- Email: [email protected]
Low molecular weight thiol content in glutathione synthetase-deficient human fibroblasts
Clink Chimica AC@ 170 (1987) 161-168 Elsevier
161
CCA 03980
Low molecular weight thiol content in glutathione synthetase-deficient human fibroblasts D. Debieu, P.J. Deschavanne and E.P. Malaise Laboratoire
de Radiobiologie
(Received
28 January
Cellulaire, Unit6 INSERM
247, Institut Gustme-Roursy,
1987; revision received 3 July 1987; accepted
Key words: Glutathione;
Glutathione
synthetase;
Human
after revision
fibroblast;
Villejuif (France) 3 August
1987)
Thiol, non-protein-bound
Summary The activity and the kinetic properties of glutathione synthetase and the concentrations of non-protein bound thiols of the y-glutamyl cycle were measured in 11 human fibroblast cell strains. Six of these strains were derived from patients suffering from 5-oxoprolinuria, a recessive genetic disease characterized by a deficiency in glutathione synthetase; the other cell strains were derived from healthy heterozygous or homozygous relatives of the patients. The glutathione synthetase activities of homozygous deficient strains were l/3 of control values while those of heterozygous strains were 2/3 of control values. The total thiol concentration was lower in only 3 of the 6 deficient homozygotes and that of glutathione (GSH) was lower in only 4 of the 6 deficient homozygotes. This lower GSH level was at least partly offset by an accumulation of y-glutamylcysteine, a precursor of GSH, which is almost completely absent from control cells. The total quantities of thiols and GSH in plateau phase cells were about 50% and 30% respectively of the levels in growth phase cells. Approximately 80% of the GSH was in the reduced form in both quiescent and growing cells.
Introduction Glutathione (GSH) is the predominant low molecular weight thiol in mammalian cells [1,2], It is involved in a variety of metabolic processes [l] and it has also been frequently shown to play a role in radioprotection [3]. The function of intracellular
Correspondence to: Dr. Danitle Debieu, Laboratoire I.N.R.A., Route de Saint Cyr, 78000 Versailles, France.
des
0009-8981/87/$03.50
B.V. (Biomedical
0 1987 Elsevier Science Publishers
Fongicides-Station
Division)
de Phytopharmacie,
162 TABLE
I
Genotype
and origin of the cell strains Genotype with regard GSH synthetase
Family GP
Origin
P healthy
SP VP Family AB
to
homozygote
heterozygote deficient homozygote
Kindly supplied by Dr Rev&z, Karolinska Institutet, Stockholm Sweden [18]
B healthy
homozygote
Kindly supplied by Dr BouC, Centre International de 1’Enfance (Inserm U.73). Paris, France (AB, OB, EB) [17]
RB JB OB EB
heterozygote heterozygote deficient homozygote deficient homozygote
Family Z AZ
deficient
homozygote
Kindly supplied by Dr Ogier, Hapital Necker Paris, France.
GM3877
deficient
homozygote
GM3878
deficient
homozygote
Kindly supplied by Dr Astor, Columbia University, New York [13,21]
GSH may be studied in cells in which the level of GSH has been lowered using chemical agents [3], but use of such agents entails the risk of producing side effects [3,4]. One way of avoiding such effects is to employ cell strains having a genetic GSH deficiency. The recessive genetic disease, 5oxoprolinuria, is characterized by a deficiency of glutathione synthetase [5-71. We carried out a systematic study of the effects of a glutathione synthetase deficiency on the concentrations of the various thiols of the y-glutamyl cycle in fibroblasts originating from patients and from healthy relatives from 4 families. The study comprised measurements of glutathione synthetase activities, followed by measurements of GSH concentrations, together with those of two of its precursors, cysteine and y-glutamylcysteine, and one of its degradation products: cysteinylglycine. A total of 11 cell strains were examined, 9 of them were deficient in glutathione synthetase (homozygotes and heterozygotes) and two were control strains (Table I). Studies on cancer cell lines have suggested that the kinetics of in vitro proliferation may influence thiol concentrations [3,8]. Non-transformed fibroblasts in plateau phase may be considered as a good in vitro model for normal tissues having a low in vivo turnover rate [9]. We therefore compared the thiol levels of exponentially growing cells to those of plateau’phase cells. As the majority of the cellular functions of GSH are performed by reduced GSH [l], the thiol status of all the cell strains was systematically studied.
163
Materials and methods Cell strains Eleven strains of non-transformed human fibroblasts were derived from skin biopsy of patients suffering from the recessive inherited disease 5oxoprolinuria and of their relatives (Table I). Monolayer cultures were maintained in Minimum Essential Medium (Earle’s salts) supplemented with 15% fetal calf serum in plastic flasks at 37°C in an atmosphere of 95% air and 5% CO,. Depending on the experiment, cells were used when in the exponential growth phase or in unfed plateau phase. Unfed plateau phase cells were obtained 10 days after plating out 5 X 10’ cells per 25 cm2 plastic flask. The medium was not renewed during the lo-day incubation period. Cells were dispersed with Ca-free saline solution containing 0.1% trypsin and 0.04% EDTA. Glutathione synthetase characteristics and Thiol concentrations The glutathione synthetase activity and the apparent kinetic parameters (Michaelis constant (K,) were determined as described by Wellner et al [7]. Low molecular weight thiol extraction, as previously described by Debieu et al [lo], was carried out in the presence or absence of dithiothreitol, to measure thiols either in reduced and oxidized forms or only in reduced form. Thiol concentrations were determined by reverse-phase high performance liquid chromatography as described by Deschavanne et al [ll]. The y-glutamylcysteine was prepared from glutatbione disulfide by treatment with carboxypeptidase A and subsequent reduction with ditbiothreitol. Cystinylglycine was obtained by dithiothreitol reduction of cystinylglycine. Protein content was measured by the method of Lowry et al [12]. Results and discussion I. Fibroblasts in exponential growth phase Glutathione synthetase activity was genotype-dependent in the 9 cell strains tested (Table II). The activity varied from 19 to 31% of mean control values in homozygous deficient strains. This result is in agreement with those obtained with other cell types [5,7,13,14] and with the generally accepted concept of a all tissues glutathione synthetase deficiency associated with 5oxoprolinuria [5,7,15,16]. The heterozygous cell strains had an activity intermediate between those of controls and of deficient homozygotes ranging between 50% and 72% of control values. This is in agreement with earlier findings for erythrocytes [7,13,17,18] and for some fibroblast strains examined [13,19]. The OB, EB and AZ strains present a special problem because, although they have similar glutathione synthetase activities, their GSH levels differ. To explain these apparently contradictory results, the apparent Michaelis constants (K,) of the enzymes were measured in the three strains using glycine and y-glutamylcysteine analogue, y-glutamylaminobutyrate, as substrates. synthetase from AZ cells were 10 times The apparent K, values of the glutathione (glycine) and 100 times ( y-glutamylaminobutyrate) greater than control values (Table III). The apparent K, values of the enzyme from OB cells was similar to
II
(2.1)
(4.8)
b
(0.4) 1.19 26.6 (4.1)
(0.4) 0.80 20.8 (2.2)
3.7
(0.1) -
(0.7)
(0.5) 0.65 24.0
3.7
(0.1)
(0.9) 0.87 0.5
9.9 19.8
OB
(5.3)
(1.7) 0.57 38.7
8.8
(0.2)
(4.6) 0.81 1.2
10.9 28.7
EB
(1.4)
(1.3) 0.46 11.6
(0.1) 0.82 4.6
(0.3) 0.80 3.9
9.1 3.1
AZ
patients
levels for fibroblasts
(2.2)
(0.5) 0.52 19.2
(1.9) 1.00 4.3
(1.7) 0.66 5.9
6.4 9.0
VP
phase
(2.4)
(0.7) n.d. 20.9
(4.6)
(W 0.48 34.6
(3.1) 0.72 5.7
(0.5) 0.65 24.4
(0.4) n.d. 11.8 (1.9) n.d. 6.9
n.d. a 2.1
GM3878
growing
n.d. 2.2
GM3877
in exponentially
a n.d. = not done. For GM3878, glutathione synthetase activity of confluent fibroblasts was found to be 6 nmol/h per mg protein [13]. b Fractions in reduced form = reduced/total. The fractions were not calculated when the level of thiol was below the detection limit (0.1 nmol/mg protein). ’ Total SH = sum of the y-ghttamyl cycle thiols (reduced and oxidized). Enzyme activity and thiol concentrations are expressed as nmoI/h per mg protein and as nmol/mg protein, respectively. The SD of the mean is given in parentheses. Values are the means of either 3 replicate experiments (enzyme activity and thiol levels) or 2 replicate experiments (fractions of reduced form).
Reduced cysteine/total Total SH = (1.1)
(0.5) n.d. 27.1
(%) 0.49 28.3
cysteine/total (2.3) 0.91 26.6
y-gfutamyl
(B:) -
Reduced Cysteine
(0.1) 7.6
b
(3.7) 0.74 0.1
(1.8) 0.65 0.0 2.1
(1.5) nd. 0.4 (0.1) n.d. 5.3
(1.7) 0.85
17.0 22.8
25.1 18.7
24.0 21.4
SP
JB
RB
forms)
5-Oxoprolinuria
and oxidized
relatives
(in reduced
Heterozygous
(2.5) 1.14 0.4
GP
controls
and cysteine
Reduced gJutathione/total y-Ghrtamylcysteine
AB
Related
y-ghttamylcysteine
34.0 23.0
b
and GSH,
35.5 18.6
activity
GSH synthetase GSH
Cell strains
GSH-synthetase
TABLE
165 TABLE K,
III
of the glutathione
synthetase
for glycine
and L-y-glutamyl-a-aminobutyrate
Controls
K, K,
0.7
0.4
0.33
2.0
glycine (mmol/l) y-glutamyl aminobutyrate (mmol/l)
a Control
cell strains
were not derived
a
OB
Cell strains
from relatives
of 5-oxoprolinuria
EB
AZ 10
2.0
20
patients.
controls with glycine as substrate, but the apparent K, value of both OB and EB cell strains were 10 times greater than normal with y-glutamylaminobutyrate. These increases in apparent K,, reflecting a loss of enzyme-substrate affinity, are much more marked for AZ than for the other strains and may, thus, explain the differences in GSH concentrations in the three strains. The glutathione synthetase activities may appear similar when measured in cell extracts with excess substrate while corresponding to very different GSH-synthesizing capacities when the cells are growing in culture. A 50 times higher K, was found for glycine in GM3878 confluent cells [13] while Larsson et al [18] found no change in the apparent K, of VP erythrocyte enzyme against its substrates, likely reflecting a lower amount of glutathione synthetase. All these findings (lower amount of enzyme and/or enzyme kinetic properties affected to varying degrees in different cell strains) strongly suggest different mutations leading to 5-oxoprolinuria. Thiol concentrations, including both oxidized and reduced forms, are shown in Table II. The GSH concentrations in heterozygous cells were similar to those of control cells, in which GSH represents 70-81% of the overall thiol content, The homozygous glutathione synthetase deficient cell strains fell into two groups: family B in which the GSH levels were similar to those of control cells and the 4 other cell strains in which the levels were lower (ll-39% of control values) (Table II). The y-glutamylcysteine concentration was negligible in the control and in heterozygous strains (< 1% of their thiols). However, the deficient homozygous cells with lower GSH level accumulated y-glutamylcysteine, so that it accounted for 31-71% of their total thiol content. Cysteine level appeared to be independent of genotype, being between 10 and 29% of the thiol control values. Cysteinylglycine was detected in only 2 cell strains, one control (AB) and one homozygous deficient strain (GM3878) and only in small quantity: 1.7 and 1.2 nmol/mg protein respectively (i.e. 6.5% and 3% of their thiol content). Of the 9 glutathione synthetase deficient cell strains (3 heterozygous and 6 homozygous), the total concentration of thiols was below control values in only 3 homozygous strains (Table II). In all the cell strains, GSH was predominantly in the reduced form, regardless of cell genotype (Table II). The fraction of reduced cysteine was between 46% and 100%. The y-giutamylcysteine, when measurable, was largely in the reduced form (72-100%) (Table II). It seems that low fibroblast GSH levels are accompanied by increases in y-glutamylcysteine. This accumulation of y-glutamylcysteine probably results from a decrease of feedback inhibition on y-glutamylcysteine synthetase
IV
b
b
’
(0.1) nd. 4.9 (1.4) nd. 16.2 (0.6)
(0.3) _ (1.3) 0.53 11.7 (2.3)
5.5
n.d. 0.1
(1.2)
(0.5) 0.85 0.5
11.2
4.3 a
(0.3)
(0.5) 0.25 14.6
7.6
0.74 0.5 (0.01)
(0.6)
5.7
(2.3)
(0.4) 0.32 18.4
6.2
0.72 0.1 (0.04)
(2.1)
12.1
JB
RB
AB
GP
Heterozygous relatives
Related controls
(1.2)
(0.5) 0.29 8.3
4.0
(0.5)
(0.3) 0.67 0.4 (0.04)
(0.5)
(0.8) 0.34 6.8
1.9
(0.1) _
0.72 0.3
4.6
OB
2.7
SP
(0.8)
(0.4) 0.39 12.9
(0.1) _ 5.5
(3.3)
(1.7) 0.25 14.1
(0.8) 0.66 7.5
(1.1) 0.76 3.3
(1.4) 1.02 0.7
AZ 3.3
patients
6.2
EB
5-Oxoprolinuria
-
(0.7)
(0.5) n.d. 8.9
(0.6) n.d. 3.3
(0.9) n.d. 3.1
2.5
VP
’ Total SH = sum of y-glutamyl cycle thiols. Values are the means of either 3 replicate experiments (thiol levels) or 2 replicate experiments (fractions of reduced form); 2 exceptions: RB and 1 for AZ. n.d.. not done.
limit. 4 experiments
for
’ Thiol concentration is expressed as nmol/mg protein. The SD of the mean is given in parentheses ’ Fractions in reduced form (reduced/total) for fibroblasts in unfed plateau phase. They were not calculated when the level of thtol was below the detection
Reduced cysteine/total Total SH ’
Reduced ghttamylcysteine/total Cysteine
Reduced glutathione/total y-Glutamylcysteine
Glutathione
Cell strains
Levels of GSH, y-ghrtamylcysteine and cysteine (in reduced and oxidized forms) for fibroblasts in unfed plateau phase
TABLE
because of the lower GSH concentration [1,5]. In contrast, feedback inhibition should be normal in the two homozygous deficient strains in which the GSH level was not lowered and the quantity of y-glutamylcysteine was very low. The heterozygotes have GSH and y-glutamylcysteine concentrations which were similar to those of controls. The different mutations of the glutathione synthetase deficiency are then reflected in the concentrations of the various thiols, each of which may be affected differently in a given mutant strain. The sum of the individual thiol concentrations measured by high performance liquid chromatography is in good agreement with values obtained by measuring the overall thiol content calorimetrically [lo]. This would indicate that the thiols of the y-glutamyl cycle represent almost the overall thiol content in control human fibroblasts. The present study has also shown that the percentage of reduced thiols is unaffected by a glutathione synthetase deficiency. 2. Fibroblasts in unfed plateau phase The transition from growth phase to plateau phase was marked by a decrease in GSH level in 8 of the 9 strains studied (Table Iv>. The extent of the decrease in GSH varied from one strain to another (Table IV), independently of genotype. The only one of the two strains which were studied in the plateau phase and which had an appreciable y-glutamylcysteine level in the growth phase showed also a decrease in y-glutamylcysteine level. In contrast to the change in GSH, the transition from growth to plateau did not appear to induce a systematic decrease in cysteine level. Cysteinylglycine was present at low level (0.8-1.4 nmol/mg protein, e.g. 6-15% of the thiol content) in 4 cell strains (AB, RB, SP, EB) in plateau phase. The overall thiol level for each cell strain (except for AZ) in plateau phase was about 50% of that of growing cells. The level of reduced GSH in plateau phase cells represented about the same percentage of total GSH as in the growth phase (72-85%) (Table Iv) and appeared to be independent of genotype. The level of reduced cysteine in plateau phase cells was generally below 50% except in control cells (Table Iv>. The lowered thiol level in fibroblasts in the plateau phase found in this study follows similar pattern to that reported for cancer cells [3,8]. GSH is the thiol which is most involved in this phenomenon. The decrease in GSH may result from a lower metabolism in plateau phase cells [9]. Genetically deficient cell strains which differ not only in their capacity to synthesize GSH, but also in their thiol profile would appear to constitute a good biological model for the study of the cellular function of GSH and the other thiols. They provide an alternative to the use of depleting agents [3] and have already been successfully employed in a number of studies, dealing with radiobiology [14,10,20-221, amino acid active transport [5] and detoxification [23]. Acknowledgements The authors wish to acknowledge Dr. A. Larsson and his group for their help in glutathione synthetase studies and Mrs. J. Encinas for typing the manuscript.
168
This work was supported in part by ‘La Ligue Nationale Franqaise Contre le Cancer’ (ComitC des Hauts-de-Seine) and by contract BOO L05/3K 3237 from ‘Le Comitt de Radioprotection de 1’ElectricitC de France’. References 1 Meister A, Anderson ME. Ann Rev Biochem 1983;52:711-760. 2 Fahey RC, Newton GL. Functions of glutathione biochemical, physiological, toxicological and clinical aspects. In: Larsson et al, eds. New York: Raven Press, 1983;251-260. 3 Biaglow JE, Vames ME, Clarck EP, Epp ER. Radiat Res, 1983;95:447-455, 4 Harris JW, Biaglow JE. Biochem Biophys Res Commun 1972;46:1743-1749. 5 Larsson A. Transport and inherited disease. In: Belton NR, Toothill C, eds. Lancaster/Boston/The Hague: MTP Press Ltd. 1981;277-306. 6 Boivin P, Galand C. Nouv Rev Fr Hematol 1965;5:707-720. 7 Wellner VP, Sekura R, Meister A, Larsson A. Proc Nat Acad Sci USA 1974;71:2505-2509. 8 Cullen BM, Michalowski A, WaIker HC, Revtsz L. Int J Radiat Biol 1980;38:525-535. 9 Hahn GM, Little JB. Curr Top Radial Res Quart 1972;8:39-83. 10 Debieu D, Deschavanne PJ, Midander J, Larsson A, Malaise EP. Int J Radiat Biol 1985;48:525-543. 11 Deschavanne PJ, Midander J, Debieu D, Malaise EP, RCvCsz L. Int J Radiat Biol 1986;49:85-101. 12 Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Biol Chem 1951;193:265-275. 13 Spielberg SP, Kramer LI, Goodman SI, Butler J, Tietze F, Quinn P, Schulman JD. J Pediat 1977;91:237-241. 14 Edgren M, Larsson A, Nilsson K, Rev&z L, Scott OCA. Int J Radiat Biol 1980;37:299-306. 15 Jellum E, Marstein S, Skullerud K, Munthe E. Functions of glutathione: biochemical, physiological, toxicological and clinical aspects. In: Larsson et al, eds. New York: Raven Press, 1983;347-352. 16 Spielberg SP, Garrick MD, Corash LM, Butler J, Tietze F, Rogers L, Schulman JD. J Clin Invest 1978;61:1417-1420. 17 Boivin P, Galand C, Schaison G. Nouv Presse MM 1978:7:1531-1535. 18 Larsson A, Zetterstrom R, Hornell H, Porath U. Clin Chim. Acta 1976;73:19-23. 19 Larsson A, Mattsson B, Hagenfeldt L, Moldeus P. Clin Chim Acta 1983;135:57-64. 20 Rev&z L, Malaise EP. Functions of glutathione: biochemical, physiological, toxicological and clinical aspects In: Larsson et al, eds. New York: Raven Press, 1983;163-173. 21 Astor MB. Br J Radio1 1984;57:717-722. 22 Malaise EP. Radiat Res 1983;95:486-494. 23 Spielberg SP, Gordon GB. Clin Pharmacol Ther 1981;29:51-55.
161
CCA 03980
Low molecular weight thiol content in glutathione synthetase-deficient human fibroblasts D. Debieu, P.J. Deschavanne and E.P. Malaise Laboratoire
de Radiobiologie
(Received
28 January
Cellulaire, Unit6 INSERM
247, Institut Gustme-Roursy,
1987; revision received 3 July 1987; accepted
Key words: Glutathione;
Glutathione
synthetase;
Human
after revision
fibroblast;
Villejuif (France) 3 August
1987)
Thiol, non-protein-bound
Summary The activity and the kinetic properties of glutathione synthetase and the concentrations of non-protein bound thiols of the y-glutamyl cycle were measured in 11 human fibroblast cell strains. Six of these strains were derived from patients suffering from 5-oxoprolinuria, a recessive genetic disease characterized by a deficiency in glutathione synthetase; the other cell strains were derived from healthy heterozygous or homozygous relatives of the patients. The glutathione synthetase activities of homozygous deficient strains were l/3 of control values while those of heterozygous strains were 2/3 of control values. The total thiol concentration was lower in only 3 of the 6 deficient homozygotes and that of glutathione (GSH) was lower in only 4 of the 6 deficient homozygotes. This lower GSH level was at least partly offset by an accumulation of y-glutamylcysteine, a precursor of GSH, which is almost completely absent from control cells. The total quantities of thiols and GSH in plateau phase cells were about 50% and 30% respectively of the levels in growth phase cells. Approximately 80% of the GSH was in the reduced form in both quiescent and growing cells.
Introduction Glutathione (GSH) is the predominant low molecular weight thiol in mammalian cells [1,2], It is involved in a variety of metabolic processes [l] and it has also been frequently shown to play a role in radioprotection [3]. The function of intracellular
Correspondence to: Dr. Danitle Debieu, Laboratoire I.N.R.A., Route de Saint Cyr, 78000 Versailles, France.
des
0009-8981/87/$03.50
B.V. (Biomedical
0 1987 Elsevier Science Publishers
Fongicides-Station
Division)
de Phytopharmacie,
162 TABLE
I
Genotype
and origin of the cell strains Genotype with regard GSH synthetase
Family GP
Origin
P healthy
SP VP Family AB
to
homozygote
heterozygote deficient homozygote
Kindly supplied by Dr Rev&z, Karolinska Institutet, Stockholm Sweden [18]
B healthy
homozygote
Kindly supplied by Dr BouC, Centre International de 1’Enfance (Inserm U.73). Paris, France (AB, OB, EB) [17]
RB JB OB EB
heterozygote heterozygote deficient homozygote deficient homozygote
Family Z AZ
deficient
homozygote
Kindly supplied by Dr Ogier, Hapital Necker Paris, France.
GM3877
deficient
homozygote
GM3878
deficient
homozygote
Kindly supplied by Dr Astor, Columbia University, New York [13,21]
GSH may be studied in cells in which the level of GSH has been lowered using chemical agents [3], but use of such agents entails the risk of producing side effects [3,4]. One way of avoiding such effects is to employ cell strains having a genetic GSH deficiency. The recessive genetic disease, 5oxoprolinuria, is characterized by a deficiency of glutathione synthetase [5-71. We carried out a systematic study of the effects of a glutathione synthetase deficiency on the concentrations of the various thiols of the y-glutamyl cycle in fibroblasts originating from patients and from healthy relatives from 4 families. The study comprised measurements of glutathione synthetase activities, followed by measurements of GSH concentrations, together with those of two of its precursors, cysteine and y-glutamylcysteine, and one of its degradation products: cysteinylglycine. A total of 11 cell strains were examined, 9 of them were deficient in glutathione synthetase (homozygotes and heterozygotes) and two were control strains (Table I). Studies on cancer cell lines have suggested that the kinetics of in vitro proliferation may influence thiol concentrations [3,8]. Non-transformed fibroblasts in plateau phase may be considered as a good in vitro model for normal tissues having a low in vivo turnover rate [9]. We therefore compared the thiol levels of exponentially growing cells to those of plateau’phase cells. As the majority of the cellular functions of GSH are performed by reduced GSH [l], the thiol status of all the cell strains was systematically studied.
163
Materials and methods Cell strains Eleven strains of non-transformed human fibroblasts were derived from skin biopsy of patients suffering from the recessive inherited disease 5oxoprolinuria and of their relatives (Table I). Monolayer cultures were maintained in Minimum Essential Medium (Earle’s salts) supplemented with 15% fetal calf serum in plastic flasks at 37°C in an atmosphere of 95% air and 5% CO,. Depending on the experiment, cells were used when in the exponential growth phase or in unfed plateau phase. Unfed plateau phase cells were obtained 10 days after plating out 5 X 10’ cells per 25 cm2 plastic flask. The medium was not renewed during the lo-day incubation period. Cells were dispersed with Ca-free saline solution containing 0.1% trypsin and 0.04% EDTA. Glutathione synthetase characteristics and Thiol concentrations The glutathione synthetase activity and the apparent kinetic parameters (Michaelis constant (K,) were determined as described by Wellner et al [7]. Low molecular weight thiol extraction, as previously described by Debieu et al [lo], was carried out in the presence or absence of dithiothreitol, to measure thiols either in reduced and oxidized forms or only in reduced form. Thiol concentrations were determined by reverse-phase high performance liquid chromatography as described by Deschavanne et al [ll]. The y-glutamylcysteine was prepared from glutatbione disulfide by treatment with carboxypeptidase A and subsequent reduction with ditbiothreitol. Cystinylglycine was obtained by dithiothreitol reduction of cystinylglycine. Protein content was measured by the method of Lowry et al [12]. Results and discussion I. Fibroblasts in exponential growth phase Glutathione synthetase activity was genotype-dependent in the 9 cell strains tested (Table II). The activity varied from 19 to 31% of mean control values in homozygous deficient strains. This result is in agreement with those obtained with other cell types [5,7,13,14] and with the generally accepted concept of a all tissues glutathione synthetase deficiency associated with 5oxoprolinuria [5,7,15,16]. The heterozygous cell strains had an activity intermediate between those of controls and of deficient homozygotes ranging between 50% and 72% of control values. This is in agreement with earlier findings for erythrocytes [7,13,17,18] and for some fibroblast strains examined [13,19]. The OB, EB and AZ strains present a special problem because, although they have similar glutathione synthetase activities, their GSH levels differ. To explain these apparently contradictory results, the apparent Michaelis constants (K,) of the enzymes were measured in the three strains using glycine and y-glutamylcysteine analogue, y-glutamylaminobutyrate, as substrates. synthetase from AZ cells were 10 times The apparent K, values of the glutathione (glycine) and 100 times ( y-glutamylaminobutyrate) greater than control values (Table III). The apparent K, values of the enzyme from OB cells was similar to
II
(2.1)
(4.8)
b
(0.4) 1.19 26.6 (4.1)
(0.4) 0.80 20.8 (2.2)
3.7
(0.1) -
(0.7)
(0.5) 0.65 24.0
3.7
(0.1)
(0.9) 0.87 0.5
9.9 19.8
OB
(5.3)
(1.7) 0.57 38.7
8.8
(0.2)
(4.6) 0.81 1.2
10.9 28.7
EB
(1.4)
(1.3) 0.46 11.6
(0.1) 0.82 4.6
(0.3) 0.80 3.9
9.1 3.1
AZ
patients
levels for fibroblasts
(2.2)
(0.5) 0.52 19.2
(1.9) 1.00 4.3
(1.7) 0.66 5.9
6.4 9.0
VP
phase
(2.4)
(0.7) n.d. 20.9
(4.6)
(W 0.48 34.6
(3.1) 0.72 5.7
(0.5) 0.65 24.4
(0.4) n.d. 11.8 (1.9) n.d. 6.9
n.d. a 2.1
GM3878
growing
n.d. 2.2
GM3877
in exponentially
a n.d. = not done. For GM3878, glutathione synthetase activity of confluent fibroblasts was found to be 6 nmol/h per mg protein [13]. b Fractions in reduced form = reduced/total. The fractions were not calculated when the level of thiol was below the detection limit (0.1 nmol/mg protein). ’ Total SH = sum of the y-ghttamyl cycle thiols (reduced and oxidized). Enzyme activity and thiol concentrations are expressed as nmoI/h per mg protein and as nmol/mg protein, respectively. The SD of the mean is given in parentheses. Values are the means of either 3 replicate experiments (enzyme activity and thiol levels) or 2 replicate experiments (fractions of reduced form).
Reduced cysteine/total Total SH = (1.1)
(0.5) n.d. 27.1
(%) 0.49 28.3
cysteine/total (2.3) 0.91 26.6
y-gfutamyl
(B:) -
Reduced Cysteine
(0.1) 7.6
b
(3.7) 0.74 0.1
(1.8) 0.65 0.0 2.1
(1.5) nd. 0.4 (0.1) n.d. 5.3
(1.7) 0.85
17.0 22.8
25.1 18.7
24.0 21.4
SP
JB
RB
forms)
5-Oxoprolinuria
and oxidized
relatives
(in reduced
Heterozygous
(2.5) 1.14 0.4
GP
controls
and cysteine
Reduced gJutathione/total y-Ghrtamylcysteine
AB
Related
y-ghttamylcysteine
34.0 23.0
b
and GSH,
35.5 18.6
activity
GSH synthetase GSH
Cell strains
GSH-synthetase
TABLE
165 TABLE K,
III
of the glutathione
synthetase
for glycine
and L-y-glutamyl-a-aminobutyrate
Controls
K, K,
0.7
0.4
0.33
2.0
glycine (mmol/l) y-glutamyl aminobutyrate (mmol/l)
a Control
cell strains
were not derived
a
OB
Cell strains
from relatives
of 5-oxoprolinuria
EB
AZ 10
2.0
20
patients.
controls with glycine as substrate, but the apparent K, value of both OB and EB cell strains were 10 times greater than normal with y-glutamylaminobutyrate. These increases in apparent K,, reflecting a loss of enzyme-substrate affinity, are much more marked for AZ than for the other strains and may, thus, explain the differences in GSH concentrations in the three strains. The glutathione synthetase activities may appear similar when measured in cell extracts with excess substrate while corresponding to very different GSH-synthesizing capacities when the cells are growing in culture. A 50 times higher K, was found for glycine in GM3878 confluent cells [13] while Larsson et al [18] found no change in the apparent K, of VP erythrocyte enzyme against its substrates, likely reflecting a lower amount of glutathione synthetase. All these findings (lower amount of enzyme and/or enzyme kinetic properties affected to varying degrees in different cell strains) strongly suggest different mutations leading to 5-oxoprolinuria. Thiol concentrations, including both oxidized and reduced forms, are shown in Table II. The GSH concentrations in heterozygous cells were similar to those of control cells, in which GSH represents 70-81% of the overall thiol content, The homozygous glutathione synthetase deficient cell strains fell into two groups: family B in which the GSH levels were similar to those of control cells and the 4 other cell strains in which the levels were lower (ll-39% of control values) (Table II). The y-glutamylcysteine concentration was negligible in the control and in heterozygous strains (< 1% of their thiols). However, the deficient homozygous cells with lower GSH level accumulated y-glutamylcysteine, so that it accounted for 31-71% of their total thiol content. Cysteine level appeared to be independent of genotype, being between 10 and 29% of the thiol control values. Cysteinylglycine was detected in only 2 cell strains, one control (AB) and one homozygous deficient strain (GM3878) and only in small quantity: 1.7 and 1.2 nmol/mg protein respectively (i.e. 6.5% and 3% of their thiol content). Of the 9 glutathione synthetase deficient cell strains (3 heterozygous and 6 homozygous), the total concentration of thiols was below control values in only 3 homozygous strains (Table II). In all the cell strains, GSH was predominantly in the reduced form, regardless of cell genotype (Table II). The fraction of reduced cysteine was between 46% and 100%. The y-giutamylcysteine, when measurable, was largely in the reduced form (72-100%) (Table II). It seems that low fibroblast GSH levels are accompanied by increases in y-glutamylcysteine. This accumulation of y-glutamylcysteine probably results from a decrease of feedback inhibition on y-glutamylcysteine synthetase
IV
b
b
’
(0.1) nd. 4.9 (1.4) nd. 16.2 (0.6)
(0.3) _ (1.3) 0.53 11.7 (2.3)
5.5
n.d. 0.1
(1.2)
(0.5) 0.85 0.5
11.2
4.3 a
(0.3)
(0.5) 0.25 14.6
7.6
0.74 0.5 (0.01)
(0.6)
5.7
(2.3)
(0.4) 0.32 18.4
6.2
0.72 0.1 (0.04)
(2.1)
12.1
JB
RB
AB
GP
Heterozygous relatives
Related controls
(1.2)
(0.5) 0.29 8.3
4.0
(0.5)
(0.3) 0.67 0.4 (0.04)
(0.5)
(0.8) 0.34 6.8
1.9
(0.1) _
0.72 0.3
4.6
OB
2.7
SP
(0.8)
(0.4) 0.39 12.9
(0.1) _ 5.5
(3.3)
(1.7) 0.25 14.1
(0.8) 0.66 7.5
(1.1) 0.76 3.3
(1.4) 1.02 0.7
AZ 3.3
patients
6.2
EB
5-Oxoprolinuria
-
(0.7)
(0.5) n.d. 8.9
(0.6) n.d. 3.3
(0.9) n.d. 3.1
2.5
VP
’ Total SH = sum of y-glutamyl cycle thiols. Values are the means of either 3 replicate experiments (thiol levels) or 2 replicate experiments (fractions of reduced form); 2 exceptions: RB and 1 for AZ. n.d.. not done.
limit. 4 experiments
for
’ Thiol concentration is expressed as nmol/mg protein. The SD of the mean is given in parentheses ’ Fractions in reduced form (reduced/total) for fibroblasts in unfed plateau phase. They were not calculated when the level of thtol was below the detection
Reduced cysteine/total Total SH ’
Reduced ghttamylcysteine/total Cysteine
Reduced glutathione/total y-Glutamylcysteine
Glutathione
Cell strains
Levels of GSH, y-ghrtamylcysteine and cysteine (in reduced and oxidized forms) for fibroblasts in unfed plateau phase
TABLE
because of the lower GSH concentration [1,5]. In contrast, feedback inhibition should be normal in the two homozygous deficient strains in which the GSH level was not lowered and the quantity of y-glutamylcysteine was very low. The heterozygotes have GSH and y-glutamylcysteine concentrations which were similar to those of controls. The different mutations of the glutathione synthetase deficiency are then reflected in the concentrations of the various thiols, each of which may be affected differently in a given mutant strain. The sum of the individual thiol concentrations measured by high performance liquid chromatography is in good agreement with values obtained by measuring the overall thiol content calorimetrically [lo]. This would indicate that the thiols of the y-glutamyl cycle represent almost the overall thiol content in control human fibroblasts. The present study has also shown that the percentage of reduced thiols is unaffected by a glutathione synthetase deficiency. 2. Fibroblasts in unfed plateau phase The transition from growth phase to plateau phase was marked by a decrease in GSH level in 8 of the 9 strains studied (Table Iv>. The extent of the decrease in GSH varied from one strain to another (Table IV), independently of genotype. The only one of the two strains which were studied in the plateau phase and which had an appreciable y-glutamylcysteine level in the growth phase showed also a decrease in y-glutamylcysteine level. In contrast to the change in GSH, the transition from growth to plateau did not appear to induce a systematic decrease in cysteine level. Cysteinylglycine was present at low level (0.8-1.4 nmol/mg protein, e.g. 6-15% of the thiol content) in 4 cell strains (AB, RB, SP, EB) in plateau phase. The overall thiol level for each cell strain (except for AZ) in plateau phase was about 50% of that of growing cells. The level of reduced GSH in plateau phase cells represented about the same percentage of total GSH as in the growth phase (72-85%) (Table Iv) and appeared to be independent of genotype. The level of reduced cysteine in plateau phase cells was generally below 50% except in control cells (Table Iv>. The lowered thiol level in fibroblasts in the plateau phase found in this study follows similar pattern to that reported for cancer cells [3,8]. GSH is the thiol which is most involved in this phenomenon. The decrease in GSH may result from a lower metabolism in plateau phase cells [9]. Genetically deficient cell strains which differ not only in their capacity to synthesize GSH, but also in their thiol profile would appear to constitute a good biological model for the study of the cellular function of GSH and the other thiols. They provide an alternative to the use of depleting agents [3] and have already been successfully employed in a number of studies, dealing with radiobiology [14,10,20-221, amino acid active transport [5] and detoxification [23]. Acknowledgements The authors wish to acknowledge Dr. A. Larsson and his group for their help in glutathione synthetase studies and Mrs. J. Encinas for typing the manuscript.
168
This work was supported in part by ‘La Ligue Nationale Franqaise Contre le Cancer’ (ComitC des Hauts-de-Seine) and by contract BOO L05/3K 3237 from ‘Le Comitt de Radioprotection de 1’ElectricitC de France’. References 1 Meister A, Anderson ME. Ann Rev Biochem 1983;52:711-760. 2 Fahey RC, Newton GL. Functions of glutathione biochemical, physiological, toxicological and clinical aspects. In: Larsson et al, eds. New York: Raven Press, 1983;251-260. 3 Biaglow JE, Vames ME, Clarck EP, Epp ER. Radiat Res, 1983;95:447-455, 4 Harris JW, Biaglow JE. Biochem Biophys Res Commun 1972;46:1743-1749. 5 Larsson A. Transport and inherited disease. In: Belton NR, Toothill C, eds. Lancaster/Boston/The Hague: MTP Press Ltd. 1981;277-306. 6 Boivin P, Galand C. Nouv Rev Fr Hematol 1965;5:707-720. 7 Wellner VP, Sekura R, Meister A, Larsson A. Proc Nat Acad Sci USA 1974;71:2505-2509. 8 Cullen BM, Michalowski A, WaIker HC, Revtsz L. Int J Radiat Biol 1980;38:525-535. 9 Hahn GM, Little JB. Curr Top Radial Res Quart 1972;8:39-83. 10 Debieu D, Deschavanne PJ, Midander J, Larsson A, Malaise EP. Int J Radiat Biol 1985;48:525-543. 11 Deschavanne PJ, Midander J, Debieu D, Malaise EP, RCvCsz L. Int J Radiat Biol 1986;49:85-101. 12 Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Biol Chem 1951;193:265-275. 13 Spielberg SP, Kramer LI, Goodman SI, Butler J, Tietze F, Quinn P, Schulman JD. J Pediat 1977;91:237-241. 14 Edgren M, Larsson A, Nilsson K, Rev&z L, Scott OCA. Int J Radiat Biol 1980;37:299-306. 15 Jellum E, Marstein S, Skullerud K, Munthe E. Functions of glutathione: biochemical, physiological, toxicological and clinical aspects. In: Larsson et al, eds. New York: Raven Press, 1983;347-352. 16 Spielberg SP, Garrick MD, Corash LM, Butler J, Tietze F, Rogers L, Schulman JD. J Clin Invest 1978;61:1417-1420. 17 Boivin P, Galand C, Schaison G. Nouv Presse MM 1978:7:1531-1535. 18 Larsson A, Zetterstrom R, Hornell H, Porath U. Clin Chim. Acta 1976;73:19-23. 19 Larsson A, Mattsson B, Hagenfeldt L, Moldeus P. Clin Chim Acta 1983;135:57-64. 20 Rev&z L, Malaise EP. Functions of glutathione: biochemical, physiological, toxicological and clinical aspects In: Larsson et al, eds. New York: Raven Press, 1983;163-173. 21 Astor MB. Br J Radio1 1984;57:717-722. 22 Malaise EP. Radiat Res 1983;95:486-494. 23 Spielberg SP, Gordon GB. Clin Pharmacol Ther 1981;29:51-55.