situ hybridization techniques will correlate the cellular site ofmRNA synthesis and protein detected by antibodies. Understanding the paltern of expressim of the Q4 gene is the first step in determining the so far elusive function of these MHC genes. and IL2R mRNA a f t e r mitogenic stimulation. Antibodies against CD45R, but not against C045 common determinants, synergise with suboptimal doses o f mitogen t o induce IL2 and IL2R mRNA expression, suggesting t h a t CD45R molecules are operative i n transmembrane s i g n a l l i n g i n immature thymocytes. There i s also an i n d i c a t i o n from Northerns using CD45 probes t h a t CD45p180 mRNA i s not induced i n activated CD348-thymocytes as i t i s i n mature T c e l l s .

These r e s u l t s support the idea t h a t CD45R+ molecules are essential f o r generating signals required f o r c e l l survival w i t h i n the productive intrathymic lineage. We examined a panel of Thl and Th2 T cell clones for the ability to induce antibody synthesis in a Mishell-Dutton culture system under cognate B-T cell conditions: Our findings indicate that both Thl and Th2 T cells are heterogeneous, i.e., some but not all Thl and some but not all Th2 clones have the capacity to induce antibody under these conditions. We examined the effect of 7-irradiation or rnitomycin-c pretreatment of our Thl and Th2 clones on their ability induce antibody synthesis. Asano et al. (J. Immunol 138:667) have reported that Th2 clones are exceedingly sensitive to 7-irradiation, with doses as low as 500 rads abrogating the ability of Th2 clones to induce antibody synthesis. We found that while the helper activity of Th2 clones was very radiation sensitive, helper activity in Thl clones was very radiation resistant. Thl clones given 2 0 0 0 rads of irradiation were as effective as unirradiated clones in inducing anti-TNP plaque forming cells (PFC). Moreover, when used at higher T cell/B cell ratios in culture, irradiated Thl clones were more effective than unirradiated clones in inducing antibody synthesis. The effect of 7irradiation on Th2 clones was not simply due to inhibition of proliferation, since mitomycin-c pretreatment of the clones had little effect on helper activity.

CONSERVATION, Alexander L. Dent, Pamela J. Fink and Ste hen M. Hedrick, Department of Biology, University of California, San Diego, 8A 92093. The p chain gene of the murine T cell receptor has been shown previously to have an alternative1 spliced form of message. This message contains a novel exon, termed C& which is inserted between the VDJ and constant region exons. We have studi6d expression of the CpO exon at the mRNA level by RNase protection. We have found that about 1% or less of of p messages in normal T cell clones contain the CgO exon, whereas p m e s a es. in the thymus contain the exon at 10-20 fold higher levels. To address t i e importance of the C 0 exon in the immune ,system, we have undertaken a phylogenetic approach. \y cloning and sequencin the rat analogue of CpO, we have found that while the rat exon is very simifar to the mouse exon, both donor and acceptor RNA splice signals are defective in the rat CpO gene. This implies that rat CpO cannot be spliced into rat p m e s a 8s. Furthermore, we have sequenced the analo ous region to mouse CpO in 8 e human p chain locus, and have found no stretct of sequence remotel homolo ous to CpO. Because CpO is not conserved evolutionarily, we beIeve that ! he CpO gene element does not sewe an important function to the immune system of most vertebrates. 1988) . We found that one arrdno acid substitution at a VDJp juntional region position found to be highly conserved in pigcon cytochrome c-specific TCR's results in a change in antigen fine specificity, while an* change abolishes all detectable responses characteristic of the D6 TCR. We will present the results of mutagenesismansfection analyses of two other pigeon cytochrome c-specific TCR'S. The murine CTL response to human class I molecules is 1-2 orders of magnitude lower than the response to murine alloantigens, due to structural differences between human and murine homologs. We investigated whether this discrepancy could be overcome by exposure of developing T cells to human class I molecules in transgenic C57BL/6 mice expressing HLA-A2.1.

! c222 ZHE BDIXUJIAR BASIS OF - . LBVe DiGiustn ard Ed P d l U b 2 r , Div- of

S ch mice expressed HLA-A2.I in spleen, bone marrow and thymus at levels similar to those of endogenous H-2 molecules. However, the frequency of CTL specific for other human alloantigens remained similar to that of normal mice. The frequency of HLA-A2.1 restricted influenza specific CTL was 1-2 orders of magnitude less than the frequency of H-2 restricted CTL. These results indicate that the poor response of murine CTL to human class I antigens is not determined by selection in the thymus, but by species-specific constraints on the interaction of MHC antigens with T-cell recognition structures. While the mice are tolerant to HLA-A2.I expressed on murine cells, they still respond to HLA-A2.1 expressed on human cells. The epitopes defined by such clones are present on HLA-A2.1 positive human cells derived from several different tissues. Such epitopeJ are not dependent upon the species of p2m associated with the class I molecule, nor upon the structure of the attached carbohydrate. The results suggest that one or more highly conserved normal human proteins contribute to the formation of such epitopes, and provide an explanation for the failure of CTL raised against class I molecules on human cells to recognize the same molecules expressed on murine transfectants. This suggests that normal endogenously expressed molecules may also be important in the formation of epitopes on class I antigens recognized by allospecific CTL. We previously demonstrated t h a t several subclones derived from a C03+, CD4-/CD8-T-cel I l i n e have undergone secondary rearrangements a t t h e T-cell receptor (TCR) a locus w h i l e maint a i n i n g i t s o r i g i n a l TCRB and Igh D-J rearrangements (Marolleau e t . al., i n press). These secondary rearrangements r e s u l t i n t h e j o i n i n g of germline Va and J a gene segments which replace the p r e -e x i s t i n g Va-Jacmplexes of t h e parental T-cell line. I n an e f f o r t t o examine t h e molecular mechanism responsible f o r these Va-Ja gene replacements, t h e s t r u c t u r e s o f TCRa cDNAs prepared from both t h e parental and subcloned T-cell l i n e s were determined. I n addition, Northern b l o t and Southern b l o t analyses were performed on both t h e parental and subcloned T-cell l i n e s using a panel o f Va and J a probes. Our r e s u l t s i n d i c a t e t h a t : 1) Secondary rearrangements r e s u l t i n both productive and non-productive Va-Ja j o i n s , 2) The mechanism whereby secondary rearrangements occur i s a d e l e t i o n event t h a t involves germline Va genes 5 ' t o t h e p r e -e x i s t i n g Va-Ja complex j o i n i n g t o J a The class II major histocompatibility complex (MHC) antigens are a family of integral membrane proteins whose expression is tissue-specific and developmentally regulated. A pair of consensus sequences, X and Y, separated by an interspace element, is found upstream to all class II genes. Deletion of each of these sequences eliminates expression of class II genes in vitro or in transgenic mice (1-3). Furthermore, the absence of a specific binding protein for the HLA DR a X box in patients with severe combined immunodeficiency disease whose cells lack class II suggests a critical role for these proteins in class II gene transcription (4). report the cloning of a Agtll cDNA encoding a DNA binding protein (human X-box binding protein, hXBP-1) which, like the proteins in whole nuclear extract, recognizes both the X box and interspace elements of the human DRa and murine Aa genes. The hXBP-1 cDNA hybridizes to two RNA species, 2.2 kb and 1.8 kb in human, that are expressed in both class II positive and class II negative cells.

hXBP-1 does not cross-hybridize to two murine Aa X box binding cDNAs recently isolated in our laboratory which also recognize the DRa and A aX boxes. These observations provide evidence for the existence of multiple X box binding proteins which recognize a common or overlapping motif. Chromosome mapping studies demonstrate that hXBP-1 arises from a multi-gene family two of whose members map to human chromosomes 5 and 22. Taken together, these data suggest a high degree of complexity in the transcriptional control of the class II gene family. FRANCE . In an attempt to analyze positive or negative in vivo regulation of clonal expansion of cytolytic T lymphocytes (CTL), we immunized BlO.BR mice with the Kb specific CTL clone KBS-CU). and we tested whether T cells obtained from such mice would influence the in vitro development of the CTL clone KBS-C?O. A clone-specific helper effect has k e n observed, which is mediated by CD4+ splenic cells from immunized mice. Control immunizations of B1O.BR mice with Ti negative variants of suggest that this growth regulation involves the recognition of the Ti of KBSC20. The precise nature of the antigen recognized on the CTL clone, the possible involvement of Ti determinants with or without MHC products u e now under investigation. We have shown that Thy-]+ dendritic cells present in the epidermis of mice (dEC) express CD3 associated V73 and V61 gene products. We have produced a monoclonal antibody directed against V73 and found that a wave of cells appearing at the earliest stages of fetal thymic development express V73. Phenotypic and functional analysis of V73+ cells in the early fetal thymus indicates that they have characteristics in common with the V73+ dEC.

Both populations express high levels of Ly-1 and are Ly-6C+. Neither express CD4 or CD8. Interestingly, the V73+ fetal cells express elevated levels of IL-2 receptor, indicating that they may have been activated. Functional analysis demonstrated that, unlike other fetal thymocytes, the V73+ cells can be stimulated to produce lymphokines and lyse a panel of target cells which are also lysed by the adult Thy-l+ dEC. These results raise the intriguing possibility that the first receptor-bearing component of the T cell system to appear in ontogeny might give rise to the Thy-]+ dEC. (3ritical to an understanding of the function of cells bearing the gamma-delta T cell receptor will be an understanding of when and where such cells function. In order to investigate this, we have used a variety of techniques (in situ hybridisation, cDNA cloning, and PCR) to examine m a s of TCR gamma delta gene expression. One conspicuous site of expression is the intestinal epithelium, which is by contrast almost devoid of TCR alpha beta expression. Interestingly, the V gene segment usage in this location is quite specific and is different to the specifcity that we have found in the spleen and in the thymus, and that others have found in the skin. This specificity suggests in turn that expression of the resmcting elements recognised by gamma delta may also be spatially nxticted. Extensive analysis of junctional diversity can pmvide infomation on the diversity of antigen nxogniscd by gamma delta. The basis for selective expnssion of V gene segments may in part lie in different requkments of the V-gannna gene pmoters. To examine this, the Wnscriptional capabilities of the various gamma gene promoters in T cells MURINE TCR GAMMA GENES: DISTINCT SPATIAL RESTRICTION OF V-GENE SEGMENT USAGE, Adrian thyday*, Susan Kyes*, Simon Carding#, Charles A. Janevay, being compared by linkage to the chloramphenicol acetyl transferase gene.

Biochemistry, University of Wisconsin-Madison, Madison, 11 53706. Murine strain A sublines A/J and A/WySnJ have a genetic polymorphism that regulates serum immunoglobulin responses to several protein antigens. Strain A/J secondary IgG1 responses to bovinv gamma globulin, oralbumin, hemocyanin and galactosidase L-ere 50-, l o -, i -and 4-fold greater, respectively, than A/WySnJ responses.

Subline A/HeJ is a low responding strain like .l/WySnJ. Analysis of H-2 class 1 and class I 1 molecules provided no evidence for a breeding error to account for the genetic polymorphism. Instead, an important immune response gene outside H-2 may hare been heterozygous when the sublines diverged, and the polymorphism resulted from sezregation and differential allele fixation. A mutation subsequent to subline divergence is also a possible source of the polymorphism, but is less likely. The high responder phenotype inheritance pattern in (A/WySnJ X A/J)Fl, F2, and backcross mice was consistent with segregation of a single, recessive gene. We named this locus L a for the strain A sublines that define it; strain A/J represents the k a h allele and strains A/YySnJ and A/HeJ represent the allele.

Hayes, Keith D. Hanson, Faye Nashold, and David J . Miller, Department of

The secondary IgGza responses Were also affected.

Several different proteolytic digests of denatured SEB have been tested for their ability to stimulate T cell hybrids to produce IL-2. A tryptic digest that retains activity has been fractionated by HPLC and the stimulatory component is being analyzed. Examination of the amino acid sequence of SEB and the proteolytic cleavage sites has led us to synthesize several peptides for analysis. These peptides, and their analogues, will be tested for their function in vivo and in vitro. To understand the interactions involved in the famation of peptide-MHC complexes, an assay has been developed to detect DR specific binding of peptide analogues of T e l l determinants to cell surfaces. EBV msformcd B cell lines (BCLs) w m incubated with biotinylated peptide followed by FlTC conjugated streptavidin, and then anaiysed by flow cytomctg. A panel of BCLs homoygous for diffmnt DR types bound analogues of peptide 307-319 from influenza virus haemagglutinin (previously shown to be a helper T cell determinant restricted through DR1) to varying degrees, w h m no binding was observed to the DR-BCL RJ225. Binding could be specifically inhibited by the natural unbiotinylated T cell determinant or other DRl restricted determinants. Competition by a range of peptides revealed quantitative diffmnces in their ability to bind DRl. The assay is currently being used to generate a detailed model of the complex formed betweenHA307-319andDRl. and Trp for Leul 6. and HLA-A2.3 differs from HLA-A2.I by the substitutions of Thr for Ala 4 9 Glu fo??a1152 and Trp for zeu156. Residues 9 and 95 in the 8-sheet of the molecule, and residues 149, 152, and 156 in the a-helix are thought to interact with bound peptide or the TCR. To evaluate the role of these residues on CTL-defined epitopes, two genes were constructed that encoded novel molecules which differ from HLA-A2.1 only at residues 9, 43, and 95, or at residue 156. The effect of a-helix substitutions on serologic and CTL-defined epitopes that varied between HLA-A2.I and HLA-A2.3 were evaluated by constructing genes that encoded the individual differences at residues 149, 152, and 156, as well as additional non-naturally occuring substitutions at these same positions. HLA-A2.1 specific CTL were found that were: (1) insensitive to substitutions at either residues 9, 43, and 95, or residue 156, but were lost when all four positions were changed; (2) dependent upon the residues 9, 43, 95, but not residue 156; (3) dependent upon residue 156, but not residues 9. 43, and 95; and (4) dependent upon residues 9, 43, 95, and residue 156. Further epitope mapping with the a-helix mutants demonstrated that a substitution at residue 152 often destroys an epitope not affected by substitution at residue 156. Even conservative substitutions at position 152 were more disruptive than nonconservative changes at residue 156. Residue 149, while important in defining an mAb epitope, had no effect on any CTL epitopes. These results indicate that spatially separate residues in the a-helix and 8-sheet of the molecule can contribute to the epitope recognized by a given CTL. Furthermore, considerable complexity must exist in the spectrum of T cell receptors utilized to recognize HLA-AZ, as 28 CTL clones exhibited 21 distinct fine specificity patterns. to follow the evolution of these class I types, to discern the chief selective pressures on its members and thus indicate the probable functional properties of the antigens. A cosmid library was screened for class I genes. 91 clones were mapped and could be grouped into 17 clusters of contiguous DNA spanning 1,264 Kb. By hybridisation studies, 61 class I genes/ gene fragments could be distinguished. Transfection analysis revealed that 10 genes could be expressed as cell surface antigens: two genes, in a block of duplicated DNA encoded serologically defined RT1.C products, the other 8 genes gave rise to novel class I antigens detected by the xeno-antibody 0x18.

Using region specific probes, we could detect clear rat homologues of the mouse Qa and H-2 genes, however there were only two rat genes with limited homology to the mouse Tla genes. The analysis showed extensive remodelling of the class I region in the evolutionary gap between rat and mouse. While the immunological role of T cells bearing the ap T cell receptor (TCR) has been well characterized, much less is known about the function of T cells bearing the $3 TCR. We investigated the role of TCR $cells in the immune response to Complete Freund's Adjuvant (CFA). After immunizing mice with CFA, we observed a greater than 26-fold increase in the number of TCR $3 cells present in lymph nodes draining the sites of Immunization, compared to a 3-4-fold increase in the number of TCR ap cells. There were at least three different species of 3TCR's expressed on these cells in the draining lymph nodes, including two protein products derived from the rearrangements of Cyl and Cp, and one product derived from Cy4. 37% of TCR yS cells from immunized lymph nodes expressed the IL-2 receptor in vivo. and these cells constituted roughly 50% of the proliferative response of total lymph node T cells to 11-2. Tse, et al. (J.lmmun..Vol.l25, p.491.1980) have demonstrated that at least three cell types are involved in the T cell proliferative response to antigen, including an antigen specific-T cell, an antigen-presenting cell, and a T cell that is found in unprimed lymph nodes or spleen, which has been termed the recruitable cell. We have utilized their approach of analyzing the slope of log cell number-log response curves to examine whether TCR @ cells can function as "recruitable" cells. We found that TCR yS cells as well as TCR ap cells can function as recruitable cells in this system. These data suggest that TCR $3 cells can participate in the immune response without being specific for the antigen.

Analysis of the membrane associated phosphoprotein profiles of B cells harvested from cultures of resting cells exposed to IL-4 for 18-24 hrs reveals the presence of phosphoprotein with an Mr in the range 75-80,000. destroy the autoradiographic signal from this phosphoprotein suggesting that it is phosphorylated upon tyrosine residues. appearance of this molecule, and LPS also apparently fails to result in the presence of a 75Kd structure in the phosphoprotein profiles. anti-IL-4 antibody. 11811, in the cultures prevents the appearance of the 75Kd phosphoprotein. The genes for T N F -a and TNF-p are tandemly arranged on mouse chromosome 17, with only 1.1 kb separating the 3' end of the TNF-p mRNA from the 5' end of the TNF-a mRNA. Yet, the two genes are independently regulated. In vitro transcription and nuclear run-on experiments indicate that the two genes are transcribed from independent promoters. In macrophages, which express TNF-a but not TNF-p, only the TNF-a promoter is active. In T lymphocytes, which can synthesize both proteins, both promoters are active. Activation of either cell type results in a moderate (up to 10-fold) increase in the level of transcription, while mRNA levels increase more than 1WfoId under the same conditions. Interestingly, the TNF-p gene is aanscribed 10-fold less than the TNF-a gene in T lymphocytes, although the corresponding mRNA is more abundant. These results indicate that the accumulation of both TNF-a and TNF-p mRNA after cell activation and their relative steady state levels are controlled mostly at a post-transcriptional step. Acanomycin D chase experiments reveal that TNF-a mRNA stability in macrophages is not significantly altered after activation by LF'S, and therefore that stabilization done cannot account for the observed accumulation of TNF-a mRNA. In order to examine more closely which elements are required for the regulation of TNF-a and TNF-P mRNA abundance, we constructed hybrid genes combining putative control regions of TNF-a and TNF-p with known constitutive control elements. Results obtained from the transfection of these hybrid genes into various cell types indicate that elements located both 5' and 3' of the coding sequence are required for the proper regulation of TNF-a and TNF-P mRNA abundance. Celiac disease is characterized by small intestinal mucosal injury and malabsorption. Disease is activated when a genetically susceptible host ingests wheat gliadin or similar proteins (i.e., prolamins) in rye and barley. D region specif icities -DR3 and -DQw2. class I1 D-region haplotype associated with celiac disease is extended and also includes genes in the HLA-DP subregion. chain gene with those encoding DR3 and DQw2 may indicate that the HLA haplotype associated with celiac disease exhibits an unusual degree of linkage disequilibrium or, alternatively, that disease susceptibility involves the gene products of more than one HLA locus. To characterize possible HLA structural variants unique to celiac disease, the polymorphic second exons of the expressed DR, DQ and DP genes were amplified from genomic DNA of celiac disease patients, and their nucleotide sequences determined. Our studies indicate the presence of a unique constellation of D region genes associated with the celiac haplotype, and exclude the presence of a disease specific DR, EQ or DP structural gene variant in this disease.

Disease susceptibility is strongly associated with the HLA class I1 We recently determined that the HLA This same population of T cells contains a high frequency 0 1 % ) of cells which will respond to a given allogeneic MHC protein, or to differences at two other genetic loci termed Mls, in conjunction with MHC. We have transfered the a and B chain genes from a pigeon cytochrome c/EL specific, alloreactive. and MIS' specific murine T cell clone into an unrelated host T cell.

We demonstrate that the genes encoding a single a B receptor chain pair can transfer the recogntion of self MHC molecules c m p l e x e d with fragments of antigen, allogeneic MHC molecules. and an M1sc (Hls-2) encoded determinant.

In this case the transfer of antigen specificity and alloreactivity requires a specific a8 receptor chain combination, whereas Mlsc reactivity can be transfered with the 6 chain alone into a recipient expressing a randomly selected a chain. Site directed mutagenesis of the Ja region has also been performed in an attempt to identify sites involved in the alloreactivity of this T cell clone.

In addition. we demonstrate that a single amino acid change in the V-J junction of the a B receptor can alter MHC restriction a s well a s antigen fine specificity.

Department of Genetics, Washington University School of Medicine, St. Louis, I(0 63110. The S49 tumor sublines are variants isolated from a sing1 parent BALB/c tumor which demonstrate locus-specific shut-off of their Kd, Dd and L8 genes. Four phenotypically different sublines were characterized at the DNA and RNA level. Southern blot analysis indicated that no major chromosomal deletions have occurred, and treatment of the sublines with 5-azacytidine had no effect on Class I expression. between loci are unlikely. None of the repressed Class I antigens could be induced with interferon even though the expressed antigens were fully inducible. Northern blot analysis revealed message only for the expressed antigens, showing that the repression mechanism is acting at the transcriptional level. Rnase protection analysis confirmed this result and demonstrated that the transcriptional repression is exquisitely specific for the Kd, Dd and Ld genes as other "class I-like'' messages are present in. all the cell lines. expressing Class I antigens from both fusion partners, but the negative Class I antigens originating from the S49 partner were not expressed. Lymphokine gene expression was examined in a panel of 116 short-term murine T lymphocyte clones derived by single-cell micromanipulation from allogeneic mixed leukocyte cultures. About 30% of clonable T cells, including both CD4+CD8-and CD4-CDW cells, could be expanded for assay at an average of 22 days after cloning. Following stimulation with concanavalin A or anti-CD3 antibody, all clones secreted detectable granulocyte-macrophage colony stimulating factor (GI(-CSF), interleukin-2 (IL-2) and IL-3, but CD4+ clones on average secreted higher 'levels of each lymphokine than CD8+ clones. clones (85%-96%) expressed detectable GM-CSF, interferon-y and IL-3 mRNA and 11% expressed IL-4 mRNA. When the frequencies of co-expression of any pair of lymphokine mRNAs vere determined, all were found to correspond to the values predicted for random assortment of the individual frequencies. For example, among 13 IL-4-positive clones, 11 also transcribed interferon-y, giving the frequency of double-positive clones expected for random association (9.6% 10.8%). Expression of the four lymphokine genes therefore segregated independently among the clones and did not allow the division of T cells into subsets vith distinct patterns of lymphokine synthesis. greater than 20-fold in the adult liver cell line, 2 to 3 fold in the macrophage cell line and just slightly in L-cells. We have subcloned the region 5' to the li gene which contains sequences that may be important to regulating expression of the li gene. This region includes a 15-mer (CCTAGAAACAAGTGA) which occurs 5' to many IFN?I regulated genes. Current research has been directed towards identifying and comparing proteins from nuclear extracts prepared from control and IFN-7 treated cells which bind to this region (-260 to -1 1). This data indicates the li molecule may be expressed in cells not known to be directly involved in the immune response. Although there has been considerable interest in the recently identified gamma, delta T cell receptor, relatively little is known as to its function. During our studies of the human immune response to autologous B cell lymphomas, we generated cytotoxic T lymphocytes (CTL) specific for tumor idiotype. These CTL lysed only autologous tumor cells and none of a large panel of other autologous and allogeneic cells. inhibitable by anti-idiotypic and anti-immunoglobulin antibodies but not by a panel of classical anti-MHC antibodies. Phenotypic analyses showed that these CTL were CD3+, CD4-, CD8-, and express the delta, and presumably gamma! T cell receptor. Such CTL can be used to gain new insights into the function of the gamma, delta T cell receptor and T cell recognition of immunoglobulin, and may prove clinically useful in adoptive immunotherapy.

Tumor lysis was

for EBV-induced antigens. Furthermore, LCL variant .221, which does not express any HLA -A, -B. or -C determinants. is killed by cultures primed to LCL-.180. Antibody blocking experiments suggested that this killing was mediated by T cells, and was not restricted by known class I antigens. Depletion of leu 19 positive cells from the effector population did not eliminate cytotoxicity on LCL-.221. Cold-target blocking studies further suggested that the class 11-nonexpressing LCL-.180 and the class I-nonexpressing LCL , 2 2 1 share residual deterninant(s) other than HLA class I or class I1 that can restrict cytotoxic T cell responses to EBV-induced antigens.

National Jewish Center for Immunology and Respiratory Medicine, Denver, CO 80206 It is uncertain to what extent lymphokines can be differentially produced by activated primary T cell populations. To determine if IL4 and IFNr were differentially regulated in uncloned human T cells from adults (Ad) and neonates (Nt), these mRNAs vere measured by in situ hybridization after maximal stimulation by ionoaycin and PMA. IL4 mRNA was detected in 1.2% of total (TL), 3.5% of CD4', 10% of CD4' CD45R-, and 0.1% of CDB' Ad T cells, but in none of the TL, CD4+, or CD8' Nt T cell populations (virtually all Nt T cells were CD45R').

In contrast, IPNr mRNA was found in 39.42% of TL, 34.36% of C D 4 ' , 51% of CD4' CD45R-, and 58% of CD8+ Ad T cells, but only 2.3% of TL, 2% of C D 4 ' , and 4% of CD8' Nt T cells. These results agreed with other estimates of IL4 and IFNr production based on RIA of cell culture supernatants, RNA blotting, and gene transcription assays. In contrast to IL4 and IPNr, IL2 was expressed in similar amounts by Ad and Nt T cell fractions, as well as the Ad CD4' CD45R' and CD45R' subsets. Thus, the capacity for increased IL4 and IFNr production by Ad T cells appears attributable, in large part, to the postnatal acquisition of the CD45R' subset (putative memory T cell population). Aowever, additional mechanisms exist which act transcriptionally to limit IL4 production by both neonatal and adult T cells. Such selective expression may be important for restricting the potentially pleiotropic effects of certain lymphokines t o appropriate responder cells. We observed significant inhibition (>70% at 800 ng/ml) of the presentation of wOVA and of OVA 323-339 by the anti-Ap 57-78 peptide mAb's. exhibited significant inhibition. 61 peptide mAb's.

after incubation with antigen +/-mAb, indicate that the inhibition occurs at the level of antigen presentation. DG11, was also observed for the anti-p chain peptide mAb's and to a lesser extent by the anti-a chain peptide mAb's. peptide sequences are capable of interfering with antigen presentation, in vitro. Supported by NIH grant, AI-14764.

The ovalbumin (OVA) I-Ad restricted T cell hybridoma, DO1l.10 was used to

The anti-I-Ad mAb, MKD6, also Much less inhibition was observed with the anti-% 43-Experiments with glutaraldehyde fixation of the B1D.p cells before or Inhibition of I-Ad allorecognition by the T cell hybridoma.

These results indicate that mAb's generated against class I1

Rijllinghoff, Institute for Clinical Microbiology, University of Erlangen-Nurnberg, 8520 Erlangen, F.R.G. and the *Institute for Clinical Immunology and Rheumatology, University of Erlangen-Niirnberg, 8520 Erlangen. F.R.G. Recently we have shown that cloned L . major-specific L1/1 T-helper cells of type 2 (Th2cells), when stimulated with antigen, are able to induce polyclonal B-cell proliferation (1). We here present evidence demonstrating that this process is dependent on a direct cellcell interaction between T-and B-cells. which in the effector phase, i.e. during stimulation of the B-cells by activated T-cells, can be mediated by a mechanism other than cognate interaction. This conclusion is derived from experiments, in which highly purified resting B-cells were polyclonally stimulated by L1/1 T-cells triggered by an anti-T3 monoclonal antibody, in the absence of antigen. The triggering process was independent of the presence of the Fc part of the antibody and occurred in cultures devoid of macrophages. Thus, the well established cognate recognition does not appear to be the only way of B-cell induction by T-helper cells of type 2. Studies show that a proportion of the peripheral blood CD3' T lymphocytes do not express CD4 or CD8 and are called double negative T cells. They normally have a 1 6 TCR. However, another population of double negative T cells exists that expresses the a@ heterodimer. W e have purified and expanded such a population isolated from the peripheral blood of a healthy individual and studied i t s phenotypical and functional characteristics.

The c e l l s are CD3'

CD4-CD8-, positive for WT31 and negative for the NK markers. They express a and p mRNA b u t lack ymRNA.

From surface iodinated cells were precipitated w i t h monoclonal pF1 two closely running bands (46 & 48 Kd) .

Functional studies demonstrate that they proliferate to antiCD3 and PHA, t h i s response was blocked by cyclosporin A. There was no NK lysis b u t antiCD3 induced l y s i s of target cells. The cells responded t o IL-2 and IL-4 as previously shown for other T c e l l s , b u t also t o IL-3, a lymphokine thought t o affect mainly stem cells and not previously shown t o stiaulate growth of mature cells. Long term growth of these c e l l s was also maintained by these cytokines .

Roberto Biassoni , Silvano Ferrini , Rafck P. Sekaly , and Eric 0. Long , Laboratory of Immunogenetics, National Institute 2f Allergy and Infectious Diseases, NIH, Beth-MD 20892, and Istituto Nazionale per la Ricerca sul Cancro , 16132 Genova, Italy. CD3-cells grown in vim in the presence of IL-2 acquire the ability to l~s e a wide variety of tumor cells in an MHC-unrestricted manner. We have previously shown that CD3-16 clones expressed the CD3 epsilon gene but no functional transcript from CD3 gamma, CD3 delta, TCR alpha, TCR beta and TCR gamma genes. This result suggested that these CD3-16' cells represented an early stage in T cell differentiation. To test for expression of the TCR delta gene in these cells, RNA from a panel of CD3-16' clones and from three highly enriched populations was hybridized with several DNA fragments of the delta locus. Abundant transcripts were detected with a C delta probe and a J delta 1 probe in 6 out of 8 clones and in all three populations.

At least four different transcripts were present with sizes similar to those found in CD3' TCR gamma-delta' cells. However, the TCR delta transcripts in CD3-16' cells are most likely derived from unrearranged genes because no rearrangement could be detected in DNA from an enriched population using a J delta 1 probe, and because these aanscripts hybridized to a DNA fragment corresponding to the unrearranged genomic sequence 5'-upstream of J delta 1. Expression of unrearranged TCR delta genes in CD3-cells provides further evidence that these cells belong to the T cell lineage. functional capabilities and by differential release of either IL2 or IL4 upon activation. We have produced a new monoclonal antibody to CD45 which has allowed us to separate normal murine CD4+ cells into two populations based on the density of expression of CD45 epitope. The separated populations seem to be analogous of subsets found in cloned T cell lines. CD4+ T cells with high density of cell surface CD45 after polyclonal activation produce IL2 and mRNA encoding IFW and IL2. It does not produce IL4 or IL4 mRNA. CD45 low density population on the other hand transcribes mRNA for IL4 and secretes IL4 protein. Data will be presented to demonstrate that the two subsets of normal CD4+ cells also differ in their proliferative response to mitogenic stimuli and to exogenously added growth factors. The substitution of V to L at 95 was the only change that could be discriminated by 2 of 9 allospecific CTL lines. suggesting that those 2 CTL lines recognize A2.1 plus a peptide whose presentation andlor binding is affected by the V to L substitution in the floor of the peptide binding site. In contrast, the L to W substitution at 156 (but not the other 2 substitutions) abolished the ability of the A2 molecule to present the viral peptide to 24 out of 25 peptide-specific A2.1-restricted CTL lines, suggesting that this substitution alters the presentation of the influenza matrix peptide but does not inhibit the ability of the peptide to bind to the A2 molecule. Although y8TCR.s have a great potential for diversity, it remains to be determined whether this potential is realized in terms of expressed y6TCRs. Preliminary studies in several laboratories have indicated that y6TCRs expressed in earlg t h r c y t e s and adult epithelial tissues are more restricted in diversity com ared to adult TC expressin thymocytes. We have derived a panel of cloned dendritic epigrmal T cells (JETC, lines and5ybridomas that express at least three types of y6 receptors -C 8, Cy26 and C N . Immunoprecipitation, Northern and Southern blot analyses, and sequence anazses of L gt 10 cloned cDNA or olymerase chain reaction ( K R ) amplified cDNA segments have been used to anal ze in detail &e extent of diversit in the expressed y and 6 chains and whether restricted airin o?y and 6 chains occurs. Our resu& indicate that for this panel of cloned cell lines 7 and ! irkg is nonrandom and that variability in certain types of receptors appears to be restricted. %owever, we have observed significant 6 chain diversity in these cells that is obtained by the use of multiple V-regions, and N-region and junctional diversity. We are investigating whether the observed y and 6 chain pairing, and pattern of 6 chain diversity are present in other $TCR bearing cells or whether they are only characteristic of DETC. Activation of CTL precursors from murine unprimed spleen cells with rIL-2 or rIL-4 results in distinct lytic spectra, depending on which lymphokine is present. We have used allo-stimulation in limiting dilution analysis with subsequent testing on an allo-specific target (A20) and an MHCdeficient, non-specific target (RlE). In the presence of rIL-4 exclusively allo-specific CTL are generated, while rIL-2 supports a proximately equal numbers of precursors that k~ll A20 and RIE targets. Dose response analysis of rIL-2-supported killing activity indicates that the lytic spectrum is independent of the amount of rIL-2 used, and therefore this IL-2 effect is intrinsic in its activity on unprimed spleen cells. Mixing experiments indicate that rIL-4 can partially override the effect of IL-2 on the generation of non-specific killer cells. Split well analysis and cold target inhibition experiments are in rogress to ascertain the actual proportion of specific killer cells which can be generated with rIL-2. be. are also testing the ability of cofactors, such as IL-1 and IL-6, to optimize the response of IL-4 generated CTL. We conclude that IL-4, not IL-2, must be used when CTL are generated from unprimed spleen cells in mice. t r a n s c r i p t s i n y/6 TCR populations. I n t e r e s t i n g l y , these same V genes, as well as a further+crosshybridizing V gene previously designated Va7.2, are expressed by peripheral a& TCR c e l l s as 1.6kb TCRa transcripts. These data suggest t h a t B2A2-DN TH represent a developmentally unique subset i n which both V6 and Vg segments are non-randomly expressed.

Furthermore they i n d i c a t e t h a t there i s considerable overlap between the V a and V6 gene repertoires .

Indianapolis, IN 46285 In order to detect the small amounts of lymphokines generated in vivo following antigen stimulation, we developed a co-culture system which allows for detection of IL-2/4, IL-3/CSF and TNF from LN cells stimulated in vivo with picryl chloride (PCL). Utilizing thb system in combination with FACS analysis and receptor binding studies, we examined the production of these lymphokines in primary and secondary immune responses. During a primary immune response, the production of IL-2 was not readily detectable on dl, peaked on d3 and was gone by d5. At no time were we able to demonstrate the presence of IL-4. Alternatively, the presence of IL-3/CSF and TNF was w i l y detected on dl, but olso peaked on d3. In comparison to primary responses, secondary immunization lead to at least two alteraticns. (I) Peak production of all lymphoki es shifted towards dl.

(2) Although most lymphokines did not demonstrate increasea in the amount produced/lO cells, the amount of lymphokine generated/LN was vastly increased due to an increased number of cells. Utilizing single and dual color FACS analysis we also examined the LN cells for alterations in T cell subpopulations. During the course of the primary response: (I) the percentage of Thy I+ and L3T4+ cells decreased until d3 and then began to recover, (2) the percentage of Thy-I+, T4-,T8-cells peaked at the time of greatest lymphokine production (i.e.-d3) and (3) the IL-2 receptor was expressed solely on Thy-I+ cells, was detected on both T4+ and T8+ subsets and peaked on d3.

Most of these alterations also occurred during the secondary response, but their timecoune was shifted so that maximal effects occurred earlier (e.g., dl). Finally. the maximal binding of radiolabeled IL-2 by the LN cells following both primary and secondary sensitization correlated with the expression of the IL-Zr as detected by FACS analysis. In addition, binding of radiolabeled IL-4 demonstrated similar patterns except for the detection of significant binding on dl. These results demonstrate that (I) an ordered timecourse of lymphokine production occurs in vivo following exposure to antigen and (2) the secondary immune response to PCL is characterrd by an accelerated tempo of lymphokine production, rather than an increased level of lymphokine production/lO cells. activation as direct G-protein activation by A1F4 PI-hydrolysis using phorbol diester stimulation of PKC restores the inhibi$$ble phenotype and the ability to upregulate c-fos. Even more interesting, sIg-linked Ca responses by VS2.12-c1.2 are equivalent to those observed in the wildtype WEHI-231. resul$g suggest that contrary to current thought, sIg-generated signals may not be coupled to Ca fluxes via inositol phospholipid hydrolysis. Thus, VS2.12-c1.2 is a new and powerful tool with which to analyze signalling through sIg at the molecular level.

Unlike the wildtype, crosslinking of sIgM on VS2.12-c1.2 did The signaling defect in VS2.12-cl.2-appears to be proximal to phospholipase C triggers PI-hydrolysis and bypassing

These latter lmmunogenicity C 302 ANALYSIS OF T CELL RECEPTOR 7 CHAINS FROM ADULT CD4-,CD8-THYMOCYTES Mark W. Moore, I. Nicholas Crispe and Michael J. Bevan, Department of Immunology, Research Institute of Scripps Clinic, La Jolla, CA 92037. The role of TCR 7 genes in T cell development has not been determined. To extend out understanding of the repertoire of TCR 7 expression, we prepared a cDNA library from CD4-,CD8-adult BALB/c thymocytes and cloned and sequenced 15 TCRq genes from this cDNA library. We found that 2 clones were transcripts of the unrearranged C7, gene and that 3 clones terminated in the J7, region. Nine of the remaining clones were V71,2-J7 C7 genes and five of these were in frame. Only one clone corresponded to C71 and was V7 -Jy C7jofned'in frame. SDS-PAGE analysis of the 7-chain proteins from the surface of both BALB/c anZCdBL)6 adult CD4-,CD8-thymocytes did not detect the 32,000 MW Vy C7 protein, but did detect the 35,000 MW V7C7 protein. These results suggest that despite the abundanc$%f Pull-length, functionally joined, V7 C7 transchpts in the thymocyte subset, the protein product is not expressed on the cell surface as the prehfcted 32,000 MW 7 protein. Finally, our analysis of the V-J jointing of the 7 genes reveals both flexibility at the V-J junction and extensive N-region nucleotide addition that lead to diversity of the predicted protein sequence. IL5 in response to the same stimuli. e identification of these two subsets of CD4' helper cells is mostly based on studies performed with long-term cultured T cell lines and it is not clear whether these two subsets exist in vivo and represent distinct lineages of T cells. In particular, the frequency, tissue distribution and ontogeny of cells capable of secreting IL4 in vivo is not known. These studies have been hampered by the fact that freshly isolated T cells from unprimed animals failed to secrete detectable amounts of ILA and IL5 when stimulated in vitro by lectins or alloantigens, whereas IU is readily detectable in these same cultures. Data presented here indicate that freshly isolated T cells from unprimed animals can be induced to produce IL4 in a receptor-de endent, antigen-independent manner upon stimulation by anti-CD3 antibodies. Our results also stow that only CD4' and not CDST cells can be induced to secrete IL4 and that cross-linking of the receptor is required for o timal activity. We believe that this approach will be useful in identifying in vivo cells recomittefto the Th2 pathway and study their ontogeny, activation requirements and tissue distriiution.

hLB, Brussels, BELGIUM. We have studied the murine TcR repertoire against the C-terminus of cytochrome c in association with certain alleles of the MHC ClaSS I1 molecule, EakEpk (IEk) and EakEpb (IEb). For mice possessing these alleles, the majority of responsive T cells utilize one member of the variable Val1 gene family in conjunction with a limited set of Vp genes. As an extension of these studies, we have examined IE specific, alloreactive hybridomas derived from IE non-expressing (Eab) cytochrome c non-responder mice to determine their usage of va and Vp genes. tion assay showed that fourteen utilized the same Val1 gene segment used by the majority of cytochrome c specific, IE restricted T cells and eight utilized a closely related Val1 gene that also is associated with this antigen response. element most commonly used by cytochrome c-specific T cells was not found among the alloreactive hybridomas tested, @ genes less frequently used in the cytochrome response were expressed by seven of the 22 alloreactive hybridomas whose Va segments were defined by RNase protection. determining recognition of IE molecules both in MHC-restricted, antigen specific immune responses and in alloreactive responses. The T-helper cells of seven mouse strains, representing 5 Class I1 haplotypes (IAs, IA4, IAb, IAkIEk, IAdIEd) were responsive to immunization and restimulation with parent peptide. The IEd determinant was shown to be a presenting element by monoclonal antibody blocking and by use of L-cell-transfectants as AF'Cs to purified T cells and to T cell hybridomas. A series of overlapping synthetic peptides identified two minimal T-cell sites within the parent peptide: Mice expressing IA and IE responded to a fragment at the N-terminus of the parent peptide (site 1) while mice expressing only IA responded to a distinct but overlapping fragment at the C-terminus (site 2 ) . These minimal sites identified in vitro could be used to immunize mice in vivo in an MHC-restricted manner. The human 6 TCR Locus is strategically located within the aTCR complex between the cluster of Va/v6 region and the Ja segments. which can be spliced to Ca in pre T cells, separates 6 from the Ja segments. pulse field gel mapping as we11 as molecular cloning link diversity (Ds), J g , c,5 and TEA within 35 Kb. considerable 6 TCR diversity is generated despite the predominant use of one v 6 and J 6 segment. D61 and D62 are 9 and 13 bp long, are frequently recombine as D,1/D62? and reveal exonucleolytic trimning with extensive "N" segment addition. Specialized 5' and 3 ' 6 deleting elements, 6 Rec and P J a , separate the 6 locus from the a locus. Cells with 6 Rec/$b Ja recombinations comprise most 6 deletion events although 6 Rec recombines with 2 other major acceptor sites in fetal and post-neonatal thymic DNA. The 5' 6 deleting element ( 6 Rec) is evolutionarily conserved in the mouse and functional comparisons are underway. delete the 6 locus may prove to be the pivotal event establishing separate y6 and aE lineages. To study the mechanism of T-cell tolerance, transgenic mice were generated that expressed the Mlsa reactive T-cell receptor (TCR) O-chain VB8 1 on -90 % of peripheral T-cells. In transgenic mice bearing Mlsd, the numbers of high TCR expressing thymocytes and of Thy 1.2+ peripheral T-cells were reduced. The CD4/CD8 ratio of peripheral T-cells was decreased fourfold compared to negative littermates. Both Mlsa and Mlsb TCR &transgenic mice were able to mount a T-cell dependent antibody response against viral antigens whereas the capacity to generate alloreactive and virusspecific cytotoxic T-cells was impaired in TCR &transgenic Mlsa, but not in transgenic Mlsb mice. RNA analysis and immunof luorescence with TCR VB-specific mAb further revealed, that the expression of endogenous TCR 0-genes in these mice was suppressed. tolerogen-reactive lymphocytes, as measured in the MLR, in spite of their long-term acceptance of a skin graft bearing the tolerated antigens.

Lymphokine production by MLR+ tolerant lymphocytes is different from that of syngeneic normal lymphocytes.

Normal lymphocytes produce only IL-2 in primary response to tolerogen, while tolerant lymphocytes produce IL-2 and IL-4. Using limiting dilution analysis, we have to estimated the frequencies of pIL-2 and pIL-4 (precursor) cells in these cultures. After primary k vitro stimulation, normal responders have a low but measurable frequency of pIL-4 cells, while tolerant responders have a much higher pIL-4 frequency. However, following subsequent & restimulations, the pIL-4 frequency of normal responders rises and begins to approach that of the tolerant responders, such that the two populations are indistinguishable based on pIL-4 frequencies following the third round of in vitro stimulation. These data suggest that the high frequency of IL-4 producers (presumably T,, cells) among the tolerant lymphocytes resembles unexpectedly a "primed" state, rather than "unprimed"as in nontolerant responders (where TH1, dominate the early response). The existence of "primed" T cells in phenotypically tolerant animals raises the possibility that precocious activation of TR1 (by neonatal exposure to tolerogen?) suppresses the later emergence of T,,, which would be expected to contain the cells responsible for graft rejection. A large number of CD4+ T-cell clones, obtained from peripheral blood T lymphocytes by direct limiting dilution, allowed us to address the question whether functional heterogeneity exists within the human CD4+ T-cell subset. Six out of 12 CD4+ clones were able to lyse Daudi or P815 cells in the presence of anti-CD3 antibodies. The remaining 6 CD4+ Tcell clones tested did not acquire this cytotoxic capacity during a culture period of 20 weeks. In the absence of anti-CD3 mAb, no lytic activity against Daudi. P815 and K562 target cells was observed under normal culture conditions. These two types of CD4+ T cells showed high reactivity with anti-CDw29 (4B4) MAb and no reactivity with anti-CD45R (284) mAb. The CD4+ clones without anti-CD3 mediated cytotoxic activities (TH2) consistently showed a higher expression level of CD28 antigens. TH1 CD4+ clones did produce IL-2, IFNgamma and TNF-alpha.beta. whereas the TH2 T-cell clones produced minimal amounts of IL-2. IFN-gamma and TNP-alpha. beta in response to anti-CD3 mAb and PMA. Not all CD4+ clones did release IL-4, but there was no correlation with cytotoxic activity. Moreover, as compared to the TH1 CD4+ clones, TR2 CD4+ clones proliferated moderately in response to anti-CD3 mAb. However, anti-CD3 mAb induced proliferation of only the TH2 CD4+ T-cell clones was enhanced by anti-CD28 mAb. Both CD4+ subsets provided help for polyclonal B-cell activation with anti-CD3 mAb. Our data suggest that the human CD4+ subset, in analogy to the murine system, comprises two functionally distinct T-cell subpopulations. In the mouse, when la antigens are isolated immunochemically, the predominant species isolated are the isotypic matched pairs, AaAp and EaEp. However, when la ap dimer expression is studied using an L cell transfection model, it is found that the isotype-mismatched dimer ApdEa is readily expressed at the cell surface. These results suggest that differences in assembly andl or transport of different la pairs may be most readily visualized in a competitive environment where multiple distinct la chains are available. To investigate this possibility, the relative efficiency of inter-and intra-isotypic dimer formation and expression was evaluated using a sequential L cell transfection system. L cells already expressing an ap dimer on the cell surface (ApdEa or ApdAad) were supertransfected with a third la gene (Aad or Ea, respectively). Synthesis of this second a or p protein led to competition for the unique partner chain. Individual clones were scored for cell surface expression of the distinct dimers (e.g., ApdEa vs EpdEa or ApdAad vs ApdEa) using FACS analysis with chain specific monoclonal antibodies. In addition, each species of mRNA was quantitated by Northern blot hybridization using bcus specific probes. Our results indicate that in the H-2d haplotype. isotype-matched dimers are expressed with 3-4X the efficiency of isotype-mismatched dimers. This result suggests that, regardless of the cell type studied, If each of the four murine la genes is expressed at equivalent levels, intraisotypic dimers will be expressed to the virtual exclusion of the interisotypic dimers. However, if chain synthesis asymmetry occurs, the isotype mismatched pairs may be expressed at immunologically relevant levels.

DIFFERENTIAL We have identified a series of discrete stages among the CD3-double negatives which seem to form a sequence, with TcR gene rearrangement and RNA expression gradually progressing, but with potential for expansion and repopulation of irradiated thymuses diminishing along the series. On this pathway CD3 must be expressed late or after the acquisition of CD4 and CD8. Cell cycle analysis shows the highest rates of cell division to be among the RSA+ IL-2R-Pgp-1-population which probably precedes the transition to CD4+CD8+ and TcR expression. Thus it seems unlikely that TcR/antigen interactions play a role in cellular events occurring among the double negative cells which lead on to mainstream T-cell development. Egr-l is a murine early growth factor inducible gene which encodes a protein with zinc fingers. Its expression was investigated in murine B-lymphocytes stimulated through their antigen receptor (sIg) with anti-recptor antibodies (anti-Ig) . Rapid (by 15 minutes) upregulatlon of Egr-l mRNA expression was observed at doses of anti-Ig sufficient to drive the majority of GO cells into cell cycle. Agonists and inhibitors of protein kinase C (PKC) showed that expression was coupled to the PKC component of receptor immunoglobulin transmembrane signalling. Interestingly, signalling through sIg on the murine B lymphoma WEHI-231 did not upregulate Egr-l expression even though similar signalling pathways are associated with this receptor in these cells. Southern analysis showed that Egr-l is not deleted or translocated in this cell line. Importantly, cell growth and proliferation of WEHI-231 is inhibited by anti-Ig stimulation suggesting a relationship for Egr-l expression and differential processing of receptor Ig signals. This notion is further supported by the finding that murine B lymphomas whose proliferation is not inhibited by anti-Ig showed receptor immunoglobulin coupled Egr-l expression.

Reeulation of EXDreSSiOn of a class 1 Wc transeene. Dinah s . the expression of the transgene product. The patterns of expression of the transgene parallels that observed in situ, indicating that regulatory elements necessary for normal patterns of expression are contained within the injected 9 kb DNA segment, and that trans acting factors involved in its regulation function between species. Included among these elements are those specifying preferential expression in B cells relative to T cells. In vivo treatment of transgenic mice with a/@-interferon results in increased expression of the transgene in a number of tissues. The response parallels that observed for the endogenous H-ZKb, but differs markedly from Qa-2. Analysis of the chromatin structure of the transgene reveals a single constitutive DNase I hypersensitive site present in both spleen and thymus, which is not altered by interferon. Both a novel negative and positive regulatory elements have been identified in the 5'flanking region of the transgene. The negative regulatory element reduced the activity of both the homologous class I promoter and a heterologous viral promoter. In vivo competition experiments indicated that the functions of the positive and negative elements are mediated by distinct cellular trans-acting factors. The negative regulatory element requires the presence of a positive regulatory element to function. This interaction between elements represents a novel mechanism for regulating gene expression.

McDevitt. Department of Microbiology and Immunology, Stanford University School of Medicine, Stanford, CA 94305. Published data show that encephalitogenic H-ZU murine T cell clones with specificity for the N-terminal eleven amino acid peptide of myelin basic protein display a restricted fine specificity when tested on substituted analogs of the native peptide. For example, substitution of alanine at certain positions in the peptide totally abolishes the response of each clone (Acha-Orbea et al.. 1988. Cell 54:263) . Recent experiments also have shown that the ability of some peptide analogs to bind to H-ZU I-A gene products does not always correlate with their ability to stimulate the T cell clones (see accompanying abstract by David C. Wraith and Hugh 0. McDevitt). This suggests that H-2U mice may lack a T cell repertoire capable of recognizing these peptides complexed to H-2U I-A gene products. To test this possibility, H-2U mice were immunized with a panel of peptide analogs, as well as the native peptide. The in vitro T cell proliferative response to each of the peptides then was measured. The results show that in vivo immunogenicity of the peptide analogs also does not strictly correlate with their capacity to stimulate the T cell clones. In this way, the polyclonal T cell repertotre of H-ZU mice for the myelin basic protein peptide analogs was examined, and could be compared with the I-A binding characteristics of the peptides. terms of antigen and MHC recognition. This response involves a limited repertoire of T cells which crossreact on species variants of the antigen. In addition, T cells specific for the antigen in association with syngeneic MHC can recognize antigen on similar allogeneic MHC molecules. The groupin of clones by functional phenotypes defined by these crossreactivities allowed us to corrcfate TCR gene usage with either antigen or MHC recognition. Some of the pigeon cytochrome c-specific clones within one functional phenotype use receptors that differ by as few as two amino acid residues. Other clones e y s s very different TCRs but exhibit similarities in antigen/MHC reco nition. The efect of these TCR differences on recognition was assessed using a pane? of anti en analogs with single amino acid substitutions presented on different MHC molecufes. Each clone exhibited a unique pattern of res onse to the antigen analog panel, even clones with very similar receptors. Also, eacE residue in the antigenic region of the peptide was critical for interaction with at least one T cell receptor. Therefore, the antigen must either be a linear molecule with each residue available to interact with the TCR or be able to assume several conformations to interact with MHC and the TCR. lmmunogenicity C323 THY-1+ CD3+ LY-5(B220)+ CD4-CD8-TCRx-6' HELPER CELLS. Anne I.

We have found that these cells can be preferentially stimulated to proliferate when cocultured with the B lymphoma, CH12. One to 2% of nylon wool non-adherent, Ia-, Jlld-, and CD8-lymph node cells from normal unimmunized mice have the phenotype Thy-l', CD3', CD4-. and CD8-. These cells proliferate when co-cultured with a syngeneic surface Ig' lymphoma, CH12, even in the absence of any added antigen, mitogen, or fetal calf serum. Prior to stimulation we find that approximately 30% of Thy 1.2' CD3' CD4-CD8-express the marker Ly-5(B220), however after culture with CH12 the majority of cells with this phenotype express the marker Ly-5(B220). After CH12 dependent proliferation the Ly-5(B220)' T cells are able to provide help for secretion of Ig by fresh CH12 B cells. Surface labelling and precipitation of T cell receptor molecules reveals that most of the Thy-1' CD3' Ly-5(B220)* CD4-CD8-cells express TCR(r-6). Furthermore, CD3 precipitation shows that as many as four different 7-6 heterodimers are utilized within the entire responding population. This suggests that a heterogeneous population of double negative TCRi-6 cells are involved in the response to CH12.

College of Kedicine at East Tennessee State University, Johnson City, TN 37614 Interferon-producing (T 1) and Interleukin 4 (IL4) producing (T 2) clones were assayed for their ability to diregtly induce cytostatic activity in macro:hages generated from splenic myeloid precursors (M -c). In the presence, but not in the absence, of antigen, T 1 clones activated the M -c to inhibit the growth of P815 tumor cells in vitro. TH2 cjlones were not able to activate such effector activity in the I4 -c. effectively present antigen to the T 2 clones as evidenced by the proliferation of T 2 cells cultured with antigen in the pFesence, but not in the absence, of M -c. Thereyore, although both T 1 and T 2 were activated by cognate interaction with antigen presenting (BA) or Nippostrongylus brasiliensis (Nb). Spleen cells from these mice were cloned at limiting dilution with alloantigen stimulation, and every two weeks, LK production in response to Con A was measured. Clones derived from, and stimulated with, cells from unimmunized mice initially tended to secrete low LK levels, with few clearly defined TH1 or TH2 clones. By 56 days after cloning, some clones had acquired TH1 or TH2 patterns. CFA, BA and Nb-imnunized mice gave rise to clones that were mostly TH1 or TH2 even at early times. CFA and BA immunizations induced almost exclusively TH1 clones, whereas Nb induced more TH2 clones. These results are consistent with a model in which resting, previously unstimulated T cells produce low amounts of LKs, and progress through stage(s) where they secrete both TH1 and TH2 LKs before finally differentiating into TH1 and TH2 cells. The results with CFA, BA and Nb-primed mice suggest that this process occurs in vivo as well as in vitro.

STRAINS AS CARRIERS OF MELIOIDOSIS ANTIGENS TO THE IMMUNE SYSTEM, Deja Tanphaichitra, Mahidol University, P.O. Box 4-217, Bangkok 10400, Thailand The attenuated galE mutant, Salmonella typhi strain, TyZla, served as the recipient in a conjugal DNA transfer experiment. Conjugal DNA transfer was obtained by the mating procedure on an appropriate blood agar medium. were examined serologically. One selected strain was found to have the serological characteristics of the recipient S . typhi, Tyfla strain and also expressed the Pseudomonas

The donor strain was a Pseudomonas pseudomallei MU107.

The resulting antigen clones were repurified by restreaking on the medium and pseudomallei antigen. the S. typhi transconjugant strain is due to the presence of the Pseudomonas pseudomallei plasmid. A group of subjects when received four doses of this bivalent vaccine strain

In this study it appears that Pseudomonas pseudomallei synthesis in developed antibodies against Pseudomonas pseudomallei up to 70%. pseudomallei, an intracellular pathogen, produces a characteristic antigen probably to be plasmid coded, we considered that the GalE Salmonella typhi TyZla oral vaccine strain, highly effective against typhoid fever, might be modified so as to be protective also against melioidosis due to Pseudomonas pseudomallei. Terminal deoxynucleotidyl transferase (TdT) is a lymphoid-specific nuclear enzyme present in early lymphocytes. To investigate the regulation of TdT gene expression, pre-B and pre-T cells were treated with phorbol 12-myristate 13-acetate (PMA) o r three analogs, and TdT steady-state mRNA levels were determined by Northern blot analysis. Treatment of early lymphocytes with PMA results in a rapid and reversible decline in steady-state TdT mRNA levels within six hours. This rapid decline can be blocked by pretreatment of the cells with a protein kinase C inhibitor, implicating protein kinase C activation in the decline of TdT mRNA. Nuclear run-off studies demonstrate that TdT transcription is rapidly down-regulated within 45 minutes after PMA treatment, indicating that this regulation occurs mainly at the level of transcription. Furthermore, cycloheximide blocks the decline in TdT in RNA showing that new protein synthesis is required for transcriptional inactivation. The nucleoprotein gene from the influenza virus A/NT/60/68 was stably cloned into the attenuated aroA-strain of Salmonella typhimurium SL3261. Nucleoprotein purified from pNP -3261 was tested for the ability to generate virus-specific immunity. Immunization with recombinant derived nucleoprotein induged immunity to all type A influenza tested but not against type B viruses. CD4 helper T cells were primed but no evidence was found for priming of class I restricted CTC. Mice immunized with recombinant nucleoprotein were protected against a subsequent challenge of influenza virus. The information obtained from the study of the immunity and protection generated by the purified recombinant protein was then used to design experiments to investigate the possibility of using the attenuated Salmonella vector to deliver the nucleoprotein molecule to the immune system by the parenteral or enteral routes. We characterized the extrachrom-1 circular I N i s in 19-day-fetal and 4-week-old m u r i n e thpmcytes and 8-week-old m u r i n e splenocytes. f popllation of circular ChIAS was clone3 into the kgtll phase vector. We screened ca. 10 tNA cl-by plaque hybridizations with all far kirds of TCR gene probes derived from Jal , Val 0, DB1, DB2, Jyl , J61 and 562 loci.

Cut of 10,000 CNA cl-from fetal and 4-week-ld thymocytes, 30 hybridized with TCR aprobes and 5 hybridized with TCR &probes. Positive cl-with TCR yand 6probes were 3 to 7 in fetal thymocyte4erived library, but few in 4-week-old thymocyte. Of 7 fetal TCR 6 clcnes analyzed, 6 cl-had DD or VD reciprocal joints and 1 clcne had VD ar DD d i n g joint. Relative frequencies of circular DNA clones for four different TCR genes are consistent with the order of the expression of the genes the T cell develOpnent. Of 10,000 tNA cl- signalling could be studied. llzmambxane signalling was maasured by 9 ability to t?z3nSlOCate FKC frcrm the cytcplasa to the nW2leus after surface I-A was banrl by a or p dxdn specific monoclcnal antibody. I(pmwing either 6 or 12 amino rids fmn the a chain cvtoplasnic (Cy) damin did not affect the ability of tkse I-A r m l d e s to trarslocate PKC to the nucleus. Normal splenic B c e l l s were rendered non-responsive t o subsequent challenge w i t h LPS, as measured by a decreased a b i l i t y t o generate antibody forming c e l l s (AFC), by incubation overnight (18-24 hours) w i t h 10 ug/ml a n t i -I g .

Both i n t a c t and F(ab)', a n t i -I g , as well as monoclonal anti-IgM (BET2 and B-7-6) , were able t o induce 6 c e l l non-responsiveness t o subsequent LPS challenge, suggesting t h a t sIg/FcR i n t e r a c t i o n s are not necessary i n the induction o f LPS non-responsiveness.

I n contrast, induction o f nonresponsiveness t o subsequent challenge w i t h FITC-prucella abortus required i n t a c t a n t i -I g .

The a b i l i t y o f mitogenic a n t i -I g (Rab F(ab)', o r 6-7-6 Northern blot analysis and bioassay data were used to analyze 9 separate lymphokines as well as the IL-2 receptor (murine Tac). Northern blot comparison of fresh and primedT4 enriched RNA revealed that primed T cells produced 10-fold more lymphokine than the fresh T cells. The only lymphokine that showed equal amounts of mRNA for both fresh and primed T cells was IL-2. A time course of fresh and primed T4+ cell lymphokine production was also analyzed. The primed cells produced a short burst of lymphokine mRNA that peaked between 7.5 and 13 hr after Con A stimulation and declined after 18 hr. The fresh T cells produced a longer burst of lymphokine mRNA that peaked 18-44 hr after stimulation. The IL-2 receptor @-2R) mRNA time course from activated primed cells showed different kinetics than lymphokine mRNA. This suggested that molecular regulation of the IL-2R might be different than lymphokine regulation. To further examine molecular regulation in the primed T cells polysome profiles were evaluated for lymphokines, L 2 R , and other cellular genes. The recently developed method of gene amplification by the polymerase chain reaction (PCR) has proven to be particularly suited for the analysis of T cell receptor (TCR) genes. We adopted existing methods for the preparation of cytoplasmic RNA from as little as 1000 cells and used this material as template for first strand c-DNA synthesis. PCR amplification of this c-DNA, using V-and C-specific oligonucleotide primers yielded enough material to produce single-stranded DNA in a second PCR which could then be sequenced without cloning. In case of unknown V-usage, the PCR was employed for screening for V-beta elements by sequential reactions with different V-beta specific primers. We have used this method to reinvestigate the H-2b restricted cytotoxic T cell response to TNP in C57B116 mice. Beta chain sequences of 26 CTL clones obtained by direct cloning of immune spleen cells were compared to sequences of 11 clones obtained by cloning of individual short-term in vitro CTL lines. It was found that a) in vitro bulk-stimulations reduced the heterogeneity of the beta-chain responses to TNP, b) similarities between different TCR-beta-chains concentrated on the usage of certain Jb-elements ( Jb2.6, 2.5, 2.1 ) rather than V-region or NID-region sequences, and c) the majority of Jb2.6 containing beta-chains was associated with alpha-chains expressing V-segments of the Val 0 family. These Expression of genes which encode the T cell antigen receptor is cenval to the generation of the T cell repemire. Our labomtory has been investigating genes for both the alpha and beta chains of this receptor in inbred strains of Runus norvqicus (the laboratory rat), a species in which several autoimmune disease models have been developed. and which is used extensively in transplantation studies. Using genomic Southern blots and mouse probes specific for five different V a subfamilies, we have estimated the size of the V a repertoire in ten inbred strains of rat. Results show a significant increase in the size of one subfamily and suggest increases in two others in all ten strains. The rat V a l subfamily has about twice as many members as the mouse, while the Va2 and Vu5 subfamilies, depending upon the enzyme used, show a similiar duplication. The Va6 and Va9 subfamilies have a comparable number of members in both species. These data are most easily explained by a single duplication event in the rat invoking at least one and perhaps three subfamilies, but not encompassing the entire V a locus. This implies that the Val subfamily (perhaps together with Va2 and Va5 subfamilies) is regionally clustered and not interspersed with either the Va6 or Va9 subfamily. Based on restriction fragment length polymorphisms, we find evidence for six distinct V a haplotypes in the ren strains tested.

We have also cloned eight unique germline V a l gene segments. One of these has been sequenced. and has a coding region 87% identical to the most closely related mouse V a l sequence. This degree of relatedness is similiar to ra#nouse Vg homologues. which share 8548% nucleotide sequence similarity. We are using these clones to generate angle copy probes from flankiig regions to further map the V a l locus. Current approaches to MHC-peptide binding studies require either large quantities of highly purified MHC protein and/or the use of sophisticated detection apparatus. I n order to simplify detection of peptide-MHC interactions we have investigated the use of photosensitive-crosslinkers. Two reagents have been successfully tested. A benzophenone derivative of peptide 1-16 from rat myelin basic protein (RMBP) was only effective after the introduction of a glycine spacer residue between peptide and crosslinker. An azido-nitro-benzoyl derivative of peptide 7.4, a heteroclitic analog of RMBP 1-11 (1). had a high affinity and bound specifically to the peptide binding site. The 7.4 photoaffinity probe has been used to test the binding properties of other analogues of RMBP 1-11 and is Currently being used to define (a) the kinetics, (b) pH and (c) temperature dependence of the binding event. This particular photoaffinity conjugate retains both the MHC binding and biological properties of the original peptide and is helping us to define the roles of "determinant" versus "T cell repertoire" selection in the MHC linked autoimmune response to MBP The antigen-specific T cell repertoire is diverse in its ability to recognize a wide universe of foreign antigens. This T cell repertoire is composed of a set of clones each of which is specific for a given foreign antigen. Therefore the precursor frequency of T cells specific for any give foreign antigen is extremely low. However, two prominent exceptions to this general rule exist, and these are the T cells present at high precursor frequency which are specific for foreign HHC products or for the products of the minor lymphocyte stimulatory (Mls) genes in the mouse. The present studies were undertaken in order to examine factors involved in T cell repertoire formation by assessing the relationship between T cell repertoire for conventional foreign antigens and for Mls products. studies indicate a striking degree of overlap between the set of T cells specific for pigeon cytochrome c and the set of T cells specific for MlsC gene products. demonstrate that the basis for this overlap lies in the predominant expression of one TCR Vp gene, VBS, by those T cells which recognize MlsC.

involvement of specific TCR afl dimers in recognition of MlsC and further suggest that T cell reactivity to these gene products may play an important role in establishing the T cell repertoire for foreign antigens. conclude that, rather than destruction of some essential APC structure, ECDI fixation prevents the APC from actively responding during the encounter with the T cell. This results in a failure to express new structures (probably located on the APC plasma membrane) that appear to be essential for stimulating T cell proliferation. These structures are distinct from Ia or IL1. The induction of these structures during T-APC interaction occurs in six hours, requires protein synthesis, and can be elicited by IL1, IL4 or LPS, but not IFN-gamma. In the absence of these induced structures, the APC stimulates a partial T cell response, IL4

release, but the T cells fail to proliferate. These induced structures on the APC may be either adhesion molecules that stabilize the T-APC interaction, or they may provide additional stimuli to the T cell. were not c m n t o the three s t r a i n s o f mice (BALB/c, regions 1, 3, 5, 6 and 7; C3H/He, regions 2, 3, 5, 6, 6', 7 and 7'; and C57BL/6, regions 2, 4, 5, 6 and 7). Immunisation with type I1 collagen (CII) leads to development of arthritis in mice with certain MHC haplotypes and is associated with an immune response against CII. We have been studying the T-cell response in the arthritis susceptible strain DBM1 (H-2q) . Analysing the proliferative response in cultures of lymph node cells from immunised mice a s well as T-cell lines and clones established from such cultures it was found t h a t lI The T-cell response after immunisation with heterologous CII was preferentially directed against foreign determinants on the CII molecule with little o r no crossreactivity against autologous CII. 2 ' Both the primary response and the reactivity of established lines and clones were directed against the CBll fragment of the CII molecule, using C B l l fragments prepared from chick, bovine or rat CII. 3/ Pepsin present in CII preparations after using pepsin digestion for solubilisation of the collagen is strongly immunogenic even in very small amounts and it was therefore necessary to use CII prepared from lathyritic cartilage without pepsin digestion for immunisation. In contrast to the pattern in lymph node cultures from immunised mice we found that when culturing spleen cells from unimmunised mice there was a T-cell response against collagen that was preferentially directed against autologus CII. Since we earlier have found that autologus CII may induce an immune response and also arthritis in DBN1 mice we conclude that there exist T-cells capable of reacting with autologus collagen and inducing an immune response as well as arthritis but that these cells are under regulation so that they not readily can be activated into proliferation but may be induced to perform certain effector functions. tested. The characterization of these two CD6 rDSAs will be presented.

ANALYSIS OF HLA POLYMORPHISM USING SEQUENCE SPECIFIC OLIGONUCLEOTIDE PROBE HYBRIDIZATION TO AMPLIFIED DNA, Lee Ann Baxter-Lowe, Jay B . Hunter, and Jack Gorski, The Blood Center of Southeastern Wisconsin, Milwaukee, Wisconsin 5 3 2 3 3 . HLA polymorphism plays a key role in antigen:MHC interaction. The polymorphism of the first domain encoding exon of the HLA-DR p chain has been studied by in vitro DNA amplification and use of sequence specific oligonucleotide probe hybridization (SSOPH) to detect polymorphic sequences. A 230 bp segment of genomic DNA was amplified and hybridized with synthetic oligonucleotide probes (12-19 bases) under conditions that detect single base pair mismatches. Identification of these mismatches can be used to predict micropolymorphism in the protein products, including single amino acid changes. Haplotype specific patterns of oligonucleotide probe hybridization were defined for a panel of homozygous typing cells. Analysis of family data demonstrated the expected inheritance patterns. Most known serological specificities are encoded by multiple allelic forms of DR p chains and SSOPH can identify these differences. This was exemplified by detection of unique SSOPH profiles for subtypes of DR4, DRw6 and DRw52 alleles. This procedure was also used for analysis of HLA-DR polymorphism in large numbers of heterozygous individuals, including an HLA-deficient SCID patient. The SSOPH data were correlated with serological specificities and will be useful for delineation of HLA restriction in alloand autoimmunity. Different cell membrane receptors have been shown to be involved in human T lymphocyte activation induced by either monoclonal antibodies or mitogenic lectins. These T cell surface molecules can be divided into two categories : a) the T cell antigen receptor (TcR) associated with the non-polymorphic CD3 antigen b) T cell differentiation molecules not linked to CD3/Ti such as CD2 (T11) and Tp44 (9.3). Monoclonal antibodies directed against these T cell surface structures triggered different T lymphocytes functions : mitogenesis, IL-2 receptor expression, IL-2 secretion. Our knowledge about early events involved in T cell membrane activation is not complete, especially involving the transduction mechanism mediated by GTP-binding proteins ; nevertheless, numerous authors have demonstrated that CD3/Ti complex triggering induces the activation of phospholipase C, leading to the phosphoinositide cascade associated with an increase of free cytoplasmic calcium ions.

In the present report, we show that different activating cell molecules (Con A , PHA and PMA) can trigger oxygen free radical liberation when incubated with the human Jurkat tumor T cell line. Since membrane oxidative metabolism has been shown to be related to the stimulation of the phospholipase A2, and to be the final consequence of a membrane NADPHoxidase : this could represent a previously undescribed pathway of T lymphocyte activation. High affinity monoclonal antibodies (MAb), specific for Staphylococcal nuclease (NASE), were produced and characterized. Competitive inhibition assays were conducted resulting in a series of complementation groups that define eight overlapping epitopes. It is estimated that these epitopes account for 70% or more of the accessible surface of NASE. Mutagenesis of the coding sequences for NASE was carried out to produce a series of variant molecules (each differing from wild-type. NASE and from each other by a single amino acid) that will enable mapping of NASE epitopes, determination of residues involved in antibody binding, and the contribution of various physical and chemical factors to affinity and fine specificity. Screening some of these mutants with the panel of MAb enabled us to map several nonoverlapping epitopes and further subdivided some of the MAb complementation groups. Oligonucleotide-directed mismatch mutagenesis has been done on codons encoding the original amino acid residue and other surface residues in its immediate vicinity. Determination of enzyme activity and structural analysis by CD spectropolarimetry of several of the mutant proteins suggests that any structural changes that may occur are local and not global. Supported by grants AI20745, l32CA09109 and S07RR05431 from the National Institutes of Health. Activation of T cell proliferation is believed to occur via the hydrolysis of inositol phos holipids, which, through the second messengers inositol-1,4,5-tris hosphate and diacylglycerof(DAG), promotes the elevation of intracellular calcium levels anjactivation of protein kinase C (PKC), respectively. The role of PKC in T cell activation was investigated by comparing the effects of stimulation by 12-0-tetradecanoyl phorbol acetate (TPA), and the DAG, oleoylacetyl glycerol (OAG), on a >99% pure population of T cells cultured in RPMI 1640 medium containing 10% autologous serum. Treatment with either TPA or OAG caused down-regulation of the T cell rece tor, a consequence of its hosphorylation, but only TPA, in syner leuiin 2 receptor (IL2-R), expression and, sgsequently, proliferation. Immunohistochemical staining with antisera specific for the PKC subspecies a, PI, PII and 7 shows that restin T cells express a, PI and pII PKC subspecies which are diffusely distributed throughout the Celt. After 20 minutes treatment with either OAG or TPA all three subspecies are redistributed to a focal area within the cell. The redistribution is transient in OAG stimulated cells, where the PKC distribution is similar to that in untreated cells after 1 hour of treatment. In TPA stimulated cells, however, the PKC redistribution is prolonged, becoming more marked until mitosis occurs after 48-72 hours of treatment. These results suggest that transient intracellular redistribution ofPKC causes phosphorylation and down-regulation of the T cell receptor, but that prolonged redistribution is required or T cell proliferation. Sm is a nucleoprotein complex associated with small RNA molecules in eukaryotic cells. The spontaneous generation of anti-Sm antibodies is specific for patients with systemic lupus erythematosus (SLE) and develops in 2 5 % of MRL mice. The response has been shown to be T-cell dependent in MRL/lpr mice. T-cells specific for Sm are found only in MRL (H-2k) mice and mice bearing H-2s and H-2f haplotypes (which do not develop anti-Sm antibodies). We are currently working to define the variable regions of the T-cell receptor genes used in the Sm response. A series of T cell hybridomas from MRL mice has been generated and are being screened for Sm positivity. A technique has been designed to amplify specific alpha and beta chain TcR genes using the polymerase chain reaction allowing for a more rapid sequence analysis. It is also our intention to locate the Sm specific epitopes of the T-cell hybridomas.

D. Bloom, P.L. Cohen, and S.H. Clarke, Department

Medical Institute and Experimental Immunology Branch, NCI. NIH, Bethesda. MD 20892.

To determine whether prior activation history affects T cell receptor mediated activation of T cell clones, the murine type I helper clone AE7 was maintained in tissue culture by stimulation every ten days with either (1) antigen (cytochrome c), irradiated H-Zk spleen cells. and IL-2 or (2) IL-2 alone. AE7 cells grown with antigen and antigen presenting cells (AE7-AG) proliferated and produced T cell growth factor activity (TCGF) in its culture supernatants following stimulation with immobilized anti-T3 antibody. The TCGF activity was shown by bioassay using indicator cell lines and specific blocking antibodies to be almost entirely due to GM-CSF with little or no IL-2 activity detectable. detectable IL-2 mRNA levels.

(AE7-ILZ) displayed substantially greater anti-T3 induced proliferation than did AE7-AG cells.

In contrast to AE7-AG cells, AE7-IL2 cells produced large quantities of IL-2 in response to anti-T3 stimulation. Furthermore. one cycle of stimulation of clone AE7-AG with IL-2 in the absence of antigen and irradiated spleen cells was sufficient to cause this clone to produce substantial amounts of IL-2 upon subsequent anti-T3 stimulation. These data suggest that T cell receptor mediated stimulation of T cell clones by specific antigen and antigen presenting cells inhibits subsequent anti-T3 induced IL-2 production. T cell proliferative responses and sera antibody levels of myasthenic patients to several synthetic peptides representing different epitopes of the human AChR were examined. We detected significant differences in the humoral and cellular responses of MG patients compared to healthy controls to peptides of the human AChR alpha-subunit with sequences p195-212, p257-269 and ~310-327. Proliferative responses of lymphocytes from myasthenic patients to p195-212 and to p257-260 correlated significantly with HLA-DR5 and with HLA-DR3, respectively. In order t o investigate further the immune responsiveness to selected sequences of the human AChR, T cell lines and clones specific for peptides p195-212 and p259-271 were established from lymph nodes of C3H.SW mice. The recognition specificities of these lines were tested by examining crossreactivity to a series of shortened and/or extended peptides of the above sequences. Deletions of amino acids in positions 211 and 212 (211=P, 212=L) resulted in a decrease of the peptides' stimulatory activity on the ~195-212 specific T cell line, whereas deletion up to position 200 on the N-terminal end had no effect on the triggering potential of the peptides. Similar results were obtained when deleting residues 270 and 271 (270=V, 271=P) in stimulation assays of the p250-271 specific T cell line. help in determining important T cell epitopes on the human AChR. The role of guanine nucleotide binding regulatory proteins (G proteins) in the regulation of phosphorylation of the y subunit of the CD3 antigen has been examined. CD3 y chain phosphorylation in isolated T cell microsomes or permeablised T cells was stimulated by the G protein activator, guanosine 5'-0 thiotriphosphate (GTPyS), but other nucleotides such as CAMP or GDPBS were ineffective. dependent. These data are consistent with the involvement of a G protein in the signalling mechanisms that regulate the phosphorylation of the CD3 y chain. The regulatory effects of calcium and GTPyS were compared in normal peripheral blood derived T cells and Jurkat cells. There were differences regarding G protein regulation of CD3 y chain phosphorylation in normal T cell and Jurkat cells and current models explaining these differences will be described. Expression of the gamma-delta T cell receptor has been thought to first occur in a population of thymocytes shortly after their precursors populate the thymus between 11 and 13 days of gestation. In the course of our studies investigating the ontogeny of T cell receptor expression in the mouse embryo we have identified an extrathymic site of gammadelta expression in a population of cells present at distinct times of gestation. Evidence will be presented demonstrating two periods of activity of the murine gamma locus in the developing embryo. are colonizing the thymus from the liver and the gene segment useage detected is different to that first expressed in cells of the developing thymus. around the time of birth) involves the functional rearrangement and expression of a gamma gene segment corresponding to the initial functional rearrangements detected earlier in gestation in the thymus, which can occur independently of thymic influence. demonstrate a new site of gamma-delta receptor expression in the liver of newborn mice that can occur in the absence of any thymic influence. The primary (in vivo) response of CS7BL/6 animals to the class I antigen Qa-1 is a helper (TH) dependent event as indicated by the requirement for copriming with a distinct antigen capable of activating helper cells. In contrast, the secondary (in vitro) response to Qa-1 demonstrates no need for costimulation with the helper antigen. In attempts to more closely examine the helper requirements for activation of primed CTLp, we have observed that depletion of L3T4 cells from spleens of Qa-1 primed mice abrogates the in vitro generation of anti-Qa-1 effectors. The response is restored by the addition of concanavalin A induced supernatant (CAS) or by the addition of syngeneic but not Qa-1 allogeneic L3T4 cells. Indeed, even in the presence of CAS, L3T4 cells expressing the Qa-1 alloantigen specifically suppress the activation of anti-Qa-1 CTL in a manner reminiscent of that seen with Lyt-2 veto cells. Although the mechanism whereby L3T4 cells exert suppression is unclear, we have determined that CTLp are susceptible to veto only within approximately the first 48 hours of culture, after which they resist suppression. Results from further studies of the nature of suppression and the L3T4 veto cell will be presented.

Group I1 proteins induce IgE ab responses in -90% of mite allergic patients. murine mAb and human IgG and IgE ab. unrelated. crossreactive epitopes on GpI and GpII allergens from different mite species. In contrast, IgG ab in BALB/c mice immunized with lOug specific" epitopes and <1% was a n t i -u I (a GpI homologue, with -80% amino acid sequence homology to I). Four non-over lapping epitopes were defined by mAb, with one species specific immunodominant site on each GpI allergen. cross-reactive GpI epitope and this mAb could inhibit human IgE ab binding by -40%. specificity of the murine anti-GpI response was not H-2 restricted, but could be altered by immunizing BALB/c mice with lower ag doses (lug) in alum or 8 . Dertussis. Using these regimes, up to 52% of the murine IgG ab responses was GpI cross-reactive. responses to GpII allergens appear to be strain dependant. unresponsive to GpII. however, BALB/b, A/J, CBA, C3H 6 C57B16 all produce GpII cross reactive IgG ab. anitgenic sites on GpI allergens are conformational, whereas those on GpII may be sequential. known to affect IgE expression in mice may also affect the epitope specificity of IgG ab.

We have compared the B cell epitopes on these allergens using panels of However, ag binding RIA on 73 sera showed that human IgG and IgE ab recognize The GpI and GpII allergens are antigenically I in CFA was directed against "species

Murine ab BALB/c are completely Thermal denaturation and reduction and alkylation expts suggest that the Results with the GpI allergens suggest that immunization regimes which are c425

This report demonstrates for the The exclusive recovery of 72-84-specific T cell clones c 426

Further molecular analysis should identify and characterize AChR reactive autoimmune clones and/or suppressor cells.

COHPLEX, Mogens H. Claesson, P e t e r BKamS a n d S t e e n D i s s i n g , L a b . E x p .

H e m a t . Immunol., D e p t . Med. Anatomy A, a n d D e p t . G e n e r a l P h y s The Ly-6 alloantigens represent a family of phosphatidylinositol anchored proteins that function as accessory molecules in the process of T lymphocyte activation. The expression of these alloantigens is often induced on T and B lymphocytes after activation by mitogens or antigens. Previous studies have shown that the induction of Ly-6 alloantigens in T cells is at least in part due to the action of IFN-a/B or IFN-7. In the present study, we have demonstrated that IFN-7 also induced Ly-6 molecules on B lymphocytes and bone marrow cells. Furthermore, we now show that TNF also participates in the induction of at least one of the Ly-6 proteins, Ly-6A/E. TNF was found to synergize with IFN-7 to induce Ly-6A/E expression in thymocytes, T lymphocytes, and bone marrow cells, but not B cells. For T lymphocytes, the synergistic induction of Ly-6A/E by TNF was restricted to cells from the Ly-6.1 haplotype whereas IFN-7 was sufficient to fully induce Ly-6A/E expression in cells from the Ly-6.2 haplotype. This result is consistent with the notion that there is more complex regulation of the Ly-6AfE molecules in T cells obtained from the Ly-6.1 haplotype. For T cells from BALB/c (Ly-6.1) mice, Ly-6A/E, but not Ly-CC, molecules were induced by IFN-7 and TNF. Furthermore, when compared to Ly-6A/E, the regulation of MHC class 1 molecules in these T cells by TNF was minimal. The induction of Ly-6AfE molecules on BALBfc T cells resulted in an enhanced capacity to activate these cells through the Ly-6 T cell activation pathway. One transformed T cell line, 5.1.2. was also identified whose Ly-6A /E molecules were synergistically induced by IFN-7 and TNF. Optimal expression of Ly-6A/E molecules on 5.1.2 cells required continuous culture of this cell line with these two cytokines and resulted in the detection of optimal levels of cytoplasmic Ly-6AfE mRNA by Northern blot analysis. This latter result suggests that IFN-7 and TNF regulate Ly-6A /E at the level of transcription and/or mRNA stabilization. 

UT Southwestern Medical Center, Dallas, Texas 75235. An IgM antLCD3 mAb (38.1) was found to modulate cell surface CD3 on highly purified human T cells within 5 hours in the absence of a secondary antibody or accessory cells. Inhibition could be overcome with accessory cells or IL2. The inhibitory effects of 38.1 could be mimicked by briefly pulsing cells with the calcium ionophore, ionomycin.

38.1 or ionomycin pulsed cells were inhibited in their subsequent capacity to resp nd to PHA even when exposures were carried out in the presence of EGTA to prevent increases in [Cap*]. from extracellular sources.

Inhibition was not the result of an inability to respond to PHA 'by increasing [Ca2+] .. Moreover the newly expressed CD3 molecules were capable of generating increases in [Ca2*].' after reacting the cells with anti-CD3 + a cross-linking secondary antibody. These studies dem'onstrate that a state of nonresponsiveness in resting T cells can be induced by modylating CD3 with an anti-CD3 mAb in the absence of co-stimulatory signals. A brief increase in [Ca ' 1. resulting from the mobilization of intracellular calcium stores appears to be sufficient to induce'this state of T cell nonresponsiveness.

CD3. Laurie S. Davis, Mary C. Wacholtz, and Peter E. Lipsky, Dept. of Internal Medicine, lmmunogenicity c a Central Lab.Blood Transf.Service, Lab. of Exp. and Clin.Immunology of the Univ. of Amsterdam, Amsterdam. The Netherlands Monoclonal antibodies (mAb) directed against the human CD3 molecular complex induce a strong proliferation of T cells, when immobilized on microtiter wells. This activation system, that was shown to be independent of accessory cells, accessory-cell derived factors or LFA-1 mediated intercellular adhesion (1). allows one to study the requirements for T-cell proliferation and differentiation in a well defined manner. IL-2 and IFN-gamma but no IL-4 could be detected i n culture supernatants of coated anti-CD3 stimulated T cells. The addition of rIL-1 or rIL-2 had only a moderate effect on T-cell proliferation, vhereas helper activity for Ig production was strongly enhanced in the presence of these factors. In this system differentiation of precursors to cytotoxic T lymphocytes (CTL), as measured in anti-CD3mediated cytotoxicity, could be demonstrated within 2 days after initiation of the activation. Allospecificity of the induced effector CTL was demonstrated using a panel of HLA class-I P815-transfectants. In this system the regulatory role of the CD28 molecule in Tcell activation and differentiation was studied. Addition of anti-CDZ8 mAb to T cells stimulated with coated anti-CD3 mAb enhanced IL-2 production, proliferation as well as Ig production. Interestingly, pCTL differentiation was also enhanced by anti-CD28 mAb. This system seems valuable for the analysis of requirements for differentiation of human T cells subsets. 1. Van Noesel et al., Nature 333, 850-852, 1988 ANALYSIS OF THE REQUIREMENTS FOR HUMAN T-CELL DIFFERENTIATION, Rolien de Jong. Vivienne Rebel, G i j s van Seventer. Miranda Brouwer, Frank Miedema, Rene' van Lier,

The newly described T cell receptor (TCR) 6 locus is located inside the TCR a locus between Va and Ja . Despite this unique situation, a highly efficient regulatory mechanism results in the complete independence of these two loci. We have recently described, in humans, a site specific recombination which joins a 5' deleting element (SRec) to the Send of the Ja's (YJa) resulting in the deletion of the TCR-6 locus in T lymphocytes expressing the a/B TCR.

Rearrangements of the TCR as well as immunoglobulin genes are mediated by a unique recombination machinery and therefore, the specificity of these rearrangements is thought to be the result of a differential accessibility of the DNA involved in the recombination process. As a consequence (and/or cause) of the opening of a segment of DNA, the region involved is fxst transcribed as a sterile transcript prior to the rearrangement. In that regard, we have found that the 2 Kb of DNA u p s a u~n of YJa are actively transcribed ('T early a" transcript, TEA) early during fetal development. The presence of the TEA transcript presumably reflects the opening of the TEA sequence prior to the TCR-6 deletional rearrangement. In order to better understand the mechanisms involved in the DNA accessibility model, we started to look for DNA-binding proteins which might play a putative role in the opening or blocking of the TEA sequence. By the technique of "gel shift assay" we found such a negative regulatory protein in the nuclear extract from a non-lymphoid cell. The binding activity appeared to be specific as it was competed out by an excess of unlabeled autologous DNA and not by an excess of irrelevant DNA. Further studies are now in progress to determine first wether the presence of this binding activity can be correlated with a "closed" configuration of the TEA region and second to determine the precise location of the DNA binding region.

Cohen, Laboratory of Chemical Biology, NIDDK, National Institutes of Health, Bethesda, Md 20892. Mabs retard lymphoproliferation as well as autoimmunity. Interesting, so does the adminisuation of a Mab to L3T4 , thus suggesting that the T helper subset, which is not part of the unusual expanding population, is required for initiating the pathology in these animals. As a means of characterizing the expanding population of abnormal cells as well as the phenotypically mature (L3T4+) cells that may be associated with them, I have generated a series of T cell hybridomas from the enlarged lymph nodes and spleen of m p r mice. In parallel, I have derived a series of control (non-lprflpr) hybridomas from MRulpr X Balb/c F1 animals (which show no sign of pathology), and a series from MRL, mice (which have a delayed onset of autoimmunity without lymphadenopathy). Very few hybridomas (20-30) were obtained in the non-lpr derived fusions. When I Con A stimulated the lymphocytes from non-lpr mice prior to fusion however, many more hybridomas were obtained(l20-220). This is in contrast to the fusion efficiency obtained from lprnpr mice which did not require in viuo lymphocyte stimulation to obtain a comparable number of hybrids. This result suggests that the M U p r lymphocytes are activated in situ. In addition, while less than 15% of the lymphoid mass is comprised of T helper (L3T4+) cells , over 50% of the hybrids are L3T4+. The fact that a dispropomonate number of T helper cells are rescued by fusion suggests that the cells activated in situ may be autoreactive T helper cells. Currently I am characterizing these T helper cells for their lymphokine production, T cell receptor gene usage and auto-specificity and will compare them to the hybridomas obtained from non-diseased animals.

Goodnow, S. Gilfillan, H-J. Garchon, J. Erikson and M. Davis. Stanford University, Stanford, CA 94305.

We have made transgenic mice bearing gene constructs encoding the T cell receptor a and p chains from a cytochrome C-reactive T cell hybridoma. Despite a lack of tissue-specificity in mRNA expression, cell surface expression of uansgene-encoded protein was limited to T cells, presumably because both chains require CD3 proteins in order to assemble on the cell surface. In mice carrying only the a chain consuuct, the transgene was expressed in the thymus as early as day 15 of fetal life, 1-2 days before endogenous a chain mRNA. The first detectable cell surface expression of a transgene was on 25% of day 15 fetal thymocytes. This vast increase in up-bearing cells in fetal thymus was due to pairing of transgenic a chains with endogenous p chains, of which a substantial number are. normally rearranged by day 15 of fetal liie. The balance between ap-expressing T cell supopulations was grossly disturbed in these mice, the most marked abnormality being an increase in the number of L3Tnyt-2-cells both in thymus and in peripheral lymphoid organs. It therefore appears that premature expression of surface ap T cell receptor may disturb T cell differentiation pathways by allowing T cells to leave the thymus without expressing L3T4 or Lyt-2. Mice carrying the p construct showed no increase in surface expression of T cell receptor in fetal life, since endogenous a chain rearrangement was limiting. In mice carrying either the a or p chain mansgenes, the number and surface phenotype of T cells expressing y& T cell receptors was unaffected in early fetal liie. suggesting that the a p and y5 T cell lineages diverge before thc rearrangement and expression of the appropriate subset of T cell receptor genes.

Department of Microbiology and Immunology,

Institute and the University of Pennsylvania, Philadelphia, Pa. 19104.The thymic stroma plays a major role in initiating the colonization, organization and differentiation of precursor stem cells into functionally mature T cells. A variety of cell types including thymic nurse cells, cortical and medullary epithelial cells, nonepithelial dendritic cells, and macrophages, combine to form the thymic stroma. The differential role of such cells in thymic development is unclear. We have isolated a number of morphologically distinct stromal cell lines from the thymuses of SV40 transgenic mice. Several of the cell lines are of epithelial origin, while others have features consistent with non-epithelial "dendritic" cells. We have focused on one of these cell lines, bearing the phenotype of a cortical epithelial cell, for its ability to support the growth and differentiation of stem cells from the fetal liver and fetal thymus, and cloned pre-T cells obtained from adult mice. The cortical epithelial cell line produces factors that induce the dramatic proliferation of fetal liver and thymic Stem cells . In addition, fetal liver cells cocultured with this cell line are induced to rearrange and express their T cell receptor (TCR) genes. A cloned pre-T cell line is also induced to rearrange its TCR oenes in resoonse to sianals mediated bv this cell line.

Gugrin, Marie C. B6n6, Corinne Amiel, Nadia Coniglio, Jacques Leclsre, Laboratoire d'Immunologie and Clinique Endocrinologique, CHU de Nancy and Facult6 de Mgdecine, 54500 Vandoeuvre les Nancy, France. The LFAl molecule, an adhesin of the LFA family involved in cell-cell interactions, is physiologically expressed on all white blood cells. It is absent in some congenital immune deficiencies ((ID), and is expressed on a decreased number of peripheral blood lymphocytes (PBL) in AIDS. We investigated its presence on PBL from 60 patients with auto-immune disorders of the thyroid. A monoclonal antibody (IOT16, Immunotech) directed to a conformational epitope involving both chains of LFAl was used in indirect immmunofluorescence on PBL from blood drawn at a similar time in all patients. A calibrated flow cytometer (Epics Profile, Coultronics) was used to measure the percentage and numbers of positive cells, as well as the mean fluorescence (MF) and shape of the fluorescent peak. Data were correlated with clinical information,therapeutic, and other PBL features such as the CD4ICDB ratio. The percentage of LFA1+ cells was significantly decreased in patients with Graves' disease, hypothyroidism and hyperthyroidism. The MF was lower and the shape of the fluorescent peak seldom displayed the bimodal characteristic noted in controls. These data suggest the participation of the altered expression of LFAl in the pathogenesis or evolution of auto-immune diseases. pt=ciilat.Pd that excess HLA r l a s s TT expression, ciinmonly found i n a i t i v e human nutoinimiine tlisrases maintdin:? the *rctivatinn of dutnreactivr T c e l l s which in turn prnducr mediators which maintain r.la:;s I T expression. This hypothesis has been tested i n many ways i n t h y r o i d i t i s . Crj t i c a l l y autoreactive T cell:? are €nilrid i n thyroid autoinnline tissues which a r e rrstimnlatrd hy thyroid f o l l j c u l a r r r l l s . More rrcently we have been exploring the s p e c i f i c i t y of the autoantigen reactive T c e l l s in Hashinoto's t h y r o i d i t i s where thyroy-lobulin s p e c i f i c clones have been found, i n contrast t o Graves' disease, where thyrocyte recognizing clones do not react wi.th tliyroglnhulin. Tn rheumatoid a r t h r i t i s , collagen type I1 clones have been found, persistently in the activated (IL-2R') T c r l l pmil over several years i n t.he same p a t i e n t . To verify t h a t antigen present,ation is involved i n rhrumatoid art.hritis ( R A ) a disease i n which, unlike thyroi.dj t i s the nature of the major antigen presenting c e l l (APC) is unknown, thc e f f e c t of ,~nticl.ass I1 dntibodies a t 1-oncentration which block *ari;j vation of T r-el Is mi the synthesis n € RLA-DR mRNA wa:j waluat.rd. The inhibitory effect supports d i.rifira1 role nf an d s yet iiriknown APC i n maintaining the I:hronj.ci t y o f R A

Conversely, all the clones were unable to respond to a substitution at 91 (tyr to asp). Nase mutant proteins were constructed with the same single amino acid substitution and T cell responses to peptides and mutants were compared. Preliminary evidence suggests that the mutant proteins like the peptides, substituted at residue 88 and 91, will not induce T cell clone responsiveness. These data suggest that the overall structure of the protein will not compensate for the lose of a particular amino acid which is necessary for T cell recognition.

Medicine, Baltimore, MD 21201. To explore the variables important in T cell priming, an adjuvantfree immunization regimen was developed. BIO.A mice were primed subcutaneously with syngeneic spleen cells that had been pulsed with high concentrations (100pM) of the peptide 81-104, a CNBr cleavage fragment of pigeon cytochrome E. The T cell response was assessed using a sensitive limiting dilution assay that measures lymphokine production with the CTL-L cell line. The precursor frequency of antigen-specific cells in the draining lymph nodes of mice primed with antigen-pulsed spleen (APS) was 1 in 4000, indistinguishable from the frequency of 1 in 3400 found in mice primed in the footpads with 10 nmol of 81-104 in complete Freund's adjuvant (CFA) (data are given as geometric means, n=5, S.E.M = x/t 1.7 and 1.3, respectively). Despite the apparent similarity in the T cell compartment of mice primed using these different regimens, antibody induction was strikingly different. Mice primed with 81-104 in CFA developed serum IgM and IgG responses against the peptide, with antibody detectable in an ELlSA assay at a 1 :3000 dilution. Mice primed with 81 -1 OWAPS, however, produced no detectable anti-peptide antibodies. Maximal T cell clonal expansion therefore appears to be possible in the absence of antigenspecific B cells. These data argue against the hypothesis that antigen-specific B cells play an obligate role in T cell proliferation in vivo. The reasons for the lack of antibody induction are currently under investigation. cell receptor (TCR) complex of Jurkat cells. The coprecipitation of these peptides with TCR requires treatment with monoclonal antibodies (mAbs) directed against TCR (C305 or R140) prior to cell lysis and immunoprecipitation. Treatment of Jurkat cells with mAbs directed against CD2 (9-1 or 9.6) or HLA (W632) does not not induce the association of these peptides with TCR.

The signal-transduction mutant cell lines, J.CaM1 and J.CaM2, have previously been described (1,Z). These cell lines, derived from Jurkat, fail to activate the inositol-phopholipid second-messenger pathway in response to anti-TCR mAbs. Treatment with mAb C305 induces the association of the 35 and 39 kD peptides with TCR in J.CaM2 cells but not in J.CaM1. J.CaM1 modulates TCR normally in response to anti-TCR mAb treatment (1). Hence, these observations suggest that the two peptides are involved in the signal-transduction pathway of the T cell receptor complex rather than receptor internalization. SLE is an autoimmune disorder associated with several different HLA class I1 antigens. We studied a large SLE patient population by sequencing of the PCR amplified first domain of the DQB and DQa chains and by sequence specific oligonucleotide probes to further define these associations. Shared DQB sequences at amino acid positions 26=1eu, 30=tyr, and 57fasp may predispose some individuals with HLA DR1,2,4, or 6 to develop SLE. A novel DQB sequence found in two DRw6 DQwl SLE patients shares these amino acids in the DQB hypervariable regions. The association of the DRZ DQwl.AZH gene was greatly increased in the SLE patients with lupus renal disease. The HLA association may be directly due to structural aspects of the HLA genes. When parent -+ F1 chimeras are prepared with supralethal irradiation (1300 rad + 900 rad), the donor-derived CD4+ cells differentiating in the chimeras show partial tolerance to host-type H-2 determinants, despite the apparent absence of host-type APC. Donor-derived CD4+ cells give only low proliferative responses to host-type APC in primary mixedlymphocyte reactions (MLR); furthermore, in I-E-+ I-E+ combinations, the donor CD4+ cells show molecules.

This finding implies that tolerance is induced intrathymically, presumably through contact with a non-marrow-derived component of the thymus, e.g. epithelial cells. In support of this possibility, thymectomized & + ( a x b)F1 chimeras given strain J? marrow cells and a strain thymus graft (irradiated) show no detectable tolerance to host-type strain b determinants: the strain & CD4+ cells differentiating in these chimeras give strong MLR to strain b and do not show deletion of v 11+ cells.

= 70% deletion of CD4+ cells expressing I-E-reactive V 11 T cell receptor B measured by these two parameters applies not only to lymph node (LN) CD4+ also to CD4+ cells recovered from the thymus. interphotoreceptor retinoid-binding protein) is a glycoprotein of 1264 residues (bovine) which localizes in the retina and pineal gland and induces inflammatory changes in these organs (EAU and EAP, respectively) in immunized animals. The experimental disease is considered a model for certain uveitic condiiions in man. We have recently shown that IRBPderived synthetic peptides can also induce EAU/EAP in Lewis rats. The present study compared two such peptides, "R4" (residues 1158-1 180) and "R14" (1 169-1 191). Peptide R14 was found to be immunodominant, shown by its being recognized by lymph node cells (LNC) or line cells sensitized against whole IRBP. In contrast, peptide R4 was not recognized by the whole IRBP-specific lymphocytes and is considered nondominant. In addition, LNC sensitized to R14, but not to R4, responded to intact IRBP. R14 was superior to R4 in producing EAU/EAP and cellular immunity (minimal doses: 0.06 vs 67 pg/rat). On the other hand, the two peptides were comparable in their capacity to stimulate presensitized lymphocytes. Moreover, LNC sensitized against R4 were similar to those sensitized against R14 in their capacity to adoptively transfer EAU/EAP to naive recipients. This study thus provides a unique system in which both immunodominant and nondominant peptides produce autoimmune disease and can be compared for their immunological features. The age-related diminution in immune responsiveness has been shown to result from increased regulatory mechanisms and not from a paucity of immunobgical recruitment (Aging: Immunology and Infectious Disease 1,47, 1988). We present evidence based on "libraries" of monoclonal antibodies (MAbs) oMained from young and aged donors that there occurs with aging an increase in autoimmunity which is possibly the result of the accumulation of liielong "original antigenic sins". The resultant increased connectivity of the immune system is represented by MAbs obtained from aged donors which are multiply anti-self cross-reactive. Furthermore increased connectivii is supported by the evidence that anti-2.4,6-trinitrophenyl MAbs are AD8 positive, 2D3 positive as determined by i n h b i i n studies using MAbs anti-idiitypic reagents. Analysis of the VH and VK region genes utilized by these MAbs indicate a nonrandom gene usage. Life bng stochastic immunological events lead to a pattern of cross-reactiities and non-random usage of VH genes. These immnological events lead to the emergence of the patterns which are partially elucidated by the data presented. These patterns mimick those seen early in ontogeny, but indicate a possible convergence to an ever-increasing connectance of the idiitypic repertoire expression. In other words, life-long immunological experiences contribute to a down-regulation resulting in both paucity of Drimaw immune reswnses and an increase in autoimrmnity which are both the earmarks of immunity in aging. (Supported by USPHS grants AG-04042 to EAG and Al 18316 to CAB) lmmunogenicity c449 Stimulation of these cell lines in suspension with saturating levels of mAb OKT3 produces total and fractional inositol phosphate accumulation linearly related to receptor number, (r > 0.9 ). This technique also allows an approximation of the minimal number of reccptors which must be engaged for second messenger generation in this system, which we estimate as 6.5~102 receptors per cell. or terminate T cell activation. Since this molecule plays an important role in human T cell development we sought to identify the murine homologue of CD28 in order to determine its expression on murine T cell and its role in activation. We have used a human CD28 cDNA clone to isolate a full length cDNA encoding the murine equivalent of CD28 from an EL4 T cell lymphoma library. This clone shows similar domain organization and a high degree of homology to the human CD28 molecule. The murine cDNA clone has been used to examine mRNA expression of CD28 in normal and activated murine T cells, and in various T cell tumors. Peptides generated from the translated sequence will be used to produce antisera to correlate the surface expression of CD28 with mRNA expression, and to biochemically and functionally characterize this molecule.

Pat Happ and Ed Palmer, Basic Sciences Division, Dept. of Pediatrics, National Jewish Center for Immunology and Respiratory Medicine, Denver, CO 80206. We have attempted to determine the frequency of rearrangement and expression of the individual a and g chain V gene segments that make up an unselected, untolerized T cell repertoire. In order to do this we generated over 500 T cell hybridomas from freshly-isolated thymocytes of newborn C57BlllO mice and subjected RNA from these hybrids to Northern dot blot analysis using 11 Va, 16 Vp. Cy and c6 probes. Comparison of the expressed repertoire of Vp gene segments in this newborn thymocyle population with similar data previously generated from an adult peripheral T cell population reveals two Vg genes, Vp12 and Vpl5, whose expression is decreased in the periphery, possibly due to the effects of tolerance. Two additional Vp gene segments were expressed more frequently in the peripheral population than in the newborn thymus, Vp5 (4.5 times higher in the periphery) and Val0 (2 times higher). It is possible that these represent Ewo instances of positive selection of T cells which is determined primarily by the receptor's Vg gene segment. Va gene segments were expressed in only 15% of newborn thymocyte hybridomas (compared to 58% expressing Vp) and determination of Va rearrangement frequencies was complicated by the unexpectedly large number (67%) of hybrids expressing CS mRNA. Further examination revealed that several Va gene probes were actually detecting rearrangements to CS. The most notable of these was Va7, which accounted for approximately 34% of the expressed Va repertoire but was rearranged exclusively to CS.

Barbara Bergman, Brenda Bradley, Kevin Lafferty and Mary Portas. Barbara Davis Center for Childhood Diabetes, U. Colo. Health Sci. Ctr., Denver, CO 80262. We have produced a panel of islet-specific T cell clones by culturing lymphoid cells obtained from non-obese diabetic (NOD) mice in the presence of NOD islet cell antigen and antigen-presenting cells (APC). These clones were selected to the panel on the basis of (a) their antigen-specific reactivity to islet cells and APC in an in Vitro proliferation assay and (b) their ability to mediate islet graft rejection in vivo in a tissue-specific manner. We have further characterized these lines for cell surface phenotype, 11-2 production, and proliferative response to non-NOD islet antigen. All of the clones tested to date are of the CD4 phenotype and make IL-2 in response to islet antigen and NOD APC. Nearly all of the clones we have tested also make good proliferative responses to islet cell antigen obtained from mouse strains other than the NOD or to a mouse beta cell tumor line. Preliminary results indicate that at least one of these clones can lead to islet cell damage in a disease transfer experiment in which the cloned T cells are injected into a non-diabetic NOD F1 recipient. We are currently carrying out tests to further characterize lymphokine production by these cloned cell lines, to analyze differences in antigen recognition and MHC restriction requirements among the clones, and to determine their effectiveness in mediating the disease process in nondiabetic animals. In an attempt to identify the epitopes on class I mo ecules recognized by alloreactive cytotoxic T lymphocytes (CTL) we have examined K -specific CTL for tpir recognition of synthetic peptides with sequences derived from the native K molecule.

Co secutive overlapping peptides molecule were tested for their capacity to inh&bit K -specific CTL clones in their recognition of cells expressing the native K molecule. In these studies inhib9ion by peptide was found to be an extremely rare event, although one peptide (K 111-122) did inhibit recognition by a particular CTL clone (clone 13). In a separate set of experiment8 it was observed that clone 13 could recopize Kblll-122 when presented by H-2 class I molecules. As clone 13 was of H-2 origin, this finding led to conclude that inhibition may be due to class I-restricted recognition of the K pegtide on the surface of the CTL clone, peptide and native Kb for the T cell receptor. We present evidence in favor of this conclusion. The PEPSCAN method is used for the systematic identification of sequential B cell epitopes i n protein molecules (Geysen, Meloen, Barteling. PNAS 81: 3998; 1984) . It was designed for the synthesis and subsequent testing for antibody binding of large numbers of overlapping peptides directly on their solid supports. The MHC dependent presentation required for T cell recognition seemed prohibitive for the use of PEPSCAN to identify T cell epitopes. However we now have shown that by a novel modification the peptides can be recovered from their solid supports and used in T cell assays. 

Holmdahl, Department of Medical and Physiological Chemistry, Box 575 Uppsala University. S-75123 Uppsala, Sweden. Both autoreactive T cells and autoantibodies play an important role in the pathogenesis of type I 1 collagen (CII) induced arthritis in mice. We have earlier reported that only strains with H-2q , H-2w3, H-2w17 and H2-r were responders to autologous mouse CII and only these strains developed arthritis after immunization with autologous or heterologous CII. However, heterologous CII induced a more acute and severe disease and a more pronounced autoantibody response. This findings indicate that 1) the ability to mount an immune response against autologous CII is a prerequisite for the susceptibility to collagen arthritis and 2) that a crossreactive autoantibody response after immunization with heterologous CII may further enhance development of arthritis.

We have now studied activation of autoreactive B cells after primary immunization of DEN1 mice with rat CII. in hybridoma collections, obtained 9-11 days after immunization, 30.80% of the hybridomas produced IgG reactive with autologous CII, 10-15% produced multispecific IgM and a significant number produced IgG rheumatoid factors. The anti-CII antibodies recognized at least 5 different epitopes on the CII molecule and originated from many different Vh and V kappa gene families. Furthermore, none out of 6 investigated anti-CII hybridoma expressed CD5 RNA message. We therefore suggest that the primary anti-CII autoantibody response involves activation of memory B-cells. These memory B cells have most likely been earlier activated by CII autoreactive T cells. In these aspects the origin of the anti-CII autoantibody response is principally different from the origin of "natural" autoantibodies. T cell receptors (TCR) recognize antigen in association with self MHC molecules, usually following processing to smaller peptides. The T cell repertoire to an antigen, therefore, reflects not only the ability of a given MHC molecule to interact with antigen, but also the affects of initial repertoire selection by self MC. We have t#tn analyzing the TCR repertoire specific for beef insulin (61) in Balb/c mice (H-2 ), which are high responders to the antigy. These studies revealed that Vp.3 is dominantly used in the TCR's specific for BI/A and our preliminary data suggests that the VgB.3 chain may be involved in MHC restriction. We have now obtained several T cell hybridmas specific for BI from (Balb/cXA/J) F1, animals. A/J mice (H-2a) are low responders to BI while the F 1 m ce a e high responders. Most of the Balb/cXA/J hybridmas were restricted to the Hin the Balb/c hybridmas. Interestingly, the analysis of V gene usage demonstrated that Vg8.3 was not used in the Balb/cXA/J hydridonas. The relevance of these results to the development of the TCR repertoire in different mouse strains will be discussed. This work was supported by the MRC of Canada. CTL specific for the q K molecule were generated from normal splenocytes by & vitro culture with A2K bearing stimulators. These CTL have been shown to lyse transfected targets expressing HLA-A2 regardless of their murine haplotype, and they specifically kill A2 bearing human target cells.

Furthermore. the effector function of these CTL can be inhibited with an HLA-A2 specific monoclonal antibody. Thus, the transgene product functions correctly as a tolerogen and is recognized directly as a class I antigen.

Although transgenic mice have been shown to be tolerant to A2K expressed by murine cells, transgenic CTL specific for HLA-A2 on the surface of human cells have been generated.

These The major virulence factor of M.pneumoniae was shown t o be a 168 kDa protein which is located in the tip structure membranes of these cells. Beside the adhesin function this protein is also involved in first massive humoral and cellular responses of the human host during the acute phase of upper respiratory tract infections and interstitiel pneumonia. Intranasal inoculation of guinea pigs with the isolated 168 kDa protein led to lympho-histiocyte infiltrations around bronchi and small vessels of the lungs which are characteristic infiltrations after an infection with live M.pneumoniae cells. Furthermore one peptide ( 1 7 amino acids long) which was synthesized according to the amino acid sequence of the adhesin, showed a proliferative activity to in vitro cultivated T-cells of bronchial washings, whereas synthetic peptides with th e sequences of the direct neighbourhood showed no in vitro activity. Most interestingly this T-cell proliferative activity is located on a surface loop of this protein which is also responsible for the adhesin f uction.

Christopher A. Smith, Gwyn T. Williams, Rosetta Kingston and John J.T. Owen. Department of Anatomy, University of Birmingham, Medical School, Vincent Drive, Birmingham 815 2TJ, UK. Rearrangement of T-cell receptor a and b chain gene segments during T-cell development results in a diverse array of receptor specificities. To avoid auto-immune responses, cells that have generated self reactive receptors are thought to be eliminated or inactivated, to produce self tolerance. Recent studies have provided compelling evidence that clonal deletion of immature receptor bearing cells within the thymus makes an important contribution to this process, although the mechanisms involved are not uderstood. of immature mouse thymocytes with anti-CD3 antibodies added to thymus organ cultures, induces DNA degradation and cell death through the endogenous pathway of apoptosis. is in marked contrast to the activation of mature T-cells by the same anti CD3 preparation and is specific to the extent that apoptosis is not induced by either anti-CD-4 or anti-Thy-1. to organ cultures suggesting a role for changes in intra-cellular C a w levels in the signalling pathway leading to the induction of apoptosis in immature cells binding. Thus activation of the process of apoptosis in immature cells binding self antigens may be the mechanism responsible for the selective deletion of cells that could generate an auto-reactive response if allowed to mature.

We have now obtained evidence that engaging the CD3/T-cell receptor complex This In addition, calcium ionophore (ionomycin) also causes apoptosis when added

Anderson Cancer Center. Houston, TX 77030 Immunization of patients with BCG and irradiated tumor cells induces specific delayed-type hypersensitivity (DTH) to tumor cells and not to normal colon cells. Since IgG antibodies may require T-cell help, we wished to characterize the IgG-defined tumor-associated autoantigens (TAAA) of human CRC so as to define a subset of the T-cell repertoire for CRC. Western blots of detergent extracts of 73 primary and metastatic human colorectal carcinomas and paired normal tissues were probed with autologous IgG. Nine TAAA were recognized by 20% or more of the sera: 74, 72, 58, 52, 45, 41, 38, 29 . and 26 kDa. These TAAA may be normal colon differentiation antigens, since they were present in extracts of normal colon. Autoantibodies are more frequently present to the 41 kDa antigen in patients with metastases (79%) than in primary tumors (47%. p<O.O5) while there is more reactivity to the 38 kDa antigen in primary tumors (38%) than in metastases (7%, pSO.05).

Only 1 of 8 patients sensitized with BCG and irradiated autologous tumor cells, developed antibody to new TAAA; whereas 2 of 12 control patients developed antibody to new TAAA without sensitization. The TAAA in the Western blots that stimulate autologous lymphocyte proliferation are greater than 200 kDa. Finally, immunization with a human CRC failed to induce DTH in mice to a syngeneic murine colon carcinoma that expresses at least 1 of the IgG-defined human TAAA. Thus, TAAA that presumably require T-cell help in patients do not appear to be part of the T-cell repertoire for DTH to human CRC. In several rodent species and strains, the emergence of specificity to a discrete immunodominant epitope of Guinea pig basic protein (GPBP) results from selection of cell lines with whole GPBP. These epitope specificities have been found to be strain-specific and define the specificity of the T helper cells which transfer Experimental autoimmune encephalomyelitis (EAE) in adoptive transfer experiments. line from ACI-strain rats which was encephalitogenic and specific for a new T cell epitope of GPBP. The cell line responded in vitro to the peptide COKKeSpOnding to the region 39-54 of GPBP but not to peptides corresponding to other regions of GPBP. This cell line is restricted by I-A but not I-E and transfers a DTH response to GPBP as well as EAE into naive recipients. These results provide support for the hypothesis that the in vitro immunodominant epitope specificity defines the specificity of GPBP-specific encephalitogenic T cells. The definition of a new encephalitogenic epitope in the ACI strain also suggests that the mechanism of EAE in rats depends on strain-specific extrisic properties of antigen recognition such as the MHC class I1 genotype or the composition of the T cell antigen-receptor repertoire.

We have produced a GPBP-specific T lymphocyte

Medical Center, Stanford, CA 94305. A wide variety of T lymphocyte surface antigens have been isolated and shown to play important roles in the T cell response. We outline here a strategy to isolate and characterize surface proteins unique to activated T cells. A rabbit anti-serum was generated by immunizing with allo-activated human T cells and subsequently absorbing with the T cell tumor HPB-ALL, a tumor that expresses all functionally relevant T cell surface antigens previously described. This absorbed antiserum was able to recognize an uncharacterized group of surface proteins on T cells that was absent from HPB-ALL. Furthermore, the absorbed anti-serum was able to inhibit i n virro cytotoxic and MLR responses; an effect that was abolished by carrying out an additional absorption with the T cells used for the initial immunization. Monoclonal antibodies to activation antigens were generated by immunizing with complexes made by adding the absorbed anti-serum to solubilized CTL. The absorbed anti-serum was also used to screen a lcgtll cDNA library. One cDNA clone so isolated is expressed on CTL as evidenced by Northern analysis. DNA sequence data indicates no similarity to other sequences in available databanks. Childrens Hospital of Orange County, Orange, CA 92668, and University of California, Irvine, CA 92717. We examined the ability of cytokines to stimulate the proliferation of PBM from 41 new-onset IDDM, 11 long-term IDDM, 10 non-diabetic pediatric control, 5 type I1 diabetic, and 3 non-diabetic adult control subjects.

Initial results demonstrated the ability of human PBM culture supernatants containing peak natural IL-1 activity to stimulate a significantly higher proliferative response from PBM of new-onset IDDM than from PBM of long-term IDDM, type I1 diabetics, or non-diabetic controls.

More recent results have shown that human recombinant IL-la, -lb, -3 , -4 and -6 , interferon gamma, lymphotoxin, and tumor necrosis factor, as well as Con P.-stimulated rat spleen supernatants, also stimulated a significantly higher proliferative respcnse from PBM of new-onset IDDM than from PBM of controls. However, human recombinant IL-2 stimulated equivalent levels of proliferation from PBM of both new-onset IDDM and controls.

These data indicate that, in new-onset IDDM, there 2.e present PBM which are in an early stage of activation not revealed by increased responsiveness to recombinant IL-2. In the case of T cells, this may represent a stage prior to an IL-2-dependent stage of activation. This would be consistent with the active induction of newly responsive cells during a stage of IDDM when islet cells are still present. This response is not the result of the metabolic aspects of IDDM but a reflection of the immunologically activated state of T cells, B cells andlor monocytelmacrophages in IDDM.

Houghten, C. David Pendleton, James L. Comette, Charles DeLisi and Jay A. Benofsky, Metabolism Branch, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892. Analysis of known helper T cell m) cpitopes suggests that Th epitopes tend to be amphipathic alpha helices, that is, helices with hydrophobic residues on one side and hydrophilic residues on the other side. Based on this hypothesis, a computer program "AMPHI" was generated which may aid in prediction of Th epitopes. Within residues 1 to 55 of spermwhale myoglobin (SwMb), a Th epitope is known to exist but has not yet been found. As an approach to further test the hypothesis, this undetermined Th epitope was sought both by an unbiased test of ten evenly overlapping 15-residue peptides that span the region, and by computer analysis. Of the ten peptides, only one was able to stimulate two clones that are specific for the 1-55 region. This peptide gave remarkably high stimulation index. Analysis of this region by the computer program also predicted the site covered by this peptide (residues 2640) to be the most likely site for Th epitope.. Therefore, unbiased in v i m testing and the computer prediction correlated well. The peptide was able to prime T cells of B1O.BR mice in vivo for in v i m response to the native SwMb as well as to the peptide fragment of residues 1-55. Immunization of mice with SwMb showed that, of the ten overlapping peptides, the major site of response is to the identified peptide. Finally, a peptide of residues 24-42 was made to increase the amphipathic score. This extended peptide induced greater proliferation of the clones. Thus, this study has identified a new dominant epitope of SwMb and confirmed the utility of the amphipathicity hypothesis in a prospective test. amino acids a t a selected site; 2 ) T-cell antigenicity, that is analysis of prirrary structures with 9-and 11-amino acid templates obtained from database of knawn Y-epitopes;

Gly, Hhydropkbic, Pplar or charged amino acid (2); 4) prediction of the secondary structure according to Gamier (3) to exclude sites having high turn or coil forming ptential. ccanparison of t h i s set with published program (1,4) shows that it has high predictive F r . So we used it fo search for probable T-epitopes i n hepatitis A virus capsid pro- and for mediators of delayed hypersensitivity (TDH ~ assayed by ear swelling induced by co-injection of responder cells with irradiated tumor cells). Lymph node cells recovered 14 days post-inoculation contained TDH cells and pTc (freq 1/1,200) , but Tc effector cells. By contrast, graft-infdtrating cells not only contained TDH and pTc (l/l,oaO), but direct Tc effector cells (21 lytic units). All three sets of T cells were antigen-spedfic. The finding of significant frequencies of pTc within the lymph nodes and the graft site, but Tc Q& at the latter, suggests that differentiation of Tc from unprimed pTc may be a multistep process, the first of which OCCUIS in the draining lymph nodes. The data further suggest that the terminal step@) in this process take place after the donally expanded pTc have disseminated to the graft site. We suspect that the final step requires help provided uniquely by TDH cells within the graft. The fmal step may not occur (as might have been expected) in the lymph node because priming of pTc and TDH take place at noncontiguous anatomic sites.

INTERFERENCE &   Seed, PNAS a, 8573, 1987) . The N-terminal amino acid sequence of normal T cell CD28 was found to match the HPB-ALL sequence except in residue 10 and 11 (...TS...) .

In principle this might reflect somatic mutation in HPB-ALL or CD28-gene polymorphism: erroneous residue assignment in the gas phase sequencing cannot be excluded entirely. To investigate the significance of this sequence variation a genomic CD28 DNA clone of a normal human T cell clone was analysed around residues 10. 11 and found to be identical with the HPB-ALL cDNA sequence. Sumtic mutation in HPB-ALL CD28 therefore can be ruled out.

The molecular characterization of T cell receptor (TCR) structure in a panel of mouse T cell clones specific for sperm whale myoglobin (SpW Mb) has shown that 6 1-Ed-restricted clones specific for the SpW Mb region 110-120 use highly homologous V Bchains. Four of these clones are indistinguishable in their response pattern to a panel of overlapping synthetic peptides spanning the SpW Mb sequence 110-120, yet use 3 different V a gene segments. We are currently trying to find fine specificity differences among these 4 clones, in order to correlate them with TCR expression. T cell hybridomas have been made from these clones, and are being screened on panels of substituted peptides. We hope that conservative substitutions at positions important for interaction with the TCR will reveal fine specificity differences among these clones.

Sloan-Kettering Institute, New York, NY 10021 We previously showed that clonal deletion, non-suppressor mediated immunological tolerance neonatally induced to Mls allo-determinants was specific in that the frequencies of T cells responding to a variety of other stimuli were unaffected while the frequency of T cells responding to Mls were drastically reduced. Here we show that Mls tolerance is detectable in the thymus prior to a detectable positive response in the periphery (spleen and lymph node). Tolerance induced by spleen cells from Mlsa congenic mice in BALB/c mice is as "potent" as that induced by whole background incompatible DBA/2 spleen cells, indicating that incompatibility at the Ml s locus ( The Ly-6 alloantigens have been shown to participate in the process of T cell activation based on the ability of anti-Ly-6 monoclonal antibodies to induce IL-2 production and proliferation of T lymphocytes. In the present investigation, we have demonstrated that peripheral T lymphocytes from A strain mice exhibited abnormally low proliferative responses after stimulation through Ly-6A/E and Ly-6C molecules when compared to responses of T cells from numerous other mouse strains. The abnormal activation of the Ly-6 pathway of A strain T cells was not due to ineffective Fc receptor cross-linking of the anti-Ly-6 monoclonal antibodies, to inappropriate cellular expression of the Ly-6A/E alloantigen in A strain T cells, or to an active suppressive phenomenon. Identification of this defect provided an opportunity to compare T cell activation by the Ly-6 and TCR pathways. Interestingly, T lymphocytes from A strain mice proliferated normally when the cells were activated by monoclonal antibodies to Thy-I or the CD3/TCR complex suggesting that this defective activation was selective for the Ly-6 T cell activation pathway. Cell separation studies of T cells and accessory cells demonstrated that this defect was quantitative rather than qualitative and that it was complex, residing at both the T cell and accessory cell levels. These results suggest that activation of T lymphocytes via the Ly-6 molecule may involve unique signalling pathways and that activation via CD3/TCR can proceed normally in the absence of a fully functional Ly-6 pathway. The H-u)d transfected L cells were recognized by the CIL when they were infected with either influenza A virus or a vaccinia recombinant virus encoding the PB2. Thirty-five synthetic peptides derived from PB2, which had either an amphipathic suucture or contained a Rothbard' epitope, were tested for their ability to render uninfected targets susceptible to lysis by influenza specific effectors. Three epitopes were identified, each of which contained a Rothbard epitope.

Department of Microbiology and Inununolag, Albany Medical College, Albany, NY 12208 Our laboratozy has developed a new, powerful technique for hVeStigating the inmplne ~n s = to peptide epitopes, involving the covalentcoupling of peptide to phosphatidylethanol-, follmed by ampleXing with additional lipids and phospholipids to form a peptide-phoqholipid complex. These cmplexes can be used to inununize mice in the absence of protein carriers or adjuvants, thus facilitating the study of the hmne response to a mall chemically defined antigen. Use of this technology has allmed us to identify two T helper cell epitopes in conserved regions of KIV gp 160 not previously identified by computer algorithims, defined by amino acids 485-518 and 585-615. Inununization with these peptides in peptide-phoqholipid mmplexes results in the prduction I G and IgG antibodies, which truss react with cloned fragments of the whole protein. Us& this tecfinolosy we have begun to characterize the irmnUne reqonse to individual peptide antigens. ?he response of H2-k mice to amino acids 494-518 of gp 160 of HIV, has been analyzed. lhe optimal dose of a peptide containing both B and T cell epitopes Was found to be 15-30 ug, depending on the mute of administtation. IM mhization required less antigen for optinarm antibody response than did IP. Anchorage in the phospholipid cmplex is a strict requirement for an antibody response. Additional variables, such as phospholipid m i t i o n and method of cross-linking have been studied and will be discussed. We believe that the use of this peptide-phospholipid ccnnplex technology will be significant both for studying the i n m u m response to single epitops and for vaccine developrent. 

Virology, National Bacteriologica; Laboratory, 105 2 1 Stockholm, Departments of Neurosurgery, Virology and Immunology, Karolinska Institute, 104 01 Stockholm, Johnson & Johnson Biomedical Research, La Jolla, California, Pentadecapeptides, sequentially overlapping by 10 a.a., were syntesized based on the HTLV-I11 sequences of gag and env proteins and used as antigens in IgG-subclass ELISAs andT-cell stimulation assays. Sera and cells were obtained from 30 asymptomatic, HIV-infected homosexuals. Reactivity in all subclasses could be found to parts of the gag-protein. Extensive areas of IgGl and 3 reactivity were indentified mainly in the p17 and N-terminal half of p24 with an additional region in the N-terminal end of p15. The highest IgGl and 3 reactivities were detected in a.a. 8-33, rich in glycine, arginine and lysine and thereby hydrophilic with a high segmental flexibility; IgGZ and 4 epitopes were found in the hydrophilic COOH-terminal of p1J; the second highest IgGl and 3 reactivities were found in the hydro-?hilic N-terminal of p15. For gp120, the IgGl epitopes identified were in the putativeloopregion (a.a. 296-331) and in the hydrophilic COOH-terminal end of gp120 (a.a.489-503). The immunodominant region of gp41 (a.a. 582-617) showed some IgG2, 3 and 4 responses inaddition to IgG1. T-cell proliferative responses were detected against many peptides usually in areas not binding IgG. Still, simultaneously T-and B-cell reactive peptides could be detected, and the showed a distinctly preference for IgG1. The variation between different patients was considerable regarding both T-cell reactivity to peptides and the subclasses of reactive IgG. A mapping of subclass restricted epitopes allows an elucidation of anti-viral imunological mechanisms. Kristin Komschlies, Laboratory of Experimental Immunology, BRMP, DCT, and BCDP, Program Resources, Inc., National Cancer Institute-Frederick Cancer Research Facility, Frederick, MD 21701 and Jackson Laboratories, Bar Harbor, ME 04609. "double positive" (DP) thymocytes are precursors for CD4+ and CD8 mature subsets have been based on indirect evidence with in vivo anti-CD8 or anti-CD4 treatments. We have approached this question directly by isolating subsets of DP thymocytes from fetal or adult thymus and have shown that only a subset of DP cells ( 4 0 % of adult thymocytes) can generate donorderived thymocytes after i.t. injection into irradiated hosts. low-density DP cells are injected i.t., up to 20-40% donor-derived cells can be seen at 6 to 18 days after transfer. CD4' cells can constitute up to 30% of these donor-derived cells at day 10. In contrast, an equivalent i.t. transfer of dull Ly-l/CD5 (dLy-1). early precursor thymocytes would not generate CD4' cells until days 12-13 but would produce some CD8' cells by 18 days. Thus we have detected differentiation of CD4+ cells and not CD8' cells from adult DP cells. depleted host and recovered donor-derived thymocytes 8 to 10 days after i.t. transfer of dLy-1 cells, when most of the donor-derived cells are DP. When retransferred, these DP cells displayed a more rapid progression of thymic localization patterns than dLy-1 cells, but were unable to produce enough donor thymocytes to analyze subsets. Thus CD4' cells appear to differentiate via a very minor (low density) subset of DP thymocytes. may not be generated from these intermediate DP precursors but directly from dLy-1 cells. The intrathymic differentiation process by which bone marrow-derived progenitor cells develop into mature, immunocompetent T lynyhocytes remains poorly understood. The majority of all intrathymic lymphocytes are CD4 8 "double positive" cells that have no well-documented function in thymic differentiation, but have been proposed to be either a critical intermediate between CD4-8precursor cells and mature T cells, or, alternatively, a "dead-end" cell type. Furthermore, although signals mediated through the TcR complex are critical for T cell tolerance induction and repertoire selection, the functions of the CD4 and CO8 differentiation antigens expressed by the vast majority of thymocytes are unknown. In the present study we addressed the possibility that CD4 functions as a signalling molecule during T cell development. We engaged cell surface CD4 on murine thymocytes by in vivo injection of anti-CD4 mAb, and monitored the effects of that engagement on distinct thymocyte subsets. C D 4 ' 8 + cells respond to CD4 cross-linking by dramatically increasing their cell surface expression of CD3 and TcR, and decreasing their cell surface expression of CD4. In contrast, mature CD4'8-cells respond to CD4 engagement by decreasing their cell surface expression of both CD3 and CD4. These results provide clear evidence that CD4 engagement regulates CD3/TcR expression on developing thymocytes, and, most interestingly, results in increased TcR expression by CD4+8+ double positive thymocytes. Thus, CD4 engagement is a membrane event to which C D 4 ' 8 ' thymocytes respond significantly, indicating that CD4 may be a signalling molecule in T cell development and may promote T cell differentiation by inducing increased CD3/TcR cell surface expression on CD4+8+ thymocytes in vivo.

F.G. Terpstra. P.Th.A. Schellekens and R.A.W. van Lier, Central Lab.Blood Transf.Service, Lab.Exp.Clin. Immunol. of the Univ.of Amsterdam, The Netherlands T cells can be activated through several membrane molecules that may use distinct intracellular signalling pathways. T-cell activatiep threygh both the CD3/TCR complex and CD2 is accompanied by riaea in free intracellular Ca (Ca ) and PKC activation. In contrast, evidence from our laboratory suggests that triggering h a CDZ8 occurs through a PKC-independent route. In this study we used Mab to various T-cell membrane antigens to localize the activation defect in T cells from asymptomatic HIV-infected men that we previously showed to have an intrinsic defect in their proliferative response to soluble anti-CD3 Nab. Compared to HIV-homosexual controls, the monocyte-dependent response of T cells from HIV+ men to soluble anti-CD3 and a mitogenic combination of anti-CDZ Mab was decreased as was the monocyte-independent resgpnse to immobilized low-dose anti-CD3. In T cells from both groups a normal rise in (Ca ) . was induced by anti-CD3 and anti-CDZ Mab. However, T cells from HIV-infected men respondad poorly to direct PKC activation by PMA in the presence of healthy donor monocytes, whereas control T cells responded vigorously. In contrast, induction of proliferation by anti-CD28 Mab was comparable in T cells from HIV+ and HIV-men. Our results indicate in HIV+ men a selective qualitative T-cell defect in activation routes that are dependent on PKC activation, whereas PKC-dependent alternative activation appears to be unaffected. These findings will be discussed with repect to HIV immunopathology and normal T-cell physiology. 

We have studied the role of the protein kinase C (PKC) in the activation of a prototype TH-2 cell, D10.G4.1. Blocking of this enzyme with the drug H-7 [1-(5-isoquinolinyl-sulfonyl)-2-methylpiperazinet leads to an increase in cellular proliferation of D10 cells activated with specific antigen or mitogen. n addition, an increase in the amount of IL-4 and IL-5 mRNA is seen under these conditions compared to the level observed in mitogen activated cells in the absence of H-7. This superinduction of IL-4 and IL-5 lymphokine mRNA is also detected in mitogen activated D10 cells that are depleted of PKC protein by pretreatment of the cells with phorbol esters. In contrast, mitogen activation of s lenic T cells measured by proliferation is blocked in the presence of H-7. These data indicate that JKC has a regulatory role in TH-2 cells by exerting a negative feedback on the events that occur after cell activation avoiding an uncontrolled response. This effect is opposite to that seen in a bulk T cell population that represents mainly TH-1 type cells. The mechanism of signal transmission is, therefore, distinct in the two types of CD4 + T cells. Aberrant expression of Class I1 MHC proteins on non-immune somatic tissue has been implicated in several autoimmune diseases, but it was unknown whether this expression was involved in the initiation of disease or resulted from induction by the immune reaction. We ha e engineered several transgenic mouse lines expressing tissues using eith r the pancreatic elastase promoter or the Class I Kb gene promoter. The I-As on the pancreas is fully capable of presenting peptide antigens to I-Ad restricted hybridomas. H-Zb mic expressing I-Ad on pancreatic acinar cells but not in thymus are not tolerant to I-As in mixed lymphocyte reactions in vitro, but do not spontaneously produce an immune respon e to the pancreas in vivo.

peripheral antigen presentation, I-A restricted immune response, I-A expression by T-cells, and widespread somatic I-Ad expression. expression, whereas another has only medullary expression. The effect that these different patterns have on MHC restriction is being examined. Ctr., Stanford, CA 94305. A transient increase in the intracellular Cat+ ion concentration has been shown to occur rapidly following activation of T lymphocytes. This response i s a more immediate and direct indication of T cell activation than either production of Interleukins or proliferation. We examined Ca++ flux analysis using Indo-l and the FACS to measure ag-speclfic T cell activation and also to sort for an ag-specific T lymphocyte subpopulation. Human T cell clones were exposed to elther allostlmulatory 6 lymphoblastoid cells (which cause prOliferatiOn of the T cells) or nonstimulatory 6 lymphoblastoid cells. In both CD4+ and CD8+ clones, we observed an ag-specific Cat+ flux response similar t o an lonomycin positive control. We then attempted to enrich for antigen specific T cells from a mixed population. Two phenotypically different human T cell clones were mixed in equal amounts. After coculture of the cell mixture w i t h 6 lymphoblastoid cells shown to stimulate only one of the T cell clones, the cell mixture was sorted on the FACS based on specific Cat+ flux. A ten fold enrichment was observed for the ag-specific clone. This approach allows rapld observation of T cell activation and provides a means of sorting antlgen reactive T cells from a heterogeneous cell population. Previously attemps to characterize the function of beta-2-microglobulin (beta-2-m) have used anti-beta-2-m antibodies, to block the functional activity of the protein. We now demonstrate that human beta-2-m and the proteolytic derivates beta-2-m Lys58 and beta-2-m desLys58 added in nanomolar concentrations to allogeneic, PHA and Con-A stimulated mouse lymphocyte cultures modulate the response induced. Thus beta-2-m and especially beta-2m desLys58 when added day 0 and 1 augment the generation of specific cytotoxic T lymphocytes in MLC, problably due to stimulation of endogeneous lymphokine production in the culture. Both the Con-A and PHA induced response were modulated. In the presence of beta-2-m and its proteolytic derivates the proliferation was optimal at 48 hours, as opposed to the 24 hours optimum in the abscence of beta-2-m. These results strongly suggest an active role of beta-2-m or proteolytic derivates hereof in the T cell activation and generation of specific cytotoxic T lymphocytes. A (1.2 fold) . The increase peaked at 2 min and returned almost to baseline by 10 mi?. 

IgM B cells showed random X inactivation, while more mature IgG and IgA B cells had only the normal X active. We conclude that the XLA gene product is required for B cell development, while the XSCID gene product is required for both T and B cell maturation. Gene mapping has placed XSCID in Xq13-21.3, linked to phosphoglycerate kinase (PGK), lod 3.1, 8=0. Mouse xid, which affects only B cells, is also linked to mouse PGK, suggesting the possibility that XSCID and xid belong to related gene families. c 530

Institute at Denver, National Jewish Center, Denver, CO 80206.

Mapping data for the BXD and AKXD recombinant inbred strains shows that the MHC and the products of at least two other genes control the expression of Vp3. One of these is linked to Ly-7, and the second for the AKXDs is linked t MTV-13 on chromos me 4.

A backcross of B1O.BR x (C3H x B1O.BR)Fl (Mls-2 x ( M l~-2~ x Mls-2 ) ) has been used to confirm the linkage of one of these controlling genes to Ly 7.

Using the Vp3-specific monoclonal antibody, exceptions have been found to the general rule, that strains which have eliminated T cells expressing a particular Vp specific for a defined self antigen in the thymus, also express this self antigen on their spleen cells, and so these cells are capable of stimulating T cell hybridomas bearing the particular Vp in quest ion. The T cell receptor (TCR) of the alloreactive T cell clone 2C recognizes the H-2Ld specificity. The cloned a (Va 3.1) and p (Vg 8.2) TCR cDNAs have been inserted into pLJ retroviral vectors containing LTR and regions of the MoMuLV and neomycin resistance genes.

Three different constructs have been used to transfect v2 cells. G418-resistant clones

showing the highest titer of viral particle psoduction have been selected, and injected subcutan ously lacking the H-2L vt cells producing retroviruses carrying the a gene have been examined for the expression of the new specificity on the infected T cells.

Northern blot analysis using the Va and the neo-probes showed, in the lymphoid tissues, a transcript with an expected size of 3.6 Kbs hybridizing with both probes. In the spleen of treated Balbfc mice, but not in dm2 mice, the expression of the exogenous Va 3 resulted in proliferation of infected T cells, as assessed by an auto-MLR proliferative assays.

into newborn Balb/c mice (H-2 L +) o r into the congenic strain dm2 loose most of their lytic activity as well as the ability to form conjugates with specific TC. Preliminary experiments show that SPDP-anti-CD3-derivatization of specific TC overcomes the trypsin effect, so that both conjugation and lysis are observed between trypsin-treated PEL and SDPD-anti-CD3-derivatized specific TC. many investigators, the Ti/CD3 complex on PEL surface is central to specific lysis. The observed inhibition of trypsin may be due to its action upon it. Since trypsin-treated PEL could be triggered toward conjugation with and lysis of TC by anti-CD3, the CD3 at least must have remained intact and is the important molecule for both the critical conjugation step and lysis. Trypsin treatment possibly disturbs the interaction of TC with Ti/CD3 of PEL so as to prevent proper engagement of the CD3. Concanavalin A (Con A) treatment of TC appears not able to bring back the lytic activity or the ability to form conjugates of trypsin-treated cells, indicating a dichotomy in the mechanism of killing by derivatization with anti-CD3 versus Con A mediation. comparing PEL and their blasts IPEB). Unlike PEL, PEB do not show LDCC or lytic activity against SPDP-anti-CD3-derivatized nonspecific TC. LDCC by Con A has been proposed to function via Ti/CD3 interaction. unlike PEL, PEB also do not show fluorescence. It may be that the Ti/CD3 complex on PEB differs from the one on PEL. At the same time, total lack of CD3 is contraindicated for PEB kill specifically. The region important in Ti interaction for LDCC and that recognized by anti-CD3 may be within the aberrant area of CD3, so neither the Con A nor the anti-CD3 trigger could be transmitted on PEB.

The importance of the CD3 is indicated further by Immunolabeling both cells with anti-CD3 revealed that c533 COMBINATION OF CD4+ AND CD8+ ISLET SPECIFIC T CELL CLONES ARE PATHOGENIC IN THE NOD MOUSE, Eva-Pia Reich and Charles A. Janeway. Jr., Department of Pathology, Section of Imunobiology, Yale University School of Medicine, 310 Cedar St., New Haven, CT 06510. Although diabetes in the NOD muse is thought to be initiated by T lymphocytes, thus far no T cell clone has been derived from NOD islets. We therefore isolated islets from acutely diabetic NOD mice and islet derived lymphocytes were propagated in IL-2 and stimulated every 2 weeks with fresh mitomycin-C treated NOD islets to maintain islet specificity (i.e. a proliferative response to NOD islets). T cell clones were generated by limiting dilution followed by recloning by single-cell manipulation. FACS analysis showed that all clones were positive for Thy-l+ and CD3+, confirming that receptor bearing T cells had been cloned. CD4+CD8-and CD4-CD8+ clones were isolated that proliferated in the presence of NOD islets. respond to NOD but not BALBlc or B1O.BR islets, while the two C D 8 ' clones respond to NOD and BALB/c islets but not BlO.BR. islet-specific, I-ANoD restricted, while the C D e clones are islet-specific and Dd restricted. Injection of individual cloned lines into young, irradiated NOD mice leads to no overt or microscopic disease. a syndrome of dehydration, rapid deterioration and death about 10-14 days after transfer. Mice sacrificed at this time show lymphocytic infiltration around the islets.

This suggests that the CD4+ clones are However, mixture of C D 4 ' and CD8+ cloned lines causes These cloned lines may be useful in the search for islet cell autoantigens. DUsseldorf 1 (FRG). Autoantibodies to nucleolar components are a common serological feature of patients suffering from scleroderma, a collagen vascular autoimmune disease. While animal models which spontaneously develop abundant anti-nucleolar antibodies have not yet been described, high titers of such antibodies may be induced by treating susceptible strains of mice with mercuric chloride. We have identified the nucleolar autoantigen against which the HgCI2-induced IgG autoantibodies from mice of strain B1O.S are directed. It is a protein of apparent molecular weight of 36 kDa and PI value of about 8.6 which is associated with the nucleolar snRNA U3, and by these criteria, is identical with a polypeptide called fibrillarin. It is striking that scleroderma patients spontaneously form autoantibodies against the same U3 RNP protein. The HgC12-induced murine and the scleroderma-specific human anti-U3 RNP autoantibodies were indistinguishable in their reactivities towards fibrillarin. They further resemble each other in so far as both recognize epitopes on the 36-kDa protein which have been highly conserved throughout evolution. Our results provide a basis to investigate at the molecular level whether similar immunoregulatory dysfunctions may lead to the preferential a n t i 4 3 RNP autoantibody formation in the animal model and in scleroderma patients. Forty BP-specific CD4+ T-cell clones were isolated from the peripheral blood o f a patient with multiple sclerosis. Studies with a panel of xenogeneic BPs of known amino acid sequence and with laree fragments of the BP molecule demonstrated the existence of at least 10 epitopes recognized by human T-cells. At least 7 of these epitopes were potentially clustered in the middle or at the C terminus of the molecule. In studies with synthetic peptides corresponding to residues 86-105 and 152-170 of human BP (HBP), 9 clones proliferated in response to 86-105 and 19 recognized 152-170. CTL assays were carried out with targets consisting of an autologous B-cell line coated with HBP. Nineteen of the 40 clones demonstrated BP-specific CTL activity; these were the same 19 that had recognized peptide 152-170 i n proliferation assays. These clones also killed targets coated with peptide 152-170 but not targets coated with 86-105. Clones that proliferated in response to peptide 86-105 or to epitopes located outside of these two regions failed to kill HBPcoated tarpets. to a lack of binding of 86-105 to the target cell, as the same B-cell line successfully presented peptide 86-105, in proliferation assays, to clones that had previously recognized this peptide when presented by E-cells. activity against BP is more restricted than is the specificity demonstrated for proliferative activity.

The lack of killinp of targets coated with 86-105 could not be ascribed Prystowsky, Department of Pathology and Laboratory Medicine, University of Pennsylvania, Philadelphia, PA As an approach to studying regulatory factors important during T cell proliferation, the expression of genes induced during IL2-stimulated T cell growth was examined. To identify genes induced by IL2, a cDNA library was made from IL2-stimulated cloned T cells in late G1 of the cell cycle and was screened by differential hybridization, selecting those clones which had higher steady-state mRNA levels after IL2 stimulation. Of 40,000 clones originally screened, 90 positive clones were identified, which represented 21 different genes. When partially sequenced and compared with sequences in the NIH and EMBL databases, 6 clones were identified as cDNAs encoding glycolytic enzymes, 4 were cDNAs of cytoskeletal genes, 3 were cDNAs coding for proteins important in protein synthesis, and 8 clones were not identified. The maximal steady state mRNA levels generally occurred during late Gl/early S phase, and the induction of most M A S was not inhibited by cycloheximide at a dose that inhibited cell proliferation. Nuclear run-on transcription demonstrated an increase in the rate of transcription for GAPDH. In addition, there is post-transcriptional regulation of expression, as some mRNAs were stabilized upon IL2 stimulation. Having cloned these mRNAs, we can now begin to study the regulation of expression of selected genes. Supported by NIH grants. Cyciosporin A (CsA). a potent immunosuppressive drug, caused organ-spedfic autoimmune disease, such as gastritis with anti-parietal cell autoantibodies or oophoritis with anti-oocyte autoanliies, in BALWc mice when the drug was administered daily for one week to newborns. Administration to adult mice did not. CsA abrogated the production of L3T4' T cells and Lyl-2' T cells in the thymus. Consequently, these T cells were substantially depleted from the peripheral lymphoid organs, especially when the drug was administered from the day of birth. Autoimmune disease was prevented when CsA-treated newborn mice were inoculated with splenic T cells from normal syngeneic mice. On the other hand, removal of the thymus immediately after neonatal CsA treatment produced autoimmune disease with a higher incidence and in a wider spectrum of organs, i.e.. thyroiditis, sialoadenitis of the salivary gland, gastritis. insulitis of the endocrine pancreas, adrenalitis, oophoritis, or orchitis. Each autoimmune disease was accompanied by the development of autoantibodies specific for the corresponding organ antigens. It is likely that CsA caused organ-specifc autoimmune disease, when administered to newborns, by depleting certain regulatory T cells (suppressor T cells) controlling self-reactive clones. (This work was supported by a grant from the Lucille P. Markey Charitable Trust.)

Artur Scherf, Philippe Dubois, Marika Plat and Luis Perelra da Sllva, Instltut Pasteur, 25-28, rue du Dr. Roux. 75015 Paris, 'U93 INSERM, Hopltal St. Louis. 2, place du Dr. Alfred Fournier. 75010 Paris, France. We Isolated from a genomlc expression library of the malarla parasite P. falciparum a clone coding for an antigen associated with the membrane of infected erythrocytes. Intensive sequence analysis of this gene, called 11-1, showed that the major part of this molecule consists of at least 200 repeats of nine amino acids. The consensus sequence of these degenerated repeats is E-E-V-V-E-E-V-V-P and secondary structure analysis predicts t h a t they have a high tendency t o form amphipathic alpha helices. The 8-and T-cell response t o the two most frequent types of repeat were analysed in five different H-2 congenlc mice strains using two synthetic peptides (PQA: Y-P-(E-E-V-V-E-E-V-V-P)n-K: PQB:

The 8-cell response t o both peptides is restricted to the H-2d and H-2k xaplotype. PCA and PQB specific antibodies do not dlscriminate between the two peptides (measured by ELISA). The MHC restriction was confirmed by anaiyslng the Tcell response to both peptides. T-cell lines speclfic for each of the peptides PQA/PQB were derived from H-2d responding mice. Although the peptides differed In four positions of the hydrophoblc part of the helices (valine replaced by isoleucine and leucine) no signiflcant dlfferences were observed in the antigen presentation between peptide PQA and PQB. In contrast, both peptides were unable to induce a proliferation of the heterologous T-cell Ilne. suggesting t h a t the variable hydrophobic amino acids of the repeats are involved in Tcell recognition. We are currently investigating if other variant repeats of the 11-1 antigen are restricted t o a different MHC haplotype and that the totality of repeats may be subjected to no restriction. Glven the type of amino acid replacements within the two repeats tested, the T-cell receptor appears t o discriminate between very similar amino acids.

Auphan, Claude Boyer, Annick Guimezanes, Claire Langlet, Cenae dImmunologie INSERM-CNRS de Marseille-Luminy, Case 906, 13288 Marseille, Cedex 9, France. The yinterferon gene is activated upon cell surface triggering of the aansmembrane T cell receptor (TcR)/CD3 complex or via phosphatidylinositol anchored Thy-1 molecules cell clones. Both these activation events are inhibited by antibodies directed at the Lyt-2jLyt-3 molecules and are lost on TcR a-chain deletion variants (1, 2). Biochemical analyses failed to reveal any association between the TcWCD3 complex and either Lyt-Wyt-3 or Thy-1 molecules. However, such studies consistently revealed a non-covalent, cis-type association between Lyt-ULyt-3 and H-2 class I products as well as between Thy-1 and H-2 class I products. The structural features of the molecules required for such interactions are being determined using hybrid engineered molecules and possible implications of such interactions for T cell activation are being evaluated.

(1) Schmitt-Verhulst, A.-M. et al. Nature 325, 628-631 (1987) .

(2) Guimezanes, A. et al. Cellular Immunol. 113,435-446 (1988) . I t i s increasingly apparent that the association of both chains is critical to reconstitute t,he 4g-MHC binding site. Combinational mechanisms for generating diversity play a major role in increasing the total size of the repertoire. However, evidence for preferential associations have been described between other heterodiaeric proteins of the immunoglobulin gene superfamilg.

I t is therefore important to determine if association between different sets of Vas and YBs occur at random or if, alternatively, preferential associations exist. We have devised a system which should enable us to determine, by DNA mediated gene transfer of cloned cDNAs encoding different Vas o r VBs, if such preferential associations do exist. We hare also used a panel of monoclonal antibodies specific for a particular Va or VB to derive T cell clones expressing these Vas or VBs.

In a polyclonally activated T cell population expressing a specific Va or VB chain, we will present data indicating that the association between these chains is preferential or random. !nteractions between T cells and MHC molecules are thought to select a T-cell repertoire skewed towards recognition of antigens in the context of self MHC molecules. In addition, T cells which react strongly to self MHC molecules are eliminated by a process called self-tolerance. We have recently described generation o f transgenic mice expressing the a6 T-cell receptor (TCR) from the cytotoxic T lymphocyte 2C (Sha et al., 1988 NATURE 335:271) .

The clone 2C was derived from a EALf3.E (H-Zb) anti-EALE/c (H-2d) MLC and is specific for the Ld class I MHC antigen. In transgenic H-2b mice, a large fraction of T cells in the periphery expressed the 2C TCR. These T cells were predominantly CD4-CD8+ and were able to specifically lyse target cells bearing Ld. In the periphery of transgenic mice expressing Ld, functional T cells bearing the 2C TCR were deleted. This elimination of autoreactive T cells appears to take place at or before the C04+CD8+ stage in thymocyte development. In addition, we report that in H-2s mice, a non-autoreactive target haplotype, large numbers of CD8' T cells bearing the 2C TCR were not found providing strong evidence for the positive selection o f the 2C TCR specificity by H-26 molecules. The requirements for CD4 expression during T cell differentiation in fetal thymus organ culture WOC) were investigated by analysis of the effects of addition of anti-CD4 Mab. Control ETOC cult& for 5 days contained predominantly CD4+CD8+ cells as well as CD4 or CD8 single positive cells expressing high levels of CD3. T cells from FTW cultured for 5 days in the presence of anti-CD4 M a b expressed markedly diminished CD4 expression. When anti-CD4 M a b was removed 20 hours prior to phenotypic analysis, it was found that CD4 was recxpressed on CD4+ CD8+ T cells. After 12 days of culture, control ROC contained high frequencies (23-3046) of CD4+CD8and CD4-CD8+ thymocytes expressing high levels of CD3. In contrast, anti-CD4 mated cultures, while containing CD4-CD8+, CD3+ T cells, were well depleted of cells which reacted with anti-CD4 antimy. When anti-CD4 Mab was removed 20 hrs prior to analysis, CD4 was reexpressed on CD4+CD8+ cells, and low frequencies (6%) of CD4+ CD8-T cells were detected, but these CD4 single-positive cells did not express high levels of CD3 or detectable ap heterodimrs. While normal frequencies of CD4-CD8+,CD3+ cells expressed Vg8 determinants, the a n t i -w mated cultures were relatively enriched in CD4-CD8+CD3and CD4-CD8-CD3-T cells compaml to control 12 day EOC. These data demonstrate that while CD8+CD3+ T cells can differentiate in the presence of anti-CD4 Mab, differentiation of CD4+CD3+ cells is inhibited and cell surface CD4 is down-modulated. CD4+CD8+ cells can develop in the presence of anti-CD4 M a b and do express T cell receptors. These results suggest that CD4 expression, while not required for development of CD3+CD4+CD8+ T cells, is necessary for differentiation of mature CD4+CD8-CD3+ T cells, and that the interaction of 0 4 with its ligand is critical for thymus selection of Class-II specific T cells.

Uppenkamp, and Alfred Singer, Experimantal Immunology Branch, NIH. Bethesda, MD 20892.

In order to investigate the role of the T cell receptor (TcR) in thymocyte maturation, we have studied thymocytes from the immunodeficient C.B-l7/scid mouse, which due to a genetic defect, is unable to express TcR genes. Despite this mutation, scid thymocytes were functional as they produced lymphokines and proliferated in response to stimuli (PMA, ionomycin, IL-2, IL-4). Phenotypic analysis revealed that, like normal immature thymocytes, sci+dd'&mocytes were large, Thy 1.2+, CD4-, CD8-, TcR-and were enriched in CD5

, IL-ZR+, Pgp-lf, and HSA+ cells. However other TcRpopulations present in normal mice, (i .e. CD4-8+TcR-and C D 4 ' 8 ' T c R -cells) were absent from scid mice. In order to determine if TcR expression was required for CD4 and CD8 expression, we analyzed thymi from the occasional scid mouse which possessed small numbers of TcR-bearing thymocytes. Interestingly, expression of CD4 or CD8 was detected only in those scid thymi that possessed TcR' thymocytes, although CD4 and CDB expression was not confined to those TcR-bearing cells. Further, the introduction of TcR' cells from Thy 1 congenic normal mice into TcR-scid mice consistently promoted the expression of CD4 and CDB (but not TcR) by scid thymocytes. Moreover, scid thymocytes from these chimeras possessed CD4+8+ cells, however no mature, single positive thymocytes of scid origin were detected. Together, these results suggest that TcR-bearing cells within the thymic milieu are required for the expression of CD4 and CD8 and for the differentiation into CD4+8+ cells. However, final maturation into mature single positive thymocytes requires that the thymocyte itself express TcR. TcR IN THYMOCYTE DIFFERENTIATION: EVIDENCE FROM THE  IMMUNODEFICIENT The CD4 T-cell surface receptor has been implicated in the stabilization of lntercellular interactlons. In addition, it Is felt to mediate the transduction of an independent lntracellular signal. We have recently shown (Veillette et al.. in press) that the CD4 molecule expressed in T-lymphocytes is complexed to the internal membrane tyrosine kinase p66'ck. suggestlng that alterations in the properties of thls enzyme mediate the CD4 related signal. To gain further Insight into the structure and function of the CD4-lck complex, we have generated NIH3T3 flbroblasts that co-express the CD4 and lck proteins. NIH3T3 flbroblasts expressing a murine lck cDNA were transfected by calclum-phosphate preclpitatlon with a construct containing a murine CD4 cDNA and the neomycin resistance gene. Stably transfected clones were selected for resistance to the amlnoglycoslde C418. We have now demonstrated that the CD4 and ~6 6 '~~ expressed in these non-lymphoid cells can form a stable non-covalent complex. The structure and function of the CD4-lck complex expressed in murine fibroblasts Is currently being evaluated. H-2b T c e l l s from H-2b+HH-2bxd responder animals fail to generate one (Type A) of these two major clonotypes. environment i s permissive for the T cell specificities of both Poly-18 clonotypes (Type A and B) and also rovides the determinants for t h e i r selection. The absence of Type A clones in the H-2b+H-2gxd chimeras can therefore only be attributed to the failure of H-2b T c e l l s to generate the appropriate T cell receptors. These results suggest a genetic hole in the T cell recognition repertoire of H-2b mice for the lysine-containing epitope of the Poly-18 antigen. We have used oligonucleotide-primed gene amplification to study the association between HLA class I1 alleles and autoimmune disease. Pemphigus vulgaris (PV) is a severe autoimmune skin disease highly associated with DR4 and DRw6.

We have shown that the nucleotide sequence of the first (and most polymorphic) domain of a new DQ allele is found in 1 3 / 1 3 Israeli DRw6 PV patients but only 1/13 OR& healthy Israeli controls. This allele differs at only position 57 from two other DQ alleles not associated with PV.

Further, we show that a shared setuence in the third hypervariable region of the first domain of two common subtypes o f DR4 and DRw6 cannot be the sole determinantDRgf susceptibility to PV. None of 7 DR4 /DRw6+ PV patients had this epitope. We also report a new DR first domain sequences from a DRw6 PV patient. The frequencly of t h ! $ allele in patients and controls is being assessed. Similar studies to identify disease associated class I1 sequences are underway in patients with multiple sclerosis and rheumatoid arthritis. The frequency of these autoimmune responses is determined genetlcally. and is essentially 100% in the projeny derived in crosses d transgenic mice to some inbred strains of mlce [e.g. those of H-2d haplotype). and yet only 1096 in others (e. g. H-2b) .

Recently. we initiated an analysis of the i m m~~l~g i d nonresponsiveness in two lineages of insulin/T antigen transgenic mice with early onset of expression of the transgene (RIP-1 Tag #2 and RIR Tag #2). While the presence of the transgenic protein durlng the prenataI/neonatal period is the primary requirement for the nonresponsive phenotype, it further appears that the levels of T antigen expression durlng embryDnic development quantJtatively influence the subsequent nonresponstvmess in adult Me. An additional modification is exerted by a genetic component whlch seems to map to the mouse major histocompatibility complex [H-2). Interesiingly, it appears that the H-2d haplotype. whlch is associated with the most complete impairment of immune responses when T antigen is expressed prenatally. is also assodated with the highest frequency of autoimmune responses when the developmental expression of the transgene is delayed. This correlation is considerably strengthened by the observation that outcrosses to C57B1/6J M-Zb) produce mlce with either a weaker tolerance in the case of the early expressers. or a less hquent and weaker autoimmune response in lines with a delayed onset of expression of the transgene. Thus. the two distinct biolcgical consequences of the presentation of this antigen (tolerance or autolmmunity) are &ly to result from differences of the developmental maturity of the immune system, and both effects appear to be similarly MHC restricted. lmmunogenicity C 549 Mapping the PZB geneti2 contribution t o lupus-like autoimmune disease.

I-Division o f Basic Science, National Jewish Center for Immunology and Respiratory Medicine, 1400 Jackson Street, Denver CO 80206; 2-The Jackson Laboratory, Bar Harbor, ME.

NZB mice carry a geneb) that contributes to the lupus-like autoimmune disease seen in (NZW X NZB) FI hybrids. However, the precise locus of this gene has yet t o be determined. To map i t s location, we are breeding a panel of NZBxSM/J recombinant inbred (RI) with NZW mice. By determining which of the RI strains produce FI animals that develop autoimmunity, the location of this NZB gene can be deduced based on the strain distribution pattern of the known genetic markers among the NZBxSMlJ RI strains. cells were then fractionated on a percoll gradient to obtain small dense 'resting" B celb and large 'activated' B cells. The small dense B celb demonstrated increased inhibition at every dose of IC examined, when compared to large B cells. For instance, pulsing with 10 ug of IC resulted in a 96% reduction in anti-FL, but not anti-trinitrophenyl, plaque forming cell (PFC) response for small dense B cells. In contrast, the PFC response of large "activated' B cells was reduced only 66%. In addition, when subtolemgenic doses of IC were used (10 ng) prostaglandin €2 (5X10-8Y) could function as a potent secondary negative signal rendering these B cells susceptible to negative signalling. We concluded that small dense B cells are more sensitive to IC induced unresponsiveness and that prostaglandin €2 may further sensitize B cells along this pathway. .5 were developed in the mouse against antigens present on humans, hamster and rat NK and T lymphocytes. Pre-incubation of lymphocytes with Moab did not block the ability of the NK cells to bind to targets but did block their ability to lyse NK sensitive targets. The pre-incubation media contained lytic factors that lysed NK sensitive targets in a 20 h chromium release assay. Interestingly pre-incubation of thymus cells or resting peripheral T lymphocytes with the Moab induced the T lymphocytes to release lytic factors and acquire NK activity against NK-sensitive targets. Lytic factors and acquired lytic activity were unable to lyse LAK targets. Signalling events did not involve cross linking or Fc receptor participation since factors were Immunoprecipitation and western blot analysis using the Moabs defined a molecule of 27 kD on the surface of thymus or lung mononuclear cells. Isolation and amino acid sequence analysis is being done to assist in the characterization of the G1. 

We have been studying in vitro activation of CD4-, CD8-thymocytes. In the present studies, highly purified double negative thymocytes were cultured with IL 1 with or without IL 2. Following a 72 hour culture with IL 1 + IL 2 (in the absence of mitogens) significant proliferation was detected, although all cells remained CD4-, CD8-. The frequency of cells staining for CD3 cell surface protein increased from 10% to 60% and TCR Vbeta8 from 4% to 40%.

FACS sorted CD3-cells from the CD4-, CD8population showed similar increases in cell surface TCR suggesting differentiation.

Northern blot analyses showed an increase in c -m RNA at 24 hours, no change in CD3 delta or epsilon expression at 24, 48, and 72 hours, and a marked increase in the ratio of 1 . 3 to 1.0 TCR beta transcript between 24 and 48 hours. These data suggest that IL 1 + IL 2 can induce CD3 cell surface protein expression on progenitor thymocytes. Furthermore, this induction is not associated with increased CD3 transcription, but rather is associated with an increased proportion of functional TCR beta transcripts. The P28 antigen of Schistosoma mansoni has been shown to be able to induce protective immunity against schistosomiasis in rodents and primates. Synthetic peptides were used to localize the major B and T-cell epitopes of this protein.

The primary structure of the P28 molecule was analyzed using various predictive algorithms : surface localization, hydrophilicity, flexibility, secondary structure, amphipathicity of predicted a-helical regions, Rothbard pattern. According to these criteria, several peptides were synthesized and tested in immunological assays in vitro and in vivo. Correlations between these results and the physico-chemical predictions will be shown. Simian virus 40 (SV40) T antigen is recognized by cytotoxic T lymphocytes (CTL) in association with class I major histocompatibility antigens.We have identified four distinct antigenic sites on SV40 T antigen defined by H-2Db restricted CTL clones. We have therefore attempted to map these antigenic sites precisely by using overlapping synthetic peptides corresponding to SV40 T antigen amino acids. The synthetic peptides of 15-20 amino acid length were incubated for five hours with 5lCr labelled-non-SV40 transformed H-2b cells in the presence of CTL clones specific for each of the four antigenic sites; and the released radioactivity was used as a measure of specific recognition of synthetic peptides by the CTL clones. The results demonstrated that the antigenic site I resides between amino acids 205-219; sites II and 111 between amino acids 220-233. Site 111 can be separated from site II as it was found to reside between amino acids 225-239. The antigenic site V was localized between amino acids 489-503. These results show that SV40 T antigen is the target protein for CTL and the regions of T antigen recognized by CTL are tightly clustered in the amino terminal half. Attempts to identify amino acids that selectively interact with the class I antigen and the T cell receptor are in progress. Peripheral blood lymphocytes of cancer patients demonstrate an early humoral and specific response to application of tumoral extracts. The response is measured as a change in the spectrum of fluorescence polarization of fluorescein molecules. This is probably a measure of sol-gel transformations induced in the cytoplasm upon cellular activation of resting lymphocytes. This measurement is carried by epifluorescent microscopy on single lymphocytes, sitted on a regular holes matrix that is scanned under computer control.

It is possible to follow changes within each cell in a population of upto ten thousand cells.

Since the cells are sitted safely in their holes, it is possible to rinse them with various reagents and follow the kinetics of individual responses by repeated scanning. Further studying of the stimulative process might be useful in identifying the responding lymphocytes and the tumoral reagents. This test of the immune system is intended for early and specific detection of cancer and for follow-up of immunotherapeutic manipulations as well.

c 559

Katharine S. Ullman, J.P. Shaw, Elizabeth Emmel, and Gerald Crabtree. Stanford University, Stanford, CA 94305.

Following antigenic stimulation, T cells undergo an orderly series of changes that result in cell division and immunologic function. In hopes of working backward in the cascade which links the membrane events to the genes essential for these differentiated functions, we have begun an analysis of the nuclear proteins responsible for gene activation in stimulated T cells. In previous studies, using a panel of internal deletion mutations of the 1L-2 enhancer, we found that when the region from -13 to -82 was mutated, the level of induction dropped 20-fold. This region is protected in a DNAase footprinting assay by a protein, NF-ILZ-A. Although the protection is seen with nuclear extracts from both stimulated and resting Jurkat cells, in vivo this is a site of DNAase hypersensitivity found only in activated Jurkats and peripheral blood T cells. A tetramer of the protected site directs transcription from an unrelated promoter in Jurkat cells only when the cells are stimulated by lectin or antibodies to the antigen receptor. Furthermore this activation can be blocked by 1 ng/ml of cyclosporin A. As a means of detecting more subtle changes in protein binding following stimulation. the methylation-interference patterns of stimulated and non-stimulated Jurkat cell nuclear extracts were compared and found to be identical. We propose that though the DNA-binding protein, NF-IL2-A, is present constitutively, it becomes functional only following specific signalling through the antigen receptor. Further characterization of this protein will help to clarify its particular role in the activation of IL-2 expression following T-cell stimulation. The negative e f f e c t of f l a n k i n g sequences on T c e l l lmmunogenicity c561 Central Lab. Blood Transf.Service, Lab. Exp. and Clin. Immunology of the Univ. of Amsterdam, The Netherlands The hydrolysis of phosphatidylinositol-diphosphate and the formation of inositol-triphosphate (IP3) and diacylglycerol (DG) re among the first events that can be monitored after T-cell triggering. IP3 mobilizes C a 2 ' from intracellular stores, whereas DG is the physiological activator of Protein Kinase C (PKC). Although it is suggested that activation of PKC is essential in the induction of mitogenic T-cell responses, no direct evidence for this hypothesis has been presented yet. In this study we used two novel PKC inhibitors, 1.e. 1-alkyl 2 methyl-glycerol (AMG) and staurosporine to assess the role of PKC activation in human T-cell stimulation. We observed that the inhibl$ors did not interfere with the anti-CD3 or anti-CDZ monoclonal antibody (mAb) induced Ca rises in peripheral blood T cells, whereas cytoplasmatic alkalinization initiated by the same stimuli or the phorbol ester PMA was strongly reduced. AMG as well as staurosporine gave a dose dependent inhibition of both Interleukin 2 (IL-2) production in Jurkat cells and T-cell proliferation. In contrast, the induction of IL-2 receptor expression was highly resistent to the action of both compounds. It was shown that anti-CD28 Mab were able to partly overcome the action of both inhibitors. These studies support the important role of PKC-activation in human T-cell triggering. Moreover, since anti-CD28 mAb were able to counteract the action of the PKC-inhibitors, it is suggested that mitogenic stimuli generated via different T-cell membrane molecules may use distinct intracellular signal transduction pathways. A f t e r incubation w i t h NDlal-26 o r al-17, t a r g e t s o f @ phenotype were lysed by a n t i -a k i l l e r c e l l s . The a n t i t h e s i s was a l s o observed: a phenotype t a r g e t s were lysed by a n t i -8 T c e l l s a f t e r incubation w i t h NDlgl-17. NDla4 We have developed a monoclonal antibody that recognizes an apparently T cell-specific antigen, LTac, expressed only in the late stages of T cell activation. When peripheral blood lymphocytes are activated with allogeneic stimulators or PHA and examined by FACS, LTac is expressed only after day 6 or 7. LTac is expressed by all T cell clones tested, both CD4' and CD8'. A wide range of cell types were tested and found negative for LTac expression, including B cells, fibroblasts, endothelium, and neural cells; thus LTac expression seems to be restricted to the T cell lineage. Immunoprecipitation of cell-surface labelled proteins and SDS-PAGE revealed a 160 kD band under reducing conditions and three bands of 240, 180, and 160 kD under non-reducing conditions. LTac is distinct from all other reported T cell activation antigens on the basis of its pattern of expression and molecular weight. We have purified LTac protein using wheat germ agglutinin-agarose chromatography followed by monoclonal antibody-agarose chromatography, and have used it to produce an antiserum suitable for screening bacterial cDNA expression libraries. The regulation and function of molecules expressed in the late stages of T cell activation is an important frontier in understanding the immune system, and reagents such as monoclonal antibodies may have important diagnostic and therapeutic uses. BALB.H.p7 mice were capable of responding to all 24 antigens tested but the magnitude of the response to some of these antigens (1 1 of 24 antigens) was reduced 2-5 fold when compared with BALB/c mice. Responses to the remaining antigens were either comparable (11 of 24 antigens) or occasionally even enhanced (2 of 24 antigens) compared to BALWc mice. These results suggest that Jp2 and Dg2 elements are required to maintain the strength of the T cell repertoire and that, in the wild, mice carrying this deletion a u l d be at a significant selective disadvantage. The identification of 11-1 was carried out in radiolabeled murine macrophages by immunoprecipitation and SDS-PAGE, using antibodies to were predominately of Mr 31.000 and 37.000, respectively. A s assessed by estimating radioactivity in each band, 11-larwas about 4 fold more abundant than 11-16. Most of the cell associated 11-1 activity, as measured by the D10.G4.1 proliferation assay, could be inhibited by anti 11-Iswhereas no inhibitory effect was observed with anti 11-1B. After removal of 11-ltc from cell lysates by immunoadsorbtion the residual activity accounted for only about 5 % of the total 11-1 activity, and was completely neutralized by anti 11-1B. However, in culture supernatants anti I1-1R reduced the 11-1 activity by approx. 50 % whereas a combination of both, anti 11-leand anti 11-10 completely neutralized the activity to background levels. Interestingly, only one radiolabeld polypeptide of M, 37.000 was precipated from the extracellular medium by the anti 11-16 antibody. These data therefore suggest, that externalization is required in order to generate bioactive 11-18 molecules, but apparently a proteolytic processing step of the precursor is not involved. Moreover, in view of previous reports, demonstrating that cellular interactions during antigen presentation can be mediated via a membrane form of 11-1, o u r experiments substantiate the hypothesis that membrane-associated, biologically active 11-1 is predominantely I1-1Q.

David L. Brandon and Joseph W. Corse. USDA Agricultural Research

Service, Western Regional Research Center, Albany, CA 94710 In order to prepare antibodies which would recognize 0-glucosides of cytokinins (purine phytohormones), we investigated an alternative to carbodiimide reagents.

Carbodiimides induce the formation of amide bonds, and thus can be used to couple haptens containing carboxyl or amino groups to proteins.

The use of water-soluble carbodiimides is limited by the sparing solubility of some haptens in water and by the formation of the intermediate 0-acylisourea, which can be less soluble than the hapten and which can itself act as an unwanted hapten. To overcome these problems, we have used 2-morpholinoethyl isocyanide and a suitable catalyst to effect amide bond formation.

These reagents are the basis of recently developed chemical procedures for peptide synthesis (Aigner et al.. 1980) . The reaction conditions permit coupling compounds which are soluble or insoluble in water to proteins in a simple, homogeneous system. We have demonstrated that this procedure is widely applicable for preparing enzyme-labeled and immunogenic conjugates of cytokinins and other biologically relevant compounds with an available carboxyl group.

We hypothesize that conjugates made with these reagents will not induce the allergic responses and unwanted crossreactivities associated with the use of carbodiimides. Supported, in part, by BARD Grant US-711-83. Mlman r I G 2 , not uwuse rIG2, uxluces a furctional1y-btut-e subset of mdlse thymqtes to. secrete -4, shichin turns iniuces cells w i t h i n this subpapllation to becane cytotcouc. Nolmal MXLS~ thymocvtes beaw CTL in to Con A * IL-2, and that the CPLresponses induced by-agentsm blocked canpletely by the addition of the anti--4 nyIb 1lB11. In addition, normal muse thymxytes produced IL-4 in response to Con A f I t 2 . The fiJmtionally-ture lobster agglutinin (mgl)-positive t h y r e p subset responded to Con A f I G 2 in a similar mmer.

ylhen the functionally-nunature subset of mgl-negative thymzyk subsets was incubated with Con A f muse r I G 2 , they did not becane CTL and &d not prcduce IL-4. Haever, when Wl-negative thymcytes were cultured w i t h Con A f human r 1~2 , cells w i t h i n this subpsxllatim developea gccd CTL activity and also secreted

'Ihese data Qmnstrate that CPL generatim and I G 4 pmauction occur mnccanitantly i n dl thymqte m a t i o r r s studied. Momever, stirmli which f a i l to induoe Cn generation also f a i l tostinulate erd0qencu.s IIr4 praduction. Immunity to the typhus group of rickettsiae is largely dependent on the effector function of several classes of T lymphocytes including those which produce gamma interferon. Since the surface protein antigen (SPA) derived from typhus group rickettsiae has been shown to be an effective immunogen in animal models, human T cell clones specific for the SPA of Rickettsia tvoh were isolated and tested for their antigenic specificity as well as for their ability to produce gamma interferon. Eighteen CM-positive clones specific for the SPA of R. t v p b exhibited considerable diversity in their response to the SPAs derived from two strains of p. orowa zekii and from R. c a n a b .

The vast majority of clones also recognized the SPA's from & or0 w w but not from R. c w . Two heteroclitic clones demonstrated significantly higher proliferative responses to the SPA's derived from one or both of the of R. orowazek i i strains than to the SPA of R., and one clone demonstrated a significantly higher response to the SPA of R. t v o k as compared to the other SPAs. All 18 clones produced gamma interferon in response to SPA stimulation. One of the clones demonstrated significant proliferative responses to a fragment of the SPA produced in E. coli infected with a recombinant lamda gt 11 phage containing a 1 kb fragment of the SPA gene from R.. We conclude that the SPA's from typhus group rickettsiae can elicit both a diverse T cell response in humans as well as the efficient stimulation of gamma interferon-mediated immunity. A monoclonal anti-mouse RBC antibody (G-8) has been prepared which appears to represent a pathogenic autoantibody related to those that arise spontaneously in aging NZB mice and which cause autoimmune hemolytic disease (AIHD). This autoantibody recognizes native erythrocytes from mice but not from other species, and Northern blot analysis revealed strong hybridization with the J558 probe but not with the 5 other Vh gene family probes tested. Thus, G-8 is clearly distinct from autoantibodies that recognize bromelain-treated MRBCs and which belong to a unique Vh gene family. When G-8-producing hybridoma cells were grown as tumors in BALB/c mice, the mice developed AIHD characterized by a decrease in the number of erythrocytes (hematwrit) and the development of Coombs-positivity. An anti-idiotypic antibody (E8) was prepared against G-8 and was found to recognize an idiotypic determinant present on most autoantibody forming cells derived from old (Coombs-positive) NZB mice. Previously, we demonstrated that treatment of spleen cells from young NZB mice with G-8 + C or with anti-Ly2 antibody + C resulted in the elimination of suppressor cells and the generation of AFC specific for MRBC. Treatment with a control NZB IgM mAb (G-4) + C or with a rabbit anti-MRBC antiserum + C had no effect on the AFC response to MRBC. Similar deletion of suppressor cells was obtained following treatment of spleen cells from BALB/c mice with G-8 + C. Greater than 80% of the resulting AFC expressed the G-8 idiotype. The results indicate that the response to self-erythrocytes is comprised of autoantibodies that express a recurrent idiotype. In an attempt to determine the optimal molecular parameters for the synthesis of anti-pathogen vaccines, we have studied the murine antibody responses to two types of model vaccines. The vaccines were: (1) Purified polysaccharide Type 3 derived from pneumococcal bacteria, and (2) Synthetic conjugates made from a peptide region of the circumsporozoite coat protein of the murine malaria pathogen, P. berghei, coupled with dextran of high molecular weight. In both cases the molecular size was systematically vaned, and an optimal size was found to be important in determining the level of immune responsiveness. In the case of the malaria peptide-dextran conjugate, epitope multiplicity (epitope density) was also found to be an extremely important parameter, immunogenicity increasing with increasing epitope density.

The James P. Allison. Cancer Research Laboratory, University of California, Berkeley, CA. 94720. We have produced an antiserum to a synthetic peptide corresponding to a portion of the murine CD2 gene product. This antiserum reacted specifically with cells expressing CD2. Flow cytometric analysis of normal lymphoid tissues revealed that thymocytes were heterogenous in the level of CD2 expression while splenic T cells expressed CD2 at uniformly high levels. CD2 expression with that of other T cell differentiation antigens (IL-2R, J l l d , CD5 and CD3) in the murine young thymus and during fetal ontogeny was analyzed using three-color flow microfluorometry. Our results indicate that CD2 expression is correlated with the stage of thymocytes maturation and demonstrate that IL-2R expression in the thymus is not induced via stimulation of the CD2 pathway.

Klinische Argeitsgruppen fur Rheumatologie/Immunologie,Schwabachanlage 10, 8520 Erlangen. Accessory cell depleted resting (la) T cells are not able to produce autocrine growth factors when stimulated via the T cell receptor complex (TCR), although IL-2 receptor expression takes place. The addition of accessory cells restores the proliferative capacity. We have tested a series of recombinant lymphokines and growth factors for their ability to substitute this accessory cell function for human CDB and CD4 T cells. Triggering via the TCR was achieved by a combinatorial stimulation with anti-CD3 and either anti-CD4 o r anti-CD8 that gives rise to IL-2 receptor expression only on the corresponding T cell subpopulation. LL-3, TNFa, IF+, IL-1, IL-4, LL-6 and GM-CSF were either negative or induced only a minor proliferative response. However, the combination of IL-1 with IL-6 was nearly as efficient as IL-2 in CD4 T cells. CD8 T cells were much less stimulated by IL-1+ IL-6. effect, which is in contrast to some reports on accessory-cell depleted mouse T cells whose function require costimulation with IL-1. This observation might be important for the understanding of T cell proliferation in chronic inflammatory tissues in which IL-1 as well as IL-6 is produced in large amounts. Structural and serological studies of these secreted effector molecules have suggested that they are disulfide-linked dimers bearing T cell receptor u and 6 chain d terminants. A T cell hybridoma, MTs 79.1, constitutively Antibody staining indicated MTs 79.1 expresses a T cell receptor containing a Vg8 element. Northern and Southern analyses indicated MTs 79.1 used V68 and Va4 genes to encode the T cell receptor. The secreted MTs 79.1 suppressor molecule was bound by monoclonal anti-V68 antibodies and by antibodies specific for constant region determinants of the u chain. To assess the role of a and 6 chain genes in encoding the secreted suppressor molecule, we have generated V68-and Vu4-negative variants from the MTs 79.1 parent. Experiments performed to date indicate that the loss of the V68 gene results in the inability to produce the suppressor molecule. The results are consistent with the idea that the T cell receptor u and 6 chain genes are used to encode these hapten-specif ic/MHC-restricted secreted suppressor molecules. producing a DNP-specific/H-ZK 5 -restricted suppressor molecule was generated and studied. Sensitivity to clonal deletion and selection for MHC restriction a m to be a feature of a particular stage of thymqte &el*, specifically l a t e mrtical CW+8+ cells. In mcent studies we have slwm that early events i n signal trarrwluction i n respnse to antigen r e c p x crosslinkiq differ in the bmdture CW+8+ p p l a t i o n and the mature CW'8-or a>rl-8 p p l a t i o n .

M t s indicate that antigen receptolg on hath inmature and mature -itive T cells tmnsdwe signals via calcium mobilization, h-er the maqnitu3e of fnflux of e&acellular Q'+ wfiich follckfi birding of antireceptor a n t q d i f f e r s tetmen these pqulaticns. Specifically, imnature cells shcw a m& reduced Q influx Espcnse carpared to mature cells. we dmw here that dligation of ~~n p has different amepexe w i t h regard to Q'+ nnbilization in mature and inmnture cells, no sucfi difference is seen f o l l a d r q ligaticm of the receptnr's transducer, a. Ihe zpsults suggest that the Signall* cascade leading to the influx Of extracellular is intact hen C D~ is ligated, but is inccnplete W I E I I m p , the physiological liw, is ligated. In addition, ligation of CW or cD8 on bmdture T cells i n a~~ influx of extracellular a ' + canparable to that sem i n mature T cells. A clonal population has been isolated frcm inmnture thymocytes whir31 has the characteristic signal tramdmtion pxqerties of the tulk of inmnture thymDcytes. 'Ihese f i r d h p suggest that "signal" transfer frcm XPnp to 03 may be inefficient in CD4+8+ cells. struchual analysis of the XPnp/CD3 canplex in hnature and mtum T cells is in progress.

Flood and Alan Friedman, Dept. of Pathology, Yale University School of Medicine, New Haven, CT, 06510. Protective immunity to the ultraviolet (W) light-induced sarcoma 1591-RE is directed toward a single tumor-specific transplantation antigen expressed by the 1591-RE tumor cells. termed the A antigen. A progressive variant line of 1591-RE, termed 1591-PRO4, lacks only the expression of this A antigen. Immunization of mice with 1591-RE tumor cells haptenated with trinitrophenyl (TNP-1591-RE) leads to the subsequent rejection of TNP-haptenated progressive tumors and to increased delayed-type hypersensitivity and CTL responses to TNP in normal, immunocompetent syngeneic mice. However, little or no humoral immunity to TNP is seen in animals injected with TNP-1591-RE tumor cells. PRO4 did not exhibit TNP-specific tumor protective or cell-mediated immunity, but rather exhibited tolerance to subsequent immunizations with more immunogenic forms of TNP. Biochemical and molecular genetic studies have revealed the A antigen to exist on the cell surface as a complex of 3 class I MHC-like molecules. Transfection studies with DNA encoding each of these molecules into 1591-PRO4 reveals that the expression of one, and only one, of these three molecules mediates this increased immunity. These experiments suggest that an MHC-like antigen expressed on the 1591-RE sarcoma acts as a natural adjuvant to increase cellmediated but not humoral immunity to linked antigens. The mechanism of this increased immunity is discussed. Efforts to immunize cattle against economically important gastrointestinal nematodes showed that inrmunity is manifested by: 1) a response that reduces the fecundity of established and subsequently acquired worms. and/or 2 ) a reduction in the number of worms developing upon challenge infection. However. the ability of individuals to mount such immunity is highly variable. Extending these studies to naturally infected populations indicate that there is a great difference in the number of eggs excreted by individual young calves on pasture. To delineate whether these differences were the result of host genetics and to begin to elucidate the mechanisms of resistance to parasite infection, a genetically defined cattle herd was assessed for parasite levels by determining fecal eggs per gram. Three years of sampling of the calves during their first grazing season indicates that: 1) certain individuals in the herd will consistently excrete high or l o w numbers of parasite eggs, 2 ) the high or low phenotype is significantly controlled by the genetic make-up of the calf, and 3 ) the high or low phenotype is highly heritable (heritability -40).

susceptibility is currently under investigation. calves have been determined and MHC class I1 typing is currently in progress. information is being used to assess the role of the bovine MHC in controlling immunoresponsiveness to parasite antigens. We have been studying the differential effects of IL 1 + IL 2 versus IL 4 on the growth and differentiation of CD 4-, CD 8-thymocytes. Culture of highly purified CD 4-, CD 8-thymocytes with IL 1 (4 U/ml) + IL 2 (100 U/ml) resulted in marked proliferation and increased cell size without change from the CD4-, CD0-phenotype. Culture with IL 1 or IL 2 alone did not cause proliferation.

A substantial contribution to the proliferation was secondary IL 4 release: addition of anti-IL 4 (11B11) blocking antibody inhibited proliferation induced by culture with IL 1 + IL 2 .

IL 4 mRNA was demonstrated by Northern blot analyses after 4 8 and 7 2 hours of culture with IL 1 + IL 2 whereas none was detected at culture initiation and very little was present at 24 hours. Effects of IL 1 + IL 2 on IL 4 transcription rate will also be reported.

Despite a marked inhibition of proliferation with anti-IL 4 , there was no affect on expansion of CD 3+ cells following culture with IL 1 + IL 2 (increase from 10% to 60% in 7 2 hours). Thus, in this system, IL 4 enhances proliferation of progenitor thymocytes but does not contribute to induction of T cell receptor. 

Carplex. lhese cwplexes can be used to inmmize mice in the absence of protein carriers or adjuvants, thus facilitating the study of the inmum to a sml1 ckmically defined antigen. Use of this tecfinology has allawed us to identify two T helper cell epitopes in cclllserved regions of HN gp 160 not previcusly identified by computer algorithims, defined by amino acids 485-518 and 585-615. Inummization with these peptides in pptide-@nqhlipid cwplexes results in the prpauction I* and I* antibodies, which CIOSS react with cloned fragmnts of the W l e protein. Us& this &logy we have begun to characterize the innume response to individual Peptide antigens. ?he reqonse of H2-k mice to amino acids 494-518 of 9p 160 of HIV, has been analyzed. lhe optimal dose of a peptide containing both B and T cell epitopes was found to be 15-30 ug, depending on the route of administration. IM d z a t i o n requir& less antigen for Opthum antibody resporrsf, than did IP. ment for an ant-response. Additional variables, as phosFholipid carpasition and method of cross-linking have been studied anl will be . Webelievethattheuseof this peptide-phospholipid canplex tehmlogy will be significant both for studying the innwE response to single epitcpes and for vaccine develolment. Based on assays in which T cell proliferation was induced via oxidative mitogenesis and exposure to MHC alloantigens, it has been reported that Langerhans cells (LC) isolated from normal mouse skin aquire maximum capacity to activate T cells only after 72 hours of culture. We have studied LC from BALB/c mouse skins for their capacity to present ovalbumin (OVA) and Ia doantigens to unprimed T cells and to antigen-specific T cell hybridomas. The data reveal that both fresh and cultured LC presented OVA and alloantigens with equal efficiency to previously primed responders (and with 10 fold greater efficiency than spleen cells or the B cell lymphoma A20.1-11). By contrast g& cultured LC displayed the capacity to present antigen to ynorimed T cells. We propose that the antigen presenting potential of freshly prepared and cultured Langerhans cells, respectively, reflect the in vivo functional properties of intraepidermal LC and of LC that have picked-up antigen in the epidermis and migrated via dermis to the regional lymph node. If so, these data suggest that resident epidermal LC are fully prepared to present cutaneous antigens to memory/effector T cells (efferent limb), whereas resident LC must leave the influence of the epidermis in order to develop the capacity to meet the more stringent conditions required for antigen presentation to porimed T cells (afferent limb). We have investigated the structural restrictions placed on residues contained within a minimal T cell determinant, using the Balb/c class I1 restricted T cell response to the site 1 determinant of the influenza hemagglutinin molecule as a model system. To delineate which of the residues comprising the site 1 determinant are involved in interaction with the T cell receptor, we have determinaed the response of a large panel of site 1 specific T cell hybridomas to a collection of peptide analogs differing by single conservative or non-conservative substitutions at 9 positions. The fine specificity patterns of the T cell panel is extremely diverse; T cells varied in both the location and number of residues within the antigenic peptide that effected recognition. Our results implicate at least 6 out of 9 residues within the antigenic pepetide as being involved in interaction with the T cell receptor. This result suggest that peptides comprising the site 1 determinant do not form alpha helical structures when in association with MHC molecules. Rubella-specific isotype and IgG subclass responses were evaluated using ELISA techniques in 44 rubella HA1 seronegative adult females undergoing rubella immunization (RA 27/3 strain). Responses were evaluated prior to immunization and at 1,2,3,4,5,6,12 and 24 wks post-immunization. Pre-immunization sera showed detectable levels of rubella-specific antibody in the IgG class (17/44); IgA class (32144) and in one or more of IgG subclasses (22/44). Post-immunization, I 1 subjects failed to develop IgM class responses by the HA1 (SDG) technique while 43/44 developed IgM antibody by ELISA techniques. IgA responses were detected at low levels in all vaccinees beginning at 3-4 wks and declining by 24 wks post-immunization. antibody in IgGl and IgG3 subclasses by 2-3 wks post-vaccine with sustained IgGl levels but significant decline in IgG3 levels noted between 6 and 24 wks post-immunization. No seroconversion was noted in the IgG2 subclass although 9 individuals had detectable pre-immunization IgGZ rubella antibody present. IgG4 levels were detected in all vaccinees post-vaccine with a delayed and progressive rise over the study period. Subsequent correlation was then performed between rubella-specific antibody responses and the presence or absence of adverse joint reactions occurring in association with rubella vaccine administration.

All individuals produced detectable C624 T LYMPHOCYTE RESPONSES TO VARICELLA ZOSTER VIRUS. Anthony Hayward, Abbas Vafai, Roger Giller & Eileen Villanueba. Departments of Pediatrics and Microbiology, University of Colorado School of Medicine, Denver CO 8 0 2 6 2 61 University of Iowa School of Medicine, Iowa City I0 The proliferative response of blood lymphocytes from varicella zoster virus (V2V)-immune donors to live VZV, extracted VZV antigens or purified glycoproteins is predominantly by CD4+, HLA-D restricted T cells but little is known of the specificies of the responder cells. We restimulated T cells cloned by limiting dilution from VZV-stimulated cultures with purified VZV glycoproteins gpI, gpII and gpIII and found that T cell clones with specificity for each of these mediated both help for antibody responses and HLA-DR restricted VZV-specific cytotoxicity. 5 polypeptides of 10 to 14 amino acids length corresponding to predicted amphipathic sequences in the primary structures of g p I, gp I1 and gp IV were synthesised. Proliferative responses were observed to 3 of these peptips (one from each glycoprotein) with responder cell frequencies in the 1:lO blood T cells range. The gp I peptide additionally defined an epitope recognised by serum antibody.

AN IMMUNOMODULATORY APPROACH TO TREATING HSV-1 CORNEAL DISEASE, Hendricks RL, Departments of Ophthalmology, and Microbiology/Immunology, University of Illinois School of Medicine, Chicago, IL 60612 Herpes simplex virus type I (HSV-1) corneal infections are a leading cause of blindness worldwide. We and others have demonstrated that the cellular immune response to HSV-1 contributes to the elimination of virus from the cornea, but in doing so causes the tissue destruction that is responsible for the blinding complications of the disease. We have demonstrated that specifically suppressing the cytotoxic T lymphocyte (CTL) response to HSV-1 renders mice resistant to corneal disease following topical corneal HSV-1 infection. In agreement with this observation was our recent finding that in vivo depletion of WT4' (T helper/inducer, and most DTH effector cells) neither reduced susceptibility to corneal disease, nor increased susceptibility to disseminated disease. The corneal lesions in WT4 depleted mice contained numerous Lyt-2 (T suppressor/cytotoxic) cells, and no L3T4 cells. The WT4 depleted mice exhibited normal HSV-specific CTL precursor frequencies.

Experiments designed to determine the effect of in vivo Lyt-2 depletion on susceptibility to corneal disease are in progress. Our goal is to identify cellular immune responses to HSV-1 that maximize protection, while minimizing immunopathology in the cornea, and identify HSV-1 epitopes that preferentially activate those responses.

Supported We found that affinity purified antibodies to BSA, KLH and diptheria toxoid all contain a substantial amount of specific anti-idiotypic activity.

against BSA react with mouse anti-BSA antibodies, which suggests that we are dealing with internal image antibodies. 9 MRL-lpr/lpr mice develop spontaneous autoimmunity. We found that these mice make anti-anti-(self H-2) antibodies prior to making appreciable amounts of pathological autoantibodies such as anti-DNA, anti-RNP.Sm, and rheumatoid factor. The anti-anti-self antibodies are detected using an inhibition of antibody mediated cytotoxicity assay, that also detects anti-anti-(self H-2) in ordinary allogeneic anti-sera. The antibodies are not rheumatoid factors, although the animals do make rheumatoid factors later in the development of the disease. anti-self activity is fully developed at 2 months, when the other autoantibodies are typically barely detectable. important role in the etiology of the disease.

The anti-We conclude that anti-anti-self antibodies could play an C 627 FEEDBACK REGULATION OF 11-1 SYNTHESIS IN MONCCYTES BY T CELL PRODUCTS: DUAL EFFECT OF 11-4. Mikko Hurme, Tessa Palkama and Marja Sihvola, Department of Bacteriology and Immunology, University of Helsinki, SF-00290, Helsinki, Finland. IL-I production of human monocytehnacrophages is regulated by several cytokines some of which are themselves able to activate the IL-1 production (e.g. TNF and IL-2) while others (e.g. IFN-y) modulate the production activated by other signals. We have now examined the effect of 11-4 on the 11-1 synthesis. 11-4 alone did not induce any 11-1 bioactivity or IL-la or 11-1 0 mRNA expression in freshly isolated peripheral blood adherent cells. In contrast, IL-4 effectively suppressed the LPS induced 11-1 production. This suppression took place without any decrease in the steady-state levels of IL-la and IL-I 0 mRNA, suggesting that this downregulative effects is posttranscriptional. Monocytehnacrophages are known to rapidly loose their ability to produce IL-I when cultivated in vitro. If IFN-yis present in the culture fluid, the cells remain capable of producing 11-1. As IFN-yand have been reported to have similar "priming" effects on macrophages (e.g. increasing the tumoricidal capacity and MHC class II antigen expression) we cultivated monocytes for 24 h in the presence of either IFN-y or 11-4, and after washing the cells they were stimulated with LPS. IL-1 activity could be detected both in the IFNyand IL-4 preincubated cultures (but not in the cultures preincubated with medium alone). These data suggest that IL-4 can also display a similar upregulatory function in IL-I production as IFN-y. Gahreston, TX 77550 Development of immunity to members of the spotted fever group of rickettsiae is a T-cell dependent response. We have used T-cell hybridomas and cloned T-cell lines from immune animals and convalescent humans to identify the rickettsial antigens that induce antigen-responsive T-cells. In these studies we found that the 155 kDa antigen of Rickettsia tickettsii. the causative agent of Rocky Mountain spotted fever, is one of the immunodominant Tcell antigens. T-cells from immune animals and humans were responded in culture to a recombinant 155 kDa antigen. Both sources of T-cells were of the T-helper type (L3T4* and 04' respectively) and produced 1L-2 and interferon. It was found that soluble antigenic material of B.

rlckettsii obtained by extraction with hypotonic buffer maximally stimulated the T-cell lines. This material was enriched far the high molecular weight polypeptides of 1S5 kDa and 120 kDa. Also.

EthiRWWi ' will induce a long-lived immunity against infection with R. rickettsii. Infected guinea pigs develop a minimally cross-reactive antibody response to B. nckettsii. In contrast. a strong cross-reactive T-cell proliferative response is produced. Studies are in progress to determine the nature of the common protective antigen of R. Humans infected with the parasitic nematode Ascaris lumbricoides vary considerably in antibody responsiveness to a 14 kDa component of the parasite. This molecule is secreted by the parasite, and is also abundant internally. This heterogeneous reactivity has been modelled in laboratory rodents, and the antibody response to it is H-2-and RT1-restricted in mice and rats, respectively. Using inbred and congenic animals, only mice of H-2' and rats of RTIU were, so far, found to be responders, and this restriction only operated in the context of infection. The specificity of the IgE response in these animals was assayed by Passive Cutaneous Anaphylaxis, and in an IgE-specific ELISA assay. The data show that the above Mhc restriction also applied to the specificity of the reaginic antibody response, although animals of all Mhc haplotypes responded to other Ascaris allergens. Amino acid analysis of the 14 kDa equates it to a previously identified "Allergen A" of the parasite, and we now have its sequence available. These findings have implications for the genetic control of allergic responses in general, and, in particular, to the hypersensitivity responses which are such a feature of infections with parasitic nematodes. There are also implications for the generation of hypersensitivity responses by recombinant vaccines involving certain parasite antigens. the CNS. Immunohistochemial analysis of both frozen sections prepared from the brains of animals immunized in this manner and of highly enriched glial cell subpopulation cultures for viral gp70 expression indicated that oligodndrocytes and possibly a subset of astrocytes were the targets of this infection. Further, microscopic analysis of frozen sections failed to reveal any overt signs of gross pathologic changes associated with the viral infection. We have been able to demonstrate the presence of virus specific antibody in the serum of these mice as well as virus specific cytolytic T cells in the peripheral lymphoid organs. ments are currently underway to determine whether the lack of pathology associated with WB91 infection in light of the previously shown virus specific immune responses in these mice is due to a failure of antigen presentation within the CNS or some other form of immunoregulatory phenomena. The T lymphocyte proliferative response to pigeon cytochrome 5 in BIO.A mice is restricted to the Egk:E,k Ia molecule and specific for the C-terminal determinant comprised of residues 93-104. BlO.A(3R) and BlO.A(SR) mice are nonresponders to pigeon cytochrome 2 Nonetheless, the T cell repertoire of BlO.A(3R) or (5R) contains some T cell clones capable of recognizing and proliferating to pigeon cytochrome c w h e n presented by BIO.A antigen-presenting cells (AFT). Therefore, one would expect to stimulate such clones in allogeneic bone marrow chimeras of the type BIO.A +BlO.A(3R) or (5R) B1O.A APCs and a BlO.A(3R) or (5R) T cell reperto4!6~.~espectively.

,en(isEa9c22RZaVe were primed with pigeon cytochrome cytochrome 5, they showed a good antigen specific proliferative response in vitro.

Surprisingly, however, if pigeon cytochrome 5 was used for priming, no response was detected, even at priming doses as high as 400 nmol per mouse. 81-104 could only be achieved by treating the allochimeras with an anti-CDB monoclonal antibody in vivo during the priming step. clones specific for purified protein derivative (PPD) in the same chimera. Thus the regulation which involves CD8 positive cells is antigen specific. Transfer of pigeon cytochrome 81-104 primed lymph node cells from the chimera into naive BIO.A mice prevented priming of the recipient for a T cell proliferative response to pigeon 81-104, but not priming to the moth synthetic fragment. chimeras of an antigen-specific suppression mechanism involving CD8 positive cells. Faculty of Medicine, Kyoto University, Kyoto 606, and Department of Oncology, Nagasaki University School of Medicine, Nagasaki 852, Japan. Sera from B6 mice immunized with a syngeneic CTL specific for FBL-3 tumor of B6 origin blocked the cytotoxic activity of only the immunizing CTL clone. Therefore, a monoclonal antibody (mAb) N9-127 was produced by fusion of the B6 spleen cells immune to a syngeneic FBL-3-specific CTL clone (No. 8). The specificity of the mAb N9-127 was confirmed by immunoprecipitation, blocking of cytolytic activity, stimulation of proliferation, and induction of TCR-mediated nonspecific cytolysis of the CTL clone No. 8. In some B6 mice, 3-13% of the anti-FBL-3 MLTC cells were positive for this N9-127-defined idiotype, and formed a well demarcated population upon examination by flow cytomehy. Even in mice in which no such population was observed some CTL clones established by limiting dilution culture were also positive for this idiotype (10 out of 89 clones from 3 mice). The cytotoxic activities of these CTL clones were blocked by N9-127, which in turn induced the nonspecific cytolysis in redirected assay. However, no positive cells were detected in non-cultured normal or FBG3-immune spleen and lymph node cells. This indicates the presence of cross-reactive (dominant) idiotype in the B6 anti-FBL-3 cytotoxic T cell responses and may provide a potent tool for analyzing the idiotype-mediated regulation of the anti-tumor immune responses. Slade andSylvie Gillard, Max-Planck-Institut fur Immunbiologie, D-7800Freiburg, Federal Republic of Germany T cells play a n essential role in t h e protective immune response to malaria and a r e associated with s o m e of t h e pathological consequences of t h e disease. However, t h e n a t u r e of their responses and t h e antigens t o which they respond a r e not well defined. W e have developed a limiting dilution assay system in which specific T cell responses to malaria antigens c a n be monitored a t t h e clonal level. I t is possible to determine t h e nature of t h e responding T cell by t h e growth f a c t o r s they s e c r e t e and by their ability to a c t as helper cells for t h e antibody response to malaria antigens. Our d a t a suggest t h a t t h e T cell response changes during t h e primary infection and in hyperimmupe animals. O n e to t w o weeks a f t e r initiation of a blood s t a g e infection t h e major CD4+ T cell which proliferates in response t o parasite antigens s e c r e t e s IL-2 and IFN-Y but is not a n efficient helper cell for antibody responses. In c o n t r a s t l a t e r in infection and in immune animals t h e r e is an e f f e c t i v e helper cell response and many of t h e s e cells a r e distinct from those secreting IFN-Y and IL-2. We a r e currently investigating whether these cells retain these phenotypes when grown in long-term in vitro culture and whether defined antigens of t h e erythrocytic parasite elicit different T cell responses. We have localized linear neutralization epitopes on the coronaviruses IBV, MHV, FIPV and TGEV. The results can be summarized as follows: 1. Linear epitopes of the spike proteins (1162-1452 residues) could be mapped to a resolution of a single residue by expression of gene fragments in the prokaryotic pEX plasmids and/or PEPSCAN peptide synthesis. 2. The length the epitopes varied from 4 to at least 20 amino acid residues. We present evidence that the larger epitopes, although conformation-independent according to operational criteria, are nevertheless discontinuous. 3 . In IBV, we localized several overlapping but different epitopes within an immunodominant region of 30 residues. This region is recognized by all polyclonal antisera tested. We propose that its immunodominancy is a consequence of its structure and function and does not depend on antigen presentation or idiotypic networks.

An immune response against the mouse testis-specific antigen LDH-C4 reduces fertility by 70 percent in female baboons. An immune reaction to human LDH-C4 would be expected to be more effective in primates. Since the human testis enzyme is not readily available in large quantities, recombinant DNA technologies were uscd to create a source of human LDH-C4. Antibodies to mouse LDH-C4 were used to screen a Xgtll human testis cDNA expression library. A full length human Ldh-c clone was identified, sequenced, and the Ldh-c cDNA was engineered for expression in E.coli.

The 5' and 3' untranslated sequences were removed by restriction enzyme digestion, and synthetic linkers were added adjacent to the start and stop codons of translation. The modified cDNA was subcloned into the prokaryotic expression vector pKK223-3 and introduced into W3110 l a c Iq cells. Cells were grown to mid-log phase, and induced with IPTG for positive regulation of the strong hybrid tac promoter. Induced cells overexpressed the 35 KD subunit which spontaneously formed the enzymatically active 140 KD tetramer.

Human LDH-C4 was purified 196-fold from liter cultures of cells by two step affinity chromatography to a specific activity of 90 I.U./mg. The 20 N-terminal amino acids sequenced were identical to those predicted from the nucleic acid sequence. Antibodies to synthetic peptide epitopes of human LDH-C4 cross-reacted with the enzyme produced in E.coli. Two mg human LDH-C4 were expressed per liter of bacterial cells. The purified protein is now available for innunogenicity and fertility studies. It is now generally accepted that the principal effector mechanism in the host's defence against leishrnaniasis is gamma-interferon (IFN-y) which activates infected-macrophage to eliminate intracellular parasites. mice by prior sublethal whole body irradiation or treatment with anti-IgM or anti-CD4 antibody.

Protection can also be induced by repeated intravenous or intraperitoneal immunisation with killed parasites or purified antigens. ly immunised mice produce little or no IL-3 or IL-4 but substantially elevated levels of IFN-y when stimulated with leishmania1 antigens in vitro. lymphoid cells from BALB/c mice with progressive disease can inhibit the MAF (macrophage activating factor) and leishmanicidal activities of the culture supernatant of lymphoid cells from mice recovered from L. major infection.

MAF appears to be IFN-y, whereas the MAF inhibiting factors are IL-3 and IL-4. system can be reproduced with recombinant IFN-y, IL-3 and IL-4 and the MAF inhibiting activity of the suppressive supernatant can be reversed by specific anti-IL-3 and anti-IL-4 antibodies. the disease by influencing the ability of macrophage to kill the intracellular parasite. The development of efficacious vaccines against malaria requires an understanding of the mechanisms involved in protective immunity. F'revious studies with Plasmodium berghei demonstrated that sporozoite immunity is dependent upon antibody responses specific for the repeat region of the circumsporozoite (CS) protein and cell mediated mechanisms involving CD8+ T cells. In this study we analyzed the splenic T cell repertoire directed against epitopes on the CS protein of P. berghei and determined whether sporozoite-immune CD4+ and CD8+ T cells respond to shared or distinct epitopes. Sporozoite-immune spleen cells, CD4+ and CD8+ enriched T cell populations of Balb/c (H-M), C3H @I-%), and C57BV6 (H-2b) mouse strains were cultured in the presence of irradiated sporozoites or synthetic peptides representing 70% of the complete CS protein. Surprisingly, none of the cultures proliferated to any of the peptides tested, although proliferative responses to sporozoites were observed in unfractionated spleens and CD4+ T cell populations. CD8+ T cells did not respond to any of the antigens tested, even in the presence of exogenously added 11-2. Titration of CD8+ cells into proliferating CD4+ cell cultures did not suppress the anti-sporozoite response. The lack of anti-peptide reactivity contrasts with uniform responses to sporozoites and may be the result of the context in which CS antigens are presented to T cells. Functional analysis of accessory splenic B cells and macrophages revealed that while the anti-sporozoite proliferative responses were not affected by the removal of macrophages, sporozoite-primed B cells were essential for the responses. These data suggest that the CS protein on sporozoites is not processed extensively by macrophages to yield many potential T cell epitopes, but instead is presented by immuncdominant B cells that resmct responses to a limited number of T cell clones. of the primary infection and is also required for optimum protection against reinfection. Current studies have demonstrated that relatively few of the viral antigens tested to date ( 7 viral envelope glycoproteins or 2 nonsmctural nuclear proteins) are recognized by HSV-1 immune CTL populations generated in several different strains of mice (H2 haplotypes H2b, H2d. or H2k). This failure of HSV specific CIZ to recognize the cloned gene products in in v i m assays was demonstrable at the clonal level and could not be attributed to a peculiarity of the recombinant vaccinia conshucts used because studies with adenovims vectors or tranfected L cell constucts yielded the same results. Surprisingly, despite their inability to be recognized by HSV specific CIZ in vim, when used to immunize mice several of the vaccinia virus constructs would induce memory CIZ populations capable of lysing HSV-1 infected autologous cells. For example, HSV-1 glycoprotein C (gC) was recognized by H2b restricted but not H2k restricted HSV specific CTL. However, immunization of either haplotype of mice with a vaccinia gC recombinant induced CTL populations which upon in v i m restimulation with HSV-1 would lyse histocompatible cells infected with HSV-1. This demonstrates that despite the presence of suitable epitopes (intrinsic factors) the context of the immunogen (extrinsic factors) will also influence it's ability to induce CTL. The results of further studies into the nature of these extrinsic factors will be presented and discussed with relevance to future sub-unit vaccine design. w d ~upponrd by public ~~l t h Service G~MU, AI 14981 md AI 24471 fran the ti-^ ~n,litu= md lnrcniour D ,~~~~~~.

Infection of mice with HSV-1 induces a brisk CIZ response which is necessary for the subsequent resolution We have investigated the structural basis for antigen mimicry by anti-idiotypic internal image antibodies. Two mouse monoclonal antibodies (mAbs) that bear internal images of a well-defined protein epitope, i.e., the rabbit immunoglobulin (Ig) a1 allotype, were produced and the variable region sequences were determined by RNA primer extension sequencing. The results showed that the mAb light chains did not contain any allotype-related residues; however, both heavy chain V regions contained a unique sequence homologous to the nominal antigen but in opposite orientation. This reversed sequence was expressed within CDR2 of both mAbs.

Synthetic peptides corresponding to the putative antigenic regions of rabbit Ig and the mAb internal images, respectively, were tested for the ability to mimic the al-like determinant. Although the homologous residues were presented in opposite orientations, both peptides completely inhibited at similar concentrations the binding of rabbit Ig to anti-a1 antibody. A paired Thr and Clu was necessary for expression of the a1 epitope as revealed by conservative substitutions in the peptide sequence. Computer-generated, energy-minimized models of rabbit Ig and the mAbs revealed that the critical a1 residue side chain placements could be almost superimposable in either context. Thus, it appears that an antigenic epitope can be determined solely by MD 20892. Proliferation of murine Type I CD4+ T cell clones quires simultaneous occupancy of the T cell antigen receptor and delivery of an accessory cell-duived costitnulamy signal In contrast, isolated T cell receptor occupancy induces the cell into a state of reduced proliferative responsiveness to antigen. Based on the observation that PKC-activating phorbl esters can at times substitute for the p s e n c e of accessory cells in T cell proliferative response. to mitogens or anti-CD3 mnodonal antibodies, we investigated the requkment for accessory cells in the antigen-and Con A-induced hydrolysis of P m and activation of PKC. The presence of normal accessory cells was found to be unnecessary for the development of PKC-dependent phosphorylations and the addition of normal accessory cells had no effect on the activity of PKC.

cell IL-2 synthesis and proliferation presents a paradox. We have studied the effects of ueatment with a calcium ionophore and p h h l ester on T cells and find that increased [Ca2+]i and PKC activation are in fact insufficient biochemid second messengers in the induction of proliferation. While pliferation was induced at high T cell density in response. to these stimuli, incubation of T cells at decrrased cell density drmonsuated markedly reduced proliferation, and single T cells failed to divide. This suggested that cellular interactions were. q u i r e d in the response. Additions of either IL2 or normal accessory cells allowed p l i f d o n at low density, consistent with a requirement for an accessory cell-derived costimulatoq signal in the induction of I L 2 synthesis, even in the plifcrative response to ionomycin and PMA. This result underscons the importance of an accessory cell-duived costimulatory signal, acting independently of T cell receptor-mediated increases in [CaZ+] i and PKC activation, in the induction of T cell proliferation. 

We describe experiments designed to determine the molecular requirements for recognition by fluorescein-specific CTLps and CTLs derived both from n a i v e and from immunized mice.

We The production of Prostaglandin E, a major immunesuppressor secreted by the macrophages was inhibited by the addition of 0.1 M indomethacin to the cultures of monocytes harvested from patients suffering from pulmonary tuberculosis and those from equal number of normal controls. The IL-I activity was estimated i n the supernatants of these cultures by their ability to proliferate mice thymocytes. It was found that the supernatants from cultures with indomethacin showed a greater IL-I activity than the ones without it(44% P 0.001). This indicates the possibility of PGE offering a negative feedback control over IL-I production. The defective cell mediated immunity i n patients with pulmonary tuberculosis may be explained through the inhibition on IL-I production by PGE whose enhanced production is reported i n our earlier studies. The results and our hypothesis on the autoregulation of IL-I production w i l l be presented and discussed. in variant viruses which differ from the parental virus (GV) at specific epitopes recognized by monoclonal antibodies directed against the env gene product, gp70. Biological clones isolated from GV express the GV phenotype suggesting that the loss of specific epitopes is the result of selective de novo processes in the immunocompetent host. Additionally, inoculation of adult mice with a biological clone expressing the GV phenotype also results in similar variant viruses. However, inoculation of GV into neonatal or nonlethally irradiated mice results in a population of viruses expressing only the GV phenotype suggesting that the emergence of antigenic variants may be influenced by neutralizing antibodies and/or cellular host res SDS-PAGE analysis of immunoprecipitates of sg~s~~,elled lysates of fibroblasts infected with clones expressing GV or variant pehnotypes shows a size difference of the gp70 precursor. Additionally, the recognition of a neutralizing epitope (E-55) associated with gp70 by mAb55 is dependent on the appropriate native conformation of the epitope which appears to require glycosylation for expression. Experiments are in progress to further examine the immunogenetic basis for the generation of these variants and to determine the molecular changes in the virus genome responsible for changes in epitope expression.

investigated the capacity of murine splenicT cells depleted of acceso cells ( AC ) t o proliferate in response t o stimulation by Con A, @ CD3 ab and activated T cells. The zepletion procedures consisted of carbonyl iron treatment, 2x "panning " on anti-Ig coated flasks, 2x anti-la cytotoxic treatments and PERCOLL gradient purification of small resting T la-cells. The appropiate concentration of Con A (10 ng/ml ) and plastic-bound (pb) @CD3 lg or its F (ab)' 2 fra ments induce proliferation, 112 r expression and 112 (but not 114 secretion in T la-cellscultured for 48% at 5x105 cells/well . Responsiveness of Tlacells t o Con A and dCD3 in low density cultures (5x104 cells/well) is restored by the addition of irradiated Th2 cloned cells but not Thl ,splenic cellsor r l l l + rll2 + r114. Likewise, responsiveness t o non activating doses of Con A (lnglml) or soluble @CD3 is restored by the addition of irradiatedTh2 costimulatory cells . These experiments demonstrate that the ability of T cells t o proliferate in the absence of AC is critically dependent on T-T interactions. T cell subsets prepared by either negative (L3T4-and Ly2'cells) or positive selection proliferate in response t o pb @CD3 lg . Although the proliferative responses of both L3T4-and Ly2-cells are maximal at 48h, the L3T4-cells require l o x more pb @CD3 lg for maximal stimulation and their res onses decline much faster than those of Ly2cells. In addition, L3T4-cells are not stimulated by pb &CD3 F(ab)'2 fra ments and their responses t o @CD3 I are inhibitable by anti FcR as well as anti-LFA abs . Re onsesof%oth L3T4-and Ly2-cells are !nhibita%le by @I12 and @I12 r abs but not by @L3T4, @Ly2 or b l l 4 . These experimentsdocument interesting differences in the triggering requirements of L3T4-and Ly2 'cells. Supported by NIH grants PO1 CA 19266, T 326M-07183 and GM 07281. The 19K glycoprotein encoded in the E3 region of Ad2 and Ad5 (gpl9K) binds to class I MHC antigens in the endoplasmic reticulum and prevents their translocation to the cell surface. This has been proposed as a mechanism by which virus infected cells can avoid recognition by the host cytotoxic T lymphocyte (CTL) response. We have shown that gpl9K can inhibit target cell lysis by adenovirus specific CTL, but the effectiveness of this inhibition varies greatly between different mouse strains. This is due in part to differences in the affinity of gpl9K for different MHC class I molecules, but this cannot account for a l l the variation observed. I t has been shown that CD4+8+ immature thymocytes fail to secrete IL-2 or express IL-2 receptors in response to activation signals. Furthermore. they cannot induce IL-2 gene transcription. Several tumor lines have now been characterized which have a CD4+8+ phenotype and fail to secrete I L -2 or express IL-2 receptors in response to stimulation with ionomycin plus PMA. These cells also do not express IL-2 mRNA after stimulation, as determined by Northern blotting and RNase protection. To determine the molecular mechanism for this lack of transcriptional activity, nuclear extracts were analysed for the presence of the D N A binding factor NFAT-I. This nuclear factor i s present only in ac- 

We have undertaken an MHC analysis using the polymerase chain-reaction (PCR) and dot blot analysis of the amplified lyme arthritis patients DNA with allele specific oligonucleotide (ASO) probes. Genomic DNA for the first domain of the DQ beta chain and of the DR/pI from 20 patients with lyme arthritis has been amplified and we are analyzing the distribution of DRIDRR and W a l l e l e s in this population to test the hypothesis that the MHC class I1 genes might be involved in presentation of selected spirochete epitopes whose recognition by T lymphocytes leads to lyme arthritis. Most inbred strains of mice do not respond to porc insulin (PIns). Experiments were conducted to elucidate the mechanism of the non-responsiveness in H-Zk mice: 1) Purification of CD-4' T c e l l s from PIns-immune B1O.MBR mice revealed PIns-specific T helper (Th) cells, 2) these PIns specific Th cells could be activated by I-Ak and I-Ek expressing L929-fibroblasts. Therefore, both I-Ak and I-Ek molecules can present PIns in an immunogenic manner and activate PIns-specific Th cells. By means of different cell-fractionation procedures, it was found that antigen-specific T suppressor (Ts) cells regulated the PIns immune response. These Ts cells were of the FcR-, CD-4-, CD-8'.

Thy-1' phenotype, and they were present in normal mice. We believe that these experiments indicate that antigen-specific Ts cells exist and are important regulators of immune and autoimmune responses. The possibility of functional inactivation of CD4+ clones by Ts-cells was investigated. M leprae-responsive CD4+ clones were preincubated with Ts CD8+ clones, APC and antigen for 1 ; hours. after which the CD8+ cells were removed from culture. The CD4+ clones were then restimulated with E. leprae and APC. CD4+ clones incubated with CD8+ cells and antigen were unresponsive to restimulation by antigen, although they were not killed and could respond well to IL-2.

Addition of IL-2 in the preor post-incubation culture neither prevented the induction of unresponsiveness nor reversed it. Earlier models of tolerance have suggested that receptor occupancy in the absence of second signals induces tolerance in B-and T-cells. We would suggest that in the presence of Ts-cells. a second signal may be negated leading to Th-cell unresponsiveness.

University of Texas Southwestern Medical School, Dallas, TX 75235 Graft versus host disease (GVHD) poses a serious threat to the survival of patients with bone marrow transplants. The state of immunosuppression established in GVHD results in a variety of immunological abnormalities at the humoral and/or cellular level. We have developed a murine model of chronic GVHD across a minor histocompatibility (mH) barrier. In this model, immunosuppression develops. Spleen cells from mice undergoing this type of GVHD are unable to respond to the polyclonal activators lipopolysaccharide and Concanavalin A .

However, the response against the B cell leukemia BCLl remains intact. The protective immune response against BCLl is directed towards the mH antigen H-40 and is mediated by cytotoxic T lymphocytes. Thus, the specific T cell response against a mH antigen can occur in the presence of chronic GVHD despite the absence of a polyclonal B and T cell response Lymphocytic choriomeningitis virus (LCMV), a member of the arenavirus family, has a biseqmented RNA genome which encodes at least three polypeptides. The smaller RNA segment encodes two virus structural proteins, the slycoprotein (GP) and the nucleoprotein (NP). Upon infection of mice with LCMV a cytotoxic T cell immune response directed against these proteins is measurable in vitro and in vivo. It can be demonstrated that depending on the haplotype of the mice, one or the other protein may play a major part in the immune response. In order to define the immunogenic epitope(s) of the nucleoprotein which are recognized specifically by the T cell receptors of cytotoxic T cells, stepwise 3 ' truncated qene fraaments encodina the nucleoprotein were cloned and expressed in vaccinia virus. With these recombinant vaccinia viruses, protection experiments in mice aqainst LCMV infection were performed in parallel with in vitro studies, namely specific recoonition of target cells expressing truncated fragments of the nucleoprotein by LCMV primed spleen cells. cDNA clones encoding the mouse and human T cell IL-1 receptors have been isolated and expressed in mammalian cells. The recombinant receptor binds IL-1 indistinguishably from the natural IL-1 receptor, and is functional in signal transduction. Deletion of the cytoplasmic portion of the receptor abolishes its signal transduction abilities. Sequence and secondary structure analysis suggest that the cytoplasmic segment of the iL-1 receptor binds a nucleotide. Experiments designed to test this hypothesis and to examine the mode of signal transduction will be presented. Also to be discussed are the mechanism of triggering of the receptor by IL-I. and the nature of IL-1 receptors expressed in other cell types such as B cells. it became evident that these subsets reflect different stages of helper T cell maturation before and after activation. Therefore, these T cell subsets have been designated as naive T cells (CD45R/2H4+, CDw29[4B4]-) and memory T cells (CD45R/2H4-, CDw29[4B4]+). We analysed the expression of these antigens in dermal lymphohistiocytic infiltrates from different benign skin diseases and cutaneous T cell lymphomas (chronic contact dermatitis (n=14), parapsoriasis en plaques (n=5), lymphomatoid papulosis (n=4), mycosis fungoides (n=15), Sezary's syndrome (n-2), pleomorphic T cell lymphoma (n=6) and high grade T cell lymphomas (~4)). In almost all cutaneous T cell infiltrates memory T cells were preferentially found whereas in the peripheral blood both subsets are equally distributed. This implicates, that T cells infiltrating the skin already have had contact with their respective antigen. Where the switch from naive to memory T cells takes place can not be answerded by our findings, as we have investigated rather longstanding skin diseases. However, these memory T cells, which can be activated more easily, make diseased skin more e f f e c t i v e in the nmDlification of an immune resnonse. organs and after several days of stimulation with antigen or mitogen and lymphokines. We find that fresh Th synthesize and secerte -IU,IFNg,IL3 and GMCSF but very little IL4 or IL.5 within 24-48 hours. This pattern resembles the pattern of lymphokines secreted by Thl cell lines. The Th responsible for this secretion are CD4 positive T cells which are long-lived since they disappear very slowly following adult thymectomy. They are also sensitive to the in vivo administration of ATS(antithyrn0cyte serum) and they express high levels of Pgp-1. The kinetics of lymphokine secretion and the phenotype of the cells suppon the hypothesis that lymphokine secretion from fresh lymphoid cells comes from a population of memory cells. In contrast we fmd that we can also stimulate a separate, ATS-resistant population to become lymphokine-secreting cells after four days of in v i m priming.These primed cultures rapidly synthesize an secrete large amounts of I U and IL5 in addition to IFNg , IL3 and GMCSF( A phenotype which could be combination of both Thl and Th2 helpers), when they are restimulated with Ag or mitogen. The cells whic are responsible are CD4 positive and have a shorter liespan since they decline considerably after adult thymectomy. We suggest that the lymphokine secreting cells detected after priming come from a population(s) of helper T cell precufiofi which have differentiated to become effectors. This generation 0: effectors requires lymphokines, especially E-4 andlor ILZ and APCs. Thus the development of helper Th appears to follow a similar pathway as that of cells of the B cell lineage developing into Ab-secreting cells and CD8 positive T cells which develop into cytotoxic effectors from precursors.

We have studied the secretion of lymphokines by helper T cells freshly obtained from lymphoid Interleukin-2 production by T cells has been shown to be required for both humoral as well as cell-mediated imune responses. Thus, IL-2 production was measured in syphilitic rabbits as a function of their imune response. Maximal IL-2 production induced by Con A at 10-14 days post-infection was only l/2 that observed for uninfected rabbits and this correlated well with a decrease in T cell proliferation ( < 35% that of normal rabbits) upon stimulation with Con A . This decrease in IL-2 production in infected rabbits was restored upon removal of most of the adherent cells. Furthermore, the IL-2 production by 10-14 day infected spleens was restored above normal levels upon addition of indomethacin. This decrease in IL-2 levels was not due to an increase in the ability of infected spleen cells to adsorb IL-2. Finally, studies assessing IL-2 production at various times postinfection indicated that at 4 days post-infection IL-2 levels were higher than normal, however as early as 10-14 days after infection IL-2 levels decreased below normal levels and continued to be depressed as late as 30 days post-infection. These results may explain why all organisms are not eliminated during primary infection w i t h ' . pallidum and why secondary and tertiary phases of the disease may develop. has been shown that especially the antigenic presentation of the fusion protein is important for eliciting a functional immune response. To study the immunolo ic properties of the F protein, we expressed the F gene in E.coli as a $galactosidase-F-fusion protein after insertion in a pEX vector. We constructed deletion mutants with fragments generated with restriction enzymes and with the polymerase chain reaction method. Using a panel of monoclonal antibodies a rough epitope mapping has been performed. Two arears were found on the protein, with one area two monoclonal antibodies react and with an other area four monoclonal antibodies react. Both area's were found in F1, the c-terminal part of the protein.

The pepscan method was used to fine map the epitopes of the monoclonal antibodies reacting with the second area on the primary sequence.

In at least one viral system, CD8+ effector cells can be induced in animals lacking CD4+ T cells. Since CD8+ effector cells are important in immunity to malaria sporozoites, we wished to know if they,too, could be generated without help from CD4+ cells. We depleted BALB/c mice of their CD4+ T cells by injection of an anti-CD4 monoclonal antibody, and then tried to immunize them with irradiated Plasmodium yoelii sporozoites. When challenged with infectious sporozoites, these mice were not protected against malaria infection. Although they did not make antibodies to sporozoites, passive transfer of hyperimmune serum into these animals still did not protect them against a sporozoite infections. CD8+ T cells from these animals functioned normally in in vitro assays against TNP labelled targets.

It appears that, unlike viral systems, the generation of CD8+ effectors in malaria requires CD4+ helper cells. Thus both CD4 and CD8 epitopes should be included in any synthetic vaccine against malaria sporozoites.

Univ. Pennsylvania, Philadelphia, PA 19104 migrate from fetal liver or bone marrow. rearrange T cell receptor (TCR) genes, express TCR. undergo thymic selection and finally emerge as mature single positive T lymphocytes.

Most studies of thymic T cell development have been performed by using polyclonal populations of T lymphocytes, which have made the interpretation of the results complicated.

cells (0 clone) from nude mice by culturing nylon wool non-adherent CD4-CD8-spleen and lymph node cells in the presence of WEHI3 supernatant and Con A supernatant. clone was Thy-1-CD3-CD4-CD8-IL2R(IL2 receptor)-and they have been maintained more than 16 months without changing phenotype.

When the C9 clone was stimulated with IL4. ILI/IL2. ILlnL6. GM-CSF, the cells were induced lo express Thy-I, TCR and IL2R proteins. However, culture of the cells with GM-CSFflL2 did not induce the expression of these molecules. Southern blotting of the DNA isolakd from GM-CSFflL2 culture suggested that they have undergone partial Db 1 -Jp 1 rearrangement. The cultured cells were then recloned twice by limiting dilution.

The cloned cells were again shown to induce expression of CD3 complex by the stimulation of IL4 enriched medium.

Therefore. we have established a system in which to induce differentiation of cloned pre T cell line into TCR+ cells in vitro. The human lymphocyte differentiation antigen CD8 is encoded by a single gene which gives rise to a 32 kDa glycoprotein expressed on the cell surface as a dimer, and in higher molecular weight forms. We demonstrate that the mRNA is alternatively spliced such that an exon encoding a transmembrane domain is deleted. that is secreted and exists primarily as a monomer. Messenger RNA corresponding to both forms is present in peripheral blood lymphocytes,(PBL), Con A activated PBL and three CD8+ T cell lines with the membrane form being the major species. ratio of mRNA for membrane CD8 (mCD8) and secreted CD8 (sCD8) exist. In addition, the splicing pattern we observe differs from the pattern found for the mouse CD8 gene. This mRNA is also alternatively spliced, but an exon encoding a cytoplasmic region is deleted giving rise to a cell surface molecule which differs in its cytoplasmic tail from the protein encoded by the longer mRNA. Neither protein i s secreted. This is one of the first examples of a different splicing pattern between two homologous mouse and human genes giving rise to very different proteins. This represents one mechanism of generating diversity during speciation. CD4' T cells in the rat can be divided into two non-overlapping subsets by theii reactivity with the monoclonal antibody MRC OX-22 which binds some of the high molecular weight forms of the CD45 antigen. Recent work, to be described has shown that the two subsets represent different stages of T cell maturation, with distinct T cell functions. The lymphokine repertoire of the memory T cell pool will be discussed with reference to the antigenic environment fmm which the cells are obtained.

PANCREATIC ISLET ALLOGRAFTS, Gill, Ronald G. and LafTerty, Kevin J., Barbara Davis Center for Childhood Diabetes I Univ. Colo. Health Sciences Center, 4200 E. 9th Ave, Box B-140, Denver, CO. 80262

We studied the cellular interactions requkd for the rejeaion of cultured MHC class I-dispiuate islet allografts. This model was suitable for studying T-T collaboration in that islet allograft immunity is CD4 dependent but rejection of the cultured islet graft is mediated by the CD8 cell. Recipient C57BY6 (B6) mice were grafted with MHC class Idisparate B6.C-H-2bml (bml) islets beneath the mal capsule. Islet grafts wen preaated for 7 days in 95% oxygen culture to reduce immunogenicity. Thirty days after grafting, recipient mice wen immunized with 1Oe6 live spleen cells from the strains indicated below. Rejection of the established graft was not trig-by challenge with donortype bml spleen cells, indicating that the MHC class I stimulus was insufficient to initiate all@ immunity. Further, immunization with a mixture of 1 M bml and 1Oe6 MHC class II-disparate B6.C-H-2bm12 (bml2) spleen cells failed to trigger host immunity. However, challenge with lOe6 (bml x bml2)Fl spleen cells t r i g g d acute rejection of the established bml islet grafts. The requirement for l i i presmtatidrecognition of class I and class 11 all* antigens to trigger allograft immunity indicates that the antigen-presenting (Apc) plays an essential role for T-T collaboration in vivo. Uature HLA-DR complexes are purported to spontaneously internalize from and recycle to the plasma membrane of B but not T lymphocytes. Using a neuraminidase protection assay, we have radiolabeled surface class I1 antigens on intact cells and cultured these cells under conditions which permit or prevent endocytosis; subsequently, surface glycoproteins on viable cells were desialylated and class I1 molecules were analyzed by iamunoprecipitation and two-dimensional gel electrophoresis. A panel of buman B lymphoblastoid cell lines and activated tonsillar B cell blasts failed to exhibit any internalization of class I1 complexes; control transferrin receptor molecules were endocytosed as ascertained by insensitivity to neuraminidase digestion. Class II+ PBA blasts and Sezary cells of the T lineage were also deficient in detectable HLA-DR internalization. Results did not vary regardless of the time allowed for efficient endocytosis (1251 OK 3~Saethionine). or the addition of anti-class I1 monoclonal antibody during the chase period for endocytosis. Therefore. within the limits of sensitivity of this assay, class 11 complexes do not appear to be internalized, either spontaneously or uben crosslinked by antibody. recycling represent a dynamic pathvay for regulating surface expression of class I1 antigens or a means of associating with and presenting foreign antigenic peptides. Supported in part by USPHS grants #5 T32 CAO9058-14 and #5 EO1 A123081-03. In previous studies of antigen-specific T cell responses in two distinct models of autoimmune tubulointerstitial nephritis (TIN), Viiia villosa lectin binding (VV+) T cells have been shown to be necessary for effector T cell expression and mediation of TIN. In anti-tubular basement membrane disease, antigen-specifi VV+ T cells direct the phenotypic selection of CD8+ nephritogenic T cells in susceptible mouse strains (J. Immunol. 141, Nov. 1,1988) . This function is mediated by an antigen-binding, I-Js+ soluble protein factor. Current studies investigate the role of the T cell glycoprotein which binds VV lectin in mediating W+ T cell function. Using the previously described effector T cell induction assay, we found that N-acetyl-d-galactosamine (GalNac) (at 25 mM but not 2.5 mM) inhibits VV+ T cell function and CD8+ effector T cell selection. When CD8+ effector T cell differentiation occurs in the presence of soluble factors derived from antigen-primed VV+ T cells, GalNAc is not inhibitory. These studies suggested that soluble Gal NAc may competitively bind to a soluble protein which stimulates VV+ T cells, in part by binding to the W lectin receptor, to synthesize andlor secrete their biologically active soluble factor. As an additional test of this hypothesis, we prepared detergent solubilized membranes from VV+ T cells and purified VV lectin binding proteins by affinity chromatography. Like GalNAc. these membrane derived lectin binding proteins also inhibit VV+ T cell function and CD8+ effector T cell selection. Inhibition by soluble 'lectin receptors' is dose dependent and is demonstrable with lectin binding glycoproteins derived from 15-20 x lo6 cells, in an assay utilizing 10 x106 VV+ cells. We are now further characterizing the lectin receptor and its endogenous ligand. Elementary bodies and outer membranes of Chlamydia trachomatis produce a high-titered IgG response in rabbits and mice as measured by ELISA and microscopic immunofluorescence assays. Western blot analysis of total elementary body protein identifies a 40kD major outer membrane protein (MOMP) as the predominant antigen. To identify the cbmical structure of the epitope, purified MOMP was subjected to chemical and enzymatic fragmentation and the resulting peptides were purified by HPLC and assayed for immunoreactivity. An immunoreactive 6kD cyanogen bromide peptide was amino-terminal sequenced and a series of overlapping synthetic peptides were synthesized and assayed for immunoreactivity. Sequential single amino acid deletions at both the NHZ and COOH termini allowed us to identify the precise epitope as a 12 amino acid peptide spanning residues 291-302 of MOMP. Two amino acid substitutions at positions 293 (PHE-GLY) and 300 (PRO-GLY) completely eliminated antibody binding. The 12-amino acid synthetic peptide is a potent immunogen producing high-titered antibody responses that are specific for the MOMP molecule.

Analysis of an independently derived mutant harboring the same defect has shorn that this trans-acting gene is not required for transport of class I1 molecules. Class I heavy chain is synthesized in this cell line and associates with B2m. Transport of the class I appears to be blocked in the ER or cis4olgi as the majority of the class I glycoproteins are not processed to the Endo B resistant form. The ability of the cell to significantly increase expression of surface class I when the incubation temperature is lowered from 37OC to 22OC suggests that this gene may function to stabilize a particular conformation of the protein. Consistent with this is the increased sensitivity of class I molecules in the mutant as compared to the parent to degradation when cell lysates are incubated at elevated temperatures. The inability to immunoprecipitate class I antigens in the mutant is possibly due to the action of endogenous proteases present in these lysates. Two complementing approaches are being employed to isolate this gene and further analyze its role in class I biosynthesis. The first involves inactivation of the trans-acting gene by insertion of a retroviral vector and subsequent PCR amplification of regions flanking the vector. In another approach a cDNA library will be introduced into the mutant cell line and the cDNA will be reisolated from cells reexpressing surface class I. Houston. TX. We have induced a panel of highly immunogenic (Imm+) vanants of the murine fibrosarcoma MCA-F using I-methyl-3niml-niuasoguanidine (MNNG). 5-aza-2'-deoxycytidine (5-azaCdR). and UV radiation. These tumors grew m immunosuppressed mice. but were complelely rejecled by normal syngeneic hosts. Mice thac had rejected large numbers of Imm+ also developed a smng, tumor. specific immunity to the parental mor. lmmunizalion with low numbers of lmm+ engendered only variant-specific immunity. The frequency of Imm+ variant g e n e d o n was similar for the three induction different protocols (64% to 92%), suggesling lhat generauon of Imm+ was more closely relalcd to the cell line used than to the inducing agent However, the swngth of the Imm+ phenaypc was related lo Ihe agent used, since MNNG induced clones had the m g e s t immunogeniciues and UV-B Ihe weakest The smng neoantigens expressed by MNNG induced Imm+ were varUnt-sppifc. while UV and S -d d R induced clones displayed significant cross-reactivilies not attributable to the parental t u n a antigen. Increased or inappropriate expression of class-l MHC antigens did not correlate with the Imm+ phenotype. We investigated the phenotypes of the spleen cells medrating tumor rejccuon using the local adopllve lransfer a s a y (LATA). Variant-specific immunity w MNNG, 5-azaCdR and UV induced Imm+ were all m d a t c d by Thy1.2+. L3T4+. LyC2.1-T cells. Afrw immunization with high numbers of Imm+ w engender both anti-lmm+ and anu-parental immunity. both CD4+ and CDW effectors rejecled the Imm+ in LATA, while only Ihe C W + T cells could wnsfer resistance w the parent Immunity 10 the parenlal tumor anugen engendered by the Imm+ suggeslcd associative recognition of the parental and neoantigem wgelher on Ihe cell surface. This hypochfsis was supported by failure of lmm+ u) pmlect againu an antigenically disunct tumor (MCA-D) admixed with it, either at the lime of immunization or at challenge. Fusion of Ihe h m * vanant with MCA-D yielded a unique, hybrid parental umor antigen that was associatively recognized with rhc original Imm+ neoanugen, demomirating the importance of antigen cocxpression. Grant RR-5511-23. W e have recently demonstrated and reported that substitution of anionic side chain carboxylic groups with aminoethylamide groups on protein antigens exhibits a pattern of enhanced immunogenicity both in vivo and in vitro. This enhanced immunogenicity was also observed in low responder strains of mice and we investigated the mechanism by which it is achieved. We examined antigen processing and presentation of native (nBSA) and modified BSA (mBSA) to T helper c e h isolated from C57/BL low responder mice. A greatly reduced amount of mBSA than nBSA was required to activate both nBSA and mBSA primed Th. Proliferation of nBSA and mBSA primed T cells increased in proportion to the amount of time of exposure of the antigen presenting cells (APC) to nBSA, peaking at 8h. Conversely, APC required less than 30 min exposure to mBSA to achieve optimal activation, indicating rapid uptake of mBSA. Paraformaldehyde fixed APC recognized mBSA without a lag phase processing, indicating that this event also occurred quite rapidly. APC processed nBSA w a s presented to primed T cells more effectively than the soluble antigen M shown by the increased rate of T cell proliferation. In contrast, mBSA was equally well presented to Th cells by APC M in soluble unprocessed form. Our data demonstrate that the reduced response in low responders is greatly enhanced by a modified antigen which is rapidly taken up and processed by APC. B cells which bear surface innunoglobulin (sIg) receptors specific for a particular antigen are abile t o present fragments of that antigen very efficiently t o T cells. This i s due. in part, t o the high affinity of the receptor, which facilitates antigen binding at low concentrations. Using TNP-ABC and specific antigen, we have demonstrated that the TNP-ABC process antigen very effectively.

W e have compared specific antigen with i t s polyclonal analog, anti-Ig, and demonstrated differences in the kinetics of degradation of anti-Ig and TNP-antigens by TNP-AEX. Both antigen and anti-Ig bound by TNP-ABC are degraded into small fragments which are released into the supernatant. However, the following differences have been found: 1) The rate of release of small fragments of TNP-antigen parallels the rate at which these cells become able to directly conjugate with T cells (a lneasure of antigen presentation), reaching a plateau between 4 and 6 hours. In contrast, the degradation of anti-Ig and release of fragnents continues for 12 hours. 2) Analysis of initial kinetics demonstrated that release of fragments of TNP-antigen begins 15 minutes after binding; there i s no significant release of anti-Ig fragrnents u n t i l about 30 minutes. 3) In contrast to anti-Ig where there i s significant accumulation of degradation intermediates within the cells, there i s very l i t t l e intracellular accumulation of intennediate-size fragments of TNP-antigen. Thus, we propose that the processing of antigen bound via specific sIg may involve a specific intracellular pathway and that intracellular routing may be determined either by the degree of cross-linking of sIg induced by antigen vs anti-Ig or the mode of interaction of the various ligands with sIg.

Alt*, Departments of Biochemistry (*) and Medicine (+), College of Physicians and Surgeons of Columbia Univerity, New York, New York 10032. We have recently analysed the structure of the 7 / 6 T cell receptor (TCR) expressed by the normal human thymocyte clone CII. CII expresses a C ' 2 constant region that is a polymorphic form lacking a copy of an izternal exon; the sequence of this constant region accounts for the size of the 7 chain and noncovalent linkage of 7 and 6 chains in the CII TCR. In order to elucidate its role, this 7/6 TCR will be reconstituted in immortalized T-cell lines. In addition, the productively rearranged human 7 / 6 receptor will be transgenically introduced into mice in order to assess the effect of the complete receptor on the development of T cells. The humoral immune response to human immunodeficiency virus has been shown to contain antibodies which act to mediate the uptake of virus through Fc receptor mediated mechanisms. It is therefore possible that vaccination with the entire envelope polypeptide may present immunologic determinants that enhance infection. One means by which to generate an immune response to HIV that shall possess neutralizing activity in the absence of infection enhancing activity is to generate anti-idiotypic Abs that bear the internal image of neutralizing human antibodies directed against HIV.

We affinity purified human antibodies from HIV+ patients on a viral lysate column. We have produced 15 monoclonal anti-idiotypic antibodies directed against these Abi's. Two of these monoclonals were shown to he Ag inhibitable by their ability to inhibit the binding of p o l y e l o n a l human antisera to HIV viral lysate on Ortho HIV Ab t,est, wells. One monoclonal, 0 B 0 , when coupled t.o KLH and used to immunize mice, produced an Abs that. bound to viral lysate in an ELISA assay. An affinity column containing RBS was used to purify an Abi that was shown to bind to p24 and p17 hy Western blot analysis.

These data suggest that 8BR may be a potential vaccine candidate. We have recently described a transgenic mouse model which co-expresses the TCR u and fl chains from the 2C cell line (recognized by the 1B2 anti-clonotype). T celh bearing the transgenic clonotype are positively selected by elements of the H-Zb MHC for expression on CD8' cells. Thus in the periphery of H-Zb animals 40-80% of the T cells are 1BZ1/CD8*. The same peripheral expression is observed when the transgenes are expressed in F1 animals bearing a "neutral" MHC haplotype (eg. H-Zb'*). However, when the transgene hi expressed in F1 animals which also express the H-2Ld gene product, negative selection occurs by clonal deletion. However, this deletion is functional rather than structural as the 1B2 clonotype is present on 10-40% of peripheral T cells. These cells are unusual in that they express neither of the characteristic peripheral molecules CD4 or CD8. The absence of CD8 expression on the 1B2* cells appears to allow these potentially self-reactive clones to exist without evidence of autoimmunity. The original clone as well as 1B2+/CD8+ cells from H-2b animals are strongly inhibited by anti-CD8 reagents.

In an effort to understand the process of negative selection and self-tolerance we have examined the capacity of these cells to be activated directly by the anti-clonotype rather than antigen (H-2Ld). The results demonstrate that the clonotype is fully functional on these double negative cells, indicating a normal maturation in the thymus. Further examination of their surface phenotype also supports the conclusion that these are fully mature cells which are phenotypically distinct from double negative cells which exist in the thymus of H-2b animals. of Imunohematology. AZL, Leiden, the Netherlands;2Praxis Biologics, Rochester, New York, USA and 'University of Southampton, UK. Immunity to disease caused by Neisseria menineitidis is associated with the presence of bactericidal and opsonic antibodies to the capsular polysaccharide (CPS), lipopolysaccharide and to outer membrane proteins (OMPs). The CPS of group A and C meningococci are proven efficacious vaccines, although the immunogenicity in infants is poor and the immunity is of short duration. The combination with T-helper epitopes will certainly improve the immunogenic properties of these T-independent (Ti,) antigens. The group B CPS is poorly immunogenic in humans probably because of tolerance due to structural similarity to host glycopeptides and/or glycolipids. We have focused our research onto the class 1 OMPs which show limited heterogeneity amongst meningococci. Murine monoclonal antibodies to these proteins are highly bactericidal in vitro and will be used to map B-cell epitopes. T-epitopes have been identified by theoretical prediction of immunodominant sites by analysis of the amino acid sequence of the OMP followed by their solid phase synthesis and subsequent testing for polyclonal activation of T-lymphocytes obtained from HL4-typed volunteers immunized with the OMP. In addition human T-cell clones are generated with OMPs and maintained with antigen, EBV-transformed B-cells, fresh feeders and rIL2. The clones are tested for antigen specificity, io vitro helper function, MHC restriction element, expression of surface markers and recognition of common meningococcal T-cell epitopes.

C 642 Demonstration o f p-azobenzene-arsonate-L-tyrosine (ABA-Tyr) Speclfic T cells in Low Responder H-26 MIce by IL-1 Supported T Cell Proliferatlon

Previous studies have shown that H-Zb mice immunized with ABA-Tyr fail to produce ABA specific delayed-type hypersensitivity and show little or no T cell proliferation in vitro to ABA-Tyr. These observations suggest that H-Zb mice are deficient in TH1 cells that respond to ABA-Tyr. By contrast, immunization of H-2b mice with TNP conjugates of ABA-Tyr revealed good cognate help, suggesting that these mice do possess ABA-Tyr specific TH2 cells and that such cells are not revealed in conventional lymphoproliferative assays. Because such assays are widely used to evaluate Ir gene control and to map T cell epitopes, the databases generated from such studies may seriously under represent the total number of responder phenotypes and T cell epitopes. Because of this concern, we established culture conditions that wlii support ABA-Tyr specific T cell proliferation in H-2b mice. In these studies, C57BU6.l mice were immunized S.C. with ABA-Tyr and 7 to 14 days later the draining lymph nodes were cultured with varying doses of ABA-Tyr or with varying doses of ABA-Tyr and varying doses of recombinant IL-1 alpha (rlL-la), a known costimulator of TH2 cells. Culture with ABA-Tyr alone produced no proliferation. By contrast, culture with ABA-Tyr and rlL-la revealed T cell proliferation that titrated with the dose of ABA-Tyr and the dose of rlL-la

specific for conalbumin presented on ryngeneic antigen presenting cells and dependent on IL-1 for its proliferation, was used a s an indicator cell for the ability of neonatal murine spleen cells to present antigen and produce IL-1 and IL-2.The antigen presenting capacity of neonatal spleen cells is low. During antigen presentation there is an augmentation of IL-1 and IL-2 production by the antigen presenting spleen cell population. However, neonatal spleen cells do not respond as well a s adult cells. The low levels o f IL-1 can not be attributed t o a low potential for producing IL-1 since neonatal cells produce high levels of IL-1 after induction by a crude IL-1 Inducer Factor (IL-1-IF).The

This impairment leads t o a decreased stimulus of the 1-helper cell to produce inducer factors which leads t o low levels of IL-1 and IL-2 production by the neonatal cells during antigen presentation. No suppressor mechanisms responsible for the l o w interleukin production were detected. Human or murine class I genomic DNA was transfected into a B-lymphohlastoid x T-lymphoblastoid hybrid cell line. This fusion hybrid has lost both T cell derived copies of chromosome six and contains deletions spanning the class I1 region on both copies of chromosome six derived from the B-cell parent. Previous data have described a transacting factor within this region that is responsible for class I antigen expression. HLA-Bw58 and B7 glycoproteins, although synthesized, were not transported to the plasma membrane in the hybrid.

were surface expressed. These data suggest a fundamental difference between human and mouse histocompatibility antigens in their requirements for intracellular transport. The role of glycans in this transport dicotomy is currently under investigation. In addition, hybrid human-murine genes are being used to identify regions of the class I molecules involved in this transport phenomenon. We probed t h e means by which t h e a n t l g e n p r e s e n t l n g c e l l (APC) handles t h e a n t i g e n produce p e p t i d e s t h a t bind t o MHC-molecules. We propose t h e e x i s t a n c e o f a new type o f i n t e r n a l image i n which immunoglobulin v-region peptides. formed by processing, imitate peptides from conventional a n t i g e n s . W e r e f e r t o such denatured i n t e r n a l images as r e s i d u e internal images, since t h e y a r e a s s o c i a t e d w i t h t h e r e s i d u e of p e p t i d e s remaining a f t e r processing. In some cases, r e s i d u e internal images may be actual sequence images, i . e . , the v-region sequence m y be i d e n t i c a l t o the conventional a n t i g e n sequence.To be Class I H-2Ka-restricted cytolytic T lymphocytes (CTL) are directed against two immunodominant sites on the A/JAP/305/57 influenza hemagglutinin (HA) that can be mimicked by synthetic oligopeptides spanning residues 202-221 in the HA1 and 523-545 in the hydrophobic, transmembrane region. Analysis of the fine specificity of HAl-specific CTL clones demonstrated that these CTL clones can be subdivided into at least two group based on their patterns of recognition of closely related influenza H2N2 field strains and a monoclonal antibody derived variant of A/Guiyang/4/57. Using a series of nested synthetic peptides spanning the 202-221 region, the minimal amino acid residues necessary for recognition by the two groups of CTL clones were defined and found to consist of two separate but overlapping sites. Sequence comparison of the HA of the A/JAP/305/57, the influenza field isolates and the monoclonal antibody derived variant has identified two amino acids, Asn at position 207 and Gly at position 215, that are critical for T cell recognition. Thus, animo acid substitutions induced either by antigenic drift or by monoclonal antibody selection can affect class I CTL recognition. pretreatment with a n t m e s reactive with Class 11 MHC anti-has previously been reported to be successful in EXOV~~KJ a n t i g e n -W i r q dendritic cells (m) fran rodent tissue grafts. We have extended these exper-to inta? whole organ grafts. ILIA pnmxses also demcnstrated a prolonged survival (17 f 2 days) ccapared to controls (11 t 1 days) (P < 0.01) tihen v l a n t e d into streprozatocin treat& DA recipients. Antigenic variation in the haemagglutinin (HA) of influenza A viruses frequently introduces new OligOsaCChEKide attachment sites ( Aan-X-SerlThreo) and carbohydrate addition prevents antibody recognition by steric hindrance. acid substitution in mutant viruses of the H3N2 subtype (HA1 63 Asp+Asn), that introduces an N-glycosylation site (hn6gCys6&Thr65), abrogates antibody and CD4+ T recognition. infected with X31 virus recognise a synthetic peptide corresponding to antigenic site E, HA1 56-76, and are sensitive to a single substitution (HA1 63 Asp-bAsn) in mutant viruses. virus infected target cells, thereby confirming that carbohydrate addition prevents CD4' T cell recognition.

Here ve show that an amino cell I-Ad restricted, HA specific T cell clones f r m BALBlc mice-reviously Recognition of mutant viruses is restored however by tunicamycin-treatment of