Inhibitors of glyoxalase I in vitro

BIOCHEMICAL

MEDICINE

29, 385-391 (1983)

Inhibitors of Glyoxalase I in Vitro’ RICHARD B. BRANDT,’ Department of Biochemistry Virginia Commonwealth Chemistry, University

MARK E. BRANDT, COLIN THOMSON

MICHAEL

E. APRIL, AND

and MCVIVCU Cancer Center. Medical College of Virginia, University, Richmond, Virginia 23298; and Department qf of St. Andrews. Scotland KY16 9ST. United Kingdom

Received December 17. 1982

Inhibition of tumor growth by extracts from normal tissues was found, at least in part, to be due to (Y, P-dicarbonyls (1,2). Methylglyoxal (MeG), the lowest class member of the ketoaldehydes is catalyzed by two separate mammalian enzymes: glyoxalase I (S-lactoyl-glutathione methylglyoxal lyase, isomerizing; EC 4.4.1 S) and glyoxalase II (S-2-hydroxyacylglutathione hydrolase; EC 3.1.2.6), with glutathione (GSH) as a cofactor for the formation of o-lactate (3). Various possible functions of the glyoxalase enzymes have been summarized elsewhere (4,5). Compounds that are inhibitors of the system have been used in an attempt to retard tumor growth, possibly through the effect of MeG on protein synthesis (6). Gillespie has suggested that increased S-lactoyl-glutathione may be related to tumor promotion (7). Clinical treatment of cancer has also been attempted with MeG derivatives (8). Various routes of both enzymatic (9) and nonenzymatic (10) synthesis of MeG have been reported. Indirect evidence for methylglyoxal synthesis by determination of u-lactate, when glycolysis was inhibited or glycolytic intermediates were present at higher than steady-state levels have been reported from this laboratory (10). Humans have small but measurable plasma concentrations of o-lactate (11) and during the rapid phase of growth in rats u-lactate is formed (5). Inhibition of glyoxalase I and growth of L1210 in culture was reported by Vince et al. (12) using S-alkylglutathiones as competitive inhibitors and in combination with MeG the effects were synergistic. However, the rapid in viva metabolism of both kektoaldehydes to their hydroxy ’ A report of this work was presented at the Sanibel Quantum Biology Symposium. March 1982, Flagler Beach, Fla. Int. J. Quantum Chemistry: Quantum Biology Symposium 9, 335-343 (1981). ’ TO whom correspondence should be addressed. 385 0004-2944/83 $3.00 Copyright ‘Q 1983 by Academic Press. Inc. All rights of reproduction in any form reserved

acids and of GSH analogues. decreases their effectiveness as chemotherapeutic agents. Nucleotides and some related compounds were reported to be cooperative inhibitors (4). but the physiological significance has not been shown. In selecting effective inhibitors of glyoxalase I rathet than substrate analogues. the use of inhibitors that fit the structure of transition-state analogues may significantly bind to the enzyme (13). An enediol has been suggested as a transition intermediate from the hemimercaptal in the glyoxalase I catalyzed formation of S-lactoyl-glutathione (14).

We report here the inhibition of human red blood cell glyoxalase 1 activity with a number of enediol compounds as possible transition-state inhibitors and their concentration for 50% inhibition t15,,). MATERIALS

AND METHODS

Red blood cell glyoxalase I. Human blood was collected in citratephosphate-dextrose anticoagulant, centrifuged in a Sorvall RC2-B at 4000g for 6 min, and the plasma discarded. The cells were washed three times with 60% of the original blood volume with 0.15 M NaCI, centrifuged at 4000g for 6 min and the supernatant fluid was discarded after each washing. The red blood cells were lysed with an equal volume of cold HZ0 by mixing for IO min in an ice bath followed by centrifugation at 10,OOOg for 20 min. The supernatant fluid was used as the glyoxalase I preparation without further purification. Hemimercaptal solutions (glyoxalase 1 substrate). GSH from Sigma Chemical Co. (St. Louis, MO.) was used to prepare 0. I M stock solutions, adjusted to pH 6.8, and stored at -20” until use. MeG was prepared as previously described (5). GSH. 3.46 mrvr, and MeG, 2.07 mM, in phosphate buffer, pH 6.8, was incubated l-2 hr at 25” giving approximately I mM hemimercaptal. Inhibitor solutions. The potential inhibitors were reagent grade and used without further purification except as noted. The compounds to be tested were diluted to about 0.02 M in potassium phosphate buffer, 0.1 M at pH 6.8. Compounds that were insoluble in buffer were dissolved in absolute ethanol and then diluted in buffer and used at 3 to 6 mM. Glyoxaluse I activity. Spectrophotometric measurements were obtained with a Gilford 252 modified Beckman DU equipped with a strip chart recorder and an automatic cuvette positioner. Hemimercaptal solutions, 2.8 ml, were added to each of four quartz l-cm2 cuvettes. The absorbance was determined for a reagent blank after addition of either 0.1 ml of phosphate buffer or 0. I ml of buffer containing the same concentration of ethanol as used in diluting the potential inhibitor. The reaction at 25” was initiated with 0.02 ml of the blood lysate and the change in absorbance/ minute (AAlmin) was measured at 240 nm which is the maximum ab-

INHIBITORS

OF GLYOXALASE

387

I

sorbance for the S-lactoyl-glutathione as reported by Racker (3). The rate of the uninhibited enzyme activity was determined at frequent intervals during the testing of inhibitors and the mean values (about 0.1 Almin) were used as the 100% activity. To the other cuvettes, 0.1 ml of various concentrations of potential inhibitor were added and the absorbance adjusted to 0.000 and the reaction initiated with 0.02 ml of the red cell lysate. The rate in Mlmin was determined in triplicate for each inhibitor concentration. The percentage activity remaining was AAlmin with inhibitor x ,oo L4lmin without inhibitor ’ The concentration for 50% inhibition (Z& was determined from interpolation of plots of percentage activity against at least five different inhibitor concentrations. RESULTS

AND DISCUSSION

The transition-state structure from the hemimercaptal of GSH and MeG has not been determined, however, the enediol shown in Fig. 1A has strong experimental support for being one of the possible intermediates and contains a structural feature common to some compounds which are inhibitor of glyoxalase I (14). The compounds that were selected to test for inhibition of glyoxalase I were selected (unless otherwise noted) on the basis of enediol similarity. The possibility of alternate intermediates of different structure type exists and must be determined by additional research. Table 1 lists the compounds tested as inhibitors of red blood cell glyoxalase I and their concentration for 50% inhibition. Tryptophan was not selected as a possible transition-state inhibitor, but was a compound that had been previously reported by Oray and Norton (4) as an inhibitor of glyoxalase I. The Z,, found here of 1.05 mM is similar to that found for mouse liver glyoxalase I (4) of 0.90. The mechanism of inhibition by tryptophan or similar compounds is not known. As reported by others A CHS

\

B I

t

c

OH

C

HO

GS’ ‘OH

RO

HO

0”

- OH u +

0

X

FIG. 1. (A) Possible enediol intermediate in the action of glyoxalase I; (B) esculin (when R is glucose and X is H), esculetin (when both Rand X are H), or d-methyl esculetin (when R is H and X is CHI); and (C) squaric acid.

I,,, CON(~ENTRATION

I’ABLE (mM) FOR

I GLYOXAIASE

INHIHITORS

Compound L-Tryptophan” L-Ascorbic aci&’ Araboascorbic acid’ 3,3-Dihydroxypyridine” 2.3-Dihydroxybenzoic acid” 3,4-Dihydroxyben~oic acid”.’ 3,4-Dihydroxybenzohydroxamic acid’ 3,45Trihydroxybenzohydroxamic acid/‘ 2,3.4-T~hydroxybenzohydroxamic acid’ Esculetin (6.7-dihydroxycoum~in)~ Esculin (6-glucono-7-hydroxycouma~n~ Coumarin” 4-Methylesculetin (4-methyl-6,7-dihydroxycoumarin) Squaric acid (3,4-dihydroxy-3-cyclobutene-1.2”dione)” -. --- -.. _ .--.-_ ~.. ” Sigma Chemical Co.. St. Louis, MO. ’ Dehydroascorbic acid had less than 5% inhibition at 0.7 mM. ’ Fisher Scientific Co., Springfield. N.J. ‘I Aidrich Chemical Co.. Milwaukee, Wise. ’ t-Dopa had less than 5% inhibition at 0.9 mM. ’ Gift from Dr. Howard Elford. p Hydrolysis product of esculin, recrystallized 2 x from ethanol.

.-- _ .

1.05 0.48 0.56 0.53 0.28 0.32 0.38 0.x 0.21 0.03 0.23 1.40 0.03 0.12

the concentration for 70% inhibition of yeast glyoxaiase I activity for squaric acid (Figure 1C and 3) and 2,3”dihydroxybenzoic acid in mlw was 0.25 and 0.58, respectively (14). Extrapolation of our graphical data for these compounds at 70% inhibition showed approximately 0.14 and 0.35 mM concentrations, which considering the differences in methodology and the difficulty in estimating inhibitor concentration in that region of the curve is a similar finding. Both L-ascorbic acid and araboascorbic acid, which fit enediol structures show ISo values around 0.5 mM. For yeast glyoxalase I 10 mM of araboascorbate was 100% inhibitory (14). Figure 2 shows the similarity of their inhibitions with ascorbic acid being slightly more inhibitory. The physiological concentration of ascorbic acid is at least a magnitude less than used here, suggesting that this is not a normal control of glyoxalase I activity in most tissues. Dehydroascorbic acid at 0.7 mM could only be estimated as about 5% inhibitory. These findings are similar to that reported by others (15) for yeast glyoxalase with an Is0 of 0.6 mM for ascorbic acid, while dehydroascorbic acid was not inhibitory at 6 mM. Ascorbic acid was shown by kinetic studies to be an uncompetitive inhibitor (15). 2,3-Dihydroxybenzoic acid with a fs,) (Table 1) of 0.28 mM is in the

lNHll3ITORS

0

2

OF GLYOXALASE

4 [INHI~I~~R~

6

8

389

I

Kl

mi~t01

FOG. 2. Inhibition of red blood cell giyoxalase I with L-ascorbic acid (B) or araboascorbic acid (A). Calculated hemimercaptal about 0.1 mM.

range of other di- or trihydroxy compounds tested of 0.21-0.53 mM with the exception of L-Dopa (3,4-dihydroxyphenylalanine) which had less than 5% inhibition of glyoxalase I at 0.9 mM. These structures and their Z,, values may provide data for theoreticai selection of compounds to be used as more effective inhibitors. The hydroxybenzohydroxami~ acids reported in Table 1 have been shown by Elford et al. (16) to inhibit ribonucleotide reductase in vitro and mouse L1210 leukemia in vivo. Of these three compounds, 3,4-dihydroxybenzohydroxamic acid was most effective in inhibiting L1210 growth (16). Figure IB shows the structure for derivatives of coumarin, esculin (a glucoside), esculetin (the aglycone), and 4-methyl esculetin depending on the R or X groups attached. As expected coumarin was least inhibitory, however, both esculetin and 4-methyl esculetin have the lowest I,, of the compounds tested. Although esculin was not as inhibitory as esculetin (see Fig. 3) this compound may have some advantages in viva due to better transport into tissue compared to the aglycone, followed by the intracellular hydrolysis of esculin. Figure 3 shows the decrease in activity of glyoxalase I with increasing amounts of esculetin and squaric acid. The different shapes of these curves suggest that the type of inhibition of these compounds may not be the same and will require detailed kinetic experiments to interpret them. Since this initial study was directed at determining compounds which were selectedfor their structural similarity to the potential transitionstate of glyoxalase I, the emphasis was on determination of the I,, concentration, with the actual type of inhibition to be determined for the most active inhibitors in a future study.

390

FIG. 3. inhibition of red blood cell glyoxaiase 1 with esculetin fl), es&in ithe glucosidef (A). or squaric acid (0). Calculated hemimercaptal about 0.1 mM.

The application of some of the compounds reported here, with and without added MeG, to test their effect on inhibition of L1210 leukemia growth in mice, is now in progress. The selection of other potential inhibitors will be based on these in ~‘itro studies, in viva L1210 leukemia resufts and correlation of the main features of the electrostatic potential maps generated for these molecules by ab initio quantum mechanical calculations. SUMMARY

Inhibition in vitro of human red blood cell glyoxaiase I activity was measured by the decrease in the rate of formation of S-n-lactoyl-glutathione as determined by the change in absorbance at 240 nm. The percentage activity remaining was determined after addition of various potential inhibitor compounds and the concentration for 50% activity was obtained by graphical interpolation. The inhibitors were selected on the basis of their similarity to a possible transition-state enediol intermediate of methylglyoxal. The most effective inhibitors were dihydroxycoumarins with a 50% inhibition of 0.03 mM. Inhibition of methyl~yoxal catabolism suggests possible application as chemotherapeutic agents based on the inhibitor characteristics of methylglyoxal. ACKNOWLEDGMENTS This research was supported by funds from the National Foundation for Cancer Research. Helpful suggestions were made by Dr. Howard L. Elford of the Virginia Associated Research Campus of the College of William and Mary.

INHIBITORS

OF GLYOXALASE

I

391

REFERENCES 1. Szent-Gyorgyi, A., Egyud, L. G., and McLaughlin, J. A., Science 155, 539 (1967). 2. Fodor, G., Sachetto, J. P., Szent-Gyorgyi, A.. and Egyud, L. G.. Proc. Nat. Acad. Sci. USA 57, 1644 (1967). 3. Racker, E., J. Biol. Chem. 190, 685 (1951). 4. Oray, B., and Norton, S. J., Biochem. Biophys. Res. Commun. 95, 624 (1980). 5. Brandt, R. B., Waters, M. G.. Muron, D. J.. and Bloch, M. H.. Proc. Sot. Exp. Biol. Med.

6. 7. 8. 9. IO.

I I. 12. 13. 14. 15. 16.

169, 463 (1982).

Szent-Gyorgyi, A., and Egyud, L., Proc. Nat/. Acad. Sci. USA 56, 203 (1966). Gillespie, E., Biochem. Biophys. Res. Commun. 98, 468 (1981). Regelson, W., and Holland, J. F., Cancer Chemother. Rep. 27, I5 (1963). Sato, J., Wang, Y., and Van Eys, J., J. Viol. Chem. 255, 2046 (1980). Brandt, R. B., and Siegel, S. A., in “Methylglyoxal production in human blood” (G. E. W. Wolstenholme, D. W. Fitzsimons. J. Whelan. Eds.), Submolecular Biology and Cancer (Ciba Foundations Symposium). Vol. 67. pp. 21 l-223. Excerpta Medica. Amsterdam, 1979. Brandt, R. B., Siegel, S. A.. Waters. M. G.. and Bloch. M. H.. Anal. Biochem. 102, 39 (1980). Vince, R., Daluge, S., and Wadd, W. B.. J. Med. Chem. 14. 35 (1971). Lienhard, G. E., Science 180, 149 (1973). Douglas, K. T., and Navadi, I. N., FEBS Lett. 106, 393 (1979). ho, M., Okabe. K., and Omura. H., J. Nutr. Sci. Vitaminol. 22, 53 (1976). Elford. H. L., Wampler, G. L., and van’t Riet. B., Cancer Res. 39, 844 (1979).