Leukemia Research Vol. 17, No. 4, pp. 305-310, 1993. 0145--2126/93 $6.00 + .00 Printed in Great Britain. © 1993 Pergamon Press Ltd

CHARACTERISTICS OF THE INHIBITION OF HUMAN PROMYELOCYTIC LEUKAEMIA HL60 CELL GROWTH BY

S-D-LACTOYLGLUTATHIONE IN VITRO

LINDA EDWARDS, JAMES D. CLELLAND and PAUL J. THORNALLEY

Department of Chemistry and Biological Chemistry, University of Essex, Wivenhoe Park, Colchester CO4 3SQ, U.K.

(Received 13 August 1992. Revision accepted 19 December 1992)

Abstract--The mechanism of the inhibition of proliferation of human leukaemia 60 (HL60) cells by S-D-lactoylglutathione in vitro was investigated. The median inhibitory concentration IC50 value was 66 ixM (95% C.I. 50-87 ixM; n = 18). The inhibition of leukaemia cell growth required exposure of HL60 cells to S-D-lactoylglutathione (and metabolites) for 12 h, with maximum growth inhibition achieved after 24 h. Removal and replacement of culture medium within the initial 12 h of culture prevented inhibition of growth and toxicity. S-D-lactoylglutathione was consumed within the initial 3 h of culture. Pretreatment of culture medium containing 10% foetal calf serum for 3 h produced no subsequent inhibition of HL60 cell growth. Incubation of HL60 cells in culture medium with low serum content (5% v/v) produced a decreased rate of cell proliferation and a decreased response to S-D-lactoylglutathione.

S-D-lactoylglutathione inhibited uptake of 3H-thymidine into DNA in the third hour of culture where the median inhibitory concentration Its0 value was 74 IxM (95% C.I. 51-102; n = 10). The mechanism of inhibition of HL60 cell growth by S-D-lactoylglutathione is unknown but may be cell cycle related, mediated by inhibition of DNA synthesis and involve an active metabolite which may be removed and/or inactivated by a change in culture medium.

Key words: S-D-lactoylglutathione, glyoxalase, HL60 promyelocytic leukaemia.

INTRODUCTION

S - D - L A C T O Y L G L U T A T H I O N E is the physiological inter- mediate of the glyoxalase system which catalyses the conversion of methylglyoxal to D-lactate (Fig. 1). It comprises two enzymes, glyoxalase I and glyoxalase II, and a catalytic amount of reduced glutathione. Glyoxalase I (EC 4.4.1.5) catalyses the formation of S-D-lactoylglutathione from the hemithioacetal formed non-enzymatically from methylglyoxal and reduced glutathione. Glyoxalase II (EC 3.1.2.6) cata- lyses the formation of D-lactic acid from S-o-lac- toylglutathione, reforming the reduced glutathione consumed in the glyoxalase I-catalysed reaction [1]. The glyoxalase system is present in the cytosol of cells and sub-cellular organelles.

In human tumour cells, the activity of glyoxalase I was below, within and above the range found in normal human tissues and the activity of glyoxalase II was within and below the range found in normal human tissues [2]. The glyoxalase system was found to be present in the cytosol of human leukaemia 60

Correspondence to: Dr Paul J. Thornalley at the above a d d r e s s ,

305

MeCOCHO Methylglyoxal

GSH

MeCH(OH)CO2H D-Lactic acid

Glyoxalase I ] 1 ~

MeCH(OH}CO-SG S-D-Lactoylglutathione

Glyoxalase III i I i ~ H20

FIG. 1. The glyoxalase system.

(HL60) cells and had a relatively high activity of glyoxalase I and a low activity of glyoxalase II, and a high cellular concentration of S-o-lactoylglutathione [4]. The activity of glyoxalase I decreased and the activity of glyoxalase II increased during drug- induced terminal differentiation [3], producing a con- comitant decrease in the cellular concentrations of methylglyoxal and S-o-lactoylglutathione [4]. S-o- lactoylglutathione is formed and further metabolised

306 L. EDWARDS et al.

intracellularly [1] and does not readily cross cell plasma membranes [5]. Its normal locus of action is therefore restricted to the cytoplasm.

Exogenous S-D-lactoylglutathione (50-5001xM) added to the extracellular medium of HL60 cells in culture induced growth arrest and toxicity [6]. No significant terminal differentiation of HL60 cells was detected in this concentration range but up to 30% was found at higher concentrations (0.75-1.50 mM). Reduced glutathione, D-lactate, and reduced glu- tathione and D-lactate together did not induce growth arrest and toxicity in HL60 cells, indicating that the thioester group is associated with the inhibition of leukaemia cell growth. During the initial 24 h of culture with S-D-lactoylglutathione, the percentage of cells in the G0-G1 phase of the cell cycle was increased and the percentages of cells in the S and the Gz-M phases of the cell cycle were decreased. S- D-lactoylglutathione also inhibited DNA synthesis in HL60 cells. The incorporation of 3H-thymidine into DNA between 2 and 3 h of culture (before decreased cell viability was detected) was decreased by S-D- lactoylglutathione where the median inhibitory concentration IC50 value was 73~tM (95% C.I. 47-116~xM) [7]. Other S-2-hydroxyacylglutathione derivatives, S-glycolylglutathione, S-L-glyceroylglu- tathione and S-D-mandelylglutathione, also induced growth arrest, toxicity and inhibited D N A synthesis in HL60 cells, although S-D-lactoylglutathione was the most effective compound studied [7].

S-D-lactoylglutathione was not toxic to corre- sponding differentiated cells, neutrophils [7], although it inhibited stimulus-activated secretion and chemotaxis at high concentrations (1-5 mM) [8, 9]. The mechanism of toxicity is still not understood.

In this report , we describe studies on the charac- teristics of the development of growth arrest and toxicity in HL60 cells exposed to S-D-lactoylglu- tathione.

M A T E R I A L S AND M E T H O D S

Trypan blue, 12-O-tetradecanoyl-phorbol-13-acetate (TPA), nitroblue tetrazolium (NBT), reduced glutathione, S-ethylglutathione, glyoxalase I from yeast (type X), gly- oxalase II from bovine liver and D-lactic acid were pur- chased from Sigma Chem. Co. Ltd (Poole, Dorset, U.K.). Tissue culture medium RPMI 1640 and foetal calf serum were purchased from Gibco Europe Ltd (Paisley, Scotland). Strong anion exchange solid phase extraction cartridges (SAX-SPE; 500mg, 2.8ml) were purchased from Alltech (Carnforth, Lancs, U.K.). High performance liquid chromatography (HPLC) cartridges (Nova-Pak octa- decylsilica ODS, 4 ~tm, 0.8 x 10 cm) and pre-columns of the same material (0.8 × 1.0cm) were purchased from Waters-Millipore U.K. (Watford, U.K.). S-D-lactoyl- glutathione, free acid, was prepared and purified as recently described [10]. Solutions of S-D-lactoylglutathione

were sterilised by micropore filtration before being added to the cell suspensions.

Cell culture conditions Human promyelocytic leukaemia HL60 cells were nor-

mally incubated at 37°C in RPMI 1640 media containing 10% foetal calf serum under an atmosphere of 5% CO2 in air, 100% humidity [6]. Cells were seeded at an initial density of 5 × 104/ml and incubated with 1-500 IxM of S- D-lactoylglutathione for up to 4 days. Cell viability was judged by the ability of cells to exclude trypan blue [11]. The differentiation of HL60 cells to neutrophil-like cells was estimated from the development of maturation-depen- dent activity of the superoxide-forming enzyme NADPH oxidase, by staining for the superoxide-mediated reduction of nitroblue tetrazolium (NBT) for formazan, in the pres- ence of phorbol ester [12].

The dependence of inhibition of HL60 cell growth on the time of exposure to S-D-lactoylglutathione was inves- tigated by incubating cells with S-D-lactoylglutathione for 0, 3, 6, 9, 12, 18 and 24 h, and thereafter replacing the culture medium with fresh medium without S-D-iac- toylglutathione and incubating the cells such that the com- bined incubation period with and without S-D- lactoylglutathione was 96 h (control cells were manipulated identically without S-D-lactoylglutathione at all times). The effect of culture medium processing with S-D-lactoyl- glutathione was investigated by incubating medium with S- D-lactoylglutathione for 3 h. This medium was then added to ceils with incubation for 96 h.

The effect of serum composition of the culture medium on the inhibition of HL60 cell growth by S-D-lac- toylglutathione was investigated by incubating HL60 ceils (5 x 104/ml) with and without S-D-lactoylglutathione (80 ~tM) in RPMI 1640 medium with 5, 10 and 15% foetal calf serum for 4 days and measuring cell growth.

The rate of DNA synthesis in HL60 cells was determined by measuring the rate of uptake of 3H-thymidine into DNA [13]. HL60 cells (5.0 x 104/ml) were incubated with 10 txM-1 mM S-D-lactoylglutathione in RPMI 1640 with 10% foetal calf serum for 2 h. 3H-thymidine (Amersham Int., Amersham, Bucks., U.K.; 2.5 ~tCi) was added and the incubation continued for 1 h. The DNA was then extracted [13] and counted.

Assay of S-D-lactoylglutathione in HL60 cells and medium S-D-lactoylglutathione was assayed by reverse phase

HPLC with spectrophotometric detection at 233 nm and sample clean up by SAX-SPE [14]. SAX-SPE cartridges were prepared by eluting 12 ml of 1 M ammonium formate, pH4.3, and then 12ml of 2mM ammonium formate, pH4.3, through the cartridge. HL60 cells (5 x 104/ml; i ml) were incubated with 100 ~tM S-D-lactoylglutathione and the medium assayed for S-D-lactoylglutathione after 0, 30, 60 and 180min. The concentration of S-D-lac- toylglutathione in HL60 cells was assayed after incubation of HL60 ceils (1 × 106/ml; 20 ml) for 3 h with and without 200 IxM S-D-lactoylglutathione. Samples (20 x 10 6 cell pel- let or 1 ml of medium) were de-proteinized by addition of perchloric acid (2 ml, 0.6 M) at 4°C, vortex mixed and left on ice for 15 min. The samples were centrifuged (6000g, 10 min, 4°C) and 1.8 ml of the supernatant was then neu- tralised to pH 5-6 with potassium hydrogen carbonate (400 Ixl, 2 M). The potassium perchlorate precipitate was sedimented by centrifugation (300g, 5 min, 4°C). The supernatant (2 ml) was applied to the SAX-SPE cartridge,

Inhibition of HL60 cell growth by S-D-lactoylglutathione 307

the cartridge was washed by eluting with 6 ml of 2 mM ammonium formate, pH 4.3, and the analyte sample then eluted from the cartridge with 6 ml of 500 mM formic acid. The eluate collected was lyophilised on a centrifugal evaporator. The residue was reconstituted with 200 p.1 of HPLC mobile phase containing 150 ~tM S-ethylglutathione (internal standard).

The HPLC mobile phase was sodium chloride (1 M), ammonium formate (10 mM), pH 3.4. The flow rate was 2 ml/min and 100 til of reconstituted sample was injected. Detection was by absorbance spectrophotometry at 233 nm with a typical full scale deflection of 0.02 A.U., integrating between 10 and 25 min post-injection. After 30 min, the column was washed by a linear gradient over 3 min to 100% methanol and then returning to the mobile phase with a linear gradient also over 3 min. HPLC was per- formed with a Waters HPLC system (2 × 510 pumps, Lambda Max 481 LC spectrophotometer with a 680 auto- mated gradient controller). The column was Nova-Pak octadecylsilica (ODS) 4 ~tm (0.8 × 10 cm) cartridge fitted with a pre-column in an 8 × 10 radial compression unit.

The S-D-lactoylglutathione assay was calibrated by add- ing 20-100 Ixl of 200 tiM S-D-lactoylglutathione to the bio- logical matrix after treatment with perchloric acid but before removal of the protein precipitate in the sample de- proteinization procedure. Intrabatch and interbatch vari- ations in the assay of S-D-lactoylglutathione were 0.7 and 12%, respectively, and the detection limit was 0.9 nmol.

R E S U L T S

Inhibition of human promyelocytic leukaemia 60 cell growth by S-D-lactoylglutathione

When HL60 cells (5 × 104/ml) were incubated for 4 days with S-D-lactoylglutathione (25-500 ~tM), there was a decrease in viable cell number with increasing S-D-lactoylglutathione concentration and a con- comitant increase in the percentage of non-viable cells, as judged by the trypan blue exclusion test. The growth inhibition data was t ransformed for a logit plot; logit (growth inh ib i t ion)= l n ( ( 1 0 0 - % viable cell n u m b e r ) / % viable cell n u m b e r ) - - s e e Fig. 2. From linear regression of these data (r = 0.983), the median inhibitory concentrat ion Ics0 was 66 ~tM (95% C.I. 50-87; n = 18). The slope of the logit plot was 3.4 +- 0.2 (n = 18).

D N A synthesis in HL60 cells was determined by measuring the rate of 3H-thymidine incorporation into D N A during the third hour of culture. During this period, no decrease in cell viability nor significant change in the cellular pool of thymidine was found. The inhibition of 3H-thymidine incorporation data gave a linear logit plot (r = 0.978; Fig. 3) where the median inhibitory concentrat ion of S-D-lactoylglu- tathione was 72 ~tM (95% C.I. 51-102; n = 10), which was not significantly different f rom the Its0 value for decrease in viable cell number (p > 0.05). The slope of the logit plot was 1.78 --- 0.13.

8

O A- Ca)

.~ 2" ~

.~ -2

-.1 -4] a

-61 3 ¼ ; ; 7

Ln([S-D-lactoylglutathione]/pM)

F1G. 2. Effect of S-D-lactoylglutathione on human leu- kaemia 60 cell growth. Logit plot. Cells were seeded at an initial density of 5 x 104/ml and incubated for 4 days with the concentration of S-D-lactoylglutathione indicated (n =

18).

o

2

2- B

0 ........................................

-1

2.5 3.0 3.5 4.0 4.5 5.0 5.5 Ln([S-D-Lactoylglutathione]/pM)

FIG. 3. Effect of S-D-lactoylglutathione on 3H-thymidine incorporation into DNA in HL60 cells. HL60 cells (5.0 × 104/ml) were incubated with the concentration of S- D-lactoylglutathione indicated in RPMI 1640 with 10% foetal calf serum for 2 h. 3H-thymidine (2.5 tiCi) was added and the incubation continued for 1 h. The DNA was then extracted [13] and counted. Logit (inhibition of thymidine incorporation) = ln(I/100-I) where I = inhibition of thy- midine incorporation. The Ic50 value was 74 tiM (95% C.I.

51-102 p.M) (n = 10).

The influence of serum composition of the culture medium on the inhibition of HL60 cell growth by S- D-lactoylglutathione

When HL60 cells (5 × 104) were incubated in RPMI 1640 medium containing 5, 10 and 15% (v/v) foetal calf serum, the viable cell number after 4 days increased with increasing serum content (Fig. 4). Addition of S-D-lactoylglutathione (80 ~tM) to cell cultures inhibited the growth of HL60 cells in medium

308 L. EDWARDS e t al.

Q.

iii 1 d~ 50 / , ~ /

0 1 2 3

Time {days)

b.

125

~ 100

x 75

i ~ 5o

"~ 25

C.

j ns[

,00i /

ol

Time (days) Time (days)

FIG. 4. Effect of serum composition of the culture medium on the inhibition of HL60 cell growth by S-D-lac- toylglutathione. HL60 cells (5 × 104/ml) were incubated with (11) and without (I-q) 80 IxM S-D-lactoylglutathione in RPMI 1640 medium containing (a) 5%, (b) 10% and (c) 15% foetal calf serum. Data are mean -+ tr for 3 inde-

pendent experiments.

I00 f

,~ 80

E

Z e~

> /

40 . . . . . . . . . t - -3 3 9 15 21 27 96

0 6 12 18 24 30

Dose period (h)

FIG. 5. Effect of the dose period on the inhibition of HL60 cell growth by S-D-lactoylglutathione. HL60 cells (5 × 104/ml) were incubated with 80 ~tM S-D-lactoylglutathione for the time indicated and then in fresh medium for a total of 96 h. Data are mean - trofa minimum of 3 experiments.

containing 10 and 15% serum (p < 0.01) but not in cultures with a slower growth rate containing 5% serum (p > 0.05).

Effect of the period of exposure of HL60 cells to S-o- lactoylglutathione and metabolites on the development of growth inhibition

S-D-lactoylglutathione (100FtM) was consumed very rapidly in HL60 cell cultures and medium. S-D- lactoylglutathione was >90% consumed within the initial 3h of culture; the half life of S-o-lac- toylglutathione was c. 35 min and surprisingly c. 40 min in the absence of cells. The concentration of S-o-lactoylglutathione in HL60 cells was not sig- nificantly increased by incubation with 200 [xM S-O- lactoylglutathione for 3 h (42.9 -+ 15.4 pmol/106 cells ( n = 4 ) vs 59.1-+7.4pmol/106 cells ( n = 3 ; p > 0.05).

Culture medium pretreated with 80 ~tM S-D-lac- toylglutathione for 3h did not inhibit the pro- liferation of HL60 cells in culture--data point shown as -3 h dose period, Fig. 5. When HL60 cells were incubated with 80 ~tM S-D-lactoylglutathione for 3, 6, 9, 12, 18, 24 and 96h, significant inhibition of growth developed only after a 12 h dose period (p < 0.01) and maximised after 24 h (Fig. 6). There- after, up to 96 h dose period, there was no further increase in growth inhibition (p > 0.05).

DISCUSSION

S-D-lactoylglutathione was previously found to inhibit the growth of HL60 cells in culture [6]. The major metabolites, reduced glutathione and D-

lactate, added separately or together, did not induce a similar effect. Inhibition of HL60 cell growth, there- fore, seems to be associated with the thioester group of S-D-lactoylglutathione; other S-2-hydroxyacyl- glutathione derivatives were also found to inhibit HL60 cell growth in culture but S-D-lactoylglu- tathione was the most potent [7]. However, alkyl thiolactates CH3CH(OH)CO-SR did not similarly inhibit tumour growth [15]. It therefore appears that the thioester and glutathione moieties are both important for antiproliferative activity. The mech- anism of growth inhibition is unknown but remains of interest, particularly because S-D-lactoylglutathione represents a prospective novel therapeutic agent with highly selective cytotoxicity to proliferating cells and with major metabolites of low toxicity and well- established metabolic profile.

S-D-lactoylglutathione inhibited HL60 cell growth and induced toxicity, as judged by the trypan blue exclusion test. It is likely that the non-viable cell number was underestimated since cell debris may be observed by microscopic and flow cytometric analysis of HL60 cell cultures incubated with S-D-lac- toylglutathione [6]. The Ics0 value for S-D-lactoyl- glutathione was 66 FtM, determined from the logit transformation of the growth inhibition data. The linearity and slope (3.41-+0.15) of this trans- formation suggests a consistent amplification of growth inhibition over the complete concentration range in which S-D-lactoylglutathione was active. Further investigation of the molecular processes involved are required but this observation suggests a controlled mechanism of cell death may be involved.

S-D-lactoylglutathione also inhibited the uptake of

Inhibition of HL60 cell growth

3H-thymidine into DNA early in the development of growth arrest and toxicity. Induction of growth arrest of HL60 cells by S-D-lactoylglutathione is a possible explanation for the observed decrease in the rate of DNA synthesis. However, the rapid inhibition of DNA synthesis relative to induction of growth arrest (as judged by changes in cell cycle distribution) [6] and the similarity of the dose-response relationships for inhibition of cell growth and inhibition of DNA synthesis implicates the inhibition of DNA synthesis in the mechanism of cell death.

The logit plot for inhibition of DNA synthesis was linear with a slope of 1.78 -+ 0.13. During the third hour of incubation there was little S-D-lactoyl- glutathione remaining in the culture medium. The linear logit plot suggests the inhibition of DNA syn- thesis is controlled by a process in which equilibrium binding of S-D-lactoylglutathione is an important step. The positive slope of the logit plot reflects co- operativity S-D-lactoylglutathione metabolite-recep- tor binding. This is consistent with a controlled mech- anism of inhibition of DNA synthesis across the active range of S-D-lactoylglutathione concentration.

The investigation of the effect of dose period required to induce inhibition of growth in HL60 cells showed that serum processing by S-D-lactoylglu- tathione did not induce toxicity and that HL60 cells had to be incubated with S-D-lactoylglutathione and metabolites for 24h to produce the maximum response. This is consistent with the inhibition of growth being produced by an interaction of S-D- lactoylglutathione with HL60 cells producing a metabolite or inducing cellular processes which may be successfully countered by replacement of the cul- ture medium. The development of growth arrest and cytotoxicity between 12 and 24 h of culture suggests that the inhibition of HL60 cell growth by S-D-lac- toylglutathione is cell cycle-related.

The time course of the inhibition of HL60 cell growth by S-D-lactoylglutathione in RPMI 1640 with 10% serum (Fig. 4b) revealed that the maximum inhibition of growth was manifes, within the first 24 h of incubation. Thereafter, the rate of growth remained the same in both cases: logarithmic trans- formation of viable cell number data produces par- allel growth curves, indicating that the surviving cells in S-D-lactoylglutathione treated cultures have pre- served their normal proliferative capacity [7].

The growth curves for the incubation of HL60 cells in culture medium containing 5% serum with and without 80 IxM S-D-lactoylglutathione (Fig. 4a) indi- cated that no significant inhibition of growth was induced by S-D-lactoylglutathione under these con- ditions. The explanation for the decreased effect may be related to the slow rate of HL60 cell proliferation

by S-D-lactoylglutathione 309

in this serum-depleted medium. S-D-lactoylglutathione was short-lived in cell cul-

tures and culture medium alone. The rate of con- sumption was far in excess of the spontaneous rate of hydrolysis [16]. Preliminary investigations suggest that the rapid consumption of S-D-lactoylglutathione in the culture medium was inhibited by c. 90% in the presence of the glyoxalase II inhibitor S-p- nitrobenzoxycarbonylglutathione [17] (100 ~tM); glyoxalase II contamination may arise from lysis of red blood cells during serum collection. This con- tamination converts most of the added S-D-lac- toylglutathione to innocuous products, D-lactate and reduced glutathione. The residual consumption of S- D-lactoylglutathione was inhibited by the y-glutamyl- transpeptidase inhibitor acivicin [18] (0.1 mg/ml) (T. W. C. Lo and P. J. Thornalley, unpublished obser- vations), y-Glutamyltranspeptidase catalyses the conversion of S-D-lactoylglutathione to N-D- lactoylcysteinylglycine [19]. The Ics0 values for the inhibition by HL60 cell growth for S-2-hydroxy- acylglutathione derivatives where found to be in inverse order of their KM values with y-glutamyl- transpeptidase [7]. Preliminary studies have shown that N-D-lactoylcysteinylglycine inhibits HL60 cell growth in culture and is more potent than S-D-lac- toylglutathione (T. W. C. Lo & P. J. Thornalley, unpublished observations). Further studies are required but it is clear that the level of glyoxalase II contamination of serum and the time of preparation of S-D-lactoylglutathione solution prior to addition to cell cultures are factors to be carefully controlled to ensure reproducibility of the response.

The mechanism of inhibition of HL60 cell growth by S-D-lactoylglutathione is unknown but may be cell cycle-related, mediated by inhibition of DNA synthesis and involve controlled cell death induced by an active metabolite which is removed and/or inactivated by change in culture medium.

Acknowledgements--The authors thank the Cancer Research Campaign for research support. This is a con- tribution from the Glyoxalase Research Group at the Uni- versity of Essex.

REFERENCES

1. Thornalley P. J. (1990) The glyoxalase system: new developments towards functional characterisation of a metabolic pathway fundamental to biological life. Biochem. J. 269, 1.

2. Ayoub F. M., Zaman M. A., Thornalley P. J. & Masters J. R. W. (1993)Glyoxalase activities in human tumour cell lines in vitro. Biochem. Soc. Trans. 21, 167F.

3. Hooper N. I., Tisdale M. J. & Thornalley P. J. (1987)

310 L. EDWARDS et al.

Glyoxalase activity during differentiation of human leukaemia cells in vitro. Leukemia Res. 11, 1141.

4. Hooper N. I., Tisdale M. J. & Thornalley P. J. (1988) Modification of the glyoxalase system during dif- ferentiation of human promyelocytic leukaemia HL60 cells in vitro. Biochem. Biophys. Acta 966, 362.

5. Rae C., Board P. G. & Kuchel P. W. (1991) Glyoxalase 2 deficiency in the erythrocytes of a horse: ~H n.m.r. studies of enzyme kinetics and transport of S-lactoyl- glutathione. Arch. Biochem. Biophys. 291, 291.

6. Thornalley P. J. & Tisdale M. J. (1988) Inhibition of proliferation of human promyelocytic leukaemia HL60 cells by S-D-lactoylglutathione. Leukemia Res. 12, 897.

7. Clelland J. D., Allen R. E. & Thornalley P. J. (1992) Inhibition of growth of human promyelocytic leu- kaemia cells by S-2-hydroxyacylglutathiones and their monoethyl ester derivatives. Biochem. Pharmac. 44, 1953.

8. Allen R. E. and Thornalley P. J. (1993) The effect of S-D-lactoylglutathione on the chemotaxis of human neutrophils. Biochem. Soc. Trans. 21, 161F.

9. Thornalley P. J., Greskowiak M. & Della Bianca V. (1990) Potentiation of secretion from neutrophils by S- D-lactoylglutathione. Med. Sci. Res. 18, 813.

10. Clelland J. D. & Thornalley P. J. (1991) S-2-hydr- oxyacyl-glutathione derivatives: enzymatic prepar- ation, purification and characterisation. J. Chem. Soc. Perkin Trans. I, 3009.

11. Kaltenbach J. P., Kaitenbach M. H. & Lyons W. B.

(1958) Nigrosin as a dye for differentiating live and dead ascites cells. Expl Cell Res. 15, 112.

12. Mendelson N., Gilbert H. S., Christian J. K. & Acs G. (1980) Effect of maturation on the response of human promyelocytic leukaemia cells (HL60) to the tumor promoter 12-O-tetradecanoylphorbol-13-acet- ate. Cancer Res. 40, 1469.

13. Wilson A. P. (1986) Cytotoxicity and viability assays. In Animal Cell Culture. A Practical Approach (R. I. Freshney, Ed.) pp. 194-216. IRL Press, Oxford.

14. McLellan A. C., Phillips S. A. & Thornalley P. J. (1992) Assay of S-D-lactoylglutathione in biological systems. Anal. Biochem. (in press).

15. Hall S. S., Doweyko L. M., Doweyko A. M. & Ryan Zilenovski J. S. (1977) Synthesis and evaluation of o:- hydroxythiol esters as anti-tumour agents and gly- oxalase I inhibitors. J. med. Chem. 20, 1239.

16. Reeves M. & Thornalley P. J. (1993) The hydrolysis of S-D-lactoylglutathione. Biochem. Soc. Trans. 21, 169F.

17. Lo T. W. C. & Thornalley P. J. (1992) Inhibition of proliferation of human leukaemia 60 cells by diethyl esters of glyoxalase inhibitors in vitro. Biochem. Phar- mac. 44, 2357.

18. Meister A., Tate S. S. & Griffith O. W. (1981) ~,- glutamyltranspeptidase. Meth. Enzym. 77, 237.

19. Tate S. S. (1975) Interaction of y-glutamyltrans- peptidase with S-acylglutathione derivatives. FEBS Lett. 54, 319.