THE JOURNAL OF BIOLOQICAL CHEMISTRY Vol. 248, No. 4,Issue of February 25, pp. 1231-1239, 1973

Printed in U.S.A.

Glycosphingolipids of Clonal Lines of Mouse Neuroblastoma

and Neuroblastoma X L Cell Hybrids*

(Received for publication, August 29, 1972)

GANESA YOGEESWARAN AND ROBERT K. MURRAY$

From the Department of Biochemistry, University of Toronto, Toronto, Ontario, Canada

MARK L. PEARSON AND B. D. SANWAL

From the Department of Medical Cell Biology, University of Toronto, Toronto, Ontario, Canada

F. ARTHUR MCMORRIS AND FRANK H. RUDDLE

From the Department of Biology, Yale University New Haven, Connecticut 06510

SUMMARY

Glycosphingolipid analyses were performed on three clonal lines of the mouse Cl300 neuroblastoma (NZA, NA, and NB41A) and on three neuroblastoma X L cell hybrid lines (NLI-15A, NLI-7A, and NLI-11A) differing in teeir membrane electrical activities. Striking variations Of ganglioside but not of neutral glycosphingolipid patterns were observed in the neuroblastoma lines. No differences were observed between the glycosphingolipid patterns of neuroblastoma cells (NZA) grown in suspension as “un- differentiated” neuroblasts or in monolayer as “differen- tiated” neurons. The neuroblastoma X L cell hybrid lines exhibited neutral glycosphingolipid patterns similar to the sum of the the parental neuroblastoma and L cell patterns, whereas their ganglioside patterns differed from those of the two parents. Despite their widely differing membrane electrical activities, the glycosphingolipid profiles of these three hybrid lines were generally similar.

The glycosphingolipids are prominent components of the membranes of neural tissues. Gangliosides, for example, are found in particularly high concentrations in fractions of brain homogenates enriched in synaptosomal elements (1). Although the functions of glycosphingolipids are unknown, it is possible that the gangliosides play an important role in synaptogenesis and neuronal differentiation, since comparative studies of the ganglioside profiles of rat and human brain during development show a progressive increase of certain higher ganglioside homo- logues (2). Others have suggested that the gangliosides may be implicated in the ion transport processes associated with electrical excitation (3, 4).

* This investigation was supported bv grants from the Medical Research Council of Canada, the National Cancer Institute of Canada and Grant 5-ROl-GM-09966 from the National Institutes of Health.

$ To whom correspondence should be addressed.

The physiological functions of GSLsl in neural cells are difficult to study because of the variety of different cell types present in the central nervous system, and because it is difficult to alter ex- perimentally the GSL content in attempts to correlate glycolipid composition with cellular function. For these reasons we wished to study variations in the GSL profiles of cloned neuronal cell lines during differentiation in culture. In this regard, the neuro- blastoma lines derived from the mouse Cl300 tumor are of par- ticular interest because these cells retain many of the characteris- tics of bona Jide neurons, including the capacity to generate neu- rites, to synthesize known neurotransmitters, and to undergo at least some of the early steps in the formation of neuromuscular junctions (6-9). In addition, these cells offer the potential of studying neuronal function at the genetic level by the isolation of appropriate mutants, and by cellular hybridization using the techniques of somatic cell genetics (9).

In the present investigation we have examined the GSL profile of several different clonal isolates of the Cl300 neuroblastoma and of three neuroblastoma X L cell hybrids differing in their membrane electrogenic response. Basically, three questions were asked. (a) Do different clonal isolates of neuroblastoma cells, like other cultured mammalian cells previously studied (lo-la), show marked differences of GSL profile? (b) Does a change of GSL pattern occur during the transition in culture of “undifferentiated” neuroblasts to “differentiated” neurons? (c) Do neuroblastoma X L cell hybrids of differing membrane electrogenic response display significant differences in ganglioside pattern, perhaps permitting a correlation to be made between complexity of ganglioside pattern and this response? Studies on the GSL composition of two of the neuroblastoma lines studied here (NB41A and N2A) have recently been reported by Dawson et al. (13, 14).

MATERIALS AND METHODS

Cell Lines and Their Properties-Table I lists some of the relevant properties of the cell lines used in these studies. N2A

1 The abbreviations used are: GSL, glycosphingolipid; HGPRT, hypoxanthine-guanine phosphoribosyltransferase; TK, thymi- dine kinase; GMI, GM*, etc., ganglioside notation according to the system of Svernnerholm (5).

1231

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

1232

TABLE I Properties of cell lines

Cell line

N2A

NB41A

NA LM (TK-) NLI-‘IA NLI-11A NLI-15A

84 5 2

n.d.e

92 f 3 47 f 3

151 f l( 102 f 11 152 f 1:

rmzoles/ min/mg n.d.

P&a

n.d.

n.d. n.d.

107 0 9 5

51

0.84 <O.lO

0.19 <O.lO

0.12

14-3-Z anti- gent

e

s _

++

++

++ -

++

+ ++

Source

Klebe and Rud- dle (15)

Augusti-Tocco and Sato (6)

KlebeZ Kit et al. (16) MeMorris McMorri@ MeMorris

0 Mean chromosome numbers and standard deviations are quoted.

b Activity is expressed in nanomoles of acetate released per min per mg of protein at 37”.3

c Neuron-specific protein, micrograms of protein per mg of soluble protein.4

d Membrane electrogenic response expressed on a graded scale of - to ++, where - indicates only passive changes in the transmembrane potential following current injection, and ++ indicates repetitive action potential generation.5

e Not determined.

and NB41A are independent cloned cell lines isolated directly from the Cl300 tumor (Jackson Laboratories, Bar Harbor, Me.) by Klebe and Ruddle (15) and Augusti-Tocco and Sato (6). NA is an 8-azaguanine-resistant hypoxanthine-guanine phos- phoriboxyltransferase (EC 2.4.2.8)-deficient (HGPRT-) mu- tant derived from N2A.2 These lines have been grown in cell culture for periods of 2 to 4 years. LM(TK-) is a 5-bromode- oxyuridine-resistant thymidine kinase (EC 2.7.1.21)-deficient (TK-) L cell (16). The NLI lines are neuroblastoma X L cell hybrids derived from Sendai virus-mediated fusion between NA and LM(TK-) cultures.3 These are independent clones iso- lated on the basis of their ability to grow in the HAT selection medium (17) and have been characterized as hybrids by the detection of both parental and hybrid bands of phosphohexose isomerase by starch gel electrophoresis, and by the presence of marker chromosomes from both parents. A more detailed description of the production and characterization of the NLI hybrids will be presented elsewhere.3

The N2A line is able to grow in suspension cdlture as well as in monolayer culture using the same medium (18). Cells grown in spinner culture as small round cells possess no observ- able neurites and are referred to here as “undifferentiated” cells, in contrast to the monolayer-grown cells which are referred to as “differentiated”, since they possess many long neurites (6, 7).

The morphology of the different neuroblastoma cells and neuroblastoma X L cell hybrids observed by phase contrast microscopy is shown in Fig. 1.

When it became apparent that different neuroblastoma lines had different GSL patterns, we selected several subclones of N2A to check the GSL variation within one cell line. These

2 R. J. Klebe, unpublished data. 3 F. A. McMorris, in preparation. *A. Kolber, F. A. McMorris, B. W. Moore, and A. S. Perumal,

in preparation. 5 D. Nelson J. Peacock, and F. A. McMorris, in preparation.

subclones, referred to as N2A-A5, N2A-A4, and N2A-D4, were subjected to two successive cloning procedures and were grown for a minimum of 20 generations prior to GSL analysis.

Cell Growth-Cells were routinely grown at 37” in monolayer culture in plastic tissue culture flasks (Falcon Plastics, Los Angeles, Calif.) in antibiotic-free Dulbecco’s-modified Eagle’s medium (Grand Island Biological Co., Grand Island, N.Y.) supplemented with 10% agamma newborn calf serum (Grand Island Biological Co.) (or fetal calf serum in the case of NB4L4) in an atmosphere of 5% carbon dioxide and 95% air. All lines were routinely checked for mycoplasma, and except in one case (see Fig. 3A) were free of any detectable contamination. For large scale preparations of cells not radioactively labeled, the cells were grown in 150-mm diameter dishes or 250-ml flasks to the late log phase of growth, the medium was decanted, the monolayer washed with phosphate-buffered saline (0.14 M NaCl, 0.015 M potassium phosphate, pH 7.0), the cells harvested by scraping with a rubber policeman, then sedimented, and the pellets frozen at - 20” until the GSL analyses were performed.

Incubation of Cells with o-[1-14C]Galactose--o[l-14C]Galactose has previously been found to be an excellent precursor for labeling the GSLs of cultured cells (19, 20). Exposure of cultured cells to this isotope for a period of 24 hours, with subsequent analysis of GSL pattern by radioautography of developed thin layer chromatograms, was shown to permit a qualitative evaluation of the GSL profile of as few as 1 x lo5 cells (20). Accordingly, this procedure was utilized to examine the GSL content of several of the cell lines studied here, particularly the neuroblas- toma X L cell hybrids.

Cells were treated with Viokase (Grand Island Biological Co.) and plated at a density of approximately 5 x lo5 cells per 25- cm2 tissue culture flask. The cultures were grown at least over- night before decanting the medium and replacing it with 4 ml of fresh medium containing 5 &i of n-[1-‘4Clgalactose (55.7 mCi per mmole, New England Nuclear Corporation, Boston, Mass.). After 24 hours of incubation with the isotope, the cells were harvested as described above.

Isolation and Characterization of Glycosphingolipids-The frozen cell pellets were extracted and the neutral GSLs and gan- gliosides isolated and analyzed according to the general method- ology described previously (11, 21). The patterns of ganglio- sides and neutral GSLs of at least three separate cultures of each cell type studied were examined qualitatively by thin layer chromatographic analyses. Quantitative analyses of ganglio- sides and neutral GSLs were performed on two of these cultures.

Radioautography-Radioactive preparations of cell ganglio- sides and neutral GSLs (containing approximately 2 X lo4 dpm) were separated as described below. The thin layer plates were covered with Kodak No-Screen x-ray film, kept in the dark for approximately 1 week and the films were then developed. The plates were sprayed with appropriate reagents to detect gangliosides and neutral GSLs and the positions of the radio- active GSLs were compared with standard GSLs.

Thin Layer and Gas-Liquid Chromatographic Analyses of

Glycosphingolipids-The gangliosides were separated and de- tected by thin layer chromatography on glass plates coated with Silica Gel G in a solvent system of chloroform-methanol- ammonia-water (60:35: 1:7, v/v) as described by Wherrett and Cumings (22).

The neutral GSL fraction was separated by thin layer chroma- tography on glass plates coated with Silica Gel G in a solvent system of chloroform-methanol-water (65:25:4, v/v) and de- tected by spraying with aniline-diphenylamine reagent (23).

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

1233

For quantitative analyses and determination of the carbo- with methanol-water (1:2, v/v). Appropriate areas of the hydrate composition of the GSLs by gas-liquid chromatography, chromatograms were scraped into screw-capped tubes, a meas- the gangliosides and neutral GSLs were separated as described ured aliquot of mannitol added as an internal standard, and

above and detected by their opaque appearance after spraying the contents subjected to methanolysis (methanolic hydrochloride

FIG. 1. Morphology of the various cell lines used in these arranged in order of decreasing membrane electrical activity. studies. A, neuroblastoma clones NdA, NA, and NB&A. The The NA S-azaguanine-resistant neuroblastoma and the LM(TK-)

bar represents 100 pm. B, Neuroblastoma X L cell hybrids and bromodeoxyuridine-resistant L cell parents are shown below the their parents. The hybrids NLI-MA, NLI-7A, and NLI-1lA are hybrids. The 6ar represents 100 pm.

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

1234

(0.75 s)) for approximately 20 hours at 80”. The sugar compo- sition of the individual GSLs was determined by gas-liquid rhromatographic analyses of the trimethylsilyl ethers of the O- methyl glycosides as described by Vance and Sweeley (24) and Clamp et al. (25). The procedure was previously utilized to analyze the GSLs of cultured mouse fibroblasts (11). The analy.ses were prformed in a Hewlett-Packard model 7610 A gas chromatograph equipped with dual glass columns (6 ft X J/4 inch outside diameter) packed with 3.8% SE-30 on 80/100-mesh diatoport S. The instrument was programmed from 120-180” at 1” per min for analyses of gangliosides and from 120-155” for analyses of neutral GSLs. The amount of glucose present in each GSL was determined by comparison with the internal man- nitol standard and the concentration of individual neutral GSLs in the different cells calculated and espressed relative to cell protein (nanomoles of GSL per mg of total cell protein). The amounts of the other carbohydrates present in the various GSLs were also determined, yielding the stoichiometry of each sugar relative to glucose.

Sfandards of (=SLs-Gangliosides and neutral GSL standards were prepared from human brain and kidney obtained at au- topsy. Standards of monogalactosyl diglyceride and digalac- tosyl diglyceride were purchased from Applied Sciences Lab- oratories, Inc. (State college, Pa).

RESULTS

Ganglioside Content oj Various Cells Stzcdied-Qualitative and quantitative data on the patterns of gangliosides in the various cell lines studied are shown in Fig. 2 and Table II, respectively. Relatively complex pattenrs of gangliosides were found in all the neuroblastorna and NLI hybrid cells excepting NB41 A (Fig. 2B, C’hamiel 1). For example, N2A cells grown in suspension as small round cells lacking neurites (“undifferentiated”) (Fig. 2.1, 0ianuel 1) exhibited ganglioside bands corresponding in

A

1 2345 1 z 1 2345 123

chromatographic migration to standards of GM?, GM?, GM,, and GDI,. Identical ganglioside patterns were observed in three independent subclones of N2A cells (each cultured for at least 20 generations after cloning) grown in monolayer as “dif- ferentiated” cells possessing extensive neurites (Fig. 2.1, Chan- nels d to 4). These results indicate that no change in ganglio- side profile accompanied the process of neurite estension and also provided a measure of the constancy of ganglioside compo- sition in subclonal lines over this time span. Contrasting with this latter finding was the marked difference in ganglioside pat- terns between N2A (Fig. 2A, Chamiels 1 to 4) and Ws41.1 (Fig. 2B, Channel 1), two neuroblastoma lines isolated inde- pendently from the Cl300 tumor. Similarly, the ganglioside pattern of NA cells (the HGPRT-deficient mutant of 52.i) differed markedly from that of the parent N2*4 aells (comparison of Fig. 2C, Channel 1 with Fig. 2A, Channels 1 to 4).

The quantitative analyses shown in Table II confirmed the differences in ganglioside pattern just described. GM:, was found to be the principal ganglioaide of N2h cells (comprising approsimately 754; of the total ganglioside on a molar bahi>), whereas it was almost absent in the NB41A and NA (bells. On the other hand, G& was a relatively minor component in the N2A cells (approximately 10 , 0; of the total), whcrcas it wah the major ganglioside of NR41A cells (90% of the total) and a prorninent component (about 257;) of the K\;;\ cells.

The ganglioside patterns of the three independently irolated neuroblastoma X L cell hybrids were also investigated. The parents were the NA(HGPR’I’-) cells and the LJ1(TII;-) c~ells. The hybrids differed from each other in their membrane electro- genie response (see Table 1) and were selected for htudy in part, to examine a possible role of gangliosides in membrane elcita- bility. i2s shown in Fig. 2C (Uiannels d to ,G) and Table II, the three different NLI hybrid cell lines showed relatively minor differences in gnnglioside profile. NLT-15.1 contained con-

D

FIG. 2. Thin layer chromat,ograms of the gangliosides of the various cells studied. A: gangliosides of: 1, N2A cells grown in suspension cult’ure (“undifferentiat,ed”) ; 2, N2A-A5 subclone groan in monolayer culture (“differentiated”); 3, N2A-A4 sub- clone grown in monolayer cult)ure (“differentiated”) ; 4, N2A-D4 subclone grown in monolayer culture (“differentiated”); 5, human brain, normal. B: gangliosides of: 1, NB41A cells grown in mono- layer cult.ure; 2, human brain, normal. C: gangliosides of: 1, NA parent neuroblastoma cells; 2, NLI-7A hybrid cells; 3, NLI-1lA hybrid cells; 4, NLI-15A hybrid cells; 5, human brain, normal

(with some Tay-Sarhs ganglioside added). I): gangliwidcs of: 1, human brain, normal; 2, Cl300 neuroblastoma solid t 1m,or 1; 3, Cl300 neur.)blastoma solid tumor 2. Aliquots of gangliosides corresponding in amount to approximately 80 nmoles of lipid- bound sialic acid were spotted in each channel. The plates were developed in chloroform-methanol-ammonia-water (60:35: 1 :i, v/v) and sprayed with resorcinol reagent. The k~or bad it, Ihe regiotl of GII’ in Fig. 2C: (Channel 4) was not a ganglioside as it did not yield a purple color on spraying with the resorcinol reagent.

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

TABLE II

Glycosphingolipid content of various cells

1235

The results presented are the averages of duplicate analyses on two cultures of each cell type. The amounts of individual GSLs in the cells studied were determined from quantitative gas-liquid chromatographic analyses of their sugar content, using mannitol as an internal standard. The values were calculated from their glucose content; all of the neutral GSLs and gangliosides contained 1 mole of glucose per mole (see Table III).

GSL

Neuroblastoma lines NLI hybrids L Cells

LM(TK-) tumor Cl300 N2A NB41A NA

I I

Neuroblastoma

1SA /

1A I

11A

Gangliosides Total. . . 4.0, 4.2 1.9, 2.3 1.2, 1.4 0.4, 0.6 2.4, 2.9 1.9,2.1 1.9,2.1 0.6, 0.6 GM3 . 2.8, 3.2 0.1, 0.1 0.1, 0.1 n.d.a,b 0.6, 0.9 0.2, 0.2 0.1,O.l 0.1,O.l GM2 . .._...... . 0.4, 0.4 1.7, 1.9 0.3, 0.3 n.d. 1.4, 1.4 1.3, 1.5 1.4, 1.6 0.1,O.l GM1 . . . 0.3, 0.3 0.1, 0.3 0.3, 0.5 n.d. 0.3, 0.5 0.3, 0.3 0.3, 0.3 0.2,0.2 GDI, . 0.5, 0.3 c 0.4, 0.4 n.d. 0.1, 0.1 0.1,O.l 0.1, 0.1 0.1,O.l GDlb. P c 0.1, 0.1 n.d. c c c 0.1,O.l

Neutral GSLs Total. . 6.0, 6.1 6.9, 6.9 7.0, 7.1 9.7, 9.6 15.2, 15.8 5.9, 5.5 15.6, 16.8 1.7, 1.5 Monoglycosylceramide 0.6, 0.8 0.9, 0.9 1.0, 1.3 2.2, 2.2 2.3, 2.7 2.0, 2.2 3.5, 3.7 0.4, 0.4 Diglycosylceramide 0.6, 0.6 0.6, 0.6 0.8, 0.8 6.7, 6.7 2.5, 2.1 1.0,l.l 3.6, 3.8 0.8, 0.6 Triglycosylceramide . 0.1, 0.1 0.7, 0.7 0.2, 0.2 0.8, 0.7 2.0, 2.4 0.5, 0.7 3.2, 3.3 0.1, 0.1 Asialo-GMz. c c c E e c c c c c 0.3, 0.3 c c c c Tetraglycosylceramide. 4.7, 4.6 4.7, 4.7 5.0, 4.8 e c 8.4,8.6 1.1, 1.2 5.3, 6.0 0.4, 0.4

- - a Not determined. b The pattern of L cell gangliosides was similar to that reported previously (21), consisting of approximately equimolar amounts of

GMI, GM2, and GM1. c Less than 0.05 nmoles of GSL per mg of protein.

siderably more GM3 than t’he other two hybrids, but otherwise the patterns were similar. The hybrid cells did, however, differ in their ganglioside patterns from those of both parents. For example, they contained more GM, than the parent NA cells and lesser amounts of the more complex gangliosides. The ganglio- side pattern of the parent L cells is not shown in Fig. 2C. It was, however, similar to that previously reported by our labora- tory for L-929 cells (21) and consisted of approximately equi- molar amounts of GM3, GMQ, and GM1. The ganglioside pat- terns of two of the hybrid cell lines (NLI-7A and NLI-1lA) thus also differed from that of the L cell parent, in that they contained only minor amounts of GM3.

The gangloside patterns of the parent Cl300 tumor grown in

viva in A/J strain mice were also studied (seee Fig. 20, Channels %’ to 3). The tumor was found to contain all of the gangliosides detected in the neuroblasoma lines derived from it, including GMS. Relative to its total ganglioside content, however, the tumor contained a greater preponderance of polysialogangliosides than was observed in the cultured cells, and thus its ganglioside pattern more closely resembled that of human brain (Fig. 20, Channel 1). (The ganglioside pattern of adult mouse brain was also studied and found to be essentially identical with that shown of human brain.)

Neutral Glycosphingolipid Content of Various Cells Studied- The neutral GSLs of the various cells were prepared by silicic acid column chromatography and subjected to analyses by thin layer chromatography. The quantitative and qualitative data obtained from these studies are shown in Table II and Fig. 3. In general, the cultured cells contained a higher amount of total neutral GSL than total ganglioside (expressed as nanomoles of GSL per mg of protein).

The neutral GSL patterns of the different lines of neuroblas- toma cells showed some quantitative differences (e.g. compare results on N2A and NBlA shown in Table II), but tended to

resemble each other much more than did the ganglioside pat- terns. The three subclones of N2A exhibited essentially identi- cal neutral GSL patterns, consisting of double zones with the chromatographic migration of standards of mono-, di-, and tetraglycosylceramides. The “undifferentiated” N2A cells grown in suspension also contained these components (Fig. 3A, Channel I). The glycolipid pattern of the batch of N2A cells shown in this figure exhibited bands corresponding in chroma- tographic migration to standards of mono- and digalactosyl diglyceride. This particular batch of cells was found to be con- taminated with mycoplasma. Glucosyl- and galactosyl-con- taining mono- and diglycosyl diglycerides are known to be the principal glycolipids of certain mycoplasma (26). Other batches of N2A cells grown in suspension and shown to be free of myco- plasma did not exhibit these components and had a pattern essentially identical with that of the N2A subclones shown in Fig. 3A (Channels 2 to 4).

The neuroblastoma line NB41A (Fig. 3B, Channel 1) ex- hibited mono-, di-, and tetraglycosylceramides but also con- tained triglycosylceramide, which was not an appreciable com- ponent of the other neuroblastoma lines.

The analyses of the neutral GSLs of the three hybrid cell lines were of interest. The parent NA cell (chemically detected GSL pattern not shown here, but the pattern as studied by radio- autography is shown in Fig. 4, Channel S) revealed a pattern similar to that of the N2A subclones shown in Fig. 3A (Channels 2 to 4)) containing mono-, di-, and tetraglycosylceramides. The parent L-cell (Fig. 3B, Channel 2) contained mono-, di-, and triglycosylceramides but no tetraglycosylceramide. The three hybrid cells (Fig. 3B, Channels S to 6) were found to contain all four neutral GSLs, including both tri- and tetraglycosylceramides. In addition, line NLI-1lA (Fig. 3B, Channel 4) contained a neutral GSL with similar chromatographic properties to asialo- GMZ.

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

1236

123456 FIG. 3. Thin layer chromatograms of the neutral glycolipids

of the various cells studied. A: neutral glycolipids of: 1, N2A cells in suspension culture (“undifferentiated”), mycoplasma con- taminated; 8, N2A-A5 subclone in monolayer culture (“differen- tiated”) ; 5, N2A-A4 subclone in monolaver culture (“differenti- ated”); 4, N2A-D4 subclone in monolayer culture (“differen- tiated”); 5, a mixture of monoglycosylceramide and sulfatide (migrating just behind triglycosylceramide) from human brain and neutral GSL standards (principally di-, tri-, and tetragly- cosylceramides) from human kidney; 6, a mixture of various glyco- lipid standards including mono- and digalactosyl diglyceride, asialo-GM%, and asialo-GMi. B: neutral glycolipids of: 1, NB41A

CT ASIALO

cl-r5

ASIA I

CD

F; T SIALO

cl-r @!i!

1234567 8 FIG. 4. Radioautogram of a thin layer chromatographic sepa-

ration of the radioactive neutral GSLs of the various cells studied. Radioactive neutral GSLs of: 1, N2A cells; d, NBUA cells; 5, NA cells (the neuroblastoma parent of the hybrid cells) ; 4, LM(TK-) cells (the L cell parent of the hybrid cells) ; 6, NLI-7A hybrid cells; 6, NLI-1lA hybrid cells; 7, NLI-15A hybrid cells; 8, mouse embryo secondary cells. Incubation of cells (approximately 1 to 2 X 106) with n-[1-Wlgalactose was performed as described in “Materials

c-r ASIALO Gf$ cl-r

12 cells; 2, LM(TK-) parent L cells; 9, NLI-11A hybrid cells; 4, NLI-7A hybrid cells; 6, NLI-15A hybrid cells. C: neutral glyco - lipids of: 1. a mixture of monoglycosylceramide from human brain and the neutral GSLs of human kidney; 9, Cl300 neuroblas- toma solid tumor. Aliquots of neutral glycolipid fractions pre- pared by silicic acid column .chromatography corresponding in amount to approximately 500 pg w/w were separated by thin layer chromatography on glass plates coated with Silica Gel G using chloroform-methanol water (65:25:4, v/v) as the developing solvent. The glycolipids were detected as blue zones by spraying with aniline-diphenylamine reagent.

The neutral GSL content of the Cl300 tumor was also studied (Fig. SC, Channel 9). The total amount of neutral GSL in the solid tumor was lower than that of the cultured cells (see Table II). It contained all four neutral GSLs and diglycosylceramide was the most prominent component.

Use of D-[I-Y]Galactose to Study Glycosphingolipid Patferns- A radioautogram of the neutral GSLs extracted from a variety of neurbblastoma cells incubated with n-[lJ%]galactose is shown in Fig. 4. It is evident that all of the cells studied possessed the capacity to synthesize their complement of neutral GSLs using this isotope as a precursor. Double zones corresponding in migration to standards of mono-, di-, tri-, and tetraglycosylcer- amides are clearly visible among the various samples applied. With regard to the hybrid cells, the almost complete absence of triglycosylceramide in the parent NA cell (Channel S) and tetraglycosylceramide in the parent L cell (Channel 4) are evi-

dent. All three hybrid lines are seen to possess the four neutral GSLs,

and in particular contain both tri- and tetraglycosylceramides. The sensitivity of the method is shown by the detection (Channel 6) of the zones corresponding in migration to asialo-GMz. The radioactive zones between tetraglycosylceramide and the origin

and Methods.” Aliquots of the neutral GSL fractions prepared by silicic acid column chromatography and containing approxi- mately 2 X lo4 dpm were separated by thin layer chromatography on a glass plate coated with Silica Gel G in a solvent system of chloroform-methanol-water (65:25:4, v/v). The x-ray plate was in contact with the chromatogram for 1 week. C&f, monoglycosyl- ceramide; CD, diglycosylceramide; CT, triglycosylceramide; CT!!‘, tetraglycosylceramide.

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

A

’ c

E

FIG

. 5.

Gas

-liqui

d ch

rom

atog

raph

ic an

alys

es

of

the

suga

r m

oiet

ies

of s

ever

al

gang

liosid

es

and

neut

ral

GSL

s of

cu

lture

d ne

urob

last

oma

cells

as

tri

met

hyls

ilyl

ethe

rs

of t

heir

O-m

ethy

l gl

ycos

ides

. Th

e ch

rom

atog

raph

ic co

nditio

ns

were

as

des

crib

ed

in “M

ater

ials

and

Met

hods

.” A,

th

e an

alys

is

of

the

gang

liosid

e of

clo

ne

N2A

corre

spon

ding

in

chro

mat

ogra

phic

mig

ratio

n to

GM

ing

to

G

M2;

C

, an

alys

is

of t

he

gang

liosid

e of

clo

ne

NZA

corre

spon

ding

to

G

D,,;

3;

B,

anal

ysis

of

th

e ga

nglio

side

of

clone

NB

41A

corre

spon

d-

D,

anal

ysis

of

th

e ne

utra

l G

SL

of c

lone

NB

41A

corre

spon

ding

to

tri

glyc

osyl

cera

mid

e;

E,

anal

ysis

of

the

ne

utra

l G

SL

of c

lone

NL

I-7A

corre

spon

ding

to

as

ialo

-GM

*; F,

an

alys

is

of t

he n

eutra

l G

SL

of c

lone

NB

41A

car-

resp

ondin

g to

tet

ragl

ycos

ylcer

amid

e (g

lobo

side)

. Th

e pe

aks

are

iden

tifie

d fro

m

left

to

right

as

the

de

rivat

ives

of

: m

ixtur

e of

LY-

and

fl-ga

lact

ofur

anos

ide

(y)

; a-

gala

ctop

yran

osid

e (a

G

AL)

; fi-

gala

ctop

yran

osid

e (0

) ;

gala

ctos

amin

opyr

anos

ide

(GAL

NA

C)

; a-

gluc

opyr

anos

ide

(o

GLU

) ;

@gl

ucop

yran

osid

e @

) ;

man

nito

l (M

AN)

; p-

N-ac

etyl-

inte

rnal

st

anda

rd.

cY-N

-ace

tyl-g

alac

tosa

min

opyr

anos

ide

(GAL

NA

C);

and

N-ac

etyln

eura

min

ic ac

id

(NAN

A).

n-M

annit

ol wa

s th

e Th

e sc

ale

on t

he

absc

issae

is

in de

gree

s.

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

1238

represent small amounts of labeled ganglioside and nucleotide sugars (21). This method was also found to be applicable to qualitative analysis of the gangliosides of similar quantities of cells; an excellent correlation between the pattern of gangliosides as detected by this method and that revealed by chemical de- tection was observed.

Carbohydrate Composition of GSLs Studied-Assignments of general structures to the GSLs studied here were based on analy- ses of the carbohydrate moieties by gas-liquid chromatography. Representative analyses of GSLs corresponding in chromato- graphic migration to GM3, GM2, GDl,, triglycosylceramide, asialo-GM2 and tetraglycosylceramide are shown in Fig. 5, A, B,

C, D, E, and F, respectively. The molar ratios of galactose and galactosamine and sialic acid (where present) normalized to glucose (taken as unity) are shown in Table III. In general, the GSLs of the various cell lines studied exhibited a carbohydrate stoichiometry that might have been anticipated from their chromatographic migration relative to standards of known com- position. The analyses on NLI-7A confirmed the presence of a species with a chemical composition compatible with the struc- ture of asialo-GMz. Of interest was the observation that glucose was virtually the sole carbohydrate in the monoglycosylceramide species of all the cells, only a trace of galactose (approximately 5% of the glucose) being present. The detection of this com- ponent by radioautography (Fig. 4) thus presumably depended on epimerization of galactose to glucose. It should be noted that no distinction between N-acetyl- and N-glycolylneuraminic acids was made in this study, as the gas-liquid chromatographic system used here does not separate them.

TABLE III

Carbohydrate stoichiometry of principal glycosphingolipids

Analyses by gas-liquid chromatography and quantitation of the sugar peaks were performed as described in “Materials and Meth- ods.” Results are averages of duplicate analyses of two prepara- tions of each GSL.

Chromatographic migration of GSL studieda

N2A GM,. ................... GM2. .................. GM1 ................... GD1,, .................

NB41A MC .................... DC .................... TC ..................... TTC ..................

NLI-7A MC .................... DC .................... TC ..................... Asialo-GMz. ............ TTC ..................

LM (TK-) MC .................... DC. ................... TC .....................

--

-

Molar ratio relative to glucose

Glucose

1.00 1.00 1.00 1.00

1.00 1.00 1.00 1.00

1.00 1.00 1.00 1.00 1.00

1.00 1.00 1.00

-

-- Galactose

1.03 0.96 1.04 1.54

Trace 0.93 1.86 1.97

Trace 0.92 1.80 1.04 1.60

1.06 1.83

Galactos amine

0.78 0.89 1.08

0.89

0.78 1.14

0.91 0.89 0.75 2.07

0 MC, DC, TC, TTC: mono-, di-, tri-, and tetraglycosylcer- amide, resnectively.

DISCUSSION

Marked differences were observed in the GSL patterns of the clonal isolates of neuroblastoma cells studied in the present investigation, particularly in the case of the gangliosides (see Fig. 2 and Table II). These variations were unlikely to be mediated by differences in cell density of the various cultures (10, 19, 27), as care was taken to harvest all cultures at approximately the same stage of growth. In contrast to these findings were the ob- servations that subclones of N2A (see Figs. 2 and 3) exhibited essentially similar GSL patterns, as did two of the three hybrid lines of cells. Clonal derivatives of NIL hamster cells (lo), SV40 virus-transformed 3T3 cells (11) and x-ray-induced mouse adrenocortical tumor cells (12) have been previously shown to exhibit variations of GSL profile. The reasons for the existence of such variations of GSL profile between certain related clones of mammalian cells are not understood. They may reflect differ- ences in the activities of synthetases (28, 29) or hydrolases (30) controlling directly the rates of synthesis and degradation of these compounds. It is interesting to note that such variations of GSL profile are not lethal as far as cell growth in culture is concerned, a result consistent with the marked GSL imbalances that are compatible with at least a limited lifespan in the human glycosphingolipidoses. It would seem that GSLs play a more subtle role in cell function than other membrane lipids whose presence is essential for the integrity of certain cellular mem- branes.

While this work was in progress, Dawson et al. (13, 14) re- ported the GSL patterns of cultures derived from NB41A and N2A. In agreement with their results, we found that GM* was the principal ganglioside in NB41A, although, unlike them, we could not detect asialo-GMz in this line. Furthermore, we found GM, to be the principal ganglioside in N2A, contrary to their ob- servations. At present the reasons for these differences are un- known.

The substantial amounts of GM1 in N2A cells and their sub- clones and of tetraglycosylceramide in all the neuroblastoma cells and NLI hybrids were unexpected, particularly as these are minor components of adult brain. The observation that these compounds were present in the parent Cl300 tumor was thus of particular significance, as it apparently dispels the possibility that their appearance in the cultured cells was solely a result of adaptation to cell culture conditions. Although one could speculate that the appearance of GM, and tetraglycosylceramide correlates with the abnormal biological behavior of the Cl300 tumor, this seems unlikely since these two compounds are the principal GSLs of certain mammalian red blood cells (cf. Refer- ence 31).

All of the GSLs observed in the various neuroblastoma lines studied here (with the possible exception of asialo-GMJ have previously been detected in other cultured cells of nonneural origin (32). Thus, the GSL profiles of these cells could not be said to be unique markers for cells of neural origin. It was of interest, however, that virtually none of the GSLs that are enriched in myelin (e.g. galactocerebroside and sulfatide (33)) were detected in the neuroblastoma cells. Galactocerebroside constituted only some 5% of the monoglycosylceramide species. Sulfatide, which migrated just behind triglycosylceramide in the thin layer chromatographic system used (Fig. 3A, Channel 5), was not detected in any of the cells studied. A component of similar migration to sulfatide was observed in the neutral GSL fraction of NLI-7A cells (Fig. 3B, Channel 4) ; however, on analysis this proved to be an asialo-GM2 species. It is possible that incubation of neuroblastoma cells with NaZ5S04 might re-

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

veal the presence of trace amounts of sulfatide. All of the neutral GSLs detected in the neuroblastoma cells resolved as double bands (Fig. 3) ; this phenomenon has been noted in previ- ous studies on the GSLs of cultured cells (11, 21, 34, 25) and is presumably due to the presence of either very long chain normal fatty acids (35) or of hydroxy fatty acids (36) in the slower mi- grating bands.

We did not detect any appreciable change in the profile of the various gangliosides and neutral GSLs of N2A cells before and after neurite outgrowth (see Fig. 2A). Our expectation at the outset of these experiments was that some shift in profile toward a preponderance of higher gangliosides in the “differentiated” state might be observed, akin to the increase of these compounds observed during neuroembryogenesis in brain (2). The present findings do not exclude GSLs from participating in neurite function, as they may have important spatial distributions along the neurite, but they do rule out gross qualitative and quan- titative changes in GSLs as prerequisites for neurite extension.

The studies on the NLI hybrids were of interest, in view of their varying membrane electrogenic potentials and also because no previous investigations of the GSLs of somatic cell hybrids have been reported. The total amount of ganglioside in the hybrid cells was somewhat greater than that of the parent NA cells and much greater than that of the parent L cells. The ganglioside patterns of the hybrid cells were different from those of the parent cells, exhibiting relatively less of the more complex gangliosides than the parent NA cells and relatively less GMI than the parent L cells. The neutral GSL pattern of the hybrid cells was noteworthy as it resembled the sum of the patterns of the two parent cells, exhibiting mono- and diglycosylceramides (present in both parents), triglycosylceramide (present only in L cells) and tetraglycosylceramide (present only in the NA cells). The GSL patterns of the hybrid cells cannot be readily inter- preted in terms of dominance or recessiveness of one or other of the parental cell patterns. This finding may in part reflect the fact that both parental cells had relatively similar and complex patterns of GSLs. Fusion of parental cells with widely dissimilar patterns of GSLs might yield more information on this matter. Further studies on other hybrid cells are thus necessary to permit possible analyses of genetic factors affecting GSL composition. Application of the technique of incubation of cells with I)- [I-‘4Clgalactose followed by radioautographic analysis of their GSLs, as utilized here, may be useful in permitting such studies in which only limited amounts of hybrid cells are available.

The analyses on the hybrid cells did not provide support for any simple relationship between ganglioside composition a.nd electrogenic response, as all three hybrids exhibited very similar patterns of gangliosides. As in the discussion of the studies on neurite outgrowth, the observations here cannot rigorously ex- clude gangliosides from participation in this response, for at least two reasons. Firstly, it is conceivable that small variations in ganglioside composition could be critical in permitting expression of this membrane response. Secondly, highly specific interac- tions of gangliosides with other membrane components could also be necessary for the electrogenic response; these hypothetical components could be deficient in the hybrids of low response. The fact that NB41A cells, which contain principally GM2, ex- hibit a full electrogenic response would argue, however, that the pattern of complex gangliosides observed in adult brain is cer- tainly not critical for expression of this response.

Acknowledgments-We are grateful to Drs. Joanna Olmsted and Dominic Lam for samples of N2A and NB41A, to Ms. Susan

1239

Barrel1 for excellent technical assistance, and to Mr. Don Somers for chromosome determinations on the N2A and NA cells. Funds from the J. P. Bickell Foundation to purchase the gas chromatograph are gratefully acknowledged.

REFERENCES 1. 2.

3. 4. 5. 6.

7.

8.

9.

10.

11.

12.

13.

14.

15. 16.

17. 18.

19. 20.

21.

22.

23.

24. 25.

26. 27.

28.

29.

30. 31.

32.

33.

34.

35.

36.

DERRY, D. M., AND WOLFE, L. S. (1967) Science 168, 1450 VANIER, M. T., HOLM, M., &MAN. R., AND SVENNERHOLM, L.

(1971) J. Neurochem. 18, 581 MCILWAIN, H. (1961) Biochem. J. 78, 24 WOLFE, L. S., AND MCILWAIN, H. (1961) &o&em. J. 78, 33 SVENNERHOLM. L. (1964) J. Likd Res. 6. 145 AUGUSTI-TOCCO, G:, AND SAT;, G. (1989) Proc. Nat. Acacl.

Sci. U. S. A. 64, 311 SCHUBERT, D., HUMPHREYS. S., BARONI, C., AND COHN, M.

0969) Proc. Nat. Acad. Sci. U. S. A. 64,316 HARRIS, A. J., HEINEMANN, S., SCHUBERT, D., AND TARAKIS,

H. (1971) Nature 231. 296 MINNA, J., GLAZER, D., AND NIRENBERG, M. (1972) Nature

New Biol. 236, 225 SAKIYAMA, H., GROSS, S. K., AND ROBBINS, P. W. (1972) Proc.

Nat. Acad. Sci. U. S. A. 69, 872 YOGEESWARAN, G., SHEININ, R., WHERRETT, J. R., AND MUR-

RAY, R. K. (1972) J. Biol. Chem. 247, 51465158 YOGEESWARAN, G., SCHIMMER, B. P., SHEININ, R., AND MUR-

RAY, R. K. (1972) Fed. Proc. 31, 438 DAWSON, G., KEMP, S. F., STOOLMILLER, A. C., AND DORF-

MAN, A. (1971) Biochem. Biophus. Res. Commun. 44. 687 DAWSON, G., STOOLMILLER, A. C., AND KEMP, S. @. (1972)

Trans. Amer. Sot. Neurochem. 3. 68 KLEBE, R. J., AND RUDDLE, F. HI (1969) J. Cell Biol. 43, 69A KIT, S., DUBBS, D. It., PIEKARSKI, L. J., AND Hsu, T. C. (1963)

Exp. Cell Res. 31, 297 LITTLISFIELD, J. W. (1966) Esp. Cell Res. 41, 190 OLMSTICD, J. B., CAELSON, K., KLEBE, R., RUDDLE, F. H.,

AND ROSENBAUM, J. (1970) Proc. Nat. Acad. Sci. U. S. A. 66, 129

HAKOMORI, S. (1970) Proc. Nat. Acad. Sci. U. S. A. 67, 1741 CHATTIBRJEE, S., AND MURRAY, R. K. (1972) Proc. Canad. Fed.

Biol. Sci. 16, 578 YOGEESWARAN, G., WHERRETT, J. I~., CHATTERJEE, S., AND

MURRAY. R. K. (1970) J. Biol. Chem. 246. 6718 WIU~RRETT, J. R., END &JMINGS, J. N. (19%) Biochem. J. 86,

378 SMITH, I. (1960) in Chromatographic and Electrophoretic Tech-

niques, (SMITH, I., ed) Vol. 1, p. 246, Interscience Publishers, New York

VANCE, D. E., AND SWEELEY, C. C. (1967) J. Lipid Res. 8,621 CLAMP, J. R., DAMSON, G., AND HOUGH, L. (1967) Biochim.

Biophys. Acta 148, 342 SMITH, P. F. (1968) Advan. Lipid Res. 6, 69 ROBBINS, P. W., AND MACPHERSON, I. (1971) Proc. Rcy Sot.

177, 49 CUMAR, F. A., BUDY, 1~. O., KOLODNY, E. H., MCFARLAND,

V. W., AND MORA, P. T. (1970) P-oc. Nat. Acad. Sci. U. S. A. 67, 757

DEN, H., SCHULTZ, A. M., BASU, M., AND ROSEMAN, S. (1971) J. Biol. Chem. 246, 2721

GLICIC, M. C. (1971) Fed. Proc. 30, 1044 H~I<OMORI, S. (1965) in The Amino Sugars (BALAZS, E. A.,

AND JEANLOZ, R. W., eds) Vol. IIA, p. 353, Academic Press, New York

HAKOMORI, S. (1971) in Dynamic Structure of Cell Membranes, .22nd Mosbach Colloquium (WALLACH, D. F. H., AND FISCHER, H., eds) p. 65, Springer-Verlag, Berlin

DAVISON, A. N., AND DOBBING, J. (1968) in Applied Newo- chemistry, (DAVISON, A. N., AND DOBBING, J., eds) p. 253, Blackwell Scientific Publications, Oxford, England

KLFNI~, H.-D., AND CHOPPIN, P. W. (1970) Proc. Nat. Acad. Sci. U. S. A. 66, 57

DAWSON, G., MATALON, R.., AND DORFMAN, A. (1972) J. Biol. Chem. 247, 5944

SVENNERHOLM, E., AND SVENNERHOLM, L. (1963) Biochim. Biophys. Acta 70. 432

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from

McMorris and Frank H. RuddleGanesa Yogeeswaran, Robert K. Murray, Mark L. Pearson, B. D. Sanwal, F. Arthur

X L Cell HybridsGlycosphingolipids of Clonal Lines of Mouse Neuroblastoma and Neuroblastoma

1973, 248:1231-1239.J. Biol. Chem. 

  http://www.jbc.org/content/248/4/1231Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/248/4/1231.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on Decem

ber 11, 2020http://w

ww

.jbc.org/D

ownloaded from