Identification and Immunolocalization by monoclonal antibody of NSP-5, a surface polypeptide of neural cells

Journal of Neuroimmunology. 6 (1984),111.-426

41 !

Elsevier JNI 00204

Identification and Immunolocalization by Monoclonal Antibody of NSP-5, A Surface Polypeptide of Neural Cells G. Rougon ~, M. H i r n I, M.K. Hirsch 1, J.L. G u e n e t 2 and C~ Goridis t I Centre d'lmmunologie I NSERM-CNRS de Marseille..Luminy, Case 906,13288 Marseille Cedex 9 and : Department of Immunology, Institut Pm'teur, 75724 Paris Cedex 15 {France)

(Received23 November,1983) (Revised,received24 February,1984) (Accepted24 February,1984)

Summary A monoc,onal antibody, termed anti-NSP-5 (anti-Neural cell Surface Protein-5) was obtained from an hybridoma generated by fusing rat myeioma cells vdth splenocytes of a rat immunized with membranes from the cerebella .of weaver mutant mice This antibody reacted with the surface membrane of a subset of neurones in cMtures from cerebella and dorsal root ganglia. In both culture systems, only tetanus toxin.positive cells were stained by the antibody. In sections of adult cerebellum a punctate pattern of staining was seen in the molecular layer, the Purkinje cell layer and th~ upper part of the granule cell layer. The white matter was strongly positi'/e whereas granule cell and Purkinje cell bodies were clearly negative. in sections from adult dorsal root ganglia anti-NSP-5 labe!ed most sensory neurones including their axones in the dorsal roots. The express~on of the antigen was developmentally regulated. It could not be detected in cerebellar cultures prepared from animals younger than 7 days, in good agreement with the data obtained on tissue sections. Similarly, T~heantiger~ could not be detected by immunoblotting in neonatal spinal cord, but a NSP-5-reactive band was )resent at ~ostnatal day 7. The antibody bound a Folypel)tide of axound MW 180000 in extracts prepared from adult mouse spinal cord or cerebellum. When purified by imnmnoaffinity chromatography the antigen co-e]luted with numerous

Thi~work was supportedby a grant fromCNRS(ATP7071/72). CorrespondenceR,: C. Ooridis,Cen,re d'lmmunok,gieINSERM-CNRSde Marsdlle-Luminy,case 906.13288 Ma~ill¢ Codex9, France. 0165-$728/84/$03.00© 1984ElsevierSciencePublishersB.V.

412 strongly associated polypeptides. Upon subcellular fractionation most of it remained associated with a Triton.-Xl00 insoluble fraction thus co-distributing with the cytoskeleton. Key words:

C J ! s u r f a c e a n t i g e n - C e r e b e l l u m - C y t o s k e l e t o n - D o r s a l root g a n g l i a Monoctonal antibody

Introduction 1he antigenic specificity of the nervous system is determined by intracellular or cell .~urface determinants that are expressed by different classes of cells (for review see Weiner and Hauser 1982). With the advent of the hybridoma technology and the availability of m o , ~ l o n a l a'atibodies (mAbs) against nervou~ tissue the immunological definition of nervous s~stem antigens has reached a new level of sophistication (for review see McKay et a~. 1981) and it was possible to ,address the question how many neuronal or glial sunpopulations can be distinguished by immunological methods. In our approach to defining antigens specific for classes of neural cells we focused on generating mAUs against cell surface components. Antibodies reacting with the surface membrane can be used not only to identify cell types but also to enrich cell populations by positive or negative selection (Brockes et aL 1979) and eventually to probe the function of tFe antigen by studying perturbations of cell physiology and differentiation caused by antibody treatment (P.,ichman et al. 1980: Rutishauser and Edelman 1980; Henke-Fahle and Bonhoeffer 1983; Sadoul et el. 1983). However, mosz o f the roADs described so far that are specific for sub.types of neural cells ~re directed against intracellular components. In our lab(;ratory, we have already prepared two mAbs (anti-BSP-2 and ~nIi-BSP3), directed at;ainst mouse neural surface glycoprotcins (Hirn et el. 1981, 19E2). They were ~btaine~ Uy immunizing rats either with glycoproteins extracted from neonatal mouse brair~. ~ with whole cells from cereheilar cultures, l~)th antigens were fonnd on all classe~ of neutones and on astrocytes in sections of the adult mouse cerebellum. In order to obtain mAbs with a more restricted cellular reactivity, a different s~rategy w~s chosen. Rats were immunized with membranes from weaver (wv/wv) mutant ~rebella which Jack granule cells, the major cell type of the cerebellum (Rezai and Yoon 1972; Rakic and Sidman 1973). Relevant mAbs were detected on the basis of a strongly pcr~itive reactivity with membrane~ from weaver cerebella and with glycoproteins extracted from the forcbrains of wild-type ~fice and of negative or weak reactivity with membnmes from pcd/pcd n~utant mouse cerebella which Jack Purkinje ceils and are consequently enriched in gran~le c~dls (Mullen et al. 1975, 1976). Using this strategy we have obtained ,,;everal mAbs reacting with the cell surface of subpopulations of neural cells. One of them, anti-NSP-4 (Neural cell Surface Protein-4) has been already described (RouBon et al. 1983). It reacts with a glycoproteir~ family pre~nt on subpopulations of neurones and astrocytes. This paper" describes the immunolocalization and preliminav: bio-

413

chemical characterization of the antigen defined by a second mAb issued from the same fusion. The antibody recognizes a high molecular weight (mol. wt.) protein and reacts with a limited number of neuronal types in the mouse central and peripheral nervous system. In continuation of the nomenclature used for r.he other surface proteins identified in our laboratory the antigen was termed NSP-5 (Neural cell Surface Protein-5).

Experimental procedures

Isolation of the hyb,,idoma Crude membranes from cerebella of 19-25 day-old wv/wv mutant mice raised on an inbred recombinant (B6 × CBA) background were prepared as described (Him ¢t al. 1981). Lou/W.~;i rats were immunized with membranes containing 2 mg protein in complet~ Freund's adjuvant following the same protocol as reported for NSP-4 antigen (Rot~gon et ai. 1983). Their spleen cells were fused with Y3 Agl-23 rat myeloma cells (ob:ained from C. Milstein, Cambridge, U.K.) and hybrid cells selected as usual (P~erres et al. 1982). Hybridoma supernatants were tested for relevant antibody using the dot immunobinding assay (Hawkes et al. 1982). The thre,." preparations used as sol~d phase antigens were (i) crude membranes from wv/wv ,:erebella (2 lag per dot), (ii) crude membranes from pcd/pcd cerebella (2/tg per dot) and (iii) a glycoprotein fraction prepared as described (Him et al. 1981) from the brains of neonatal C57BL/6 wild-type mice (0.5 lag per dot). The pod mutants were maintained on an inbred recombinant (B6 × CBA) background at the Institut Pasteur (Paris). Thc.v were used between the age of 22 and 28 days. Supernatant:; of hybridoma cultur¢~ strongly positive on wv/wv cerebella and on normal brain glycoproteins, but weakly positive on ped/pcd cerebella were selected. The cells were cloned twice by limiting dilution. Supernatants from mass cultu/es grown to high density were used for further study. A mAb thus derived and termed anti-NSP-5 was used in the present study. I~ was identified as a rat IgG2c by u~ing class-specific antibodies (kindly provided by H. l~azin, Louvain, Belgium).

lmmunolocali,ation of NSP-5 antigen Primary cultures of dissociated cells were prepared on polylysine-coated glass coverslips from cerebella (Goridis et al. 1978) or dorsal root ganglia (DRG) (Fields et ai. 1978) of 1-7-day-old C57BL/6 mice and grown in vitro for 2-10 da)~. Indirect immunofluorescenee was performed on live cultures essentially as described (Raft et aL 1979). Cell types were identified using double-labeling with established marke~s. Tetanus toxin (TTX)-anti-tetanus toxin serum was used to id¢~ttify neurones (Raft et al. 1979), glial fibrillary acidic protein (GFAP) as a markc~. for astrocytes (Bignami et al. 1972) and galactocerebroside as a marker for , ~ o dendroglia (Raft et al. 1979). Although some TTX + astrocytes have been detected in dissociated cultures from the CNS, they seem to be absem from cerebeUar cultures (Raft et al. 1983 and unpublished observations). TTX was a gift from B. Bir~.iv.i

414

(Institut Pasteur, Paris), rabbit anti-GFAP, anti-Tl~X and anti-galactocerebroside sera were kindly provided by M. Raft (University College, London). Species-specific tetramethyl rhodamine isothiocyanat.e (TRITCplabeled goat anti-rat lg antibodies and fluoreseein isothiocyanate (FITC)-Iabeled goat anti-rabbit It4 antibodies (both from Nordic, Tilburg, The Netherlands) were used as developing antibodies, in order to detect GFAP the ctdtures were fixed and permeabiliz~..dwith 95% ethanol/5% acetic acid (Raft et al. 1979) before application of :he antibody. The slides were viewed under a Zeiss photomicroscope equipped with an epifluore~ence attachment and photographed on Kodak Ektachrome 400 film. For in ~ivo studies, cerebella or DRG were quickly dissecled, frozen in isopentane, chilled in liquid nitrogen and cut sagitally into 8-1~m thick sections with a Bright cryomicrotome. In some experiments, the sections were slightly fixed by dipping them into acetone at - 4 0 ° C for 2 rain. Sections were then incubated with raAb overnight at 4°C. After 3 washings in saline, TRITC-iabeled goat anti-rat Ig antibodies were applied for 20 rain. When acetone-fixed sections were processed 0.1% Triton-X100 was added to the antibody solutions. After 3 more washings the sections were mounted in Ekanol. in control experiments, anti-NSP-5 antibodies were replaced by an irrelevant rat mAb of the same lgG2c class.

lmmunochemical detection of NSP-5 antigen Extracts of various p,~rts o1" the central nervous system were prepared in the prc,;ence of 2% sodium dodecy!sull'ate (SDS). The techniques for SDS-Bel elect~rophoresis and immune blot detection of ~parated polypeptide chains have been described elsewhere (Rougon et al. 1982). The apparent raol. wt. of individual bands was estimated from the raigrati~m of dansylated marker proteins. Purification of NSP-5 polypeptide front adult morose ¢es,ebella Membranes were prepared from the cerebella of adult mice and extracted with 2% Triton or 2% sodi,Jm deoxychorate as described in IFig. I. The 100000 x g sups;rnatant was then pas~d through two columns operated in tandem, the firsl containing rat igG coupled to Sepharose and the sexond Ihe Scpbarose-bound mAb. To prepare the mAb column, anti-NSP-5 mAb was fir,it purified on prol(~in A-Sepharose and coupled to Sepharose CL-4B at 10 mg/ml of beads using the procedure described by Campbell ct al. (1981). After passing the detergent extracts through the columns containing 3 ml ,)f Se.pharose at 15 rai/h, the columns were disconnected and washed until the A2~0 was reduced ;o that of the buffer. The columns were eluted separately with 50 mM diethylaraine IpH 11.5)/0..5% Triton-Xl00 or :~liura deoxyeholate, The elated fractions containing the protein peak were l~Ooled and immediately neutralized with solid glycine. The ¢iuted material, which came off in approximately 6 ml, was concentrated, dialyzed against 10 mM Tris pH 8 containing 0.5% detergent and subjccte~t to SI}S-polyaerylaraide gel analysis. The elaates from the rat lgG column were used as controls. Quantitative measurement of antigen in subceilular fra~'tions Initial titrations of purified NSP-5 antibody were performed to dett:rmine the

41.5 dilution of antibody which fell on the steepest part of the binding curve in :llc~ dot assay. Then, aliquots of equal volume of antibody at the chosen concentration (1.2 pmole/ml) were incubated for 4 h at 4 ° C with serial dilutions of the fractkms to b¢ tested. The mixture was then centrifuged at 10000 X g for 10 min, and the supcrnatants tested for residual antibody activity in the dot ass;ly. The amount of protein needed to reduce antibody activity by about 50% was recorded i s one inhibitory unit. A flow diagram of the fractionation is preseated in Fig. 1.

Results

Expression of NSP-5 antige,~ in cultures from early posmatai mouse cer{'bella mul DRG Monolayer cultures from l-7-day old C57BL/6 mouse cerebella were examined by indirect immunofluorescence for the expression of the I~'SP-5 antigen ift~:r 1 - 6 ttomogenotes inlO mM TrLs-HCI 0.32 I~ Sucrose pH:Z5 3ram MgCIa Proteose ~hibitors

Spin 15min,15~O xg

~Spin 1 I%1~3(::)000x g

O~motlc sr~¢k

5raM Tns- HCI 0 5 r~M EDTA

[

pile5

[ Spin 1 n, 'K:XXX:~Ox 9

P3 /

~

~

SN 3

:)o~ul:.lization 20 mM Tris-015 M NoEl 2% TrLon or SDC DH "/5 [ ~.~

Spin 1 I1. IO0::XX) x g

P4 4r ~"~'~ SN4 Trlt on- ~l~oluL~e IVlemt~-1~nes sokd~hzed ~or imr~unopuriftcotion

Fig. !. Row diagram of the frac~ionation of tissue homogenates for the preparation of on.de membranes.

416 d;~',s in culture. %¢hcrc:~, ctaJtur~s ~d" pcrrnc;~hilizCd ~ , . ! n~,~-pcrm¢~bllizcd c¢ll~, fr~u~ culture, cultures nrcpart'd t'r~m .'-d:~',-~ld mice e x h i b i t e d some r~:;~.'ti,,il~,

.'~ttcr

2.

,.~mc ~,I thdm w~.rc seen t¢, be l:~F,¢lcd (l-i~3. 2(' and I)). "I'h~ I;~b¢lcd ~.'¢lls~'¢r~id~.'ntil'icd m, iletlr,)Ylds on the basis o f ~.Jl¢ir ~ ] ) o r p h o h ~ x arid their rc~.'li~il~, ,,,,ith l I X m d o u b l e - l a b e l i n g exp~-ri::t~tznt~+ ( n ~ I ~.11,)~,1+1). Orll~' a nm-'~jl subpt4+ul;~lit,n t~f

rq,, I ,v,.IFql.'d tt t,,~lVII %,ull.lr-k~ ,,l~ll,..

417

l l g 3. Indirect mlmunofluore,.cence of 6-da~-old I)R(; cultures fr(,m 2-day-oh.l mice. ×: l(~ll3..-I B: I a b d i n g '~.~tll anti-NSl'-5 anlib,~dy, tt: I:luoresc~nce and B: NomarsLi dlfferemhd ,merferencc ~.'on|;'a~[ ~|lowing d ~l~up o f ¢¢1~ with neuronal nlorpllOl.~g~, only one being NSP-5' . l he focus is on |he ,:¢11 b~dJ¢~. (" ~f: I)oubl¢ immumdaheimg for NSP-:~ a[ ligen ( I ) ) and T T X (t::). (" Nonlar~ki differential I+|lcrl'¢rUnc¢ .~)lllfilsl. All the N.~."+-5 ' celln or n:urltcs art.' +I+Ix ' . F.,otc tl3e unl;lbeled flat cell In the b.lc k g r o u nd ( a rr~,v. )

41~;

ceils with neuronal m o r p h o l o g y could be stained with the a n t i - N S P - 5 a n t i b o d y : usually the staining was rathe" weak o n individual cell bodies a n d s t r o n g e r o n t~heqr neurites (Fig. 2C). N o difference was seen w h e t h e r the cultures were permeabiliz~,d or not. In D R G cultures from 1-2-day-old mice maintained in vitro for 6 da;y.~. NS]-'..5 was detected on the surface c.f around 5 0 % of the celh; (Figs, 3 A and B) which cot~Id bc identified as neurones by m o r p h o l o g y and by double labeling, with T T X (Fi~gs. 3 C E). U n d e r our culture conditions not all cells of ~'~euronal m o r p h o l o g y could be

stained with T T X but all N S P - 5 * cells could. E x t e n d i n g the culture time to 10 da:/s did not affect the p e r c e n t a g e o f NSP-5 ~ cells. H o w e v e r . w h e n the cu]tures were p r e p a r e d from older animal,. (5--6 d a y s old) a r o u n d 7 0 ~ o f neL~rone-like cells could be stained with a n t i - N S P - 5 m A b . T h r o u g h N o m a r s k i o p t i c s 2 types o f n e u r o n a l cells were distinguished namely, large, light n e u r o n e s a n d small dark ones. Only the large n e u r o n e s were labeled by a n t i - N S P - 5 (see Figs. 3A a n d B). In both cerebellar or D R G cultures n o n - n e u r o n a l cells were always negative. In fact. in d o u b l e - l a b e l i n g e x p e r i m e n t s . NSP-5 staining never o v e r l a p p e d with antiG F A P , a n t i - f i b r o n c c t i n o r a n t i - g a l a c t o c e r e b r o s i d e reactivity ( n o t shown).

Localization of NSP-5 ant,:gen in sections from cerehellur,~ and dor.~al root gangha T h e expression of NSP-5 antigen in vivo was studied by indirect immunofluorc!,cence on sections from c e r e b e l l u m and D R G . Since forn~aldehyde fixation d e s t r o y e d the antigen reactivity, all studies were d o n e o n fresh frozen sections slightly fixed in a c e t o n e or noL A c e t o n e ~.reatment did not affect the staining p a t t e r n . In th,: cerebellum the antigen was first d e t e c t e d o n po:,tnatal d a y 7. This was in g..x)d a g r e e m e n t with the in vitro observations. At this age. a general weak staining o f the Purkinje cell layer and the thin molecular' layer was o b s e r v e d (Fig. 4A). In . ~ lions from the adult cerebellum (Fig. 4B), small f l u o r c ~ e n t p~tches were visible in the molecular lay,:r. Small, synapse-size d o t s were also obse,.-ved in the Pnrkin.ie cell layer. Labeling was particularly d e n s e a r o u n d Purkh,je cell bodies, this is reminiscent of basket cell axons. F l u o r e ~ e n t dots o f grcat~.'r d i a m e t e r t h a n t h o s e seen ~p. the molecular layer occurred t h r o u g h o u t the granular layer. l ' h e fluorescence int¢~sity seemed to follow a gradient, the most intense labeling o~:curring in tbe o u t e r m o s t region b e n e a t h the Purkinje cell layer. T h e f l u o r e ~ e n t d*)ts ;~ppeared to ctmlesce in s(,nae places to form p a t c h e s of the size of synaptic ~,lomeruli. T h e g r a n u l e .t:ell!i stood out as negatively stained structures b,;tw,:en them. N o clear o u t l i n i n g o f their" cell

|:1~. 4. Indirect trnmun()fluorc~c'¢n,:¢ ]abelJn~ I'*~1 ~';SP-5 allall~¢n irl tn.~h fr~/,:¢~ ~ . J l l J

~.'~.-IIor~ o~ ~ u ~ "

cerebella and I)R(L > 44H). I'(JL - external ~r.lnular lls~.¢r;M L = nh,~:¢uhr J~,ef: PL ~ Porkinj:: eel! la'.cr;(it. : granular la!,cr:WI, I,= v,huc mailer. A:('crcbc||ur;1of a %d~.-~)J,Jn~u~t~, B D ('¢rcb¢llun, of aduh ~ild-t~pc mice ( B. D ) and I~.d mutant mi¢¢ ( ( ) . N,~tt' the punc~a t¢ s~a~nJn~ in t ~ ' ~.tL ;~ir,:d Pl.

and the fluorescent dots with the ~,J~eof synaplic glomcruli in Ih¢ GL t..~rro~,~) t B ). ]'he W~ ~.htl~~, patchy fluorescence wh~cl~appears to f¢,llo~ ~,ometime~~be t nurse (~f indig~ded~hb¢~ ~afn¢~ ) I D ). I r~pod cerebeilum ((') the slimmingi', confined to the Crl.. Nolo Ih~ ah~'ncc o~" I~kd~n~ o~{ Ih¢ M;,. and g'M F F: In adult I)RG ~cclions NSP -~, antigen can be d¢l~:ced ,}n the ,,ur!,e~:~,-~3[ ,a~m¢ ~n~a n, ne~r~}r,~ (arrov.~) ( I" ) and on fibers m scct~tv,so ~ 1he nx~ts (/" ).

420 bodies was seen. Patchy fluorescence was also observed in the white matter which appeared to follow the course of individual fibers (Fig. 4D). W h e n sections ~'r~m adult pcd cerebella were examined, a strikingly different s~aining p~ttern was observed. T h e molecular layer a n d the white m a t t e r were completely negative (Fig. 4C) a n d there was ~iffuse label throughout the granular layer. In D R G sections from adult wild-type mice, cell bodies of sensory nenronc,s were clearly stained at their surface {Fig. 4E). Some cytcglasmie staining was observed which was appareatly non-specific since it cxvdld also b e observed in controls {not shown). O n sagitt21 sections of the roots, the antigen was detected o n neurona,~ fibers (Fig. 4F).

Biochemical identification of the NSP-5 antigen T h e i m m u n e b~ot technique was used to identify the NgP-5 antigen in $ D ~ e x tracts from various regions of the adult a n d neonatal mouse central nervous system. In extracts of adult tissues analyzed u n d ~ reducing conditions the anti-lqSP-5 antibodies reacted with a 180K polypcptid¢ (Fig. 5). i n some experiments a fainter b a n d at 165K aL,~o was revealed. The 165K b a n d was not always present and might have been created by degradation d~ring preparation.

1 2 3

4

5

6

7

8

top 220-

1 2094-

68-

Fig. 5. Immune ~lot analysis of N~r~-5 an~i~'n in SDS-cxIr~cIsfrom m c ~ C~S. Lanes I. 2. 3: spinal cord fro~r, adul'~, l-day-okl and 7~lay-old ~ r~'liv~. ~ ,4. 5. 6: cerebella from wild type mice. w, avcr and lind mutantr~ respectively. Lan~ 7. g: ~rahb ,~¢~bral hemispheres fro., 7,.day.old ar,d a(h. ;t mice. IFAI,Jalamolmts of ~olein 12~ ~ were Ioade¢~on ¢a~.'hrdrip. Mol. wt. marker~ used were ~al,~)khlmide-labded ?20K fratgmml of I h ~ j~alacto~i&~¢ (120K). phosphorvlase b (94K) an~! mvam alb~umin~68K).

421 In good agreement with our immunohistological observations, NSP-5 antigen was more abundant in adult tissues. A typical result is shown in Fig. 5 (lanes 1 . 2 . 3) where NSP-5 could not be detected in 1-day-old mouse spinal ~,~d whereas 7 days after birth it was well represented, although the band was stil~ weaker lhan iin the adult. Reactivity was strongest in the adult spinal cord (lane 1). The adult cerebellum (lane 4) seemed to contain relatively less antigen and the cerebral henuspheres were always negative (lanes 7 and 8) at ieast under our experirae:ntal conditions. The cerebellum of weaver mutants, ~hich supplied our intmunogen, contained NSP-5-reactive bands of similar mol. wt. as wild-type cerebellum (lane 5). To our surprise, an immunoreactive band was also revealed in the cerebellum of pod mice (lane 6).

Subcellular distribution of the NSP-5 antigen A semi-quantitative estimation of the distribution of the NSP-5 antigen in different fractions from adult cerebellum is shown in Table 1. The antigenic activity of the fractions was measured by the inhibition assay described in Methods. For the sake of comparison the data was expressed in inhibitory units, one unit being defined as the amount of protein required to give around half-maximum inhibition of antibody binding in our assay. Nea~rly no antigenic activity was detected in hypotonic extracts of crude membraries indicating that the antigen is entirely membrane-bound. Less than 15% of the antigen activity was solubilized by TritonXI00. Consequently, the detergent-insoluble material was highly enriched in antigenic activity.

Purification of the NSP-5 antigen As stated in the preceding paragraph, only 10-15% of the initial antigen activity was recovered in Triton-X109 or sodium deoxycholate extracts. Although this yield was very low, these detergents have the advantage of allowing the use of aff'mity chromatography to purify the', antigen. When NSP-5 was purified from such extracts of adult cerebellum by immunoaffinity chromatography, approximately 90% of the TABLE 1 SOLUBILITY CHARACTERISTICSOF NSP-5. Adult mouse cerebella were homogenizedand crude membranes prepared as described in Fig,. l, which were then shocked osmoticallyor solubilizedin Triton X100 Fractions

1O- 4 × activity (units)

Total homogenate soluble Pellet 1 hypotoni¢ soluble Pellet 2 Triton soluble Pellet 3

13 0.6 12.8 0.13 11.8 1.8 8.0

Specificactivity (10"~units/g protein) 3.5 0.22 7.7 0.5 15.25 3 33.0

Relative specificactivity 1.0 0.06 2.2 0.14 4.3 0.85 9.5

422 dete.ctabie antigenic activity ~ a s r e m o v e d f r o m the e x t r a c t after r e p e a t e d p a s s a g e o v e r the c o l u m n a n d a r o u n d 59% o f this activity was r e c o v e r e d in the eluate with a 150-fold e n r i c h m e n t in a n t i g e n activity. In a d d i t i o n to a 180K b a n d sc~.n u n d e r reducing o r n o n - r e d a c i n $ c o n d i t i o n s a n d w h i c h m o s t likely c o r r e s p o n d e d to the NSPob-reaetive b a n d d e t e c t e d in i m m u n o b l o t s o f c r u d e m e m b r a n e s , a n u m b e r o f p o l y p e p t i d e s with the following mol. w t ' s were revealed: 115 a n d 110K, 68K, a triplet b e t w e e n 55 a n d 50K a n d a s t r o n g b a n d m 4 4 K . M o s t o f these b a n d s were no~ p r e s e n t in control p r e p a r a t i o n s o b t a i n e d b y p a s s i n g the extract over a o01u-nn

1

2

,

3

4

5

.

Fig, 6. SDS-polyacryla~fide gels ~ho,-~[n$prcparalion~ of imm~.~nopurifw,~NSP-.5 anti$~cn a~d ~f~trol ~tained with Coomassie blue, and • c, rrespondin$ immune bl~t. Lane I = moLwt, markers which ate RHA polymera~¢ O.55 ara:l 165K). ~ - ~ ~la~to~dlas¢ (12~1K). phosphorylas¢ b (94K). bovine serun: albumin [6gK) and m alb,amia (43KJ Lane 2 - ma~ctial e[uted from the rat IgG control column -"n~iyzedafter rc4Juction. L,me 3 ~ NSP-$ purified Ifom adult ~m~u~,~ ¢ereh¢l|a and anaJy~,gd after reduct~,m. Larm 4 - t"le ~am¢ prcpaf,~l~an ~,ithou~ rcdugliot. Lane corresponds In an immune blot of the same prcparat;on ~l~wifl$ only one NSP.5.react~ve band at ISOK.

,~23 containing normal rat lgG (Fig. 6). lmmunoblot experiments were done on the purified material to determine whether the additional bands bore the antigen/c determinant. As shown in Fig. 6, a single polypeptide chain with an apparent tool. wt. of 180K bound the mAb. However, when the purified material was ic~linaw.d and i.,~lmunoprecipitated with anti-NSP-5, most of the additional polypeptides were co-precipitated indicating that the antigen exists in close association wi~h other polypeptides that do not bear the antigenic determinant (no: shown).

Discussion

The anti-NSP-5 mAb described in this report reacts wilh a single 180K polypeptide chain in SDS extracts from cerebellum and spinal cord. The NSP-5-reaefi,,,e band migrates with an apparent mol. wt. of 180K with or without r~luctk$ indicating that it does not form disulfide-linked complexes. !n tile dot assay, the mAb binds to a concanavalin A-binding glycoprotein fraction extracted from whole mouse brain (Hirn et al. 1981) (not shown). However, more work will be needed ~:o ascertain that the protein recognized is glycosylated and not bound to the concanavalin A column via another glycoprotein. There are two arguments for classifying NSP-5 as a plasma membrane protein associated with cytoskeletal structures. First, the antigen, which clearly has a significant surface localization as shown by immunofluorescence on live c~lls, remains to a large extent associated with a detergent-soluble pellet after extraction with Triton-X100 or s ~ i u m dcoxycholate, which tend to preserve protein-protein interactions (I-lelenius et al. 1979). It is. however solubilized by SDS and thus co-distributes with the cytoskeleton which ~ n be defined operationally as the cellular constituents that are insoluble in non-d~,naturating detergent (Starger et al. 1978). Secondly, during purification on a mAb column, NSP-5 co-elutes with. a number of other polypeptides whose tool wt's. might correspond to those of some intermediate filament subunits (Lazarides 1980) and actin. It is now well established that cytoskeletal structures can interact with each other and also with the plasma membrane (Pober et al. 1981: Sheterhine and Hopkins 1981), and high mol. wt. glycoproteins, possibly expressed at the cell surface, have been found associated with cytoskeleton proteins in cultures of embryonic chick brain and sympathetic ganglia (Moss 1983). However, these and our own results do not exclude the possibility that such associations occur only after cell lysis. A considerable amount of data can be found in the literature ~3ncerning cytoske~eton-associated membrane proteins. For instance, a spectrin-like protein family including brain fodrin (Levine and Willard 1981), intestinal brush border proteins (Glenney et ai. 1982), brain spectrin (Bennet et ai. 1982) and a novel 230K polypeptide widely shared by different types of cells has been described (Lehto and Virtanen 1983). The tool. wt. of NSP-5 as well as its biochemical properties do indeed suggest a relationship with these molecules. However, the spectfin-like mole~'ules have a peripheral but subplasmalemmal localization (Lehto and Virtanen 1983) and are only detected by immunofluores,.:ence of cultured cells after penmen-

42d bilization. None of the glycopolypeptides described by Moss ,1"198:)) migrates at or near 180K, although the presence of not further specified high rnoL wt, bar~d~ has been indicated by tl~is author. Expression of lqSP-5 follows a ~:evelopmcntal time course, in ,.~tions ;rom cerebellum, NSP-5 can first be detected on postnatal day 7. But the fluorescence intensity is still low at this age and increases considerably in the adult cerebellum. In good agreement with these c~ata, the antigen cmmot be detected in cerebellar caltares prepared from animals younger than 7 days. In ¢:ultures from older mice, some neuronal cell bodie~ and fiber bundles are labeled by anti-NSP-5 mAb. Similarly. the ~umber of NSP-5 + cells in DRG cultures it.creases with the age of the animals from which they are prepared. However, the developmentally regulated appearance of NSP-5 cannot be reproduced in vitro, since, cecebellar cells from younger animals do not acquire the antigen when maintained in culture for longer periods of time. and the number of NSP-5 ~ neurones in D R ( ; cultures remains constant rcgardles,~ of culture time. It is not possible at the optic level to identify the structures hearing the antigen in cerebellar sections, ~lthough the punctate pattern of staining in ~.e molecular layer is reminiscent of synaptic sites. Granule cell bodies and their axons, t ~ parallel fibers of the molecular layer, as well as Purkin~e cell bodies are not stained by the mAb in~icating iis specificity for a sub-class of neurons. The presence of NSP-5 in pcd cerebella was unexpected sine¢ we used tl~es¢ mutam cerebella as a n~ative target fi,r monoclonal antibody selection. H¢,wever, the antigen is less u idely distributed in the cerebellar cortex of ped muta~tts than in wild-type mice and may thus give a negative signal ia the dot assay. The lack of staining of the molecular layer in pcd mutant cerebella is intriguing. One po~sibility is that the put, crate staining seen in the molecular layer o f wild-type n~ice c~mresponds to synaptic structures on Purkinje c¢11 dendrites, which are absent in the Purkinje cell-d¢fick:nt mutants. Electron.microsc~¢ic data will be ne.~ .~1 to identify the structures bearing the antigen in normal and mutant cerebella. Because of its cell surface reactivity, and its specificity for a sub-population of neurones in DRG cultures our raAb may be a valuable tool not only for identilying classes of sensory neurone:~ buz also for separating them by Ix~itive or negative selection. Moreover. NSP-5 s~¢ms to underg~ multiple associations with other proteins ~vhJch correspond p e r h p s to cytoskele~al com~nents. Work is ,mderway to test the specificily of these ihtcractions and to identify the prote/ns involved by specific antibodies.

References

Bennct, HV., J. Davis and W.E. ~"o)Jcr, Br~in H)ectnn ~ ~

r

~

~m~n ~ ¢ ~

.~

structure and function to ¢tylhroc.yl¢ sl:~cctrin, Nature (L~,I1KI.). ~ (IgS,~.) 126-131. Bignami, A., L.F. En K, D. Dahl and C,T. Uycda, L,ocal~liOn 04" I M SJ~ f d ~ l l a r y acidic i~rot¢~ ~ in astrocytes by immunofluoresccnce. B-ain R¢~.. 43 (19"/2) 429-435. Brockcs, J.P,. K.L. Fields and M.C. Raft. Studies cn cuhur~l f i l .¢,chwan+l cells - - F+.~lJl~,d~llls.'lqI Of purified populations from culluf¢~, o[ pcriphclrai r~crvc. Brain R¢~.. |65 II~r'/g) I O ~ - I l l ,

425 Campbell. D.. J. Gagnon, K. Reid and A. Williams. Rat brain Thy-I glycoprotein, Bloc.bern. Jr 195 (1981) 15-30. Fields. K.L., J.P. Brockes, R. Mirsky and L.M. Wendon, Cell surface markers for distingnif,hhaBdifferem types of dorsal root ganglion cells in culture. Cell, 14 (1978) 43-51. Glenney, Jr.. J.R., P. Glenney and K. Weber, Ezythroid spectrin, brain fodrin and intestinal bfuth border proteins (TW 250/240) are related molecules containing a common ealmodulirJ-binding sub*rail bound to a variant cell-type specific subunit, Proc. Nat. Acad. Sci. (USA), 79 (1982~) 4002--4005. Goridis. C., M.A. Jober, M. Hirsch and M. Schachner, Cell surface proteins of culutred brain ct41s and their recognition by anti*cerebellum (NS-4) antiserum, J. Neorocbem., 31 (1978], 531-539, Hawkes. R., E. Niday and J. Gordon. A dot-immunobinding assay for monoclonai and other antibodies. Analyt. Biochem., 119 (1982) 142-147. Helenius. A., D.R. Milaskin, E. Fries and C. Tanford, Properties of detergent. Methods in Enzyp~oio~-. 56 (1979) 734-749. tlenk~,~Fahle, S. and F. Bonhoeffer. Inhibition of axonal growth by a monoclo:~al antibody. Nature (Lond.). 305 (1983) 65-67. Hirn. M.. M. Pierres, H. Deagostini-Bazin. M. Hirsch and C. Goridis, Monoclonal antibody al~urtf~t surface $1ycoprotein of neurons, Brain Res., 214 (1981) 433-439. Hirn, M.. M. Pierres. H. rJeagostini-Bazin. M.R. Hirsch, C. Goridis. M.S. Ghandour. O.K. Lanlk.~j and G. Gombos, A new brain cell surface glycoprotein identified by monoclonal antibody, N m 7 (1982) 239-250. Lazarides. E., Intermediate filaments as mechanical integrators of cellular space, Nature (Lond.k 2~3 (1980) 249-256. Lehto. U.P. and I. Virtanen. lmmunolocalization of a novel cytoskeleton associated polypeptide of Mr 230000 daltons (p. 230), J. Cell Biol.. 96 (1983) 703-716. Levine, J. and M. Willard, Fodrin - - Axonally transported polypeptide~ associated with the internal periphevj of many cells, J. Cell Biol., 90 (1981) 631-643. McKay, R., M.C. Raff and L.F. Reichardt, Monoclonal antibodies to neural an:igens (Ct-4d Spring Harbor Reports in the Nenrosciences, No. 2), Cold Spring Harbor Laboratory, New York, 1981. Molts, DJ.. Cytoskeieton-associated glycoproteins from chicken sympathetic neurons and chickcm embryo brain, Europ, J. Biocbem.. 133 (1983) 291-297. Mullen, R.J.E.M. Eicber and R.L Sidman, Two new types of retinal degeneration in cerebellar m m a a t mice, N~Lture (Lond.). 258 (197.~) 528-530. Mulkm, R.J., E.M. Eicber and R.L. Sidman, Purkinje cell degeneration, a new neurological mutatio~ in the moose, Proc. Nit. Acad. Sci. (USA), 73 (1976) 208-212. Pierre& M., C. Goridis and P. Golstein, Inhibition of routine T cell-mediated cytolysis and T cell pro|iferation by a rat monoclonal antibody immunoprecipitating two lymphoid ceil surface polypeptides of 94000 and 180000 molecular weight, Europ. J. lmnmnol., 12 (1982) (~O-69. Pobeto J.S., B.C. Guild, J. Strominger and W.R. Veatch, Purification of HLA-2 antigen - - Fluor~ceaz labelina of its inlra-cenular region and demonstration of an interaction between flumcscently labeled HI,,A-2 and cytoskeleton proteins in vitro, Biocbemi,~;try, 20 (1981) 5625-5633. Raff, M.C., K.L. Fields, S.I. Hakomori, R. Mirsky, R.M. Pruss and J. Winter, Cell-type specific markets for dit tinguishin$ and studying neurons and the major classes of glial cells in culture, Brmn ~ |74 (1979) 283-308. Raff, M.C.. E.R. Abney. J. Cohen, R. Lindsay and M. Noble, Two types of astrocytes in cultures developing rat white matter - - Differences in morphology, surface gangliosides and ~zmfth characteristics, J, Neuro~i., 3 (1983) 128,o-1300. Rakic, P. and R.L. Sidman. Sequence of developmental abnormalities leading to granule ceil*deficit m cerebetlar cortex of weaver mutant mice, J. Comp. Neurol., 152 (1973) 103-132. Rezai, 7. and C H. Yoon, Abnormal rate of granule cell migration in the cerebellum of weaver m u ~ t mice, Develop. Biol. 29 (1972) 17-25. Richman, D.P., C.M. Gome~ P.W. Berman, S A. Burres, F.W. Fitch and B. Arnasono Monockmal anti-a~tyk:holine receptor antibody can cause experimental myasthenia, Nature (Lond.), 286 (19go) 738-?39.

426 Rougon, G., H. Dea~fini-Bazil~, M. Him and C. Gm'idis, T~aw.- and developmental stage-sl~.cific forms of a neural cell surface antigen linked ¢o differences in glycosylation of a comm~n pcp~ide. EMBO J., ! (1982) 1239-124~. Rougon, G., M.R, Hirsch, M. Him, J.L. Guenet and C. Goridis. Mocc~lonad antibody to neural cell surface protein - - Identification ol a glycoprotein family of restricted cellular localization. '~euroscience, 10 f1983) 511-57J). Rutishauser, U. and G. Edclman, Effects of fascicolation on the outlg'owth of mmzi~s from spinal ganglia in culture, J. Cell Biol.. 87 (1980) 370-378. Sadoul, R., M. Him, H. Deagostini-Bazin. G. Rougon m ~ C. Gotidi$. Adult and embryonic moth,to netJral cell adhesion molecules have different binding pfopertk"s. Nature (l~nd.). 304 (1983) 347-349. Schlaepfer, W. and R.G. Lynch, l m m u n o f l u o f ~ ~tudies of neurofilaments in the rat and human peripheral and central nervon,~ system. J, Cell Biol.. 74 (1977) 241-247. Sheterhine, P. and C.R. Hopkins. Transmembrane linkage between surface glycoprt~teins and components of the cytoplasm in neutrophil leuc4~-ytes. J. Cell Biol.. 90 (1981) 747-754. Starger, J.M., W.E, Brown, A.E. GoMv.tan and R. Goldman, Biochemical and immunological analysis of rapidly purified 10 nm filamen from baby hamster kidney (BHK21) ~ l h . J. Cell It;ol 78 (1978) 93-7.09. Weiner, H.L. and S,L. Hauser. ,gntisenic specificity of the nervous system. Annal. NemcL. 12 (1982) 499- 509.