key: cord-0010402-2k2sc354 authors: Fleming, John O.; Ting, Jenny Y.P.; Stohlman, Stephen A.; Weiner, Leslie P. title: Improvements in obtaining and characterizing mouse cerebrospinal fluid(): Application to mouse hepatitis virus-induced encephalomyelitis date: 2002-12-09 journal: J Neuroimmunol DOI: 10.1016/0165-5728(83)90017-6 sha: 3f3f6e459432582506aee02c5ad16fe99d8433f9 doc_id: 10402 cord_uid: 2k2sc354 This report describes advances in techniques for analyzing cellular and humoral immune components in the cerebrospinal fluid (CSF) of the mouse that are applicable to other laboratory animals. CSF studies undertaken during experimental infection of mice with JHM strain virus (JHMV) of mouse hepatitis virus are presented. A critical pitfall which can lead to erroneous or invalid results is contamination of the CSF by even minute quantities of blood. Means of avoiding this contamination are attention to anatomical reference points, the use of a micropipet, and prior intracardiac perfusion of animals with phosphate-buffered saline. Cells in the CSF were typed as either B, T, polymorphonuclear, or mononuclear cells by the combination of a microcytotoxicity assay and histologic stains. A radioimmunoassay (RIA) allowed quantification of antibodies to JHMV in the CSF and indicated the presence of intrathecal synthesis of antibody in chronically infected mice. The combined use of these sensitive methods makes possible CSF analysis in individual mice rather than in pooled groups. The central nervous system (CNS) has been referred to as an immunologically privileged site because of the relative isolation of brain and spinal cord from the systemic vascular and immune systems (Medawar 1948) . By contrast, the cerebrospihal fluid (CSF) is in intimate contact with the CNS parenchyma and is considered an excellent indicator of metabolic (Cutler 1980) , neoplastic (Wertlake et al. 1972) , and immune (Lisak and Zweiman 1977; Trotter and Brooks 1980) processes occuring within the CNS itself. Thus, exanfination of the CSF is often of considerable clinical and theoretical importance in assessing neurological diseases, especially those with immunological features, such as multiple sclerosis or neurosyphilis. CSF analygis is also of value in the study of experimental neurological disease. With large animals such as sheep or primates, essentially the same techniques used in human CSF examination may be applied (Nathanson et al. 1979) . However, in small laboratory animals, obtaining and characterizing CSF is often fraught with technical difficulties. Because mice are commonly employed in animal models of disease, it would be useful to have a reliable means of assessing CSF in this species. Carp et al. (1971) first introduced a method ot withdrawing CSF from mice by means of cisterna magna puncture. Zinkcrnagel an,~ Doherty 0973) adapted this method to the study of sensitized iymphocytes in the CSF during lymphocytic choriomeningitis virus encephalitis. Gerhard et al. (1978) analyzed antiviral antibodies in the CSF following infection with Parainfluenza type 1 (6/94) virus, and Griffin (1981) has studied the entrance of immunoglobulin classes into the CSF of mice during acute viral encephalitis. More recently, Doherty and Gerhard 0981) have used antiviral monoclonal antibodies as probes in the aw~iysis of the blood-brain barrier. The present paper draws upon and extends these ~tudies. JHM virus (JHMVL a neurotropic member of the mouse hepatitis (MHV) group of coronaviruses, has been extensively used to study viral-induced &myelination in mice (Weiner 1973; Robb and Bond 1979; Stohlman and Weiner 1981) . As part of an on-going study of pathogenesis in this model, CSF was obtained during several experiments. A modification of the method of Carp et al. (1971) for obtaining CSF and a specially designed micropipet were utilized. A microcytotoxicity assay, together with a nonspecific esterase stain, allowed enumeration of all major cellular types within the CSF. Determination of CSF antiviral ar.tibody by RIA was performed. The means of avoiding contamination of CSF with blood or tissue, as well as the importance of non-contamination of CSF, are stressed in the techniques presented. Using these methods, CSF from individual mice may be characterized. Animals used in these experiments were 6-8 week.old C57BL/6 (136) and A/J mice obtained from Jackson Laboratories, Bar Harbor, ME. Mice were spot-checked by the solid phase RIA described below for the presence of antibody to JHMV before use in these studies. Chimeric mice were con.~tr,actecl by the metho:l of yon Boehmer et al. (1975) . Bone marrow recipients, 6 monthg of age. were irradi~tted wi~h 950 rads. Femur man'ow cells were removed from donors, treated with Aeti Thy-I plus complement as previously described (Stoh~man et al. ] 980) to remove mature T lymphocytes, and then injected into recipients at 2 x 10 7 ceils/animal. End1 animal was checked for ehimerization by tissue typing its splenic cel'ts using the m~crocytotoxicity assay described in Meth~xls. The JHM strain of mouse hepatits ~nls OHMV) was originally isolatod from a mouse with hind leg paralysis trod demyelina~ien (Cheerer et al. 1949 ). The DS strain, a small-plaque variant of JHMV, was used in this study. This virus was plaque-purified and propagated in DBT ~ells as described previously (Stohlman et al. 1982) . For studies of acute viral inf~.'fion, mice were given 1000 plaque-looming units (pfu) intracerebrally (i.c.) (chimeric mice) or 3600 p|u intranasaliy (i.n.) (B6). For studies of chronic viral infection with JHMV, B6 mice were given 1000 pfu Lc., and CSF was taken 12 months later. Control viruses used in the RIA were influen~ virus type A, mouse-adapted, kindly supplied by Dr. Peter Braylon, and herpes simplex virus type I kindly supplied by Dr. Dru Willey, both of the University of Southern California, School of Medicine. Mice were first anesthestized with methyoxyflurane inhalation (Metafane, Pitman..Moore, Washington Crossing, N J) or intraperitoneal pentobarbitoL and a thoracotomy was performed with the animals sup;ne on a dissection rack. The fight atrium was opened and the mice perfusvt through the leh '¢entricle by hand or by ~_ pressure apparatus set at 3 p.s.i, with approximately 30 nd of phosphate-buffered saline pH "].2 (PBS) viia a 26-g~mge needle. Ade~luate perfusion was signaled by blanching of the liver. The skin over the posterior n,~ek was removed by two incisions, the first in the midline from low cervical .area ~:o the ~mterior cranium and a second across the craniocervical junction, just belc,w the level of the ears. Subcutaneous tissue and nuchal muscles were exposed, sectionefl along the rim of the occipital bone, and then removed laterally unveiling the glistening clear arachnoid membrane overlying the cisterna magna (Fig, 1) . By utilizing a ,~tereo dissecting microscope, a micropipet was guided into the cisterna ma~a a:~ sl,~o,sn in Fig. 1 . "[he nfi-cropipet was comtmcl~ by cutting a 30-gauge needle witl~ scissors (Yale Incorporated, Rutherford, NJ) to approximately 6-7 mm; a thin guide vfire was pla,:ed through the needle whm cutting it, so as to avoid collapsing the lumen. The needle wgs then insert~! 4-5 mm into the lumen of a 10-or 20-/d micrcpipet (Van-Lab micropipets, VWK Sc~fific Inc., Los Angeles,CA). The junction wa~; sealed with a &'op of epoxy ~ (Foxy Poxy, Krazy Glue, New York, NY). The needle-pipet ~Lsserahly was then attached to :!h¢ mouth suction app~u atus supplied with the micropilpets. An in-line trap provided with the micropipet pre, lented aspiration of CSF. Care was taken not t,~ puncture the fourth ventricle acr,~s the inferior medullary 132. velum. Although the inferior medullary velum has only 2 cell layers at its most inferior aspect, it is a much thicker tissue as it rises to join the cerebellum, and at this level it contains ependyma, glial cells, myelinated axons, and capillaries (Williams and Warwick 1975) . Puncture of this structure thus risks blood and tissue contamination of CSF. CSF was aspirated by gentle mouth suction, and volumes were estimated by comparison with the markings on the pipet. For critical experiments, the volume may be determined by weighing tubes prior to and after addition of fluid. Usually 5-15 ttl of clear CSF were obtained. In most instances, CSF was immediatel2v diluted in tissue culture media for cell analysi~ or PBS (1:5) for antibody determine.fion (se~ below). A combination of Wright's stain (Diff-Quik, Harleco. Gibbstown, N J). nonspecific esterase stain for monocytes. (Yam et al. 1971) , and antibody-mediated. complement-dependent mict ocytotoxicity assay (F~elinger et ~1. 1974) were used. For the sgains, cells were mixed with a drop of fetal calf serum diluted I : 5 in PBS, smeared on a glass slide, and then air-dried. Alternatively, they were prepared by cytocentrifugation. The CSF sample, 10-15 FI, w~t .,tdded ~ 0.3 ml of tissue culture :23 ~tedia with 10% fetal calf serulat, vortexed, aml spun onto ;t clean dido at 800 rpm f(¢5 min with a ',~handon-,Sou them Cytospin (Sewickley, PA) It w~ found tha~t the latter method gave superiormorphololgy. For the microcytotoxicity au~y. ~ were im~Lediately counted and resuspended at appro:~imately 106/Ini. Two itl of antibody were incubated with 2/tl of cell!s for 15 rain at room temperature ht a Teratullki tray. The antibodies used were ant:i-It (C~tppei Laboratories, West Chester, PA) gad anfi..T.hy 1.2, (PL/Jx A.AKIEt)F; anti-A.AL,, kindly p;ovid~ by Dr. Jeffrey Frelinger~ (University of South,,,:rn California, S~ool of Medichle) for B cells amt T cells, respectively. Both antiser;t were tested or.~ cells from sple~;n and thymus and were ~ound to be, appropriate][y Specific. Following incubation, the antibody was rer,3oved and then replaced with 2 ~tl of rabbit complement. After 30 ~ incub~;u~on at 37°C, the cells were fixed with forn~aldehyde and viability determined by p~tse contrast microscopy. The microcytotoxity assay is ide~al for CSF analysis siw© ~o few ce~ls are required. The solid phase radioimmtmoassay (RIA) developed in our laboratmy is a modification of the methods of F [osenthal et al. (1973) and Griffin (1981) . Sezatm-l~e JHMV, typically 104-105 pfl~L/ml, was used as an immunoadsorbant. Pol:/-vinylchlorid~: flexible microtitrafion 96-well plates (Dynatech Laboratorie~ Alexandria, VA) were coated wiith 100 ttl per well of virus incubathm overnight at 4°C. The plates were next in culbated for 1 h at room temperature (R'r) with each well containing 100/LI of 0.'3~ bovine serum albumin (BSA) in PB3 pH 7.2 with (l~h'armacia, Uppsala, Sweden) was radioiodinated with 1251 by the lacio~-r0xidase method (Enzymobeads, Bio-Rad, Richmond, I CA) to achieve a specific activity of approximately 20 ~Ci/~tg. Although a 2oh incub~.tion with [1251]protein A w~s adequate, overnight incubation gave optimal resulL~. Following incubation, plates were washed 5 × m4th RIA diluent, 5 x with PBS, and 5 × with distilled water. Individual wells were cut ¢ut and counted in a gamma counter. A representative assay is shown in Fig. 2 . The interpretation and validity of results obtained from CSF examination is critically dependent on the absence of blood contamination. As a preliminary effort to minimize contamination, the effect of cardiac perfusion by PBS was examined using erythrocytes (rbc) as markers of CSF: bleed mixin& During initial studies, CSF was obtained from mice both with and without prior per'fusion. Gross contamination of CSF by rbc was apparent by vi~,Jal inspection in only 2 of 11 mice not perfused and in none of 13 mice perfused. Nevertheless, as shown in Table ! , micrc3copie analysis showed contamination in 82~ and 15~ of the non-perfused groups, respectively. Furthermore, the magnitude of contamination in the non-perfused group was significant (1 part in 1000), even when the 2 grossly bh~dy samples were excluded from consideration. The perfused group, by contrast, showed insignificant (< I : 105 parts) contamination by microscopic criteria. Therefore, in the subsequent studies described all mic~ were perfused prior to CSF sampling. Acute oiral infection B6 mice were first infected using the intr~nasal route of infection, which, while less efficient than the intr~¢erebral route at causing di~ase, does not tv.sult in direct trauma to the blood-brain barrier. Twenty-three of 79 mice inoculated devdop~ pfu of JHMV i.e. and subject to CSF analyses I year later. b Ratios were calculated by the following method: CSF ¢x~unts were detemlined at I : 5 dilution. Serum samples were measured at several dilutions, typic~ly undil:~ted, i :5, I : 10, I : 100o I : 1000. A stand~d curve was constructed for each serum sample. ~ p~t~ were linear on ~'miio 8 paper in the relevant ranges. The dilution of serum which would give the s,tmm cpm as the CSF w~ found and t;m tiler expressed as a ratio of the two dilutions. For example, if a CSF sample yielded 100 ¢pm at I : 5 dilution and the corresponding serum gave 100 clnn at a i :500 ~Ji|utioll. the CSF:~ru:ra ratio was expreumd as I : I00. etai. 1974) . Mice with,nut clinical signs had no detectable cells m the C~F. Histologic examination of ~dl mice infected by the intr~nasai route showed ¢¢iden~ of interstitial pneumoniti.,~; sham-infected animals sho~ ed only minimal imlmoLuT clmnges. Examination of the brains of these groups of mice showed that mdy mice with clinical signs of illnc~.~s (see TabLe 2 ) had histologic evidence of em:epl~ri~ correlating with the finding; of pleocytosis in the CSF. To determine the origin of cells found in CSF, chimeric mice were infected ~th JHMV and the cells entering the CSF examined for H-2 histocompatibility type_~ Since donor • and recipient i[tave different H-2 types, the origin of the cells found in the CSF can be determined. In the following experiment, 2 ser~ of chimeric m~¢,.were tested: (A/J × B6) 1=' I recipients transplanted with B6 cells:, and t~5 recilpi~tt~ transplanted with (A/J × :E6) Fn cells, Both groups were inoculal~d ix. 16~h JHMV, and. CSF was removed at c|ay 3 post inoculation (p.i.). Table 3 shows ~e iympb)cytes in the CSF were exclusively of donor H-2 type, confirming that the c¢lhdar infiltrate during acute v~ral mfec~ion is originally from a hematogenous soutc~ It has been previously shewn that approximately 605 o~ 136 mice surviving JHM',' infection show evidence of chronic CNS in~:ection (Stohlman and Weiner 1981) . We used our method for obtalxfing contamination-free CSF to examine a group of the~ chronically infected .mice. Table: 4 shows that all the mice tested one y,~ur post infection had significant serum antibody tkers to JHMV. In 5 o~ the I 1 mk:e antibody to JHMV was defrosted in the CSF. In at least 2 of these m/ce the ratio o~ the's.erum titer to CSF titer was less than that expected by diffusion, suggesting the possibility of specific antiv~iral antibody synthesis within the CS]F. Control mice not show detectable anti-JtiMV antibody in serum or CSF. Obtaining and analyzing~ CSF often are important in the study of clinked and experimental neurological disease. A critical problemin this procedure is contamination of CSF by blood. In noxmal humans and most other sgecies stud~t, there are CSF to blood ratios of approximately 1/350, 1/2000, and > 1/106 for inmnmoglobulins, leukocytes, and erytkrocytes, respectively (Davson 1967; Tourtt~lotte 1970; Fishman 1980) . In view of the relatively low concentrations of CSF components, small degrees of contamination by blood may.lead to hmccurate or ivvahd results. For example, mixinlg ! part blood in 350 parts CSF may falsely double the value obtained for immunoglobuilin concen~ation: In this reco~rd, the pcesem~ o[ rbc's serve as a sensitive marker of blood mixing with CSF; thua, as little as g-9 rbe's/pl of CSF indicate contamination on the order of one part in a millkm. In the past, CSF has been obtained from mice by first puncturing the aracbaoid membrane and then removing the CSF that flows onto its surfa,~e. In out view, th~ procedure entails a high risk oJF CSF contamination with blood or tissue cong~aents remaining on the arachnoicl ~ffter dissection. The original reF~rt o[ this (Carp et al. 1971 ) stated that more than 70~ of samples were virtually free of blood as judged by macroscopic appearance and hemoglobin electrophoresis. This implies that approximately 30~ of samples were not free of gross contamination and that a lesser degree of mixing may have been present in other samples. When a similar technique was applied to neonatal rats, it was reported that CSF was rarely grossly bloody but most samples did contain small numbers of rbc's (range 0-1000//~1) (Moxon and Ostrow 1977) . This degree of contamination may be unacceptable for many experiments, such as those deafing with the integrity of the blood-brain and blood-CSF barrier systems. Contamination of CSF by blood can be reduced by first exsanguinating mice; however, as Gerhard et al. (1978) have shown, there remains a residual mean mixing of blood and CSF on the order of !/500 parts. In the procedure employed in the present experiments, mice were exsanguinated and then perfused with PBS (Lipton and Gonzales-Scarano 1978) prior to removal of the CSF by direct arachnoid puncture (Griffin 1981) . As shown in Table I , blood/CSF contamination was reduced to less than a mean of 1 part in 10 s. In non-perfused mice, CSF was usually clear on visual inspection; however, microscopic analysis showed unacceptably high rbc contamination in these samples. Thus, our results confirm the usefulness of the exsanguination-perfusion procedure and point out the unreliability of macroscopic inspection as the only guide to contamination. This procedure was used in conjunction with other microtechniques to analyze CSF during acute and chronic viral encephalomyelitis. To avoid trauma to the blood-brain barrier, mice were initially infected via the intranasal route and ~.he relative numbers of T cells, B cells, PMN, and monocytes were monitored. Experiments with chimeric mice, infected i.e., showed that this methodology could demonstrate that the cells recruited into the CSF were of hematogenous origin. Finally, antiviral antibody was measured in the blood and CSF of chronically infected mice. Five of 11 mice had detectable CSF anti-JHMV antibody, and in at least 2 of these mice CSF : blood ratios implied local synthesis of antibody, a phenomenon noted in other experimental infections (Nathanson 1979), as well as human diseases such as subacute sclerosing panencephalitis and multiple sclerosis (Tourtellotte 1970). Ultrastructural examination of the chronically infected mice used in tiffs study has shown plasma cells in their subarachnoid space (Erfich, personal communicatton). The methods employed in thi~ study are appfieable to many other studies. The microcytotoxicity assay may be used to enumerate any cell type for which cytolytic antibody exists. The RIA may be utilized to determine other immunoglobulin classes and sp~ificities (Griffin 1981) . Finally, the microteclmiques described have been shown to be sensitive enough to analyze CSF from individual mice, rather than pooled s~mples. In some experiments, such as those shown in Tables 2 and 4 , pooling CSF would have obscured important individual differences. A routine ~ OHM) caa~ disseminated encephalomyelitis with extensive destruction of myeli n Ne~irochemical ~tspects of blood-brain-cerebrospinal fluid ba.rriers Gerha~'d, Backdown of the blood-cerebrospinal fluid barrier to i~ in mice injected intracerebr.'dly with a neurotropic influenza A virus Pa|:lak, The exchange of material between cerebrospinal fluid and inahL hi Cerebrospinal Fluid in Diseases of the Nervous System Evidence for the ~ of la The central nervous sysl:em-associated immmte sespome to i,arainflnenza type I virus in mi Raft', T. and B lymphocytes --Origins. properties, ~ roles in i~nmune responses lmmuaogiobulins in 1:he cerebrospinal fluid --Chang~:s during acute viral encephalitis ia ,1lice Centr',d nervous system imme~nity in mice infected with Theiler's wrus, Part I (Localizing neutnllizin 8 antibody respons,~) Immunity to homologous grafted skin, Part 3 (Fate of skin homographs tv',.ns~ to th: blain, to subcutaneous tissue, and to the anterior chamber of the eye H~'emophilus influenzae meningitis in inf,.at rats --Role of b~lesem~a in pat~ho~enesis of a~e-dependent inflammatory responses in cerebrospinal fluid Padmlgaesis of visi!a. Part 4 (Spinal fluid s~'.udies) Pathoge,ic routine coronaviruses, Virolc~y Comparison of direct ~utd indirect solid-phase minrmadioimn!unoa~says for the detection of viral antigens and antiviral antibody Comparison of normal blood picture of ymmg adults frem 18 it.bred strains of mice Chronic central n~:rvous system demyelination in mice after JHM v:nm imOec)ion Frelinger an:l L.P. Weiner, Resistance to fatal central nervous system disetse by roues., hepatitis virus, strain JHM, Part 2 (Adherent cell mediated protection) Murine coronavh-uses --lsolali,.m and cl~aracterization of two plaque morphology variants of a neurotropic strain JHM On cerebrospinal fluid immunoglobulin-G (lgG) quotien~ in multiple sclerosis and other ~iseases Pathophysiolo~ of cerebrospina~ f~'J~d immunoslobuli~ Tolerar~ to hi~uibility detmuinaats in tetr, tparental bone marrow chimeras Pathogenesis of demyelination induced by ~' . mouse hepatith virus (JHM virus), ~rch. Neurol Cytologic evaluation of cerebrospinal fluld with clinic~ and histolof, ic correlation Cytochemical identification of monocytes ,'rod gjrarmlocytes, Araer Cytotoxic thymus-derive~ lympho~tes in cerebrospinal fluld of mice with lymphocytic choriomening, itis We wist: to thank Raymond Mitchell and Josie Lopez for editorial assistance and Carol Futr,.'li for the illustration.