Generation of monoclonal antibodies specific for developmentally regulated antigens of the chicken retina

Developmental Brain Research, 59 (1991) 197-208 (~ 1991 Elsevier Science Publishers B.V. 0165-3806/91/$03.50 ADONIS 0165380691512505

197

BRESD 51250

Generation of monoclonal antibodies specific for developmentally regulated antigens of the chicken retina Burkhard Schlosshauer and Monika Wild Max-Planck-lnstitut ffir Entwicklungsbiologie, Tfibingen (E R. G~) (Accepted 8 January 1991) Key words: Chicken visual system; Explant culture; Immunization strategy; Monoclonal antibody; Neural development; Protein purification; Retinal cell type

In order to raise monoclonal antibodies specific for neural antigens that might be involved in regulatory events during histogenesis of the chick visual system, 5 different immunogen fractions were employed in conjunction with 4 immunization regimes. By screening 17,000 hybridoma supernatants from 58 fusion experiments, empirical guidelines were formulated for generation of antibodies with distinct specificities. A number of positive and negative results would not have been predicted from theoretical considerations and emphasize the relevance of empirical data, To substantiate our observations we present 10 monoclonal antibodies that bind to distinct cell types or layers of the retina. Single cell as well as explant cultures in conjunction with Western blot analysis were employed to determine subcellular localization and cell type specificity of selected antibodies. Six antibodies bind specifically to the inner limiting membrane or the optic fiber layer or the inner plexiform layer or the outer plexiform layer. Four recognize selectively ganglion cells or Miiller glial cells or presumably subpopulations of amacrine cells. These antibodies promise to be helpful tools to evaluate mechanisms of neural differentiation. INTRODUCTION Over the past decade various new techniques have been invented to elucidate the relation between the structure and function of distinct neural elements. Although these approaches have allowed thorough investigations of morphological details as well as transmitter distributions in CNS regions such as the neural retina 1°'31 the developmental mechanisms that give rise to such sophisticated receptor organs are still poorly understood. It is not known how postmitotic neurons identify their final destinations during the early period of cell translocation nor which molecular processes determine cell specificity, oriented neurite extension and recognition of synaptic partners. In neuroembryology only minute amounts of material are available impairing biochemical and functional analysis of distinct molecules. The invention of hybridoma technology, however, has provided the possibility to generate immunological reagents that allow the detection and characterization of distinct neural antigens, even if they represent only a very limited fraction of all components in a given tissue 42. Initial attempts to produce neurone specific monoclohal antibodies used unfractionated tissue homogenates as immunogens. Monoclonal antibodies were successfully

generated against the rat and chicken retina 2'26, distinct neurones of the leech or drosophila melagonaster nervous system 23'44. In our hands, unfractionated immunogen mixtures, that were used without any suppression technique, led to a poor immune response with the result that the vast majority of monocional antibodies were of no relevance. In contrast to these approaches, biochemical purifications of distinct immunogen fractions lead to higher efficiencies in hybridoma generation 16,32. In addition, different in vivo interventions were employed to direct the immune response in a more predictable fashion. Two methods of differential immunization (immune suppression and tolerance induction) have been designed to eliminate lymphocyte populations specific for irrelevant antigens. I m m u n e suppression is based on the use of the alkylating agent cyclophosphamide, which is toxic for stimulated lymphocytes 22m. Therefore sequential immunization with irrelevant immunogens and subsequent cyclophosphamide injection is supposed to eliminate the immune response against these components, even if they are present in later immunogen mixtures. After the initial phase of immune suppression, unfractionated mixtures which include relevant and irrelevant antigens can be used for further inoculation, leading theoretically to an immune response preferentially directed against relevant immunogens. One disadvantage

Correspondence: B. Schlosshauer, Naturwissenschaftliches und medizinisches Institut an der Universit~it Tfibingen, 7410 Reutlingen, F.R.G.

198 of this m e t h o d is due to the transient effect (approx. 4 w e e k s ) o f i m m u n e s u p p r e s s i o n , which h a m p e r s i n d u c t i o n of a s t r o n g s e c o n d a r y i m m u n e r e s p o n s e against r e l e v a n t i m m u n o g e n s 29. This p r o b l e m can be c i r c u m v e n t e d using the a p p r o a c h of t o l e r a n c e

induction.

This t e c h n i q u e

relies on the

e n d o g e n o u s e l i m i n a t i o n o f l y m p h o c y t e p o p u l a t i o n s specific for self a n t i g e n s d u r i n g a critical p o s t n a t a l p e r i o d of m o u s e d e v e l o p m e n t . I r r e l e v a n t i m m u n o g e n s are i n j e c t e d during

this

period

and

a

mixture

of

relevant

and

i r r e l e v a n t i m m m u n o g e n s a r e g i v e n at later stages. A g a i n , this p r o c e d u r e c o u l d result in a r e s p o n s e specific o n l y for r e l e v a n t i m m u n o g e n s 19'39. In an e f f o r t to g e n e r a t e m o n o c l o n a l a n t i b o d i e s ( M A b s ) against d e v e l o p m e n t a l l y r e g u l a t e d a n t i g e n s of the chick visual system, we h a v e e m p l o y e d 4 d i f f e r e n t i m m u n i z a tion r e g i m e s in c o n j u n c t i o n with purification of i m m u n o g e n s by 5 d i f f e r e n t b i o c h e m i c a l m e t h o d s .

Although

g e n e r a t i o n o f M A b s with d e f i n e d specificities is not yet c o m p l e t e l y p r e d i c t a b l e , a n u m b e r of basic rules for the m o r e efficient g e n e r a t i o n o f M A b s c o u l d be d e t e r m i n e d . Ten M A b s are p r e s e n t e d that r e c o g n i z e distinct tissue layers and cell types o f the d e v e l o p i n g chick retina. MATERIALS AND METHODS

Materials Embryonic retinae and optic tecta were dissected from white leghorn chickens. Triton X-114 (Sigma) was preeondensed as described 5. Enzyme inhibitors from Sigma (2,3-dehydro-2-deoxyN-acetylneuraminic acid, aprotinin, leupeptin, pepstatin and phenylmethylsulfonyl fluoride) were added to all homogenization and solubilization buffers as recently described 33. Concanavalin ASepharose, lentil iectin (lens culinaris agglutinin)-Sepharose supplied from Pharmacia; peanut agglutinin and wheat germ agglutinin coupled to agarose (Sigma). Affi-Gel 10 was purchased from BioRad Laboratories. Hapten saceharides (N-acetyl-D-glucosamine, 1-O-methyl-D-glucopyranoside and D-galactose) were bought from Sigma. Rhodamine- and fluorescein-conjugated F(ab')2 fragment goat anti-mouse IgG (heavy and light chain) as well as conjugated isotype specific anti-IgM and anti-IgG and anti-rabbit (Jackson Immunoresearch Laboratories) were used for 1/2 h at a dilution of 1:100 at 22 °C. HRP-conjugated F(ab')2 fragment goat anti-mouse IgG and M (Jackson Immunoresearch Laboratories) were applied at 4 °C at a dilution of 1:2000 for 5 h for Western blot analysis and 1 h for ELISA. Antiserum specific for glial fibriUary acidic protein (GFAP) (Dakopatts) application: 1:500, 2 h, 22 °C.

Methods lmmunogen purification. Cytoplasmic membranes were purified from neuronal tissue of embryonic chicken (embryonic day 10 in most cases (El0)) by sucrose gradient centrifugation based on the method described in refs. 20 and 43. Solubilization of membranes with Triton X-114 was performed as described 33. After precipitation of insoluble material by ultracentrifugation at 4 °C, the clear supernatant was applied individually or in sequence to MAb D3and MAb Dl-columns against different isoforms of N-CAM 33 and thereafter to lectin affinity columns (ConA-sepharose, lentil lectinsepharose, peanut agglutinin-agarose, wheat germ agglutinin-agarose). The corresponding ligands were eluted at pH 11.5 and by hapten saccharides, respectively as outlined recently 33. Proteins

which did not bind to any of the above affiniy matrices, were subjected to temperature-induced detergent phase separation in two consecutive centrifugation steps as describedS'~:~. Hydrophilic proteins which were not pelleted with the detergent at 30 °C were finally precipitated by the addition of acetone (20 times the detergent buffer volume at -20 °C for 16 h). Proteins were sedimented by centrifugation (25,000 g, 30 min, 4 °C) and after the removal of the supernatant, the precipitate was vacuum-dried and thereafter solubilized by boiling in SDS sample buffer 24. Preparative gel electrophoresis of these samples, to which prestained marker proteins had been added on 5 mm thick gels, was performed according to Laemmli 24. After electrophoresis, unfixed 10 cm long gels were cut into horizontal strips (5 mm), and the corresponding protein fractions were electroeluted in a biotrap chamber BT 1000 (Schleicher and Schfill) according to the manufacturer's instructions. Separation of detergent insoluble cytoskeletal components from enriched cytoplasmic membranes of embryonic retinae was performed according to TM. Protein quantification was carried out according to Lowry et al. zS. Analytical polyacrylamide gel electrophoresis 24, silver staining of SDS gels ~ and Western blot analysis z8'4° were employed for qualitative characterization of protein fractions. Antibody generation. Standard immunization was performed over a period of approx. 3 months using 50-500 ~g protein per injection in conjunction with adjuvant, as detailed in Schlosshauer 33. Immunogen activation was based on the method of St/ihli et al. 38. Typically 150/xg protein (tissue homogenate or purified glycoconjugates) was coupled to 750 /xg carrier protein (hemoglobin or bovine serum albumin or hemocyanine; all from Sigma) with the aid of 0.08% glutaraldehyde in 500/tl phosphate-buffered saline for 5 h at 22 °C. Still reactive glutaraldehyde was blocked by addition of 15 /xl 2 M Tris-HCl, pH 8.8 for 1 h, 22 °C. Each carrier protein was used only for a single injection per mouse to avoid an elevated secondary immune response against the carrier. For immune suppression, cyclophosphamide was used as described eg. One hundred to 500/~g protein of fraction A (retinal homogenate or denervated tectal tissue homogenate or purified cell membranes from eyeless chick embryos) without adjuvant was injected followed by daily intraperitoneal injections of 2 mg cyclophosphamide (Sigma) in 200 ~1 PBS for 3 days. Three weeks after distinct suppression cycles (see below) fraction B (an immunogen fraction that had not been used for suppression: tectum or retina, respectively) was applied several times as adjuvant emulsion with trehalose dimycolate and monophosphoryl lipid A (Ribi) in a 1 : 1 ratio. In a number of cases intercalating suppression cycles were started 2-4 weeks after regular immunization injections. Sequences of inoculation with fraction A plus suppression (A-S) and fraction B (B) were as follows: (1) A-S, A-S, A-S, A-S, B, B, B. (2) A-S, A-S, A-S, A-S, B, B. (3) A-S, A-S, A-S, B, A-S, B, B, B. (4) A-S, A-S, A-S, B, A-S, B, B. (5) A-S, A-S, B, A-S, B, A-S, B, B. (6) A-S, B, A-S, B, B, B. (7) B, B, B, A-S, A-S, A-S, B, B. Tolerance induction ~3 was initiated by two intraperitoneal injections of paraformaldehyde-fixed (PFA) denervated tectum El0 (approx. 500 #g protein in 50/~1 PBS) into postnatal mice at days 1 and 2. Six weeks later, for two months, mice were inoculated several times with paraformaldehyde-fixed retinae E8 (approx. 500 /xg protein per injection) using Freund's adjuvant as for standard immunizations 33. Cell fusions were performed as described H using spleen cells and NS-1 myeloma cells in a ratio of 8:1. Fused cells were plated in most cases in microtiter plates containing macrophages from rodent intraperitoneum. In order to determine the influence of Ewing sarcoma growth factor (Costar, 1:40 dilution) and Insulin (Serva; Zn-complex, A- and B-chain 0.5 mU/ml) on the fusion efficiency and cell survival, initial fusion- or subclone-cultures were split into two groups and tested separately. The effects of cell density were tested in the range of 0.5-5 × 105 cells/100/xl per well. Neuronal cultures. Retinal single cell cultures from chick embryos on laminin coated (20/~g/ml for 16 h, 4 °C) 8 well slides (Flow) were performed as described earlier 33'36. For explant cultures, E6 chick

199 retinae were flat mounted on nitrocellulose filters, cut into 0.3 mm strips and placed onto laminin coated glass cover slips17. After a 1-2 day incubation period, cultures were fixed and immunolabeled. The culture medium was based on F12 medium (Gibco) and contained serum as well as methylcellulose. Details are given in Walter et al. 43. ELISA. Neuronal tissue from embryonic chicken was homogenized in homogenization buffer (see above) and the suspension (protein concentration 10/~g/ml; 100 ~1 per well) added for 1 h to microtiter plates without further fractionation. Using a standard procedure2~ primary and secondary antibodies were applied for 1 h each. Mouse sera including heparin were employed in a two-fold serial dilution of 1:200 to 1:200,000. lmmunofluorescence assays. Cell surface antigens were identified in vitro by addition of hybridoma supernatants to viable retinal or tectal cells. After washing and fixation fluorescently labeled secondary antibodies were added34. For immunodouble labeling all antibodies were employed in sequence rather than in parallel, which was found to cause some artefacts. For colabeling with phalloidin to visualize actin filaments, Fluorescein conjugated phalloidin was added to specimens after application of the primary and secondary antibodies, postfixation (4% paraformaldehyde, 30 min, 22 °C) and permeabilization (1% Triton X-100, 1/2 h, 22 °C) and subsequent washing. For whole mount labeling, E7 retinae or tecta were fiat-mounted on adhesive nitrocellulose filters37, fixed for 16 h, 4 °C in 4% PFA washed and incubated with primary and secondary antibodies without detergent. For screening of fusion cultures special compound sections were created7, to be able to visualize antigen expression simultaneously in various tissues from 3 different developmental stages using only 40/A hybridoma supernatant. For this purpose, PFA-fixed half chick retinae, including pigment epithelium, sclera, surrounding connective tissue and muscles of embryonic day 6 and posthatch stages were arranged around an El0 eye still attached to a head half. After embedding in tissue-tek (Miles) and rapid freezing in liquid nitrogen, sectioning was performed in such a way that each 10/~m section included besides all 3 retinae also El0 head structures, such as the tectum opticum. A grease lining (PAP PEN; SCI) between 3 adjacent sections mounted on gelatine chromalaun-coated glass slides allowed application of 3 different supernatants on a single object slide. Consequently antibody screening was accelerated considerably.

RESULTS General i m m u n e responses The m a j o r goal of our investigations was to generate immunological p r o b e s that would facilitate the identification of d e v e l o p m e n t a l l y regulated antigens. In o r d e r to prevent losses of potentially i m p o r t a n t antigens, we inoculated mice with unfractionated tissue. Because the i m m u n e responses were unsatisfactory in these cases, we started to enrich distinct subcellular fractions using various different biochemical p r o c e d u r e s (Fig. 1). In the first step, cell m e m b r a n e s were enriched by sucrose density gradient centrifugation, which eliminates most of cytoplasmic and nuclear components. Solubilization of m e m b r a n e constituents with detergents allowed subsequent affinity c h r o m a t o g r a p h y using different lectin and a n t i b o d y columns. In some experiments we e m p l o y e d a sequence of 7 columns. In those cases, where TX-114 was used, d e t e r g e n t phase separation at a physiological t e m p e r a t u r e range could be p e r f o r m e d after affinity

c h r o m a t o g r a p h y to enrich integral m e m b r a n e proteins. During this p r o c e d u r e a h o m o g e n e o u s p r o t e i n solution is f o r m e d at 0 °C whereas above 20 °C a phase separation is induced. A b o v e 20 °C TX-114 forms the lower phase and preferentially incorporates integral m e m b r a n e proteins 5. We found that also some cytoskeleton c o m p o n e n t s were included in this fraction. M o r e than 90% of the initial cell m e m b r a n e proteins r e m a i n e d unfractionated after these procedures. These were subjected to p r e p a r a t i v e gel electrophoresis, fractions comprising approx. 10 k D a ranges were electroeluted and the individual fractions used for immunization. By means of i m m u n o g e n fractions purified by p r e p a r ative gel electrophoresis the highest fusion rates of all applied p r o c e d u r e s were realized: 942 viable h y b r i d o m a clones in one fusion. A strong i m m u n e response was also reflected in the high serum titer as evaluated by E L I S A and the extremely enlarged spleen. Quantitative evaluation of indirect immunofluorescence d a t a revealed also a high n u m b e r of antibodies (269 positive clones). The generated antibodies were characterized by very different specificities such as the ganglion cell specific monoclonal antibody 3110 (Fig. 2). This p r o c e d u r e p r o v i d e d additional advantages such as selection for protein specific antibodies that are often suitable for Western blot analysis and therefore allow rapid identification of the corresponding antigen. A strong i m m u n e response was also found after immunization with soluble glycoconjugates that had been isolated by lectin affinity c h r o m a t o g r a p h y . Long term immunization ( 2 - 6 months) with several b o o s t e r injections yielded serum titers that were one o r d e r of magnitude higher than short term immunizations (2-3 weeks) (Table I). The n u m b e r of initial clones after short term immunizations were in the range of 103-380 (clones 108-380). The n u m b e r of i m m u n o f l u o r e s c e n c e positive clones were b e t w e e n 102 and 371 (imm. ft.: 102-371). The respective d a t a for long term immunizations were: clones: 387-897; imm. fl.: 308-543. Lentil lectin t u r n e d out to be the best lectin for affinity c h r o m a t o g r a p h y because the total yield of eluted proteins was 10-fold higher than with 3 o t h e r lectins (Concanavalin A , wheat germ aggglutinin and p e a n u t agglutinin). Only 0 . 0 1 - 0 . 0 3 % of the starting tissue protein could be recovered in the latter cases. This is most likely due to the relative weak saccharide affinity of lentil lectin (association constant K a = 2.3 x 102 l/mol). Consequently ligand elution is easier than with Concanavalin A as has been r e p o r t e d earlier 27. Purification of integral m e m b r a n e proteins by detergent phase s e p a r a t i o n p r o v i d e d a less immunogenic fraction that induced s o m e w h a t lower serum titers and a relatively small n u m b e r of h y b r i d o m a clones with CNS

200 binding antibodies (clones: 175-509; imm. ft.: 68-235) (see Table I). Surprisingly this approach provided a considerable number of cell type specific antibodies, which deserve special attention (see below). We found that immunogen activation did not significantly enhance the immune response. Neither hemocyanin, hemoglobin nor serum albumin covalently coupled to neural chick proteins increased the number of CNS antibodies (data not shown). In addition to this complex set of biochemical manipulations two biological interventions in vivo were also tested. Both methods attempted to decrease the number of irrelevant hybridoma clones and therefore increase the general efficiency. To induce immune suppression mice

were inoculated with irrelevant immunogens (e.g. unfractionated tectum opticum from eyeless chick embryos), and the initial immune response was inhibited by application of cyclophosphamide. Immunization with relevant tissue (retina) was started afterwards as detailed in 'Materials and Methods'. Far less than 1% of 2872 different hybridoma clones in this group were found to produce antibodies that bound preferentially to retina tissue. Although a number of variations of intercalating immunization and suppression cycles were tested (lasting for up to 7 months for some animals) no significant improvement in the suppression of irrelevant antigens was found. Therefore the efficiency of this method appears to be very limited. In addition serum titers as

Standard Immunisation Immunogen Activation Immune Suppression Tolerance Induction

1

Homogenisation

Density Gradient Centrifugation Cell Nuclei ] ~ - - - - - - - ~

M.m0raoe I Solubilisation Ultracentrifugation

ICytoskeleton1.-----Antibody Affinity Chromatography Lectin A f fini t yt C hromatography

I

t

1G,ycocon,u ate;1

Detergent Phase Separation

"1

Protei.s

I

6)

Preparative Gel Etectrophoresis

I

" Protein Classes i ~" J(Molecular Weight) I

(~)

P

Fig. 1. Experimental design of immunogen purification and immunization procedures. The basic biochemical methods are interconnected :by thick arrows in the vertical axis (for details see Materials and Methods). Most fractions depicted in frames on both sides of the diagram were used for immunization. These immunogen fractions were used for immunization in four different ways as indicated (top right) for the g e n e r a t i o n of monoclonal antibodies. The combination of immunogen fractions and immunization procedures is shown by encircled crosses o n the right hand side.

201

C

2U1

C4

OFL=,..GCL=,- IPL = " . .

. . . . ' \ . ..... - ~

I N L =,,.-

OPL=,- -

1C8

3110

lZ15

3F2

1B12

1G2

0

Fig. 2. MAb specificties for retina layers. Indirect immunofluorescence staining of paraformaldehyde fixed sections of embryonic day 10 chick retinae by different monoclonal antibodies. The designation of each antibody (2U1, C4, 1C8, 3110, 1Z15, 3F2, IB12, 1G2) is given below individual micrographs. All pictures are aligned and correspond to the phase contrast image of the retina on the left top. Each antibody is specific for distinct tissue layers or cell types. Retina layer markers are given on the left hand side: GCL, ganglion cell layer; INL, inner nuclear layer; IPL, inner plexiform layer; OFL, optic fiber layer; ONL, outer nuclear layer; OPL, outer plexiform layer. Bar = 75/~m.

202 well as the average clone number were in the lowest range of all procedures outlined in Table I (clones: 85-644; imm. ft.: 9-300). The second biological intervention used tolerance induction to eliminate distinct lymphocyte populations in vivo. A b o u t 2.5% of the antibodies of this group bound preferentially (but not exclusively) to the retina as the relevant tissue. With regard to tissue specificity comparison of both methods clearly show that tolerance induction is better than immune suppression. The general immune response using tolerance induction was not very pronounced as judged from serum titers. The average n u m b e r of hybridoma clones per fusion was not very high (Table I) (clones: 127-204; imm. ft.: 48-120). As depicted in Fig. 1, various immunogen fractions were used to generate hybridoma cell lines. In about 80,000 inoculated wells from 58 fusion experiments, approx. 17,000 hybridoma cultures were established. Of these 7500 were found to produce antibodies that bound to chick tissue. The number of propagated cultures were reduced by more than one order of magnitude after the initial immunofluorescence screening on cryostat tissue sections comprising only the most promising hybridoma clones. In this way a hybridoma library was established that allows the identification of a variety of developmentally regulated antigens, which are restricted to distinct retina layers, cell types or cell compartments.

TABLE I General immunization efficiency

a TO indicate the employed immunogen fraction the major biochemical purification method or immunization procedure is given. For immune suppression as well as for tolerance induction different immunogen fractions were used (see Fig. 1), which however caused similar results and are therefore not stated separately. Prep. Gel. El., preparative gel electrophoresis; Lec. Affi. (!), iectin affinity chromatography and long term immunization; Lec. Affi., (s), iectin affinity chromatography and short term immunization; Det. Pha. Sep., detergent phase separation; Immun. Supp., immune suppression; Tol. Induct., tolerance induction, b Serum titers of immunized mice as determined by ELISA. The highest serum dilution is given that still provided a positive signal above background, c Spleen cell number of nonimmunized mice were in the range of 1 x l0 s. d Numbers of hybridoma clones of individual fusions. ~ Numbers represent supernatants that were positive in immunofluorescence tests on fixed cryosections of embryonic chick tissue. All numbers given represent averages of 3-12 fusion experiments, except for Prep. Gel. El.: only 1 fusion. Procedure a

Titer b

Spleen cells"

Clones d

lmm. Fluoresc. e

Prep. Gel. El. Lec. Affi. (1) Lec. Affi. (s) Det. Pha. Sep. Immuno. Supp. Tol. Induct.

1:16,000 1:50,000 1:4000 1:4000 1:8000 1:4000

1.5.109 4,0-10s 4,1.108 4,3.10 s 4,2-10s 3,4.10s

942 580 242 342 261 153

269 363 237 151 113 95

Eight of these monoclonal antibodies are depicted in Fig. 2, representing their staining specificities of the central chick retina at embryonic day 10. The specificities are: M A b 2U1 - - inner limiting membrane, M A b C4 - ganglion cell axons (see also Fig. 4), M A b 1C8 - - neurite fiber layers, M A b 3110 - - ganglion cells, M A b 1ZI5 - optic fiber layer and possibly a subset of amacrine cells, M A b 3F2 - - inner plexiform layer, M A b 1BI2 --possibly a second subpopulation of amacrine cells, M A b 1G2 - - outer plexiform layer. These monoclonal antibodies are likely to facilitate developmental studies with regard to cell type differentiation and lamina formation of the retina. Glia specific antibodies

The identification of glial cells in the chicken retina is easily performed using morphological criteria. All glial cells are radially oriented and span the width of the retina 3~. In contrast to other species, in the chicken all other retinal cells are neurones. Using fractions of integral membrane proteins enriched by detergent phase separation for immunization, resulted in a surprisingly high number of antibodies that bound preferentially or even exclusively to gtial cells. In some fusions more than 25 glial specific antibodies were found. Using other approaches~ the yield of glia specific antibodies was only marginal. The integral m e m b r a n e protein fractions were purified from chick retinae of the second week of incubation. Because retinal glia only starts to develop at this early stage of embryogenesis 25, the specific immune response suggests that the detergent phase separation enriches glial components considerably. The reason for this p h e n o m e n o n is unknown. M A b 2M6 represents a typical antibody of this group. In retina tissue sections immunolabeled cells are radially oriented across the width of the retina 36. Single cell cultures indicate that G F A P positive glial cells do indeed express the 2M6 antigen. Interestingly not all mitotic glial cells, which synthesize G F A P are also 2M6 positive. This became obvious in triple labeling experiments (2M6, GFAP, D A P I ) of mitotic cells with highly condensed chromosomes (Fig. 3). Neurone

specific antibodies

The recommended approach to generate neurone specific antibodies is the purification of glycoconjugates by lentil lectin affinity chromatography. Screening approx. 4300 growing hybridoma clones in this group, several hundred were found that stained preferentially neurones of the retina but not radial glial cells, nonneuronal pigment epithelium and mesenchymal cells of the sclera. A relative high percentage of antibodies turned out to be specific for the neural cell adhesion molecule

203 (N-CAM). Many antibodies as M A b C4, 1C8 and 1B12 revealed high specificities for distinct retina layers (Fig. 2). M A b 1B12 is possibly specific for a subset of amacrine cells. A considerable number of these antibodies recognize antigens that were exclusively expressed on axonal surfaces of projection neurones, and not on cell bodies. Using M A b C4 for in situ labeling of retina and tectum whole mounts demonstrates that only ganglion cell axons are marked (Fig. 4). This observation is supported by results from retinal cell cultures in which the antigen is found exclusively on a minority of long cell processes, but not on cell bodies (Fig. 4). Several hundred randomly selected monoclonal antibodies of this group revealed that 90% were specific for cell surface antigens. Purified membranes as immunogens, e.g. induced only in 41.3% of the cases cell surface specific antibodies. Using other methods, additional, presumably neurone specific, antibodies were generated. The high molecular weight fraction (>100 kDa) of the preparative gel

Fig. 3. MAb 2M6 identifies glial cells. Single cell cultures of embryonic chicken retinae were kept for 1-2 weeks on laminin coated multiwell slides. After fixation triple labeling was performed with MAb 2M6 (a), rabbit antiserum specific for the glial fibrillary acidic protein (GFAP) (c) and the nucleus marker DAPI (d). The corresponding phase contrast image is shown in (b). MAb 2M6 stains the large GFAP-positive cell, but not the small (mitotic) cell with condensed chromosomes (arrows). Bar = 20/~m.

electrophoresis (Fig. 1) induced the ganglion cell specific antibody 3110 (Fig. 2), cytoplasmic membranes as immunogens led to a photoreceptor specific antibody, and a detergent phase separation fraction resulted in an antibody, which is possibly specific for horizontal cells (data not shown). However, these alternative methods did not provide reproducibly neurone specific antibodies.

Antibodies specific for extracellular matrbc (ECM) and cytoskeleton components Monoclonal antibodies specific for the E C M or the cytoskeleton could not be identified unequivocally by histological screening of neural tissue sections. Therefore no quantitative data can be given. However, identification of monoclonal antibodies specific for E C M antigens by indirect immunofluorescence is facilitated in those cases where these antigens are also present in some tissues of mesenchymal origin such as the sclera with

Fig. 4. MAb C4 labels specifically ganglion cell axons. Fiat mounted fixed retina (a) and tectum (b) of day 7 embryos were stained with MAb C4 and a rhodamine labeled secondary antibody. Only fascicles of ganglion cell axons located on the tissue surfaces are marked, but no inner layers as revealed by an incision into the retina tissue (arrows in (a)). Cell culture of retina E6 on laminin after 2 days of incubation: MAb C4 staining (c), corresponding phase contrast image (d). Note that immunoreactivity is restricted to cell processes, whereas the large population of cell perikarya remains unstained. Bars = (a) 70¢tm, (b) 100~m, (d) for (c) and (d) 40/~m.

204 relatively large ECM spaces between single cells. Typical staining of ECM like in the sclera around the eye ball was infrequently observed with all employed immunization protocolls with two exceptions. Substantially more ECM specific monoclonal antibodies were generated by tolerance induction and immune suppression. A typical example is MAb 3F2, which stains specifically the inner plexiform layer of the El0 chick retina (Fig. 2) and cartilage (Fig. 5). Comparison of the immunofluorescence staining with the DAPI labeling of cell nuclei and the phase contrast image clearly reveals extracellular space being marked by MAb 3F2. In addition in retinal cell cultures partially floating fibers, reminiscent of ECM fibrils, could be identified by virtue of this antibody (unpublished observation). Although most ECM components are glycosylated glycoconjugate fractions purified by lectin affinity chromatography did not induce an obvious immune response against ECM antigens. Investigation of neural cells in vitro as well as retinal explant cultures with the aid of randomly selected antibodies from different immunization groups suggested that differential immunization leads also to more monoclonal antibodies specific for cytoskeletal antigens. The question arises whether these similar immunological effects are due the fibrous character of distinct ECM components and cytoskeleton proteins. Some antibodies that bound to the cytoskeleton, displayed also a cell compartment and cell type specificity, MAb 2A1 for example. Double staining of permeabilized retinal axons in vitro with MAb 2A1 and phalloidin to mark actin filaments in the growth cone

indicates that MAb 2A1 labels only the axon (Fig. 6). Histochemical results suggest that the corresponding antigen is exclusively localized in axons of ganglion cells 35. Purification of cytoskeletal proteins according to Go et al. TM and subsequent Western blot analysis indicates that a 140 kDa protein is recognized by MAb 2A1 (Fig. 6).

Species specificity Out of the presented 10 monoclonal antibodies (MAb 1B12, C4, 1C8, 1G2, 1Z15, 2A1, 2M6, 2U1, 3F2, 3110), none crossreacted with mouse, rat or goldfish tissue. Therefore the antibodies appear to be species specific for the chicken.

Culture conditions Besides testing the consequence of different immunization strategies, culture conditions for maintenance of hybridoma clones were also modified in several ways. Plating of cells on a feeder layer of phagocytotic macrophages results in a considerable improvement of the survival rate right after fusion. This way the survival rate increased by a factor of 3. The number of antibody producing hybridoma clones was also increased by a factor of 2 by addition of Ewing sarcoma growth factor to the culture medium. In contrast to Banal et al. 3 no positive effect could be detected for insulin as medium additive. Neither the fusion rate nor the survival rate after subcloning was improved in the presence of insulin. Finally the fusion rate was found to be inversibly proportional to the cell density after fusion. Therefore it

Fig. 5. MAb 3F2 stains extracellular matrix. Paraformaldehydefixed horizontal cryosectionsof embryonic chick head were double labeled with MAb 3F2 (a) and the nucleus marker DAPI (b). Corresponding phase contrast image (c). MAb 3F2 stains exclusivelythe extracellular space of cartilage (arrows) but no other surrounding tissue. Bar = 50/~m.

205

iii:¸¸

i

--saw

Fig. 6. MAb 2A1 binds to a cytoskeletal protein of axons. Embryonic retinae were explanted onto a laminin substrate to induce axon outgrowth. After fixation and permeabilization, explant cultures were double labeled with phalloidin (a) to visualize actin filaments of the growth cone and MAb 2A1 (b). MAb 2A1 stains only the axon, not the growth cone (arrow in (a)). Bar in (b) = 10/~m. Western blot analysis of a purified cytoskeletal fraction reveal a single band at 140 kDa (c). Molecular weight standards are given in kiloDalton (kDa). is recommended to keep the cell density possibly low (less than 2 × 105 cells per microtiter well). DISCUSSION

Methodological considerations Although the invention of hybridoma technology revolutionized neurobiological analysis, it quickly became obvious that the vast majority of potentially important lymphocyte clones of a given animal remain inaccessible because of low fusion rates. In addition monoclonal antibody production remains very time consuming due to the extended immunization periods, the cell culture work and the screening procedures. Therefore it is important to evaluate strategies that improve the overall efficiency. As a basic principle the enrichment of immunogens is highly recommended, even if the biochemical nature of an antigen is unknown. Especially in this case preparative gel electrophoresis is most appropriate for several reasons. Applying whole tissue homogenates to SDS-gel electrophoresis and using separately the different molecular weight fractions including the lipid containing dye front for subsequent fusion experiments guarantees that essentially no immunogen fraction is lost. Using this method, individual molecular components can easily be enriched 20-fold. In conjunction with an elaborate but quick test system, which allows one to reduce the number of propagated hybridoma clones to 5% during the first screening, several fusions can be performed in a short

period. The use of cryosections composed of various tissues of different developmental stages for screening by indirect immunofluorescence enabled us to perform up to two fusions per week. We found that the immune response induced by SDS-denatured immunogens is very pronounced and yields a more divers set of antibodies than any other procedure. It should be taken into consideration that the efficiency of electroelution declines with increasing molecular size of proteins. For example for 10% gels we found that proteins smaller than 60 kDa were readily eluted, whereas the recovery of proteins above 70 kDa was one order of magnitude lower. Highly glycosylated extracellular matrix components were found to be not very immunogenic - - possibly due to the presence of distinct T-suppressor cells (see below). To counteract the bias of endogenous immune regulation Matthew and Sandrock z9 employed the immunosuppressant cyclophosphamide to eliminate distinct inhibitory lymphocyte populations. Using repeated cycles of immune stimulation and immune suppression a number of neural glycosaminoglycan specific antibodies could be generated 29. Although our results confirm this observation, we found that tolerance induction which is believed to eliminate similarly distinct T-suppressor cell clones ~3, is superior to immune suppression with respect to the number of monoclonal antibodies specific for ECM components. In our hands immune suppression techniques com-

206 pared with regular immunization procedures resulted also in a higher number of cytoskeleton specific antibodies. Therefore immune suppression together with prepurification of cytoskeletal proteins by differential solubilization promises to provide new monoclonal antibodies specific for yet unknown cytoskeleton-associated proteins. The number of such proteins has been estimated to be in the range of 1000 (ref. 6). The reason for a higher frequency of compartment specific antibodies is unknown and emphasizes that generation of monoclonal antibodies still depends on empirical data. This conclusion is also drawn from our negative results concerning the lack of a strong immune suppression with the aid of cyclophosphamide. In order to gain CNS specific MAbs, in a series of methodological experiments Matthew and Sandrock z9 immunized mice first with PNS tissue, followed by cyclophosphamide and only afterwards with CNS tissue. The specificity of this approach varied between 0% and 85 % concerning specific suppression. Our data are less positive than the results from Matthew and Sandrock 29 and indicate a very limited suppression, despite the fact that we took into consideration that the suppression is only transient (4-5 weeks). Consequently intercalating cycles of inoculation with tectum tissue, cyclophosphamide suppression and retina immunization were performed. However, the number of retina specific antibodies remained only marginal. Tolerance induction proved to be more reliable with respect to a predictable immune response. The specificity can possibly be further improved by reduction of the time interval between suppression and fusion from 12 to 4 weeks which, however, results in hybridomas secreting predominantely IgM t9. IgG isotypes are preferred for most bio- and histochemical experiments because IgMs are relatively unstable macromolecules 21. Combination of this approach with the use of lymph nodes instead of spleen as the source of lymphocytes for hybridoma generation, immunoglobulin synthesis can be directed to IgG isoforms 19. In summary we recommend the use of immunogen fractions and immunization techniques for the indicated specificities as depicted in Table II for a more efficient design of hybridoma generation. Despite these guidelines it should be stressed that further empirical data are needed to predict distinct immune responses more precisely. Monoclonal antibody specificities Using the retina as a model system we focused our interest on two major aspects of neural histogenesis: namely cell determination and directed neurite extension. Therefore it has been of considerable importance to

generate immunological markers for distinct cell types and distinct developmental stages of neural pattern formation. Some of the monoclonal antibodies presented here appear likely to be useful markers for distinct cell types. Monoclonal antibody 3110 binds predominantly to perikarya of the ganglion cell layer. This monoclonal antibody is of interest, not only because it might serve as a perikarya marker of ganglion cells but also because the corresponding antigen is subject to strict developmental regulation and, therefore, defines a distinct period of cellular differentiation. The antigen is detectable at El0, but not at very late or early embryonic stages when presumptive ganglion cells become postmitotic and arranged in the ganglion cell layer (our unpublished observation). In this respect MAb 3110 is clearly different from Thy-1, which has been described as a ganglion cell marker but is not under such stringent control during development lz. Of similar interest is MAb 2A1, which binds to ganglion cell axons (Fig. 6). Consequently the 2A1 antigen is cell type and cell compartment specific. Some of these features have also been described for axonal proteins such as fasciclin 1, 2 and 3 during the neuronal development of Drosophila 4"3°. Another cell type specific monoclonal antibody is MAb 2M6, which binds only to glial cells, as revealed by immunodouble labeling of single cell cultures with MAb 2M6 and an antiserum against GFAP (Fig. 3). MAb 2M6 does not stain the early embryonic retina (E6), but labels radial structures typical of Miiller glial cells, beginning at

TABLE II Immunization recommendations To increase the percentage of MAbs with defined specificities (MAb specificity) the fraction (Tissue fraction) to be used for immunization is given. The major biochemical method (Purification method) to gain these fractions as well as the immunization procedure (Immunization) is indicated on the right hand side. Affi. Chrom., affinity chromatography; Det. Pha. Sep., detergent phase separation; Diff. Solubil., differential solubilization; Extra. Mat., extracellular matrix; Glyco., glycoeonjugates;Imm. Supp., immune suppression; Int. Memb. Pro., integral membrane proteins; Moi. Weight, molecular weight fractions; Post. Ant., postulated antigens; Prep. GeL El., preparative gel electrophoresis; Tiss. Homo., tissue homogenate; Tol. Induct., tolerance induction. MA b specificity

Tissue fraction

Purification method

Immunization

Neuron Glia Cytoskeleton Cell Surface Extra. Mat. Post. Ant.

Glyco. Int. Memb. Pro. Cytoskeleton Glyco. Tiss. Homo. Mol. Weight

Affi. Chrom. Det. Pha. Sep. Diff.Solubil. Affi. Chrom. no Prep. Gel. EL

Standard Standard Imm. Supp. Standard Tol. Induct. Standard

207 the end of the second w e e k of incubation, when retinal glial cells start to differentiate. A t later stages labeled cell bodies b e c o m e evident in the middle of the inner nuclear layer and radial processes extend to the inner and o u t e r limiting m e m b r a n e 36. T h e absence of 2M6-antigen expression during early stages distinguishes this antigen from vimentin 26. In addition 2M6 immunoreactivity was not found in o t h e r parts of the brain (unpublished) in contrast to G F A P 8 and the glial R4 and R5 antigens, the latter two being expressed also in some cells of mesenchymal origin 9. M A b 1Z15 and IB12 promise to be also helpful m a r k e r s for possibly distinct amacrine cells, because both antibodies stain different s u b p o p u l a t i o n of small cell bodies in the inner half the inner nuclear layer. These cells extend processes into the inner plexiform layer. The second set of monoclonal antibodies comprise those that are specific for d e v e l o p m e n t a l l y regulated antigens of neurite layers of the retina. M A b 2U1 i m m u n o r e a c t i v i t y is restricted to the region of the inner limiting m e m b r a n e separating the inner surface of the optic fiber layer from the vitreous body. Future experiments will focus on w h e t h e r the antigen could possibly serve as an e n d o g e n o u s substrate for ganglion cell axons, since axonal outgrowth preferentially takes place at the inner limiting m e m b r a n e 15.

M A b C4 selectively stains ganglion cell axons as shown by indirect i m m u n o f l u o r e s c e n c e of cryosections, retina whole mounts and single cell cultures (Figs. 2 and 4). O u r unpublished observations suggest that this antigen is not only cell type and cell c o m p a r t m e n t (axon) specific, but also subject to d e v e l o p m e n t a l regulation with a maximal expression during the p e r i o d of axonai outgrowth and target recognition. F u r t h e r m o r e , p r e l i m i n a r y in vitro assays suggest that the antigen represents a fasciculation protein, which might be involved in directed axonal outgrowth (Schlosshauer and Diitting, unpublished). In this respect the C4 antigen is similar to N g - C A M / G 4 as well as to FI132. Biochemical analysis should reveal the relationship b e t w e e n these p r e s u m p t i v e adhesion molecules. In the future it will be i m p o r t a n t to investigate w h e t h e r other monoclonal antibodies which are specific for distinct layers of the retina such as M A b 1C8, 3F2 and 1G2, affect neural histogenesis. Acknowledgements. We want to thank Dr. F. Bonhoeffer for generous support of this project as well as Dr. S. Henke-Fahle for introducing us to hybridoma technology, Dr. D. DOtting for preparation of eyeless chick embryos, Dr. P. Layer and Dr. C. Stiirmer and colleagues for providing microscopes. The technical assistance of D. Grauer, K.H. Herzog, I. Sommer and H. Weber is greatfully acknowledged as well as inspiring discussions with Drs. D. St. Johnston and S. Korsching.

REFERENCES 1 Ansorge, W., Fast and sensitive detection of protein and DNA bands by treatment with potassium permanganate, J. Biochem. Biophys. Methods, 11 (1985) 13-20. 2 Barnstable, C.J., Monoclonal antibodies which recognize different cell types in the rat retina, Nature, 286 (1980) 231-235. 3 Bartal, A.H., Feit, C. and Hirshaut, Y., Modification of HAT medium and hybridoma formation. In A.H. Bartal and Y. Hirshaut (Eds.), Methods of Hybridoma Formation, Humana, Clifton, 1987, pp. 163-179. 4 Bastiani, M.J., Harrelson, A.L., Snow, P.M. and Goodman, C.S., Expression of fasciclin I and II glycoproteins on subsets of axon pathways during neuronal development in the grasshopper, Cell, 48 (1987) 745-755. 5 Bordier, C., Phase separation of integral membrane proteins in Triton X-114 solution, J. Biol. Chem., 256 (1981) 1604-1607. 6 Chen, L.B., Rosenberg, S., Nadakavukaren, K.K., Walker, E.S., Shepherd, E.L. and Steele, G.D., The cytoskeleton. In T.A. Springer (Ed.), Hybridoma Technology in the Biosciences and Medicine, Plenum, New York, 1985, pp. 251-268. 7 Constantine-Paton, M., Blum, A.S., Mendez-Otero, R. and Barnstable, C.J., A cell surface molecule distributed in a dorsoventral gradient in the perinatal rat retina, Nature, 324 (1986) 459-462. 8 Dahl, D., Bj6rklund, H. and Bignami, A., Immunological markers in astrocytes. In S. Fedoroff and A. Vcrnadakis (Eds.), Astrocytes -- Cell Biology and Pathology of Astrocytes, Academic Press, Orlando, 1986, pp. 1-24. 9 Draeger, U.C., Edwards, D.L. and Barnstable, C.J., Antibodies against filamentous components in discrete cell types of the mouse retina, J. Neurosci., 4 (1984) 2025-2042. 10 Ehinger, B. and DoMing, J.E., Retinal neurocircuitry and transmission. In A. Bj6rklund, T. H6kfelt and L.W. Swanson

11 12 13

14

15 16

17 18 19 20

(Eds.), Handbook of Chemical Neuroanatomy. Vol 5: Integrated Systems of the CNS, Part 1, Elsevier, New York, 1987, pp. 389-447. Fazekas de St. Groth, S. and Scheidegger, D., Production of monoclonal antibodies: strategy and tactics, J. Immunol. Methods, 35 (1980) 1-21. Fields, K.L., Cell type-specific antigens of cells of the central and peripheral nervous system, Curr. Top. Dev. Biol., 13 (1979) 237-257. Gammon, G., Dunn, K., Shastri, N., Oki, A., Wilbur, S. and Sercarz, E.E., Neonatal T-cell tolerance to minimal immunogenic peptides is caused by clonal inactivation, Nature, 319 (1986) 413-415. Go, M.J., Tanaka, H., Obata, K. and Fujita, S.C., Developmental changes in phosphorylation state of neurofilament proteins in the chick embryonic optic nerve, Dev. Biol., 134 (1989) 85-102. Goldberg, S. and Coulombre, A.J., Topographical development of the ganglion cell fiber layer in the chick retina. A whole mount study, J. Comp. Neurol., 146 (1972) 507-518. Grumet, M., Hoffman, S. and Edelman, G.M., Two antigenically related neuronal cell adhesion molecules of different specificities mediate neuron-neuron and neuron-glia adhesion, Proc. Natl, Acad. Sci. U.S.A., 81 (1984) 267-271. Halfter, W., Claviez, M. and Schwarz, U., Preferential adhesion of tectal membranes to anterior embryonic chick retina neurites, Nature, 292 (1981) 67-70. Hawkes, R., Niday, E. and Gordon, J., A dot-immunobinding assay for monoclonal and other antibodies, Anal. Biochem., 119 (1982) 142-147. Hockfield, S., A Mab to a unique cerebellar neuron generated by immunosuppression and rapid immunization, Science, 237 (1987) 67-70. Hoffman, S., Sorkin, B.C., White, P.C., Brackenbury, R.,

208

21

22

23

24

25

26 27

28

29

30

31

32

Mailhammer, R., Rutishauser, U., Cunningham, B.A. and Edelman, G.M., Chemical characterization of a neural cell adhesion molecule purified from embryonic brain membranes, J. Biol. Chem., 257 (1982) 7720-7729. Johnstone, A. and Thorpe, R., Enzyme linked immunosorbent assay (ELISA). In lmmunochemistry in Practice, Blackwell, Oxford, 1982, pp. 252-255. Kipps, T.J., Benacerraf, B. and Dorf, M.E., Regulation of antibody heterogeneity by suppressor T cells: dimishing suppressor T cell activity increases the number of dinitrophenyl clones in mice immunized with dinitrophenyl-poly(Glu,Lys,Phe) or dinitrophenyl-poly(Glu,Lys,Ala), Proc. Natl. A cad. Sci. U.S.A., 75 (1978) 2914-2917. Kotrla, K.J. and Goodman, C.S., Transient expression of a surface antigen on a small subset of neurons during embryonic development, Nature, 311 (1984) 151-153. Laemmli, U.K., Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature, 227 (1970) 680--685. Lemmon, V. and Gottlieb, D.I., Monoclonal antibodies selective for the inner portion of the chick retina, J. Neurosci., 2 (1982) 531-535. Lemmon, V. and Rieser, G., The developmental distribution of vimentin in the chick retina, Dev. Brain Res., 11 (1983) 191-197. Lis, H. and Sharon, N., The biochemistry of plant lectins (phytohemagglutinins), Annu. Rev. Biochem., 42 (1973) 541574. Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J., Protein measurement with the folin phenol reagent, J. Biol. Chem., 193 (1951) 265-275. Matthew, W.D. and Sandrock Jr., A.W., Cyclophosphamide treatment used to manipulate the immune response for the production of monoclonal antibodies, J. lmmunol. Methods, 100 (1987) 73-82. Patel, N.H., Snow, P.M. and Goodman, C.S., Characterization and cloning of fasciclin III: a glycoprotein expressed on a subset of neurons and axon pathways in Drosophila, Cell, 48 (1987) 975-988. Ramon y Cajal, S., The Structure of the Retina (Translated by S.A. Thorpe and M. Glickstein), Charles C. Thomas, Springfield, II1, (1933). Rathjen, EG., Wolff, J.M., Frank, R., Bonhoeffer, E and

33

34

35

36

37

38

39

40

41

42

43

44

Rutishauser. U., Membrane glycoproteins involved in neurite fasciculation, J. Cell Biol., 104 (1987) 343-353. Schlosshauer, B., Purification of neuronal cell surface proteins and generation of epitope-specific monoclonal antibodies against cell adhesion molecules, J. Neurochem., 52 (1989) 82-92. Schlosshauer, B., Blum, A.S., Mendez-Otero, R., Barnstable, C.J. and Constantine-Paton, M., Developmental regulation of ganglioside antigens recognized by the Jones antibody, J. Neurosci., 8 (1988) 580-592. Schlosshauer, B., D~itting, D. and Wild, M., Target-independent regulation of a novel growth associated protein in the visual system of the chicken, Development, 109 (1990) 395-4(19. Schlosshauer, B., Grauer, D., DOtting, D. and Vanselow, J., Expression of a novel M/iller glia specific antigen during development and after optic nerve lesion, Development, in press. Schlosshauer, B., Schwarz, U. and Rutishauser, U., Topological distribution of different forms of neural cell adhesion molecule in the developing chick visual system, Nature, 310 (1984) 141-143. St~ihli, C., Staehelin, T. and Miggiano, V., Spleen cell analysis and optimal immunization for high-frequency production of specific hybridomas, Methods Enzymol., 92 (1983) 26-36. Stohrer, R., Lee, M.C. and Kearney, J.E, Analysis of the anti-al~3 dextran response with monoclonal anti-idiotype antibodies, J. lmmunol., 131 (1983) 1375-1379. Towbin, H., Staehelin, T. and Gordon, J., Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications, Proc, Natl. Acad. Sci. U.S.A., 76 (1979) 4350-4354. Turk, J.L., Parker, D. and Poulter, L.W., Functional aspects of the selective depletion of lymphoid tissue by cyclophosphamide, Immunology, 23 (1972) 493-501. Valentino, K.L., Winter, J. and Reichardt, L.F., Applications of monoclonal antibodies to neuroscience research, Annu. Rev. Neurosci., 8 (1985) 199-232. Walter, J., Kern-Veits, B., Huf, J., Stolze, B. and Bonhoeffer, E, Recognition of position-specific properties of tectal cell membranes by retinal axons in vitro, Development, 101 (1987) 685-696. Zipser, B. and McKay, R., Monoclonal antibodies distinguish identifiable neurons in the leech, Nature, 289 (1981) 549-554.