Biochemical and Genetic Studies of Peripheral Myelination in Normal Development and in the Mouse Mutant Trembler Citation Fryxell, Karl Joseph (1983) Biochemical and Genetic Studies of Peripheral Myelination in Normal Development and in the Mouse Mutant Trembler. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/eahp-bx93. https://resolver.caltech.edu/CaltechTHESIS:10112019-091636747 Abstract I developed a radioimmunoassay for P 0 , the major peripheral myelin protein, and adapted an existing radioimmunoassay for myelin basic protein. Results from these assays showed that Schwann cells do not make either protein before myelination begins, and accumulate P 0 and myelin basic protein with the same time course during development. Schwann cells cultured in the absence of neurons do not express detectable levels of either protein, but they do continue to synthesize sulfatide, a myelin sulfolipid, apparently indefinitely, as shown by biosynthetic labeling. The trembler (Tr/+) mouse mutant has a pronounced reduction in peripheral myelin, while central myelin and peripheral unmyelinated nerves appear normal. This myelin deficiency is caused by an autonomous Schwann cell defect. Since the Trembler phenotype is expressed in heterozygotes, most previous studies were confined to Tr/+ mice. Using two alleles, I show here that the six possible genotypes can be ordered as follows: +/+ > Tr j /+ > Tr/+ > Tr j /Tr j > Tr j /Tr > Tr/Tr, based on myelin basic protein radioimmunoassays of sciatic nerve extracts. The amount of compact myelin in each genotype, as shown by electron microscopy, is in good agreement with the radioimmunoassay results, with two important provisos. First, Tr/Tr mice have 1% of wild-type myelin basic protein levels but essentially no compact myelin. Secondly, the first steps in the above genetic series reduce both the average myelin sheath area and the number of myelin sheaths by similar amounts, while the last steps primarily reduce the number of myelin sheaths. These results suggest that, if the activity of the Tr + gene product is reduced below a certain point, myelin protein synthesis can be induced without forming a myelin sheath. Visualization of P 0 by indirect immunofluorescence shows that Schwann cells without compact myelin express much lower levels of P 0 than adjacent Schwann cells, associated with the same axon, that do form compact myelin. Finally, although Trembler Schwann cells proliferate excessively in vivo, their proliferation is not affected by Tr gene dose. In vitro Tr j /Tr j Schwann cell proliferation is apparently normal. Thus, it is likely that the function of the Tr + gene product is not directly concerned with Schwann cell proliferation. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: (Biology) Degree Grantor: California Institute of Technology Division: Biology Major Option: Biology Thesis Availability: Public (worldwide access) Research Advisor(s): Revel, Jean-Paul Thesis Committee: Revel, Jean-Paul (chair) Hood, Leroy E. Hudspeth, A. James Owen, Ray David Defense Date: 17 May 1983 Funders: Funding Agency Grant Number NSF UNSPECIFIED NIH UNSPECIFIED Evelyn Sharp Foundation UNSPECIFIED Jean Weigle Memorial Fund UNSPECIFIED Record Number: CaltechTHESIS:10112019-091636747 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:10112019-091636747 DOI: 10.7907/eahp-bx93 Related URLs: URL URL Type Description https://doi.org/10.1111/j.1471-4159.1980.tb09026.x DOI Article adapted for Chapter 1. https://doi.org/10.1111/j.1471-4159.1983.tb11316.x DOI Article adapted for Chapter 2. ORCID: Author ORCID Fryxell, Karl Joseph 0000-0002-2975-7897 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 11812 Collection: CaltechTHESIS Deposited By: Mel Ray Deposited On: 16 Oct 2019 14:27 Last Modified: 31 Jul 2025 22:00 Thesis Files Preview PDF - Final Version See Usage Policy. 28MB Repository Staff Only: item control page

BIOCHEMICAL AND GENETIC STUDIES OF PERIPHERAL MYELINATION IN NORMAL DEVELOPMENT AND IN THE MOUSE KIJTANT TREMBLER Thesis by Karl Joseph Fryxell In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy California Institute of Technology Pasadena, California 1983 (Submitted May 17, 1983) ii ACKNOWLKDGEMEBTS I would like to thank my great deal advisor, about Jeremy sc1ence and Brockes, teaching me a thereof. I should also acknowledge his participation 1n some of the experiments described in Chapter 4. of my connnittee for their time particularly Jim Hudspeth and and philosophy I thank the members helpful Jean-Paul the for suggestions, and Revel, who generously provided equipment and technical expertise. Pat Koen and Richard Jacobs helped ease my transition from immunofl uor e scence Milton Grooms to electron microscopy. provided competent and Teresa Stevens and reliable help with cell culture and mouse breeding, respectively. Susan Maxwell helped with almost everything, and particularly with my mouse colony, proof-reading, photography, and moral support. Many of my colleagues in the Biology Division have helped to enrich my education. At the risk of omitting most of them, I Dave would like to thank Balzer, Herman Gordon, Kintner, Greg Lemke, James McCasland, Arthur Roach, Stygall for their suggestions, unpublished experimen ts. reagents, Chris and Katie company and/ or I would also like to thank my parents for their support--intellectual, moral, and financial. I g · adly a c knowledge fellowship support from the National Science Foundation, the National Institutes of Health, and the Evelyn provided Sharp Foundation. financial The assistance Jean 1n thesis, which 1s greatly appreciated. the Weigle Memorial preparation of Fund this iii ABSTRACT I developed per i pheral radioimmunoassay a myelin protein, adapted and radioimmun oassay for myelin basic protein. assays showed that Po, for Schwann cells do an major existing Results from these not make before myelination begins, and accumulate the Po either protein and myelin basic protein with the same time course during development. Schwann cel ls express cu l t ur ed detectable 1n the absence of neurons levels of e i ther protein, synthesize sulfa tide, a but myelin do not they do continue sulfolipid, to apparently indefinitely, as shown by b ios yn thet i c labeling. The trembler (Tr /+) reduction 1n peripheral unmyelinated deficiency 1s Since t he Trembler most pr evio us allele s , I ordered as Tr/Tr, peripheral caused mutant has myelin, while central nerves by an follo ws : +/ + on myelin sciat i c n e rve extr acts . pronounced myelin and normal. This myelin autonomous Schwann cell defect. 1s expressed s t u d ies we r e confined that a appear phenotype show here based mouse the in heterozygotes, to Tr/+ mice. Using two s1x possible genotypes can be > Trj/+ > Tr/+ > Trj/Trj > Trj/Tr > basi c protein radioimmunoassays of The amount of compact myelin in each geno type , as shown by electron microscopy, is in good agreement wi th the prOV1SO S . protein the radioimmunoassay with two important First, .I:_/Tr m1ce have 1% of wild-type myelin basic levels first results, steps but essentially no compact myelin. 1n the above genetic series Secondly, reduce both the iv average myelin sheath area and the number of myelin sheaths by similar amounts, while the number of myelin sheaths . activit y of the Tr+ gene last steps primarily r educe the These results suggest that, if the product certain is reduced below a point, myelin protein synthesis can be induced without forming a myel i n Visualization sheath. shows immunofluore s cence myelin express much cells, a sso ci ated mye l in . Finally, that Schwann lower lev e ls of wit h t he of same cells by indirect without compact Po than adjacent Schwann axon, although Trembler Po that do form compact Schwann cells proliferate excessively in vivo, their proliferation is not affected by Tr gene d ose. In vit r o apparently normal. Tr+ gene product proliferation . TrJ /Tr J Schwann cell proliferation 1s Thus, it i s likely that the function of the is not direct l y concerned with Schwann cell v TABLE OF COIITENTS Acknowledgements .......•.•...•..••..•..•.....••........•.. ii Abstract .• •• . ....•...•.•..•••.•.........••••••..•.....•.. iii Table of Contents ............. . ............................. v Introduction ..••.•.....••.•...••.. • ......•....•.••......••. 1 Chapter 1: Synthesis of sulfatide by cultured rat ....•.•. 12 Schwann cells Chapter 2: Development and applications of a •.•..•...... 18 solid-phase radioimmunoassay for the Po protein of peripheral myelin Chapter 3: The Trembler locus of the mouse: two ........ 28 semidomin ant alle l es determine peripheral myelination wi t hout a direct effect on Schwann cell proliferation Chapte r 4: Discontinuous myelination in the Trembler .•••. 58 mouse: gene dose dependence suggests the existence of a threshold in peripheral myelination Append ix ~ Tn~;mbler Schwann cells in culture •..•..•......• 93 1 INTRODUCTION 2 Many cellular interactions adjacent cells; examples formation of junctions, gap occur only between physically include and synapse formation, the myelination. We only are beginning to learn the molecular details of these interactions, but it prov e the seems likely that to be of general int e rac tion results between i n myelination. some of the details will ultimately significance. the axon I and have chosen to study the Schwann cell that The dramatic biochemical consequences of this interaction, and the variety of available mutants, are Moreover, this interaction has important practical advantages. two alternative unmyelin a ted . outcomes: The axon 1s the axon may be myelinated known to determine or this outcome. composed of two types of These po iDt s ar e d i scuss ed in more detail below. Vertebrate nerve axons : periph e ral n e rve l.S rapi dly conducting myelinated con ducting unmyelinated axons. axons and slowly Myelinated axons are surrounded by a compact, multilame ll ate myel i n sheath, which is formed by a Schwann cell (SC). Groups of unmyelin a t ed axons are enfolded by a layer of SC cytoplasm (1,5). Myelin functions primarily as an electrical insulator; the presence o f axonal th is impulse insulation grea tly increases propagation (2). The the velocity of synthesis of the myelin sheath r e sults in an enormous and metabolically costly increase in the membrane surface area of the SC (3) . reason, must be myelination transmitted is limited rapidly to (i.e., those Pe rhaps for this axons motor who~;e axons signals innervating 3 voluntary muse les, or sensory axons mediating the sensation of "sharp" pain), while less urgent signals (i.e., "dull" pain, or commands to involuntary muscles) are generally carried by unmyelinated axons (4). Peripheral SC cytoplasm. ~s unmyelinated axons are enfolded by a layer of The functional significance of this relationship unknown but is usually assumed to be protective or nutritive In contrast , (6). unmyelinated ~n axons the central nervous system are only occasionally wrapped by astrocytes, by oligodendrocytes central nervous in adult [the system; myelin-forming see ( 6)] • peripheral nerve myelinated axon, or several are axons ~n nervous dendrocyte s may SCs are not ~n ~n associated axons, only system but not both with (6). single myelinated Thus, oligo- The difference in vivo has in behavior between SCs and clear molecular consequences vitro (8), as discus sed in Chapters 1 and 2 of this thesis. In the developing nerve, a ll init i a lly of the unmyelinated type. be the some sense be committed to form myelin while (7). oligodendrocytes of More precisely, a given SC unmyelinated Oligodendrocy tes central cells is associated with either a ( 1) . the glial and never myel i na ted, which th~ n it will become axon-SC relationships are If an axon is destined to separately enfolded wraps many times around the axon. by a SC, The SC membranes in this sheath come into very close and stereotyped apposition [ca. 12 nm repeat (5)], to form the distance mature, in fixed compact peripheral myelin myelin; sheath. It see is 4 possible to surgically confront SCs f rom nerve, another without from one nerve with axons ambiguity [i.e., without significant amounts of SC migration, proliferation, or ingrowth of axons of this from other nerves--see Elegant experiments (9-12)]. sort have shown that the percentage of axons that are myelinated depends only on the source of the axons, and not on the source of the SCs That is, (9-12). any SC can form (and maintain) myelin but will do so only if continuously induced by the appropria t e sort of axon. Myelin 1s ve r y plentiful peripheral nervous systems . 1n Moreover, both the lipid/ protein ratio, and hence the density, any membrane nervous fact s all ow one and puri ty ( 3 ). The by to 1n iso l ate s ubcellular composition of illus t rated in Figure 1, is rathe r are found on l y in mye lin . lowest buoyant system the myelin membrane c o nve ntional protein the and the myelin membrane has the highest of central These (3). 10 high yield isolation techniques the simple. myelin membrane, The major proteins Some of the myelin-specific proteins are shared between central and peripheral myelin (for example, the large and small myelin basic pr oteins), while others are not (for example, P 0 ). Specific antiser a directed against many of the s e ava ilable. are an t i -P proteins are 0 ( 13) and Those that concern us here anti-myelin basic protein. The latter antibody is often raised against a mixture of large and small myelin basic proteins, as they immunological crossreactivity (21). show almost complete These reagents have been 5 extensively characteri zed immunoautoradiography sodium Using gels (13). these was able I indirect of polyacrylamide sera , by to immunofluorescence and dodecylsulfate radioimmunoassays show quantitatively based that on Schwann cells do not make either protein before myelination begins, and accumu la te course Po during and myelin basic d evelopment protein (Chapter 2) . with the same The levels of time both myelin basic pro t ein and Po are reduced in the Trembler mutant mouse , a ma j or topic of thi s t hesis. In many experiments I have me as ured only one or the other of these proteins, however, for technical fluorescent abundant reasons . labeling, sensitive gives presumably peripheral myel i n accessible to antibody. most Anti-P 0 the because protein brightest it (Figure 1) 1s unmuno- the and/or most 1s more Anti-myelin basic protein provides the radio i mmunoassay, per haps because it 1s much more antigenic than Po (22) and produces higher titer antisera . Myelin al s o cont ains some unusual lipids. One of these, sulfatide, is particular l y easy t o identify, since it 1s one of the fe w li pids that can be biosynthetically labeled with 35 s- sulfate . I fo und that SC cultured in the absence of neurons continue to synthesize first example of a sulfatide (Chapter 1) . characteristic myelin This was component the that continues to be expressed by purified SCs in culture. SCs in the Trembler (Tr/+) mouse have a defec t in phenotype the process includes of specific genetic peripheral myelination. peripheral hypomyelination, The reduced Tr /+ axonal 6 conduction and diameter cervical 1.n myelin the trunk, appear completely central nervous system sc increased and Normally unmyelinated nerves, proliferation (14). superior velocity, such as the normal , as does (14). Surgical experiments, 1.n which SCs of one genotype are confronted with axons of another genotype, have shown that the myel in deficiency in Tr/+ nerve is caused by a defect not in the axon (15,16). This conclusion is also supported by tissue culture different absence experiments, genotypes of are fibroblasts Tremb1e~/ wi1d-t.y pe 1.n which SCs and confronted with each as well as (17), in the SC and neurons other by of 1.n the studies of chimeric mice (14). Although no rmally ~mn yelina ted nerves 1.n Tr/+ m1.ce appear identical to their wild-type counterpar ts , if one confronts SCs from thes e myelina t e d (18). Tr/+ nerves n erves), with the competent complete Tr/+ axons (from phenotype normally 1.s produced Since these abnormalities are uncovered only if SCs are challenged to viability of apparently myel inate Tr SCs concerned 1.s by competent axons, essentially normal, 1.n some way with and the s1.nce the Tr defect is myelination itself ( 14) . I t is not known at what stage of myelination the Tr gene acts, a l though the histolog i cal evidence indicates t h at Tr /+ SCs do wrap competent axons individually and are then blocked, apparently at an earlier stage than other known SC autonomous myelination mutants was ( 14 ). Nor it known if the Tr gene product 1.s essential for myelination or not, or, for example, 7 what the phenotype of a homozygous null mutation at the Tr Gene dose experiments should help to answer locus would be. these questions; Tr homozygotes had not previously been studied at the cellular or molecular levels. experiments, lethal. I found that the Encouraged by such a In preliminary breeding Trj /Trj genotype l.S actually clear-cut gene dose effect, I studied this problem more closely, and found that Trj /Trj mice do live to at least 12 days of age and can be identified at this age by gross nerve appearance. Similar criteria of nerve appearance al so appl y to the Tr allele. These results allowed me to do detailed gene dose experiments, reported in Chapter 3, which suggest myel ina t ion that and proliferation. the is also show Trembler SCs. no t gene an product directly Immunofluorescence observat i ons, reported and Tr and 1s essential concerned electron with for sc microscopic in Chapter 4, support these conclusions interesting heterogeneity in myelination by 8 REFERENCES 1. (1975) Webster, H. de F. in Peripheral Neuropathy, eds. Dyck, P. J., Thomas, P. K. & Lambert, E. H. (W. B. Saunders, Philadelphia), Vol. 1, pp. 37-61. 2. Kuf f ler, S. W. & Nicholls, J. G. (1976) From Neuron to Brain: Cellular Approach !£ the Function of the Nervous System. a (Sinauer Associates, Sunderland, Mass.). 3. Morell , P., ed. (1977) Myelin. 4. Guyton, A. 097 1 ) Textbook of Medical Physiology. C. (Plenum Press, New York, N. Y.). (W. B. Saunders, Philadelphia). 5• ( 1976) Landon , D . N. , ed . The Peripheral Nerve. (Chapman and Hall, London) • 6. Peters, A., Palay, S. L. Structure the Nervous Cells. 7. of & Webster, H. de F. System: the (1976) Neurons and The Fine Supporting (W. B. Saunders, Philadelphia). Br ocke s , J. P . , Fryxell , K. J. & Lemke, G. E. (1981) .:!..!. ~ Biol. 95, 215-230. 8. Mirsky, R., "Wint er, J. , Abney, E. R., Pruss, R. M., Gavrilovic, J. & Raff, M. C. (1980) .:!..!_ Cell Biol. 84, 483-494. 9. Weinberg, H. J. & Spencer, P. S. 10. Ag uayo, A. J ., Epp s , J., Charron, L. (1976) & Brain Res. 113, 363-378. Bray, G. M. 0976) Brain Res. 104 , 1-20. 11. Aguayo, A. J., Charron, L. & Bray , G. M. (1976) .:!..!_ Neurocyt ol. 5, 565-573 . 12. Spencer , P. S. & Weinberg, H. J. (1978) Pathobiology of Axons , ed. Waxman, S. G. N. Y.) pp. 389-405. 1n The Physiology and (Raven Press, New York, 9 13. Brockes, ~ (1980) 14 . Bray, P., J. G. Raff , M. C., Nichiguchi, D. J. Winter, & J. Neurocyto1. 9, 67-77. M., Rasminsky, M. Aguayo, & J. (1981) J., Perkins, S. (19 7 9) A. Ann. Rev. Neurosci. 4, 127-162. 15 . 16. Aguayo, A. G. M. (1977) Attiwe11, J., Aguayo , A. M., Trecarten, S. & Bray, Nature (London) 265, 73-75. Bray, G. M. & Per kins, J., C. Ann . N. Y . Acad. Sci. 317, 512-533. 17. Bunge, R. P ., Bunge, M. B. , Okada, E. & Cornbrooks, C. J . (1980) in Neurob iol ogica l Mu t at i ons Affecting Myelination, ed. Baumann, N. (Elsevier/North Holland Biomedical Press, Amsterdam) pp. 18 . Perkins , C. S., Aguayo, A. J. & Bray, G. M. (1981) S., S. E. 433-446. ~ Neurol. 71, 515-526. 19. Greenfield, Brostoff, W. & Hogan, (1980) L. J. Ne ur ochem. 34, 453-455. 20. Winter, J. 21. Milek, (1982) Natur e (London) 298, 471-472. D. J.' Sarva s , H. Brostoff, s. w. (1981) Brain Re s. 208' 387-396. 22. Whitaker, J . N. (1980) Ann. Neurol. 9, 56-64. 23 . Eyl a r, 0.' Greenfield, E. H.' Ishaque, A. & C hemis t r~ and Biolog~, ed . Hashim, New York, N.Y.) pp. 37-53. Szymanska, G. I. A. s.' Weise, (1980) (Alan R. M. J. & l.n Myelin: Liss, Inc., 10 Figure 1. rat The protein content of purified rat central myelin (C), and peripheral c onventional myelin (P). techniques electrophoresis, and Myelin (3), stained was purified and delipidated by separated by NaDodso 4 polyacrylamide gel with Coomassie LBP, Blue. large basic protein of centra l myelin; SBP, small basic protein of central myel i n. SBP is related to LBP by an internal deletion of 40 amino acids ( 3). The basic proteins of peripheral nomen clature. This figure corre s pond ing exactly in extreme l y fa in t band o f shows myelin a have a complicated myelin protein SBP, as well s li ghtly lower molecular weight. The molecular peripheral more weight to "P 2 ," actually applies only to the latter band (19-2 1 ). band as an label, P 2 represents only about 1% of t he protein in rat or mouse peripheral myelin, but is 10% m myelin r abb it pe ripheral mye l i n and almost (23 ). P2 has a basic pi, but relation to LBP or antigenic SBP 20% has (3, 23) . in bovine no known Ra t peripheral structural and mouse or peripheral myelin also contain two additional basic proteins, however : (the l a tter 1.s shown but not labeled as such), identical to LBP and SBP, respectively (3,19). protein , " which are apparently The term "myel i n basic as used in the literature and in this thesis, does not r efer to all of the basic proteins of myelin, but rather to the subset that react with anti-LBP; namely LBP, found in central myelin. SBP, P 1 and Pr. Photo c ourt esy of J. P . Po and P 2 are not Br oc ke~> . 11 C_E PO 12 CHAPTER ONE Journal of N~uroch~mistry 35(6): 1461-1464, December. Raven Press , New York © 1980 International Society for Neurochemistry 13 0022-3042/8(){ 1201-1461/$02.0010 Short Communication Synthesis of Sulfatide by Cultured Rat Schwann Cells Karl J. Fryxell Division of Biology, California Institute of Technology, Pasadena, California 91125 Abstract: The 3''S sulfolipids synthesized by purified cultures of rat Schwann cells, fibroblasts, and a rat cell line (RN2) were studied. Schwann cell 35 S sulfolipids were almost entirely [3;;S]sulfatide, as shown by TLC in two different solvent systems with unlabeled authentic sulfatide run in the same track . RN2 and fibroblasts did not synthesize significant amounts of sulfatide, by the same criteria. Previous studies failed to detect any characteristic myelin components, including sulfatide, on Schwann cells after several days in culture (Brockes et al., 1980a; Mirsky et al ., 1980). My results show that Schwann cells continue to synthesize some sulfatide in the absence of neurons . Key Words: Sulfatide-Schwann cells-Sulfolipids . Fryxell K. J. Synthesis of sulfatide by cultured rat Schwann cells. J . Neurochem . 35, 1461-1464 (1980). Galactosylceramide 3-0-sulfate (sulfatide) is considered to be a "myelin-typical" lipid, because it is found in relatively high concentrations in purified myelin, but is probably not specific to myelin (Norton, 1977) or even to the nervous system (Green and Robinson, 1960). The synthesis and accumulation of sulfatide have been correlated with myelination, both in vitro and in vivo (Silberberget al. , 1972 ; Norton and Poduslo , 1973 ; Fry et al., 1972, 1974; Wiggins et al., 1975; Tennekoon et al., 1977), and may become a useful assay for in vitro myelination . Among cultured primary central nervous system (CNS) cells, only oligodendrocytes are sulfatide-positive by indirect immunofluorescence criteria (Raff et al ., 1979). In dissociated cultures of the dorsal root ganglion or sciatic nerve, only the Schwann cells (SC) are sulfatide-positive. Not all of the SCare sulfatide-positive , however, and all become sulfatide-negative within a period of 4 days in culture , as determined b y immunofluorescence (Raff et al., 1979; Mirsky et al. , 1980) . [3 ''S]Sulfatide synthesis has been observed in cultures of partiall y purified oligodendrocytes (Poduslo et al., 1978) and in some clonal cell lines derived from the CNS (Dawson et al ., 1977) , but not by the fibroblasts , astrocytomas , or neuroblastomas so far examined (Dawson et al., 1972, 1977). It seems very likely that SC synthesize the sulfatide that is found in peripheral myelin , although this has not been demonstrated directly. Sulfatide may be synthesized (in varying amounts) by all SC. Alternatively , sulfatide synthesis by SC may occur only after interaction with an appropriate class of axon, as seems to be the case for several peripheral myelin proteins (Brockes et al., 1980a; Mirsky et al ., 1980) . I have exploited recently developed techniques for the purification and growth of SC in culture (Brockes et al ., 1979) to distinguish these alternatives by studying the sulfolipids synthesized by rigorously purified cultures of SC. Cultures of RN2 cells and fibroblasts were also studied. RN2 is a clonal cell line derived from an ethylnitroso urea-induced tumor of rat cervical spinal nerve root (Pfeiffer and Wechsler, 1972). It is thought to be a Schwannoma, because of its origin and its synthesis of 2':3'-cyclic-nucleotide 3'-phosphodiesterase (EC 3. 1.4.37) and S100 protein (Pfeiffer et al ., 1977). RN2 was included in this study because the question of whether or not it synthesizes sulfatide, a third potential SC characteristic, is presently controversial (Dawson et al. , 1977; Lucas et al., 1977) . Fibroblasts were included as a control, since present evidence indicates that fibroblasts do not synthesize detectable levels of sulfatide (Dawson et al . ' 1972). Received February 14, 1980; accepted June 9, 1980. Address correspondence and reprint requests to K . J. Fryxell at the above address. Abbreviations used: CM , chloroform :methanol (proportions by volume); MEM , Modified Eagle's medium ; PBS, Dulbecco's phosphate buffered saline without Ca and Mg; SC, Schwann cells . Sulfatide is used in this paper to denote galactosylceramide 3-D-sulfate. MATERIALS AND METHODS Dissociated SC were prepared from neonatal rat sciatic nerve. Such cultures are composed of only two cell types, 1461 14 1462 K . J. FRYXELL SC and fibroblasts (Brockes et al., 1977). Fibroblasts were eliminated by treatments with cytosine arabinoside and monoclonal anti-Thy 1.1 plus complement , essentially as described (Brockes et al ., 1979; 1980a). This technique produces cultures that are routinely >99 .5% SC (Brockes et al., 1979). The SC used in these experiments were cultured for at least 14 days , and passaged 1-3 times. The culture medium was supplemented with a partially purified bovine pituitary factor (Brockes et al. , 1979; 1980b) in order to stimulate division. This factor was washed out of the medium at the time of isotope addition , or 4 days before isotope addition . RN2 cells (originally obtained from Dr. S. Pfeiffer) were used at passages 1 1-17. Fibroblasts were obtained from thigh muscle of neonatal rats, and were used at passages 4-7. All cells were maintained in plastic petri dishes (Falcon), and grown in 10% fetal calf serum in Dulbecco' s MEM (Gibco) with added penicillin and streptomycin , in a humidified atmosphere of 10% C0 2 at 36°C. Confluent cultures were incubated with added N~­ ssso. (700-800 Ci!mol ; New England Nuclear) for 48 h, washed 3-5 times with HEPES buffered MEM, once with PBS , and removed by gentle rinsing with PBS plus EDT A (5 x JQ- • mol/liter). The cells were collec.ted by centrifugation, resuspended in PBS (0.2-1 ml), and extracted with 20 vol. of CM (I : 1). Aliquots for protein analysis were removed either before or after the addition of CM (I : 1), stored at 4°C , and assayed essentially by the method of Lowry et al. (1951). The remainder was centrifuged to remove insoluble material, and the pellet was washed once with CM (1 :1) . Lipids were extracted from the combined ,supernatants essentially by the method of Folch et al . (1951) , except that 0.43% K2S0 4 was used in place of H 20 , as described by others (Silberberget al., 1972; Fry et al., 1974; Wiggins et al., 1975). Following the initial solvent partition, the lower phase was re-extracted twice with pure upper phase, and then transferred to a clean test tube and evaporated to dryness under a stream of N2 gas in a 37°C water bath . The residue was redissolved in CM (2: 1) . An aliquot of this solution (0. 1-ml) was either diluted directly into Aquasol-2 (15 ml; New England Nuclear), or evaporated to dryness and redissolved in Aquasol, and counted in the 14 C channel of a scintillation counter. Another aliquot of the sample was used for TLC. Silica G plates (250 J,Lm; Analtech) were used, in either of two solvent systems: CHCI 3 -CH 30H-H 2 0 (70:30:4 by vol. , solvent 1), or CHCI 3-CH 3 0H-15 mol/liter NH 4 0H (14:6: I by vol., solvent II) . Solvents were from Mallinckrodt or Matheson Coleman and Bell, and were all reagent grade or better. After solvent development , the plates were marked with radioactive ink and exposed to Kodak X-ray film at -70oc with a DuPont "Cronex " intensifying screen, to detect [3 ''S)sulfolipids . To detect total lipids, the plates were then sprayed with 50% H 2S0 4 , charred on a hot plate , and photographed. Preliminary experiments (not shown) demonstrated that the small amount of cellular lipids used here produced negligible charring in the vicinity of the sulfatide bands , and that the sulfatide standard alone produced no image on the autoradiograph . Thus, the standard and unknown could be detected separately even when they were mixed together before spotting, allowing precise comparison of mobilities . Therefore, authentic sulfatide from bovine brain (P-L Biochemicals Inc ., Milwaukee, Wisconsin) in CM (2 : I) was added to each sample before it was spotted onto the TLC plate. RESULTS Both SC and RN2 incorporate significant amounts of '• S into lipid, whereas fibroblasts do not ( <2% of SC incorporation per microgram of protein ; see Table 1). Considering the 3 ''S specific activity used (Table 1) , the normalized 3 ''S incorporation by SC is roughly comparable to the [3''S)sulfatide synthesis previously reported in cultures of oligodendrocytes (Poduslo et al., 1978) or cell lines (Dawson et al. , 1977), and about an order of magnitude lower than in myelinating spinal cord ex plants (Fry et al., 1972). The much lower counts consistently obtained in the fibroblast samples show that inorganic 3''S contamination of the lipid fractions was negligible , and strongly suggest that the occasional fibroblast contaminating the SC cultures did not affect the results . After TLC, SC lipids show three bands of 3 ·'S sulfolipids . The two major species comigrate precisely with the two bands of authentic sulfatide, in the same track, and in each of two different solvent systems (Fig . 1). The identification of these species as sulfatide rests not only on their TLC mobilities, but also on the fact that they are sulfolipids and are synthesized by SC but not by fibroblasts or RN2 . Further structural studies to confirm this identification would be of interest. RN2 lipids show only one band of 3 :;S incorporation , which does not comigrate with either band of the sulfatide standard (Fig. 1). 3 TABLE 1. Incorporation of 3·'S04 into extractable lipids by various cultured cells Sample SC culture I SC culture 2 RN2 culture I RN2 culture 2 Fibroblast culture I Total radioactivity incorporated (c .p.m.) Protein (mg) Normalized incorporation (c.p .m./mg) 302 267 0.33 0.32 834 489 439 1.52 1.52 289 19 1.18 16 915 322 Cultures were incubated with 0.4 mCi of 35 S0 4 per 10 em petri dish (estimated final specific activity in the media = 60 Ci/mol) . Cellular lipids were extracted and counted as described in Materials and Methods . Each culture represents two petri dishes. The total radioactivity incorporated is corrected for background. J . Ne,urochem ., Vol . 35 , No . 6 , /980 15 SVLFATIDE SYNTHESIS BY RAT SCHWANN CELLS 1·0 ;... . ! ,. ( ...... •- ' ... , ....._ ... !- •. 0 I. ,.. • ...... '~ ,.- sc I sc II t ...... ~ ~ RN2 RN2 I II FIG. 1. Thin layer chromatography of sulfolipids from cultured SC and RN2. Cultures were labeled with 2 mCi of 3'S0 4 per 6-cm petri dish and extracted as described in Materials .and Methods. Approximately 475 c.p.m. of cellular lipids were mixed with 15 1-Lg of nonradioactive, authentic sutfatide, applied to the TLC plates, chromatographed, exposed to X-ray film for 13 days, and charred as described in the text From left to right: SC lipids, developed in solvent I; SC lipids, developed in solvent II; RN2 lipids, solvent I; RN2 lipids, solvent IL Within each column, the autoradiograph (i.e., 35 S sulfolipids) is shown on the left , and the same track of the TLC plate after charring (i.e. , total lipids) on the right Arrows show the authentic sulfatide bands on the charred plate. The absolute mobility of the sulfatide bands has varied between experiments (upper band R, of 0.32-0 .37 in solvent I; 0.36-0.36 in solvent II). Thus, the difference in the absolute mobility of sulfatide between the two solvents shown above may not be significant This does not affect the interpretation of the results, which are based solely on relative mobilities within a given track. DISCUSSION In previous studies, Dawson et aL (1977) found that solid tumors of RN2 (or its subclone RN22) did not synthesize significant amounts of sulfatide, and RN22 grown in monolayer culture did not incorporate 35 S0 4 into the sulfatide spots on their TLC system. The general question of whether or not RN2 synthesizes any sulfolipid(s) was apparently not investigated . Lucas et aL (1977) found that RN2 did synthesize detectable amounts of a sulfolipid, which appeared to comigrate with sulfatide in their TLC system. Under somewhat similar conditions, I found that the mobilities were difficult to distinguish unless the standard and unknown were run in the same track. In another solvent system (II), however, the difference in mobility is much greater (Fig. 1). My results thus appear consistent with both earlier reports, and suggest that RN2 makes a single major species of sulfolipid, which is not identical to sulfatide. Mirsky et aL (1980) found that purified SC lose the ability to be labeled by antibody to sulfatide after a few days in culture . My results show, however, that SC continue to synthesize measurable amounts of sulfatide after more than 2 weeks in culture . Whether the rate of sui- 1463 fatide synthesis changes with the age of the culture , or some other change is responsible for the immunofluorescence results, is unclear. Sulfatide synthesis appears to be a long-lasting property of oligodendrocytes in culture as well as SC (Raff et al., 1979; Mirsky et aL, 1980). This suggests the need for caution in the interpretation of sulfatide synthesis data from myelinating cultures, as factors other than myelination-e.g., glial cell proliferation-may affect the net rate of sulfatide synthesis. A continuing interaction with a myelin-inducing axon seems to be required for SC to synthesize detectable amounts of the peripheral myelin proteins P0 , PI> and P2 (Brockes et aL, 1980a; Mirsky et aL, 1980). Sulfatide, in contrast, is a characteristic component of the myelin membrane that is synthesized in detectable amounts by SC in the absence of neurons . ACKNOWLEDGMENTS I thank J. P. Brockes for many helpful suggestions and financial support, D. J. Nishiguchi for providing some of the cells used in this study, and A. J. Hudspeth for the use of photographic facilities. This work was supported by National Institutes of Health grant NS 14403 to J . P. Brockes. I am a National Science Foundation Graduate Fellow. REFERENCES Brockes J.P., Fields K. L., and Raff M. C. (1977) A surface antigenic marker for rat Schwann cells. Nature , Lond. 266, 364-366. Brockes J.P., Fields K. L., and Raff M. C. (1979) Studies on cultured rat Schwann cells. I. Establishment of purified populations from cultures of peripheral nerve . Brain Res. 165, 105-118. Brockes J . P., Raff M. C., Nishiguchi D. J., and Winter J . (1980a) Studies on cultured rat Schwann cells. III. Assays for peripheral myelin proteins. J. Neurocytol . 9, 67-77. Brockes J . P., Lemke G. E., and Balzer D. R. Jr. (1980b) Purification and preliminary characterization of a glial growth factor from the bovine pituitary. J. Bioi. Chern ., in press . Dawson G., Matalon R., and Dorfman A. (1972) Glycosphingo1ipids in cultured human skin fibroblasts. I. Characterization and metabolism in normal fibroblasts. J. Bioi. Chern. 247, 5944-5950. Dawson G., Sundarraj N ., and Pfeiffer S. E. (1977) Synthesis of myelin glycosphingolipids (galactosylceramide and galactosyl (3-0-sulfate) ceramide (sulfatide)) by cloned cell lines derived from mouse neurotumors . J. Bioi. Chern. 252, 2777-2779. Folch J., Lees M., and Sloane Stanley G. H . (1951) A simple method for the isolation and purification of total lipides from animal tissues. J . Bioi. Chern. 226, 497-509. Fry J. M., Lehrer G. M., and Bomstein M. B. (1972) Sulfatide synthesis: Inhibition by experimental allergic encephalomyelitis serum . Science (NY) 175, 192-194. Fry J . M., Weissbarth S., Lehrer G. M., and Bomstein M. B. (1974) Cerebroside antibody inhibits sulfatide synthesis and myelination and demyelinates in cord tissue cultures. Science (NY) 183, 540-542. Green J. P. and Robinson J.D. Jr. (1960) Cerebroside sulfate (sulfatide A) in some organs of the rat and in a mast cell tumor. J. Bioi. Chern . 235, 1621-1624. Lowry 0. H ., Rosebrough N.J., FarrA. L., and Randall R. J . (1951) Protein measurement with the folin phenol reagent. J . Bioi. Chern. 193, 265-275 . J . N~urochem ., Vol. 35 , No. 6 , /980 16 1464 K. J. FRYXELL Lucas A., Flintoff W., Anderson R., Percy D., Coulter M., and Dales S. (1977) /n vivo and in vitro models of demyelinating diseases : Tropism of the JHM strain of Murine Hepatitis Virus for cells of glial origin . Cell 12, 553-560. Mirsky R., Winter J., Abney E. R., Pruss R. M.• GavriloviC J., and Raff M . C . (1980) Myelin-specific proteins and glycolipids in myelin-forming cells in culture : Differences between rat Schwann cells and oligodendrocytes . J . Cell Bioi. 84, 483-494 . Norton W. T. (1977) Isolation and characterization of myelin, in Myelin (Morell, P., ed), pp. 161-199. Plenum Press, New York . Norton W. T. and Poduslo S . E. (1973) Myelination in rat brain : Changes in myelin composition during brain maturation . J . Neurochem . 21, 759-773 . Pfeiffer S . E . and Wechsler W. (1972) Biochemically differentiated neoplastic clone of Schwann cells . Proc. Nat/ . Acad. Sci., USA 69, 2885-2889. Pfeiffer S. E., Betschart B., Cook J ., Mancini P., and Morris R. (1977) Glial cell lines, in Cell, Tissue, and Organ Cultures in J . Neurochem ., Vol. 35 , No . 6, /980 Neurobiolog y (Federoff S . and Hertz , L., eds) , pp . 287-346. Academic Press , New York. Poduslo S. E ., Miller K., and McKhann G. M. (1978) Metabolic properties of maintained oligodendroglia purified from brain. J . Bioi. Chern . 253, 1592-1597. Raff M. C., Fields K . L. , Hakomori S.-1., Mirsky R ., Pruss R. M., and Winter J . (1979) Cell-type-specific markers for distinguishing and studying neurons and the major classes of glial cells in culture . Brain Res. 174, 283-308 . Silberberg D. , Benjamins J ., Herschkowitz N ., and McKhann G . M. (1972) Incorporation of radioactive sulphate into sulphatide during myelination in cultures of rat cerebellum . J . Neurochem . 19, 11-13 . Tennekoon G . 1., Cohen S. R. , Price D. L., and McKhann G. M. (1977) Myelinogenesis in optic nerve: A morphological , autoradiographic, and biochemical analysis . J . Cell Bioi. 72, 604-616. Wiggins R. C., Benjamins J . A., and Morell P. (1975) Appearance of myelin proteins in rat sciatic nerve during development . Brain Res . 89, 99 - 106. 17 Thin Layer 35 Chromatography of s - Sulfolipids 1· 0 •( >- .: : 0·5 .D 0 .... .... .... .... ~ 0 . .... .... sc sc RN2 RN2 I II I II Figure 1. For legend, see page 15. .... .... 18 CHAPTER TWO Journal of N~urochemistry Raven Press , New York © 1983 International Society for Neurochemistry 0022-3042/83/020 1-05381$03.00/0 19 Development and Applications of a Solid-Phase Radioimmunoassay for the P0 Protein of Peripheral Myelin Karl J. Fryxell, David R. Balzer, Jr., and Jeremy P. Brockes Division of Biology, California Institute of Technology, Pasadena, California, U.S.A. Abstract: Thi s is the first report of a quantitative radioimmunoassay for P0 • The assay uses antigen-coated plastic microwells, with antibody binding detected by 12 5J-labeled protein A. Either peripheral myelin proteins or purified P 0 may be used as the antigen. Optimal extraction of tissue samples for P 0 immunoassay requires careful attention to the sodium dodecyl sulfate-to-protein ratio . Sodium dodecyl sulfate interference with antibody binding can be minimized by adding an excess of non ionic detergent and carrier protein to the incubation buffer. This method allows the detection of 0.8 ng of P 0 (20 ng/ml) . Results from this assay showed little or no immunoreactivity in extracts of brain, central myelin, liver, purified myelin basic proteins, cultured, purified second- ary Schwann cells, or membrane preparations from these cells . P0 was clearly detectable in Schwann cell cultures from 3- to 4-day-old rats at 12-18 h after dissociation (4'7c of the level in adult sciatic nerve) and in extracts of oneday-old rat sciatic nerve (2% of the level in adult nerve). Myelin basic protein radioimmunoassays showed that the ratio of P 0 to myelin basic protein is essentially constant in extracts of sciatic nerve from one-day-old, four-dayold, and young adult rats . Another result was that P0 levels are reduced in the trembler mouse sciatic nerve. Key Words: Myelin proteins-Radioimmunoassay. Fryxell K. J. et al. Development and applications of a solid-phase radioimmunoassay for the Po protein of peripheral myelin . J. Neurochem . 40, 538-546 (1983) . P 0 is the major protein of peripheral myelin (Greenfield et al., 1973; Wood and Dawson, 1973) . It accounts for 50-60% of the protein in mammalian peripheral myelin (Greenfield et al., 1973), but has not been detected in central myelin (Braun and Brostoff, 1977; Brockes et al. , 1980). P0 is a glycoprotein with an apparent molecular weight of 28,000-30,000 by sodium dodecyl sulfate (SDS) gel electrophoresis (Brostoff et al., 1975; Benjamins and Morell, 1978 ; Roomi et al., 1978; Eylar et al., 1979). It is soluble in aqueous solutions of detergents such as SDS or in acidic organic solvents, but not in simple aqueous solutions or in neutral organic solvents (Greenfield et al., 1973; Brostoff et al., 1975). Previous measurements of P0 levels have relied on densitometry of SDS-polyacrylamide gels stained with Fast Green (Greenfield et al., 1973; Wiggins et al., 1975) or immunoautoradiography of SDS gels (Brockes et al., 1980). These methods are time consuming, and the former is particularly limited in sensitivity and specificity. So far as we are aware, no quantitative radioimmunoassay (RIA) for P0 has been described. Rabbit antisera have been produced against P0 purified from bovine (Trapp et al., 1979) and rat (Brockes et al., 1980) peripheral myelin. The use of such antisera for quantitative measurement of Po levels would be of interest for several reasons. Unlike myelin basic protein assays (Schmid et al., 1974; Cohen et al., 1975; Delassalle et al., 1980; Groome, 1980), .which provide a quantitative index of both central and peripheral myelination (Cohen and Guarnieri, 1976), a P0 assay would be specific for peripheral myelin . An assay for P0 might be a more sensitive way of detecting peripheral myelin: for example rat peripheral myelin contains roughly 10-fold more P0 than basic protein 1 (Greenfield et Received May 12, 1981: revised August 27, 1982; accepted August 30 , 1982. Address correspondence and reprint requests to K. J. Fryxell, Division of Biology, 216-76, California Institute of Technology, Pasadena , California 91125 , U.S .A . The present address of Jeremy Brockes is MRC Cell Biophysics Unit , 26-29 Drury Lane , London WC2B 5RL , England. Abbreviations used: BSA, Bovine serum albumin; HMEM , HEPES-buffered Eagle's modified minimum essential medium ; PBS , Phosphate-buffered saline: PBS/TW, Phosphate-buffered saline plus 0.5 mg/ml Tween 20 ; RIA, Radioimmunoassay : SDS, Sodium dodecyl sulfate . 1 The term myelin basic protein, as used here , includes the large and small basic proteins of rodent central myelin, as well as the immunologically related basic proteins of rodent peripheral myelin (P,. PR, and P8 -see Greenfield et al., 1980: Milek et al. , 1981). It does not include the P 2 protein of peripheral myelin . which is basic, but appears to be immunologically unrelated to the above (Milek et al., 1981) and is only a minor component of rat and mouse peripheral myelin (Greenfield et al. , 1980). 538 20 P 0 RADIOIMMUNOASSAY al., 1973, 1980). The issue of sensitivity is particularly important to us, since we are interested in measuring the production of myelin by cultured rat Schwann cells in the presence and absence of neurons. Finally, measurement of both P0 and myelin basic protein in the same sample may be useful in some cases. We therefore set out to define the conditions under which our rabbit antiserum (Brockes et al., 1980) could be used in a quantitative immunoassay. In this report we describe a RIA for P0 , as well as assay results from developing nerve, cultured cells, and the mouse mutant trembler. MATERIALS AND METHODS Bovine serum albumin (BSA) , essentially globulin-free, polyoxyethylene sorbitan monolaurate (Tween 20), octyl phenoxy polyethoxyethanol (Triton X-100), and gelatin, type I from swine skin, were from Sigma. SDS was from Bio-Rad or Matheson, Coleman and Bell. Phosphatebuffered saline (PBS) was prepared as 0.05 moVL sodium phosphate pH 7.3, plus 9 mg/ml NaCI and I mg!ml NaN 3 • PBS plus Tween 20 (PBS/TW) contained 0.5 mg/ml Tween 20. Wistar rats were obtained from Hilltop Lab Animals (Chatsworth, CA), and were bred in this laboratory . Trembler mice (Rex Tremblerj/ + + in the C57BL/6J strain) and wild-type controls (C57BL/6J) were obtained from the Jackson Laboratory (Bar Harbor, Maine) and bred in this laboratory. Myelin basic proteins, purified from rat brain, were a gift from Dr. D. McFarlin. Anti-P0 was prepared by immunizing rabbits with the Pu band from SDS gels of rat peripheral myelin proteins and was the same reagent as previously described (Brockes et al., 1980, 1981; Mirsky et al., 1980). Protein A iodination Staphylococcus aureus protein A (Pharmacia) was iodinated by the chloramine T method essentially as described (Brockes et al., 1980), except that the 12 5J-Iabeled protein A was desalted by centrifugation (Neal and Fiorini, 1973) through a small column of Sephadex G-25 medium beads (50-150 JLm, Sigma). The specific activity of the 12 5J-Iabeled protein A was determined by precipitation of an aliquot of the iodination mixture (prior to desalting) with 10% ice-cold trichloroacetic acid. The specific activity was typically 300-350 Ci/mmol. In more recent experiments, protein A has been iodinated to a higher specific activity (2000-3500 Ci/mmol), by iodinating tenfold Jess protein A (5 JLg) with slightly more chloramine T and 12 5J as described (Bolton, 1977). The reaction was quenched with tyrosine and desalted by centrifugation as described above . Both preparations of 125I-Iabeled protein A were 50-60% retained by rabbit immunoglobulin G coupled to Sepharose 4B. Myelin preparation Peripheral myelin was prepared from rat sciatic nerves (Pel-Freez Biologicals, Rogers, Arkansas) by a shortened version of the method of Everly et al. (1973); two discontinuous sucrose gradient steps and two osmotic shock steps were generally used. T~e myelin was lyophilized, delipidated with 2:3 CH 3 CH 2 0H:(CH 3 CH 2h0 (voUvol) 539 and (CH 3 CH 2 ) 2 0 (Everly et al., 1973), dried with a stream of N 2 gas or brief lyophilization, and suspended in 2.5 mg!ml SDS. The SDS solution was then immersed in a boiling water bath for 60 s and centrifuged to remove insoluble material (Eppendorf microfuge: 15,600 g x 1 min). The supernatant was stored at -20°C until needed . Solubilized peripheral myelin proteins were boiled for 4 min in sample buffer and electrophoresed in discontinuous, Tris-buffered 15o/c polyacrylamide SDS slab gels essentially as described (Hubbard and Lazarides, 1979). The gels were stained with Fast Green and scanned at 580 nm (Greenfield et al., 1973) for determination of the proportion of P 0 • No attempt was made to correct for differences in dye-binding capacities between proteins . A shoulder presumably representing Y was sometimes observed and was included in the P0 peak when we quantitated these gel-scans, as was also done by Greenfield et al. (1973). Cammer et al. (1980) recently found that theY band, which has a slightly lower apparent molecular weight than P0 (Benjamins and Morell, 1978), represents unreduced P 0 • Peripheral myelin proteins, solubilized and assayed as described, typically contained 2.5-7.5 mg!ml protein, of which about 30% was P0 • These myelin proteins can be used as a source of P 0 antigen for antibody competition and for coating plastic microwells (see below). A preliminary experiment indicated that repeated freeze-thawing of solubilized peripheral myelin did not reduce its ability to compete for anti-P0 binding. Therefore, all solubilized samples were stored at -2ooc between assays. Central myelin was prepared from adult rat brain (dissected as described under Sample preparation) by the method of Norton and Poduslo (1973). The central myelin was then lyophilized, delipidated, and solubilized as described above for peripheral myelin, except that I, rather than 2.5, mg!ml SDS was used. Purified P 0 preparation Solubilized rat peripheral myelin protein (2 mg) was electrophoresed on a preparative scale (13 em wide x 7 em x 0.5 em), 15% polyacrylamide, SDS gel. The P 0 band could be visualized directly; its location was checked by staining a side track . The band was cut out , electroeluted, and concentrated (Amicon Bl5) . The purified P0 showed a single major band by analytical SDS-gel electrophoresis. The only visible minor bands, at approximately 60 ,000 and 90,000 daltons, presumably represented aggregates of P0 • Densitometry of these gels showed that 63% of the Fast Green stain was well localized to the P0 band. Sample preparation Adult rats and mice were anesthetized with ether and killed by cervical dislocation . Various tissues were dissected out and stored at -7ooc. When dissecting out rat brain, a slice of only the dorsal-most cerebral cortex (white and grey matter) was taken so that there would be no contamination with the peripheral myelin of cranial nerves. After thawing, the tissues were homogenized with a motor-driven glass-on-glass homogenizer (Kontes) in 40 volumes (40 mUg wet weight) of I mglml SDS. The homogenates were then immersed in a boiling water bath for 60s and centrifuged to remove insoluble material (15 ,600 g x I min). The supernatant was stored at -20°C until needed. Dissociated Schwann cells were prepared from J . Neurochem., Vol . 40, No.2. /983 21 540 K. J. FRYXELL ET AL. neonatal rat sciatic nerve. Sciatic nerve cultures are composed of only two cell types , Schwann cells and fibroblasts (Brockes et al .. 1977). The fibroblasts were eliminated by treatments with cytosine arabinoside and monoclonal antiffhy 1.1 plus complement, essentially as described (Brockes et al., 1979, 1980). We used nerves from 3- to 4-day-old rats. as they yield more Schwann cells than those of 1- to 2-day-old rats. Some of the nerves used in this study were stored in liquid N 2 before dissociation for cell culture. This maneuver selects against fibroblasts. Cultured Schwann cells were rinsed 3 times in PBS without Ca and Mg (GIBCO), incubated for 10 min at room temperature in PBS without Ca and Mg plus 5 x 10- 4 mol/L EDTA (Sigma), and removed by gentle rinsing . The cells were collected by centrifugation, sonicated for 5 s (Virtis sonicator model no . 16-850) in PBS plus I mg/ml SDS, and stored at -20°C . After thawing, the extracts were boiled and centrifuged as described above for tissue samples. Schwann cell membranes were prepared by the Tween 40 method of Standring and Williams (1978) except that the cells, which were removed with EDT A as above, were suspended in Tris saline buffer (Standring and Williams, 1978) at 3 x 10 6 cells/mi. The final membrane pellet was dissolved by boiling in I mg/ml SDS as described above. Total protein was determined with a BSA standard (Sigma) by the method of Lowry et al. (1951). Myelin basic protein RIA Myelin basic protein was assayed essentially as described (Cohen et al., 1975; Delassalle et al., 1980), with minor modifications as described below. The incubation buffer was 0.2 mol!L Tris (pH 7.3) + 0.5 mg/ml histone (from calf thymus, Sigma) + 0.5 mg/ml Tween 20 + 1 mg/ml NaN 3 . All tubes in a given assay contained identical SDS concentrations. Bound antigen was separated using heat-killed , formalin-fixed Staphylococcus aureus bacteria (Bethesda Research Labs) . After the usual antibody incubation, the bacteria were added for I h at 4°C, pelleted and washed by centrifugation, and the pellet was counted in a gamma counter. P 0 Radioimmunoassay We used two kinds of round-bottomed microtiter plates in the RIA-rigid polystyrene plates (Flow Labs . No. 76-213-05), which require NaOH for removal of the adsorbed protein (see below), or flexible acrylic plates (Flow Labs . No . 76-362-05) , which may simply be cut up and counted directly in a gamma counter. During all incubations. the microwells were covered and placed inside a humidified plastic box to minimize evaporation. Microwells were coated with antigen by diluting peripheral myelin protein in PBS and incubating 30 ~I per well (approximately 25 ng of P 0 ) for 2 h at 37°C. Competing, soluble antigen and antibody were preincubated together for 1.5 h at room temperature in uncoated microwells . The preincubation solution contained the following: anti-P0 serum ( x 533), Triton X-100 (I mg/ml), BSA (I mg/ml), SDS (0 . 1 mg/ml), and competitor in a final volume of 40 ~I of PBS/TW. The final SDS concentration, including the SDS present in the competitor supernatant, was always constant within a given assay. Each competitor dilution was usually done in triplicate . J. Neurochem., Vol. 40, No . 2. /983 The coated microwells were washed three times with PBSffW, and 30 ~I from each preincubated antibodycompetitor microwell was transferred to a corresponding coated microwell. These were incubated for 18 h at 4°C. They were then washed three times with PBS/TW for removal of unbound antibody, and I 00,000 cpm of 12 5J-Iabeled S . aureus protein A (300- 350 Ci/mmol, about 10-s mol of protein A monomer per L) was added to each well in 30 ~I of PBS/TW plus 2.5 mg/ml gelatin . A preliminary experiment indicated that omitting the gelatin significantly increased the amount of nonspecifically bound protein A. After incubation for 1 h at 37°C, the microwells were washed three times with PBS/TW for removal of unbound protein A . Bound protein A was solubilized with an excess (125 ~I) of 2 mol!L NaOH. After a few minutes (5-10) at room temperature , the NaOH solution was transferred quantitatively to plastic tubes and counted in a gamma counter. Alternatively, when flexible microliter plates were used, the wells were simply cut out and counted directly. Another P0 RIA protocol was used in recent experiments to increase sensitivity . Anti-P0 was used at a final dilution of x 21,333 and was preincubated with competitor (and the usual detergents, etc.) for 6 hat 4°C. Coating and overnight incubation were the same as previously . 100,000 c.p.m. of 1251-labeled protein A (2000-3500 Ci/ mmol, about I0- 9 mol/L) was incubated in each microwell in 30 ~I of PBSffW plus gelatin for 2 h at room temperature, and the bound 12 5J-labeled protein A radioactivity was determined as above. RESULTS Assay methods Anti-P 0 binding to antigen-coated microwells and its dependence on antibody concentration are shown in Fig. I. Maximal binding was reached at an antibody dilution of about x 5 (earlier experiments showed a more abrupt plateau than Fig. 1). Binding was reduced twofold at an antibody dilution of approximately x 100 and fourfold at approximately the antibody dilution ( x 533) that was routinely used for RIA. Binding by normal rabbit serum was negligible (see, for example, Fig. 2). Anti-P0 binding to uncoated wells or to wells coated with BSA or myelin basic protein was Jess than or equal to normal serum binding (not shown). Experiments on the time course of antibody binding to coated microwells (not shown) indicated that, at an antibody dilution of x21,333, binding was complete within about 2 h at 4°C. We used an 18-h incubation at 4°C in the experiments described in this paper, however, partly for convenience. SDS-solubilized peripheral myelin proteins were generally used to compete with the myelin proteincoated microwells for anti-P0 binding. The P 0 content of the myelin protein was measured by densitometry of Fast Green-stained gels. The myelin protein was also compared with purified P 0 in the RIA (see below). Our standard assay protocol gives a working range of about 1 to 100 ng of P 0 (Fig. 2A). To determine the threshold sensitivity, we com- 30 00 S2 " :E ~ FIG. 1. Anti-P0 titration . Various dilutions of antibody were preincubated without competitor (but with detergents) and then incubated in coated microwells for 18 hat 4°C as described in Materials and Methods. After washing to remove unbound antibody, each microwell was incubated with 317,000 cpm of ' 25 1-labeled protein A (about 3 x 10· • moi/L) for 1 hat 3rC and processed as described in Materials and Methods. Prelim inary experiments (not shown) indicated that this amount of protein A was not limiting at the highest antibody concentration used. Each point represents the mean of three determinations; standard deviations averaged 4% of the mean and were all less than 10% of the mean. 20 !d C( ~ w ba: a. 10 I .;, ~ c z ;:) 0 Ill 3 LO~ ANTIBODY 4 DILUTION 5,-------------------------------------------~ f 4 X E 0. ~ 3 <{ c -~ e D.. I 2 "0 c :::> 0 a:l o ---~ 0 0.5 5 10 50 100 NRS P0 Competitor (ng) +3 +2 FIG. 2. Standard competition curve . Anti-P0 was preincubated with various dilutions of peripheral myelin proteins containing the indicated amount of P0 . Samples were subsequently incubated and processed as described in Materials and Methods. (A, above) Raw data showing the cpm of 125 1-labeled protein A bound versus the amount of soluble P0 competitor on a log scale . Each point represents the mean± SO of 12 determinations. NRS, normal rabb it serum (rather than anti-P 0 ) . (8, left) The same data linearized by applying the logit transform to the bound cpm Logit (B) = ln(B/1-8) , where B = (cpm of sample - cpm normal serum) / (cpm of 0 competitor- cpm normal serum). Each po int represents logit (mean cpm); the error bars ind icate logit (mean cpm ± SO). These error bars are not symmetrical because of the mathematics of the logit transform. +1 E 0. " "0 c :::> 0 0 ~ ·;:;, _,0 -1 -2 -3 o·s 5 10 Po Competitor (ng) 50 100 23 542 K. J. FRYXELL ET AL. pared several dilutions of peripheral myelin competitor with wells incubated without competitor. P0 at I.6 or 0.8 ng was significantly different from no competitor, by Student's t test (p < 0.01, n = 12 for each dilution), but P 0 at 0.4 ng was not (p > 0.3, n = I2). A second RIA protocol, with a much higher antibody dilution and a longer preincubation at lower temperature (see Materials and Methods), gave slightly less than twofold extra sensitivity based on 3-4 standard curves from each protocol. Although the standard curves were linear throughout much of their range on plots of specifically bound 1251-labeled protein A versus log competitor, the logit transform (Rod bard et al., I968) provided a somewhat better fit to the data and was used for all data reduction (Fig. 2B) . A desk-top computer (Hewlett-Packard 9820A) was programmed to perform a linear regression fit to the logit-transformed standard curve, plot the results, and print out the estimated Po content of the unknowns . Myelin proteins and purified P 0 were compared as soluble competitors by use of myelin-coated wells and gave parallel straight lines (Fig. 3). Wells coated with purified P 0 gave similar results. Thus none of the myelin components remaining after delipidation interfere with antibody binding to P0 • Anti-P0 binding was strongly inhibited by O. I mg/ml or more SDS. When additional non ionic detergent (Triton X-H)()) and carrier protein (BSA) were added, inhibition by SDS was reduced (Fig. 4) . It seemed de sirable to be able to assay weakly reac- tive samples at the lowest possible dilution and, hence, the highest possible SDS concentration . Therefore, we routinely performed the RIA with O.I mg/ml SDS, I mg/ml Triton X-100, and I mg/ml BSA . Some detergents do not interfere with antigenantibody binding to the same extent as SDS . Preliminary experiments showed that, even at a concentration of 40 mg/ml, sodium deoxycholate interfered only slightly (77% of control c.p.m .) and Triton X-100 not at all (103%). On the other hand , lithium diiodosalicylate interfered strongly (51 % of control c.p.m. at I mg/ml). Therefore, it was of interest to see if high concentrations of nonionic detergents or sodium deoxycholate would extract P0 quantitatively from tissue samples. Rat sciatic nerves (Pel-Freez) were homogenized in 12 volumes of H 2 0, and aliquots of the homogenate were centrifuged after being boiled for I min in 2% of one of the following detergents: SDS , sodium deoxycholate, NP40, Triton X-100, or Tween 20. After P0 RIAs (run separately in SDS, deoxycholate, or Tween 20 with controls for detergent interference) , it was found that SDS extracted two- to fourfold more P0 immunoreactivity than Tween 20, Triton X-100, or sodium deoxycholate, and about 10-fold more than NP40. When rat sciatic nerve homogenates were extracted with several concentrations of SDS and assayed by SDS gel electrophoresis and densitometry , maximal amounts of P 0 were extracted at an SDS/ total protein ratio of 1.5 to 2.0 or more. Maximal •2,---------------------------------------------~ • ', FIG. 3. Comparison of peripheral myelin proteins (e) and purified P0 (+) as soluble competitor in the RIA. Samples were prepared and assayed (in peripheral myelin protein-coated wells) as described in Materials and Methods. The P0 content (abscissa) was measured by densitometry of SDS gels. Data points represent the mean of two separate experiments, each run in triplicate. Lines calculated by the method of least squares are shown for both myelin proteins and purified P0 . Neither slope was significantly different from the theoretical value of -1 (p > 0.1 by Student's t test) . The purified P0 showed only 75% of the expected immunoreactivity, presumably due to the unavoidable exposure to excess SDS during electroelution (see text). An experiment using wells coated with purified P0 gave similar results (not shown). +1 • •• 'E fr "0 c: 5 0 ~ •• -1 ', • • -2~-----r------~---------------r------~~ 5 50 10 Po Competitor (ng) J . Neurochem ., Vol . 40, No . 2, 1983 100 24 543 P 0 RADJOIMMUNOASSA Y basic protein did not change significantly with developmental age (Table 1). In preliminary RIAs of adult mouse sciatic nerve , Po levels were 24 ± 4% of total protein in wild-type nerve and 4 ± 0.4% in trembler nerve (mean ± SEM , n = 2 for each genotype) based on the usual rat myelin standard. These mouse nerve extracts were also as sayed directly by densitometry of SDS gels (see Materials and Methods), which gave 23 ± 1% Po for wild-type and 7 ± 0.4% for trembler. The cross-reaction of our antiserum with mouse P 0 is apparently nearly complete. c z :::l 0 Ill ""'> l\\ 40 ;:::: < ...J ""'a: 20 0 0.025 0.05 0 .1 •·-- -... ..• 0.2 0.4 SDS CONCENTRATION (mgtml) FIG. 4. Detergent effects on anti-P0 binding . Ant i-P 0 was preincubated with the indicated concentration of SDS in the absence of competitor either with (l.) or without (e) 1 mg/ml Tnton X-100 and 1 mg/ml BSA. Other details of incubation and processing were as described in Materials and Methods. Each point represents the mean~ S.D. of a total of 12 determinations from two experiments . The experiments were combined by arbitrarily scaling the mean cpm bound at 25 1-Lgi ml SDS, with Triton X-100 and BSA, to 100. amounts of immunoreactivity, assayed by P 0 RIA, were extracted at a lower SDS/protein ratio (0.4 to 0.8) . These observations suggested a compromise extraction protocol aimed at achieving an SDS/total protein ratio of 0.4 to 0.8. Assay results Assays of central myelin or brain extracts (Table I) ga~e less than 0.1 % of the P0 level in peripheral myehn (Greenfield et al., 1973), providing further evidence that P 0 is confined to the peripheral nervous system (Trapp et al., 1979; Brockes et al., 1980). Liver extracts or purified myelin basic protein showed little or no P 0 immunoreactivity . Schwann cells cultured for more than one week or membrane preparations from these cells gave P0 RIA results similar to the above negative controls (Table 1), confirming that Schwann cells do not synthesize detectable amounts of P 0 under standard culture conditions (Brockes et al., 1980; Mirsky et al., 1980). Po was clearly detectable in Schwann cells cultured for only 12 to 18 h, however, as expected from the fluorescence results of Mirsky et al. (1980). Extracts of sciatic nerves from 1-day-old, 4-dayold , or young adult rats gave P0 levels (as a percent of total protein) of 0.2%, 2%, and 9% , respectively (Table 1). The assay was linear within the range of nerve extract dilutions examined (Fig. 5) . Myelin basic protein RIAs revealed that the ratio of P 0 to DISCUSSION The specificity of our rabbit anti-rat P 0 antiserum has already been extensively characterized by immunofluorescence as well as immunoautoradiography of SDS gels (Brockes et al., 1980; Mirsky et al., 1980). Immunoreactivity on SDS gels was limited to the Po band 2 • In the sciatic nerve, immunoreactivity increased dramatically during development. In the cervical sympathetic trunk, onl y occasional P 0-positive fibers were observed. No immunoreactivity was detected in brain by either technique. Results in this report provide additional confirmation of the specificity of our anti-P0 serum (Table 1). Anti-P0 binding to uncoated wells or to wells coated with BSA or myelin basic protein was negligible. RIAs showed negligible levels of P 0 immunoreactivity in brain, central myelin, liver, and purified myelin basic protein . In preliminary experiments , it proved difficult to purify large amounts of P 0 from the rat peripheral nervous system. We therefore found it more convenient to use purified peripheral myelin as a source of Po for coating the wells and for soluble competitor. The use of impure antigen for these purposes is justified only if the antiserum is monospecific for P0 , but the results in this and the preceding report (Brockes et al., 1980) support this assumption . Further, the use of purified Po rather than myelin proteins does not alter the standard curve significantly (Fig. 3). Initial experiments with 12 SJ:-Iabeled P 0 (specific activity about 3 p.,Ci/p.,g) did not show a significant amount of radioactivity specifically bound by anti-P0 , under any of several conditions. The reason for this is unclear. Solid phase assays do avoid the possibility of iodination damage to the antigen , however, and the use of excess solid phase antigen may increase the signal-to-noise ratio in cases of low antibody titer (see also Lessard et al., 1979). Several solid phase assays gave better results with 2 Another very minor band with a molecular weigh I slightly lower than Po was also labeled. It may be a breakdown product of Po (Eylar and Roomi , 1978: Brockes el al., 1980) or unreduced P, (Cammer et al., 1980). J. Neurochem., Vol. 40. No. 2. /983 25 K. J. FRYXELL ET AL. 544 TABLE 1. P 0 immunoreactivity in extracts of rat tissues and cultured rat Schwann cells Sample Young adult sciatic nerve 4-day postnatal sciatic nerve 1-day postnatal sciatic nerve Schwann cells, cultured 12 to 18 h Schwann cells , cultured 8 days Schwann cells, cultured > 7 days , membrane preparation Brain Central myelin Liver Myelin basic proteins P0 immunoreactivity (p.glmg protein extracted) Po/MBP immunoreactivity (p.glp.g) Number of samples Total protein extracted (mglml) 87 ~ 4 4.7:!: 0.8 4 1.3:!: 0. 1 19:!: 2 4.4:!: 0.2 3 1.5 :!: 0.04 1.8 :!: 0.6 4.7 ~ 0.8 3 1.6:!: 0.1 3.7:!: 0.6 4 1.0 :!: 0.3 <0.6" 3 2.6:!: 0.3 < 0.6° < 0.6" < 0.6° < 0.6" 0 2 3 3 3 0.28:!: 0.05 2.0:!: 0.2 0.48 :!: 0. 10 3.8:!: 0.3 2.5 I Samples were solubilized and assayed as described in Materials and Methods . The results are expressed as means :t SEM for the indicated number of samples. For young adult sciatic nerve , we used one nerve per sample from 2 300-330 g female rats . For 1-day and 4-day sciatic nerves, we pooled nerves from 6-7 animals per sample. Schwann cells were obtained from 3- to 4-day-old rat sciatic nerve as described in Materials and Methods . Total protein was measured by the method of Lowry et al. (1951) , except for myelin basic proteins, which were measured gravimetrically . MBP, myelin basic protein . 0 Mean values were less than 0.6 !J.g/mg , but not necessarily significantly different from 0.6 !J.g/mg. The actual values were: Schwann cells, 0.3 :t 0. 1; Schwann cell membranes, 0.5 :t 0.2; brain , 0.2 :t 0.002 ; central myelin, 0.3 :t 0.01; liver, 0. 1 :t 0.02 . antigen covalently linked to diazo paper or Sepharose beads, noncovalently bound to concanavalin A-conjugated Sepharose beads , or passively coated onto plastic microwells . The passivecoating method was chosen because of its convenience. The extraction of P0 with SDS involves an apparent conflict between solubilization and retention of antigenicity. Loss of antigenicity after extraction with higher amounts of SDS is unlikely to be due to proteolysis, since the amount of P0 observed by SDS-gel electrophoresis and densitometry is gener- 150 FIG. 5. The linearity of P0 RIA values from nerve extracts . Samples were prepared and assayed as described in Materials and Methods. Each point represents the mean of determinations on 3 or 4 samples . Lines calculated by the method of least squares are shown for each set of points. e. young adult rat sciatic nerve; +. 4-day rat sciatic nerve ; A , 1-day rat sciatic nerve. [;, ..s 0 c.. 6 0 Total Protein (pg) J . Neurochem., Vol. 40. No . 2. /983 26 P 0 RADIOIMMUNOASSAY ally increased (to 12-20% of total protein) under these conditions. Direct interference by SDS with the antigen-antibody reaction can be ruled out , since the SDS concentration was standardized within the assay, and in any case additional SDS interference would increase the apparent P 0 immunoreactivity . The remaining possibilities include interference by the large amount of extracted lipids and denaturation of P0 during extraction. Preli minary experiments with delipidation and/or other detergents, however, were not promising. Therefore , we decided to use intermediate levels of SDS for P0 e xtra ction. The amount of P0 immunoreactivity extracted from adult rat sciatic nerves by this method (87 ~-tg/mg ; see Table l) is only 62% of the value reported by Wiggins et al. (1975), who used SDS-gel electrophoresis and densitometry. Our own observations on nerves extracted with high level s of SDS are similar to those of Wiggins et al. (1975). This suggests that, at the moderate levels of SDS optimal for retention of antigenicity, extraction of P0 from rat nerve is incomplete. The extraction of P 0 is quantitatively reproducible, however (RIA standard deviations between samples were typically 10% of the mean for adult mouse and rat nerve). Preliminary experiments with mouse nerve, using our standard extraction protocol (see Materials and Methods), gave apparently complete extraction of P 0 and complete retention of antigenicity (see Results) . This suggests that the difficulty in extracting P 0 may reflect difficulty in adequately homogenizing the larger, tougher rat nerves. In any case, when this assay is used to study a single problem, standardization of the extracted protein concentrations would be desirable. The presence of SDS in the RIA buffer interferes with antibody binding (Fig. 2) . We were able to minimize this interference by adding Triton X-100 and BSA to the RIA incubation buffer. Both the inhibition of antigen-antibody binding by SDS and the reduction of this inhibition by Triton X-100 have been observed previously (Dimitriadis, 1979) and are consistent with other observations (Clarke, 1975; Robinson et al. , 1980). This mixture of ionic and nonionic detergents seems to be quite effective in solubilizing proteins (Dimitriadis, 1979) and presumably is sufficient to keep P0 solubilized . Our standard RIA protocol is capable of detecting 0.8 ng of P0 (20 ng!ml) , given sufficient replicates . Routine assays with only three data points at each competitor dilution , however, typically gave detection limits of about 1.5 ng P0 • Both our standard protocol and a modified protocol that was roughly twofold more sensitive (see Results) could detect P0 clearly in extracts of rat sciatic nerve at one day after birth, when myelination only just begins . Variation of readings between assays was usually less than 10% of the mean if the unknowns could be measured at several dilutions within each assay. 545 The dependence of P 0 levels in rat sciatic nerve on age (Table l) is similar to that reported previously by others (Wiggins et al., 1975 ; Brockes et al., 1980). The ratio of P 0 to myelin basic protein did not change with developmental age (Table 1). This indicates that the synthesis of these proteins may be initiated at the same early stage of myelination, contrary to previous suggestions (Mirsky et al., 1980). P0 was clearly detectable in Schwann cells after 12 to 18 h in culture, but not after 8 days in culture (Table 1). These results support previous reports (Brockes et al., 1980; Mirsky et al., 1980) that rat Schwann cells synthesize myelin proteins only when they are specifically signaled to do so by the appropriate class of axon (Brockes et al., 1981) , although we cannot yet rule out the possibility that Schwann cells stop synthesizing myelin proteins artifactually as a result of exposure to standard cell culture conditions. This RIA should prove to be useful for studying the induction of P0 synthesis in cultured Schwann cells. Acknowledgments: We thank D. McFarlin for the gift of myelin basic protein , and T. Stevens for help with cell culture and discovering that prefreezing nerves selects against fibroblasts . K. J. Fryxell was a National Science Foundation Graduate Fellow during part of this work and is now supported by a National Research Service Award (5 T32 GM07737) from the National Institute of General Medical Sciences . This research was also supported by National Institutes of Health grant NS 14403 and a grant from the Pew Charitable Trust to J . P. Brockes. Some of these results have been incorporated into a review (Brockes eta! ., 1981). REFERENCES Benjamins J. A. and Morell P. (1978) Proteins of myelin and their metabolism. Neurochem. R es. 3, 137-174 . Bolton A . E. (1977) Radioiodination Techniques , pp. 45-48 . Amersham Corp ., Arlington Heights , Illinois . Braun P. E. and Brostoff S. W. (1977) Proteins of myelin, in My elin (Morell P. , ed) pp. 201-231. Plenum Press, New York. Brockes J.P. , Fields K. L., and Raff M. C. (1977) A surface antigenic marker for rat Schwann cells . Natur e 266, 364-366. Brockes J . P ., Fields K . L., and Raff M. C. (1979) Studies on cultured rat Schwann cells I. Establishment of purified populations from cultures of peripheral nerve . Brain Res. 165, 105-118. Brockes J . P., RaffM . C . , Nishiguchi D . J ., and Winter J . (1980) Studies on cultured rat Schwann cells. Ill. Assays for peripheral myelin proteins. J . Neurocyrol . 9, 67-77. Brockes J.P., Fryxell K . J., and Lemke G . E . (1981) Studies on cultured Schwann cells-The induction of myelin synthesis. and the control of their proliferation by a new growth factor. J. Exp . Bioi. 95, 215-230. Brostoff S . W ., Karkhanis Y . D., Carlo D . J., Reuter W ., and Eylar E. H. (1975) Isolation and partial characterization of the major proteins of rabbit sciatic nerve myelin. Brain R es. 86, 449-458 . Cammer W., Sirota S. R. , and Norton W . T . (1980) The effect of reducing agents on the apparent molecular weight of the my- J. Neurochem ., Vol . 40, No . 2, /983 27 546 K. J. FRYXELL ET AL. elin P0 protein and the possible identity of the P0 and "Y" proteins. J. Neurochem. 34, 404-409. Clarke S. (1975) The size and detergent binding of membrane proteins . J. Bioi. Chern . 250, 5459-5469. Cohen S. R. and Guarnieri M. (1976) Immunochemical measurement of myelin basic protein in developing rat brain: An index of myelin synthesis. De1·. Bioi. 49, 294-299. Cohen S. R ., McKhann G. M . , and Guarnieri M. (1975) A radioimmunoassay for myelin basic protein and its use for quantitative measurements. J. Neuro chem. 25, 371-376. Delassalle A., Jacque C., Drouet J., Raoul M., Legrand J . C., and Cesselin F. (1980) Radioimmunoassay of the myelin basic protein in biological fluids, conditions improving sensitivity and specificity. Biochimie 62, 159-165. Dimitriadis G. J. (1979) Effect of detergents on antibody-antigen interaction. Anal. Biochem. 98, 445-451. Everly J. L., Brady R. 0 . , and Quarles R. H. (1973) Evidence that the major protein in rat sciatic nerve myelin is a glycoprotein. J. Neurochem. 21, 329-334. Eylar E. H. and Roomi M. W. (1978) The action of trypsin on central and peripheral nerve myelin, in Myelination and Demyelination (Palo J., ed), pp. 307-328 . Plenum Press, New York. Eylar E. H., Uyemura K. , BrostoffS . W., Kitamura K., Ishaque A., and Greenfield S. (1979) Proposed nomenclature for PNS myelin proteins. Neurochem. Res. 4, 289-293. Greenfield S. , Brostoff S., Eylar E. H., and Morell P . (1973) Protein composition of myelin of the peripheral nervous system. J. Neurochem. 20, 1207-1216. Greenfield S ., Brostoff S . W., and Hogan E. L. (1980) Characterization of the basic proteins from rodent peripheral ner· vous system myelin . J. Neurochem. 34, 453-455 . Groome N. P. (1980) Enzyme-linked immunoadsorbent assays for myelin basic protein and antibodies to myelin basic protein. J. Neurochem. 35, 1409-1417. Hubbard B. D. and Lazarides E. (1979) Copurification of actin and desmin from chicken smooth muscle and their copolymerization in vitro to intermediate filaments . J. Cell Bioi. 80, 166-182. Lessard J. L., Carlton D., Rein D. C., and Akeson R . (1979) A solid-phase assay for antiactin antibody and actin using protein-A . Anal. Biochem. 94, 140-149. J. Neurochem .. Vol . 40. No.2. /983 Lowry 0. H., Rosebrough N . J., FarrA. L., and Randall R. J . (1951) Protein measurement with the Folin phenol reagent. J . Bioi. Chern. 193, 265-275. Milek D. J., Sarvas H . 0 ., Greenfield S., Weise M. J., and Brostoff S. W. (1981) An immunological characterization of the basic proteins of rodent sciatic nerve myelin. Brain Res. 208, 387-396. Mirsky R. , Winter J., Abney E. R., Pruss R. M., Gavrilovic J ., and Raff M. C. (1980) Myelin-specific proteins and glycolipids in rat Schwann cells and oligodendrocytes in culture. J. Cell Bioi. 84, 483 - 494. Neal M. W. and Fiorini J. R. (1973) A rapid method for desalting small volumes of solution. Anal. Biochem. 55, 328-330. Nonon W. T . and Poduslo S. E. (1973) Myelination in rat brain: Method of myelin isolation. J . Neurochem. 21, 749-757 . Robinson J. B. Jr., Strottman J. M., Wick D. G. , and Stellwagen E . (1980) Affinity chromatography in nonionic detergent solutions. Proc. Nat/. Acad. Sci. USA 77, 5847-5851. Rodbard D. , Rayford P. L., Cooper J. A., and Ross G. T . (1968) Statistical quality control of radioimmunoassays. J. C/in. Endocrinol. Metab. 28, 1412-1418. Roomi M. W., Ishaque A., Khan N. R. , and Eylar E . H. (1978) The P0 protein: The major glycoprotein of peripheral nerve myelin. Biochim . Biophys . Acta 536, 112-121. Schmid G., Thomas G., Hempel K . , and Grueninger W. (1974) Radioimmunological determination of myelin basic protein (MBP) and MBP antibodies . Eur. Neural. 12, 173-185. Standring R. and Williams A. F . (1978) Glycoproteins and antigens of membranes prepared from rat thymocytes after lysis by shearing or with the detergent Tween 40. Biochim. Biophys . Acta 508, 85-96. Trapp B. D. , Mcintyre L. J., Quarles R. H., Sternberger N.H ., and Webster H . deF. (1979) Immunocytochemical localization of rat peripheral nervous system myelin proteins. P, protein is not a component of all peripheral nervous system myelin sheaths. Proc. Nat/. Acad. Sci. U.S.A. 76, 3552-3556. Wiggins R . C., Benjamins J. A ., and Morell P. (1975) Appear· ance of myelin proteins in rat sciatic nerve during development. Brain Res. 89, 99-106. Wood J . G . and Dawson R. M. C. (1973) A major myelin glycoprotein of sciatic nerve . J. Neurochem. 21, 717-719 . 28 CHAPTER 'IBREE THE TREMBLER LOCUS OF THE MOUSE: TWO SEMIDOMINANT ALLELES DETERMINE PERIPHERAL MYELINATION WITHOUT A DIRECT EFFECT ON SCHWANN CELL PROLIFERATION 29 ABSTRACT Trembler m1.ce show a pronounced reduction 1.n peripheral myelin, while central myelin and peripheral unmyelinated nerves appear normal. autonomous This myelin deficiency l.S caused Schwann cell defect. Falconer found by in 1951 an that Tr/+ and Tr/Tr mice could not be distinguished by behavior or fertility; more recent studies have used Tr/+ mice. studies at the cellular and molecular levels, be very informative. I Gene dose however, should show here that Tr /Tr mice have a 10- fo ld l ower level of myelin basic protein (in the sciatic nerve at 12 days of age, mice, as measured by radioimmunoassay) and 100-fold lower than +/+ mice. than Tr/+ A similar result was obtained with a second allele, Trj, whose effects are 2-3 fold weaker in These each cas e. two mutations do not complement each other. That is, the s1x possible genotypes can be ordered as +/+ . follo ws: . . > TrJ/+) Tr/+ > TrJ/TrJ > TrJ/Tr > Tr/Tr, allowing a detailed titration of the extent of myelination over a 100-fold range. Although increased in Trembler nerves , 1n Schwann cell and TrJ/TrJ re spec t between TrJ/+ results indicate that Trembler gene for peripheral myelination and Schwann cell proliferation. 19 day s viability. of age, while l.S there is little or no difference this the proliferation is no t nerves. These product is essential directly concerned with Curiously, Trj /Trj mice die at 12Tr/Tr m1.ce have apparently normal 30 IlffRODUCTIOR Vertebrate is composed of two types of rapidly conducting "myelinated" axons and slowly nerve axons : conducting peripheral nerve "unmyelinated" Myelinated axon8. axons are surrounded by a compact, multilamellate myelin sheath, which is formed by a Schwann cell Groups of unmyelinated axons ( SC). are enfolded by a layer of SC cytoplasm. In the mature nerve, any given SC a axon , or sev e r a l developing the 1s associated nerve, unmyelinated myelinated , with either unmyelinated axons, all relationships are an destined axon-SC type. If but single myelinated axon not is both. In the initially of to be it will become separately enfolded by a SC, which then wraps many times around specific proteins the axon and such as Po and P1 (1-7). synthesizes myelinI t is possible to surgically confront SCs from one nerve with axons from another nerve, without ambiguity [i.e., without significant amounts of SC migration, proliferation, nerves--(8-11)]. percentage of or ingrowth of axons from other These el egant experiments have shown that the axons that are s ubsequently myelinated depend only on the source of the axons, and not on the source of the SCs (8-11). That is, any SC can form (and maintain) myelin but wil l do so only if continuously induced by the appropriate sort o f axor . SCs m defect 1n phenotype the Trembler (Tr /+) mouse have a specific genetic this process includes of peripheral myelination. peripheral hypomyelination, The Tr/+ reduced axonal 31 and diameter proliferation (15). super1or cervical in myelin the experiments velocity, conduction and sc increased Normally unmyelinated nerves, such as the trunk, appear completely central nervous system similar to those discussed normal , as (15) . Surgical above have does shown that the my elin deficiency in Tr/+ nerve is caused by a defect SC and not 1..n the ax on (12-13). the supported by tissue cultur e in This conclusion 1s also experiment s , in which SCs and neurons of diff erent genotypes are confronted with each other in the ( 14), as well as by studies of absence of fibroblasts Trembler/wild-type chime ri c mice (15). Although normally unmyelinated nerves 1n Tr/+ m1ce appear identical to their wild-type counterparts, if one confronts SCs from these myelinated (16) . Tr/+ nerves n e rv es ), with the competent comple t e Tr/+ axons (from phenotype normally 1s produced Since these abnormalities are uncovered only if SCs are challenged to viability o f appar ently myelinate Tr SCs competent axons, is ess entially normal, concerned 1n (15) . It is n ot known a t acts , although the by some way with and s1nce the the Tr defect myelination is itself what stage of myelination the Tr gene histological evidence indicates that Tr /+ SCs do wrap competent axons individually and are then blocked, apparently at an earlier stage known if the product is essential for myelination . or not, or, for example, myelination mutants what the phenotype (15) . of a Nor than other known SC autonomous u homozygous it null mutation at Tr gene the Tr 32 locus would be. Gene dose experiments should help to answer these but questions, available to the only Tr /+ distinguish criteria from Tr /Tr that are readily m1ce are breeding experiments (which are hampered by the low reproductive success of Tr males In (17)) . preliminary breeding experiments, found that the Trj/Trj genotype is actually lethal . a c l e ar-cut gene dose by such more closely , effect, I I Encouraged studied this problem and f o und that Trj/Trj mice do live to at least 12 days o f age and can be identified a t this age by gross nerve a ppearance. the dose Similar criteria of nerve appearan ce also apply to The s e allele. Tr e x periments, res ult s reported allowed me to do de t ailed gene below, which suggest function of the Tr+ gene product is essential that the for myelination and is not directly concerned with SC proliferation (an early step in the axon-SC interac ti on ). MATERIALS ARD METHODS Genetics. outbred Th e stock of Tr m1ce mutation arose (17 ) wa s an d partially inbred, triple mutant Falconer, communicati on). this pers onal (Ins t itute stock Edinburgh, u. K.) ' and of spontaneously later stock (Tr , Animal maintained 1n an 1n a D. S. maintained Re, and nu; I obtained Tr /+ mice from Genetics, them by University crosses C57BL/6J strain as described 1n the legend to Table 3. to of the The Trj mutation arose spontaneously 1n the DBA/2J strain and was then repeated ly backcrossed to the C57BL/6J strain (18,19). I 33 obtained Re Trj I++ m1ce (back-crossed to C57BLI6J for a total of 9 generations) and control C57BLI6J mice, Laboratory Harbor, (Bar Maine), and from the Jackson maintained them by interbreeding as described in the text. Myelin basic protein (MBP) radioimmunoassay (RIA) . MBP from rat brain (a generous gift from D. McFarlin) was iodinated with chloramine T to US1ng buffers the a specific described 50 ~Cilug, activity of about by Cohen et al. (20), and then desal ted b y t wo centrifugation steps (21) through Sephadex G-25 medium beads (50-150 ~m . Sigma) . NaDodS0 4 I 100 mM LiCl in 1 mg/ml for 4 min. antibody Mouse nerves were homogenized I 0.1 mglml NaN 3 , and boiled Diluted tissue homogenates were pre incubated with (rabbit anti- rat MBP di lu ted 0.1 m1 of the f oll owing buffer: calf thymus histone (Sigma) 1: 1000) for 6-12 hr in I 2.5 mg/ml 0.2 M Tris pH 7.3 I 0.05% Tween 20 (wlv) I 1 mglml All wells within an assay contained the same amount of NaDodS0 ; this and all subsequent steps were done at 4°C. Two 4 thousand c.p.m. of 125 r-labeled MBP was added to each tube and incubated for an addit ional 12-24 hr. Ten ~1 of 5% (vlv) fixed Staphylococcus aureus bacteria (Bethesda Research Labs) was added tube, vort exed brief ly , incubated for 1 hr, to each layere d over 4 ml histones), of 24% sucro se ( i n incubation buffer without cen tri f uged s o l u tion aspirated, solution) counted labeled MBP was at and 1200 g the for 30 min, pellet in a gamma counter . thus left above (plus the top 1 ml of remaining sucrose Residual unbound 125r- the sucrose solution, outside 34 the counting window of the gamma counter . The standard curve used as known normal amounts rabbit labeled MBP. logit of unlabeled rat MBP competitor, and binding of 12 5r- serum to measure non-specific The results were processed using an unweighted transform, as described (6). The competition of mouse nerve e x tr a c ts against 1251- labeled rat MBP was linear over t he range f r om 0 . 3 t o 7 ng MBP/ml; all nerve extract data points were confined to the range from 0.3 to 3.5 ng/ml. In v i v o pr olifera t ion assay. using st erile technique (22) Eagle's medium ethanesulfonic buffered acid and placed in 0.5 ml of modified wi t h (HEPES) Sciatic nerves were removed N-2-hydroxyethylpiperazine-N'-2- in one we ll (per animal) of a 24 well plate (Linbro). The nerves were briefly examined under a dissecting microscope, 1n order or fat to remove fragments of muscle and to record the nerve appearance (see Results). The medium was then changed to 1 ml of Dulbecco' s modified Eagle ' s med ium with 10% fetal calf serum and 2 ~Ci of 125 r-deoxyuridine (1 2 5rUDR; New Engl and Nuclear). The nerves were incubated for 6 hr at 37°C in a humidified atmosphere of 10% co 2 . The nerves were t hen rinsed three t imes with phosphate-buffered saline and stored 1n l iquid N2 .. h omoge ni zed vigorously , ice-cold 20 mM ATP. measure total homogen ate DNA were After thawing, the nerves were with a motor-driven Teflon pestle, 1n Aliquots of this homogenate were used to (see added below). To an volume equal the remainder of of the ice-cold 20% trichloroacetic acid (w/v), and 10 ~1 of 10 mg/ml bovine serum 35 dH 2 o. in albumin 2.5 min x 15,000 g at 4 o C, and the was centrifuged pellet counted 1n a for gamma The precipitable radioactivity was reduced 90% when counter. 1 mM solution This hydroxyurea, red uct ase [and a inhibitor specific hence of DNA synthesis of ribonucleotide (23)], was present throughout the 6 hr incubation. Total DNA Hoechst 33258 DNA assay. (Calbiochem) levels were measured with in a high salt buffer as described (24), with controls for the background fluorescence due to the ATP homogen i:&ntion solution. RESULTS TrJ/TrJ m1ce die before 19 days of age. produced in conjuction with the linked genetic marker rex (Re) , which produces a wavy coat (25), in the following breeding scheme: ( i) Re TrY++ x Re TrJ /++ + Re TrJ /?? (ii) Re Tr J /?? x ++/++ + Re Trj/++ and??/++ Re TrJ /++ males and females were mated to each other in cross -(i) , a nd breeding the resulting experiments phenotypes*. progeny only if were they selected showed both for further Re and The exact genotype of these Re Trj /?? progeny was deduced by mating them to wild-type mice in cross ( ii). both Re and Trj are (at progeny of the Since l eas t: i ncomplete 1y) d ominant , ~ TrJ parents that are homozygous fo r 100% Trj eitber mutant gene will yie ld cor r esponding phenotype; heterozygotes, 36 Of 61 50%. Re Trj /?? parents Trj/Tr j, although 22 Trj/Trj scored 1n this way, none were were expected. This difference is highly significant (Table 1). It is also suprising, since Tr/Tr m1ce (17) . are viable and fertile In control matings, Re +/Re + C57BL/6J mice showed apparently normal viability and fertility Thus (not shown), consistent with previous reports (25). the phenotype of Trj /Trj mice prevented them in some way from completing t he above breeding experiment. One would like a stricter ac count i ng, however , of exactly what happened to the TrJ/TrJ mice--whether t he y died, were sterile, or had an unrecognized phenotype that cause d them t o be not selected for use in cross ( ii). Several lines of ev i d ence indicate that verified TrJ/TrJ adults is not due to sterility. the lack of One argument concern s the sex di stribution of the Re TrJ /?? mice that failed to produce scorable progeny. males, but not success ( 17), female s , as found Tr is an autosomal locus. have markedly reduced Tr/+ reproductive in other peripheral neuropathies (38). Thus we would expect many Re TrJ /?? male parents not to produce in scorab l e progeny ·: ested , however s- progeny . Of apparently normal nerves from e x amined an y case . Of only three fail ed these one three lit ters of the Secondly, female 51 Re TrJ /? ? - -- females to produce scorable parents , t wo produced Sciatic t hat were cannibalized. latter adult Re histologic ally and (see be l ow). the showed a Trj /?? typical females Trj /+ were phenotype a more direct approach is to examine 37 the phenotypic ratios of all of the progeny of cross ( i). If Trj /Trj mice surv1ve, then one would expect the classic 3:1 ---ratio of Rex mice vs. mice with wild-type fur, while if Trj /Trj die, one would expect a 2:1 because of crossing over) . ratio (actually 2.048 : 0.952 This experiment is not difficult if the m1ce are s cored at the age of weaning ( 19 days or more). The observed agreement rat io '"ith at this "2: 1 11 a age rat io (313:160) and rules is in reasonable out a 3:1 ratio (P <<O.OOl by the chi 2- t e st) . Si mi l arly, an unexpected phenotype of Trj /Trj m1ce can be ruled out. If these m1.ce a 3:1 survived but other assumptions), together, these expected whi ch results can non-Rex would be If the TrJ /TrJ mice were scored as not Rex, then a be to non- expected. wou ld Rex as then ratio of scored Trembler , "2:2 11 ratio were (the exact value also be ruled depending ou t . on Taken show that Trj /Trj m1.ce die before 19 days of age. The results 1.n Table 1 al so suggest that the Re TrJ /Re + genotype may be p artially lethal , and thus that the Trj allele is assoc iat ed with one o r more recess1.ve lethal mutations lying be t ween. Re and Tr on the genetic map. border of statistical significance, This effect 1.s at the however, and cannot be regarded as well established. TrJ /TrJ mice survive to at least 12 days of age. Among the 12 to 16- d:ty-old progeny of cross ( i), there is a distinct subset of young with a severe, rigid spastic paralysis, quite 38 distinct from Trj /+ behavior (which can barely be recognized at this age). phenotype These (25) spastic while young do Re/Re expected if the spastic phenotype represents Trj /Trj mice. At their body weight averages 25% less than that of their littermates. m1ce usually the as Re/Re littermates have not, 12 days of age, their usually do appear Whole sciatic nerves from these spastic translucent viewed b y diffuse tr ansmitted nerves are completely Further experiments opaque showed and light light, and that when while their littermates' grey the yellow-brown to (Figure 1). black "clear nerve" phenotype occurred with the frequency expected of TrJ /TrJ mice, among the 12- to 13-day-old progeny of many different (Table 2) . This correspondence between Trj /Trj genotype, indicates and the both "clear that nerve" kinds of crosses 1s a 1:1 phenotype and the there that Trj /Trj mice generally live to at least 12 days of age. All Tr genotypes may be uniquely identified at 12 days of age by b ehavior cross (i), or other~ fa l l into and nerve the appearance. progeny t hree of Tr/+ C57BL/6J hybrids non-overlapping sc i atic nerve levels of MBP (Table 3). progeny crossed classes 1n to of each terms of In all cases tested 1n this way (11/11 mice for TrJ; 27/27 for Tr), the class to which each mouse belonged was correc tly predicted, based on behavior and nerve appearance . Trj /+ sciatic nerve s appear light to medium grey at 12 days of age, while +/+ nerves appear dark grey to black, when 39 viewed by diffuse transmitted light. Although this is a subtle d i fference, the it was reinforced by Re phenotype and by TrJ /+ mice show a steady ~ncrease in tension of the behav ior. leg muscles after decapitation , while +/+ mice show only rapid and chaotic kicking. The increase in leg tension is much more pronounced in Tr J /TrJ mice, whose nerve appearance also makes them to easy days), diffe r ences (Figure 1). ~n younger ages (3 or 6 are more obvious, bu t behavioral d i fferences are not apparent. Deduced genotypes were verified ages these nerve At appearance at the i dentify by histology Tr/+ sciatic ~n 6/6 cases (not appear much shown). the By method, s ame nerves lighter than TrJ/+ nerves, consistent with their lower myelin content (Tab l e 3) . Tr /+ sci a t ic nerves appear light grey to yellow and are only slightly opaque at 12 days of age. Tr/+ nerves are easy to d i stinguish alone. Tr/1:£. nerves are more transparent than Tr/+, and do not have any grey c o loration by from +/+ nerves by appearance t ransmitted light. This ~s a subtle distinction, and it is not accompanied by any behavioral Wi th pr a c ti c e , it c an be reliably scored, however differences. (see ab.ov;e). "f wo arguments between TrJ background. and suggest that Tr (Table 3) the difference ~s not an in MBP content artifact of genetic First, +/+ mice from t h ese two genetic backgro unds have similar MBP levels (Table 3). ' Secondly, the _IE.. allele was measured on a hybrid genetic background that should vigorous than C57BL/6J, and yet showed lower MBP levels. be more 40 The allelic difference in MBP content does not explain the death of Trj /Trj m1ce. Trj /Trj mice have ~ MBP than Tr/Tr, but weaning, Trj /Trj die before while Tr /Tr mice survive at least to the age of weaning (in the Tr /+ x Tr /+ cross described in 106/137 Table 3, surviving progeny had the Tr phenotype). Thus the death of Trj /Trj mice is not simply due to a lack of It peripheral myelin. is also not due myelin; cerebellar MBP levels normal (not survival shown). to a lack of central in Trj /Trj mice are essentially The possibility that between Trj /Trj and Tr/ Tr are the differences due to in the different genetic backgrounds has not yet been ruled out. Schwann cell proliferation Trembler p ro li feration. which extends myelination Tr/+ mice show abnormal adulthood (40). into deficit 1s caused by SC If the abnormal SC proliferation, then Trj /Trj mice should have a much higher rate of SC proliferation than Trj/+ mice do. The identification of Tr genotype by nerve appearance allows me to use much younger mice than previous studies that relied on behavior to identify genotypes times, (13) . to Therefore, facilitate prolifer a tion rat e s I the chose relatively short labeling measurement with age. of the change 1n Preliminary experiments showed that i n t act nerve s could be labeled with 1 2 5ruDR to an easily level in vitro. The rate of incorporation declined the incubation 6 hr reflects the (ca. 10,000 detect ab le period cpm per animal) (Figure 2A), after 6 hr somewhat during which presumably limitations of our culture system in maintaining 41 intact Nevertheless, nerves. this assay provides a simple, semi quantitative measurement of cellular proliferation. The rate of proliferation decreased steeply from 3 to 12 days of age, the high variation as previously reported (41). v ar iability J.n between developmental litters age) did definite conclusion. In differ e d significantly, particularly older At 3 days of age, (presumably justify not Trj/+ litters, in and due to any more +/+ mJ.ce 1 2 5IUDR incorporation at 6 day s of age and DNA levels at 12 days of age (Figure 2B, 2C) . Trj /+ and Tr J /TrJ were not significantly different m either respect. DISCUSSION The genetic was Trj mutation was named mapping previously locus. Among and as an allele of Tr based on histopat hologi c no basis f or the progeny features testing of but there complementation at this Trj /+ x Tr /+ (19), mice, I found a distinct subset of progeny having the "clear" nerve appearance and correspondingly low MBP levels (Table 3). do not complemen t mut ations in between these the each same two other gene, and are although semidominant Thus TrJ and Tr confirmed as being lack of complementation alleles could have other interpretations. The Trj mutation has several nnusual advantages for study in cell culture. Many other cell autonomous, developmental mutants have been identified as such partly because they cause 42 the degeneration of the cell [as E in or rd, type 1n which they are expressed see (26)] understood molecular deficit or [as cause a relatively in shi or twi, well- see (15,27)] . In contrast , Trj/Trj SCs are viable, and the molecular basis of their lesion 1s unknown, so ce 11 culture experiments should be inte resting. Powerful puri ficati on culture an d maintenance (22,28,39). nerves may be methods The iden tified of dissociated present as study SCs shows the for available are 1n tissue that Trj /Trj such before being dissociated or homogeniz e d, so that "genetically pure" cultures of dissociated TrJ /TrJ SCs could be grown. Actually, the yield of viable SCs from Trj/Trj nerves is higher that from wild-type nerves (K. J . & J. P. Brockes, unpublished). Fryxell The 100-fold deficit of myelination 1n this stud y, between the Tr/Tr m1ce, st rengthens the already mutation (on chromosome 11, Tr peripheral myelin) affec t s 10 the .1.£. mutation and onl y centr al myelin). Of the interesting shown parallels affects only (on the X chromosome, f ive best-characterized myel in-s pecific genetic loci in t he mouse (IS), only these two r espec t the border b etween the central systems, and it and peripheral nervous is now clear that mutations at these two loci also have by far the most potent, and the earliest, effects on myelination. All five of than the glialt cell rather recent work suggest s proliferation of t hat these ~ axon 1s oligorlendr ocytes loci are expressed 1n the (15,27,29,30). Moreover, associated increased (the with glial cell that 43 generates normal numbers in the central nervous of oligodendrocytes ( 31), system) and roughly somewhat similar Both i£ alleles lNere described as being recessive Tr. and myelin both Tr alleles clear, however, as being dominant It (17,19). to (26), l.S now that Tr l.S semidominant at the cellular level, while ..il:. l.S sex-linked and cell autonomous (29). Thus only one cop y of the ..lE. gene dominance at the l.S active l.n any given cell, so its level l.S not known and may not cellular r e fl ect any fundamental di ff erence between J.E. and Tr. case, these considerations glial gen es contr olling tend t he to early In any suggest that many of the stages of myelination are central/ peripheral specific, while the glial genes involved 1.n the later s t eps are premature, s1.nce characterized and phenotypes. myelin Such mo st of these have two mutant Moreover, d ysmyelinating myelin (32). not . mutation 1.n loci rat l. s may are alleles only the the speculation not fairly be well with similar comparably potent specific central to Given the fea s ibility of the molecular cloning of protein structura l ge ne s, clarified in the next few years. these 1.ssues should be 44 ACKNOWLEDGKMERTS I thank J. P. Brockes for advice and support, U. McFarlin for myelin basic protein, and S. for help with mouse breeding records. R. Maxwell G. F. Richards for help with RIAs, This work was supported by grants from the NIH to J. P. Brockes. was supported Institute of by a pr.edoctoral General Medical Graduate Fellowship . I traineeship from the National Sciences, and an Evelyn Sharp 45 FOOTNOTES *since + TrJ /+ Trj progeny will arise only by double cross1.ng over, they should be rare. Given 22% recombination between Re and Trj (33,34), 95% of the Trj/Trj mice will also have the Rex phenotype. Thus, the search for Trj /Trj progeny was rest ric ted to Re progeny of convenience. t Prec isely Wh ich k ind of glial cell is directly affected by the i.E_ muta tion is not yet clear ( 30,35,36). 46 REFERENCES 1. Webster, Dyck, H. de P. J.' F. (1975) Thomas, in Peripheral Neuropathy, K. P. E. H. (W. & Morell, P. 0975) Lambert, & eds. B. Saund e rs, Philadelphia), Vol. 1, pp. 37-61. 2. Wiggins, T. c. ' Benjamins, A. J. Brain Res. 89, 99-106. 3. Bro ckes, J. P . , Raff, M. C., Nichiguchi, D. J. & Winter , J. (1 9 80 ) ]_:_ !tle uroc y t ol . 9, 67-77 . 4. Mirsky, R., Gav r ilov i c ~ "W i n ter , .J. J., Ra ff, & Abn ey, M. E. R. , Pruss, R. M. , 0980) J . Cell Biol. 84, C. 483-494. 5. Brockes , ~ Biol. 6. Fryxell, K. J. & Lemke, G. E. (1981) J . 95, 215-230. Fryxell, K. ~ 7. P., J. J., Balzer, D. R. Jr. & Brockes, J. P. (1983) Neurochem. 40, 538-546. Politis, M. J., Sternberger, N., Ederle, K. & Spencer, P . S. (19 82) ~ Neurosci. 2, 1252-1266. 8. Weinberg, H. J. & Spenc er, P. S. (1976) Brain Res. 113, 363-378 . 9. Aguayo, A. J., Epps, J., Charron, L. & Bray, G. M. (1976) Brain Res . 104 , 1-20 . 10. Aguayo, A. J., Charron , L. & Bray, G. M. 0976) J. Neurocytol. 5 , 565-573. 11. Spencer, P. S. & Weinberg, H. J. 0978) in The Physiology and Pathobiology of Axons, ed . Waxman, S. G. (Raven Press, New York, N. Y.) pp. 389-405. 47 12. Aguayo, A. J., Attiwell, M., Trecarten, J., Perkins, S. & Bray, G. M. (1977) Nature (London) 265, 73-75. 13. Aguayo, A. Bray, G. M. J., & Perkins, C. (1979) Ann . S. !.:_ !.:_ Acad. Sci. 317, 512-533. 14 . Bunge, R. P., Bunge, M. B., Okada, E. & Cornbrooks, C. J. (1980) e d. in Neurological Baumann , N. Mutations Affecting Myelination, (Elsevier /North-Holland Biomedical Press, Amsterdam) pp. 433-446. 15. Bray, G. M.. & Aguayo, A. J. (1981) Ann . S., Aguayo, A. J. & Bray, M. (1981) ~ M., Rasmin s ky , Rev. Neur osc i . 4, 127-162 . 16 . Perkins, C. G. Neurol . 71, 515- 526. 17. Falconer, D. S. (1951) J. Genetics 50, 192-201. 18. Heiniger , H. - J. & Dore y , J. J., eds. (1980) Handbook on Genetically Standardized JAX Mice (The Jackson Laboratory, Bar Harbor, Maine). 19. Sidman, R. L., Cowen, J. S. & Eicher, E . (1979) Ann. M. !.:_ !.:_ Acad. Sci . 317, 497-505. 20. Cohn, S. R., McKhann, G. M. & Guarnieri, M. (1975) J. J. R. ( 1973) Anal. Biochem. 55, Neuroc hem. 25, 371-376. 21. Neall M. W•. & Fiorini, 328-330 . 22. Brockes, J. P., Fields, K. L. & Raff, M. C. (1979) Brain Res . 165, 105-118. 23. Castellot, J. J. Jr., Miller, M. R. & Pardee, A. B. (1978) Proc. Natl. Acad. Sci. USA 75, 351-355. 48 24. Labarca, C. Paigen, K. & (1980) Anal. Biochem. 102, 344- 352. 25. Carter, T. C. (1951) ~Genetics 50, 268-276. 26 . Green, M. C., ed. (1981) Genetic Variants and Strains of the Laboratory Mouse (Gustav Fischer Verlag, Stuttgart). 27. Beraducci, A., Peterson, A., Aguayo, A. & Marler, J. (19 82) Neurol. 32, A84. 28. Brockes, J . P., Lemke, G. E. & Balzer, D. R. Jr. (1980) J. Biol. Chem . 255, 8374-8377. 29. Skoff, . R. P. & Montgomery, I. N. (1981) Brain Res. 212, 175-181. 30. Wolf, M. K., Schwing, G. B., Adcock, L. H. & Billings-Gagliard i, S. (1981) Brain Res. 206, 193-197. 31. Skoff, R. P. (1982) Brain Res. 248, 19-31. 32. Dentinger, M. P., Barron, K. D. & Csiza, C. K. (1982) J. Neurocytol . 11, 671-691. 33. Fal coner, D. S. & Sobey, W. R. (1953) J. Hered. 44, 159160. 34. Flanagan, S. P. (1966) Genet. Res. 8, 295-309. 35. Skoff, R. P. (1976) Nature (London) 264, 560-562. 36. Hertz, L., Chaban, G. & Hertz, E. 0980) Brain Res. 181, 48 2- 487. 37. Low·-y, 0. H., Rosebrough, N. .J., Farr, A. L. & Randall, R. J . (1 951) J. Bi ol. Chem. 193, 265-275. 38. Sout hard, J. L., Wolfe, H. G. Nature (London) 208, 1126-1127. & Russell, E. S. (1965) 49 39. Wood, P. M. & Bunge, R. P. (1975) S., Mizuno, K., Aguayo, Nature (London) 256, 662-664. 40. Perkins, A. & Bray, Neurol. 27, 377. 41. Asbury, A. K. (1967) J. Cell Biol. 34, 735-743. G. (1977) 50 Figure 1. old TrJ/TrJ Comparison of whole sciatic nerve appearance from a 12-daymouse (left) diffuse transmitted light. and +/+ littermate (right), as viewed by 51 100 l.IID 52 Figure 2. Specific DNA synthesis rates and total DNA levels in sciatic nerves from C57BL/6J mice of various ages and genotypes. 125 ruDR incor- poration into DNA, and total DNA content, were measured as described 1n Materials and Methods. described under The genotypes of the mice were Results . Each point represents individual determinations on 3-7 mice, with a the identified as mean ±SEM of few exceptions as noted below. (A) In vitro were course of 125 ruDR incorporation . time incubated 1n 1 2 5rUDR-containing medium, Wild-type nerves with or without hydroxyurea, for v arying times. Two litters were used; one litter was 6 days days o ld and t he other was 8 scaling the maximal The old. data were combined by 12 5IUDR inco rpor ation with i n each litter to 100%. Each point r e pr e sent s 2-3 m1ce . A., no added hydroxyurea ; e, 1 mM hydroxyurea . (B) Two litters of each age Total DNA content as a function of age . •, Trj /+; • , +/+; (fr om cross (i), see Result s ) were assayed. •• Trj/Tr j . ---(C) Specific DNA s ynthes is r a t es as a function of age. were the mice , 12 5 I UDR incorporation could same as in (B) above, except that only be measured incorporation was significant at 3 days of age. i nto this limited amount for in 2 mice, while Variation between litters in 125 runR total DNA was measured in 3 m1ce. bias The m1ce used of data, the To avoid introducing a 3-day-old litters were weighed equally (that is, mean and SEM were calculated between litters rather than between individuals). • , +/+; •, TrJ/+; e,Trj/Trj. 53 A B a: 100 gao ~--;so ... 0 ~6 0 40 ~ ·-c. c; o2o -u G>c ...._ .. 4 ~2 c 3 6 9 12 age, days 2 4 6 incubation,· hr c 0)~10 ::::J.O ... ,.. 8 ~ >< 6 ... E~ 4 frc 2 3 6 9 12 age, days 54 Table Deduced 1. genotypes of Re Trj /?? m1ce and comparison with theoretical expectations. Observed Expected if all if Tr j /Trj if Trj /Trj and survive die 1/2 of Re Trj /Re + die -Re-Trj/++ - 28.5 44.8 50.0 55 Re TrJ / Re Trj 14 . 2 0.0 0.0 0 Re TrJ /Re + 8.0 7.0 3 . . - - - -- 12.6 Re TrJ /+ TrJ 8.0 0 .0 0.0 0 ~ +/+ TrJ 2. 3 3.6 4.0 3 <<0.001 <0 . 05 NS P (chi 2 test) Re Trj/?? C57BL/6J m1ce were produced from cross (i) and their genotype deduced fr om cross ( ii ) Seventy-four Re as described under Results. Trj /?? mice were tested, of which 61 produced scorable progeny ( 13/23 male parents and 4-8/ 51 female parents). The deduced genotypes of the 61 Re the Tr j/?? parents are listed under he ading "Observed . " The compar able numbe r s of animals lis ted under "Expected" were calculated by assuming 22% rec ombination between Re and TrJ (33,34). have approximately the same genetic 1.1ap location. Trj and Tr do For example, using the appropriate subset of the above bLeeding data,* I found a recombination frequency of 203/1008, or 20.1%. The mean number of progeny scored 55 per~ TrJ/?? parent was 18 (range 2-46). The data from each Re Trj/?? --- parent was compared individually to the next most likely genotype, by the chi 2 test. 1 case, 0.01 cases are parents In 55 cases, P < 0.001; in 2 cases, 0.001 < P < 0.05; and in 3 cases, P > 0.05. ~n included with marginal offspring. for emphasize reproductive all that success did of produce the Re TrJ /?? -- non-Trembler measuring NS, not significant. recombination Re Trj/?? parents for which: P to The latter marginal Omission of these cases, however, would not affect any of the other conclusions. *Data order < P < 0.01; ~n frequency were limited to those (1) the complete genotype was known at the < 0.001 level of significance, (2) recombinant progeny could have been identified (i.e., parental genotype (3) at least 5 progeny were scored. of Re Trj/++ or Re +/+ Trj), and 56 Table 2. Observed frequency of the "clear nerve" phenotype among the progeny of various crosses. parental genotypes % having number of clear nerves progeny scored Re TrJ/++ X Re TrJ /++ 28% 79 + Trj /++ X + Trj /++ 24% 51 ++/++ X Re TrJ/++* --- 0% 31 Re +/++ X Re +/++ t 0% 34 Sciatic nerves were dissected from 12- to 13-day-old C57BL/6J m1.ce, and viewed with a x3 objective, using diffuse transmitted illumination. The nerves as were scored as having a "clear" or "opaque" appearance described under Results. *one of the parents in t his group was actually Re TrJ/Re +. t Some of t:he parents in this category were actually Re +/Re +. 57 Table Myelin 3. basic protein immunoreactivity ~n sciatic nerve extracts from 12- to 13-day-old mice. gene tic MBP (~g/mg protein genotype background mean ± SEM) +I+ C57BL/6J 16.3 ± 0.9 100% Trj/ + C57BL/6J 4.9 ± 0.2 30% Trj/Trj C57 BL/6J 0.33 ± 0 . 06 2% +I+ CE x CE 16.9 ± 0.9 100% Tr/+ CE xCE 1.65 ± 0.04 10% Tr/Tr CE X CE 0.10 ± 0.02 1% TrJ/Tr CE 0 . 11 ± 0.02 1% The genotypes of their nerves Methods. m~ce assayed * a cros s wild-type were identified as described under Results, for MBP content as described 3-5 an i mals were assayed in each category. progeny o f colony . C57BL/6J X % of ~n Materials and and CE denotes the F1 between C57BL/6J mice and mice from the Edinburgh The differe nt geneti c backgrounds did not affect the MBP levels i n wi ld-type mice (P > 0.9, Student's t-t est). *The r e sults are stated ~n terms of the ~g of rat MBP required to give equivalent competition. Total protein was measured essentially by the method of Lowry et al. (37). 58 CHAPTER FOUR DISCONTINUOUS MYELINATION IN THE TREMBLER MOUSE: DOSE DEPENDENCE SUGGESTS THE EXISTENCE OF A THRESHOLD IN PERIPHERAL MYELINATION GENE 59 ABSTRACT The dose , dependence demonstrated of myelination in Chapter 3, 1n Trembler m1ce allows one on gene to do a detailed titration of the extent of myelination over a 100-fold range. This range has been studied by s t aining for t he myelin protein and mo un t s whol e indirect immunofluorescent Po 1n frozen sections of nerve of cornea, as well as frozen sec tion s , the Schwann by m1croscopy. _ In ( SCs) of Trj /Trj nerve show a remarkable heterog eneity 1n staining: a f e w SCs are brightl y stained, very faintly, sections while most of the SCs are only but significantly, stained. suggest t hat these brightly generally only one SC 1n length. Longitudinal frozen stained regions Thus, the heterogeneity 1n content is not due to diffe rences between axons. micrograph s show that, myelinated genotype to affect both ave rage as the one progresses sheath Electron from least myelinated, myelin are A similar result was obtained from whole mounts of Tr/+ cornea. Po cells electron the the most first steps cro s s-sectional area and myel in sheath number to a similar extent, while the last steps affect primar il y myelin sheath e x plained by the sheath s . SC numbers first 1n this genetic series. f t ep diffi c u l ty 1n n umber. remain This recognizing essentially The Trembler SCs respond to an axonal not very easily thin myelin constant last step, genotype, had no recognizable compact myelin . all 1s after the the Tr /Tr I conclude that signal to initiate 60 myelination by synthesizing small amounts of P0 • A minority of SCs 1n Tr/+ or Trj/Trj mice express much larger amounts of P0 , which is assembled into compact myelin. 61 INTRODUCTION As decribed more phenotype includes Schwann cell of the se not ( SC) other symptoms. fully in the previous chapter, only hypomyelination, proliferation, reduced but the Tr/+ also excess axonal diameter, and A detailed study of the gene dose dependence v a r ious aspect s of the Tr phenotype may insights into the f un c t ion of the Tr+ gene product. provide Moreover, the ability to reduce the amount of peripheral myelin over five d i stinc t steps may reveal other aspects of myelination , apart from the function of the Tr+ gene product itself. I have chosen this problem. sensitive P0 , regardless locali zin g myel in- s pecific of It 1s particular axon a follo wed over are few whether almost or not they impossible, over long di stances Immunofluor escent reveals complementary methods to approach Immun ofluores cence on frozen sections provides a way of myel in. several staining myelinated r el a tively much t hicker sec.t ior:s, however, of lon g assembled however, mounts axons d is t ance s. (c a. 150 lJID) are such to as into follow a in longitudinal sections. whole sensory proteins than of ( 1), These the which cornea can be preparations conventional frozen and have a correspondingly higher level of background fluorescence. Finally, electron microscopy provides a clear view of morphological details, such as the frequency of compact myelin, wrapped by SCs . and the extent to which axons are individually 62 MATERIALS AHD METHODS Immunofluorescence on S-6 ~ frozen sections was performed essentially as previously described (2). were fixed 10 min, with ice-cold neutralized, 9S% with anti-P 0 at a dilution of x40) washed, incubated goat anti-rabbit with I S% acetic primary antibody ethanol incubated Briefly, the sections acid for (rabbit for 3S min at room temperature, secondary antibody [rhodamine-labeled IgG, F( ab') 2 fragment (Cappel), adsorbed with bovine liver powder and used at a di 1 ut ion of x40] for 35 min at 20°C , washed, mounted, and viewed with Zeiss epifluorescence In som:e cases, optics. the nerves were also prefixed (before sectioning) with 0.1% para form ald ehyde Dulbecco 1 s phosphate- 10 buffered saline (PBS) without calcium or magnesium (Gibco) for 30 min at 20°C . Corneas for whole mount staining were removed innnediately sacrifice with fine Contaminating tissues (lens, watchmaker's forceps. The corneas were fixed for 30 min l.n ethanol I 5% acetic acid, and neutralized by MEM overnight at 4°C , antibod y (anti-P 0 , x2S) were same reagents after ice-cold 9S% washing 1n HEPES-buffered changes . Th.e antibody (goat above. The buffer used SO mM sodium NaN3 Tween primary anti-rabbit, scusors 1r1s, x40) and placed etc.) were the 10 PBS. removed with and with many secondary used for antibody dilution and washing was phosphate pH 7.0 I 9 mglml NaCl I 1 mglml I 1 mglml bovine serum albumin (IgG free, Sigma) I 0 .OS% 20. After each antibody incubation, the corneas were 63 washed at least 4 times, at least 2 hr per wash, by rocking gently at 4°C. Electron microscopy. The breeding and Tr mice was described in Chapter 3. identification of Sciatic nerves from 12- to 14-day-old mice were exposed in situ by carefully slitting the The exposed nerve was bathed overlying muscle with a scalpel. in primary fixative for 10-15 minutes at room temperature. foll o wing pr1mary fixative was used: par a fonoa ldehyde I 0.1 M 0 . 5% In CaC1 2 • a later glutaraldehyde 1.5% cacodylate pH 7.4 s eries of experiments, The I 0.002% 0.1 M sucrose 0.05% CaC1 2 were added to the primary fixative (3). I and The area and number o f mye l in sheaths were similar in wild-type nerves fixed under combined . when t h ese two The volume nerves were conditions, of fixed SC 1n and so cytoplasm was the presence these data reduced, of were however, sucrose. These results are in agreement with those of Dyck and colleagues (4). After fixation in situ, the nerves were gently removed and primary fixation continued for 1-4 hr at room temperature. The nerves were postfixed i n 2% osm1um tetroxide in the same buffer for 1-4 hr a t Epon, and 4 °C, washed, dehydrated 1.n ethanol, embedded 1n thin diamond kn ife . sections ( silver to light gold) cut with a The sections were stained with uranyl acetate and lead citrate, and examined on a Philips 201 or 301 electron microscope. Photographic p rints were made at a final enlargement of x7933 or x8354, and the cross-sectional area of compact myelin measured with a digitizing tablet (Tektronix), under nerves a magnifying glass. Two (from two animals) of 64 each genotype except +/+ were examined. Four +/+ nerves (from four animals, two of each genetic background) were examined. RESULTS By indirect less staining few bright immunofluorescence, Trembler nerves show much for (Figure 1). Po than do wild-type nerves "hot spots" of fluorescence are present, A however (Figures 1, 3 and 4), which are sometimes as brightly labeled as a wild-type myelin sheath (Figures 2 and 3). Immuno- fluorescent staining for myel in basic protein (MBP) indicates a qualitatively similar reduction and heterogeneity in MBP ~· Po staining in Trj/Tr j (not shown ) . Radioimmunoassays on extracts of TrJ /+ mo u s e sci a ti c nerves also indicate that Po and MBP are both reduced 3-5 fold in this genotype, depending presumably on ag e (see Chapters 2 and 3). Although on l y a fe w SCs are brightly labeled for Po in transverse sections of TrJ/TrJ sciatic nerve, most of the rest of the cells are faintly labeled (Figure 4). that this tha t faint enfold labeling is spec i fic, unmyel i nat e d axons 1n we also stained the SCs the cervical trunk of t he +/+ mouse (Figures 5 and 6). labeled by anti-P axons in this 0 (Figure 5), In order to show These SCs were not although occasional myelinated section were very brightly labeled Incubation wit h normal rabbit sympathetic (Figure 6). serum instead of rabbit anti-P 0 produced staining indistinguishable from that 1n Figure 5. seen Two arguments suggest that the brightly stained profiles by immunofluorescence 1n TrJ/TrJ correspond to myelin 65 First, sheaths. the 1n transverse sections at high magnification, brightly stained profiles generally appear as thin-walled cylinders, While the lightly stained cells do not. The halo of fluorescence around the brightest "hot spot" in Figure 1 is due to the vertical orientation of the cylinder; most fluorescent cylinder is not in the plane of focus. too overexposed to demonstrate this point. of the number 1000 l.lm 2 , of based these brightly two com p lete on of the Figure 4 is Secondly, estimates stained profiles Trj /Trj per (1 · sciatic nerve transverse sections ) are 1n rough agreement with the number of myelin sheaths seen electron micrographs 1n (2 1000 l.lm2 , per see Table 1). In longitudinal brightly stained internodes are oriented Trj/Trj sections, the entire wild-type nerve (Figure 2), but only occasional Trj/Trj --- brightly stained longitudinal (Figure 3). sections, In the our phase is best- contrast image suggested to us that most axons could be followed for at least 75 l.1ID before t hey left the plane of section. less, the brightl y stained profiles observed Never the- in these Trj /Trj longitudinal sections averaged on l y 32 l.1ID in length (range: to 45 l.liD; N = ll), suggesting profil e s each represent the axon same single that SCs, and are not brightly stained. the brightly that stained adjacent SCs the on The preservation of morphology 1n frozen sections is not good enough, however, completely rule out 23 possibility that the to apparent ends of these brightly stained profiles correspond to points at which a myelinated axon leaves the plane of section. This objection 66 was removed isolated by staining brightly whole stained mount profiles of preparations, a similar 1n which length were also observed (Figure 7) . The Tr /+ genotype was selected for the cornea whole mount experiment, rather observa tions (see than Tr j /Trj, below) because electron microscopic suggested that heterogeneity 1n sta i ning would al so be present in this genotype, but with much more frequent "hot spo ts . " the cornea, Given the small number of axons 1n brightly stained profiles should be difficult to fin d in TrJ/TrJ cornea. ---- In whole mounts of the +/+ mouse cornea, bright areas of staining are c l ose ly apposed (Figure 7B). Bright areas of staining in :!:5._/+ cornea are generally separated by an tmstained gap tha t is equivalent in length to one or more of the brightly stained profiles (Figure 7A). Although longer gaps between myelin internodes one can find a few in +/+ mouse cornea, as well as a f ew c losely apposed internodes in Tr/+ cornea, these are not the rule. of 2-4 axons, the cornea, The above observations were made on bundles which run circumferentially around n ear t he cornea/ s cleral border. the edge of Visualization of a ll axons in the cornea with indirect hmmunofluorescence, using a monoclonal an ti-neurofilament periodical ly leave these bundles antibody, showed that and run radially in the cen ter of the cornea (not shown). axons towards In the + /+ cornea , the radially oriented axons cease to be mye linated soon after they leave the circumferential bundles. Where more than one radially oriented internode of myelin 1s present on a given +/+ 67 axon, each internode is shorter than the last. t he radially orientated axons do not Interestingly, seem to be myelinated at all in the Tr/+ cornea. In electron micrograph s of +/+ sciatic nerve, quantitation of the cross-sectional area of myelin sheaths showed that their sue is broadly distributed from about 1 to 10 }Jm2 ; histograms (not shown ) pe aks. of this distribution did Eac h step in the genetic not sen.es show any from +/+ definite to Trj /Trj mic e results in about a 2-·f o ld decrease in the average crosssectional area per myelin sheath ( Table 1). The number of myelin sheaths showed 2-fold decreases in the first steps, but a more than 10-fold decrease profiles were scored as from Tr /+ to TrJ /TrJ. Myelin such only when associated with axons. Debris resembling myelin which was clearly not associated with axons (or whose case an axon , was not l.n qualitatively in which if presen t, mi g ht not have been clearly seen), counted. similar ~, inner diameter was less than 0.4 The and Tr/+ freque n cy of such debris Trj /Trj nerves. ---- es t i mated to be This similar to the was roughly frequency was frequency of my elinated axons in TrJ /TrJ (about 2 per 1000 }Jm2 ), although it must be a dmitted. t hat betwe e n the smallest staining organelles . Trj /Trj nerves was 1.s is not a debris of this sort and periodicity of compact The found 11). This not period in+/+ nerves Student's t-test). there sharp dividing other line darkly myelin in to be 14. 1 ± 0 . 8 nm (mt:: a.n ± SD, N = significantly different (13.6 ± 0.8 nm, N = 11; frohl p the myelin > 0.1 by the 68 No myelin sheaths were found entire sciatic nerve cross sections at a (from although two two mice) were One of these nerves was also scanned 1.n photographic prints. completely scanned in Tr /Tr mice, higher magnification directly on the electron microscope, without finding any myelin. The total amount of myelin, represented by the product of myel in sheath size and myelin sheath number, the MBP content of these nerves agrees well with (compare the last column in Table l wi th Tab le 3 of Chapter 3). This suggests, as do the fluo r escence the results, tha t most of nerves is found 1.n compact mye l i n . compact myelin l.n Tr J /Tr J r espectively), values , an d (1% values, Along respec tive l y). these and 0% of wild-type somewhat corresponding measurements of MBP levels type in Estimates of the amount of Tr/Tr were and MBP Po lower (2% with than the and 1% of wildthe fluorescence results, this implies that a significant part of the Po and MBP in not these nerves form l.S compact present 1.n lightly staining cells that do Several myel i n . electron microscopic observations suggest that these lightly staining cells have, in general , enfolded a single axon. enfolded" plus myelinated (Table 2)~ sugge sting axons that t his The total number of "singly is constant total in all represents genotypes the same population o f axons in al l genotypes, namely those competent to induce myelination. similar l.n all (Total genotypes.) nerve cross-sectional Secondly, the area ~~s "singly enfolde-:d" axons, in general, had a larger diameter than axons 1.n the same 69 section that were wrapped m bundles (see Figure 8) suggesting again that the singly enfolded axons represent the "competent" subpopulation. Some of actually wrapped by a spiral that had not condensed the "singly enfolded" axons were of many layers of SC cytoplasm into compact myel in. This phenomenon appeared to show a gene dose effect, 1.n that such spirals were quite extravagant 1.n TrJ/+ nerves (Figure 8B), and rare 1.n Tr/Tr nerve s (Figure 8E). As expected from the results of Chapter 3, no difference 1.n SC numbers was f ound between the various Trembler genotypes (Table 1). DISCUSS lOB Two sheaths arguments indicate that in _:!E.. homozygotes found the small 1.s not recognize very thin myelin sheathB. due number of myelin to failure a to First, sheaths as small as O.l llm 2 were no t difficult to rec ognize, and were, in fact, not infrequent Tr/+ 1n and TrJ/TrJ nerves. This corresponds approximately to a sheath of 2 myelin lamellae around an axon 1 ~ in recogni zed, see 0.1 llm 2 could be the mean sheath size in Trj /Trj nerves (0 .66 ~2 ' diameter. Table 1) sheaths of ---- was " sing ly enfolded" examined at myelin. Although a much larger. axons higher 1.n Secondly, Tr/+ and magnification a few examples Trj/Trj and showed nerves no of were compact 70 The heterogeneity of myelination 1.n Trembler mouse nerves shows a superficial resemblance to the myelin degradation caused by diphtheria toxin. Waksman et al . (5) showed that the injection diphtheria of degradation low of doses of isolated internodes toxin of can myelin, cause the well as as abnormal SC proliferation, while most myelin internodes and all axons remain intact. This phenomenon was subsequently termed "se gmental demyel ination " ( 6) , wh ich term was applied to the Trembler phenotype by early workers (7) because of the spar1.ng of axons, t he h e t erogen eity 1.n myelination evident in electron micrographs of Tr / + nerve, and the relatively short internodes [suggested in Tr/+ ne rv es by a generalized lipid stain (8,9) and also character i st i c of remyelination after diphtheria toxin injection (10)]. resemb l a n ce more l.S demyelination It af ter 1.s than far from clear, superficial. di p theria toxin however, The that heterogeneity injection is the in presumably simply due to t he po tenc y of diphtheria toxin at the molecular level (11). Two ot he r myelinat i on po ss i ble interpretations of the heterogeneity of 1.n Trembler can be ruled out. The t o d i f f er ences between axons, s1.nce h eterogene i ty :ts n o t ad j acen t on the same axon (Figure 7) . It l.S also notable that myelinated unmyelinated axons 1.0 Trj /Trj nerves often have SCs diameter myelination due nerves (Figure 8D). l.S Secondly, behave the quite differently a similar heterogeneity not caused by SC proliferation, and l.n s1.nce these two 71 phenomena have a completely different dependence on gene dose (Table 2). The morphological ensheathment does, however, the heterogeneity reflect heterogeneity 1.n expression the molec ular heterogeneity l.S l.n axonal a corresponding molecular of myelin proteins. due to differences protein synthesis or degradation is unclear. Whether in myelin Indeed, the cause of this heterogeneity could be different from (i.e., subsequent to) the function of the Tr+ gene product. note, however, that a been observed myelination has adding another yet (Chapter 3). defined para lle l similar l.n the between heterogeneity jimpy these mouse two 1.n (12,13), genetic loci The problem posed by this heterogeneity is better l.n Trembler axons tha t somewhat It is interesting to fail than in the jimpy mutant. In Trembler, to be mye l ina ted not only contact mutant SCs, but also induce the SCs to synthesize myelin-specific proteins such as P0 , which the SCs would not otherwise make. axons that fail In jimpy, to be myel i na te d may not be contacted by mutant oli godendrocytes . In conclusion, to perform able interaction. SC one the results first few show that steps in a Trembler SC are normal axon-SC Trembler SCs wrap competent axons in a 1:1 (one to synthesize detect ar le levels of the myelin proteins Po and MBP. Depending on to these axon) fashion, the exact genotype, stage, while a minority of SCs levels of mye l in myelin. most and p ro t ·e ins , of are the induced SCs go on which are a re blocked at this to express much h igher assembled into compac t 72 REFERENCES 1. Rhodin, J. A. G. (1974) Histology: a Text and At las. (Oxford Univ. Press, London). 2. Brockes, J. P., Raff, M. C., Nishiguchi, D. J. & Winter, J. (1980) ~ 3. Neurocytol. 9, 67-77. Arborg, B., Bell, P., Brunk, U. & Collins, V. P. (1976) J. Ult ra struct. Res. 56, 339-350. 4. Ohnish i, A., O'Brien, P. C. & Dyck, P. J. (1976) J. Neurol. Sci. 27 , 193-199 . 5. Waksman, B. H., Adams, R. D. & Mansmann, H. C. (1957) .:G_~ Med. 105, 591-614 . 6. Web ster, H. de F. (1964) Prog. Brain Res. 13, 151-174. 7. Ayers, M. M. & Anderson, R. McD. (1973) Acta Neuropath. (Berlin) 25, 54-70. 8. Low, P. A. (1976) J. Neurol. Sci. 30, 327-341. 9. Low, P. A. ( 1976) J. Neurol. Sci . 30, 343-368 . 10. Allt, G. (1976) in The Peripheral Nerve, Landon, D. N., ed. (Chapman and Hall, London) pp. 666-739. 11. Yamaizumi, M., Mekada, E., Uchida, T. & Okada, Y. 0978) Cell 15, 245-250. 12. BiH ings-Gagliardi, S. ~ Adcock, L. H. & Wolf, M. K. ~~~· 13. ( 1980) Brain 194 , 325-338. Bi11ings-Gagliardi, S., Adcock, L. H., Schwing, G. B. & Wolf, M. K. (1980) Brain Res. 200, 135-150 . 73 Figure 1. · Frozen transv erse sections of mouse sciatic nerve, with rabbit an t i -P 0 f o llowed by r hod amine-labeled goat anti-rabbit IgG, and viewed with rhodamine epi-fluorescence optics. "brightly labeled" cells stain as as br i gh tly exposure. (B) +/+. stained are present, but none +/+ myelin sheaths, (A) 1.n this printed here TrJ /TrJ. Six field of v1.ew at the same 74 10 lliD 75 Figure 2. Frozen longitudinal st ained with anti-Po as before. contrast view of t he same field . muscle, evident here a t i n s e ctioning . section of +/+ mouse sciatic (A) Rhodamine fluorescence. nerve , (B) Phase The nerve was embedded in a piece of the upper left and lower r i ght, for convenience The muscle is not labeled by anti-P 0 • 76 10 JliD 77 Figure 3 . Fro ze n l ongitud in a l section of TrJ/TrJ mouse sciatic nerve, stained wi t h anti-P 0 as before. is evident , (A) Rhodamine fluorescence. A single stained roughly as brightly as +/+ brightly labeled cell myelin sheaths. (B) Phase contrast v1ew of the same field. 78 10 pm 79 Figure 4 . Frozen The was nerve stained with (A) Rh od amine t r ansve rse prefixed with an ti-P 0 as fluorescence, section of TrJ /TrJ mouse 0 .1% paraformaldehyde, de sc ribed printed ~n after sciatic nerve. sectioned, Materia l s and a exposure longer Figures 1-3, to more clearly show the faintly labeled cells . contrast view of the same field. and Methods. than (B) Phase 80 10 Jlffi 81 Frozen transverse section of +/+ mouse cervical sympathetic Figure 5. trunk. The nerve was prefixed with 0.1% paraformaldehyde, and sections stained with anti-P . 0 exposure as Figure (A) Rhodamine fluorescence, printed at the same 4A. Faintly labeled (B) Phase constrast view o f the same field. cells are not present. 82 B 10 Jlffi 83 Figure 6. Fro zen transverse section of +/+ mouse cervical sympathetic trunk, processed as in Figure 5. (A) Rhodamine fluorescence, prin t ed at the same exposure at the same exposure as Figures 4A and SA. myelinated axons are very strongly labeled. not presen t . Occasional Lightly labeled cells are (B) Phase contrast view of the same field. 84 10 Jlffi 85 Figure 7. Po Whole mount preparations of mouse cornea, stained with anti- as describ ed in Materials and Methods. rhodamine fl uoresc e nce opti cs. axons 1s shown. Brightly A bundle of two or more (probably three) stained separated by unstained reg1ons regions. (A) Tr/+ cornea, viewed by patches axon s any given axon are as long as, or longer than the stained (B) +/+ cornea, viewed by rhodamine bundle of two myelinated on 1s shown. fluorescence optics. A Although the axons cross 1n and out of the plane of focus, one can see four brightly stained, normal internodes of myelin on each axon, (normal nodes of Ranvier). separated by small A third myelinated bundle, out of the plane of focus. unstained axon runs gaps below this Other out-of-focus nerve bundles can be seen at upper left and lower right. 86 A 10 I-'m 87 Figure 8. Electron micrographs of transverse sections of sciatic nerves from of var ious mice (D) Trj /Trj. genotypes. (E) .IE./Tr. (A) +/+. (B) Trj/+. (C) Tr/+. 88 A 5JJm 89 c 5pm E 51Jm 91 Table 1. Quantitation of myelin 1.n electron micrographs from sciatic nerves of 12- to 14-day-old mice of various gentotypes. average myelin area ( jlm2) number of total myelin per sheath myelin sheaths area (% of per 1000 jlm 2 wild-type) genotype mean ± SEM N = +/+ 4. 2 ± 0.2 146 70 100% TrJ/+ 2.0 ± 0.2 70 40 27% Tr/+ 1.1 ± 0.1 50 25 9% Trj/Trj 0.66 ± 0.14 22 2 1% Tr/Tr 0.0 0 0 0% The nerves were processed and analyzed as described in Materials and Methods. The last column, "total myelin area," represents the product of columns 2 and 4, expressed as a percentage of the wildtype value. 92 Table 2. Counts of axon and cell numbers 1n electron micrographs from sciatic nerves of 12- to 14-day-old mice of various genotypes. genotype axons * % axons * per 1000 lJm2 myelinated Schwann cells N= N +I+ 73 96% 313 6.0 26 TrJ/+ 52 78% 88 18 30 Tr/+ 69 36% 151 16 35 TrJ /TrJ ---Tr/Tr - 69 3% 254 16 58 71 0% 444 17 109 The nerves were processes and analyzed as described Methods. "N" referes to the 1n Materials total number of axons or and Schwann cells counted , respectively. *Axons were counted only if they were: individually enfolded (i.e., one axon enfolded by one SC), not enfolded (bare), or myel ina ted . of axons enfolded by a single e xtr emel y rare in all genotypes. SC were not counted. Bundles Bare axons were 93 APPENDIX TREMBLER SCHWANN CELLS IN CULTURE 94 MATERIALS AND METHODS Mouse methods Schwann cells similar to (SCs) were purified and cultured by used for rat (1), those SCs with minor Trj /Trj sciatic nerves were ---- modifications as described below. identified as described in Chapter 3, slowly cooled in a medium con tai ning 20% fetal calf sertun and 8% dimethyl sulfoxide, and stored in liqui d nitrogen . This allows one to freeze nerves from single Tr_.:/ Tr J animals and later use the nerves in larger b atches. The nerves were qu ick ly thawed, and dissociated with trypsin and collagenase (l ) . We f ound that fibroblasts were preferentiall y k illed, for unknown reasons, by this freeze/ thaw Wild-type nerves were dissected from 3- to 5-day- procedure. old C57BL/6J mice and frozen as above, as controls. Although the TrJ /TrJ nerves were obtained from 12- to 15-day-old animals fo r ease of identification, 12- to 15-day-old +/+ nerves did not give satisfac tory SC yields . After a treatment few days with a England Nuclear) positive cells 1.n cul t u re , monoclonal and anti-Thy complement (1). the se immunofl uoresc enc e , using fibroblasts a antibody (New England Nuclear) 1.2 were killed by IgM antibody (New Visualization of Thy-1 cultures monoclonal by anti-Thy indirect 1. 2 IgG showed that, as observed in rat sciatic nerve cultures, all of the Thy-1 positive cells have a flat mo rphology, positive. but not all of the flat cells are Thy-1 95 Mouse ~n SCs were maintained Dulbecco 1 s modified Eagle 1 s medium, supplemented with 10% fetal calf serum and partially- purified glial (2), containing 10% growth factor co 2 , at 37°C. in a humidified Cellular atmosphere proliferation incorporation of 125 r-labeled deoxyuridine, assayed by the was as described (2) . RESULTS Wild-type, (3,4), freshly dissociated their wh ich may Ultrastructural (3) have this myelin debris shown culture. that af fect and SCs contain myelin debris behavior immunofluorescence is (4) rapidly ~n vitro. observations eliminated in Therefore, proliferation experiments were confined to SCs that had been cultured for at least one week, and typically two weeks, ~n order to test both genotypes in the absence of myelin debris. Both +/+ and Trj /Trj mouse SCs show a similar sensitivity to partially purified glial growth factor SCs, however, show a l ower c oncent r ation of than rat plateau response at (Figure 1). Mouse a 1000-fold about partially purified glial growth factor SCs do (not shown) . It has been reported that mouse SCs are stimul ated to divide by fibroblast growth factor (5), altho ugh rat SCs are not ( 6) • Since the initial step in the pur ific at ion of fibroblast growth factor from bovine pituitary (7) is similar to the initial step used in the purification of glial growth factor from the same tissue (2), it is likely that this extract growth factor. also contains substantial amounts of fibroblast 96 SCs did not differ greatly from +/+ SCs Trj /Trj background rate of unstimulated proliferation, or 10 their 1n their maximal amount of proliferation (Table 1). Intriguing differences between genotypes 1n morphology were of t en observed. +/+ SCs were generally more elongated , and showe d a greater tendency to orient themselves parallel to each other, than di d Tr j/Trj SCs (Figure 2). The morphology of cells however, attempt 1n these was difference. · cu lt ures made to 1s quite v ariable, quanti f y this apparent and no morphological 97 REFERENCES 1. Brockes, J.P., Fields, K. L. & Raff, M. C. 0979) Brain Res. 165, & Balzer, D. R. Jr. (1980) J. Biol. 105-118. 2. Brockes, J. P., Lemke, G. E. Chem. 255, 8374-8377. 3. Dub ois-Dalcq, M. , Rentier, B., Baron, A. & Burge, B. (1980) Exp. Cell Re_:_ . 131, 283-297. 4. Mirsky , R., Winter, J., Abney, E. R., Pruss, R. M., Gavrilovic, J. & Raff, M. C. (1979) .:!..:_Cell Biol. 84, 483-494. 5. Varon, S. & Manthorpe, M. (1982) Adv. Cell. Neurobiol. 3, 35-95. 6. Raff, M. C .~ Abne y, E., Brockes, J. P. & Hornby-Smith, A. (1978) Cell 15, 813-822. 7. Gospodarowicz , D. ( 1975) J. Bi ol . Chem. 250, 2515-2520. 98 Figure 1. Proli f e r a t i on assayed by 125 ruDR denotes the following of mo use Proliferation was "Fold observed indic a ted pituitary extract dilution) I (cpm observed wit hout pituitary extract). . , +/+ mo use Tr J /TrJ •, determination s. ratio: SC. as c e l l s. described. SC. incorporation Schwann (cpm at Each point represents stimulation" the mean of five 99 10 • 9 • 8 c • •• • 7 ... 6 0 ctJ :;, E 5 ... 4 U) "0 3 0 2 - LL 1 3 4 Log CM 5 6 Dilution 7 8 100 Figure 2. Living mouse SC cultures, shown 1n phase contrast optics. 101 +/ + 100 l.lffi 102 Table 1. Mouse Schwann cell proliferation. Trj/Trj +/+ Trj/Trj +/+ Experiment background background fold stim. fold stim. 1 1675 1098 12.3 23.0 2 1100 643 7.9 9.1 3 528 1283 16.4 21.4 SC proliferati on "Backgro und" extract. was assayed represents t be by c pm 12 5IUDR incorporation incorporated without "Fold stim." represents the follo wing ratio: pituitary extrac t d il uti o n ) / (c pm without entry the mean of at represents experiment. least as described. added pituitary (cpm at optimal pituitary extract). three determinations Each in one