- Email: [email protected]
Inhibition of bovine dihydrofolate reductase and enhancement of methotrexate sensitivity by N4-(2-acetoxyethoxymethyl)-2acetylpyridine thiosemicarbazone
81
Biochimida et Biophysica Acta, 1034 (1990) 81-85 Elsevier BBAGEN 23292
Inhibition of bovine dihydrofolate reductase and enhancement of methotrexate sensitivity by N4-(2-acetoxyethoxymethyl)-2-acetylpyridinethiosemicarbazone Evelyne Lebrun 1, Yongxue Tu 1, Roland van Rapenbusch 1, All R. Banijamali 2 and William O. Foye 2 1 NMR Laboratory, Department of Organic Chemistry (CNRS URA 464), University of Haute Normandie, Mont Saint-Aignan (France) and 2 Massachusetts College of Pharmacy and A flied Health Sciences, Boston, MA (U. S.A.)
Key words: Thiosemicarbazone; Dihydrofolate reductase; Activity modulation; Methotrexate activity enhancement
N4-(2-Acetoxyethoxymethyl)-2-acetylpyHdine thiosemicarbazone (AATSC) belongs to a series of molecules known to have broad antimicrobial inhibitory activity. These molecules contain the 2-acetoxyethoxy moiety which could conceivably take up a conformation analogous to that of the ribosyl group. Moreover, the thiosemicarbazone moiety, when in the presence of a suitable enzymatic site, could mimic the triazine group, which is found in a number of antifolate drugs. AATSC, which has beth bacterial inhibitory activity and water solubility, was accordingly evaluated for its antifolate activity against the bovine liver dihydrofolate reductase. AATSC is shown to be a fully uncompetitive inhibitor of that enzyme. Furthermore, AATSC enhances the activity of methotrexate. Such a potentiation could be useful for therapeutic purposes.
Introduction N4-(2-Acetoxyethoxymethyl)-2-acetylpyridine thiosemicarbazone (AATSC) was shown [1] to have, among a fairly large number of thiosemicarbazones, the highest inhibitory activity against the growth of the following series of microorganisms: Staphylococcus aureus (a Gram-positive bacterium), Escherichia coli and Pseudomonas aeruginosa (two Gram-negative bacteria), Candida albicans (a yeast) and Aspergillus niger (a mold). It also had activity against resistant strains of W-2 lndochina Plasmodium falciparum and D-6 African Plasmodium falciparum. The molecular targets of AATSC are not known at present. Various mechanisms, such as inhibition of ribonucleotide reductase or DNA binding, have been suggested to explain the inhibitory activity of thiosemicarbazones. AATSC is composed of two open chains (the 2acetoxyethoxymethyl and the thiosemicarbazone moieties) which are open chain analogues of ribosyl and
triazine groups, respectively. The standard interpretation of a competitive inhibition of an enzyme involves a flexibility of the inhibitor which could adopt, at the active center of the molecule, a conformation similar to that of the substrates or cofactors. Fortunately, the triazine group, which could be conceptually mimicked by AATSC, is found as one of the four major classes of inhibitor [2] of the dihydrofolate reductases. These enzymes (5,6,7,8-tetrahydrofolate NADP + oxidoreductases, EC 1.5.1.3) catalyze the reduction of dihydrofolate to tetrahydrofolate. They have received considerable attention due to the fact that their inhibitors are used for control of several disease states, as antitumor drugs or as antimicrobial or antimalarial agents. The cyclized inhibitors, including triazines, generally compete with the natural DHF substrate, as they mimic its structure. In this way, AATSC was considered to be a potential inhibitor of DHFR. Furthermore, AATSC
Abbreviations: AATSC, N4-(2-acetoxyethoxymethyl)-2-acetylpyridine thiosemicarbazone; DHFR, dihydrofolate reductase; DHF, dihydrofolic acid; MTX, methotrexate; DMSO, dimethyl sulfoxide.
Correspondence: E. Lebrun, NMR Laboratory, Department of Organic Chemistry (CNRS URA 464), University of Haute Normandie, B.P. 118, 76134 Mont Saint-Aignan, France. 0304-4165/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
~
CH3 S I H 11 C~-N~N~C..~NH / i CH31OCH2CH2 0 ~ H 2 Scheme I.
82 was also expected, because of its 2-acetoxyethoxy moiety, to be a possible competitor of the ribosyl group of the coenzyme (NADPH). There, we present the chemical synthesis, 1H-NMR resonance characteristics and biological evaluation of AATSC using D H F R from bovine liver. Material and Methods
Biochemicals Bovine DHFR, D H F and MTX were purchased from Sigma, St Louis, MO, U.S.A. The purity of D H F and MTX was checked by I H - N M R analysis. The bovine D H F R enzyme was tested as described below under Enzyme assays. Chemical synthesis of AA TSC N4-(2-Acetoxyethoxymethyl) thiosemicarbazide (1.04 g, 5 mmol) was added to a solution of 2-acetylpyridine (0.60 g, 5 mmol) was added to a solution of 2-acetylpyridine (0.60 g, 5 mmol) in 10 ml of ethanol. The mixture was stirred with reflurdng for 45 min and cooled in an ice bath. The pale green precipitate was filtered, recrystallized from ethanol and dried at 60 ° C under reduced pressure, giving 1.20 g (77%); m.p. 134135 ° C. Analysis (C13HlsN403S) C,H,N. ~H-NMR spectra High-resolution aH-NMR spectra were run at 297 K on a Bruker WM 500 spectrometer operating at 500 MHz and upgraded with an Aspect 3000 computer. The solutions were 5 m M in DMSO-d 6. The field strength of the various proton resonances is expressed as parts per million; peak multiplicity is given by the spectrometer. The intensity of peaks is depicted, below in the Results as follows: vs, very strong; s, strong; m, medium; w, weak. Enzyme assays The bovine D H F R was assayed spectrophotometrically using the decrease in absorbance at 340 nm that occurs when N A D P H and D H F are converted to N A D P and tetrahydrofolate. Velocities were derived from the change in absorbance continuously recorded (c = 12300 M -1. cm -a for both N A D P H and D H F at 340 nm as reported [3]). Ultraviolet spectra were obtained from a PYE U N I C A M SPS250 spectrophotometer thermostated at 30°C. In order to take into account the spontaneous decrease of the absorbance of N A D P H and D H F (both were found to have, under our experimental conditions, a half-life of around 2 h), differential records (with a blank cell containing N A D P H and D H F ) were obtained. The standard assay buffer, in a volume of 3 ml,
contained 50 m M phosphate buffer (pH 5.8) with 600 mM KC1. Bovine D H F R (final concentration: around 0.5 nM as determined by the method of Dixon [4]), N A D P H and inhibitors (MTX or AATSC) were added to buffer and then incubated at 30 ° C for 5 min. The reaction was initiated by adding DHF. No time-dependent changes in velocity were observed over the 5-8 min reaction. These steady-state velocities were measured and used to determine the values of the constants of inhibition 150 and g i. The determination of 15o values was engaged only in the preliminary screening among the series [1] of thiosemicarbazones. These 150 values were simply determined from plots of percent inhibition versus the logarithm of inhibitor concentration. In experiments in which MTX and AATSC were used simultaneously as inhibitors, MTX was added first and incubated as above. Thereafter, AATSC was introduced into the buffer and was again incubated at 30 ° C for 5 rain before initiation of the reaction by DHF.
Control experiments The p H activity profile of bovine liver D H F R in phosphate-KC1 buffer was diphasic between pH 4.2 and pH 8.4 with modes located at p H 5.8 and pH 7.2. The sharp maximum centered at pH 5.8 was chosen for kinetic analyses. In the control analyses of bovine D H F R , the data from the double-reciprocal plots of the enzyme (initial velocities- ~ versus substrates concentrations- 1) were computer fitted to the sequential Bi-Bi rate equation of Cleland [5]. The D H F concentrations were varied from 1 ~tM to 48/~M (at a constant N A D P H concentration: 96/~M) and those of N A D P H were varied from 1 #M to 96 t~M (at a constant D H F concentration: 48 /~M). The Michaelis constants K M for N A D P H and D H F were calculated to be equal to 3.1/~M ± 0.3 # M and 2.9 /~M + 0.3 #M, respectively. Consequently, inhibition experiments were performed with N A D P H present at a concentration of 30-times its apparent Michaelis constant. So the inhibition studies on MTX and AATSC are related to the D H F R - N A D P H binary complex and not to the free enzyme: their conclusions are irrespective of whether the reaction of bovine D H F R has an ordered or random mechanism. At an N A D P H concentration equal to 96 /~M, the plot of the reciprocal of the steady-state velocities versus the reciprocal of the D H F (from 1 to 48 /~M) at four MTX concentrations (0, 5, 10 and 20 nM) exhibited the well-characterized [6] competitive inhibition (not shown here) for such a tight-binding inhibitor. The methotrexate inhibition data were then fitted to the MichaelisMenten equation for competitive inhibition and analyzed according to Spector and Hajian [7]. A value equal to 1.6 nM was obtained for the inhibition constant Ki of bovine D H F R by MTX. This results agree with the expected value [8].
83 Results
575 "
Among the series [1] of 20 thiosemicarbazones, AATSC was found, under our experimental conditions to have the best inhibitory/so value with respect to the bovine DHFR. All other compounds of that series appeared to have lesser inhibitory properties by a factor of at least 10.
/1 4 7 5 V~ 3 7 5 o
275
i t y
175
o
75
1H-NMR characterization and behavior of A A T S C towards substrates The 1H-NMR spectrum (DMSO-d 6 at 297 K) was unambiguously assigned: 6ppm 1.998 (1, vs, CHACO); 2.415 (1, vs, CH3CN); 3.743 (5, s, OCH2); 4.126 (4, s, CH2OCO); 5.108 (2, m, CH2N); 7.410 (7, m, Hs~); 7.837 (7, m, H4~); 8.479 (2, m, H2~); 8.593 (2, m, H6~); 9.321 (3, w, NHCS) and 10.657 (1, w, NNH). Besides peaks at 6ppm: 10.657, 9.321, 5.108 and 2.415, extra resonances appeared at 10.555 (singlet), 9.275 (triplet), 5.066 (doublet) and 2.412 (singlet), respectively. They were interpreted as showing the presence of cis-trans-isomers in the ratio of 1 : 4. The trans-isomer is the most populated. No perturbation (chemical shift or profile modification) of the spectral lines of the 1-D 1H-NMR spectrum of AATSC was observed following the addition of NADPH or DHF (both at a concentration of 150 #M) to an initial sample of AATSC (150/~M) dissolved in a H 2 0 / D 2 0 (96:4) phosphate buffer 50 mM (pH 5.8) with KC1 600 mM. For this reason we discarded the hypothesis of a direct interaction between AATSC and the substrates of the bovine DHFR. Kinetics of inhibition of bovine DHFR by AA TSC The Lineweaver-Burk double reciprocal plots (Fig. 1), at different AATSC concentrations, yielded a series
5501
0
1
2
3 4 5 6 AATSC C o n c e n t r a t i o n {tiM)
+
OHF:48~M
+
DHF:12pM
~
DHF:6wM
+
DHF:31JM
--~-
OHF:21JM
+
DHF:IpM
7
+
8
DHF:4pM
Fig. 2. Uncompetitive inhibition of the bovine D H F R assay by AATSC (Dixon plots). The reciprocal of the steady-state velocity (103 m i n . v M -1) was plotted versus the AATSC concentration ( 0 - 8 / t M ) at various D H F concentrations (1-48 /LM) with constant N A D P H concentration (96/~M).
of parallel straight fines showing that the compound could be defined [9] kinetically as an uncompetitive inhibitor. The inhibition appeared as reversible as judged by the possibility to recovering the free enzyme with its initial characteristics after extensive dialysis and freezedrying. When replotting the steady-state velocities -1 data versus the AATSC concentrations (Fig. 2), we obtained again the pattern of parallel fines expected from an uncompetitive inhibitor. Such an uncompetitive inhibition kinetics could also be observed in the very special case of mixed inhibition [10] where the ratio of the dissociation constants for inhibitor binding to enzyme and enzyme-substrate complex is equal to the ratio of breakdown rates of these complexes when inhibitor is bound to give products. We rejected this special case with help of the profile of the graph of Fig. 2 because in such an eventuality the reciprocal of the velocity is not linearly dependent on the inhibitor concentration.
500 ~ 1 /
450 400. 350 -
300
250
Ve 300 t 250 o 200 150 yt I00
I e
50 0
r
-1000
-soo -600 -400 -200
0
200
400
600
800
looo
t / (oraF)
200 -
150
c e
pt
100 i
--O-
AATF,C:01JM ~
AAYSC:IIJM
--£--
AAI'SC:4uM
AA'rSC:StJM +
+
+
AAZSC:2pM +
AA~'SC:31JM
A.,~rsc:6pM
A&lrSC:71~M
"+-
Fig. 1. Uncompetitive inhibition of the bovine D H F R assay by AATSC (Lineweaver-Burk double reciprocal plots). The reciprocal of the steady-state velocity (103 m i n . ~ M -1) was plotted versus the reciprocal of the D H F concentration (1-48 /~M), expressed as 103 /~M -1, at various AATSC concentrations (0-7 t~M). The N A D P H concentration was maintained constant (96/~M).
50-
0 - -4
-3
-2
-1
0 1 2 3 AATSC C o n c e n t r a t i o n
4 5 (IJM)
6
7
8
Fig. 3. Uncompetitive inhibition of the bovine D H F R assay by AATSC (replot of intercepts on vertical axis of Fig. 1 versus AATSC concentration). The intercept on the base line is equal to - K i.
84 2000-
1 / V •
I
o c i t Y
1500 -
1000 -
500
r
i
0
1
--
2
4
r
i
i
3
4
5
6
M T X Concentration A A T S C : 0 MM Fig.
4.
Enhancement
of
the
~ MTX
i 8
(nM)
A A T S C : 1 I.IM inhibition
-7
---W- A A T S C : 2 IJM of
bovine
DHFR
by
AATSC. The reciprocalof the steady-statevelocity(10 3 min-gM -1) was plotted versus the total MTX concentration (0-8 nM) at two AATSC concentrations (1 and 2 gM). DHF concentration, 3 #M; NADPH concentration, 96 I~M. At the bovine DHFR concentration used (0.5 nM), the free (unbound) MTX is, in fact, generallyclose to that of the total MTX concentration.
Furthermore, the case of a partially uncompetitive inhibition (where inhibitor is released from the ternary complex at the same time as the products of reaction) could be also discarded: such a mechanism would display [9] on the double reciprocal plot, at various inhibitor concentrations, a set of convergent straight lines (not seen in Fig. 1). Moreover, this partially uncompetitive case gives rise to a curved graph when replotting the intercepts of Fig. 1 versus AATSC concentrations: this is not seen in Fig. 3 where the graph appears as a straight line. Finally, the steady-state data velocities were fitted [11] to equations for a fully uncompetitive inhibition. AATSC appeared as a weak binding inhibitor: K i = 2.9 + 0.3 #M when varying D H F concentration at constant N A D P H level. Conversely, at constant D H F concentration (48 #M), but varying that of N A D P H (from 1 to 96 #M) the inhibition of bovine D H F R by AATSC also appeared (not shown here) as a fully uncompetitive one and the K i of AATSC by was calculated to be equal to 3.3 M _ 0.3 # M with respect to NADPH.
Enhancement of M T X sensitivity by AA TSC When AATSC was added to the standard assay buffer together with MTX, before starting the reaction with DHF, the inhibition of bovine D H F R by MTX was enhanced (Fig. 4),
as to the possibility of a hypothetical competition either with the catalytic ribosyl site of the nicotinamide mononucleotide ribose or that of the adenine mononucleotide ribose (or both) of the N A D P H . We also questioned the analogous possibility of hypothetical mimicry of a triazine group by the thiosemicarbazone moiety of AATSC. Such hypotheses of direct attachments to the catalytic sites of the enzyme appear to be untenable due to the uncompetitive type of inhibition observed with AATSC. As MTX is a strong inhibitor of D H F R , the nonlinear shape of the graphs in the Fig. 4, when AATSC is present, reinforces the hypothesis that the binding of AATSC takes place at a suitable site different from the site of MTX and D H F location. An extra-binding site is enough to explain, the decreased amount of enzyme via a conformational change of the enzyme from the breakdown operation as stated by Cleland [5]. Such an extra-site, whose functionality has yet to be demonstrated clearly, is suggested by observations based upon distance geometry comparisons [13] or cross-reactions of antiserums [14] to peptides which do not include any active-site residue in the human D H F R . Our experiments indeed suggest that AATSC binds the N A D P H D H F R binary complex and also the MTX-NADPH-enzyme ternary complex. However, further investigations are necessary to demonstrate the possibility of direct binding of AATSC with the free enzyme. When AATSC is used together with MTX, the nonlinearity of the kinetics suggests the involvement of some cooperative effect on the inhibition by MTX. The enhancement of MTX activity by AATSC could be indicative of the location of the interacting site of AATSC inside bovine D H F R . The enhancement of MTX activity was reinvestigated [8] with the D H F R from L1210/R6, prepared from an MTX-resistant subline (R6) of L1210 mouse leukemia cells. An enhancement of MTX activity was obtained, in particular, with SH-modifying agents such as 5,5'-dithiobis(2-nitrobenzoic acid). The suggested mechanism for activation of MTX activity was recently confirmed [15] with experiments of site-directed mutagenesis: this mechanism was related to the single cysteine located within the NHE-terminal region of the eukaryotic enzymes. It was considered that cysteine could control, via an H-bond, the dynamics of the so-called fl-strand A [16,17] which contributes to form the ration of the catalytic site of the enzyme. Such a remote mechanism could also be imagined for a thiosemicarbazone such as AATSC. In order to locate precisely the building site of AATSC in bovine D H F R , N M R experiments of molecular interaction are in progress.
Discussion and Conclusions Acknowledgment Since the acetoxyethoxymethyl moiety of AATSC, already used in antiviral compounds, was reported [12] to be an open chain analog of ribose, the question arose
The authors are grateful for support by the John R. and Marie K. Sawyer Memorial Fund, M.C.P.A.H.S.
85
References 1 Foye, W.O., Banijamali, A.R. and Patarapanich, C.J. (1986) Pharm. Sci. 75, 1180-1184. 2 Blaney, J.M., Hansch, C., Silipo, C. and Vittoria, A. (1984) Chem. Rev. 84, 333-385. 3 Hillcoat, B., Nixon, P. and Blakley, R.L. (1967) Anal. Biochem. 21, 178-182. 4 Dixon, M. (1972) Biochem. J. 129, 197-213. 5 Cleland, W.W. (1979) Methods Enzymol. 63A, 103-138. 6 Morrison, J.F. (1969) Biochim. Biophys. Acta 183, 269-286. 7 Spector, T. and Hajian, G. (1986) Anal. Biochem. 115, 403-409. 8 Dully, T.H., Sato, J.K., Vitols, K.S. and Huennekens, F.M. (1986) Adv. Enzyme Regul., 24, 13-21. 9 Dixon, M. and Webb, E.C. (1979) Enzymes, 3rd FAn., Longmans Green & Co, London and Academic Press, New York, pp. 341-343.
10 Frieden, C.J. (1964) Biol. Chem. 239, 3522-3527. 11 Williams, J.W., Morrison, J.F. and Duggleby, R.G. (1979) Biochemistry 18, 2567-2573. 12 Kelley, J.L., Kelsey, J.E., Hall, W.R., Krochmal, M.P. and Scharffer, H.J. (1981) J. Med. Chem. 24, 753-759. 13 Tu, Y.X., Lebrun, E., Bon, F. Liu, Z.J. and Van Rapenbusch, R. (1988) C. R. Acad. Sci. Paris, III, 385-389. 14 Ratnam, M., Tan, X., Prendergast, N.J., Smith, P.L. and Freisheim, J.H. (1988) Biochemistry 27, 4800-4804. 15 Prendergast, N.J., Delcamp, T.J., Smith, P.L. and Freisheim, J.H. (1988) Biochemistry 27, 3664-3671. 16 Matthews, D.A., Bolin, J.T., Burridge, J.M., Filman, D.J., Volz, K.W., Kanfman, B.T., BeddeU, C.R., Champness, J.N., Stammers, D.K. and Kraut, J. (1985) J. Biol. Chem. 260, 381-391. 17 Matthews, D.A., Bolin, J.T., Burridge, J.M., Filman, D.J., Volz, K.W. and Kraut, J. (1985) J. Biol. Chem. 260, 392-403.
Biochimida et Biophysica Acta, 1034 (1990) 81-85 Elsevier BBAGEN 23292
Inhibition of bovine dihydrofolate reductase and enhancement of methotrexate sensitivity by N4-(2-acetoxyethoxymethyl)-2-acetylpyridinethiosemicarbazone Evelyne Lebrun 1, Yongxue Tu 1, Roland van Rapenbusch 1, All R. Banijamali 2 and William O. Foye 2 1 NMR Laboratory, Department of Organic Chemistry (CNRS URA 464), University of Haute Normandie, Mont Saint-Aignan (France) and 2 Massachusetts College of Pharmacy and A flied Health Sciences, Boston, MA (U. S.A.)
Key words: Thiosemicarbazone; Dihydrofolate reductase; Activity modulation; Methotrexate activity enhancement
N4-(2-Acetoxyethoxymethyl)-2-acetylpyHdine thiosemicarbazone (AATSC) belongs to a series of molecules known to have broad antimicrobial inhibitory activity. These molecules contain the 2-acetoxyethoxy moiety which could conceivably take up a conformation analogous to that of the ribosyl group. Moreover, the thiosemicarbazone moiety, when in the presence of a suitable enzymatic site, could mimic the triazine group, which is found in a number of antifolate drugs. AATSC, which has beth bacterial inhibitory activity and water solubility, was accordingly evaluated for its antifolate activity against the bovine liver dihydrofolate reductase. AATSC is shown to be a fully uncompetitive inhibitor of that enzyme. Furthermore, AATSC enhances the activity of methotrexate. Such a potentiation could be useful for therapeutic purposes.
Introduction N4-(2-Acetoxyethoxymethyl)-2-acetylpyridine thiosemicarbazone (AATSC) was shown [1] to have, among a fairly large number of thiosemicarbazones, the highest inhibitory activity against the growth of the following series of microorganisms: Staphylococcus aureus (a Gram-positive bacterium), Escherichia coli and Pseudomonas aeruginosa (two Gram-negative bacteria), Candida albicans (a yeast) and Aspergillus niger (a mold). It also had activity against resistant strains of W-2 lndochina Plasmodium falciparum and D-6 African Plasmodium falciparum. The molecular targets of AATSC are not known at present. Various mechanisms, such as inhibition of ribonucleotide reductase or DNA binding, have been suggested to explain the inhibitory activity of thiosemicarbazones. AATSC is composed of two open chains (the 2acetoxyethoxymethyl and the thiosemicarbazone moieties) which are open chain analogues of ribosyl and
triazine groups, respectively. The standard interpretation of a competitive inhibition of an enzyme involves a flexibility of the inhibitor which could adopt, at the active center of the molecule, a conformation similar to that of the substrates or cofactors. Fortunately, the triazine group, which could be conceptually mimicked by AATSC, is found as one of the four major classes of inhibitor [2] of the dihydrofolate reductases. These enzymes (5,6,7,8-tetrahydrofolate NADP + oxidoreductases, EC 1.5.1.3) catalyze the reduction of dihydrofolate to tetrahydrofolate. They have received considerable attention due to the fact that their inhibitors are used for control of several disease states, as antitumor drugs or as antimicrobial or antimalarial agents. The cyclized inhibitors, including triazines, generally compete with the natural DHF substrate, as they mimic its structure. In this way, AATSC was considered to be a potential inhibitor of DHFR. Furthermore, AATSC
Abbreviations: AATSC, N4-(2-acetoxyethoxymethyl)-2-acetylpyridine thiosemicarbazone; DHFR, dihydrofolate reductase; DHF, dihydrofolic acid; MTX, methotrexate; DMSO, dimethyl sulfoxide.
Correspondence: E. Lebrun, NMR Laboratory, Department of Organic Chemistry (CNRS URA 464), University of Haute Normandie, B.P. 118, 76134 Mont Saint-Aignan, France. 0304-4165/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
~
CH3 S I H 11 C~-N~N~C..~NH / i CH31OCH2CH2 0 ~ H 2 Scheme I.
82 was also expected, because of its 2-acetoxyethoxy moiety, to be a possible competitor of the ribosyl group of the coenzyme (NADPH). There, we present the chemical synthesis, 1H-NMR resonance characteristics and biological evaluation of AATSC using D H F R from bovine liver. Material and Methods
Biochemicals Bovine DHFR, D H F and MTX were purchased from Sigma, St Louis, MO, U.S.A. The purity of D H F and MTX was checked by I H - N M R analysis. The bovine D H F R enzyme was tested as described below under Enzyme assays. Chemical synthesis of AA TSC N4-(2-Acetoxyethoxymethyl) thiosemicarbazide (1.04 g, 5 mmol) was added to a solution of 2-acetylpyridine (0.60 g, 5 mmol) was added to a solution of 2-acetylpyridine (0.60 g, 5 mmol) in 10 ml of ethanol. The mixture was stirred with reflurdng for 45 min and cooled in an ice bath. The pale green precipitate was filtered, recrystallized from ethanol and dried at 60 ° C under reduced pressure, giving 1.20 g (77%); m.p. 134135 ° C. Analysis (C13HlsN403S) C,H,N. ~H-NMR spectra High-resolution aH-NMR spectra were run at 297 K on a Bruker WM 500 spectrometer operating at 500 MHz and upgraded with an Aspect 3000 computer. The solutions were 5 m M in DMSO-d 6. The field strength of the various proton resonances is expressed as parts per million; peak multiplicity is given by the spectrometer. The intensity of peaks is depicted, below in the Results as follows: vs, very strong; s, strong; m, medium; w, weak. Enzyme assays The bovine D H F R was assayed spectrophotometrically using the decrease in absorbance at 340 nm that occurs when N A D P H and D H F are converted to N A D P and tetrahydrofolate. Velocities were derived from the change in absorbance continuously recorded (c = 12300 M -1. cm -a for both N A D P H and D H F at 340 nm as reported [3]). Ultraviolet spectra were obtained from a PYE U N I C A M SPS250 spectrophotometer thermostated at 30°C. In order to take into account the spontaneous decrease of the absorbance of N A D P H and D H F (both were found to have, under our experimental conditions, a half-life of around 2 h), differential records (with a blank cell containing N A D P H and D H F ) were obtained. The standard assay buffer, in a volume of 3 ml,
contained 50 m M phosphate buffer (pH 5.8) with 600 mM KC1. Bovine D H F R (final concentration: around 0.5 nM as determined by the method of Dixon [4]), N A D P H and inhibitors (MTX or AATSC) were added to buffer and then incubated at 30 ° C for 5 min. The reaction was initiated by adding DHF. No time-dependent changes in velocity were observed over the 5-8 min reaction. These steady-state velocities were measured and used to determine the values of the constants of inhibition 150 and g i. The determination of 15o values was engaged only in the preliminary screening among the series [1] of thiosemicarbazones. These 150 values were simply determined from plots of percent inhibition versus the logarithm of inhibitor concentration. In experiments in which MTX and AATSC were used simultaneously as inhibitors, MTX was added first and incubated as above. Thereafter, AATSC was introduced into the buffer and was again incubated at 30 ° C for 5 rain before initiation of the reaction by DHF.
Control experiments The p H activity profile of bovine liver D H F R in phosphate-KC1 buffer was diphasic between pH 4.2 and pH 8.4 with modes located at p H 5.8 and pH 7.2. The sharp maximum centered at pH 5.8 was chosen for kinetic analyses. In the control analyses of bovine D H F R , the data from the double-reciprocal plots of the enzyme (initial velocities- ~ versus substrates concentrations- 1) were computer fitted to the sequential Bi-Bi rate equation of Cleland [5]. The D H F concentrations were varied from 1 ~tM to 48/~M (at a constant N A D P H concentration: 96/~M) and those of N A D P H were varied from 1 #M to 96 t~M (at a constant D H F concentration: 48 /~M). The Michaelis constants K M for N A D P H and D H F were calculated to be equal to 3.1/~M ± 0.3 # M and 2.9 /~M + 0.3 #M, respectively. Consequently, inhibition experiments were performed with N A D P H present at a concentration of 30-times its apparent Michaelis constant. So the inhibition studies on MTX and AATSC are related to the D H F R - N A D P H binary complex and not to the free enzyme: their conclusions are irrespective of whether the reaction of bovine D H F R has an ordered or random mechanism. At an N A D P H concentration equal to 96 /~M, the plot of the reciprocal of the steady-state velocities versus the reciprocal of the D H F (from 1 to 48 /~M) at four MTX concentrations (0, 5, 10 and 20 nM) exhibited the well-characterized [6] competitive inhibition (not shown here) for such a tight-binding inhibitor. The methotrexate inhibition data were then fitted to the MichaelisMenten equation for competitive inhibition and analyzed according to Spector and Hajian [7]. A value equal to 1.6 nM was obtained for the inhibition constant Ki of bovine D H F R by MTX. This results agree with the expected value [8].
83 Results
575 "
Among the series [1] of 20 thiosemicarbazones, AATSC was found, under our experimental conditions to have the best inhibitory/so value with respect to the bovine DHFR. All other compounds of that series appeared to have lesser inhibitory properties by a factor of at least 10.
/1 4 7 5 V~ 3 7 5 o
275
i t y
175
o
75
1H-NMR characterization and behavior of A A T S C towards substrates The 1H-NMR spectrum (DMSO-d 6 at 297 K) was unambiguously assigned: 6ppm 1.998 (1, vs, CHACO); 2.415 (1, vs, CH3CN); 3.743 (5, s, OCH2); 4.126 (4, s, CH2OCO); 5.108 (2, m, CH2N); 7.410 (7, m, Hs~); 7.837 (7, m, H4~); 8.479 (2, m, H2~); 8.593 (2, m, H6~); 9.321 (3, w, NHCS) and 10.657 (1, w, NNH). Besides peaks at 6ppm: 10.657, 9.321, 5.108 and 2.415, extra resonances appeared at 10.555 (singlet), 9.275 (triplet), 5.066 (doublet) and 2.412 (singlet), respectively. They were interpreted as showing the presence of cis-trans-isomers in the ratio of 1 : 4. The trans-isomer is the most populated. No perturbation (chemical shift or profile modification) of the spectral lines of the 1-D 1H-NMR spectrum of AATSC was observed following the addition of NADPH or DHF (both at a concentration of 150 #M) to an initial sample of AATSC (150/~M) dissolved in a H 2 0 / D 2 0 (96:4) phosphate buffer 50 mM (pH 5.8) with KC1 600 mM. For this reason we discarded the hypothesis of a direct interaction between AATSC and the substrates of the bovine DHFR. Kinetics of inhibition of bovine DHFR by AA TSC The Lineweaver-Burk double reciprocal plots (Fig. 1), at different AATSC concentrations, yielded a series
5501
0
1
2
3 4 5 6 AATSC C o n c e n t r a t i o n {tiM)
+
OHF:48~M
+
DHF:12pM
~
DHF:6wM
+
DHF:31JM
--~-
OHF:21JM
+
DHF:IpM
7
+
8
DHF:4pM
Fig. 2. Uncompetitive inhibition of the bovine D H F R assay by AATSC (Dixon plots). The reciprocal of the steady-state velocity (103 m i n . v M -1) was plotted versus the AATSC concentration ( 0 - 8 / t M ) at various D H F concentrations (1-48 /LM) with constant N A D P H concentration (96/~M).
of parallel straight fines showing that the compound could be defined [9] kinetically as an uncompetitive inhibitor. The inhibition appeared as reversible as judged by the possibility to recovering the free enzyme with its initial characteristics after extensive dialysis and freezedrying. When replotting the steady-state velocities -1 data versus the AATSC concentrations (Fig. 2), we obtained again the pattern of parallel fines expected from an uncompetitive inhibitor. Such an uncompetitive inhibition kinetics could also be observed in the very special case of mixed inhibition [10] where the ratio of the dissociation constants for inhibitor binding to enzyme and enzyme-substrate complex is equal to the ratio of breakdown rates of these complexes when inhibitor is bound to give products. We rejected this special case with help of the profile of the graph of Fig. 2 because in such an eventuality the reciprocal of the velocity is not linearly dependent on the inhibitor concentration.
500 ~ 1 /
450 400. 350 -
300
250
Ve 300 t 250 o 200 150 yt I00
I e
50 0
r
-1000
-soo -600 -400 -200
0
200
400
600
800
looo
t / (oraF)
200 -
150
c e
pt
100 i
--O-
AATF,C:01JM ~
AAYSC:IIJM
--£--
AAI'SC:4uM
AA'rSC:StJM +
+
+
AAZSC:2pM +
AA~'SC:31JM
A.,~rsc:6pM
A&lrSC:71~M
"+-
Fig. 1. Uncompetitive inhibition of the bovine D H F R assay by AATSC (Lineweaver-Burk double reciprocal plots). The reciprocal of the steady-state velocity (103 m i n . ~ M -1) was plotted versus the reciprocal of the D H F concentration (1-48 /~M), expressed as 103 /~M -1, at various AATSC concentrations (0-7 t~M). The N A D P H concentration was maintained constant (96/~M).
50-
0 - -4
-3
-2
-1
0 1 2 3 AATSC C o n c e n t r a t i o n
4 5 (IJM)
6
7
8
Fig. 3. Uncompetitive inhibition of the bovine D H F R assay by AATSC (replot of intercepts on vertical axis of Fig. 1 versus AATSC concentration). The intercept on the base line is equal to - K i.
84 2000-
1 / V •
I
o c i t Y
1500 -
1000 -
500
r
i
0
1
--
2
4
r
i
i
3
4
5
6
M T X Concentration A A T S C : 0 MM Fig.
4.
Enhancement
of
the
~ MTX
i 8
(nM)
A A T S C : 1 I.IM inhibition
-7
---W- A A T S C : 2 IJM of
bovine
DHFR
by
AATSC. The reciprocalof the steady-statevelocity(10 3 min-gM -1) was plotted versus the total MTX concentration (0-8 nM) at two AATSC concentrations (1 and 2 gM). DHF concentration, 3 #M; NADPH concentration, 96 I~M. At the bovine DHFR concentration used (0.5 nM), the free (unbound) MTX is, in fact, generallyclose to that of the total MTX concentration.
Furthermore, the case of a partially uncompetitive inhibition (where inhibitor is released from the ternary complex at the same time as the products of reaction) could be also discarded: such a mechanism would display [9] on the double reciprocal plot, at various inhibitor concentrations, a set of convergent straight lines (not seen in Fig. 1). Moreover, this partially uncompetitive case gives rise to a curved graph when replotting the intercepts of Fig. 1 versus AATSC concentrations: this is not seen in Fig. 3 where the graph appears as a straight line. Finally, the steady-state data velocities were fitted [11] to equations for a fully uncompetitive inhibition. AATSC appeared as a weak binding inhibitor: K i = 2.9 + 0.3 #M when varying D H F concentration at constant N A D P H level. Conversely, at constant D H F concentration (48 #M), but varying that of N A D P H (from 1 to 96 #M) the inhibition of bovine D H F R by AATSC also appeared (not shown here) as a fully uncompetitive one and the K i of AATSC by was calculated to be equal to 3.3 M _ 0.3 # M with respect to NADPH.
Enhancement of M T X sensitivity by AA TSC When AATSC was added to the standard assay buffer together with MTX, before starting the reaction with DHF, the inhibition of bovine D H F R by MTX was enhanced (Fig. 4),
as to the possibility of a hypothetical competition either with the catalytic ribosyl site of the nicotinamide mononucleotide ribose or that of the adenine mononucleotide ribose (or both) of the N A D P H . We also questioned the analogous possibility of hypothetical mimicry of a triazine group by the thiosemicarbazone moiety of AATSC. Such hypotheses of direct attachments to the catalytic sites of the enzyme appear to be untenable due to the uncompetitive type of inhibition observed with AATSC. As MTX is a strong inhibitor of D H F R , the nonlinear shape of the graphs in the Fig. 4, when AATSC is present, reinforces the hypothesis that the binding of AATSC takes place at a suitable site different from the site of MTX and D H F location. An extra-binding site is enough to explain, the decreased amount of enzyme via a conformational change of the enzyme from the breakdown operation as stated by Cleland [5]. Such an extra-site, whose functionality has yet to be demonstrated clearly, is suggested by observations based upon distance geometry comparisons [13] or cross-reactions of antiserums [14] to peptides which do not include any active-site residue in the human D H F R . Our experiments indeed suggest that AATSC binds the N A D P H D H F R binary complex and also the MTX-NADPH-enzyme ternary complex. However, further investigations are necessary to demonstrate the possibility of direct binding of AATSC with the free enzyme. When AATSC is used together with MTX, the nonlinearity of the kinetics suggests the involvement of some cooperative effect on the inhibition by MTX. The enhancement of MTX activity by AATSC could be indicative of the location of the interacting site of AATSC inside bovine D H F R . The enhancement of MTX activity was reinvestigated [8] with the D H F R from L1210/R6, prepared from an MTX-resistant subline (R6) of L1210 mouse leukemia cells. An enhancement of MTX activity was obtained, in particular, with SH-modifying agents such as 5,5'-dithiobis(2-nitrobenzoic acid). The suggested mechanism for activation of MTX activity was recently confirmed [15] with experiments of site-directed mutagenesis: this mechanism was related to the single cysteine located within the NHE-terminal region of the eukaryotic enzymes. It was considered that cysteine could control, via an H-bond, the dynamics of the so-called fl-strand A [16,17] which contributes to form the ration of the catalytic site of the enzyme. Such a remote mechanism could also be imagined for a thiosemicarbazone such as AATSC. In order to locate precisely the building site of AATSC in bovine D H F R , N M R experiments of molecular interaction are in progress.
Discussion and Conclusions Acknowledgment Since the acetoxyethoxymethyl moiety of AATSC, already used in antiviral compounds, was reported [12] to be an open chain analog of ribose, the question arose
The authors are grateful for support by the John R. and Marie K. Sawyer Memorial Fund, M.C.P.A.H.S.
85
References 1 Foye, W.O., Banijamali, A.R. and Patarapanich, C.J. (1986) Pharm. Sci. 75, 1180-1184. 2 Blaney, J.M., Hansch, C., Silipo, C. and Vittoria, A. (1984) Chem. Rev. 84, 333-385. 3 Hillcoat, B., Nixon, P. and Blakley, R.L. (1967) Anal. Biochem. 21, 178-182. 4 Dixon, M. (1972) Biochem. J. 129, 197-213. 5 Cleland, W.W. (1979) Methods Enzymol. 63A, 103-138. 6 Morrison, J.F. (1969) Biochim. Biophys. Acta 183, 269-286. 7 Spector, T. and Hajian, G. (1986) Anal. Biochem. 115, 403-409. 8 Dully, T.H., Sato, J.K., Vitols, K.S. and Huennekens, F.M. (1986) Adv. Enzyme Regul., 24, 13-21. 9 Dixon, M. and Webb, E.C. (1979) Enzymes, 3rd FAn., Longmans Green & Co, London and Academic Press, New York, pp. 341-343.
10 Frieden, C.J. (1964) Biol. Chem. 239, 3522-3527. 11 Williams, J.W., Morrison, J.F. and Duggleby, R.G. (1979) Biochemistry 18, 2567-2573. 12 Kelley, J.L., Kelsey, J.E., Hall, W.R., Krochmal, M.P. and Scharffer, H.J. (1981) J. Med. Chem. 24, 753-759. 13 Tu, Y.X., Lebrun, E., Bon, F. Liu, Z.J. and Van Rapenbusch, R. (1988) C. R. Acad. Sci. Paris, III, 385-389. 14 Ratnam, M., Tan, X., Prendergast, N.J., Smith, P.L. and Freisheim, J.H. (1988) Biochemistry 27, 4800-4804. 15 Prendergast, N.J., Delcamp, T.J., Smith, P.L. and Freisheim, J.H. (1988) Biochemistry 27, 3664-3671. 16 Matthews, D.A., Bolin, J.T., Burridge, J.M., Filman, D.J., Volz, K.W., Kanfman, B.T., BeddeU, C.R., Champness, J.N., Stammers, D.K. and Kraut, J. (1985) J. Biol. Chem. 260, 381-391. 17 Matthews, D.A., Bolin, J.T., Burridge, J.M., Filman, D.J., Volz, K.W. and Kraut, J. (1985) J. Biol. Chem. 260, 392-403.