42

BBA 52434

Biochimicu et Biophysics Acta 917 (1987) 42-47

Eisevier

Metabolic activities in human skin fihroblasts preloaded

with labeled G ,,-ganglioside

Srinivasa Raghavan, Timothy A. Lyerla *, Allan Krusell and Edwin H. Kolodny

3epurtment of Biochemistry, Eunice Kennedv Shrioer Center for Mentul Returdution, Iw., Wulthnm, MA,

and the Neurology Depurtment, Mussuchusetts General Hospitul und Harvurd Medico1 School, Boston, MA (U.S.A.)

(Received 7 May 1986)

Key words: Ganglioside G,, metabolism; Gangliosidosis; (Human skin fibroblast)

Confluent cultures of human skin fibroblasts were maintained for 10 days with sphingosine labeled

13H1G,. Labeled medium was then replaced with normal medium and the cells maintained for 42 days with weekIy medium changes. Cells were harvested at regular intervals and cells, medium, and trypsin digest supematant analyzed for i3H]G,, and its metabolic products. The ganglioside can be membrane associated and removed by trypsin, or membrane incorporated and trypsin insensitive. The membrane incorporated material is apparently transported to the lysosomes slowly by membrane flow, where 80% of the cellular GM, can be metabolized by day 42. (3H1G, as well as its metabolic products in control cells is continuously released into the m~ium, during which it can also become associated with the cell surface membr~e- There is no detectable metabolism of the 1 3H]G M2 in G M2 g~gliosidosis cell lines over the extended ~st-labeling period, indicating that there is no residual enzyme activity in these cells. Undegraded G,, is continuously’ released into the medium and remains associated with the cell surface membrane as well.

Induction

The neurological disorders Tay-Sachs and Sandhoff disease arise from an inability to hydro- lyze G,,-ganglioside [l] due to a deficiency in the activity of hexosaminidase A (EC 3.2.1.52). The hydrolysis, in vivo, of G,, aiso requires an activa- tor protein f2] whose deficiency leads to the AB

variant neurolo~cal disorder [3]. Hydrolysis of

radio-labeled G M2 in vitro by hexosaminidase A

requires the presence of detergent [4,5] or the activator protein [6] for mediating substrate-en- zyme interaction. Detergent alters the substrate specificity for the isozyme [7-lo], while the activa-

tor is difficult to prepare and available in only a few laboratories [6].

We reported that the difficulties in. using in

* Present address: Department of Biology, Clark University, Worcester, MA 01610, U.S.A.

Nomenclature: The ganglioside nomenclature used here is

that of Svennerholm (J. Neurochem. (1963) 10, 613623)

G M3r I13NeuAc-LacCer; G,,, II’NeuAc-GgOse,Cer; G,,, I13NeuAc-GgOse+Cer.

Correspondence: Dr. Srinivasa Raghavan, Department of Bio-

chemistry, Eunice Kennedy Shriver Center, 200 Trapelo Road,

Waltham, MA 02254, U.S.A.

vitro assays to diagnose G,, gangliosidoses can be circumvented by an in vivo method using cul- tured skin fibroblasts challenged with 3H-labeled G,, [ll]. These studies showed that in normal fibroblasts maintained for 10 days in medium containing 20 PM [3H]GM2, about half the labeled substrate taken up by the cells is metabolized. Cells deficient in hexosaminidase A activity or in activator factor, however, show little hydrolysis of

00052760/87/$03.50 0 1987 Elsevier Science Publishers B.V. (Biomedical Division)

43

the substrate over the same period. We find that

the method is especially helpful in the evaluation

of atypical cases of hexosaminidase deficiency for the definitive diagnosis of G,,-gangliosidosis [12].

In order to follow the long-term kinetics of

complete metabolism of G,,, normal cells were preloaded with [ 3 H]G M2 ganglioside and the cel- lular radioactivity was followed in the absence of

[3Hl%z in the medium for varying periods of time up to 42 days in culture. Radioactivity was

examined in the cells, the media and trypsin

removable cell surface material. Similar studies were performed in cells from patients with G,,- gangliosidosis to determine whether residual G,, metabolism occurs in these mutant cells.

Materials and Methods

Labeling cells with [-‘H]G,, Fibroblasts from controls and patients with

various types of G,, gangliosidoses were first labeled with [ 3H]G,, ganglioside as previously

described [ll]. Briefly, 6 cm dishes were seeded

with 0.5 . lo6 cells and 3 days later culture medium containing about 20 PM [3H]GM2 (specific activ- ity, 48000 cpm/nmole) were added to each dish.

The medium is Dulbecco’s modified Eagle’s medium containing 4.5 g glucose/l and supple- mented with 10% fetal calf serum, 5 pg/ml gentamycin antibiotic, 1 mM pyruvate, 2 mM glutamine, 35 mM NaHCO, and 1 x minimal

essential medium non-essential amino acids. 3H- labeled ganglioside medium is prepared by drying an appropriate volume of a labeled ganglioside solution in chloroform/methanol (2 : 1, v/v) with

a sterile stream of nitrogen, resuspending in the culture medium and sonicating for 30 min in a bath-type sonicator. After 10 days of labeling the

cells were washed free of radioactivity by four successive rinses with phosphate-buffered saline and placed in fresh non-radioactive culture medium without G,, (considered as day 0). Cul- tures were maintained in most cases for another 42 days with weekly medium changes. All sterile work is done in a vertical laminar flow hood and cells are maintained at 37°C in 95% sir/5% CO, under 100% relative humidity. Cells are harvested by trypsinization in 2 ml enzyme medium, the reaction stopped with 0.2 ml fetal calf serum,

pellets collected by centrifugation and rinsed two times in phopshate-buffered saline. They are

analyzed immediately or kept frozen at - 20” C until 3H-labeled lipid analyses are made.

(‘H]glycine incorporation into proteins

Protein synthesis of cell cultures in 6 cm dishes between day 0 and day 42 was determined as

[ 3 Hlglycine uptake into hot trichloroacetic acid

precipitable material. Duplicate cultures on days 0, 21 and 42 were rinsed twice in phopshate-

buffered saline and fresh culture medium was added containing 5 PM [ 3 Hlglycine per dish (New England Nuclear, specific activity 15 Ci/nmol).

Cells were labeled for 4 h at 37°C in 95% sir/5% CO,. [3H]glycine medium was removed, the cul- tures rinsed twice with phosphate-buffered saline and cells collected in 3 ml of cold 10% trichloro- acetic acid with a rubber policeman. They were

allowed to stand for 15 min on ice, then rinsed

twice in cold 10% trichloroacetic acid by centrifu- gation at 1500 X g. The final pellet was resus-

pended in 1.5 ml cold 10% trichloroacetic acid,

heated at 100°C for 15 min and subsequently chilled on ice for 15 min. The hot acid precipitable

material was collected by centrifugation after one rinse in 95% ethanol and another in diethyl ether. The ether was dried under nitrogen and material solubilized in 0.5 ml 0.5 M NaOH overnight at

37°C. Samples were counted in a Packard Tricarb scintillation spectrometer. Total protein was de- termined from a 10 ~1 sample of solubilized pellet

according to the method of Lowry et al. [13].

Extraction and analysis of sphingolipids

Media (10 ml) pooled from duplicate dishes during weekly changes and trypsin digests (4.4 ml) combined from two dishes at the time of harvest- ing cells were extracted overnight with 20 ~01s. of chloroform/methanol (2 : 1, v/v). The extract was filtered through a sintered glass funnel and residue washed with another 5 ~01s. of chloroform/ methanol. The clear filtrate was taken to dryness

in a rotary evaporator with the addition of toluene to remove water present in the lipid extracts. The sample was taken up in 10 ml of chloroform/ methanol and partitioned as follows to remove the labeled ganglioside (approx. 95%) into an aqueous upper phase and to retain the lipid metabolites in

44

the lower organic phase. The extract was mixed with 2 ml of water, vortexed and centrifuged

briefly to separate the phases. The aqueous phase was withdrawn and saved. The lower phase was

washed with 4 ml of theoretical upper phase con-

sisting of chloroform/methanol/water (3 : 48 : 47,

v/v/v) and the resultant upper phase combined

with the first upper phase. The lower phase was taken to dryness, residue redissolved in 10 ml of

chloroform/methanol (2 : 1) and the above parti- tioning procedure was repeated. We found it nec-

essary to perform the partitioning procedure twice to remove more than 95% of labeled G,, from chloroform/methanol extracts of culture medium containing the ganglioside. Due to the high salt concentration of the medium, G,, is not com- pletely removed into the aqueous phase with only cycle of the partitioning procedure. From the

pooled upper phases (17.3 ml) containing labeled ganglioside and the final lower phase (7.2 ml) containing labeled lipid metabolites, 400 ~1

aliquots were taken in small counting vials for

determining radioactivity as previously described [ll]. The ganglioside present in the aqueous phase

was purified on a small BondElut Cl8 cartridge as previously described [ll] and radioactive compo-

nent identified as G,, by thin layer chromatogra- phy [ll]. The total lipids from the lower phase were subjected to alkaline hydrolysis and fractionated on a small silicic acid column by successive elution with chloroform, acetone/

methanol (9 : 1, v/v) and then methanol, and ra- dioactivity in each fraction determined as de- scribed previously [ 111.

The fibroblast cell pellets were sonicated in 200 ~1 water. After removing a sample for protein determination [13], the homogenate was extracted

with chloroform/methanol (2 : 1) and analyzed for radioactivity as previously described [ll].

Results

In order to ensure the cells maintained meta- bolic activity over the long experimental time in culture, [ 3 Hlglycine incorporation into protein was measured between day 0 and day 42. [ ‘Hjglycine uptake into hot trichloroacetic acid precipitable material ranged from 28000 to 53000 cpm/mg protein over the 4 hour labeling period, with lowest

TABLE I

RELEASE OF RADIOACTIVITY INTO THE MEDIUM

FROM CONTROL CELLS PRELOADED WITH LABELED

c MZ DURING THE CHASE PERIOD

The medium collected during weekly changes was counted for

radioactivity. Chloroform/methanol (2 : 1) extracts of media

were partitioned to remove unhydrolyzed G,, into the aque-

ous phase, while retaining the metabolic products in the organic

phase.

Time

Day 0- 7

Day 7-14

Day 14-21

Day 21-28

Day 28-35

Day 35-42

Total O-42

Counts per min Total Percentage in per dish percentage aqueous phase

65 760 36 39 35165 19 39

28 306 15 42 20601 11 28

19140 10 39 14500 8 19

183472

values found on day 0, some 13 days after the cells were originally plated onto 6 cm dishes. Maximal incorporation was found on day 21, but day 42

had higher cpm/mg protein than day 0 (38 000 vs. 28 000), indicating that overall metabolic activity,

as reflected in short-term protein synthesis, was not hampered as cell remained on the dishes for this extended period. No cell disruption was ob-

served during routine examination with a micro-

scope over the course of the experiment. Total cellular protein increased throughout this period

from 0.33 mg on day 0 to 0.42 mg on day 21 and 0.58 mg per dish on day 42,

The continuing metabolic activities of the cells in handling i3 H]G M2 was observed throughout the culture period after removing the labeled ganglio- side from the medium. Radioactivity was released from control cells into the medium during this entire time (Table 1). More than half of the total released radioactivity were recovered during the first and second weeks of the post-labeling period (36 and 19% of the total, respectively), which declined to only 8% of the total released during the last week. The lipid extracts of the medium were partitioned with water to remove the unhy- drolyzed [3H]G,, ganglioside into the aqueous upper phase. The radioactivity in the aqueous phase was present only in gangliosides and none as small water-soluble molecules as determined by

45

TABLE II

RADIOACTIVITY RELEASED IN THE TRYPSIN DI-

GEST WHEN CELLS WERE HARVESTED BY TRYPSINI-

ZATION

For each time point, two dishes were harvested by trypsiniza-

tion and the combined trypsin digests were counted for radio-

activity. Chloroform/methanol (2 : 1) extracts of the trypsin

digests were partitioned to remove unhydrolyzed GM2 into the

aqueous phase, while retaining the metabolic products in the

organic phase.

Time Counts per min Percentage

per dish in aqueous phase

Day 0 37070 X3

Day 21 10813 60

Day 42 6486 31

chromatography on Bond Elut C,, cartridge (111. The unhydrolyzed [3H]G,, in the medium dropped from 39% in the first week to 19% by the last week in the unlabeled medium (Table I), indicating that metabolic products of labeled G,,

were continuously released from the cells into the medium.

Radioactivity in the trypsin-removable cell- surface material also declined steadily throughout this time, but unhydrolyzed [3H]GM2 pre- dominated in this fraction for the first half of the

post-label period (Table II). At the end of 6 weeks in non-labeled medium, almost a third of the

radioactivity recovered in this component were

still unhydrolyzed [ 3H]GMz (Table II).

TABLE III

RADIOACTIVITY IN THE METABOLIC PRODUCTS OF

LABELED GM2 PRESENT IN THE MEDIA AND TRYP-

SIN DIGESTS

Radioactive lipids from the lower organic phase of a washed

lipid extract were fractionated on a silicic acid column.

Chloroform Acetone : methanol methanol

fraction fraction fraction

Media

Day 21 39% Day 42 43%

Trypsin digest

Day 0 35%

Day 21 36%

Day 42 24%

24% 37%

25% 32%

40% 25%

33% 30% 35% 40%

Fractionation of the lipids in the lower organic phase by siiicic acid column chromatography showed that both media and trypsin digest super- natants contained a variety of G,, ganglioside metablic products (Table III). Thin-layer chro-

matography analysis of the chloroform fraction indicated that the radioactivity was in fatty acids

originating from ester linkage in glycerolipids. All of the radioactivity of the methanol fraction was present as sphingomyelin. Cells from variant forms

of G,, gangliosidoses were also examined simi- larly in these same experiments. Like control cells, these cells also continuously released cellular ra-

dioactivity into the medium. However, unlike con-

trols, this radioactivity consisted essentially of

[3H]GM2 ganglioside with very little of its hydro-

lytic products (data not shown). Because of this continuous release of radioac-

tivity into the medium by both control and disease cell lines, the cell-associated radioactivity con-

tinued to decline during the course of the post- labeling period. Nevertheless, at any time point, the percent distribution of remaining intracellular radioactivity between the aqueous and organic phases of washed lipid extract reflected the extent

TABLE IV

G M2 HYDROLYSIS IN SITU IN FIBROBLASTS PRE-

LOADED WITH THIS LABELED GANGLIOSIDE

Confluent cultures of normal human skin fibroblasts were

maintained in the medium containing sphingosine labeled

[3H]G,, ganglioside (20 PM; 48000 cpm/nmol) for 10 days,

After this time, labeled media were replaced with a regular

medium without any ganglioside, and the cultures were main-

tained for an additional 21 and 42 days with periodic changes

of medium. The cells were extracted with chloroform/methanol

(2 : 1) and analyzed for unhydrolyzed ganglioside. Values for

duplicate dishes are given separately for each time point. Day 0

refers to the day on which the cells were changed into non-ra-

dioactive medium.

Days after Protein (mg)

G M2 per dish

removal

0 0.245 0.209

21 0.362

0.398 42 0.495

0.495

Counts Percentage per min aqueous per dish phase

132 170 55 102654 51 69595 33 68418 33 40411 20 47352 21

46

1 0 10 20 30 40

Days

Fig 1. [3H]GM2 metabolism in control and GM2 gangliosido-

sis skin fibroblasts. Cells were labeled for 10 days in [ 3H]Gdvz

when the medium was removed and replaced with unlabeled

medium (day 0). Cells were harvested at day 0, and days 21

and 42 (or 14 and 28) during the post-labeling period. Chloro-

form/methanol (2: 1) extracts of the cells were partitioned

with water to remove u~ydrol~ed GM, into the aqueous

upper phase, while retaining the metabolic products in the

organic lower phase. The values on the ordinate represent

unhydrolyl.ed GM1 in the aqueous phase. All points are aver-

ages of duplicate or triplicate analyses of the same cell line.

Two different control lines were used, and one line for each

G M2 gangliosidosis.0 - 0, Tay-Sachs; m-R, Sandhoff; 0 -0, Juvenile Tay-Sachs; A ------A, Juvenile

Sandhoff; A -A, AB Variant; X __ x , Control (No.

2493); + - +. Control (No. 4).

of the metabolic breakdown of the ingested

13H]GMz ganglioside in the cell (Table Iv). Nor- mal cells effectively metabolize the labeled gang- lioside during the post-label period with only about

20% present as G,, at the end of the experiment.

In fibroblasts from various forms of G,, gang- liosidosis, intracellular radioactivity remained es-

sentially as unhydrolyzed G,, (approx. 90%) at all time points in the post-label period (Fig. 1). The small amount of radioactivity seen in the organic phase from day 0 to day 42 in extracts from these disease cell lines is not due to metabo- lism, but due to the partioning of a small amount

of GM, into the organic phase as determined by thin-layer chromatography (data not shown).

Discussion

In the earlier studies on uptake and metabolism

of 13WG,, g an lioside g by human skin fibrob- lasts, a IO-day labeling period provided a method

to distinguish between those individuals capable of metabolizing G,, and those incapable of hy-

drolyzing this substrate [ll]. Sonderfield et al. [14]

noticed that inco~oration of labeled G,, by

fibroblasts depended strongly on the concentra- tion of fetal calf serum in the culture medium. Optimizing at 0.3% serum concentration with 50

PM GM, in the medium, they have shown that 40% of G,, ingested during 72-92 h was metabolized by control cells, whereas in fibrob- lasts from patients with G,, gangliosidosis, 96% of cellular radioactivity remained as unhydrolyzed G M2'

In the present study, metabolism of ingested

[31-Il%z was followed over an extended period of time after removal of labeled ganglioside from the

medium. Both control and disease cell lines con-

tinuously release radioactive material into the medium for this entire period: quickly at first,

then more slowly towards the end. The released

material from control cells mostly contained metabolic products with smaller amounts of unhy- drolyzed ganglioside. The metabolic products con- susted of esterified fatty acids, sphingomyelin and other neutral sphingolipids. The media collected from disease cell lines contained essentially unhy- drolyzed ganglioside.

In contrast to the results found for medium of control cells, radioactivity in the trypsin remova- ble pericellular surface component of these cells

was present mostly as G,, for the first half of the post-labeling period. These results suggest that

exogenous [ 3 H]G,, is strongly attached to the trypsin-removable surface material. It is ap-

parently not incorporated readily into the plasma membrane for internalization and metabolism by lysosomes in control cells. After incorporation,

radioactivity in the cell associated ganglioside de- clined with time, and there was an increase of its metabolic products. However, cell-associated G M2 was not completely hydrolyzed, suggesting that exogenous G,, incorporated into plasma mem- brane is not directly transported to the lysosomes for complete hydrolysis. Instead, the ganglioside may move slowly within intracellular membranes by membrane flow until its eventual transfer to lysosomes where it can be catabolized. Also, these results suggest that as the [ 3H]sphingosine-labeled G,, was metabolized in control cells, the labeled

products (consisting of ester-linked fatty acids in

glycerolipids and sphingolipids, including sphin- gomyelin) were exocytosed into the medium, when

they could be partly bound to the pericellular surface membrane and found as trypsin remova- ble components. It can be calculated that on day

0, before trypsinization, there are 154482 cpm per dish and at the end of the experiment, day 42,

there are 50368 cpm per dish showing that about 70% of ingested radioactivity is released from the cells during this time. This is recovered in the medium as cpm/dish collected from day 0 to day 42 (Table I).

In the disease cells, very little metabolism of

the ingested labeled G,, was seen in the entire period after removing labeled ganglioside from the medium. This suggests that hexosaminidase B in

Tay-Sachs disease, hexosaminidase S in Sandhoff disease and normal hexosaminidase A without the activator protein in AB variant do not have any

residual slow rate of hydrolytic activity on the

intracellular G,,. It appears that both normal hexosaminidase A and the activator protein are essential to detect even a small amount of metbaolism of G,, in vivo using these methods.

Studies on sulfatide metabolism in fibroblasts from variant forms of metachromatic leuko- dystrophy have shown that there is no residual metabolism in cells from infantile metachromatic leukodystrophy, but cells from juvenile and adult metachromatic leukodystrophy could be dis- tinguished by their residual slow metabolism of the ingested sulfatide [14-171. Conzelmann et al.

[18] have reported that when G,, hydrolysis was assayed in vitro in the presence of activator pro- tein, fibroblasts from adult type G,, gangliosido- sis had measurable residual activity compared to the total deficiency observed in cells from infantile

G,, gangliosidosis. We have also reported that cells from adult-type G M2 gangliosidosis exhibited a small amount of metabolism when preloaded with labeled G,, and maintained for 42 days in

the absence of [ 3H]Gr,,r2 [19]. This observation must be confirmed, however, using cells from several patients with adult type G,, gangliosido- sis. The kinetics of G,, metabolism as described above, should be evaluated carefully in these adult variants to determine if this milder clinical course

47

is explicable using these techniques and, thus,

distinguishable in vivo from severe infantile G,, gangliosidosis.

Acknowledgements

This study was supported by grants HD-05515 and HD-04147 and from the National Institutes of Health grant GM-28207. We thank Denise C. McDonough for her excellent assistance in prepar- ing this manuscript.

References

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

I9

O’Brien, J.S. (1983) in Metabolic Basis of Inherited Dis-

eases (Stanbury, J.B., Wyngaarden. J.B., Fredrickson, D.S.,

Goldstein, J.L. and Brown, M.S., eds.), pp. 945-969, Mc-

Graw-Hill, New York

Conzelmann, E. and Sandhoff. K. (1979) Hoppe-Seylers Z.

Physiol. Chem. 360, 1837-1849

Conzelmann, E. and Sandhoff, K. (1978) Proc. Natl. Acad.

Sci. USA 75, 3979-3983

O’Brien, J.S., Norden, A.G.W., Miller, A.L., Frost, R.G.

and Kelly, T.E. (1977) Clin. Genet. 11, 171-183

Poulos, A., Holding, J. and Carey, W.F. (1982) Clin. Chim.

Acta 108, 361-368

Erzberger. A., Conzelmann, E. and Sandhoff, K. (1980)

Clin. Chim. Acta 108, 361-368

Tallmann, J.F., Brady, R.O.. Quirk, J.M., Villaba, M. and

Gal, A.E. (1974) J. Biol. Chem. 249, 3489-3499

Sandhoff, K., Conzelmann, E. and Nekhorn, H. (1977)

Hoppe-Seylers Z. Physiol. Chem. 358, 779-787

Li, S.-C., Hirabayashi. Y. and Li, Y.-T. (1981) J. Biol.

Chem. 256, 6234-6240

Harzer, K. (1983) Clin. Chim. Acta 135, 89-93

Raghavan, S., Krusell, A., Lyerla. T.A., Bremer, E.G. and

Kolodny. E.H. (1985) Biochim. Biophys. Acta 834, 238-24X

Raghavan, S., Krusell. A., Krusell. J., Lyerla, T.A. and

Kolodny. E.H. (1985) Am. J. Hum. Genet. 37, 1071-1082

Lowry, O.H., Rosebrough, N.J.. Farr, A.L. and Randall,

R.J. (1951) J. Biol. Chem. 193. 265-275

Sonderfeld, S.. Conzelmann, E., Schwarzmann. B.J.,

Hinrichs. U. and Sandhoff, K. (1985) Eur. J. Biochem. 149,

241-255

Porter, M.T., Fluharty. A.L., Harris, SE. and Kihara. H.

(1970) Arch. Biochem. Biophys. 138. 646-652

Kihara, H., Tsay, K.K. and Fluharty, A.L. (1983) Biochem. Med. 29, 278-284

Kudoh, T. and Wenger, D.A. (1982) J. Clin. Invest. 70, 89-97

Conzelmann, E., Kytzia. H.-J., Navon, R. and Sandhoff, K. (1983) Am. J. Hum. Genet. 35, 900-913

Kolodny, E.H. and Raghavan, S. (1983) Trends Neurosci. 6, 16-20