Abstract
This study was performed to evaluate whether selenocysteine Se-conjugates are substrates for human cysteine conjugate β-lyase enzymes. By testing kidney cytosols of three different humans, we studied interindividual differences in β-lyase enzymes in humans. A series of 22 selenocysteine Se-conjugates were tested in rat and human kidney cytosols to compare their ability to form selenol compounds by β-elimination. All compounds appeared to be good substrates for rat and human cysteine conjugate β-lyase enzymes. The β-lyase activity toward the selenocysteine Se-conjugates was comparable with those of the known nephrotoxic cysteine S-conjugateS-(2-chloro-1,1,2-trifluoroethyl)-l-cysteine in rats and humans. In rat kidney cytosol, between 22- and 877-fold higher β-elimination rates were observed compared with human kidney cytosol. Significant correlations (P < .0001) between three human kidney cytosols in β-lyase activities were found within the tested series of 22 compounds. Specific β-lyase activities and intrinsic clearances of β-elimination reactions ranged up to 3-fold, indicating that there are quantitative rather than qualitative interindividual differences in β-eliminating enzymes in humans. Furthermore, Se-alkyl selenocysteine conjugates showed a sterically dependent bioactivation to selenol compounds in humans but not in rats. The present study supports the hypothesis that selenocysteine Se-conjugates may be useful as prodrugs to target pharmacologically active selenol compounds (e.g., antitumor or chemoprotective) to the kidney in humans.
Previously, we demonstrated that selenocysteine Se-conjugates (Se-Cys conjugates) were β-eliminated at a high rate when incubated with rat renal cytosol (Andreadou et al., 1996). The β-elimination rates were up to 100-fold higher than their corresponding sulfur analogs. For this reason, these compounds may potentially be used as prodrugs to generate pharmacologically active selenols, as depicted in Fig.1. Recently, the enzyme cysteine conjugate β-lyase/glutamine transaminase K was shown to be active toward these substrates (Commandeur et al., 2000). Relatively high concentrations of β-lyase in rat kidney make these compounds promising kidney selective prodrugs (Commandeur et al., 1995). A similar renal bioactivation-targeting approach was already tested byHwang and Elfarra (1989). After dosing rats withS-(6-purinyl)-l-cysteine, the renal concentration of the metabolite 6-ercaptopurine, a known antitumor and immunosuppressant drug, was nearly 25-fold higher in kidney than in plasma and 2.3-fold higher than in liver. On the administration of S-(guanin-6-yl)-l-cysteine, a similar tissue distribution of 6-thioguanine was observed (Elfarra et al., 1995).
It has been reported that miscellaneous selenol compounds possess antitumor effects. Thus, selenopurines, like 6-selenopurine and 6-selenoguanine, for example, were shown to be effective against leukemia L5178Y, sarcoma 180, and Ehrlich ascites tumors in mice both in vivo and in vitro (for a brief review, see Shamberger, 1983).p-Methoxyphenylselenol was found to possess anticarcinogenic activity against benzo(a)pyrene-induced forestomach tumors in mice (El-Bayoumy, 1985). Furthermore, Se-methyl-l-selenocysteine was demonstrated to have anticarcinogenic activity against dimethylbenzo(a)anthracen-induced tumors in rats. The formation of methylselenol has been shown to be implicated in this activity (Ip et al., 1991).
For other organoselenium compounds, protection against toxic side effects of drugs has been observed. Thus, ebselen, a nontoxic anti-inflammatory agent, was shown to inhibit the cytotoxicity of doxorubicin in MCF-7 human breast cancer cells in vitro (Doroshow, 1986). As demonstrated by Li et al. (1994), ebselen also protected against the cytotoxicity of paracetamol in rat hepatocytes. Furthermore, ebselen and sodium selenite protected rats and mice against cisplatin-induced nephrotoxicity without interfering with its antitumor activity (Baldew et al., 1989, 1990). The mechanism of this protection is not clear, but it has been proposed that ebselen and sodium selenite are reduced by glutathione to their selenol metabolites, followed by reaction with the covalently bound reactive hydrolysis products of cisplatin (Baldew et al., 1992; Vermeulen et al., 1993). Taken together, it seems of interest to target selenol compounds selectively to the kidneys by using Se-Cys conjugates as prodrugs.
So far, no information is available on the bioactivation of Se-Cys conjugates by human renal β-lyase enzymes. It is known that human kidneys as well as human renal carcinomas do possess β-lyase activity, as demonstrated by Nelson et al. (1995). β-Lyase activity in normal human kidney tissue differed up to 14-fold between individuals withS-(2-benzothiazolyl)-l-cysteine as a substrate. Furthermore, Green et al. (1990) demonstrated thatS-(1,2,2-trichlorovinyl)-l-cysteine was β-eliminated in human kidney cytosol, although the intrinsic clearance (Vmax/Km) of β-elimination reactions was 28-fold lower compared with rat kidney cytosol. Similar results were reported by Lash et al. (1990) withS-(2-benzothiazolyl)-l-cysteine as a substrate.
To delineate whether the concept of targeting of Se-Cys conjugates to the kidney and local β-lyase-mediated bioactivation to selenols would also be applicable in humans, we tested in the present study a series of 22 Se-Cys conjugates for their ability to form selenols by measuring the amount of pyruvate in human kidney cytosols. Andreadou et al. (1996) demonstrated that formation of pyruvate is a good reflection of the amount of selenol formed. It is well known that sometimes large interindividual differences exist in enzymes likeN-acetyltransferases, cytochromes P450, and glutathione-S-transferases, as reviewed by Wormhoudt et al. (1999). These often genetically based interindividual differences in human have been associated with differences in pharmacological activity and side effects of drugs in individuals. In using a prodrug concept, it is therefore important to know whether there are quantitative and/or qualitative differences in enzyme expression of the bioactivation enzymes involved. Because of this, in the present study we also made preliminary tests of the β-elimination of Se-Cys conjugates in the kidney cytosol of three humans. By testing a large range of substrates with varying activities, we were able to screen for qualitative and quantitative differences of β-eliminating enzymes in humans. Finally, to get more insight into the substrate specificity of β-lyase enzymes, we synthesized and tested a novel series ofortho-substituted Se-phenyl- and Se-benzyl-l-selenocysteines (n = 7).
Experimental Procedures
Materials
α-Keto-γ-methiol-butyric acid (KMB) was purchased from Sigma Chemical Co. (St. Louis, MO). Di-tert-butyl dicarbonate was obtained from Fluka (Zwijndrecht, the Netherlands).o-Phenylenediamine was obtained from Janssen Chimica (Geel, Belgium). β-Chloro-l-alanine,l-selenocystine, Se-ethyl-l-selenocysteine (1), Se-(n-propyl)-l-selenocysteine (2), Se-(n-butyl)-l-selenocysteine (3), Se-phenyl-l-selenocysteine (5), Se-(p-methylphenyl)-l-selenocysteine (10), Se-(p-methylphenyl)-d-selenocysteine (11), Se-(p-chlorophenyl)-l-selenocysteine (12), Se-(p-methoxyphenyl)-l-selenocysteine (13), Se-benzyl-l-selenocysteine (14), Se-(p-methylbenzyl)-l-selenocysteine (18), Se-(p-methylbenzyl)-d-selenocysteine (19), Se-(p-chlorobenzyl)-l-selenocysteine (20), Se-(p-methoxybenzyl)-l-selenocysteine (21), and Se-(3,4-dichlorobenzyl)-l-selenocysteine (22) were prepared as described by Andreadou et al. (1996).S-(2-Chloro-1,1,2-trifluorethyl)-l-cysteine (CTFE-Cys) (23) was synthesized as described by Commandeur et al. (1988). All other chemicals were of the highest grade commercially available.
Apparatus
Gas Chromatography-Mass Spectrometry (GC-MS).
GC-MS analyses of methylated extracts were carried out on a Hewlett Packard model 5890 gas chromatograph equipped with a 25-m BPX5 column (0.22 mm i.d., 0.25 μm film thickness; SGE, Amstelveen, the Netherlands) coupled to a Hewlett Packard model MSD 5970 mass spectrometer (E.I. mode, electron energy of 70 eV). Temperatures of the injection port and transfer line were 270°C. The column temperature was programmed from 60°C (2 min) to 270°C (20°C/min) and maintained at 270°C for 10 min.
1H NMR.
1H NMR spectra were recorded on a Bruker AC 200 (200.1 MHz) spectrometer with tetramethylsilane as an internal standard.
HPLC.
Samples were analyzed on two ChromSpher C18 columns (5-μm particles, 100 × 3 mm; Chrompack, Bergen op Zoom, the Netherlands) that were eluted isocratically at a flow of 0.4 ml/min. The eluent consisted of 54% demineralized water, 45% methanol, and 1% acetic acid. Detection was accomplished on a Shimadzu fluorescence detector (model RF 530) at an excitation wavelength of 336 nm and an emission wavelength of 420 nm.
Syntheses
N-tert-Butoxycarbonyl-l-chloroalanine.
l-Chloroalanine (0.1 mol, 12.4 g) was added to a solution of 4.4 g of (0.11 mol) sodium hydroxide, 110 ml of water, and 75 ml of tert-butyl alcohol. To the well-stirred clear solution, we added di-tert-butyl dicarbonate (21.8 g, 0.1 mol) dropwise within 1 h. The reaction mixture was stirred overnight at room temperature. The pH was lowered to 1 to 1.5 by the addition of 22.4 g of (1.65 mol) potassium hydrogen sulfate in 150 ml of water. The reaction mixture was extracted four times with 40 ml of ethyl acetate. The combined organic layers were dried over anhydrous sodium sulfate, filtered, and evaporated to give a white solid. The yield was 73%, and the purity, as determined by1H NMR, was >98%: 1H NMR (CDCl3): δ (ppm) 1.5 [9H, s, C(CH3)3], 3.9 (2H, dd,CH2-CH), 4.7 (1H, m, CH2-CH); GC-MS after methylation with diazomethane: retention time 9.6 min; m/z(relative intensity, assignment) 178 (29, M⋅+-COOMe), 138 (15), 136 (17), 78 (29), 59 (60), 57 [100, C(CH3)3⋅+].
N-tert-Butoxycarbonyl-l-selenocystine.
The same procedure as for the preparation ofN-tert-butoxycarbonyl-l-chloroalanine was used with 10.9 g of (0.032 mol) selenocystine and 14.1 g of (0.065 mol) di-tert-butyl dicarbonate. Instead oftert-butyl alcohol, methanol was used as a solvent. After evaporation of the combined organic layers a yellow solid was obtained. The yield was 63%, and the purity, as determined by1H NMR, was >98%: 1H NMR (CDCl3): δ (ppm) 1.5 [9H, s, C(CH3)3] 3.5 (2H, dd,CH2-CH), 4.6 (1H, m, CH2-CH); GC-MS after reduction with sodium borohydride and methylation with diazomethane: retention time 9.4 min; m/z (relative intensity, selenium isotope, assignment) 297 (17, 80Se, M⋅+), 180 (71,80Se), 165 (23, 80Se), 138 (13, 80Se), 109 (26, 80Se), 57 [100, C(CH3)3⋅+].
o-Methyldiphenyl Diselenide.
o-Methyldiphenyl diselenide was prepared according to a method of Reich et al. (1988), using o-bromotoluene instead of bromobenzene. The yield was 86%, and the purity, as determined by1H NMR, was >98%: 1H NMR (CDCl3): δ (ppm) 2.5 (6H, s, CH3), 7.0 to 7.6 (8H, m, Ar-H); GC-MS: retention time 13.1 min; m/z (relative intensity, selenium isotope, assignment) 342 (0.5, 80Se, M⋅+), 171 (19, 80Se, Me-Ar-Se⋅+), 91 (100, Ar-Me⋅+), 65 (40).
o-Methoxydiphenyl Diselenide.
o-Methoxydiphenyl diselenide was prepared according to a method of Reich et al. (1988), using o-bromomethoxybenzene instead of bromobenzene. The yield was 87%, and the purity, as determined by 1H NMR, was >98%;1H NMR (CDCl3): δ (ppm) 3.9 (6H, s, CH3), 6.7 to 7.6 (8H, m, Ar-H); GC-MS: retention time 14.7 min; m/z (relative intensity, selenium isotope, assignment) 374 (0.6,80Se, M⋅+), 186 (11,80Se), 157 (41, 80Se, Ar-Se⋅+), 117 (20,80Se), 107 (77), 93 (46), 77 (100).
o-Chlorodiphenyl Diselenide.
o-Chlorodiphenyl diselenide was prepared according to a method of Reich et al. (1988), using o-bromochlorobenzene instead of bromobenzene. The yield was 82%, and the purity, as determined by 1H NMR, was >98%;1H NMR (CDCl3): δ (ppm) 7.0 to 7.6 (8H, m, Ar-H); GC-MS: retention time 14.3 min;m/z (relative intensity, selenium isotope, assignment) 382 (20, 80Se, M⋅+), 191 (98, 80Se, Cl-Ar-Se⋅+), 156 (100,80Se, Ar-Se⋅+), 117 (38, 80Se).
Se-Allyl-l-selenocysteine (4).
l-Selenocystine (1.5 mmol, 500 mg) was dissolved in 8 ml of 0.5 N NaOH and 2 ml of ethanol. At 0°C, 0.4 g of (15 mmol) sodium borohydride was added while the reaction mixture was stirred. The mixture was allowed to reach room temperature, during which the color of the solution changed from yellow to colorless. After cooling again to 0°C, 4 ml of 2 N NaOH and 6 mmol of allylbromide were added, and the mixture was stirred for 3 h at room temperature. Concentrated HCl was added until pH 5 to 6 at 4°C. Se-(allyl)-l-selenocysteine precipitated as a white crystalline solid. No racemization had occurred as determined by HPLC. The purity, as determined by HPLC and 1H NMR, was >98%; The 1H NMR spectrum obtained was almost identical with that ofS-allyl-l-cysteine (Freeman et al., 1994): 1H NMR (D2O, Na2CO3): δ (ppm) 2.95 to 3.22 (2H, m,CH2-CH-NH2), 3.30 (2H, d,CH2CH-CH2), 4.25 to 4.40 (1H, d of d, CH2-CH-NH2), 5.05 to 5.25 (2H, m, CH2CH-CH2), 5.80 to 6.08 (1H, m, CH2CH-CH2).
Se-(o-Methylphenyl)-l-selenocysteine HCl (6).
o-Methyldiphenyl diselenide (0.55 g, 1.6 mmol) was dissolved in 4 ml of dimethylformamide, 1 ml of water, and 0.76 g of (7.2 mmol) of sodium carbonate under nitrogen atmosphere; 0.30 g of (7.9 mmol) sodium borohydride was added to the well-stirred solution. The color of the solution changed from yellow/orange to colorless. After 30 min, 0.68 g of (3 mmol) ofN-tert-butoxycarbonyl-l-chloroalanine dissolved in 5 ml of dimethylformamide was added dropwise. The disappearance ofN-tert-butoxycarbonyl-l-chloroalanine was determined by GC-MS. The solution was stirred for 2 h at 50°C, after whichN-tert-butoxycarbonyl-l-chloroalanine had disappeared. The solution was acidified with 2 N HCl and extracted twice with ethyl acetate. The combined organic layers were washed with 2 N HCl to remove dimethylformamide, dried over sodium sulfate, and evaporated after filtration. The residue was dissolved in 7.5 ml of ethyl acetate and 7.5 ml of ethyl acetate saturated with hydrogen chloride and stirred overnight. The product precipitated as a white crystalline solid and was washed with ethyl acetate to give the pure product as described by Moore and Green (1988). No racemization had occurred as determined by HPLC. The yield was 24%, and the purity, as determined by HPLC and 1H NMR, was >98%;1H NMR (D2O, Na2CO3): δ (ppm) 2.5 (3H, s, CH3), 3.5 (2H, dd,CH2-CH), 4.3 (1H, m, CH2-CH), 7.1 to 7.4 (3H, m,m-, p-Ar-H), 7.7 (1H, d, o-Ar-H).
Se-(o-Chlorophenyl)-l-selenocysteine HCl (7).
The same procedure as for Se-(o-methylphenyl)-l-selenocysteine HCl was used with 1.6 mmol of o-chlorodiphenyl diselenide and 3 mmol ofN-tert-butoxycarbonyl-l-chloroalanine. The yield was 22%, and the purity, as determined by HPLC and1H NMR, was >98%; white crystalline solid;1H NMR (D2O, Na2CO3): δ (ppm) 3.6 (2H, d, CH2-CH), 4.4 (1H, m, CH2-CH), 7.2 to 7.4 (2H, m,m-, p-Ar-H), 7.5 (1H, d, m-Ar-H), 7.7 (1H, d, o-Ar-H).
Se-(o-Methoxyphenyl)-l-selenocysteine HCl (8).
The same procedure as for Se-(o-methylphenyl)-l-selenocysteine HCl was used with 1.6 mmol of o-methoxydiphenyl diselenide and 3 mmol ofN-tert-butoxycarbonyl-l-chloroalanine, except that the reaction was carried out in dimethylformamide and stirred for 5 h at 50°C. The yield was 39%, and the purity, as determined by HPLC and 1H NMR, was >98%; white crystalline solid; 1H NMR (D2O, Na2CO3): δ (ppm) 3.9 (3H, s, CH3), 3.4 (2H, dd,CH2-CH), 4.2 (1H, m, CH2-CH), 6.9 to 7.1 (2H, m,p-, m-Ar-H), 7.4 (1H, d, m-Ar-H), 7.6 (1H, d, o-Ar-H).
Se-(o-Nitrophenyl)-l-selenocysteine HCl (9).
N-tert-Butoxycarbonyl-l-selenocystine (0.50 g, 0.9 mmol) was dissolved in 8 ml of dimethylformamide and 0.45 g of (4.2 mmol) of sodium carbonate under nitrogen atmosphere; 0.17 g of (4.5 mmol) sodium borohydride was added to the well-stirred solution. The color of the solution changed from yellow to colorless. After 30 min, 0.32 g of (2 mmol) ofo-nitrochlorobenzene dissolved in 2 ml of dimethylformamide was added dropwise. The disappearance ofN-tert-butoxycarbonyl-l-selenocystine was determined by GC-MS. The solution was stirred for 1 h, after whichN-tert-butoxycarbonyl-l-selenocystine had disappeared. The solution was acidified with 2 N HCl and extracted twice with ethyl acetate. The combined organic layers were washed with 2 N HCl to remove dimethylformamide, dried over sodium sulfate, and evaporated after filtration. The product was purified by preparative TLC (solvent n-propanol, Rf= 0.4) and dissolved in 5 ml of ethyl acetate and 5 ml of ethyl acetate saturated with hydrogen chloride and stirred overnight. The product precipitated as a yellow crystalline solid and was washed with ethyl acetate to give the pure product as described by Moore and Green (1988). No racemization had occurred as determined by HPLC. The yield was 65%, and the purity, as determined by HPLC and1H NMR, was >98%; 1H NMR (D2O, Na2CO3): δ (ppm) 3.6 (2H, dd, CH2-CH), 4.4 (1H, m, CH2-CH), 7.5 (1H, d,m-Ar-H), 7.6 to 7.9 (2H, m, m-,p-Ar-H), 8.3 (1H, d, o-Ar-H).
Se-(o-Methylbenzyl)-l-selenocysteine HCl (15).
N-tert-Butoxycarbonyl-l-selenocystine (0.50 g, 0.9 mmol) was dissolved in 7 ml of dimethylformamide and 0.45 g of (4.2 mmol) sodium carbonate under nitrogen atmosphere; 0.17 g of (4.5 mmol) sodium borohydride was added to the well-stirred solution. The color of the solution changed from yellow to colorless. After 30 min, 0.53 g of (3.8 mmol)o-methylbenzylchloride dissolved in 1 ml of dimethylformamide was added dropwise. The disappearance ofN-tert-butoxycarbonyl-l-selenocystine was determined by GC-MS. The solution was stirred for 1.5 h, after whichN-tert-butoxycarbonyl-l-selenocystine had disappeared. The solution was acidified with 2 N HCl and extracted twice with ethyl acetate. The combined organic layers were washed with 2 N HCl to remove dimethylformamide, dried over sodium sulfate, and evaporated after filtration. The residue was dissolved in 5 ml of ethyl acetate and 5 ml of ethyl acetate saturated with hydrogen chloride and stirred overnight. The product precipitated as a white solid and was washed with ethyl acetate to give the pure product as described byMoore and Green (1988). No racemization had occurred as determined by HPLC. The yield was 41%, and the purity, as determined by HPLC and1H NMR, was >98%; white crystalline solid;1H NMR (D2O, Na2CO3): δ (ppm) 2.4 (3H, s, CH3), 3.1 (2H, dd,CH2-CH), 4.0 (2H, s, Ar-CH2-Se), 4.2 (1H, m, CH2-CH), 7.3 (4H, m, Ar-H).
Se-(o-Chlorobenzyl)-l-selenocysteine HCl (16).
The same procedure as for Se-(o-methylbenzyl)-l-selenocysteine HCl was used with 0.9 mmol ofN-tert-butoxycarbonyl-l-selenocystine and 3.8 mmol of o-chlorobenzylchloride. The yield was 57%, and the purity, as determined by HPLC and 1H NMR, was >98%; white crystalline solid; 1H NMR (D2O, Na2CO3): δ (ppm) 3.1 (2H, dd, CH2-CH), 4.0 (2H, s, Ar-CH2-Se), 4.2 (1H, m, CH2-CH), 7.3 to 7.5 (4H, m, Ar-H).
Se-(o-Nitrobenzyl)-l-selenocysteine HCl (17).
The same procedure as for Se-(o-methylbenzyl)-l-selenocysteine HCl was used with 0.9 mmol ofN-tert-butoxycarbonyl-l-selenocystine and 3.8 mmol of o-nitrobenzylchloride. The yield was 49%, and the purity, as determined by HPLC and 1H NMR, was >98%; white crystalline solid; 1H NMR (D2O, Na2CO3): δ (ppm) 3.1 (2H, dd, CH2-CH), 4.2 (2H, s, Ar-CH2-Se), 4.3 (1H, m, CH2-CH), 7.5 to 7.7 (3H, m,m-, p-Ar-H), 8.1 (1H, d, o-Ar-H).
Human Tissues
Kidneys from three Dutch men were isolated within 12 h after death at the Pathology Department, Academic Hospital Vrije Universiteit, Amsterdam, the Netherlands. The cortex of the kidneys was cut into small pieces and stored at −80°C until use. The causes of death were adenocarcinoma with metastases in the kidney for kidney I (donor died at the age of 53; pathologist selected the tumor-free portion), colon carcinoma for kidney II: (donor died at the age of 77), and lung tumor for kidney III (donor died at the age of 78). Kidney cytosols were prepared as described by Stijntjes et al. (1992). Protein concentrations as determined with the BioRad (Hercules, CA) protein assay were 17.51 mg/ml for kidney I, 14.80 mg/ml for kidney II, and 10.04 mg/ml for kidney III.
Animals
Male Wistar rats (200–250 g) were obtained from Harlan (Zeist, the Netherlands); they were fed a standard laboratory diet from Hope Farms (Woerden, the Netherlands) and had access to food and water ad libitum. Rats were sacrificed by decapitation, and kidneys were isolated, sliced, and used directly for cytosol preparation as described by Stijntjes et al. (1992). Protein concentration as determined with the BioRad protein assay was 16.57 mg/ml.
Enzyme Kinetics
Initially, the time course of the β-elimination in kidney cytosol was determined to assess linearity; 700 μl of 2.0 mM Se-Cys conjugate dissolved in 50 mM sodium borate buffer (pH 8.6) and 0.71 mM KMB (cofactor) was reequilibrated for 3 min at 37°C. For rat kidney cytosol, 280 μl of 50 mM borate buffer (pH 8.6) was added before reequilibration. The incubation was started by the addition of 20 μl of rat or 300 μl of human kidney cytosol. The final concentrations of the substrate and KMB were 1.4 and 0.50 mM, respectively. At several time points, 100-μl samples were mixed with 500 μl of 12 mMo-phenylenediamine in 3 N HCl and heated for 60 min at 60°C to complete the derivatization. The incubation vials were centrifuged for 15 min (4000g), and the amount of pyruvate was analyzed by HPLC (see earlier) as described by Stijntjes et al. (1992). Nonenzymatic degradation was investigated by parallel incubations in the absence of cytosols. All incubations were performed in triplicate. For human cytosol, the protein dependence of β-elimination was tested between 1 and 8.5 mg/ml using 0.50 mM Se-phenyl-l-selenocysteine (5) and a 20-min incubation time.
Enzyme kinetic parameters, apparent Kmand Vmax values, were determined by incubating substrates at concentrations ranging from 0.14 to 1.40 mM (six concentrations and a blank). Due to poor solubility, Se-Cys conjugates up to 2.0 mM were dissolved in 50 mM borate buffer (pH 8.6) and 0.71 mM KMB and subsequently diluted. The incubation was started by the addition of 2 μl (rat) or 30 μl (human) of cytosol. The total incubation volume was 100 μl, and the final KMB concentration was 0.5 mM. After 5 min for rat kidney cytosol or 20 min for human kidney cytosol, the reaction was terminated by the addition of 500 μl of 12 mM o-phenylenediamine in 3 N HCl and heated for 60 min at 60°C to complete the derivatization. The incubation vials were centrifuged for 15 min (4000g), and the amount of pyruvate was analyzed by HPLC (see earlier) as described by Stijntjes et al. (1992). Nonenzymatic degradation was investigated by parallel incubations in the absence of cytosols. All incubations were performed in triplicate. Enzyme kinetic parameters were obtained using Lineweaver-Burke plots.
Results
Time Course and Protein Dependence of β-Elimination Reactions.
The formation of pyruvate from all 22 selenocysteine Se-conjugates (Se-Cys conjugates) and CTFE-Cys (23) was linear up to 10 min in rat kidney cytosol. In case of incubations with human kidney cytosols the pyruvate formation began to deviate from linearity after 20 min. For this reason all specific activities as well as the enzyme kinetic parameters, Km,Vmax and intrinsic clearances (Vmax/Km), were obtained using 20-min incubation times in case of human cytosol and 5 min in case of rat cytosol (at that time sufficient pyruvate was formed and sufficient KMB was left). In human kidney cytosols, the activity was linear up to 8.5 mg/ml of protein. No nonenzymatic degradation was observed as determined by parallel incubation in the absence of cytosol.
Kinetics of β-Elimination Reactions.
Saturation kinetics was not obtained for all Se-Cys conjugates within the concentration range studied (0.14–1.40 mM) (Table 2). Due to low solubilities of the Se-Cys conjugates, the enzyme kinetic parameters, apparentKm andVmax values, could not be determined accurately for some compounds. In these cases, intrinsic clearances (Vmax/Km) were calculated from the slope in the Michaelis-Menten plots. The apparent Km values in rat and human kidney cytosols did not correlate with each other, although they were of the same order of magnitude. The apparentVmax values, however, were significantly lower in human kidney cytosol than in rat kidney cytosol.
As can be seen in Table 1, the specific activities for all compounds tested were significantly higher in rat than in human kidney cytosol. CTFE-Cys (23), a known nephrotoxic compound with a high β-elimination rate (Commandeur et al., 1995), had a specific activity comparable with those of the Se-Cys conjugates in both rat and human kidney cytosols, indicating that the Se-Cys conjugates are good substrates for β-lyase enzymes. As shown in Fig.2A, the intrinsic clearances of β-elimination (Vmax/Km) for the Se-Cys conjugates in rat kidney cytosol were between 4 (compounds 18) and 92 (compound 15) times lower than for CTFE-Cys (23), due to a lower apparent Km value for CTFE-Cys in rat kidney cytosol (Table 2). However, in human kidney cytosol, the intrinsic clearances of β-elimination (Vmax/Km) for some of the Se-Cys conjugates (5, 6, and 7) were comparable with that of CTFE-Cys (23) (Fig. 2B).
By comparing the specific activities obtained in rat and human cytosols (Table 1), significant correlations were observed for phenyl-substituted Se-Cys conjugates (rat/human kidney cytosol I,P = .0083, n = 9; rat/human kidney cytosol II, P = .011, n = 9) and benzyl-substituted Se-Cys conjugates (rat/human kidney cytosol I,P = .0019, n = 9; rat/human kidney cytosol II, P = .0006, n = 9; rat/human kidney cytosol III, P = .044, n = 9) but not for alkyl-substituted Se-Cys conjugates. Furthermore, by comparing the intrinsic clearances of β-elimination (Vmax/Km) in rat and human cytosols, significant correlations were obtained for phenylselenocysteine Se-conjugates (rat/human kidney cytosol I,P = .0037, n = 5; rat/human kidney cytosol II, P = .0012, n = 5) but not for benzyl-substituted Se-Cys conjugates.
The specific β-lyase activities of three human kidney cytosols tested were in the same order of magnitude (Table 1). As can be seen in Fig.3, A and B, and Table3, there were significant correlations between the specific activities of β-lyase enzymes in human kidneys. As demonstrated in Fig. 3, C and D, and Table 3, the intrinsic clearances (Vmax/Km) between different human kidneys also correlated. The deviation from the line indicates small quantitative differences between these human kidney cytosols (Fig. 3).
Substrate Selectivity.
By comparing stereoisomers of Se-Cys conjugates (10 of 11 and 18 of 19), it can be seen that stereoselective β-elimination of Se-Cys conjugates previously found in rat kidney cytosol (Andreadou et al., 1996) was also present in human kidney cytosol (Table 1). The β-lyase activity in human cytosols decreased by an increase in alkyl chain length (compounds 1, 2, and 3), whereas in rat cytosol, this effect was not observed.
As demonstrated in Fig. 2a, para-substitution at the phenyl-group (10 and 12) decreased the β-elimination rate in rat kidney cytosol, whereas the corresponding ortho-substituted phenyl Se-Cys conjugates (6 and 7) did not change the β-lyase activity considerably. Surprisingly, para-substitution of benzyl-substituted Se-Cys conjugates (18 and 20) resulted in an increase instead of a decrease in β-elimination rate, whereasortho-substitution (15 and 16) decreased it. In human kidney cytosols, a similar structure-activity relationship was found for phenyl-substituted Se-Cys conjugates, whereas for benzyl-substituted Se-Cys conjugates, an increase in activity was also found onpara-substitution; no decrease was observed onortho-substitution. In human kidney cytosols, phenyl-substituted Se-Cys conjugates (5, 6, 7, and 12) were better β-eliminated than the corresponding benzyl-substituted Se-Cys conjugates (14, 15, 16, and 20). For the para-methyl substituents in aromatic Se-Cys conjugates (10 and 18), the opposite was found (Fig. 2B).
Discussion
This study was performed to evaluate by measuring the formation of pyruvate whether Se-Cys conjugates undergo β-elimination reactions in human kidney cytosol. With this assay, we wanted to delineate whether β-elimination of Se-Cys conjugates into selenols is occurring in human kidney cytosol. Furthermore, we compared the β-elimination activity in human kidney cytosol with that in rat kidney cytosol. Several selenol compounds, which are anticipated as products, are known to possess antitumor and/or chemoprotective effects (Parnham and Graf, 1991). Furthermore, we tested kidney cytosol of three human individuals to preliminarily screen for interindividual differences in β-elimination reactions.
The present study indicated that all tested Se-Cys conjugates (n = 22) indeed underwent β-elimination reactions in human renal cytosol, although the activity was lower than that in rat kidney cytosol. Between 41- and 857-fold lower intrinsic clearances (Vmax/Km) were observed in human kidney cytosol compared with rat kidney cytosol. In analogy with our present findings, Green et al. (1990) demonstrated, usingS-(1,2,2-trichlorovinyl)-l-cysteine as substrate, that the intrinsic clearances of β-elimination (Vmax/Km) were 28 times lower in human kidney cytosol than in rat kidney cytosol. Using S-(2-benzothiazolyl)-l-cysteine (BTC) as a substrate, Lash et al. (1990) observed that β-lyase activity as well as the apparent Vmaxvalue was 10 times lower in human kidney cytosol than in rat kidney cytosol. The apparent Km value of BTC also differed between rat and human kidney cytosol. MacFarlane et al. (1989) and Yamauchi et al. (1993) both demonstrated that highly purified cytosolic rat kidney β-lyase/glutamine transaminase K, the main β-lyase enzyme in rat kidney cytosol, possesses little β-lyase activity toward BTC. However, Lash et al. (1990) observed excellent β-lyase activity toward this substrate in purified cytosolic human kidney β-lyase. These results clearly indicate that significant qualitative and quantitative differences exist between β-lyase metabolism in rats and humans. Perry et al. (1993, 1995), who isolated full-length cDNAs of human and rat kidney β-lyase and expressed them in Cos-1 cells, found comparable Kmvalues for β-lyase activity usingS-1,1,2,2-tetrafluorethyl-l-cysteine as a substrate. A comparison of the amino acid sequences of rat and human showed 82% overall similarity, with 90% similarity around the pyridoxal binding site. Lash et al. (1990) demonstrated that both rat and human cytosolic kidney β-lyases were inhibited by amino-oxyacetic acid, indicating that the rat and human enzymes are pyridoxal-dependent enzymes. Differences in responsiveness toward KMB, an exogenous 2-keto acid and a cofactor for β-lyase, were reported between purified human and rat kidney cytosolic enzymes: the rat enzyme was stimulated 30-fold by KMB, whereas the human enzyme was stimulated only 1.3-fold (Stevens et al., 1986; Lash et al., 1990).
In the present study, we observed significant correlations between specific activities and intrinsic clearances (Vmax/Km) in rat kidney cytosol and two of three human kidney cytosols for phenyl- and benzyl-substituted Se-Cys conjugates but not for alkyl-substituted Se-Cys conjugates, indicating significant species differences in substrate selectivity between human and rat enzymes. From human renal cytosol, two isoforms of β-lyase have been purified (Buckberry et al., 1990). Both isoforms possessed physicochemical and biochemical properties comparable with cytosolic rat renal β-lyase. Through activity staining in a nondenaturing gel system, Abraham and Cooper (1991) demonstrated that in rat kidney cytosol, two β-lyase enzymes are present with apparent molecular masses of 90,000 and 330,000 Da, respectively. The 90-kDa β-lyase previously has been shown to be a homodimeric protein of two subunits. Cloning studies performed by Perry et al. (1993) and Abraham and Cooper (1996), however, suggest that rat renal cytosol may contain two 90-kDa β-lyase enzymes. Perry et al. (1993) cloned and sequenced rat kidney cytosolic β-lyase/glutamine transaminase K with a calculated mass of 47.8 kDa. This sequence is smaller than that obtained by Abraham and Cooper (1996), who cloned and expressed another β-lyase enzyme with glutamine transaminase K activity from rat kidney cytosol with a calculated mass of 48.5 kDa.1 This clearly indicates that rat kidney cytosol possesses at least three β-lyase enzymes.
In the present study, variability in β-elimination of Se-Cys conjugates was tested by comparing enzyme kinetics in three different human kidney cytosols. Between these three kidney cytosols, significant correlations were obtained for specific activities and intrinsic clearances of β-elimination (Vmax/Km). Significant correlations that were obtained between substrate specificities in different individual human kidney cytosols can be explained either by the fact that bioactivation of Se-Cys conjugates is catalyzed by only one β-lyase enzyme or by a constant expression ratio between individuals of different β-lyase enzymes. Although no qualitative differences were observed, quantitative differences were up to 3-fold, suggesting that the same enzymes are expressed differently in human kidneys. Nelson et al. (1995) reported interindividual differences in specific β-elimination activity in normal human renal cortex tissue for up to 14-fold with BTC as substrate. Whether the presently observed significant interindividual differences in β-elimination activity in three human kidney cytosol are present in larger numbers of human kidneys, and to which extent, remains to be established. Because of the large range in specific activities toward renal β-lyases, the present Se-Cys conjugates might be used to delineate both quantitative and qualitative differences in β-elimination activities.
Intrinsic clearances of β-elimination (Vmax/Km) were higher for phenyl-substituted Se-Cys conjugates than for benzyl-substituted Se-Cys conjugates in human kidney cytosols. The newly synthesized ortho-substituted aromatic Se-Cys conjugates (n = 7) also appeared to be substrates for human renal β-lyase enzymes, and they were comparably active as or even exceeding the other Se-Cys conjugates. The fact thatortho-substitution is permitted may imply the possibility of generating selenols comparable with that formed by the glutathione peroxidase mimetic compound ebselen (Cotgreave et al., 1992). Targeting of these compounds to the kidney may be an approach to protect the kidney against nephrotoxic agents (Baldew et al., 1990, 1992).
In conclusion, Se-Cys conjugates are β-eliminated extremely well by human kidney cytosol, although at a lower activity compared with rat kidney cytosol. Se-Cys conjugates are therefore potentially useful as prodrugs to target selenol compounds with antitumor and/or chemoprotective activities to kidneys in humans. This is the first report, to the best of our knowledge, that shows strong correlations between the substrate specificity of β-elimination reactions in individual humans.
Footnotes
-
Send reprint requests to: Prof. Dr. Nico P. E. Vermeulen, Leiden/Amsterdam Center for Drug Research (LACDR), Division of Molecular Toxicology, Department of Pharmacochemistry, Vrije Universiteit Amsterdam, De Boelelaan 1083, 1081 HV Amsterdam, the Netherlands. E-mail vermeule{at}chem.vu.nl
-
↵1 Abraham and Cooper (1996) reported a mass of 45.8 kDa, whereas on recalculation of the mass from cDNA sequence, it appears to be 48.5 kDa, which was confirmed by Abraham (personal communication).
- Abbreviations:
- Se-Cys
- selenocysteine Se
- CTFE-Cys
- S-(2-chloro-1,1,2-trifluoroethyl)-l-cysteine
- BTC
- S-(2-benzothiazolyl)-l-cysteine
- KMB
- α-keto-γ-methiolbutyric acid
- GS-MS
- gas chromatography-mass spectrometry
- Received January 6, 2000.
- Accepted April 19, 2000.
- The American Society for Pharmacology and Experimental Therapeutics