key: cord-0041824-5junebrl
authors: nan
title: Molecular pathways of cytokine action
date: 2004-02-19
journal: J Cell Biochem
DOI: 10.1002/jcb.240440602
sha: 176af69f5a34f7b9a85e9d423d72e07cb22824e4
doc_id: 41824
cord_uid: 5junebrl

nan

can induce DNA synthesis in isolated, quiescent nuclei, whereas extracts prepared from resting lymphocytes fail to do so. This activity is mediated through the action of a cytoplasmic protease (ADR). ADR activity can be induced in resting cytoplasmic extracts, in a dose-dependent fashion, by brief incubation with a membrane-enriched fraction from spontaneously proliferating MOLT-4 cells, suggesting that ADR is already present in the resting cytoplasm in a inactive or precursor form. This precursor or inactive ADR is precipitated between 30 and SO% ammonium sulfate saturation, similar to active ADR isolated from dividing lymphocytes. Furthermore, ADR activity induced in resting cytoplasmic extracts by MOLT-4 membranes can be inhibited by the protease inhibitor, aprotinin, suggesting that this induced ADR has a similar mechanism of action to the factor normally present in cytoplasmic extracts from actively dividing lymphoid cells. Thus, the induced ADR activity appears similar to or identical with "spontaneous" ADR activity. The interaction between MOLT-4 membranes and resting cytoplasm that generates ADR activity appears to be dependent upon protein kinase activity, as this reaction is inhibited, in a dose-dependent manner, by H7, which is an inhibitor of CAMP-dependent protein kinases and protein kinase C, but not by a synthetic peptide, IP-20, which appears to be a specific inhibitor of CAMPdependent protein kinases. These results suggest that a membrane-associated protein kinase C participates in the conversion of an inactive or precursor form of ADR already present in the resting cell cytoplasm to one that is capable of inducing nuclear DNA synthesis. The EGF receptor belongs to a class of growth regulatory proteins in which a single membrane-spanning segment separates the extracellular (ligand-binding) domain from the intracellular (tyrosine kinase) domain. We have studied the expression and stability characteristics of two forms of receptor mRNAa 5.6 kb form that encodes for the intact transmembrane 170 kDa receptor and a smaller 2.6 kb form that encodes for a 100 kDa secreted EGF-receptor. The latter is identical (u to nucleotide 2079) to the normal intact EGFreceptor mRNA of 5. 6 kb, but it diverges downstream in t i e 3' region, and lacks the regions encoding for transmembrane spanning and cytoplasmic kinase domains. The studies indicate that synthesis and stability of EGF-receptor ene transcri ts are stimulated by EGF and cycloheximide. The stability of the 5.6 kb transcript IS enlanced by E 8 F and cloheximide in a non-additive manner, suggesting certain common features in their pathways of action.%e smaller 2.6 kb mRNA shows higher basal stability in corn arison with the 5.6 kb enti . It is further stabilized by cycloheximide, especially in combination with EGgacting in a synergistic manner.%he results imply the involvement of 3' located structural features in the control of receptor mRNA stability, and suggest the involvement of labile proteins in the regulation of EGF-receptor mRNA expression Next we tested whether EGF influenced receptor expression at the level of post-translational maturation and externalization. We used the 100 kDa EGF-receptor as the model. This receptor is identical to the external domain of the transmembrane EGF-receptor in its polypeptide structure, glycosylation state and conformation; but because it lacks a membrane anchor, biosynthetic transport and externalization of this soluble receptor is easier to monitor than that of the membrane-bound receptor, and is moreover uncomplicated by any possibility of internalization. The studies indicate that exocytosis of the 100 kDa protein is constitutive, but slow, and de endent upon events within endo lasmic reticulum (ER) that are rate-determining from the viewpoint o8ransport. One of the ER-locateieventscore N-linked glycosylation although essential for transport, occurs soon after the translation of the protein. However, a subsequent slow event(s) controls its maturation, processing, exit from ER and entry into golgi. The rate determining step may be regulated by the ligand EGF, which is now found to have an accelerating effect upon receptor maturation, transport and secretion. This EGF-induced acceleration of biosynthetic transport occurs independent of ongoing translation. Lymphotoxin (LT; TNF p) and tumor necrosis factor (TNF a) are functionally and genetically related molecules. Though originally descibed a s cytotoxic factors, they can affect target cells in several different ways including killing, induction of proliferation, and induction of differentiation. LT and TNF's effects are dictated in large measure by the differentiated state of the target cell itself. LT and TNF can induce release of target cell DNA. This induction of DNA fragmentalion termed apoptosis or programmed cell death is characteristic of many differentiating systems including digit formation, resorption of the tadpole tail, and glucocoflicoid and antiLCD3 induction of thymocyte suicide. The role and mechanism of LT and TNF in several of these systems is under investigation. It has been postulated that the laddered pattern of DNA isolated from cells undergoing apoptosis is due to fragmentation at internucleosomal intervals suggesting the possibility that histone proteins might be released in association with the DNA fragments. We have determined that the kinetics of histone release from nuclei of cells undergoing apoptosis corresponds to the kinetics of DNA release. The size of these fragments suggests that the DNA and histones are associated in the form of intact chromatin. We have determined that LT and TNF are not themselves endoncleases but must activate or release cellular endonucleases perhaps through their effects on lysosomes. The effect and mechanism of LT and TNF induction of lymphocyte differentiation is under investigation They may act as autocrine and paracrine growth factors or as inducers of suicide. A variety of murine B cell lines representing stages of B cell differentiation from fetal liver to plasrnacytornas have been studied. LT and TNF expression is associated with early B cells which expand by responding to proliferating signals or may undergo programmed cell death In the case of inappropriate gene rearrangements.

Institute, Boston, MA 02115

By sequence comparison, GRO can be placed in a family of genes which encode secretory proteins involved in the inflammatory response. Previous studies of GRO in fibroblasts have shown that it is an early response gene. GRO mRNA is elevated by serum, shortly after fos but before E , when serum-starved cells enter the cell cycle. GRO mRNA is also elevated in starved or growing cells in response to TNF or IL-1, suggesting that GRO has a r o l e in cytokine signalling.

New findings support and extend these studies.

In order to determine other functions of GRO, recombinant protein has been produced through harvest of conditioned medium from cos cells transiently transfected with a plasmid containing GRO cDNA. A purification scheme has been developed which allows f o r production of homogeneous protein. The recombinant protein has an apparent molecular weight of 6.5 kDa, in close agreement with the predicted molecular weight. Current studies demonstrating the biological activity of recombinant GRO protein will be presented. In addition, a polyclonal antibody has been raised against a peptide sequence of GRO and has been shown to be specific for the GRO protein. It is capable of recognizing GRO in Western analysis, immunoprecipitation, and RIA. Using these three assay systems it is possible to monitor GRO protein production as an adjunct to expression levels noted by Northern analysis.

CAT-assays and gel retardation analysis shows that the TNF and IL-1 responses are mediated by the NFkB binding site located at -65 to -75 5 ' to the GRO coding region. A protein which is induced by IL-1 and TNF is capable of binding to a fragment containing the NFkB site. this binding is eliminated by a 72 base pair SV40 DNA containing the NFkB sequence. Competition experiments with oligonucleotides containing wildtype o r mutated NFkB sites are in progress.

In competition experiments

EXPRESSION, William L. Farrar', Gonzalo Garcia Garcia', Gerald A. Evans', Dennis F. Michiel' and Diana Linnekin', Cytokine Mechanisms Section, Laboratory of Molecular Immunoregulation, Biological Response Modifiers Program, National Cancer Institute, Frederick, Maryland 21701-1013 and 'BCDP, Program Resources, Inc., NCI-FCRF, Frederick, Maryland 21701-1013. Proliferation and differentiation of lymphohematopoiesis and the regulation of the mature immune system is under the exquisite control of a variety of polypeptic factors collectively referred to as cytokines. With the exception of CSF-l/c-&, all the cytokine receptors cloned to date have no cytoplasmic catalytic domains, thus separating them from previously described growth factor receptors such as those found for insulin, EGF, PDGF, etc. W e have examined the protein kinase coupling mechanisms and cellular substrates for a number of lymphohematopoietic growth factors including IL 2 , IL 3 and GM-CSF. Each factor possesses a unique receptor but nevertheless, trigger many similar biochemical events and program of gene expression. All the proliferative cytokines a p p~a r to regulate a similar pattern of gene expression including, c-fos, c-myc, c-myb, ODC. cyclin, IL 2R alpha, and some members of the heat shock gene family

The cytokines appear to regulate both serine and tyrosine kinase activity. One substrate at 68 kDa is rapidly phosphorylated on serine by all mitogenic cytokines tested. The p68 protein is highly conserved and appears central to the proliferative response Human IL 3 and GM-CSF stimulate the catalytic activity and autophosphorylation of two tyrosine kinases at 140 and 92 kDa. The 9 2 kDa tyrosine kinase also appears to be regulated by the 11. 2R beta chain suggesting a conservation between myeloid and lymphoid cell types. The data suggests that c e r t a i n cytokine receptors may use generic kinases common to several receptors in mediating the signal transduction process. The adherence of inflammatory or neoplastic cells to endothelium is recognized as an event which appears to be mediated by molecules on the circulating cells which recognize specific sites on endothelid cells. The adherence of lymphocytes, macrophages, neutrophils and mast cells to endothelium has been identified as important in the generation of inflammatory reactions. The identification of the factor(s) which causes adhesion to occur has been approached in several ways: molecules on the surface of the circulating cell involved in the process have been identified as well as molecules on the endothelid cell surface. Some of the former have been shown to be specific to a cell type while others are shared by several types. Studies of endothelium have suggested that some of these cells express organ-specific ligands on their surface. Other studies have focused on conditions in the local microenvironment which result in increased binding. Of panicular interest in this regard is the role played by cytokines in cell binding phenomena.

It has also been suggested that the first critical step in metastasis is arrest and adhesion of the circulating tumor cell in the microvasculature followed by exuavasation. It appears that this process involves mechanisms similar to those operative in normal inflammatory cell-endothelid cell interactions. We have therefore studied the role of cell adhesion in tumor cell metastasis. We have found that adherence to endothelium is mediated by glycoproteins on the surface of the tumor cell and that binding is cation-dependent, To purify adhesion molecules on the tumor cell we utilized methods that had been successful in identifying adhesion molecules on neural and liver cells. In preliminary studies we have been able to achieve approximately 450-fold purification of adhesion molecules from membrane preparations of tumor cells. In other experiments we incubated endothelium with cytokines and found increased numbers of neoplastic cells bound to endothelial monolayers which had been incubated with TNF, LPS, or PMA. Unlike the situation for inflammatory cells, IL-I and E N -7 produced no changes in binding. In addition, exposure of tumor cells to these factors prior to incubation with endothelium had no effect. This suggests that there is an active process by which surface molecules on endothelial cells may be induced (or normal molecules increased) to mediate adherence of neoplastic cells and that some cytokines may play an important role in their expression. Whether these smctures are identical to those involved in inflammatory cell adhesion remains to be determined.

Smith Charitable Trust. Some of this work was supported by NIH grants CA-32319 and CA-39723 and a grant from the W. W.

Rosenman, Joey Meyer, and Thomas St. John, Division of Basic Sciences, Fred Hutchinson Cancer Research Center. Seattle. WA 98104. Many cells of hemopoietic origin. particularly lymphocytes, move continuously between the various lymphoid tissues to reach specialized n&?roenvironmentS presented by each organ. Transmigration across post-capillary venules in organized lymphoid organs and at rites of inflammation is a multistep process that can be divided conceptually into (1) attachment to endothelial cells (EC), (2) lateral movement on t h e surface of EC, (3) transmigration between EC, (4) penetration of the basement membrane, ( 5 ) release, and (6) intraorgan sorting into appropriate niicroenvironments (e.g. T vs. B). The initial phase of this movement, attachment, is orchestrated by organ-specific adhesion of the trafficking cells to the endothelial lining of the niicrovasculature at each site. Migration competence is modulated as a function of lymphocyte activation by varied expression of at least three distinct classes of lymphocyte cell adhesion molecules (CAMS): integrins (e.g. LFA-I), the CD44 glycoproteins, and selectins (e.g. Mel-14, LAM-I). For example, lack of h?el-14/LAM-l on IL-2 dependent, CD8+ cytotoxic T cells propagated invitro, may contribute to poor recirculation and limit therapeutic utility vivo. We have used a primate model to analyze the structure and function of both CD44 and (LAM-1) type CAMs. In macaques expression of these structures by normal PBL correlated with their ability recirculate and to adhere to lymphokine induced cuitured EC in. T cells could be placed in a maturational sequence of noncycling virgin, actively cycling, and resting memory cells according to expression of these receptors. These stages of activation/iilaturation were anatomically partitioned invivo. High-level CD44 expression was also correiated with a n increased responsiveness to mitogenic stimuli transmitted via CD3. This may involve a direct contribution of CD44 to the activation cascade. To facilitate structure/function analyses of CD44 in each of these processes and to test the relative contributions of CD44 and Mel-14/LAM-I type CAMs to EC binding, we isolated cDNAs encoding the macaque homologues of each. In both cases cross-species comparison of primate, human, and murine genes reveals a high degree of conservation. Functional analyses of cells expressing CD44 or macaque LAM-1 (LAM-l,,,) g e novo following plasmid and retrovirus mediated gene transfer indicate that CD44 expression can promote cell-cell adhesion in other contexts and may serve as a facilitating CAM for lymphocyte-EC atfacbment. In contrast, a more direct functional correlation with LAM-I,,, expression exists.

Constitutive expression of the latter in 1L-2 dependent T cell clones may provide a means to enhance recirculation and effector function in a clinical setting. The sites of distant metastases in many clinical cancers and animal tumors are nonrandom, and their distributions cannot be rationalized by anatomical or mechanical hypotheses Using animal models for metastasis based on murine melanoma, large cell lymphoma and fibrosarcoma and rat mammary adenocarcinoma we have examined tumor cell and host organ properties, such a s differential tumor cell adhesion to organ-derived microvessel endothelial cells or their subendothelial basement membrane-like matrix and organ parenchymal cells, dlflerential tumor cell invasion of host organ tissues and extracellular matrix and their expression of degradative enzymes, and differential tumor cell responses to organ-derived growth-stirnulatory and growth-inhibitory factors. These appear to be collectively important in determining the organ specificity of metastasis. In the interactions of tumor cells with microvessel endothelial ceils we have identified several tumor cell and endothelial cell surface components involved in organ-preference of adhesion These include endogenous lectins, integrin-like, calcium-dependent and other adhesion molecules Malignant cells possess another set of cell surface adhesion components for basement membranes, including receptors for fibronectin. laminin. collagen. heparan sulfate pfoteoglycan, and other components. During invasion metastatic cells release and display on their cell surfaces specific degradative enzymes, such as metaloproteases, proteinases and endoglycosidases, that are used to dissolve basement membranes and other structures. Finally, we have found that highly metastatic cells differentially respond to organ-derived. secreted, paracrine growth factors which are present in dissimilar amounts in different organs. One of the most potent of these paracrine growth factors for lung-colonizing epithelial tumor cells has been purified to homogeneity and shown to be a unique Mr-66,D00 glycoprotein that has been isolated from rodent and porcine lungs and binds to a specific receptor on lung-metastasizing tumor cells Another differentially expressed growth factor that stimulates highly metastatic tumor cells is transferrin. These and other tumor cell surface and host properties may eventually be used to predict and expiain the unique metastatic distributions of certain malignancies [Supported by NCI grant R35-CA443521 (joint)   CD013   THE ROLE OF HLA MOLECULES I N THE SUSCEPTIBILITY OF TUMORS TO   NATURAL KILLING , J e f f r e y R. Dawson, Walter J. S t o r k u s , J e f f r e y Alexander, J. Alan Payne and P e t e r C r e s s w e l l , Department of Microbiology and Immunology, Duke U n i v e r s i t y Medical C e n t e r , Durham, NC 27710, USA. The c y t o t o x i c i t y mediated by t h e n a t u r a l k i l l e r ( N K ) c e l l is n o t r e s t r i c t e d by class I major h i s t o c o m p a t i b i l i t y complex (MHC) a n t i g e n s ; t a r g e t c e l l s may be l y s e d a c r o s s a l l o g e n e i c and, i n some i n s t a n c e s , xenog e n e i c b a r r i e r s . Recent r e p o r t s s u g g e s t t h a t reduced t a r g e t c e l l class I e x p r e s s i o n may c o i n c i d e w i t h an i n c r e a s e i n s e n s i t i v i t y t o NK. T h i s hypot h e s i s was t e s t e d i n d i r e c t l y by examining a s e r i e s of r e l a t e d lymphoblast o i d c e l l l i n e s d i f f e r i n g i n t h e e x p r e s s i o n of c l a s s I HLA a n t i g e n s ; sens i t i v i t y of t h e s e c e l l l i n e s t o NK was i n v e r s e l y p r o p o r t i o n a l t o t h e c e l l s u r f a c e e x p r e s s i o n of class I a n t i g e n s . The e x p r e s s i o n of c l a s s I HLA a n t i g Natural killer (NK) cells and killer (K) cells mediating ADCC have been shown to be a lymphocyte subset designated large granular lymphocytes (LGLa). The mechanism of cytotoxicity by LGLs can be divided into two diitinguishable stages. involves recognition and binding to target cella, while the second includes post-binding events that lead to target cell lysis. Studies in thii laboratory are proceeding to define the receptors and structures involved in NK recognition. anti-idiotypic antibody (anti-ID) against a monoclonal antibody that blocks LCL binding anticipating that such an anti-idiotypic antibody might recognize the NK receptor and aid i n its identification. This anti-ID antibody is reactive with 85-100 and 150 kD proteins and blocks LCL binding and NK target cell lyiia. F(ab')2 anti-ID antibody significantly enhanced lysis of target cells and production of interferon-7 by LGL suggesting triggering through this molecule that results in intracellular signalling events. receptor and its cloning are in progress. factors, such a8 pore forming protein (PFP) and natural killer cytotoxic factor (NKCF), in this cell-mediated lysis has been examined to gain an understanding of the mechanism of lysis. We have analyzed PFP mRNA expression in resting and stimulated human peripheral blood CD3-LGL, CD3' T cells and their CD4' or CDBt subsets. occurred in 2-4 hours, with peak levels occurring after 6 hours in the absence of DNA o r protein synthesis. After stimulation of CD3' lymphocytes with IL-2, the CD8' T cell subset was predominantly induced to express PFP mRNA. antibodies to the anti-p75 IL-2 receptor indicated that It-2 signaling was via the p75 IL-2 receptor. The cytotoxic potential of peripheral blood T cells and LGL induced in response to IL-2 correlated with IL-2 induced PFP mRNA levels in these cells and was consistent with PFP being one of several important molecules involved in the effector function of cytotoxic lymphocytes. Using monoclonal antibodies we have demonstrated NKCF to be distinct from recombinant lymphotoxin, tumor necrosis factor, and leukoregulin. Elucidation of the site and mechanism of action of these lytic factors should expedite our understanding of NK cell-mediated killing. Macrophage activation for tumor cell killing is a multiphase reaction sequence in which responsive macrophages require sequential interaction with multiple stimuli for the development of full lytic capability. Interferon gamma (IFN-y) has been recognized as the prototypic activating factor responsible for initiation of this complex pathway and, in combination with bacterial lipopolysaccharide (LPS), this molecule can stimulate the development of macrophage anti-tumor activity. The multi-step nature of this pathway suggests the induction of this functional activity requires the completion of a series of strictly regulated biochemical and molecular events. However, investigation of these mechanisms has been hindered by the lack of objective markers for identification of intermediate stages i n the activation pathway. Our experimental approach to the analysis of the process of activation for tumor cell killing involves the use of murine macrophage cell lines as model systems for the study of normal macrophage function. In our recent work, we have demonstrated that, unlike heterogeneous populations of normal cells, a given macrophage cell line will express a distinct subset of functional activities in response to a defined activating signal such as IFN-y. Thus, the RAW 264.7 cell line exhibits a strict requirement for interaction with IFN-y and LPS for the development of tumor cytolytic activity. In contrast, the WEHI-3 cell line, although IFN-responsive as assessed by the induction of MHC antigen expression, does not develop cytolytic activity. The identification of homogeneous macrophage populations capable of differential cellular responses to IFN-y provides a novel system for analysis of the mosaic of changes associated with the activation process. As a first step toward understanding that process at the molecular level, we have produced monoclonal antibody probes which identify cell surface protein changes occurring in macrophages during activation. We also are using a variety of molecular techniques to detect and clone gene products that arc uniquely expressed in cells activated by IF"-1, in order to assess the molccular profile of macrophages activated for particular functions, Finally, our third approach to the analysis of mscrophage activation has been through the identificatian of an alternative activation slinlulus.

gamma radiation, which has many effects on macrophage function similar to those initiatcd by IFN-y. Our efforts in this area are centered on the identification of signal transduction rncchanisms and protein or molecular genetic changes expressed in common by cells wtivatcd by thcse distinct stimuli. These combined approaches should allow disscctian of thc funclional rolcs of gencs and proteins induced in macrophages activated for tumor cell killing. Cytotoxic lymphokinss such as TNF, LT and IFN-7 produced by T Cells participate at several steps in the molecular pathway leading to target cell destruction including CTL differentiation, enhancement of target cell recognition, and destruction of the target cell. Membrane receptors Specific for TNF/LT initiate the biological activities of these proteins. Specific receptors are expressed following T cell activation where they appear to initiate selected growth and differentiation functions in T cells. In a model system the receptor for TNF expressed on the 11-23 T cell hybridoma functioned to induce MHC class I expression and was -80 kDa. LT was unable to enhance MHC expression and functioned as a partial agonist indicating these two related cytokines have different functions. Delivery of these toxins to the target cell is controlled by their biosynthetic pathway in CTL. TNF synthesis in CTL is rapidly induced following contact with a triggering signal and follows a constitutive-type pathway as compared to the regulated synthesis of perforin and esterases which accumulate in granules. Using Con A as a reversible triggering signal the initial transcription of TNF mRNA and secretion of TNF required the continuous presence of the triggering signal to maintain synthesis and secretion. Within the realm of the CTL-target interaction these results suggest that TNF can be delivered to the target cell in a directed and precise fashion helping to maintain specificity. Disengagement of the CTL from the target would cease to generate the membrane signal, halting TNF gene transcription and secretion, sparing the release of the toxin onto bystander cells. Target cell susceptibility to the cytolytic activity of LT and TNF depends on several factors including the expression of specific receptors and the development of a resistant state. TNF-receptor interaction induces DNA fragmentation in target cells while concurrently acting as a growth factor to accelerate target cell progression through G! when cell death is manifested at G2/M. These results indicate that the pleiotropic activities of TNF/LT are integrated at several points in the CTL lytiC pathway to acheive target cell lysis. Supported by NIH grant CA35638. Viral infections of the CNS provide a number of excellent experimental models of demyelinating disease. Among the most widely studied of these are Theiler's murine picornavirus, lentiviruses of domestic animals and man, Semliki Forest vlrus, and the murine coronaviruses. We have studied demyelinating disease induced by mutants of the murine coronavirus, MHV-4, in an effort to understand the molecular determinants of disease pathogenesis.

Neuroattenuated, demyelinating mutants of MHV-4 were selected for resistance to monoclonal antibodies against the malor spike glycoprotein, S . These mutants and their wlld type parents have been characterized by northern and western blotting and direct RNA sequencing.

Neuroattenuated mutants were shown to Ezntain large deletions of over 450 nucleotides in the S gene sequence, and these deletions were consistently mapped to the N-terminal domain,of the S polypeptide.

Replication of the mutant and wild type viruses was compared in culture and in the mouse brain, and it was observed that,the deletion mutants grew more efficiently in culture but not in vivo. Sepences and biological properties of S proteins of additional MHV-4 point and deletion mutants as well as other MHV strains have been compared in an effort to understand the molecular and biologic basis for the observed,polymorphism in S.

Using in situ hybridization we have studied the trafficking of virus during infection and find that in contrast to wild type MHV-4 which causes massive early infection of gray matter tracts, the neuroattenuated mutants only mildly affect gray matter, but spread avidly in white matter tracts. Animals surviving chronic infection with the variant viruses show evidence of viral genome in cells in the white matter and show persistent chronic demyelination and remyelination.

We are currently studying the state of the persistent viral genome as well as activation of host MHC and myelin associated genes in these persistently infected mice in an effort to understand the pathogenic mechanisms.

Hayes.* Monica Summers,* Sunil P. Sreedharan, Division of Allergy and Immunology, University of California Medical Center, Institute, San Francisco, CA 94143-0724; *Department of Pediatrics and Immunology, Ohio State University, Children's Hospital, Columbus, Ohio 4 3 2 0 5 . Neuropeptides and structural variants of neuropeptides are generated by diverse cells of the immune system, and affect proliferative and synthetic functions of T-and B-lymphocytes. eosinophils generate substance P (SP) and VIP,.,, identical to those in the neuroendocrine system. In contrast, other immune cells produce variants of neuropeptides, such as the VIPc-,,-,, and VIP,,.,, derived from mast cells through alternative genetic mechanisms, and the VIP,-z, and VIP,,.,, which come from lymphocytes through post-translational peptidol sis. Some blood lymphocytes and cultured lines of B-and T-lymphocytes express 2 x 10 -40 x lo3 receptors for VIP with binding constants of 0.5-13 nM. School of Mediclne, Los Angel-s, CA 90024 While Interactlons between tt central nervous system (CNS) a n d immune system have been documented over 100 years ago, the molecular biological, biochemical, and anatanical bases for these interactions have been elucidated only in t h e l a s t decade. The function of shared receptors a n d t h e i r ligands in the CNS and immune systems can be examined by i l l u s t r a t i n g s i m i l a r i t i e s and differences i n the biological events I n t h e two systems resulting f r a signal transduction or cell-cell Interactlons. As examples, interleukin 1 ( I L l ) , tumor necrosls factor alpha (TNFg), prostaglandin E (PGE), and interleukin 2 ( I L 2 ) are immunanodulators which are found in normal brain and may play a role in disease s t a t e s subsequent t o increased production either by inflammatory Cells or resident glial c e l l s and an upregulation of t h e i r receptors. IL1. TNFo(, a n d PGE may be c r i t l c a l in regulating neurophysiological and inflammatory events in t h e CNS of multiple sclerosis (MS) and AIDS patlents. IL2 and IL2 receptors are prominent in brelns of patients with a variety of neurological diseases including MS and Alzheimer's disease a n d have recently been detected in normal human, r a t , and mouse brain white matter and grey matter. I L 2 affects a variety of brain cell types, both gl la1 and neuronal. and may have an enaogenous counterpart in the CNS. monokines, and t h e i r endogenous brain-derlved analogues will elucidate t h e i r particlpatlon I n the h a e o s t a t l c response of brain c e l l s t o injury and abnormallties resulting from chronlc Infections, tumors, and autoimmune disease of the CNS. In addition t o the TCF/CD3 complex activation signals can also be transduced via CD2 and CD16.

(Xlr recent data also demonstrate that the lymphocyte function associated antigen-1 (LFA-1) coactivates the TCR complex, and under physiolgical conditions the LFA-1fiCAM-1 interaction is a prerequisite which however can be circumvented by the use of bifunctional &s.

Using bifunctional d s , which recognize W C D 3 and renal c e l l or ovarian c e l l carcinoma associated antigens, we have shown that t h i s retargeting of CPL represents a powerful system for adoptive i m o t h e r a p y of these cancers. Experiments were performed t o investigate the recycling capacity of bifunctional mAb retargeted (cloned) Cn. requires the addition of "fresh" bifunctional d s . Cytokines have great potential for the development of new approaches to improve drug targeting and pharmacologic action. Leukoregulin is a lymphokine that selectively increases the permeability of many tumor cells. The kinetics of membrane destabilization and high degree of specificity for tumor cells suggest that leukoregulin treatment might be useful for increasing the intracellular entry o f pharmacologically active molecules into leukoregulin-sensitive cells. The ability of leukoregulin to facilitate the uptake of tumor inhibitory drugs, therefore, was assessed by flow cytometry measurement of the entry of the fluorescent anthracycline doxorubicin and several other antitumor antibiotics into leukoregulin-treated human K562 erythroleukemia cells. K562 cells were exposed to 0.25-0.5 pg of the metabolic inhibitors/ml for up to 60 minutes at 37%. Commencing within 15 minutes, leukoregulin increased the entry of doxorubicin approximately two-fold and the uptake of mitomycin, mithrarnycin, and propidium iodide two-fold to ten-fold. The P value for the 1.9-fold increase in uptake of doxorubicin in the presence of 1 unit of leukoregulin was <0.05 and <0.005 for the 2.3-fold increase produced by 4 units of leukoregulin. A leukoregulin concentration-dependent enhancement of doxorubicin uptake was also obtained in U937 histiocytic lymphoma and in RPMl 2650 squamous carcinoma cells. Neither interferon, tumor necrosis factor, IL-I, nor 11-2 alone or in combination influenced doxorubicin uptake Leukoregulin enhancement of doxorubicin uptake was also specific for tumor cells in that no significant increase was observed in leukoregulin-treated, freshly isolated human mononuclear leukocytes No significant increase in doxorubicin uptake was observed in leukoregulin-treated proliferating lymphocytes activated by 3-day exposure to mitogenic monoclonal antibody to the CD3 epitope or by culture for 7 days in the presence of interleukin-2. A synergistic decrease in proliferation was observed in K562 cells treated for 30 minutes with leukoregulin and doxorubicin in comparison to cells treated with either agent alone. This indicates that leukoregulin enhances both the cellular uptake and the tumor cell growth inhibitory action of doxorubicin. Identification of cytokines that increase the entry of pharmacologically active molecules into the desired target cells will greatly facilitate the pharrnacotherapeutic goal of achieving sufficient intracellular uptake of the pharmacologically active agent to obtain the desired cellular action. The increased permeability of the plasma membrane and concurrent enhanced uptake of metabolic inhibitors into leukoregulin-treated target cells constitute a new approach in this direction The ability of cytokines, e.g., interferon (IFN). interleukins ( I k ) and colony stimulating growth factors (CSFs), to stimulate the proliferation and activity of leukocytes including natural killer (NK) and lymphokine activated killer (LAW lymphocytes is well recognized. Cytokines also influence the net outcome of cytotoxic reactions by directly altering the sensitivity of target cells to the effector lymphocytes. Two cytokines possessing anti-thetical actions at the target cell level are leukoregulin and interferon. Leukoregulin, a 50 Kd cytokine up-regulates the sensitivity of tumor cells and human cervical epithelial cellsimmortalized by human papilloma virus 16 DNA transfectionto NK and IAK lymphocyte cytotoxicity. Tumor cells, ranging in sensitivity from resistant to highly sensitive, become more sensitive after one hour pretreatment with 2.5 u/ml leukoregulin in a 4hr 51Cr release assay. Moreover, leukoregulin renders the tumor cells two to six-fold more sensitive to LAK. This upregulation does not occur when the cells are treated with other cytokines, e.g., rGM-CSF, rlL-l(.&e) rlFN , rTNFa or combinations of the latter two. In fact treatment with rlFN for 16-18 hr, or with GM-CSF and IL-16 for 1 hr makes the cells resistant to both NK and LAK lymphocyte cytotoxicity. Leukoregulin may be a significant regulatory factor in NK and LAK lymphocyte immunocytotoxicity. Its biotherapeutic role in the prevention and control of cervical dysplasia and neoplasia is being studied using the HPV16 DNA-immortalized cells developed as a model for studying the role of HPVl6 in the development of cervical neoplasia and lymphokine modulation of cervical epithelial cell sensitivity to natural immunocytotoxicity.

CD 021 NEW ADVANCES IN CANCER THERAPY WITH RADIOLABELED ANTIBODIES, David M. Goldenberg, Immunology, UMDNJ, Newark, NJ 07103. Anticancer antibodies have been conjugated with gamma-emitting radionuclides for external scintigraphy (radioimmunodetection, or RAID), or with beta-or alpha-emitters, drugs, OK toxins for antibody-guided therapy. A major limitation for therapy, however, is the very low uptake of antibody in tumor, usually less than 0.01% injected dose per g tumor (1). In order to compensate for the low tumor uptake of antibodies in therapy, first generation therapeutic immunoconjugates are more likely to be radioimmunoconjugates, because the energy of the isotope will kill bystander cells. However, the high blood levels of radioisotopes deliver high rad doses to the bone marrow, which is the major organ of toxicity. Nevertheless, radiosensitive tumors have responded t o radioimmunotherapy (RAIT) (1,Z). Another approach is to combine high-dose radioimmunoconjugates with methods to protect o r rescue the marrow elements. studies have shown that use of antibodies directed against the primary antibody (anti-antibodies) can reduce bone marrov toxicity ( 3 ) . Also, certain cytokines, such as Interleukin-1, can prevent OK reverse the leukocytotoxicity of radioimmunoconjugates, thus enabling the administration of higher body rad doses (4). There was no s i g n i f i c a n t change i n HT-29 c e l l e x t r a c t s , however. To determine whether r I F N can s t a b i l i z e t h e covalent t e r n a r y complex formed by TS, FdUMP, Med CD 024 CYTOKINES MEDIATE NEURAL CELL GROWTH AND DIFFERENTIATION, Barbara Detrick, Charles H Evans, Gerald J Chader and John J Hooks, National Eye Institute. National Cancer Institute, NIH, Bethesda, Maryland 20892 Retinoblastoma, the most common primary intraocular tumor of childhood, is an important model for exploring human malignancy and differentiation. Although the histogenesis of this tumor remains controversial, current data support the concept that retinoblastomas originate from neuroectodermal tissue of the retina and consist of multipotential embryonic cells capable of differentiating into neuronal, glial and epithelial-like components. Using analytical flow cytometry we have been able to quantitate cellular proteins associated with the identification of selected cells growing within the retinoblastoma population. The undifferentiated human retinoblastoma cell line, Y-79, was used in this study. Cytokines, plurifunctional proteins which modulate many vital biological processes, were evaluated for their potential to alter retinoblastoma growth and development. We found that the cytokine. Interferon (IFN)-gamma, can selectively alter both MHC class I and II antigens as well as the neuronal protein, S-antigen. Gene transcription techniques indicate that IFN-gamma is acting at the level of post-transcriptional control to enhance S-antigen expression.

This system was also used to monitor and quantitate attachment cultures of retinoblastoma cells differentiated by butyrate. Butyrate-poly-D-lysine treatment results in enhanced expression of neuronal cellular markers while butyrate-laminin treatment results in enhanced expression of retinal pigment epithelial cellular markers. Herein, we describe that modifications in the developmental pathway of these cell types can be achieved by co-joining such cytokines, as IFN-gamma, IFN-alpha. IFN-beta and TNF-alpha, to butyrate. This study demonstrates that cytokines can influence the expression of selected cellular proteins and it underscores the possibility that cytokines may be potent modulating agents in cellular development. S t u d i e s a r e underway t o d e t e r m i n e t h e r o l e , i f a n y , o f AP-1 i n t r a n s d u c i n g n u c l e a r s i g n a l s t h a t r e g u l a t e t r a n s c r i p t The events that occur after the interaction of granulocyte colony stimulating factor (G-CSF) with target cells were analyzed in the interleukin (IL)3 dependent 32DC13(G) hemopoietic progenitor cells. These cells differentiate into granulocytes 1 2 days after G-CSF treatment and, at the molecular level, this phenomenon is associated with appearance of myeloperoxidase (MPO) mRNA at day 4, disappearance of c~myb mRNA at day 7, appearance of lactoferrin (LF) mRNA at day 6-10 and increases of beta actin and ras associated (rho) B mRNAs at day 7-10. G-CSF treatment also results in the activation of a number of early response genes including the rapid transient increase of both c-fos and c-jun mRNA between 30' and 1 hr of G-CSF exposure. Following the early response gene activation and prior to the activation of late response genes the receptor for granulocytic-macrophagic (GM)-CSF becomes expressed on the surface of 32D cells. The expression of this new receptor IS an irreversible phenomenon that provides the cells with an alternative option: to differentiate not only into granulocytes but also into macrophages upon addition of GM-CSF. The orderly program of terminal differentiation in 32DC13(G) cells can be disrupted by the introduction of viral oncogenes into these cells. v -m and E&I transformed 32D cells do not require growth factors for proliferation and they become unresponsive to G-CSF and unable to differentiate. In the case of & transformed cells G-CSF still elicits the early response of c-fos activation, indicating that the oncogene acts downstream to some of the early events in interfering with the response to G-CSF.

v-Ki-ras also interferes with the growth factor dependent proliferation and differentiation by interfering with the signal transduction pathway. Ki-ras 32D are blocked at the promelocytic stage of rnyeloid differentiation and proliferate rather then differentiate in the presence of G-CSF or GM-CSF but atypically differentiate and stop proliferating in IL3.

Thomas L. Leto, Karen J. Lomax, Stuart L. Abramson, and John I. Gallin, Bacterial Diseases Section, Laboratory of Clinical Investigation, National Institute of Allergy and Infectious Diseases, Bethesda, MD 20892. Phagocytic cells contain a multicomponent, latent superoxide-generating NADPH oxidase which is assembled into an active complex at the cell membrane following cellular interaction with opsonized particles or certain soluble stimuli. Activated oxidase catalyzes synthesis of superoxide leading to generation of potent oxidants (respiratory burst). Chronic granulomatous diseases of childhood (CGD) results from defects in activation of the phagocyte respiratory burst, leading to recurrent infections and granuloma formation. Normal monocytes cxhibit a physiologic loss of respiratory burst capacity in culture that can be restored by interferon-gamma (IFN-y). suggesting that components of the oxidase system may be responsive to IFN-y and IFN-y might be useful in CGD. autosomal recessive and some X-linked CGD patients (1). Similarly, IFN-y administered subcutaneously enhanced the oxidase activity of neutrophils and monocytes from some CGD patients. IFN-y associated increases in normal monocyte membrane cytochrome b558 (terminal electron donor of the NADPH oxidase complex) has been reported and such increases were also seen in neutrophils from some IFN-y treated CGD patients. Since monocytes from CGD patients not deficient in cytochrome b558 were particularly IFN-y responsive, this and other factors suggested that non-cytochrome components of the oxidase are also regulated by the oxidase. We have defined and cloned cDNA encoding two phagocyte oxidase factors (p47-phox and p67phox) located in cytosol and missing from phagocytes of patients with either of two autosomal forms of CGD (2,3). The p47-phox is a basic protein of 390 amino acids containing a tandem repeat of a 51 amino acid motif in the middle of the molecule with similarity to the control region A of the src family of oncogenes. This motif is present in phospholipase C gamma, fodrin, and p21-ras GTPase activating protein. The p47-phox carboxyterminus is arginine and serine-rich, containing several potential phosophorylation sites, and phosphorylated p47-phox appears to be the 47 kD phosphoprotein associated with respiratory burst activation. The p67-phox is an acidic protein of 526 amino acids which shares with p47-phox a src control region A motif at its carboxyl terminus. Using cDNA probes and recombinant protein antibodies to these oxidase components, further studies are in progress to determine the effects of IFN-y on these cytosolic oxidase components. in mice and man, by IFN?, and in vitro by TNFa, TNFp, CSF-GM, and CSF-M. Activation of normal macrophages can be explained by a decrease in the Km for NADPH of the respiratory burst oxidase, without a change in its Vmax o r content of its cytochrome b558 component (though IFN7 can increase cytochrome b558 in phagocytes with partial cytochrome b558 deficiency). ROI releasing capacity is suppressed by TGF-p1, TGF-82, and macrophage deactivation factor (MDF). MDF increases the Km for NADPH of the respiratory burst oxidase. The mechanism of deactivation by TGF-p is unknown.

Capacity to release RNI is enhanced by IFN7, LPS, IFN7 + LPS, IFNy + TNFa/p, and IFNa/p + LPS. IFN7 activates RNI-producing capacity by inducing expression of the enzyme that oxidizes NG-L-arginine to nitric oxide (.NO), and may also enhance cellular levels of tetrahydrobiopterin, a cofactor of the .NO-producing enzyme.

Induction of RNI release is suppressed by TGF-p's and MDF. AP: AP in PMN include 4 defensins, lysozyme, major basic protein (MBP), cathepsin G , elastase, bactericidal/ permeability increasing protein (BPI), cationic antimicrobial protein of 3 1 kD, and two apparently novel proteins of 29 kD termed azurocidin and p29b. Mononuclear phagocytes in man make lysozyme but reportedly lack defensins, MBP, cathepsin G , PMN-type elastase, and BPI. The possible presence in human monocytes of a protein related to azurocidin, a broad spectrum antibiotic, is under study. Also under study is the impact of IFN7 therapy on levels of AP in PMN. Complex interactions among R O I , RNI and AP can be predicted but remain to be defined.

Administration of IFN7 has helped treat and/or prevent protozoan and bacterial infections in mice and man. It has not been established if enhanced production or action of ROI, RNI, o r AP participate in these beneficial responses.

Phagocytes kill microbes either by depriving them of nutrients, such as tryptophan Capacity of macrophages to release ROI can be enhanced in vitro and in vivo,

RESISTANCE TO MALARIA SPOROZOITES, V. Nussenzweig, Department of Pathology, New York University Medical Center, New York, NY 10016.

Malaria sporozoites are introduced into the host's circulation by mosquito bite and within a few minutes they enter hepatocytes. There they multiply by schizogony and develop into the exoerythrocytic forms (EEFs). After a few days each EEF contains thousands of newly formed parasites (merozoites) which invade red cells. The EEF development is inhibited by 7-interferon and by tumor necrosis factor, raising the possibility that lymphokines, together with other effector mechanisms, mediate resistance to malaria in individuals living in endemic areas.

Antibodies to the repeat domain of the circumsporozoite (CS) protein inhibit entry of the parasite into liver cells.

However, to achieve satisfactory protection, antibody titers must be high. We recently reported' the identification of an epitope contained within amino acids 249-260 of the Plasmodium berahei CS protein, which is recognized by H-2Kd-restricted CD8' cytotoxic cells (CTL). Passive transfer of CTL clones directed against this epitope conferred a high degree of protection against sporozoite challenge. The CTLs are most likely attacking the EEFs, either by destroying the infected hepatocytes, or by releasing inhibitory lymphokines.

Either mechanism implies recognition of processed CS peptides in association with class I major histocompatibility complex molecules. 

VIVO, Steven G. Reed, Seattle Biomedical Research Institute, Seattle, WA 98109 Infections with intracellular protozoan parasites Trypanosoma cmzi and Leishmuniu are characterized by proliferation of the organisms within host macrophages. Here they are protected from the host antibody response, but are susceptible to killing by "activated macrophages. Delivery of the appropriate activation signals to the infected host cells is an alternative a proach to conventional chemotherapy which is toxic and often inefficient. A second aspect of iniction with these parasites is an accompanying depression of immune responsiveness to parasite and/or non-parasite antigens. Thus the marked decrease in T cell responsiveness which occurs during infection provides further rationale for choosing an immunotherapeutic approach in these infections.

In vitro, the most effective cytokine for the activation of mouse m a c r o p h a y to $hibit T. cruzi or LeisAmaniu is IFN-,. In vivo administration of IFN-, was also shown to be ef ective In reducing parasite numbers in infected mice. In mice with visceral leishmaniasis, liposome encapsulation of the IFN-, was necessary for effectiveness. The most dramatic effects were seen in mice infected with a uniformly fatal inoculum of T. cmzi. When these animals received repeated injections of IFNr during the course of infection, 100% of the mice survived. Furthermore, treated mice did not have depressed immune responsiveness. Several other cytokines mediated some degree of intracellular killing of these parasites, particularly in human monocytes and monocyte-derived macrophages. This included GM-CSF, CSF-1, and, to a lesser extent, IL-3. In some cases, combinations of IFN-, and a CSF had additive effects for parasite killing. The inhibitory effect of IFN-, was found to be partially blocked when the macrophages were pre-incubated with TGF-8. Pre-incubation of human macrophages with TGF-p also significantly reduced their respiratory burst activity. The induction of parasite killing by IFN-, also occurred in HIV-infected macrophages.

Other studies have focused on therapy of depressed immune responses in mice with parasitic infections. In vitro, the inability of spleen cells to respond to non-parasite antigens by producing antibody is a dramatic aspect of T. cmzi infection which begins during the acute phase and persists through the chronic phase of infection. Although the splenic B cells are capable of antibody production, the necessary T cell signals appear to be absent or blocked. Cytokine therapy is very effective in restorin depressed antibody production in infected mice. Most effective are GM-CSF, IL-1, and IL-2, all ofwhich function both in vitro and in vivo. Anti-1JA antibody (11B11) was also effective in restoring normal antibody production in vitro. Mechanisms relating to cytokine action in this system of depressed immune responsiveness will be discussed. When adenovirus infects human or rodent cells -, expression of products of the viral E1A region, which contains the functions required by the virus for the earliest phases of its replicative process, alters many cells i n sucn a way that. they become susceptible to the lytic effects of TNF.

Adenovirus encodes at ledst two, and possibly three, other proteins elsewhere in its gen0rr.e that function to prevent TNFmediated destruction of virus-infected cells.

The functioning of these proteins is highly dependent on the infected cell.

Thus, most mouse fibroblasts become susceptible to TNF when infected with adenoviruses that delete only a single gene in the E3 region that specifles a 14.7kD protein (E3-14.7K). Infection of these cells with viruses that produce E3-14.7K prevents TNF-mediated lysis.

In contrast, infect.ion of many human fibroblasts (HEL299, for example) or cultured human turror cells with viruses that delete only E3-14.7K does not produce the TNF-sensitive phenotype despite the fact that E l A induces TNF sensitivity in these cells. In addition, these cells are actively protected from TNF lysis when susceptibility is induced by other means, such as inhbition of protein synthesis.

Using double deletion mutants we have mapped a second gene involved in prevention of TNF lysis to the E 1 B 21kD protein. Although E3-14.7K appears to protect the vast majority of cells from TNF, ElR-21K is only effective in a subpopulation of cells. We have recently identified but have not precisely mapped a third gene, also witnin che E3 region, that prevents TNF lysis of yet another distinct subpopulation of cell lines. The two identified genes, E3-14.7K and EIB-21K, appear t.0 act by completely different mechanisms in preventing TNF lysis. The immunodeficiency virus (HIV), the causative agent of the acquired immunodeficiency syndrome (AIDS), frequently establishes in CD4' cells of the immune system a state of chronic infection which is characterized by low levels of virus expression. In order to investigate the physiologic mechanisms that convert a latent to a productive infection we have established a panel of chronically infected cell lines of both T lymphocytic and monocytic origin, reflecting the two immune cell targets of HIV infection. Using these systems, we have demonstrated that certain cytokines can upregulate the expression of HIV. In particular, we first observed that tumor necrosis factor alpha (TNF-a) was capable of inducing HIV expression in both chronically infected T lymphocytic (ACH-2) cells and promonocytic ( U l ) cells In addition, colony stimulating factor such as GM-CSF selectively upregulated virus expression in U1, but not in ACH-2 cells. More recently we have identified an autocrine mechanism of HIV upregulation which is mediated by the synthesis and release of endogenous TNF-a, which contributes both to the constitutive and to the inducible expression of virus in these lines. At the molecular level, we and others have identified the cellular transcription factor NF-kB as the mediator of TNF-a induced effects on virus expression. Furthermore, we have also correlated the viral inductive actlvity of certain cytokines, such as TNF-a, with the activation of a protein kinase C dependent pathway of second cellular messengers. Another pleiotropic cytokine, interleukin-6 (IL-6). similar to GM-CSF. was capable of inducing HIV expression in monocytic cells, but not in the T lymphocytic line ACH-2. Of interest, in promonocytic cells, TNF-a. GM-CSF and IL-6 synergized in the induction of viral expression. The molecular basis of these synergistic effects is currently under investigation. Preliminary results suggesl that different cytokines can affect different steps of viral replication or expression. In fact, in contrast with TNF-a, stimulation of infected cells with IL-6 results in a significant induction of HIV expression measurable at the proteinivirion level, which is not associated with a significant transcriptional activation of HIV. In conclusion, our in vitro studies support the concept that HIV has evolved to infect and persist in the CD4+ immune cells where virus expression is modulated by the cytokine network which normally regulates the homeostasis of the immune system. We have carry out a series of studies that ask why T lymphocytes from old mice, like those from old humans, respond poorly to mitogenic lectins and to antibodies specific for components of the T cell receptor complex. Our results show: a) that many T cells from old mice fail to produce an increase in cytoplasmic free calcium concentration b) that the inability of most T cells from old donors to generate this calcium signal largely accounts for the c) that the defect in calcium signal generation is a caused by defects in the production of inositol d) that T cells from old mice are relatively resistant to increases in [Ca"] even when these are caused by e) that the defective T cells that accumulate in old mice express high levels of the Pgp-l suface antigen, a These findings have led to a model for senescent decline in T cell function, in which thymic involution leads to a decrease in production of virgin T cells, leading to the relative enrichment of memory T cells that resist changes in intracellular Ca2+ levels, possibly by a change in the plasma membrane calcium extrusion Pump.

after stimulation with either Concanavalin A or anti-CD3 antibody; decline, with age, in tests of helper, cytotoxic, and proliferative T cell function; phosphate second messengers; agents that bypass the signal transduction pathways (e.g. ionomycin challenge); and marker for memory cells. Institute on Aging, NIH, Baltimore, MD 21224. The age associated decline of the immune response is well documented, but the cellular and molecular basis of this decline are not well characterized. While animal models have suggested age-associated alterations in the synthesis of the lymphokines interleukin-2 (L2), interleukin-3 (IL3). interleukin-4 (IL4), and interleukin-6 (IL6), data in the human is less clear. Decreased IL2 production and expression of cell membrane CD25, as well as the uptake and utilization of IL2 are known to characterize T cells from elderly persons. However, this defect appears dependent upon the manner in which the cells are activated. The in v i m production of IL2 and the expression of the signal transducing high affinity interleukin-2 receptor (HA-LZR) by mononuclear cells from aged individuals activated with cell membrane dependent mitogens such as phytohemagglutinin (PHA) or concanavalin A is decreased, as is the amount of specific IL2 and p55 IL2R mRNA. While p70 ILSR is an essential and lirmting component of the HA-IL2R, its mRNA expression does not undergo a significant age-related decline. Conversely, T lymphocytes activated with agents that do not require binding to the cell membrane, such as the combination of the protein kinase C activator phorbol myristate acetate (PMA) and the calcium ionophore A23187, demonstrate no age-related differences in either IL2 or p55 IL2R specific mRNA accumulation. Yet these T cells, while producing up to 100 times more L 2 than PHA stimulated cells, proliferate less than comparably treated cells from young individuals. This decreased proliferation, as quantified by [3H]TdR incorporation into cellular DNA, appears confined to an as yet uncharacterized subset of T cells that either die during culture or do not respond to mitogen or antigen stimulation. This results in fewer L 2 synthesizing and IL2 responding T cells. While present data indicates that the specific genes for IL2 and p55 and p70 IL2R are not constitutionally defective in the elderly, there is not sufficient information to account for either the decreased proliferation observed in T cells from both aged animals or elderly humans or for the distinctive effects of different activating agents on T cell mRNA expression. Despite demonstrating important interactions with IL2 and in some cases age-related changes in animal models, the production and utilization of other important lymphokines such as IL3, IL4, and IL6 have not been systematically examined at the molecular level in humans. Splenocytes from aged mice synthesized less DNA than young counterparts following stimulation in culture with antibodies against the T cell receptor/CD3 complex. This difference was not due to gross changes in anti-CD3 dose o tima response kinetics, accessory cell function, numbers of T cells cultured, C D 4 ' : CD8' cell ratios or levels of CD3c membrane molecules.

Studies using two-color immunofluorescence (IF) staining and multiparameter flow cytofluorometry (FCF) revealed that smaller proportions of both CD4+ and CD8' splenic T cells from aged versus young mice were stimulated to enter and progress through the cell cycle. This was reflected by their lower expression of pre-S phase, early activation molecules including I L -2 , membrane RL388 Ag, and receptors for IL-2 and transferrin. Importantly, cells that did respond to stimulation were found to express the full range of cell-associated, early activation characteristics regardless of donor age.

Multicolor IF and FCF anal ses of CD4+ cells have revealed substantial phenotypic differences between C D 4 ' subpopulations obtained from young o r aged mice. A larger proportion of CD4'

T cells from aged donor spleen was found to express high levels of membrane Pgp-l (CD44) molecules, but lower levels of CD45RB and 3G11 molecules when compared to young splenic CD4' cells. The phenotypic pattern of splenic CD4' cells from aged mice was consistent with the pattern that defines Th2 cells in young mice.

Consistent with this finding, splenic CD4' cells from aged mice were found to produce significantly greater levels of IL-4 and lower levels of IL-2 upon primary stimulation with anti-CD3 in culture than young counterparts. These latter results indicate that analysis of age-related changes in immunocompetence should take into account shifts in the proportion of phenotypically and functionally distinct subsets within the major CD4+ (and CD8+) T cell subpopulations. The ability to transfer new genetic information into hematopoietic cells provides a new and promising approach to address questions concerning stem cell commitment and proliferation. It should be possible, for instance, to introduce developmentally regulating genes. oncogenes and genes encoding for growth factors into various hematopoietic cells, thereby modulating the system in a precise way. In addition, certain human genetic defects may be corrected by the insertion of a functional gene into primitive bone marrow cells of the hematopoietic' system. Growth factors are soluble polypeptides that interact with cellular receptors controlling cell proliferation and differentiation. A number of oncogenes have been found to encode for growth factors or for the receptor of such a growth factor. Some receptors exhibit tyrosine-kinase activity, wich might lead to an overstimulation of cells due to increased autophosphorylation. Unregulated growth by the autocrine production of growth factors has been proposed as a mechanism of tumorgenesis. Colony stimulating factors (CSFsl are known to be essential for the survival, proliferation and differentiation of hematopoietic cells "in vitro". The exact physiological role still remains unclear but seems, for exemple, to include a proliferative response to an infection chellange. Autocrine loops involving CSFs in leukemia have generated an immense interest, in particular for GM-CSF and Interleukin -3 (IL-3 ) . Retroviral mediated gene transfer of IL-3 into normal murine cells resulted in myeloid proliferation "in vitro" under serum free conditions, and demonstrated "in vivo" an elevated white cell count with normal differentiating myeloid cells, a hypercellular bone marrow, and infiltration of the liver and the spleen. Exploiting the CSF-activity of a gibbon lymphosarcoma cell-line a gibbon-IL-3 and its human homolog could be cloned. The function of the gibbon -1L-3 is similar to that of human IL-3 with a significant homology between the two genes. Studies of the leukemogenic potential with regard to the autocrine production of human I t -3 have to be evaluated, since experiments in the murine system indicated that IL-3 alone is able to induce abnormal proliferation of mature myeloid cells. and tumorgenicity can be obtained in conjuction with other transforming events. The current working hypothesis is that in the M07E cell line. internal stimulation of receptors is qualitatively different than when stimulated on the cell surface. Suggesting that another 11-3 receptor associated component normally present at the cell surface is not operational internally. This data demonstrate that autocrine stimulation might result in autonomous growth of the human cell line M07E. The results also suggest that direct internal stimulation may also occur as has been demonstrated in the murine sys tem. Recent investigations have provided evidence that lipoxygenase and cyclooxygenase products may play an important role in immunoregulatory processes. In particular, leukotrienes have been reported to augment, while prostaglandins have been shown to inhibit IL-1 biosynthesis. Likewise, lipoxygenase inhibitors have been shown to inhibit and cyclooxygenase inhibitors reported to augment IL-1 biosynthesis. Further studies have suggested an obligatory role for lipoxygenase products in the functional activities of interleukin-1 or interleukin-2. Two difficulties in interpretation of these data are the use of a bioassay (known to be down-regulated by PGE2) to measure interleukin-1 and the lack of a highly selective LO inhibitor ( i c , one having little or no CO inhibitory activity). The purpose of the present studies, therefore, was to determine whether a selective LO inhibitor would inhibit interleukin-1 biosynthesis, as measured in a receptor assay for IL-1 activity. Compounds tested included the CO inhibitor, indomethacin; the mixed LO/CO inhibitor, tenidap; and the selective 5-LO inhibitor, WY-50,295. When culture supernatants from murine peritoneal macrophages were assayed for IL-1 activity in a receptor assay on BALBtc 3T3 fibroblasts, indomethacin did not potentiate and WY-50,295 did not inhibit IL-1 biosynthesis at concentrations 10-3Ox greater than their CO or LO inhibitory concentrations. Tenidap had no effect at its CO inhibitory concentration but inhibited IL-I biosynthesis by 50% at 30 pM, five times its LO inhibitory concentration. Neither WY-50,295 nor tenidap enhanced or inhibited mitogen-induced splenocyte proliferation or IL-1 or IL-2-induced thymocyte proliferation at concentrations up to 30 pM. We conclude that at the LO or CO inhibitory concentrations for these compounds there is no direct effect on either the biosynthesis or activity of IL-1. High and low passage HL-60 populations containing at least 90% GI cells were obtained by centrifugal elutriation, exposed to 100 u recombinant human GM-CSF and/or 0-1.25% DMSO and phosphoprotein changes quantified on autoradiograms of [32Pl-orthophosphate-labeled cell proteins separated by giant 2-D gel electrophoresis. Results were correlated with intracellular pH, determined by measurement of BCECF fluorescence; [ 32P]-orthophosphate uptake; and cell cycle progression, determined by flow quantitation of DNA content, cell volumes, and cell concentrations. GM-CSF stimulated, and DMSO inhibited the phosphorylation of two proteins (N60 kDa, p.i. 5.6; 50 kDa, p.1. 5.1) within one minute of exposure. These changes were sustained for at least four hours, were not associated with changes in intracellular pH and preceded similar antagonistic effects on phosphate uptake (15-30 min), cell volume change (18-20 hr), and cell concentration increase (28-32 hr). GM-CSF accelerated, and DMSO inhibited G1 to S transit with the most marked antagonism observed in the second cycle following synchronization (28 to 40 hrs). Cell maturation (morphology, NBT reduction) was dominated by DMSO and not antagonized by GM-CSF. These results suggest that GM-CSF and DMSO exert antagonistic effects o n growth through a common mitogenic pathway not activated by change in intracellular pH; that protein phosphorylation is an important early event, and that the major locus of cytokinetic effect is on G I to S transit. The relative ability of TNF and lymphotoxin [LT] to inhibit the growth of five human tumor cell lines was examined both in the presence and absence of IFN-y. Two adenocarcinoma lines, HT-29 and SK-CO-1, were 20-and 320-fold more sensitive to the inhibitory effects of TNF than LT in 3to 4-day proliferation assays. In contrast, the breast carcinoma line BT-20 showed only a one-to twofold difference. The MCF-7 and ME-180 cell lines exhibited intermediate behavior. These results parallel the reported disparate potencies of TNF and LT in their effects on endothelial cells, hematopoietic development and their abilities to sustain a mixed lymphocyte response. Radiolabeled TNF binding studies showed two classes of receptors (K 0.04 t o 0.15 nM and 0.2 to 1.0 nM) on the highly sensitive SK-CO-1 line.

HT-29 cells alsg appeared to possess some high affinity-binding sites, whereas the BT-20 line completely lacked the high affinity form. Thus the presence of high affinity-binding sites correlated with increased sensitivity to the antiproliferative effects of TNF. Cold TNF competed with the binding of radiolabeled human TNF three-to fivefold better than LT for binding to all three lines. These relatively small differences between the TNF and LT receptor-binding characteristics are insufficient to explain the dramatic disparity in their antiproliferative properties. Thus, the receptor binding data conflict with the growth inhibitory effects. Finch-Arietta, Department of Allergy and Inflammation, Hoffmann-La Roche, Nutley, NJ 07110. Although interferon-y has been shown to effectively prime macrophages to secrete tumor necrosis factor w (TNFw) i n response to trigger stimuli such as LPS, it i s reasonable to assume that other cytokines present in the extracellular environment may cooperate with bacterial products to activate cytokine production by macrophages. Interleukin 6 (IL-6), f o r example, has been detected i n rheumatoid synovial effusions and is produced by synovial cells and tissues. Therefore, the purpose of the present study was to examine whether IL-6 may serve as a priming stimulus for macrophage activation as assessed by TNFu production. THP-1 monocytes were differentiated by treatment with phorbol ester (TPA) for 30 minutes, then "primed" for 18 hr in the presence o f 0.1 -10 ng/ml IL-6, followed by overnight incubation with 200 ng/ml x. minnesota LPS to "trigger" TNFw secretion. TNFa contained in cell supernatants was quantitated by specific ELISA. Although IL-6 alone was observed to be a weak stimulus for TNFo secretion, I L -6 demonstrated a marked and dose-dependent potentiation of TNFa production when added prior to LPS. The priming effect of IL-6 could be reversed by the addition o f a neutralizing antibody against IL-6. Taken together, these data demonstrate that I L -6 can prime macrophages for enhanced TNFa release when exposed to a sufficient trigger signal, such as LPS. Although IL-6 exerts diverse regulatory effects within and outside o f the immune system, this is the first report to describe the macrophage-activating properties of this cytokine. Imnunoprecipitation of EGF-R from J2P-equilibrated A431 cells demonstrated that TNF treatment (10 nM) resulted in a time-dependent stimulation of EGF-R phosphorylation which was maximal (-3fold) after 10.20 min of TNF exposure. Equal concentrations of ECF resulted in a similar stimulation of EGF-R phosphorylation and phosphoamino acid analysis revealed that both factors resulted in increased phosphorylation on tyrosine residues. Tryptic phosphopeptides of ECF-R isolated from A431 cells treated with TNF, EGF or phorbol ester demonstrated that TNF induced phosphorylation of EGF-R which was quite distinct when compared to the actions of phorhol ester but was similar to the effects of EGF on receptor phosphorylation. However, TNF stimulated phosphorylation of tyrosine residues on a distinct phosphopeptide when compared to the tyrosine phosphopeptides resolved from EGF-treated cells. Unlike EGF, TNF was unable to directly stimulate EGF-R tyrosine kinase activity in membranes prepared from A431 cells. Our results suegest that TNF stimulates phosphorylation of EGF-R in a manner distinct from that of both phorbol ester and EGF. Future studies of the differential phosphorylation of EGF-R by TNF may aid in the understanding of the complex process of the molecular mode of TNF action. Anderson Cancer Center Modulation of the tyrosine kinase activity of pp60c-src has been implicated in several cellular processes. In human colorectal carcinomas the kinase activity of pp60c-src is greatly elevated with respect to normal colonic epithelial cells, leading to the hypothesis that increased activity m y be associated with loss of growth control. We have studied the relationship of kinase activity to growth control by analyzing the effects of several biologic response modifiers, including tumor necrosis factor alpha ( W ) , retinoic acid (RA), and the antibiotic herbimycin A (herb A) on the growth and soft-agar colony formation of colon cells. In cells growth-inhibited in monolayer by "NF, c-src kinase activity was decreased by greater than 19 fold within 10min; in cells resistant to TNF, no effect on pp60c-src kinase was observed. RA exerted different growth-control effects, decreasing pp60c-src kinase activity in some cell lines in the absence of qrowthinhibition in monolayer. However, in these cell lines, a decrease in soft agar colony fomtion was observed. While TNF and retinoic acid affect pp6flc-src kinase activity indirectly, herb A specifically inhibits tyrosine kinases, and directly inhibits pp60c-src kinase in vitro. We have found that herbimycin A inhibits every colon carcinm cell line thus far examined, and rapidly reduces c-src kinase activity, but not EGF receptor kinase activity. These studies suggest that specific dulation of the tyrosine kinase activity of pp60c-src is likely to directly affect growth properties of colon tumor cells.

YtjTOCENY OF INTERLEUKIN-1-LIKE ?ACTORS IN NORMAL RAT TISSUES. Tina Granholm and 'Olof SOder, PediatFic Endocrinology Unit, Karolinska Hospital, 5-104 01 Stockholm, and Department of Pediatric Surgery, S:t Gtkan's Children's Hospital, Stockholm, Sweden. Introduction: Interleukin-1 (IL-1) was originally described as a potent inflammatory mediator secreted by activated macrophages. IL-1 has also been demonstrated in the skin, and we have recently detected IL-1-like factor(s) in the testis, esophagus, stomach and tongue of adult rats. In the present project we have studied the ontogeny of IL-1-like factors in the rat. Methods: Tissues from Sprague-Dawley rats were collected at fetal day 18, and day 7, 17, 30 and 42 postnatally. IL-1 activity of aqueous tissue extracts was measured in a murine lymphocyte proliferation assay. Results and conclusions: Fetal esophagus, stomach, tongue and skin were devoid of IL-1 bioactivity, whereas the liver showed high activity. 7 days after birth all studied tissues were positive. The activity increased during postnatal development, and reached plateau levels in adult animals, except in the liver, where the activity decreased. Fetal liver is active in hematopoiesis, during which IL-1 has been shown to act as a growth factor. Our findings of IL-1-like factors in other tissues is as yet unexplained, but since these tissues have a high proliferation rate it might be suggested that they function as growth factors. Expression of the IL-6 gene was studied in I2 cases of primary human CNS neoplasia by RNA dot blot hybridization Transcription of the IL-6 gene was detected in RNA isolated from fresh tumor tissue in one case of meningioma, 1 of 2 cases of primary neuroectodermal tumor, and 2 of 4 cases of high grade glioma In addltion, one case of low grade and 2 of 4 cases of high grade glioma in tissue culture expressed IL-6 In order t o establish the influence of IL-6 upon the proliferation of primary CNS neoplasms, the same five cultured tumors were challenged w i t h increasing concentrations of recombinant human 11-6 (rlL-6) up t o lOOuiml and 3H-thymidine incorporation was measured at various intervals up t o 7 days Results indicate that only the cultured cells that did not express the 11-6 gene were stimulated by exogenous rlC-6 I f the tumors transcr!b!ng the IL-6 gene were maximally stimulated by endogenous production of this cytoklne, these results are consistant w i t h 11-6 serving as an autocrine growth factor for some Primary CNS neoplasmc and suggests that not all gliomas are transformed by the same mechanism University, Logan, UT 84322-0300, and Division of Infectious Diseases, University of Utah Medical Center, Salt Lake City, Utah 84148. It has been shown previously that some mammalian cells, including CAM's, produce nitric oxide (NO) from L-arginine and that N-monomethylarginine (NMMA) is an inhibitor of NO production by this pathway. Previous studies have also shown that CAM's induce loss of intracellular iron and iron-containing enzyme function in target cells. We show here the production of a signal detected by electron paramagnetic resonance (EPR) spectroscopy in CAM's attributable to iron-nitrosy1 (Fe-NO) complex formation. The production of this signal is enhanced by L-arginine addition and prevented by NMMA. Hyperfine splitting differences are observed in the iron-nitrosyl signal when 'IN is substituted for "N in the guanidino group of L-arginine. These data show conclusively that NO arises from the guanidino group of L-arginine and that this NO forms a tight complex with intracellular iron. We also have shown previously that TNF alone causes loss of target cell mitochondria1 electron transfer. EPR examination of mouse fibroblast C3HA cells reveals the presence in these cells of the iron-nitrosyl signal when they are killed by TNF plus cyclohexirnide. These data suggest that iron-nitrosyl complex formation via the L-arginine-dependent pathway and consequent loss of iron-containing enzyme function plays an important part in cell killing by immune effectors. Since CT-4R cells are also highly responsive to interleukin-4 (IL-4), we examined whether GSH regulates the cellular actions of IL-4.

Thymidine incorporation assay showed that incubation of CT-4R cells with GSH (0.64-3.2 mM) for 72 h enhanced the proliferative effect of IL-4. This effect of GSH was concentration dependent and down modulated by L-buthionine-(S,R)sulfoximine, an inhibitor of de novo GSH synthesis. Receptor binding studies showed that GSH treatment caused a 2-fold increase in the amount of radiolabelled IL-4 bound to the CT-4R cells. In addition, GSH increased the percentage of bound IL-4 that was internalized into the cells. These results suggest that GSH treatment increases the level of intracellular GSH and potentiates the binding and internalization of IL-4 which may in turn affect the growth and proliferation of the cells.

Winslow, Camille N. Abboud, Departments of Medicine and Pediatrics, University of Rochester Medical Center, Rochester, NY 14642. The monocyte-macrophage system is thought to contribute to host anti-tumor immunological defense. To investigate the role of cytokines on human macrophsge-mediated tumor cytotoxicity in vitro, we have utilized a model system employing the human colon cancer cell line target, SW116. This cell line is recognized by an IgG2a monoclonal antibody, 17-1A (Centocor, Inc., Malvern, PA) which mediates effective antibody dependent cellular cytotoxicity (ADCC). Monocytes isolated from normal donor buffy coat preparations were kept in continuous cultu e with maximum ADCC capability occurring after 5-7 days of culture as measured in a H-thymidine release assay. The most effective E:T ratio was 3O:l to 50:1, and cytotoxicity was determined after 18-20 hours. Nonspecific cytotoxicity was on average less than one fifth that observed in ADCC assays. Augmentation of ADCC was not seen with interleukin-1 alpha o r interleukin 6 . Human recombinant macrophage CSF in doses up to 1000 u/ml did not increase ADCC above levels seen with 10% fetal calf serum alone. Recombinant human (rh)GM-CSF, rhIL-4, and rhIL-3 were all capable of increasing ADCC in this system in a dose-dependent fashion during the 7 days of monocyte culture. These effects were seen whether or not the cytokine in question was present during the ADCC incubation. These data indicate that macrophage-mediated ADCC against a human solid tumor cell line can be augmented by several hemstopoietic cytokines. Whether such augmentation will be seen in in vivo o r with other tumor types remains an open area of investigation. We have cloned and expressed in recombinant form a cell growth regulatory molecule which can cause growth arrest at nanogram concentrations. The factor, which belongs to a class of proteins classified as soluble vertebrate lectins, plays a part i n the autocrine system of cell growth control as a negative regulator. Both as a regulatory molecule and as a growth arrest factor i t controls exit of cells from GO and traverse from S phase through G2 with a mode of action attributable to that of a cytokine rather than that of a lectin.

Ricerche Farmacologiche "MARIO NEGRI", Via Eritrea 62, 20157 Milan, Italy, hiversiti di Torino, Via Santena 5, 10126 Torino, Italy.

G and GM-CSF are hematopoietic growth and differentiation factors which have been considered restricted to the hematopoietic lineage. We observed that G and GM-CSF induce migration and proliferation of EC. Endothelial cells have receptors for G and GM-CSF similar i n number and affinity to those present in myelomonocytic cells. The rapid intracellular events initiated by these cytokines on binding to their receptors on HEC are not defined. Addition of C-or GN-CSF to HEC produced a rapid activation of Na+/H+ exchanger resulting in an increase in intracellular pH (pHi). Both cytokines induced an alkaline displacement in the pH; dependence of the exchanger, without affecting the affinity for external Na+ (Na,) Increased concentration of Cat+ and activation of protein kinase(s) C (PKC) seem critical in the triggering of human T cells. We used human peripheral blood lymphocytes (PBL) and human T cell lines H33 and J32 to study further the role of extracellular ca" and PKC in production of interleukin 2 (IL2), which ultimately drives proliferation of these cells. PBL stimulated by PHA as well as either cell line stimulated by PHA + PMA, produced IL2. Chelating extracellular Ca++ from 300 uM to 1 uM, to lower the out-in Ca++ gradient 100fold, abrogated IL2 gene expression in PHA-stimulated PBL; it was fully reconstituted by restoring the gradient. MgCl,, but not choline chloride, could partially replace Call to allow IL2 production by PHA-stimulated PBL. Inhibiting PKC with Staurosporine ( " S s ' * ) decreased IL2 gene expression in mitogen-stimulated cells by about 50% at a concentration of ~5 nM. IL2 production by mitogen-stimulated PBL or either cell line was also inhibited by Ss with an 15,, of 5-10 nM. Also, with purified PKC from H33 cells, 10 nM Ss inhibited autophosphorylation by 70% and ?istone phosphorylation by 50%. Thus, an out-in Ca*+ gradient >10 (i.e., [Cao ] > > 1 uM) is required for IL2 production by mitogen-stimulated human T cells, and PKC activation also affects IL2 gene expression. Preliminary observations suggest that PKC may also influence post-transcriptional events leading to secretion of IL2. Biochem. Biophys.252, 549, 1987) . binding of TGF-d to the EGF receptor and inhibits EGF/TGF-e induced generation of inositoll ,4,5 triphosphate in carcinoma cells. In addition, MAb 425 inhibited growth stimulation of carcinoma cells induced by exogenous EGF/TGF-a. Using a culture medium free of exogenous growth factors we found that MAb 425 inhibited growth of 5 out 7 carcinoma cell lines that express surface EGF receptors and produce TGF-w-ltke mitogens. This effect was independent of the level of EGF receptor expression, dosedependent and could be overcome by exogenous EGF/TGF-0. Our results suggest that tumor-derived, secreted TGF-0 is an autocrine growth factor for carcinoma cells under growth factor-deprived conditions.

In this study, we showed that MAb 425 also blocks

The Acidic FGF Receptor is Related to flg Tyrosine Kinase M. Ruta*, N. Neiger', J. Epstein', W. Burgess+, and J. Schlessinger** Laboratory of Retrovirology, FDA, +American Red Cross, and '*Rarer Biotechnology Inc. We have previously isolated a novel human gene from an endothelial cDNA library encoding a putative tyrosine kinase which we designated Qg (&-like gene). In order to analyze the gene product(s) of flg we generated rabbit polyclonal antibodies directed against a synthetic peptide from & and used it to immunoprecipitate biosynthetically labeled 48 protein(s) from a variety of human cell lines. This antiserum recognizes specifically glycoprotein(s) of 100-130 kd with a protein core of 90,000 and 110,000 daltons. Acidic fibroblast growth factor stimulated tyrosine kinase activity of & protein both in vitro and in living cells, suggesting that Qg encodes the membrane receptor for acidic FGF. Further supporting evidence came from cross-linking experiments on intact cells with the covalent cross-linking agent disuccinimidyl suberate and '251-labeled acidic FGF as a specific probe. The cross-linked '"Ilabeled-FGF receptor complex was specifically immunoprecipitated with & anti-peptide antibodies. It appears therefore that the receptor(s) for FGF is related to flg gene product. with the expression of TGF-a leads to transformation in model systems. We have examined the expression of EGF-R and TGF-a by several independent methods in primary human brain tumors. The amplification of EGF-R gene was observed in about 40% of the glioblastoma multiforme tumors examined, which is in agreement with previous reported studies. The TGF-a gene was found amplified in some anaplastic astrocytomas. TGF-a-like material was found to be present in the majority of tumors examined, albeit at different levels by competitive binding assays with iodiiated EGF. Western blot analysis confmed the presence of TGF-a, although the majority of the immunoreactive protein material appeared to be of large (> 10 Kd) molecular weight. Alternatively, the activity of EGF-R from the various tumors was examined in immunocomplex kinase autophosphorylation assays. Quantitation of the kinase activity per receptor, as determined by Western blotting, revealed three distinct patterns, 1) low basal activity with no stimulation by exogeneously added EGF; 2) increased basal level of activity with stimulation with EGF; and 3) increased basal level and no stimulation with EGF. The increased basal kinase activity was independent of the expression of EGF-R, although all samples exhibiting high levels of kinase activity were derived from glioblastoma tumors. These results suggest the TGF-a is expressed in many glial-derived brain tumors, and that the intrinsic tyrosine protein kinase activity per receptor varies dramatically among the tumors examined.

CD 118 ERYTHROPOETIN INHIBITORY AND POTENTIATING ACTIVITIES, G. Steiner, E. Pavlicek, U. Barnas and W. Woloszczuk. 2nd Dept. of Medicine, University of Vienna, and L. Boltzmann-Institute for Clinical Endocrinology, Gamisongasse 13, A-1090-Vienna, Austria. Erythropoetin (EPO) is considered as the key regulator of red blood cell formation. Othergrowthlactors. suchas lymphokines and colonystimulatinglactors(CSF) may modulatethe effectsof EPO. IL-1 isof interest because it can induce the production of CSFs both in vifro and in vivo, but this lymphokine is also suspected to be partially responsible for anaemia in some rheumatoid arlhritis patients. To measure EPO modulating activities we developed a cellular in vifro assay using EPO-dependent erythroid precursor bone marrow cells from rabbits treated with phenylhydrazine and actinomycin D. Cells weregrown in serum-free medium containing between 3 and 100 mU/ml EPO. Cell proliferation as measured by [ 'Hlthymidine incorporation was strictly dependent on the presence of EPO CSFs (GM-CSF, G-CSF, M-CSF, IL-3) and lymphokines (IL-1, IL-2, IL-4, IL-6, TNF-alpha. IFN-gamma) did not have stimulating activity either alone or in combination with EPO. EPO-induced-growth was reduced in the presence 01 100 U/ml IL-1 or IL-6 and almost completely inhibited by 50 Ulml IL-2 or 5 m m V l succinylacetone (an inhibior of heme bbsynthesis). To measure EPO-inhibitory or -potentiating serum activities, cellswereincubatedwith25mU/ml EPOand1O%to0.1%humanserum. At lO%serumconcentrationproliferationwascompletely abolished by most sera, inhibition disappeared when the serum concentration was reduced and at 1% concentration sera from healthy subjects EPO-potentiating activities (EPO-P) caused a 1.4 to 2-fold increase in prolileration. In contrast, only 50% of sera from patients with impaired kdney function exerted such activities. Sera from 10 hemodialysis patients treated with recombinant human €PO were assayed over 5 months of treatment. No change in inhibitory activities could be observed at 10% serum concentration. However, a rise in EPO-P was observed in seven cases during the initial phase of therapy which was followed by a pronounced decrease after two months treatment when hemoglobin levels were >9mg/dl. These data have to be interpreted carefully, but we may draw the following conclusions: serum contains activities which inhiba or potentiate the effects of EPO on erythroid precursor cells; lymphokines IL-1, IL-2 and IL-6 inhibit EPO-induced growth in vifro; therapy of anaemic patients with rEPO stimulates erylhropoesis and potentiates and inhibiis serum activities. These serum activities may be cylokines or cytokinelike factors and could be important in regulation of erythropoesis.

In view of the supposed role of mononuclear phagocytes in growth control and tumor defense an important question concerns the principles which determine the quality and quantity of a cell's response to the complex cytokine milieu created during interaction with these effector cells.

In an approach to this question the population development of 8 adherent human tumor cell (TC) lines, which grow continously in a defined medium, was studied under the influence of elutriated, unstimulated human monocytes (MO) or of supernatants from MO/TC-co-cultures. The analysis, performed by electronic counting of tumor cell nuclei and flow cytometry yielded the following results: At subconfluent TC-density MO cart cause growth stimulation in EGF-dependent cell lines, which have been deprived of EGF, whereas in the presence of EGF growth inhibition is the general TC-response observed to a challenge by MO. On the other hand, at confluent and higher TC-density the induction of cell death is a prominent feature of the interaction, leading to maximum cell loss at rather low effect0r:target-ratios (E:T=1:21, whereas high MOnumbers (E:T=10:1) cause a reversible GI-arrest of the cell cycle, and prevent the induction of cell death. EGF-withdrawal decreases the Gl/S-transition rate in EGF-dependent TC populations and concomitantly renders them less susceptible to the induction of cell death. Addition of hydrocortisone reduces the rate of cell loss. The data suggest, that in MO/TC-cocultures the decision on target cell lysis is made by the target cell and that the criterium for this decision is the target cell's ability to respond to a MO-challenge by arresting the cell cycle in G1. This ability can be modulated by growth factors and hormones. Interactions between target cells play an important role in determining the result of this decision process. and Poornima Kumar, Genelabs Incorporated, Redwood City, CA 94063. This study has employed TNF-resistant tumor cell variants to analyze the mechanism by which TNF lyses sensitive tumor cells. TNF resistant variants were selected by the prolonged culture of U937 tumor cells in the presence of TNF. Experiments were designed to test the hypothesis that TNF-induced activation of the DNA binding protein, NF-kB, results in DNA fragmentation and cytolysis in sensitive targets. Normal U937 tumor cells will undergo DNA fragmentation into pieces that are multiples of 200 base pairs in response to TNF. In contrast, the DNA of the TNF-resistant variant (UgTR), does not fragment, thus supporting a role for DNA fragmentation in the TNF lytic mechanism. Other investigators have shown that in some cells, TNF can activate NF-kB. We postulated that this response may activate the transcription of a gene(s) encoding proteins that function in DNA fragmentation and cell death. However, gel mobility shift assays demonstrated that TNF will activate NF-kB in the resistant variant as well as in the sensitive u937.

Two possible models could explain these results. One is that activation of NF-kB is not involved in the TNF lytic pathway.

Alternatively, if activation of NF-kB is essential, then some subsequent steps muat be blocked to prevent cytolysis in the U9TR variant. It is concluded the activation of NF-kB alone is not sufficient to lead to TNF-induced DNA fragmentation and target cell lysis. (Supported by grant CA 47669). Some vaccines have been shown to alter hepatic drug metabolism. The mechanism of this inhibition is unknown. Studies suggest that the immune system may be involved in the mediation of the inhibition of drug metabolism. Increased serum levels of interferons (IFN) have been reported following administration of vaccines, and IFN inducers such as p l y i:C are known to inhibit several drug metabolizing enzymes. In order to elucidate the mechanism by which vaccines alter hepatic detoxication and the role of immunomodulators, we have studied hexobarbital-induced sleep time, as well as cytosolic and microsomal enzyme activities in mice treated with DTP vaccine, poly I:C. and several cytokines. Sleep times were significantly increased in mice which had received DTP vaccine, p l y I:C. or IL-2. Treatment with IFN-a alone did not affect sleep time, but in combination with IL-2 it prolonged the sleep time over that of animals treated with IL-2 alone. Cytochrome P-450-dependent enzyme activities were significantly inhibited (up to 75%) in mice treated with DTP vaccine, poly I:C, or IL-2. IFN-a alone caused marginal (1 0-20%) decreases in cytochrome P-450-dependent enzyme activities. Mice pretreated with gamma-irradiation to suppress the immune system reduced the inhibition of drug metabolism following administration of cytokines but not the inhibition following administration of DTP vaccine. These studies suggest that cytokines may play a complex role in the inhibition of drug metabolism and furlher suggest that radiosensitive cells may mntribute to the IL-2 induced suppression of drug metabolism. The generation of active cytotoxic T lymphocytes (CTL) requires both antigen and lymphokine signaling. To investigate lymphokine requirements we cultured thymocytes from C57BL/6 mice f o r three days in the presence of 3 ug/rnl of Con A plus various lymphokines. In a typical experiment lectin mediated target cell lysis by responders cultured in the presence of IL-4 was 43+5% compared to 1751% lysis for responders cultured in the presence of IL-2 and 3121% lysis f o r responders cultured in the presence of IL-2 + IL-6. To understand the relationship between the different lymphokines we added to the thymocyte culture mAbs able to block the activity of: IL-4 (1lBll): IL-2 (S4B6); or IL-6 (6B4). Anti-IL-4 mAb, able to block the response to IL-4, had no effect o n the generation of CTL in the presence of IL-2 + IL-6. Thymocytes cultured in the presence of IL-2 + IL-6 gave 2822% lysis while culture in the presence of IL-2 + IL-6 + llBll gave 3021% lysis. In contrast mAb to either IL-2 or IL-6 was able to inhibit the generation of CTL in the presence o f IL-4. In this assay culture with IL-4 resulted in responders eliciting 4622% lysis which fell to 0% lysis when S4B6 was included in the thymocyte culture.

In another assay, culture in IL-4 generated 76%13% lysis while culture in IL-4 + 684 gave 25513% lysis. Using lymphokine-dependent cell lines we have shown that thymocytes cultured in Con A were capable of producing low levels of IL-2, IL-6, and possibly IL-4. Our data suggest that the response generated in the presence of exogeneous IL-4 is dependent upon the presence of endogenous IL-2 and IL-6. Phorbol esters (TPA) are known to induce the expression of a number of cytokines in T cells by transcriptional mechanisms. We noted that TPA induced the expression of GM-CSF mRNA in EL-4 thymoma cells, a mouse T cell line >lo0 fold without a corresponding increase in transcription as measured in nuclear run-on assays. Furthermore, we noted that untreated EL-4 cells contain detectable amounts of GM-CSF mRNA. Actinomycin D chase experiments show that GM-CSF mRNA half-life in untreated cells is <30' and after TPA treatment it exceeds 6 hours. Changes in mRNA ti12 appear to be the principal explanation of the large increase in GM-CSF mRNA after TPA treatment of these cells. We have begun using RNA mobility shift gel assays to characterize cytoplasmic proteins which bind to the B'UTR of the GM-CSF message that may modulate the changes in message stability. We also have noted the presence of unspliced and partially spliced GM-CSF mRNA precursors in nuclear RNA from untreated cells. Interestingly, these GM-CSF mRNA precursors rise > 5-fold after 30 minutes of TPA treatment and then fall below their initial levels after 3 hours at which point mature GM-CSF mRNA appears in the cytoplasm. The rapid rise in GM-CSF precursor RNA is out of proportion to the increase in transcription which occurs after phorbol ester treatment. These latter data suggest that GM-CSF mRNA levels are also modulated by stabilization and processing of nuclear RNA precursors. We have been studying the IL-l responsiveness and receptor expression by murine T cells. The T h l cell line DIOG4.1 will proliferate in response to both IL-la and IL-1p in the presence of suboptimal concentrations of ConA and this proliferation is dependent upon the production of IL-4. Over a period of several months of repeated stimulation with antigen/APC, the sensitivity of D10 cells to IL-I triggered proliferation was found to increase approximately 50-fold. This increased sensitivity could be reversed by a 2d culture in the absence of any added stimuli and was lost on cryopreservation. Binding studies revealed a single class of binding site and showed that the number of these IL-I receptors (IL-IR) was upregulated following stimulation, although the affinity remained the same. Crosslinking studies using 1261-IL-1 and the homobifunctional crosslinking agent DSS revealed 2 forms of IL-IR: an -80kDa form similar to the well characterized 1L-IR found on EL-4 cells and a -60kDa form. Partial peptide map analysis suggested that these 2 forms represent differently processed forms of the same receptor. Early passage, less sensitive, D10 cells expressed predominantly the -60kDa IL-1R whereas the more sensitive, continuously passaged cells showed a marked increase in -8OkDa IL-IR expression. Resting of such sensitive cells led to a down-regulation of the -80kDa IL-IR but had little effect on the -6OkDa form. In contrast, stimulation with ConA or anti-CD3 in the absence of APC led to a 10-fold increase in -60kDa IL-IR but had no effect on expression of the -80kDa form. The maintenance of both IL-IR forms on resting cells was blocked in the presence of anti-IL-4 antibody. A Th, cell line, MTglZB, was also studied. This cell line failed to respond to IL-I, either alone or in the presence of ConA or anti-CD3, and did not express -80kDa IL-IR. However. this cell line did express the -60kDa form of the IL-IR in a stimulation dependent manner. Although expression of the -60kDa form of the IL-IR cnn clearly be regulated in both Th, and Th, cells, whether it can transmit biological signals to these cells remains to be determined. Lymphocytes infiltrating human ovariancarcinoma were isolated from intraperitoneal ascites of patients with ovarian tumors. The presence of mRNAs of T cell-associated cytokines was determined by the polymerase chain reaction using specific primers after reverse transcription of cellular RNA. Using different numbers of amplification cycles, it was possible to quantitate the level of mRNA in the T-lymphocyte preparations. We have identified the presence of large amounts of mRNAs for I12 , Il3 , I16 , INF-6 and GM CSF in the tumor infiltrating T-lymphocytes. These cytokines may be involved in paracnne and/or autocrine regulation of the cellular immune response directed against the ovarian tumor.

CRYSTAL STRUCTURE OF RECOMBINANT HUMAN GRANULOCYTE-MACROPHAGE COLONY-STIMULATING FACTOR, William J , Cook, Steven E. Ealick, & Charles E. Bugg, Center for Macromolecular Crystallography, University of Alabama at Birmingham, Birmingham, AL 35294 and Paul Reichert, Gerald S. Hammond, Hung V. Le, Tattanahalli L. Nagabhushan, & Paul P. Trotta, Schering Corporation, Bloomfield, NJ 07003. Crystals of recombinant human granulocyte-macrophage colony-stimulating factor (GM-CSF) have been grown using either polyethylene glycol 8000 or solutions of sodium/potassium phosphate. The protein used in this study was derived from E. coli and is similar to naturally-occurring human GM-CSF.

The crystals are orthorhombic, space group P212121; the axes are a = 4 5 . 5 ( 1 ) , b = 58.7(1) and c = 127.3(1) A . Although the molecule exists as a monomer in solution, it crystallizes with two or three molecules in the asymmetric unit. Based on a molecular weight of 14,477 daltons as predicted from the cDNA, this corresponds to solvent volume fractions of 59% and 39% for two or three molecules, respectively.

The crystals are stable to x-rays at room temperature for at least three days, and x-ray diffraction data for the native crystals have been collected to 2 . 8 resolution using an area detector.

Crystallographic analysis of the structure of GM-CSF is in progress, and one potential derivative has been identified.

Current structural results will be presented.

William L. Crump 111, Laurie Owen-Schaub and Elizabeth Ann Grimm, Department of Tumor Biology, University of Texas M.D. Anderson Cancer Center, Houston, TX 7 7 0 3 0 . TNF-cu is postulated to play an obligatory role in the IL-2 driven generation of cell mediated, MHC unrestricted cytotoxicity. Furthermore, we have recently reported that IL-2 regulates lymphocyte production of TNF-ct and expression of tumor necrosis factor receptors (TNF-R) on lymphocytes. Our present work is directed towards characterizing TNF-R on activated human peripheral blood lymphocytes. Using PBL stimulated for 1 2 days with 5 0 0 U/ml IL-2, Scatchard analysis demonstrates a Kd of 40-60pM and approximately 2500 TNF bindincj sites per cell.

1251-TNF crosslinking studies with the DSS analog BS , demonstrate a receptor:ligand complex of approximately 1 0 9 . 5 kDa. An additional band at 46 kDa, which we postulate to be crosslinked trimeric TNF, is also observed. Photoaffinity crosslinking with 1251-ASD-TNF demonstrates a band of 95 kDa. The absence of the 46 kDa band in photoaffinity crosslinking studies supports our conclusion that this band is trimeric TNF.

Lectin precipitation experiments demonstrate that the TNF-R is a glycoprotein.

HUMAN PERIPHERAL BLOOD LYMPHOCYTES. SEA-induced T cell activation has been shown to be strictly accessory cell dependent and requires the binding of SEA to the MHC Class I1 molecule on the accessory cell and subsequent interaction with the TCR-CD3 complex on the responding T cells. We now demonstrate that SEA is a potent inducer of TNF in cultured human T cells and monocytes of peripheral blood. SEA at 1 pg/ml induced TNF activity with peak value after 48-72 hours of culture. IL-2 and IL-4 were growth promoting for human T cells, but only IL-2 induced TNF. TNF-alpha was produced already after 6 hours, whereas TNF-beta was produced after 48-72 hours of culture. TNF-alpha was detected in monocytes and T cells, whereas TNF-beta was seen exclusively in T cells. Production of TNF-alpha and TNFbeta required the yresence of both monocytes and T cells. CD4' and CD8 T cells produced TNF-beta, but the frequency of TNF producers was higher among the CD4+ cells. The CD4+45R-T cell subset produced TNF and interferon, whereas the CD4+45R+ cells produced only TNF. Three different antibodies against a human TNF receptor (htr-1, htr-5 and htr-9) have been examined for their binding pattern to U937 cells and ability to mimic TNF-a activity in U937 cells, Fs4 fibroblasts and human endothelial cells. Flow cytometric analysis revealed that htr-5 and htr-9 bound specifically to a TNF receptor on U937 cells which could be blocked by pretreatment with rTNF-a. Pretreatment of U937 cells with rTNF-p blocked the binding of htr-9, but to a lesser extent htr-5 binding. Pretreatment with htr-5 inhibited the binding of htr-9 to U937 cells while pretreatment with htr-9 did not inhibit htr-5 binding. These results indicate that htr-5 and htr-9 recognize distinct but overlapping epitopes of a human TNF receptor on U937 cells and that htr-5 may be close to a TNF-a specific domain of the binding site. Pretreatment with htr-5 or htr-9 only minimally reduced binding of BrTNF-a to U937 cells, however, these antibodies were much more effective in inhibiting BrTNF-n binding to HL-60 cells. Furthermore, it was found that htr-1 and htr-9, but not htr-5, had TNF-a activity on U937 cells, on Fs4 fibroblasts and on endothelial cells and that the TNF-Q activity induced by htr-9 was completely inhibited by htr-5. However, the cytotoxic activity of TNF-a was only partially inhibited by htr-5 on U937 cells while htr-5 had no effect on TNF-Q activity on Fs4 cells. The data suggest that a common epitope is involved in inducing TNF activity in three different cell systems. This investigation sought to determine whether leukoregulin secretion is up-regulated or truly inducible following mitogenic stimulation. The kinetics of leukoregulin secretion were examined to define the optimal time for collection of mRNA for cloning of the gene for leukoregulin synthesis. Human peripheral blood leukocytes were isolated on a discontinuous density gradient and cultured at a concentration of lo6 cellsfml at 37OC. Stimulation with 10xg/ml PHA revealed the inducible nature of this lymphokine as no constitutive secretion was detectable. Leukoregulin secretion commences at 4-6 hours, reaches a peak about 8 hours and plateaus by 18-24 hours. Stimulation with lo5 human K562 erythroleukemia cells/ml in place of PHA, although accompanied by a continued rise in secretion for 48 hours, is at a lower level than that obtained with PHA stimulation. The lower level of secretion may reflect a lower initial degree but continuing recruitment by tumor cell stimulation. The biochemical nature of K562 cell induced leukoregulin was analyzed to determine whether it is the same molecule as that produced with PHA stimulation by studying its molecular size, isoelectric point and biological activities. An examination of the possibility of negative feedback of leukoregulin secretion was made to determine whether the kinetics of leukoregulin secretion is a function of the lymphokine itself or of the stimulus.

S U R G I C A L O N C O~. O G Y , U C L A MEDICAN. C E N T E R , L o s ANGEI-ES, C A 90024. TNF i s A N I M P O i l T A N T 1 M M U N O R E G U I . A T O R Y C Y T O K I N E .

Culture of leukocytes in the presence of conditioned media containing leukoregulin showed no downregulation of leukoregulin production during 18 hours culture in the presence of PHA.

Division of Cytokine Biology, Food & Drug Administration. Bethesda, MD 20892. Interleukin 4 (IL-4), is a T cell-derived cytokine that regulates the induction of proliferation of resting B cells, and appears to act on various other cells involved in the immune response. The pluripotential effects of IL4 are dependent upon the interaction of IL4 with its receptor (IL-4R). IL-4R have recently been reported to be present o n the surface of a variety of inflammatory cells. both human and murine. Although the regulation and metabolism of these receptors have been examined on human and murine B and T cells, little is known about the metabolism or regulation of the IL-4R on macrophages. In studying the dynamics of IL-4R expression on the murine macrophage-like cell line 1774.16, we detected the presence of low numbers of high affinity IL-4 receptors on the surface membrane. However, upon exposure to interferon-7 ( I F N -71, a potent macrophage activating cytokine, there was a rapid induction of IL-4R on the cell surface. The increase in IL-4R density was first noted within 45 minutes after exposure of the cells to IFN-7. reaching its peak within 4 hours. No further increase in IL-4R expression occured over the next 48 hours. This induction of receptor required the continued presence of IFN-7. In experiments wherein the cells were treated for short periods of time with IFN-7, washed. and continued in culture, only those cells that were exposed to IFN-y for at least 2 hours displayed maximum IL4R expression. l h e induction of IL-4R by IFN-7 was dose-dependent. as little as 0.5 ng/ml of IFN-7 was capable of inducing IL-4R expression, with optimal induction at 10 ng/ml. The expression of murine class II genes can be induced on the surface of macrophages by both IFN-y and TNF-a. This induction is mediated via increases in levels of class II mRNA. We have demonstrated in WEHI-3 cells, using the bacterial chloramphenicol acetyl transferase reporter gene, that the promoter of the Aa gene is transcriptionally activated by IFN-yand TNF-a. Using 5 deletion and nested linker-scanner mutations, we have demonstrated that 105 nucleotides 5 of the initiation site are required for induction by each cytokine. This region includes the X,Y and H boxes. In addition, a unique sequence has been found, located between -31 and -1 7, which appears to play a role in TNFa-mediated, but not IFN-)-mediated An promoter activation. This sequence contains an SV40 core enhancer element and is also found in the human HLA-DQa and the rat RT1 .Ba genes, both Aa homologues. The core enhancer sequence is found in the En and the HLA-DRa genes at the same distance 3 of the Y box. The -31 to -1 7 sequence, which we have tentatively called the T box, also has homology to a region in the second intron of the c-mycgenes of mouse, human and rat. Gel mobility shift and methylation interference assays are in progress to analyze trans-acting factors which bind to this region and to determine whether the SV40 enhancer or the entire T box region is essential for binding . 

ANTIGENS. Gail Goodman-Snitkoff, Leslie E. Eisele, and Raphael J. Mannino. Department of Microbiology and Immunology, Albany Medical College. Albany. NY 1220R Understanding how to present specific, defined epilopes to the immune system in order to induce bolh humoral and ccI1 mediated immune responses is a major challenge in the dcsign of modern vaccines. Thc approach we havc taken has been lo attempt 10 determine minimal essential componenls required to construct a well defined immunogenic composite. Synthetic peptidcs, representing immunologically interesting epitopes from pathogenic microorganisms, i.e. HIV and Malaria, are provided with a hydrophobic tail through crosslinking to a phospholipid and then assembled into a pcptide-phospholipid ctmprisite. We have found that an immunogenic mixture is comprised of two essential components: 1. Each composite must contain peptides representing both B-cell determinants and T-helper (Th) determinants presented either contiguously, i.e. part of the same synthetic pcptidc, o r o n individual peptides. 2. To he immunogenic these peptidcs must be covalently coupled to phospholipid and inoculated as a peptide-phospholipid conjugate. No other carriers or adjuvants are required to induce an immune response. Additionally, trace amounts of a detergenl extract of the envelope of Sendai virus to act as a "generic" provider of Th-cell determinants can be included in thc ctimpositcs . Following immunization with these prcparalinns, antibody liters increase with each subsequent inoculation. Addilicmal parameters being studied includc dose OF immunogen, route of immunization, the nature of the crosslinker used to couple the peptides to phospholipid, and the pliospholpid composition. The ability o f peptide-phohpholipid composites to inducc aniigcn specific cy~olylic T lymphocytes is alsri undcr invcsligatitin.

Jane Bynum, and Frederick C. Kull, Jr. Divisions of Molecular Genetics and Microbiology and Cell Biology, Burroughs Wellcome Co., Research Triangle Park, NC 27709 The carboxyterminus of tumor necrosis factor (TNF) is a highly conserved, hydrophobic region. The role of the carboxy-terminal region in the structure and function of TNF was examined by designing changes in the codon that in the natural human TNF gene codes for the carboxy-terminal residue, leucine, (Leu156). Codons for other hydrophobic residues (Val, Phe, Ala, He) as well as for residues with hydrophilic side chains (Asn. Thr, Lys, Glu) were introduced at this position using an in-vitro site-specific mutagenesis method. The modified genes were expressed in an . Q & T7 expression system. Mutant TNF in crude bacterial extracts was tested for cross-reactivity with polyclonal anti-TNF antiserum, competitive binding with recombinant mature human TNF (rhTNF), and cytotoxicity against murine L-M cells. Substitution of Leu156 by amino acids with short hydrophilic side chains (Thr or Asn) produced TNF that was present as a 17 Kdalton protein in Western blots; however, these mutants were not competitive with rhTNF in a radioimmunoassay or a radioreceptor assay, and had no cytotoxic activity. On the other hand, mutant TNF in which hydrophobic residues (Val) or residues with long, charged side chains (Lys) were substituted for Leu156 displayed characteristics identical or very similar to those of control TNF. Zealand. The precise cellular origin of the LA28 cell line derived from a patient with nodular sclerosing Hodgkins Disease is unclear, but our recent data suggests notable similarities with cells of the human dendritic cell (DC) series.The LA28 cell line shares a similar cell membrane phenotype with human DC and stimulates allogeneic T cells strongly in MLRs during which L428-T cell clustering is evident. We have shown that tonsil DC stimulate allogeneic and autologous T cells in the absence of interleukin 1 a or fl (IL-la or 8).

Therefore the expression of cytokine mRNA in LPS stimulated and unstimulated LA28 cells was investigated by Northern Blot analysis using specific probes for IL-la and IL-Ifl. No IL-lfl mRNA was detected in LPS stimulated or unstimulated L428 cells although small amounts of IL-la message were detected (approx. 20% of control monocyte preps). However no IL-1 activity was identified in L428 supernatants using either biological or ELISA assays. Futhermore, the possible involvement of membrane bound IL-la in T cell activation was excluded by the finding that L428 cells stimulated allogeneic T cells effectively in the presence of polyclonal antisera to IL-la and IL-Ip, which were known to block functional cytokine activity. We conclude that IL-1 is not essential for primary T cell activation by either D C or the L428 cell line. Instead, it appears from experiments on isolated human cells that IL-1 may increase the antigen presenting capability of DC for autologous immune responses.

Hodes, Experimental Immunology Branch, NCI, NIH Bethesda, MD 20892 Previous work from Fitch and co-workers (Nau, G . J . , et. a1 , J.Immuno1. 141:3557-3563, 1988) has shown that stimulation of murine helper cell clones through the T cell receptor (TCR) can result in inhibition of the proliferative response to exogenous IL2. The mechanism underlying this inhibition is unclear since it has been shown that inhibition is not associated with a decrease in endogenous IL2 production, release of soluble inhibitory factors, or a decrease in expression of IL2 receptor a-chain (IL2R-a, p55) . We have studied cell surface expression of IL2R-a and IL2R-B (p70-75) by covalent cross-linking of radioiodinated IL2 to its receptor, followed by protein electropheresis. We have found that the ratio of ILZR-p to IL2R-a is decreased in murine TH1 and TH2 clones stimulated in vitro with immobilized 145-2611, an antibody to murine CD3-a. The ratio of ILZR-p/ILZR-a expression was increased slightly (1.21 times the ratio observed in unstimulated controls) in cells stimulated with IL2 alone, and was reduced to 0 . 4 8 times control in cells stimulated with anti-CD3, and 0.56 times control in cells stimulated with a combination of anti-CD3 and IL2. Since IL2R-P is an integral component of the high affinity receptor, and may be necessary f o r signal transduction, we propose this finding as a possible mechanism for the inhibition of the proliferative response to exogenous IL2 seen with supra-optimal TCR mediated stimulation. In addition, because endogenous IL2 plays a role in TCR mediated stimulation, this finding may also explain the inhibition of proliferation noted in response to supra-optimal TCR stimulation in the absence of exogenous lymphokines. In the presence of anti+ antibodies (anti-pAb), monoclonal B lymphocytes from patienls suffering from B type chronic lymphocytic leukemia (B-CLL) can respond to interleukin (IL) 2 , B cell growth factor and interferon alpha. In contrast to the effect it exerts on normal B cells, IL4 does not promote DNA synthesis by B-CLL lymphocytes. Rather this interleukin inhibits the response to IL2 in all patients' cells which responded to this interleukin. We thus examined whether IL4 would modulate the number and/or the affinity of IL2 receptors. A 3 day activation of cells by anti-pAb induced a few hundred high affinity IL2 receptors (HA-IL2R) on B-CLL cell surface, as determined by Scatchard analysis. Treatment of cells wilh IL4 caused a marked decrease in the number of HA-ILPR without affecting the amount of p55 chain of IL2 receplor (determined by flow cytometry). Moreover, after HA-IL2R were induced in a 3 day culture in the absence of IL4, one hour incubation al 37°C with IL4 still causes a decrease in their number, without interfering with the binding ability of IL2 at 4°C. Thus, IL4 appears able to influence the binding function of HA-ILZR. The clear correlation we observed between the HA-ILPR expression and the functional effect of 114 on B-CLL lymphocytes responsiveness to IL2 suggests that this interference is of functional significance. Lymphokine (LK) production following exposure to foreign antigen is a central, but poorly understood determinant of the subsequent immune response. Naive splenic lymphocytes exposed to Leishmania parasites proliferate weakly during 1" (3-5 barely detectabley-IFN and no IL3 or IL4. 1" blasts proliferate in response to exogenous IL2 but weakly to IL4 or ILI(+ILl.f-IFN production is unaltered under such conditions. Cells recovered from this 1" culture mount a vigorous 2 ' proliferative response when provided with fresh antigen and APC (30DORk Thyl-spleen cells). This is accompanied by IL2,IL3 and /-IFN production but still no IL4. Depletion of CD8+ cells during the 1" culture led to elevated 2" proliferation and increased IL2, IL3 and)'-IFN production, but IL4 was still not detected i n 1" or 2 ' culture supernatants. Increased responsiveness to IL4 was seen in CD8-1' blasts, but without any synergistic effect of IL1. These results were not due to an increased proportion of CD4+ cells in the 1' cultures. Such data illustrate the restricted LK repertoire of the developing immune response and that CDB+ T cells, though downregulating expansion of the CD4+ compartment, do not appear to influence the balance of Y-IFN vs. IL4 production by them. Experiments in progress using defined APC and anti-LK MAb will determine whether this balance reflects innappropriate APC signalling or LK (e.g. f -1FN) mediated regulation. These considerations will be of importance if restricted activation of T effector pathways is required for efficient vaccination against this and other parasites. day) culture and produce IL2, We have previously shown that induction of IL-1 p mRNA in normal human monocytes during anti-CD3 mitogensis occurs rapidly (1 h) , requires activated T cells and is not mediated by binding of anti-CD3 mAb/T cell complexes to Fc receptors on monocytes. We have further characterized the IL-1 inductive signal and now demonstrate that cell contact between T cells and monocytes is required for optimal early ( 5 4 h) induction of IL-1 a and p mRNA, although a later phase ( 2 12 h) of low level IL-1 mRNA induction occurs as a result of soluble factors released during anti-CD3 mitogenesis. The late phase of IL-1 mRNA induction coincides with the appearance of detectable IL-2, IFN-8 and TNF-a in culture supernates. Class I restricted (CD8+) and class I1 restricted (CD4+) T cell populations induce comparable levels of IL-1 p synthesis in the anti-CD3 mitogenesis model. Furthermore, both naive (CD45RA*) and memory (CD45RO') T cells are able to induce IL-1 J3 synthesis.

These studies demonstrate that the T cell IL-1 inductive signal 1) involves direct cell contact between T cells and monocytes, 2) does not require obligate binding of class I 1 MHC antigen on monocytes for signal transduction and 3) can be mediated both by naive and previously primed T cells. Studies are in progress to determine the molecular identity of the T cell IL-1 inductive signal involved in anti-CD3 mitogenesis. A highly sensitive technique [MAPPing] has been developed to simultaneously analyse the array of messenger RNAs made by small numbers of cells. MAPP utilizes a micro-procedure for isolating RNA from one to one million cells followed by two enzymatic steps: reverse transcription of total cellular RNA to produce cDNA followed by enzymatic amplification of cytokine-specific DNA fragments using specific primers and the polymerase chain reaction. Application is made to populations of lymphoid cells (T, B, NK) and macrophages. Thirty cytokines (and their receptors) and other inflammatory proteins were simultaneously MAPPed; [including: interleukins 1 -8, TGFs, TNFs, interferons, various growth factors, heat shock proteins, extracellular matrix proteinases, etc.]. The presence of individual mediators was compared to the production of specific cytokines using functional bioassays and ELISAs. IL-2 and interferon gamma can be detected from single T lymphocytes. The role of various second messenger pathways in the expression of cytokines is easily studied using the technique. The patterns of cytokine mRNAs expressed by cells derived from tissue of patients affected with inflammatory diseases such as rheumatoid arthritis and ARDS will be displayed. MAPPing permits definition of complex mixtures of cell activation molecules without the need to perform individual functional assays. This will faciliate an understanding of which factors alone or in specific combinations mediate the inflammatory response. Interleukin 2 (IL-2) was originally thought to be the only T cell growth factor, but subsequent work has shown that IL-4 also has this property. IL-I, GM-CSF, IL-6 and TNF have also been shown to influence the proliferation of some mature T cell populations. Recently we observed that IL-3, a growth factor that acts on multilineage precursor and progenitor cells, stimulates a subpopulation of peripheral T cells with the rare phenotype of C D 4~ CDX-TCR up. We therefore tested whether IL-7. a newly cloned cytokine, which similarly stimulates lymphoid precursors of the B and T lineage. acts on mature human T cells. We observed that T cell clones of four different phenotypes, namely CD4+ TCR up. CDB+ TCRaP, CD4-CDX-TCR ab and yS were stimulated to proliferate by recombinant IL-7, as were peripheral blood mononuclear cells (PBM) and purified T cells, with or without costimulation with anti CD3 antibody. Our results thus show that IL-7 acts on both the mature and the immature pool of human T cells. T cell growth may he influenced by cytokines either directly, or indirectly by stimulating the production of other growth factors and their receptors. There is evidence that IL-7 may act trough an IL-2 depenedent machanism. In view of the potency of the IL-7 effect on T cell growth. we investigated whether this may he mediated indirectly by IL-2 acting on IL-2 receptor. However experiments performed in the presence of saturating concentration of anti IL-2 antibody, anti IL-2 receptor antibody andlor Cyclosporin A were unable to block the proliferative response mediated by IL-7. These data indicate that other lymphokines, besides IL-2 or IL-4. control the clonal expansion of human mature T cells. Forskolin and cholera toxin, two activators of adenylate cyclase, as well as 3-isobutyl-1-methylxanthine (IBMX), a cyclic nucleotide phosphodiesterase inhibitor, and Prostaglandin E2 ( P G E z ) , an arachidonate metabolite known to increase the levels of cyclic AMP in fibroblasts. induced a dose-dependent Inhibition of collagen synthesis and had additive effects with IL-I. The inhibition of collagen synthesis exerted by Interleukin 1 was accompanied by a huge secretion of PGE2 in the culture medium ; however. the inhibitory effect of IL-I on collagen production persisted when PGE2 release was blocked by indomethacin. This latter drug could not further prevent the inhibitory effect exerted by forskolin, cholera toxin or IBMX. suggesting :he involvement of cyclic AMP, even when PGE2 secretion was abolished.

Taken together, these results suggest that IL-1 inhibition of collagen synthesis may be mediated in part by PGE2 and subsequent CAMP accumulation. but also by a direct action of IL-l on intracellular cyclic AMP levels. through a direct activation of adenylate cyclase or through Gscoupled proteins. For t h e s e s t u d i e s , t o t a l c e l l u l a r RNA and c u l t u r e s u p e r n a t a n t s were o b t a i n e d from lymphocytes a c t i v a t e d w i t h h r I L -2 ( r a n g i n g from 1 0 -5 0 0 U / r n l ) and r h T N F -/ j ( r a n g i n g from 0 -5 0 0 0 u/ml). S t e a d y s t a t e TNF-u mRNA was measured 1 -3 d a y s a f t e r t h e a d d i t i o n of r h T N F -0 u s i n g S1 n u c l e a s e a n a l y s i s Neonatally-induced tolerance of Class I1 alloantigens in mice of the A strain background is maintained primarily by active suppression, with little or no contribution from clonal deletion. While a majority of Class 11-tolerant mice (A.TL to1 A.TH, A.TH to1 A.TL) retain tolerogen-reactive lymphocytes detectable in the mixed lymphocyte reaction, they fail to develop tolerogen-specific cytotoxic lymphocytes (CTL). The MLR responding cells of tolerant mice differ from those in the normal mice, in that the tolerant MLR is dominated by a high frequency of IL-4-producing cells while the normal MLR is dominated by IL-2-producing cells. Recent experiments have suggested that this shift in the functional properties of T,, cells may play a role in the active suppression of specific immune effector function. In coculture MLR experiments, lymphocytes from tolerant mice suppressed two important effector functions of normal lymphocytes responding to tolerogen: the production of tumor necrosis factor (TNF) and the generation of CTL. The addition of recombinant IL-4 to normal MLR was shown to suppress the production of TNF, but not the generation of CTL (in the presence of exogenous IL-2). Therefore, i t appears that suppression in neonatallyinduced Class I1 tolerance is mediated by a unique profile of lymphokines secreted by tolerogen-specific cells. Current experiments are exploring the possibility that suppression of tolerogen-specific CTL responses is dependent on a lymphokine other than IL-4. CD 233 HUMAN BLOOD-DERIVED DENDRlTlC CELLS DO NOT PRODUCE IL-1. IL-6 OR TNF-a, Jukka M. Vakkila. Marja Sihvola and Mikko Hurme, Department of Bacteriology and Immunology. University of Helsinki, Helsinki, Finland. Little is known about the necessary non-antigen specific signals delivered to T cells by DC. Recent data shows that several monocyte-derived factors like IL-1-a, IL-1-p, IL-6 and TNF-a have enhancing activity in T cell proliferative responses. We studied the production of the above-mentioned cytokines and their mRNA by DC and monocytes separated from human peripheral blood. The intracellular expression of the proteins were studied at a singie cell level. which allowed a precise identification of the cytokine producing cell. The supernatants and cell lysates were studied with ELISA (IL-l -a, IL-1-p and TNF-a). Northern blotting analysis were used to detect the mRNA (IL-1-a, IL-1-p and IL-6). Several approaches were taken to stimulate the production of IL-I-a. IL-1-p, IL-6, and TNF-a by DC. These included the incubation of the DC either in the presence of LPS, rlL-1 or monoclonal anti-HLA-DR antibody, or the stimulation of cells with resting allogeneic T cells. None of the stimuli were able to induce the production of I L -l a , IL-1-p, IL-6 and TNF-a by DC, whereas LPS-stimulated monocytes (fresh and 24 h cultured) were strong producers of all of the above-mentioned cytokines and their mRNA. Thus we concluded that IL-1-a, IL-1-8, IL-6 and TNF-a are primarily of monocyte-derived factors and that these factors are not needed or produced during the activation of resting T cells by DC. (P170) is a 170 kD cell membrane protein associated with multidrug resistance. The quantity of this protein in tumor cells may be important in the sensitivity of cells to drug uptake and metabolism. The doxorubicin drug sensitive human myeloma cell line, 8226/S, and its doxorubicin resistant derivative cell line, 8Z6/D0x4,, provide an excellent model to study the relationship of PI70 to chemotherapeutic drug resistance. Leukoregulin (LR), a 50 kD cytokine, increases plasma membrane permeability and uptake of doxorubicin within minutes in many tumor cells. Membrane permeability is increased within 5 minutes and doxorubicin uptake is increased within 15 minutes in LR treated K562 human erythroleukemia cells. 8226/S and 8Z6/Dox4, cells are also sensitive to the membrane permeabilizing effect of leukoregulin. The sensitive cell line (8226/S) requires less LR and less time to achieve maximum membrane permeability than the resistant cell line (8226/Dox4,). P170 glycoprotein expression was studied by flow cytometry using P-glycoCHEK FITC-C21 S,(Centocor, Inc., Malvern, PA), a fluorescein labeled murine monoclonal antibody which reacts with the P170 epitope expressed on the internal surface of mammalian cell membranes. K562 cells incubated with 2 units of LR for 2 hours showed a 40% decline in the expression of P-glycoprotein. The kinetics and reversibility of LR modulation of P-glycoprotein in K562 and 8226 cells and the effects of other cytokines on epitope expression were also examined.

CD 301 NEUROBLASTOMA CELLS PRODUCE IL-6, Florence M. Hofman, Neil Sidell, David R.

Hinton, Jean E. Merrill, Department of Pathology, USC, Los Angeles, C A 90033 and Departments of Pathology and Neurology, UCLA, Los Angeles, C A 90024. Human IL-6 is a 26 K d protein originally identified as a T cell derived lymphokine inducing antibody production in B cells. Recent studies have shown that glial cells of the central nervous system can also produce IL-6. In this study we report that in the two out of three human neuroblastoma cell lines tested, IL-6 was produced constitutively, and that following differentiation induced by retinoic acid (RA), the number of cells producing this cytokine was greatly reduced. LA-N-5, IMR32, and SHF were either left unstimulated or exposed to 5X10-5M RA. LA-N-5 and IMR32 when exposed to the R A demonstrated typical neurite outgrowth, while unstimulated cultures remained in cluster formation. SHF did not exhibit neurite outgrowth in the presence of RA. IL-6 production was quantitated using immunocytochemistry with a polyclonal anti-human IL-6 reagent on cell preparations grown i n culture chambers or flasks. The results show that for LA-N-5 and IMR32, there was significant staining for IL-6 (30% and 10% respectively). In the presence of RA, the number of I t -6 positive cells dropped markedly (5% and 0% respectively). No IL-6 positive cells were detected in SHF cultures in the presence or absence of RA. Cryostat sections of normal human adult brain did not exhibit IL-6 positive neuronal cells. These data indicate that primitive, poorly differentiated neuroblastoma cells produce IL-6, while the more differentiated cells show a reduced production of this cytokine. This suggests that IL-6: (1) may be a marker for the immature neuroblastoma cells; and (2) may be involved in tumor cell growth and dissemination. Dept. of Inununology, Daniel den Hoed Cancer Center, Rotterdam, The Netherlands, Dept. of Trrmor Immunology, Radio Biological Institute, ?No Health Organization, Rijswijk, The Netherlands. We studied the prerequisites for the in vivo therapeutic application of Ln vitro expanded lymphocytes, retargeted with the bispecific mAb (bsAb) against CD3 and a human ovarian carcinoma antigen EIov18. Cytotoxic T cells ( C I Z ) were generated from PBL (starting with 2 lo8 lymphocytes) derived from both healthy donors and ovarian carcinoma patients by repeated stimlation with PHA or anti43 mAb (OKT3) and IL-2. Between days 12-20 (medium, day 14) yield was 5*109 lymphocytes required for therapy. The triggerability of lymphocytes for cytolysis was tested using anti-3 expressing hybridom cells and bsnb plus ovarian carcinoma cells (IGROV) or fresh ovarian carcinoma cells. Lytic activity was optinel between days 10-20 of culture. Remval of excess bsAb prior to cytolysis does not affect the level of target cell lysis. Moreover, triggerability of the lymphocytes is not inhibited when the lysis takes place in human serum or in ascitic fluid derived from ovarian carcinoma or non-ovarian carcinoma patients (upto 100% concentration). The purpose of this siudy was 'to generate lymphokine-activated killer activity (LAK) via alternative pathways using combinations of biologic agents. The effect on established pulmonary metastases (PM) produced from 816, K1735M2 or MCA-203 cell lines was assessed using the monoclonal antibody anti-CD3, followed by IL-2 and TNF in B6C3F mice. Treatment of PM began with a single 5 pg i.p. dose of anti-CD3 on day 3 , followed by'either IL-2 alone or 4fold less IL-2 with TNF (25,000 U/day) given at 3 day intervals. PM were counted on day 14. A single dose of anti-CD3 followed by low dose IL-2 and TNF caused the greatest reduction of metastases compared to higher doses of IL-2 alone, or IL-2 t TNF. Reduction of metastases (median 91%) using the three agents in combination was observed in all tumor models (e <0.008), and was equal to or exceeded that achieved by 9-fold higher concentrations of IL-2 alone. Treatment with anti-CD3/1L-Z/TNF significantly prolonged survival, and resulted in 60 t of,mice with 816 or K1735M2 Ptl achieving long-term survival ,120 days. This was superior to single agents or other combinations. LAK and natural killer activities of splenocytes in increased following anti-CD3/IL-Z/TNF treatment, and were consistently greater than that generated with 4-times more IL-2 alone. The anti-CD3 activated splenocytes were a heterogeneous population of T-cells, with more Lyt 2 ' cells than splenoc tes from mice treated with oses of IL-2 alone. Analysis of turnof infiltrating lymphocytes TTIL) obtained from MCA-203 PM showed a greater proportion of Lyt2 cells in anti-CD3 treated mice compared to IL-2 alone (50.7 vs 38.1t), but a lower proportion of L3T4' cells (19.0 vs 4 6 . 6 8 ) . These results indicate that the sequential use of anti-CD3, IL-2 and TNF for LAK induction and maintenance potentiates anti-tumor activity, and suggests novel strategies for combination imunotherapy. Oswaldo Cruz and Instituto Nacional de Canrer, Rio d e .Janeiro, Brazil The Eosinophil Cytotoxicity Enhancing Factor (ECEF) is secreted by human monocytes and by PMAand LPSstimulated U937 histiocytic lym?homa cells, and regulates Eirosanoid produrtion and cytotoxic activity of eosinophils. Monoclonal antibodies to ECEF recognize a 13 kDa acidic protein in secretion products as well as an integral membrane protein of macrophages and resting U937 cells. Membrane ECEF (mECEF) expression correlates with acquisition of mature macrophage characteristics by U937 cells after P?IA stimulation. Positive selection of mECEF-positive U937 cells led to the isolation of a variant cell line that responds more efficiently than unselected cells t o PYA as a differentiation stiinulus, as shown by an average 4-fold increase in the number of fullv differentiated, adherent macrophages in cultures of PMA-stimulated, positively selected cells relative to PMA-stimulated controls. This is associated with an increased sensitivity to PYA and with a faster kinrtirs of r e s p o n s e mECEF is a surface marker stronglv expressed i n some myeloid leukemias, hoth of the monocvtic a n d granulocytic lineages, hhit ahbrnt from a variety o f other hemopoictic malignancies. In peripheral blood, strong d C E F expression is found i n 70'2 n f monocytes; weak expression was seen in purified, resting T lymphocytes; positivity is therefore associated with cell types known to secrete ECEF. These findings suggest that, besides regulating arachidonate metabolism and cytotoxicity in target cells, this monokine,, in a surrace-associated form, plays a r o l e in myelomonocytir differentiation and h e h a v e s a s a n activation m a r k e r . Helsinki, SF-00290, Helsinki, Finland. Both elevation of CAMP and activation of protein kinase C function as differenfiation inducing signals in the prornyelocytic H L~6 0 leukemia cells. resulting in appearance of cells with monocytic markers. We have now used these two differentiative signals to study the appearance of the inferleukin-1 (IL-1) receptor as well as the IL-10 mRNA during monocyfe differentiation. The data obtained demonstrated that both dibutyryl (db) CAMP. a soluble structural analog of CAMP, and protein kinase C activating phorbol rnyristate PMA) were able to induce the expression of the IL-1 receptor (as detected by the binding of 1'51-labeled IL-1). In contrast to this, only PMA treatment induced the expression of IL-1 0 mRNA. However, when the effect of the PMA plus dbcAMP combination was tested, it was evident that dbcAMP greatly enhanced the PMA -induced IL-10 rnRNA expression. Taken together, these data indicate that expression of IL-1 and IL-1 receptor are differentially regulated during the differentiation of monocytes.

Obrecht, Division of Oncology, Department of Research, University Hospital Basel, Switzerland CD14 antigen expression increases in monocytes differentiating into macrophages, and it is reduced by rIFNg in monocytes in vitro (Landmann, Cell Immunol 117: 45, 1988 ). In the present study CD14 membrane expression was investigated in cultures of human mononuclear leucocytes (PBL), of elutriated, purified monocytes and of blood monocyte derived macrophages which had been incubated with rIL1, rIL2, rIL3, rIL4, rIL5, rIL6, rTNFa, rGM-CSF, rIFNa and rIFNg. rIFNa and rIL2 reduced CD14 in PBL cultures, their effect was mediated by endogenous IFNg, because it was abolished by simultaneous addition of an anti-IFNg antibody. rIFNa was ineffective in purified monocytes or macrophages. rIL2 slightly enhanced CD14 expression in purified monocytes. rIFNg and rIL4 reduced CD14 expression strongly in purified monocytes, weakly in macrophages. rIL1, rIL3, rIL5, rIL6, rTNFa and rGM-CSF did not change CD14 expression in monocytes or macrophages. In summary, CD14 antigen expression is modulated only by rIFNg, rIL4 and rIL2, the cytokine effect is stronger in monocytes than in differentiated macrophages.

Grabstein. 'Seattle Biomedical Research Institute, Seattle, WA 98109, t Cornell University Medical College, New York, NY 10021 and * Immunex Corporation, Seattle, WA 98101.

We have cultured peripheral blood monocytes (PBM) from normal donors and from patients with the parasitic infection, leishmaniasis, for several months in the continued presence of CSF-1. A 100-fold increase or greater in cell numbers has been obtained with several donors. Cells have been maintained in culture for up to three months. We have measured several monocyte functions, including surface antigen expression, antigen presentation and oxidative burst activity. Progressive changes have been noted in monocyte functions with time in culture. Fresh monocytes have high respiratory burst activity upon stimulation with PMA, ionomycin and F-met Leu-Phe (FMLP). With time in culture, the ability of these cells to undergo increased oxidative metabolism in response to ionomycin and FMLP was significantly reduced although the response to PMA remained strong. Antigen presentation by long-term (3 month) cultured mono tes from a leishmaniasis patient was also observed. Cultured monocytes from this patient were exective in presenting leishmania antigen to an autologous T-cell line in a proliferation assay. The ability of these cells to present antigen did not decrease with time in culture. FACS analysis of cultured monocytes revealed that the expression of monocyte markers varied. After approximately 8 weeks in culture, the cells became UCHMl negative, and approximately 50% of the cells were LFA-1 positive, and 60% expressed DR. Recombinant leukemia inhibitory factor (rLIF) and recombinant interleukin-6 (rlL6) induce expression of Fc receptors (FcR) and interleukin-2 receptors (IL2R) during the process of differentiation of murine myeloid leukemia M I cells to mature macrophages. In the present study, we show that the expression of these surface markers is regulated by distinct mechanisms involving metabolites of the arachidonic acid pathway. Cytofluorometry utilizing monoclonal antibodies specific to FcR and IL2R revealed different kinetics of expression of these surface markers. Expression of FcR increased continuously after induction with rLlF or rlL6 and plateaued at 48 hours. In contrast, expression of the IL2R reached its peak 24 hours after induction and declined thereafter almost to background levels at 48 hours. Addition of prostaglandins PGEz, PGDz, and PGFz a with these cytokines suppressed IL2R but not FcR expression at all time points. Addition of indomethacin, an inhibitor of prostaglandin synthesis, together with rLlF and rlL6 enabled continuous IL2R expression up to 48 hours, presumably by blocking prostaglandin synthesis in differentiating M I cells. Interestingly, indomethacin had no effect on FcR expression. Adding back PGEz, PGDz. and PGFza after inhibition of prostaglandin synthesis restored the original kinetics for IL2R expression, while FcR expression was not affected. Our data suggest that expression of IL2R in differentiating M I cells is influenced by arachidonic acid metabolites and follows a different regulatory pathway than FcR expression.

NY STIMULATING FACTORS. Kathleen T. Shiverick, Theresa Medrano, Susan Ogilvie and James R. Zucali*. Department of Pharmacology and Therapeutics and Department of Medicine*, University of Florida College of Medicine, Gainesville, FL 32610, USA Recent evidence suggests a role for hematopoietic colony stimulating factors (CSFs) in placental growth and development. The present study investigated whether rat placental tissue secretes factors which stimulate granulocyte-macrophage colony (GM-CSF) formation in v i m . Gestation day 15 placentas were dissected into separate decidua basalis, basal zone, labyrinth and yolk sac tissues which were cultured in Eagle's Modified MEM for 24 hours. Placental-conditioned medium was then bioassayed for CSF activity using rat bone marrow cells grown in 2% methylcellulose with 10% fetal calf serum; positive controls for CSF activity were endotoxinstimulated rat serum. Decidua-conditioned media (DEC-CM) had the highest CSF activity, which was retained after dialysis. The morphology of DEC-CM-stimulated colonies showed 60% M, 6% G and 34% mixed GM. Explant culture of DEC in the presence of tunicamycin resulted in a concentration-dependent loss of CSF activity in DEC-CM. Ammonium sulfate fractionation of DEC-CM isolated CSF activity in the 80% precipitate and supernatant fractions. A possible paracrine role for DEC-CSFs was further examined in experiments which characterized the effects of recombinant CSFs on trophoblast protein synthesis. Explants of basal zone tissue were cultured with [35S]methionine in the presence of 100 U/ml rec murine GM-CSF or rec human CSF-1 for 24 hours. Analysis of radiolabeled proteins in the culture medium by 2-dimensional polyacrylamide gel electrophoresis showed that mGM-CSF markedly increased the secretion of Mr 22,000 rat placental lactogen 11, while hCSF-1 increased secretion of the Mr 30,000 prolactin-like protein-A. Thus, data suppon the hypothesis that maternal decidual tissue produces CSFs which may have regulatory effects on adjacent fetal trophoblast cells in the rat placenta.

Yael Porat, Department of Human Microbiology, Faculty of Medicine, Tel-Aviv University, Tel-Aviv, Israel. Murine B cell lymphoma 1.29 cell lines can be stimulated by LPS, IL-4, IL-5 apd other reagents to secrete Ig and to differentiate from I@+ to IgGZa', IgGl', IgE , or I@+ cells. In the present work the effect of IFN-r on the Ig secretion, proliferation, and differentiation of 1 . 2 9 cell lines were examined. No direct or indirect effect of IFN-r on the cells growth were detected. IFN-r induces in LPS stimulated 1.29 cells isotype switch and secretion of IgGl. In addition IFN-r induces augmentation in the secretion of IgM, and IgA. and has no effect on the secretion of IgGPa, IgGPb or IgG3. Simultaneous stimulation of the lymphoma cells with IL-4, IFN-r. and LPS reveals that IFN-r antagonized the effects of IL-4 in elevation of IgM and total Ig secretion and abrogates the secretion of IgE. IFN-7 synergizes IL-4 in IgGl secretion. Preincubation of the cells with IFN-r for 48 hrs prepares them to secrete IgCl on subsequent stimulation of the cells with LPS. Similar observation of preeducation of the lymphoma cell lines with IL-4 or IL-6 were obtained. INTERFERON-GAMMA CONTRIBUTES TO THE LETHALITY OF ENDTOXEMIA, Frederick P.Heinze1, Department of Medicine, UCSF Medical Center, San Francisco, CA 94143. Experimental animals injected with lipopolysaccharide (endotoxin) develop widespread and fatal tissue damage that is mediated by proinflammatory cytokines. Pretreatment of CD-1 mice with 20 wg of rMuIFN-y for 2 consecutive days significantly diminished survival compared to control mice upon challenge with 1 mg of Salmonella enteriditis LPS (60% vs 0%). MuIFN-y pretreated, endotoxemic mice had significantly greater serum levels of TNF (10.4 k 3.1 units) compared to control endotoxemic mice (1.2 k 1.5 units). Spleen, lung and liver TNF mRNA levels were either unchanged or slightly diminished in rMu1FN-ypretreated mice relative to endotoxemic controls, suggesting that rMuIFN-y enhanced the translation or secretion of TNF; IL-1 beta mRNA levels were consistently decreased in these same tissues. Recombinant MuIFN-y did not induce monokine message during endotoxemia in formerly nonexpressive tissue (brain, intestine and kidney), Because the modulatory functions of administered rMuIFN-y may not predict the function of IFN-y produced endogenously after endotoxemia, we pretreated animals with monoclonal anti-IFN-y. This significantly increased survival (71% vs 14%), suggesting that IFN-y may mediate both harmful modulatory and effector functions during endotoxemic shock. Preliminary studies show that the induction of short-term endotoxin tolerance correlates with greatly decreased IFN-y expression in vivo following endotoxin challenge, further supporting a deleterious role for this factor during sepsis.

Jerrells, V. Brown, and J. Aronson, Dept Path, Univ Texas Med Branch, Galveston, TX 77550 Infection of strain 13 guinea pigs with Pichinde virus results in a progressive infection characterized by cachexia and ultimately death. Studies have shown that this virus has a prediliction for macrophages [Mq) . The present study was done to determine if infection of susceptible animals results in production of monokines, especially tumor necrosis factor (TNF) and interleukin-1 (IL-l), and to correlate monokine production with viral pathology. At various times after infection of strain 13 guinea pigs with virus levels of TNF and IL-1 in sera and M@ culture supernatants were determined. It was found that progressive viral infection was associated with increasing levels of s e r m monokines. Further, M@ isolated from infected animals produced demonstrable levels of TNF and IL-1 without further stimulation. M@ isolated from noninfected guinea pigs or mice were found to support viral growth that was associated with the production of TNF. Our data support the idea that uncontrolled viral growth in permissive MC# results in over production of TNF and IL-1. It is possible that these monokines in large amounts are at least partially responsible for the pathology associated with arenavirus infections. Supported in part by a McLaughlin Fellowship to JA.

Heidrun Moll, Christian Bogdan, Kerstin Binoder and Martin Rollinghoff. Institut fur Klinische Mikrobiologie, Universitat Erlangen-Nurnberg, Wasserturmstrasse 3, 8520 Erlangen, Federal Republic of Germany. We have assessed the role of tumour necrosis factor-& (TNF) during murine cutaneous leishmaniasis. No TNF activity could be detected in the serum of mice infected with Leishmania major but significant levels of TNF were released by spleen cells from infected mice after in vitro restimulation with L.ma'or promastigotes. After challenge with bacterial endotoxin, TNF activity titers appeared to correlate with the course of cutaneous disease in susceptible and resistant mice. The evidence for TNF production by L ma'or activated spleen cells in the absence of endotoxin suggests t h a t & , produced in response to infection and thus is involved in the immunoregulation during cutaneous leishmaniasis. Furthermore, our study indicated that the elicitation of L.ma'or-induced TNF activity by macrophages is dependent on the p r e s e n c h l cells. TNF did not exert a direct leishmanicidal effect in vitro. In combination with IFN-r, TNF induced elimination of parasites in infected macrophages, whereas it promoted parasite burden in the presence of I L -4 . These findings suggest that TNF acts in concert with other cytokines produced during L.major infection and that its role depends on the composition of T cell subsets and cytokines present. Both rodents and humans demonstrate decreased mitogen induced lymphoproliferation with increasing age. However, the level of decrease is not consistent among all individuals. Most elderly humans (mean age 85) demonstrate about 50% the level of response of the young: while 10-15% demonstrate <20% the proliferative response of young and 15-20% demonstrate prolifertive responses comparable to young.

Using Brown Norway (BN) rats, heterogeneity of lymphoproliferation was also observed. However, greater heterogeneity was observed between BN and Fisher 344 rats, raised under comparable dietary and environmental condistions, than within each strain, suggesting that genetic background may have the greatest influence upon this heterogeneity. Mean IL-2 production and IL-2 receptor expression is decreased in both humans and rats; mean IFN-g production during mitogen stimulation is decreased in humans, but increases in rats, with increasing age. Lymphokine production, however, is heterogenous among individuals in both rats and humans. Further, addition of exogenous IL-2 and/or IFN-gamma only restores the level of proliferation in 30% of elderly humans and 10% of elderly rats.

Therefore, the heterogeneity observed in individuals may reflect various levels of defects in different individuals. (Supported by AGO3934 and AG07719). Drag=*. m w t %f ~mrmnology, ~n s t i t u t e for ~x p e r i m m a Gerontology TKI, Rijswijk, The Netherlands and Aged mice have a diminished capcity to pzxduce IL-2 a d this is probably one of the mjor causes of the decrease in m e reactivity during senescence. To obtain mre insight into the exact nature of this @encm?non, w= studied purified CD4+ T cells fran young and aged W R i j mice (15 and 130 weeks of age, respectively) w i t h rqard to their ability to secrete various qtokines in response to alloantigens or m i t q e n s . In response to allcantigens a 4fold 1 -IL-2 prcduction was found with cells fran aged mice as canpareed to I L 2 @uction by young cells, and this was only in prt explained by a decrease in the frequency of antigen-spcific CD4+ T cells. I L 2 praluction i n respmse to anti-CD3, Con A or the cambination of PMA and ionanycin was also I -.

These latter findings could be attributed to a diminished expression of CD3 and were in prt also explained by intrinsic defects : cells f m old mice were less capble to increase cytoplasmic calcium in respnse to a n t i -CD3 or ioncsnycin. In addition, we obsenred a shift in the functional capacity of CD4+ T cells since supernatants of stirrmlated old CD4+ T cells consistently contain~A in-sed arcpunts of IL4 and interferon-gm. This was independent of the activation pathway tested. Similarly, an incrased prcduction of cyt~kines active on the IL-3/IL5 dependent LyH7-Bl3 cell line was observed. We therefore conclude that an age-related shift fran primq to mre xmture types of CD4+ T cells, and a s a consequence a disturbed lynphokine network, is also respnsible for the declined I L 2 pduction found in aged mice.

Holland Biotechnolcgy bv, Leiden, The Netherlands.

TNO Primate Centre, Rijsjwijk, The Netherlands Acquired immunity to blood stage infection wilh certain murine malaria species has been demonstrated to occur by an antibody-independent, cell-mediated mechanism which requires T cells and an intact spleen. The role of interferon-y (IFN-y), a pluripotent lymphokine capable of activating macrophages, in acquired immunity leading to control and elimination of the intraerythrocytic parasites was investigated during infection with P. chabaudi AS. C57BL-derived, C57BUIOScNHsd (El 0). which are resistant to infection wilh P. chabaudi AS, were treated with anti-IFN-y monoclonal antibodies (mAbs)

However, mAbs-treated animals also completely cleared the infection by 4 weeks. Thus, these results suggest that OF INTERFERON INDUCERS AND OTHER BIOLOGICAL RESPONSE

Both of these interferon inducers were more effective when treatment began after virus infection, rather than before. In contrast, MVE-2 had antiviral activity only when given prophylactically and, on several occasions, actually exacerbated the disease when given therapeutically. Suboptimal concentrations of MVE-2 (single dose of 3 or 12 mg/kg one day before infection) and AZT (3 or 10 mgikg. qd), when combined, produced an additive antiviral effect. Another BRM, soluble glucan, did not have antiviral activity. This lack of activity might be explained by the observation that, even though glucan boosted NK activity in uninfected mice, it depressed NK activity in RLV-infected mice. if this explanation is correct, it suggests a role for NK ceiis in resistance to retroviruses. Although poly [I.C]-LC had therapeutic activity in the RLV model, short-term treatment of LP-BM-5-infected mice with poly [I,C]-LC did not reverse the immunosuppression caused by this MAIDS virus

CA 92121. A Consensus peptide sequence "QNRRGLDxLxxxxG

This 5ynthetic peptide, called pISEp':ISP, is composul (it the ISP q u e n c e troni teline leukemia vinis attached at the NH2-termmuc to a polar psitive q i i e n c e also predicted to hL. exposed at the viral membrane wrface. The activities include 40.50% niaxmial siippression (6pM to 19pM peptide) of I3HlTdR uptake hy concanavalin A stunulated mmme splenocytes, inhihition (50% at 6vM peptide) of heniolytic plaque tomiation and inhihition of VUUI release from chronically mfected cell lmes. Anti-vual activity was evaluated in FeLV-mfected and Rauscher MuLV-mfected cell lmes hy two methods: protein hlot cumparirm ot cellular and vinis-localuul p27gag/p30gag. and inhihition of \'inis release, as measured by RT activity, mto conditioned growth mediuni. Both a/p-interferiin and pl5Ep':ISP. hot not an inactive peptide homolog (plSEp+:PSI), hlirkul in a niintoxic manner the cdlular release of p308"Z. Half maximal mhihition ot viral RT actwlty m conditioned growth mcdluni was detected at 6vM to 19vM peptide. The pl5Ep':ISP. hut not @-intrrteron or pl5Ep':PSL a l w directly mhihited viral RT activity We suggeqt that b u d pl5E and alpha interferon may iitilue a siniilar mechanism t o effect some anti-vual and cytostatic effects

STIMULATION OF CYTOKINE RELEASE, L.-T. Rasmussen and R. Seljelid, Institute of Medical Biology, University of Troms~, 9000 Norway. Aminated D1-3D polyglucose (AG) and 01-3D polyglucose attached to plastic microbeads (GDM) are immunomodulators with strong effects in vivo and in vitro.(Seljelid, R. Tumor regression after treatment with aminated 01-3D polyylucose is initiated by circulatory failure. Scand. J. Immunol. 29, 181-192, 1989 ; Rasmussen, L.-T., Seljelid, R. The modulatory effect of lipoproteins on the release of interleukin 1 by human peritoneal macrophayes stimulated with 01, 3-D-polyglucose derivatives. Scand. J. Immunol. 29, 477-484, 1989 .) AG causes total regression of Meth A sarcoma in mice. This effect can be understood in terms of a concerted action of local and systemic cytokines.AG and GDM protect against otherwise lethal infections in mice. Also this can be explained partly by the effect on AG and GDM on the release of cytokines and prostaglandin E2.

IL In order to evaluate the role of I L -1 as a potential adjuvant for vaccines we determined the ability of this cytokine to enhance the humoral immune response to Hepatitis B surface antigen (HBs Ag). Groups of Balb