Proc. Natl. Acad. Sci. USAVol. 80, pp. 7019-7023, November 1983Neurobiology

Lipid metabolism of isolated oligodendrocytes maintained in long-term culture mimics events associated with myelinogenesis

(primary culture/myelin maturation/sulfatide/cell polymorphism/tight junctions)

SARA SZUCHET, SUNG HYE YIM, AND SCOTT MONSMADepartment of Neurology, Brain Research Institute, The University of Chicago, 950 East 59th Street, Chicago, IL 60637

Communicated by A. A. Moscona, August 1, 1983

ABSTRACT Oligodendrocytes isolated from ovine white mat-ter according to a published procedure [Szuchet, S., Stefansson,K., Wollmann, R. L., Dawson, G. & Arnason, B. G. W. (1980)Brain Res. 200, 151-164] were cultured for up to 35 days and theircapacity to incorporate precursors into lipids was investigated. Atvarious times, cultures were double labeled with [3H]glycerol/[14C]acetate or [3H]galactose/'SO'T. The cells were harvested72 hr later and lipids were fractionated using standard proce-dures. The time course of incorporation for each precursor wasdistinct. In the days after attachment to substratum, oligoden-drocytes preferentially incorporated [3H]glycerol into phospho-lipids and [14C]acetate into cholesterol while uptake of 5so4- and[3H]galactose into glycolipids was modest. A switch in phospho-lipid metabolism from preferential incorporation into phospha-tidylcholine to incorporation into phosphatidylethanolamine,phosphatidylserine, and phosphatidylinositol occurred at about the10th day in vitro. After 20 days, uptake of [3H]glycerol into phos-pholipids and ['4C]acetate into cholesterol had stabilized but in-corporation of 35SO4- into glycolipids had increased. 35SO4- in-corporation into glycolipids was even greater at 35 than at 20 days.Uptake of [3H]galactose did not change over time. An attempt wasmade to correlate changes in lipid metabolism with morphologicdevelopments. High incorporation into phospholipids and choles-terol coincided in time with the extensive membrane synthesis re-quired for cell attachment and process extension. Differentiationof these newly formed membranes, as assessed by the incorpo-ration of myelin-characteristic glycolipids, galactocerebrosides,and sulfatides, occurred at a time when an intricate network ofprocesses had already been established. The sequence of meta-bolic events observed in vitro parallels that observed at the onsetof myelinogenesis in vivo. We postulate that mature oligodendro-cytes can reenact those early events associated with myelinogen-esis.

Oligodendrocytes (OLG) synthesize membranes that enwrapaxons; by fusion these membranes form myelin (1). Many issuespertaining to this function of OLG are poorly understood. Inparticular, little attention has been paid to the fact that OLGare not a homogeneous population of cells.

During studies on the morphology of OLG and their relationto axons, del Rio-Hortega (2) discovered pleomorphism amongthese cells in terms of size, the disposition of processes, and ininteraction with axons. OLG range from small cells connectedwith numerous axons to large cells attached to single axons. Ithas been calculated that an OLG wrapping a large axon maysynthesize an order of magnitude more membrane than a smallcell ensheathing 39 axons (3). This poses a question as to themetabolism of these two cell types: Are there quantitative dif-ferences between their metabolisms or do qualitative variationsdistinguish them? This question can be rephrased to address

yet another puzzle: Can OLG polymorphism also account forthe compositional heterogeneity of adult myelin (4, 5)?

Another type of OLG heterogeneity, based on ultrastruc-tural features, has come to light: three prototypes have beendescribed (6, 7). It is believed that these types represent stagesin the process of maturation of OLG. Although these obser-vations have been made only on murine species and their gen-eral validity remains unproven, the fact that myelin changes incomposition after its initial deposition, a process referred to asmaturation (4), has led to the hypothesis of specialization amongthe OLG subgroups. It has been suggested that young OLGassemble myelin while mature ones maintain it (8, 9). This for-mulation presupposes that OLG change from a metabolismsuitable for membrane synthesis to one geared to myelin main-tenance. Whether these changes are reversible is a question ofsignificance, because it bears on the issue of the capacity ofOLG to remyelinate.

There is evidence that the onset of myelination is heraldedby a cascade of events, morphological, ultrastructural, and bio-chemical, in OLG. These events reach their maxima midwaythrough myelination and then subside. Once the framework ofmyelin has been set up, architectural remodeling goes on formonths to years depending on the species (4). Thus, perhapsthe proper question is not whether there is a change in me-tabolism; clearly there must be, but rather, is it reversible? Canmature OLG be induced to relive the early episodes of myelin-ation? The availability of procedures to isolate OLG from ma-ture (10-14) and immature (15, 16) brains and to keep them inlong-term culture permits addressing some of these questions.We have improved our procedure for OLG isolation to yield

a highly purified (99%) population that can be kept in culturewith no loss in purity over time (17). Cells are obtained fromlambs that are mature in the sense that they have already formedmyelin. Grazing animals have well-developed myelin at birth.Thus, at the time of their isolation, these cells should have ametabolism geared more to myelin remodeling and mainte-nance than to myelin formation. Taking this as an assumption,we have asked whether or not OLG kept in vitro can revert toa metabolism compatible with myelinogenesis. We have ex-amined the lipid metabolism of cultured OLG by followingthe incorporation of four precursors-[3H]glycerol, ["4C]OAc,[3H]galactose, and 3SO--as a function of time in culture upto and including 35 days. Our results suggest that mature OLGcan reenact the events that precede the initial laying down ofmyelin. It remains to be shown that this preparation for re-myelination can in fact be followed by remyelination. In thispaper, we present biochemical evidence to support our con-

Abbreviations: OLG, oligodendrocytes; PL, NL, and GL, phospho-,neutral, and glycolipids, respectively; PtdCho, phosphatidylcholine;PtdEtn, phosphatidylethanolamine; PtdSer, phosphatidylserine; PtdIns,phosphatidylinositol; GalCer, galactocerebrosides.

7019

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertise-ment" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 9,

202

0

7020 Neurobiology: Szuchet et al.

tention; ultrastructural (18) and immunocytochemical (17) dataare reported separately.

MATERIAL AND METHODSIsolation of OLG. OLG were isolated from ovine white mat-

ter according to a published procedure (9, 11) with the follow-ing modification: the crude cell suspension (P-i) was applied toa linear gradient of 1.0 M sucrose to 1.15 M sucrose/4% (wt/vol) dextran 70, all dissolved in Hanks' balanced salt solution(Hanks' solution) at half strength, and centrifuged at 1,020 Xg for 10 min. Three bands separate on this gradient (11); onlycells from band 3 were used for the experiments described here.Cells were removed from the gradient, diluted by slow additionof 1 vol of Hanks' solution, centrifuged at 700 x g for 5 min,washed once with Hanks' solution, and centrifuged again.

Cell Culture. Detailed descriptions have been given else-where (12, 17). Briefly, a pellet of freshly isolated cells was sus-pended in Dulbecco's modified Eagle's medium/20% horseserum/2 mM glutamine supplemented with antibiotics (Am-photericin B at 0.3 ,ug/ml and garamycin at 12.4 pug/ml) andplated at 2 x 106 cells/ml in Petri dishes. Cultures were main-tained in an incubator at 370C in 95% C02/5% air at 90% hu-midity. Approximately 40% of the cells attach to the Petri dishunder these conditions; the remainder form small floating clus-ters that we refer to as B-3,f OLG. These cells were left to floatfor 96 hr, then the supernatant containing them was removedand centrifuged, and the.cell pellet was collected. Pelleted cellswere resuspen.ded gently in culture medium and replated ontopoly(lysine)-coated plastic Petri dishes. The cells attached within24 hr and extended processes. Cultures were fed twice weekly.

Isotope Labeling. At selected 'times; OLG cultures weredouble labeled with carrier-free H235S04 at 15 ,uCi/ml (782..4mCi/mmol; 1 Ci = 37 GBq) and [3H]galactose at 1 ttCi/ml (11.5Ci/mmol) or with {2-3H]glycerol at 1.5 ,uCi/ml (500 mCi/mmol)and [1-'4CIOAc at 2 ,Ci/ml (56.7 mCi/mmol). All radiochem-icals were from Amersham. Cultures were harvested after 72hr of labeling and washed thrice with 10 ml of Hanks' solution,centrifuging after each wash. The final pellet was processed im-mediately or frozen at -25°C for subsequent analysis..

Analysis of Lipids. Pellets were suspended in a small volumeof H20 and sonicated, and 30-t1k aliquots were removed forprotein determination by the BioRad microassay. The remain-ing cell homogenates were adjusted to 0.5 ml with H20. Lipidswere extracted by addition of 6 ml of CHCl3/CH3OH (2:1) (19),spun on a Vortex three times, and centrifuged for 10 min at2,000 x g. The organic phase was removed with a pipette, threedrops of 0.01% 2,6 di-tert-butyl'p-cresol in CHCI3 was addedto each tube to prevent autoxidation of lipids, and the sampleswere taken to dryness under. N2 on a 40'C water bath. Eachdried pellet was suspended in a drop of CHC13 and applied ona 20 X 6 mm silicic acid column. Neutral lipids (NL), glyco-lipids (GL), and phospholipids (PL) were separated by se-quential elution from this column with 10 ml of CHCl3 (NL),12 ml of acetone/methanol, 9:1 (vol/vol) (GL), and 10 ml ofCH30H (PL) (20). All samples were dried on a water bath un-der N2. Final separation was achieved by TLC'on 250-mm silicagel G plates. (Analtech) 'as follows: NL, 20 X 20 cm plates,developed in n-hexane/diethylether/acetic acid (90:10:1);GL, 20 x 20 cm plates, -developed in CHC13/CH3OH/H20(100:42:6); PL, 10 x 10 cm plates, developed' in n-propanol/methylethyl ketone/H20/HCOOH (48'.36:18:6) for the. firstdimension, dried 30 min in a hood and developed in.CHCl3/CH30H/H20 (100:42:6) for the second dimension. All plateswere activated at 110HC for 1 hr prior to use. Phosphatidylser-ine (PtdSer) and phosphatidylethanolamine (PtdEtn) were vis-

ualized with a ninhydrin spray [0.2% ninhydrin in acetone/HOAC/pyridine (99:0.5:0.5)]; all other-lipids were -visualizedby 12 vapor staining. NL and PL were. identified by cochro-matography with lipid standards (Sigma); GL were identifiedby cochromatography with. GL isolated from sheep brain (21).The lipid spots were scraped from the plates and placed in 7-ml minivials, and 400 ,1 of. H20 and 5 ml of ACS. fluid wereadded (Amersham 196290). Radioactivity was determined in aBeckman LS-7000 scintillation counter. All datawere normal-ized' by dividing cpm by mg of protein.

RESULTSThe incorporation of 35SO2-, [3H]galactose, [14C]OAc, and[3H]glycerol into GL, NL, and PL by B-3,f OLG was followedas a function of time in culture for up to 35 days. 35SO2- and[3H]galactose were metabolized into GL; [14C]OAc went pref-erentially into NL and [3H]glycerol, into PL. The uptake of [3H]-galactose was essentially constant for the first 20 days, with amoderate upsurge thereafter (Fig. 1)., In contrast, uptake of theother precursors registered increases fromn the start (Fig. 1).

Examining first the metabolism of [3H]glycerol into PL, we-find a 4-fold increase in incorporation from day 4 to day 7 (Fig.1). The increase in the uptake of [I4C]OAc chiefly into cho-lesterol (see below), over this same time span, is modest but,by 10 days, the intake of [14C]OAc is equivalent to that of [3H]-glycerol. From then on, the two precursors run a parallel coursethat peaks at day 20. The incorporation of 35SO2- into sulfatide-is still increasing at day 35.

It is important to correlate these observations with morpho-logical events: the measurements taken at day 4 correspond toB-3,f OLG in the floating state (Fig. 2a); at day 7, the cells haveattached and begun to extend processes. Thus, the burst in syn-thesis of PL coincides with drastic morphological changes thatrequire substantial formation of membranes. The short lag inthe synthesis of cholesterol at this time is, per se, intriguing.

2.5 [

r.

a,2.00bo

o 1.5x

0ig' 1.0Iaq$Lo0S.,0

0 0.5

0 5 10 15 20 25 30Time in culture, days

35

FIG. 1. Incorporation of labeled precursors into PL, NL, and GL by,OLG over time in culture. At the indicated times cultures were labeledwith-either [3H]glycerol/[14C]OAc or H235S04/[AHigalatse for 72hr.Lipids were isolated following standard procedures and the label foundin each fraction was normalized, per mg protein. o, 1[3H]Glycerol intoPL; A, [14C]OAc into NL; *, 35SO2- into sulfatides; a, [3H]galactoseinto total GL. (Bar = SEM; n = 6-9.)

Proc. Natl. Acad. Sci. USA 80 (1983)

Dow

nloa

ded

by g

uest

on

Dec

embe

r 9,

202

0

NSProc. Natl. Acad. Sci. USA 80 (1983) 7021

a

FIG. 2. Morphological changes accompanying attachment of OLG to substratum and their further development with time in culture. (a) OLGafter 4 days in vitro plated under conditions in which they do not attach. Note the absence of processes. (Stain, orceine; bar = 5. gm.) (b-d) Phase-contrast micrographs of OLG after 18-25 days (b, 18; c, 20; d, 25) in culture. Note the profuse network of processes and alignment (--) of cells inrows. [Bar (see a) = 20 ,um.]

The morphological development of the cells peaks at 20-25 daysas judged by the profuse network of processes seen at this time(Fig. 2 b-d) and. by the formation of tight junctions between

adjoining cells (18). Events at the biochemical level are con-. sistent with the idea that this represents a turning point towardsmembrane differentiation.

Neurobiology: Szuchet et al.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 9,

202

0

7022 Neurobiology: Szuchet et al.

The distribution of [3H]glycerol among the various PL is alsoof significance (Fig. 3A): forthe first 10 days in vitro, B-3,fOLG direct the bulk of this label into phosphatidylcholine(PtdCho), only small amounts being incorporated into other PL.Beyond day 10, dramatic changes take place: incorporation of(3H)glycerol into PtdCho decreases while incorporation intoPtdEtn, PtdSer, and phosphatidylinositol (PtdIns) increases.After 20 days, the intake of label for most PL examined remainsrelatively constant except for PtdIns, which decreases, andPtdEtn plus PtdSer, which increase, though more slowly thanbefore (Fig. 3A). Little [3H]glycerol is incorporated into eithersphingomyelin or lysolecithin.A different perspective is obtained when incorporation of

["4C]OAc is followed (Fig. 3B). Here, sphingomyelin appearsas an important acceptor of label, second only to PtdCho. Thetime course of the latter is basically the same for both labels,though the magnitudes differ. The same holds for PtdIns. Little[14C]OAc is incorporated into PtdEtn and PtdSer (Fig. 3B).Among the NL, cholesterol is the leading compound for both

precursors-i.e., t14C]OAc and [3H]glycerol-the patterns ofincorporation with time are similar but the magnitudes are dis-tinct (Fig. 4). Although in terms of cpm/mg of protein, thereis a significant intake of radiolabel into other NL, they are over-shadowed by cholesterol (data not shown).

Turning now to myelin-characteristic GL, we note the fol-lowing. (i) The uptake of 3SO2- into sulfatides is much higherthan that of [3H]galactose into either sulfatides or galactocer-ebrosides (GalCer) (Fig. 5). (ii) The time course of incorpo-ration of these precursors differs. Thus, whereas the incor-poration of 35SO2- registers a significant increase, very littlechange is seen in the incorporation of [3H]galactose over thesame time span. (iii) Within experimental error, the ratio ofradiolabel taken up into GalCer and a-OH-GalCer is main-tained at 1 (at least during the period examined; Fig. 5B Inset).(iv) The constancy of this ratio suggests that both pools-i.e.,a-OH-GalCer and GalCer-contribute comparably to the syn-thesis of sulfatides. Bands corresponding to hydroxysulfatideand nonhydroxysulfatide were visualized on the chromatogram;no attempt to separate them was made.

Even if only relatively small amounts of the other precursorsare taken up by these GL, the manner in which this is done isinteresting. [3H]OAc intake exceeded that of [3H]glycerol (datanot shown), presumably because the former comes in via fattyacid chain elongation, an important premyelination step.

It has been suggested that the GL/PtdCho ratio is a measureof myelin maturation (4). In this context, it is pertinent to com-pare the evolution of sulfatide taken as a representative for GL

O

-IS,A B

0

,15 50

4

10x 3E.:

041

0 5 10 15 20 25 0 5 10 15 20 25Time in culture, days

FIG. 3. Incorporation of [3Hlglycerol (A) and ["4C]OAc (B) into PLby cultured OLG. *, PtdCho; *, PtdSer/PtdEtn; ci, PtdIns; o, lysoleci-thin; o, sphingomyelin. (Bars = SEM; n = 6-9.)

a

20._

0hoI-

06x 10

0JC.)

ae0~-+,_

I.

0 5 10 15 20Time in culture, days

25 30

FIG. 4. Uptake of ['4C]OAc (e) and [3Hlglycerol (a) into choles-terol by OLG maintained in culture. (Bars = SEM; n = 6-9.)

(the contribution of GalCer would merely add a constant andtherefore not affect the shape of the curve) and PtdCho as OLGmature in culture (Figs. 1 and 3A). While the plots representincorporation and not necessarily de novo synthesis and thenumbers are cpm and, hence, not directly comparable, eachcurve is self-consistent. Hence, the fact that the uptake of labelinto sulfatide increases with time in culture, whereas that ofPtdCho decreases, must be significant.

DISCUSSIONWe have investigated the incorporation of precursors into PL,NL, and GL over time by cultured OLG. Lipid metabolismparallels morphological development. Thus, at the time of cellattachment and process extension, a period of extensive mem-brane synthesis, OLG synthesize PL and cholesterol prefer-entially (Fig. 1). Further differentiation of these newly formedmembranes, as assessed by incorporation of myelin-character-istic GL, GalCer, and sulfatides, occurs at a slower rate. These

a1)

0~*4.0

0

0~

bID

C.)-.'x

.o

0ccsw

8.)a

a1)

0

0

S.,a2)

CZ

0 5 10 15 20 25 30 35Time in culture, days

FIG. 5. Incorporation of 35SO2- and 13H]galactose into sulfatides(A) and GalCer (B) by cultured OLG as a function of time in vitro. (A):, [35S]Sulfatide; --, [jHasulfatide. (B) 3, l3H]GalCer, 1, a-OH-[3H1-GalCer; e, [3H]GalCer/a-OH{3H]GalCer. (Bars = SEM; n = 6-9.)

Proc. Natl. Acad. Sci. USA 80 (1983)

Dow

nloa

ded

by g

uest

on

Dec

embe

r 9,

202

0

Proc. Natl. Acad. Sci. USA 80 (1983) 7023

findings fit well with what has been observed in vivo (4, 22, 23).Insight into the lipid metabolism of cultured OLG and its

relatedness to that expressed by OLG in situ is obtained fromexamination of the variation in PL composition with time inculture (Fig. 3A). PL, NL, and GL account for almost all lipidspresent in OLG and myelin (4, 24, 25). However, their relativeproportions differ: myelin is enriched relative to OLG in cho-lesterol and GL at the expense of PL. Within the latter, thereis an unequal preponderance of classes; PtdCho is the major PLof OLG, PtdEtn is that of myelin (4, 24-26). Lipid compositionchanges slowly during development and transforms what wasan extension of OLG plasma membrane into myelin. Matura-tion of myelin can be measured by two indices: the molar GL/PtdCho and PtdEtn/PtdCho ratios. Both increase with matura-tion (4). Although these conclusions stem from studies on mu-rine species, the fact that myelin composition varies little acrossspecies (4, 22) justifies the extrapolation made here.

Figs. 1 and 3A indicate that cultured OLG recapitulate, atleast qualitatively and in the proper sequence, events associ-ated with myelinogenesis. This conclusion is strengthened bythe findings that the ultrastructural characteristics of these cellsalso match those found in situ (1, 27, 28). Thus, the culturedcells have numerous microtubules and a highly developed Golgi,align themselves in rows as do interfascicular OLG, and formtight junctions between adjoining cells and at points where pro-cesses make contact (18). The cells express myelin basic protein(12) and myelin-associated glycoprotein (17). We have mea-sured the specific activity of 2',3' cyclic nucleotide 3'-phos-phodiesterase as a function of time in culture; specific activitystays high throughout, typically, 5-6 (gmol/min)/mg of pro-tein (17). Drastic changes in the levels of incorporation of [ H]-leucine into myelin-characteristic proteins follow cell attach-ment to substratum and process development (unpublished re-sults).

It is important to stress that these myelinogenetic activitiesof cultured B-3,f OLG are expressed in the absence of neuronsor astrocytes. Thus, if a signal is required to alert OLG thattheir metabolism has to be changed, it must be programmedwithin the OLG. This is not to say that OLG might not functionbetter in the presence of other cell types. Bhat et al. (16) havereported that OLG need a non-OLG signal for enhancedexpression of their activities. No myelin is formed in these cul-tures. However, when B-3,f OLG are cocultured with rat dor-sal root ganglion neurons, myelin is formed (29). Whether thisis myelination or remyelination remains to be elucidated.The metabolism of myelin-typical GL is interesting, if puz-

zling. Note that incorporation of 35so- into sulfatide is muchhigher than that of [3H]galactose into either GalCer or sulfa-tide. This may indicate a complex mechanism of sulfation. Re-sults obtained by Benjamins and co-workers (30, 31) using ratssupport this idea. Those authors found little labeling of sulfa-tides with [3H]galactose and concluded that there must be twoGalCer pools. Our data are also consistent with a more rapidturnover of sulfatides than of GalCer. Rapid turnover of sul-fatides has been postulated (32, 33).

B-3,f OLG incorporate similar levels of [3H]galactose into a-OH-GalCer and GalCer, thus maintaining their ratio at essen-tially 1. A value of 1.4 was reported for this ratio in sheep whitematter (34), but the composition of mature white matter isdoubtless different from the membranes synthesized initiallyby the OLG.We have shown that OLG isolated from white matter of young

brains and maintained in vitro exhibit a lipid metabolism thatqualitatively and temporally mimics that seen at the onset ofmyelination. From these results, we postulate that mature OLGcan revert to a lipid metabolism compatible with myelinogene-sis. Preliminary observations (unpublished) suggest that this

holds also for protein metabolism. These findings are relevantto remyelination and, hence, bear on diseases such as multiplesclerosis. In addition, our results show that OLG function wellin the microenvironment provided and that the absence of othercell types or specific humoral factors does not lead to metabolicaberrations. We believe that this culture system offers a uniquemodel to address questions concerning OLG and the molecularand cellular interactions that lead to myelination.

We thank Drs. B. G. W. Arnason and G. Dawson for reading themanuscript and making useful suggestions. We are grateful to Miss K.Schlaug and Mr. P. E. Polak for their excellent technical assistance.This research was supported by Grant RG 1223-B3 from the NationalMultiple Sclerosis Society.

1. Peters, A., Palay, S. L. & Webster, H. F. (1976) Fine Structureof the Nervous System (Saunders, Philadelphia), pp. 232-263.

2. del Rio-Hortega, P. (1928) R. Soc. Espanola Hist. Nat. 14, 4-122.3. Blakemore, W. F. (1981) in Development in the Nervous System,

eds. Garrod, D. R. & Feldman, J. D. (Cambridge Univ. Press,Cambridge, England), pp. 289-308.

4. Norton, W. T. (1977) in Myelin, ed. Morell, P. (Plenum, New York),pp. 161-199.

5. Waehneldt, T. V. & Linigton, C. (1980) in Neurological MutationsAffecting Myelination, ed. Baumann, N. (Elsevier/North-Hol-land, Amsterdam), pp. 389-412.

6. Mori, S. & Leblond, C. P. (1970)J. Comp. Neurol. 139, 1-30.7. Kruger, L. & Maxwell, D. S. (1966) Am. J. Anat. 118, 411-436.8. Varon, S. S. & Somjen, G. C. (1979) Neurosci. Res. Prog. Bull. 17,

1-239.9. Szuchet, S. & Stefansson, K. (1980) in Advances in Cellular Neu-

robiology, eds. Fedoroff, S. & Hertz, L. (Academic, New York),Vol. 1, pp. 313-346.

10. Szuchet, S., Wollmann, R. L. & Arnason, B. G. W. (1978)J. Neu-ropathol. Exp. Neurol. 37, 349 (abstr.).

11. Szuchet, S., Arnason, B. G. W. & Polak, P. E. (1980)J. Neurosci.Methods 3, 7-19.

12. Szuchet, S., Stefansson, K., Wollmann, R. L., Dawson, G. & Ar-nason, B. G. W. (1980) Brain Res. 200, 151-164.

13. Gebicke-Hairter, P. J., Althaus, H. H., Schwartz, P. & Neuhoff,V. (1981) Dev. Brain Res. 1, 497-518.

14. Lisak, R. P., Pleasure, D. W., Silberberg, D. H., Manning, M.C. & Saida, T. (1981) Brain Res. 223, 107-122.

15. McCarthy, K. D. & deVellis, J. (1980) J. Cell Biol. 85, 890-902.16. Bhat, S., Barbarese, E. & Pfeiffer, S. E. (1981) Proc. Natl. Acad.

Sci. USA 78, 1283-1287.17. Szuchet, S. & Yim, S. H. (1984)J. Neurosci. Res., in press.18. Massa, P. T., Szuchet, S. & Mugnaini, E. (1983) Biophys. J. 41,

69a (abstr.).19. Folch, J., Lees, M. & Sloane-Stanley, G. H. (1957)J. Biol. Chem.

226, 497-509.20. Dawson, G. (1979) J. Biol. Chem. 254, 155-162.21. Mack, S. R. & Szuchet, S. (1980) Brain Res. 214, 180-185.22. Davison, A. N. (1970) in Myelination, eds. Davison, A. N. & Pe-

ters, A. (Thomas, Springfield, IL), pp. 80-182.23. Benjamins, J. A. & Smith, M. E. (1977) in Myelin, ed. Morell, P.

(Plenum, New York), pp. 233-270.24. Poduslo, S. E. & Norton, W. T. (1972)J. Neurochem. 19, 727-736.25. Farooq, M., Cammer, W., Snyder, D. S., Raine, C. S. & Norton,

W. T. (1981) J. Neurochem. 36, 431-440.26. Cohen, S. R. & Bernsohn, J. (1978)J. Neurochem. 30, 661-669.27. Massa, P. T. & Mugnaini, E. (1982) Neuroscience 7, 523-538.28. Webster, H. deF. & Sternberger, N. H. (1980) in Neurological

Mutations Affecting Myelination, ed. Baumann, N. (Elsevier/North-Holland, Amsterdam), pp. 73-86.

29. Wood, P., Szuchet, S., Williams, A. K., Bunge, R. P. & Amason,B. G. W. (1983) Trans. Am. Soc. Neurochem. 14, 212a (abstr.).

30. Benjamins, J. A. & Iwata, R. (1979) J. Neurochem. 32, 921-926.31. Benjamins, J. A., Hadden, T. & Skoff, R. P. (1982)J. Neurochem.

38, 233-241.32. Burkart, T., Hofmann, K., Siegrist, H. P., Herschkowitz, N. N.

& Wiesmann, U. N. (1981) Dev. Biol. 83, 42-48.33. Burkart, T., Caimi, L., Siegrist, H. P., Herschkowitz, N. N. &

Wiesmann, U. N. (1982) J. Biol. Chem. 257, 3151-3156.34. Nonka, G. & Kishimoto, Y. (1979) Biochim. Biophys. Acta 572, 432-

441.

Neurobiology: Szuehet et al.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 9,

202

0