0 Pharmacology & Toxicology 1998, 83, 95-103. Printed in Denmark . All rights reserved

Copyright 0

Pharmacology QToxieology ISSN 0901-9928

6-Aminolevulinate Dehydratase Inhibition by 2,3=Dimercaptopropanol is Mediated by Chelation of Zinc from a Site Involved in Maintaining Cysteinyl Residues in a

Reduced State* Tatiana Eman~e l l i ' , ~~~ , JoPo Batista T. Rocha', Maria Ester Pereira', Pado C. Nascimento', Diogo 0. G. Souza2 and

Flavio A. Beber'

'Department of Chemistry, Centre of Natural and Exact Sciences. Federal University of Santa Maria, 97105-900, Santa Maria, RS, 'Department of Biochemistry, Institute of Basic Health Sciences, Federal University of Rio Grande do Sul, 90035-003, Porto Alegre, RS, and 3Department of Alimentary Technology and Sciences, Centre of Rural Sciences,

Federal University of Santa Maria, 97105-900, Santa Maria, RS, Brazil

(Received December 4, 1997; Accepted April 21, 1998)

Abstract: The mechanisms underlying mouse 6-aminolevulinate dehydratase (ALA-D) inhibition by a chelating agent used in the treatment of heavy metal poisoning, 2,3-dimercaptopropanol (British Anti-Lewisite), were investigated. ALA-D inhibition by 2,3-dimercaptopropanol was totally reversed by 25-100 pM Zn2+, indicating that inhibition was due to chelation of zinc by 2,3-dimercaptopropanol. Our data suggested that zinc bound to a labile site (displaced by 2 5 4 0 pM EDTA or 500 pM 2,3-dimercaptopropanol) is involved in maintaining the sulfhydryl groups of ALA-D in a reduced state (essential for enzyme activity), since inhibition by these compounds was reversed by 10 mM dithiotreitol (a reducing agent). On the other hand, 10 mM dithiotreitol did not reverse ALA-D inhibition by a higher concentration of EDTA (100 pM). Accordingly, 2,3-dimercaptopropanol appears to inhibit ALA-D through a mechanism similar to that of low EDTA concentrations. Neither oxidized 2,3-dimercaptopropanol nor reactive oxygen species appeared to contribute for ALA-D inhibition by reduced 2,3-dimercaptopropanol. Taken together, these results suggest that 2,3-dimercaptopropanol inhibits ALA-D by chelating Zn2+ from a labile site that is involved in maintaining enzyme sulfhydryl groups in a reduced state. This site is compatible with the ZnB or Znp previously described in mammalian and bacterial ALA-D.

6-Aminolevulinate dehydratase (ALA-D) is a sulf'hydryl- containing enzyme (Gibson et al. 1955; Barnard et al. 1977) which catalyzes the asymmetric condensation of 2 mol- ecules of 6-aminolevulinic acid (ALA) to porphobilinogen (Gibson et al. 1955). The enzyme-catalyzed condensation occurs via the formation of two successive Schiff-base inter- mediates (Jaffe 1995). This reaction is fundamental in the biosynthesis of tetrapyrroles (such as chlorophyll and haem), the prostetic group of various proteins. Due to its sulf'hydrylic nature (Shemin 1976; Gibbs et al. 1985), ALA- D activity is highly sensitive to the presence of heavy metals such as mercury (Rocha et al. 1993 & 1995), lead (Goering 1993; Rodrigues et al. 1996; Bergdahl et al. 1997), copper (Nelson et al. 1981), beryllium (Sakaguchi et al. 1996), sel- enium (Barbosa et al. 1998) and tin (Chiba & Kikuchi 1984), which possess high affinity for -SH groups. The en- zyme from mammalian tissues and prokaryotes requires Zn2+ for maximum activity (Beyersmann & Cox 1984;

Author for correspondence: Tatiana Emanuelli, Department of Ali- mentary Technology and Sciences, Centre of Rural Sciences, Feder- al University of Santa Maria, University Campus - Camobi, 97105- 900 - Santa Maria, RS Brazil (fax +55 55 220-8695). *Preliminary results of this study were presented at the 24th Annual Meeting of Brazilian Society of Biochemistry and Molecular Bi- ology, Caxambu, MG, Brazil, May 4 7 , 1995 and 8th PABMB Con- gress, November 1996, Chile.

Gibbs et al. 1985; Chauhan & O'Brian 1995; Jaffe 1995), but the precise function of Zn2+ in ALA-D is not clear. There are several findings implying a direct catalytic func- tion for Zn2+, whereas other results support the assumption that Zn2+ stabilizes the active structure of the enzyme pro- tein without directly interacting with the substrate (Bevan et al. 1980; Beyersmann & Cox 1984; Sommer & Be- yersmann 1984; Hasnian et al. 1985; Block et al. 1990; Dent et al. 1990; Jaffe 1995). In mammalian and bacterial ALA- D, Zn2+ appears to be bound at two distinct sites (Dent et al. 1990; Jaffe 1995; Spencer & Jordan 1994; Beber et al. 1998). Bovine ALA-D presents an A metal-ion-binding site with a single cysteine residue among its ligands. Zn2+ bound to this site seems to be essential for catalytic activity and was proposed to play a role in inter-ALA bond forma- tion and in product binding (Jaffe 1995). Additionally, bo- vine enzyme presents another site for zinc binding called B metal-ion-binding site with four cysteine residues among its ligands (Dent et al. 1990). Each Zn2+-binding site is present at a number of 4 per octamer. ALA-D from Escherichia coli presents two zinc-binding sites per subunit designated c1 and j3, rather than A and B. Similarly to that described for the A site of bovine ALA-D, the a-site contains one cysteine, whereas the P-site has three cysteine residues among the metal ligands when Co2+ is present at this site (Spencer & Jordan 1994). Spencer & Jordan (1994) also provided some

96 TATIANA EMANUELLI ET AL.

evidence indicating that the two metal-ion-binding sites of ALA-D from E. coli have different functions, and suggested that the a-site is essential for catalytic activity, while the p- site is thought to be involved in the stabilization of protein conformation.

In a previous report we observed that 2,3-dimercapto- propanol, a dithiol compound therapeutically used as a chelating agent in the treatment of heavy metal intoxi- cation, inhibited the activity of renal ALA-D both in vivo and in vitro, while the hepatic enzyme was inhibited only in vitro (Emanuelli et al. 1996). We have proposed two hypotheses to explain this mechanism of inhibition: 1) the chelation of Zn2+ ions essential for enzyme activity; and/ or 2) the oxidation of essential -SH groups of the enzyme mediated by reactive oxygen species generated during 2,3- dimercaptopropanol oxidation or by the oxidized com- pound.

The objective of the present study was to investigate the role of the above mechanisms on ALA-D inhibition by 2,3- dimercaptopropanol. The results clearly showed that 2,3- dimercaptopropanol inhibits ALA-D by removing essential Zn2+ bound to a more labile ZnZf site, probably the ZnB or Znp site, and provided functional evidence that the Zn2+ bound to this site is implicated in maintaining essential cys- teinyl residues in a reduced state.

Materials and Methods Chemicals. S-Aminolevulinic acid (ALA), 6-aminolevulinate de- hydratase purified from bovine liver, 5,5'-dithio-bis(2-nitrobenzoic) acid, zinc chloride, 2,3-dimercaptopropanol, p-dimethylaminoben- zaldehyde, and DL-dithiothreitol were purchased from Sigma (St. Louis, MO, U.S.A.). Nitric acid, trichloroacetic acid and sodium carbonate were purchased from Reagen (Rio de Janeiro, RJ, Brazil). Sodium chloride was obtained from Vetec (Rio de Janeiro, RJ, Bra- zil). All other chemicals were purchased from Merck (Darmstadt, Germany). Oxidized 2,3-dimercaptopropanol (minimum 99%) was prepared by keeping a 10 mM 2,3-dimercaptopropanol solution at room temperature, until it no longer showed a positive reaction to Ellman 's reagent.

Tissue preparation. Adult mice from our own breeding colony were killed by decapitation. Liver and kidney were quickly removed, placed on ice and homogenized in 150 mM NaCl 10 and 7 volumes, respectively The homogenate was centrifuged at 4000Xg at 4°C for 10 min., to yield a low speed supernatant fraction (S,) that was used for enzyme assay and for the determination of the rate of 2.3- dimercaptopropanol oxidation.

Preparation of partially-purified ALA-D. Liver and kidney were homogenized in 2.5 and 6.5 volumes of a medium containing 50 mM potassium phosphate buffer, pH 6.4, 3 mM DL-dithiothreitol and 30 pM ZnC12. The homogenates were centrifuged at 4000Xg at 4" for 15 min. The supernatant fraction obtained (S,) was heated to 60" for 15 min. and then rapidly cooled on ice. Denatured protein was removed by centrifugation at 4000Xg at 4" for 15 min. The supernatant fraction (S2) was treated with (NH4)$04 (35% satu- ration for liver and 30% saturation for kidney). Precipitated protein was removed by centrifugation at 4000Xg at 4" for 20 min. The supernatant fractions (S3) of kidney and liver were treated with (NH4)2S04 to give 55% saturation, and the precipitate was collected by centrifugation at 4000Xg at 4" for 20 min. The pellet, which

contained the enzyme ( P 4 , was resuspended in a medium contain- ing 20 mM potassium phosphate buffer, pH 6.4, I mM DL-dithio- threitol and 10 pM ZnCI2. This procedure permits 2-3 and 8-10- fold purification of renal and hepatic ALA-D, respectively. The par- tially purified enzyme was stored at -20".

Reactivation of partially purified ALA-D. ALA-D was reactivated just before use by incubation with 10 mM DL-dithiothreitol and 20 pM ZnC12 at 45" for 15 min. and diluted with distilled water immediately before the enzyme assay, in order to obtain final con- centrations of 100 pM DL-dithiothreitol and 0.2 pM ZnCI2 in the incubation medium.

Enzyme assays. ALA-D activity was assayed according to the method of Sassa (1982) by measuring the rate of product (por- phobilinogen) formation, except that 76 mM potassium phos- phate buffer, pH 6.4, and 2.2 mM ALA were used. The reaction product was determined using modified Ehrlich's reagent at 555 nm, with a molar absorption coefficient of 6.1X104 M-l cm-l for the Ehrlich-porphobilinogen salt. All experiments were carried out after 10 min. of pre-incubation. The reaction was started 10 min. after the addition of the tissue preparation by adding the substrate. Incubations were carried out for 0 . 4 2 hr at 39". The amount of protein incubated in the assays using S , tissue prepara- tions was 0.8-1.1 mg for kidney and liver, while in the assays using partially purified ALA-D this amount was 0.2-0.3 mg for kidney and 0.25-0.35 mg for liver. Protein was determined ac- cording to Bradford (1976), using bovine serum albumin as stan- dard.

Determination of the rate of 2,3-dimercaptopropanol oxidation. The rate of 2,3-dimercaptopropanol oxidation was determined under the same conditions used for the ALA-D assay. Since ALA (2.2 mM) did not alter the rate of 2,3-dimercaptopropanol oxidation in the absence or in the presence of tissue supernatant (data not shown), it was not used in oxidation experiments. The rate of 2,3-dirnercap- topropanol(500 pM) oxidation was evaluated by measuring the dis- appearance of -SH groups. Free -SH groups were determined ac- cording to the method of Ellman (1959). The incubation was in- itiated by adding tissue preparation (S,) immediately after 2,3- dimercaptopropanol addition. At 0, 10, 30 and 60 min., aliquots of the reaction mixture were checked for the amount of -SH groups at 412 nm (absorption coefficient 1.36X lo4 M-' cm-' ). The concen- trations of the -SH groups of 2,3-dimercaptopropanol in the pres- ence of tissue supernatants (S,) were calculated as the difference between the -SH concentration in the tube containing 2.3- dimercaptopropanol+tissue and the -SH concentration in the tube containing tissue alone.

Determination of zinc concentration. Zinc concentration was deter- mined by differential-pulse polarography, with a Pokarographie- Stand VA 646 with an automatic sampler 675 (Metrohm). Measure- ments were performed with a three-electrode arrangement (mercury dropping electrode, saturated silverisilver chloride electrode and platinum electrode). Dissolved oxygen was removed from the solu- tions prior to the polarographic measurements by purging with ni- trogen for 10 min. Polarographic determination was based on the fact that zinc is active in aqueous solution at -1100 mV (versus AgIAgC1). The metal concentration was determined by comparison with a standard curve constructed from known metal-ion concen- trations.

Statistical analysis. Data about the effect of superoxide dismutase (SOD) and NADH on ALA-D activity were analyzed by one-way analysis of variance followed by Duncan's multiple range test when appropriate. All other results were analyzed by two-way ANOVA, followed by Duncan's multiple range test when appropriate. Differ- ences between groups were considered to be significant when P50.05.

6-AMINOLEVULINATE DEHYDRATASE INHIBITION BY 2.3-DIMERCAPTOPROPANOL 97

150 i 1

120

90

60

A

120 1

0 ' I

0 100 200 300 400 500 600

Fig. 1. Effect of 2,3-dimercaptopropanol and ZnCI, on renal (A) and hepatic (B) ALA-D activity. ALA-D was pre-incubated with 2,3-dimercaptopropanol and/or ZnCI, (simultaneously added) for 10 min before the enzyme reaction was started. The reaction was initiated by adding the substrate to a final concentration of 2.2 mM in a medium containing 76 mM potassium phosphate buffer, pH 6.4. ALA-D activity is expressed as percent of control activity (0 pM BAL + 0 pM ZnCI,) (n=5). S.E.M. were less than 15% of respective means. -.- 0 pM ZnCl,, -A- 25 pM ZnCI,. -0- 50 pM ZnC12. -7- 100 pM ZnCI2. BAL-2,3-dimercaptopropanol. ALA-D-6-aminolevulinate dehydratase.

Results

The role ojzinc on ALA-D inhibition by 2,3-dimercaptopro- panol. In order to determine if 2,3-dimercaptopropanol inhibits ALA-D through the chelation of essential Zn2+ ions, we investigated the effect of ZnC12 on the inhibition of ALA- D activity by 100-500 pM 2,3-dimercaptopropanol (fig. 1A and B). The renal enzyme was significantly inhibited by 250-500 pM 2,3-dimercaptopropanol and Zn2+ (25 to 100 pM) fully reversed the inhibition caused by 2,3-dimercapto- propanol. Renal ALA-D was activated by 25 pM ZnC12

(fig. 1A). As also observed for renal ALA-D, the hepatic enzyme was inhibited by 250-500 pM 2,3-dimercaptopro- panol. In contrast to renal ALA-D, 100 pM ZnC12 inhibited the activity of hepatic ALA-D. As observed in kidney, all tested ZnCI2 concentrations completely restored hepatic ALA-D activity inhibited by 2,3-dimercaptopropanol (fig. IB).

Fig. 2 shows the concentration of free zinc ions, polaro- graphically determined, in the presence of 500 pM reduced 2,3-dimercaptopropanol, 20 pM ethylenediaminetetraacetic acid (EDTA), or 500 pM oxidized 2,3-dimercaptopropanol after successive additions of ZnC12. The amount of free zinc ions was smaller than the amount of ZnC12 added at every concentration and in the presence of all compounds ana- lyzed, indicating that these compounds chelate Zn2+ under the conditions used in this assay (76 mM potassium phos- phate buffer, pH 6.4). Furthermore, at low Zn2+ concen- trations (up to 15 pM), 500 pM reduced 2,3-dimercaptopro- panol chelates approximately as much Zn2+ as does 20 pM EDTA (fig. 2), a chelating agent that inhibits the activity of hepatic ALA-D from 25 pM onwards (fig. 4). Although oxidized 2,3-dimercaptopropanol also chelates Zn2+, it ap- parently presents less affinity for this metal than reduced 2,3-dimercaptopropanol (fig. 2). These results suggest that reduced 2,3-dimercaptopropanol inhibits ALA-D by chela- tion of Zn2+.

There are some evidence in the literature showing that zinc ions may be responsible for maintaining the -SH groups of the enzyme in the reduced state (Tsukamoto et al. 1979; Gibbs et al. 1985; Beber et al. 1998). In order to verify the function of Zn2+-ions removed by 2,3-dimercap-

40

30

20

10

0 0 15 30 45 60

Added ZnCI, (pM) Fig. 2. Concentration of free zinc ions. Zinc concentrations were determined after successive additions of ZnCI, to a solution con- taining 76 mM potassium phosphate buffer, pH 6.4, in the presence of 500 pM reduced BAL (-A-), 20 pM EDTA (-V-) or 500 pM oxidized BAL (-0-). Data from one experiment representative of three. S.E.M. were less than 15%) of respective means. For abbrevi- ations, see fig. 1.

98 TATIANA EMANUELLI ET AL.

0 ’ I 200

1 50

100

50

0

6

0.00 0.50 1.00

BAL (mM) Fig. 3. Effect of DL-dithiothreitol (DTT) and BAL on renal (A) and hepatic (B) ALA-D activity. ALA-D was pre-incubated with BAL and/or DL-dithiothreitol (simultaneously added) for 10 min. before the enzyme reaction was started. The reaction was started by adding the substrate to a final concentration of 2.2 mM in a medium containing 76 mM potassium phosphate buffer, pH 6.4. ALA-D specific activity is expressed as nmol of porphobilinogen formedih/mg protein (n=3). S.E.M. were less than 15% of respective means. -.- control, -0- 100 pM DTT, -0- 500 pM DTT, -0- 10 mM DTT.

topropanol, we examined the effect of DL-dithiothreitol (a dithiol compound used for maintaining -SH groups in the reduced state) on the inhibition of ALA-D by 2,3-dimercap- topropanol (fig. 3). Renal and hepatic ALA-D inhibition by 2,3-dimercaptopropanol was completely reversed by DL- dithiothreitol (10 mM), suggesting that ALA-D inhibition by 2,3-dimercaptopropanol may involve displacement of Zn2+ bound to a site implicated in the stabilization of -SH groups from the enzyme (fig. 3A and B).

Since 2,3-dimercaptopropanol and DL-dithiothreitol are dithiol compounds, their effects may be masked by redox reactions. To test more clearly the existence of a labile Zn2+ site in ALA-D, a detailed EDTA curve was constructed in order to remove this labile Zn2+ from enzyme both in the

absence and in the presence of 10 mM DL-dithiothreitol (Fig. 4). This compound (10 mM) completely reversed ALA-D inhibition caused by EDTA up to 40 pM, indi- cating that at this concentration EDTA removes zinc ions involved in maintaining -SH groups from ALA-D in a re- duced state (fig. 4). The inhibitory effect of EDTA on he- patic ALA-D was not enhanced by 1 mM 2,3-dimercaptop- ropanol indicating that 2,3-dimercaptopropanol inhibits ALA-D by acting in a site related to that of EDTA. This site is compatible with the ZnB or Znp previously described in mammalian and bacterial ALA-D. This hypothesis was reinforced by the absence of an inhibitory effect of 2,3- dimercaptopropanol on the activity of ALA-D from cu- cumber leaves (data not shown). In plants, the B site is char- acterized by the substitution of some cysteinyl residues by acidic amino acids, which apparently renders the enzyme less susceptible to oxidation at this site (Jaffe 1995). On the other hand, ALA-D inhibition by 100 pM EDTA could not be reversed by 10 mM DL-dithiothreitol (fig. 4), indicating that, at this higher concentration, EDTA removes zinc ions from a site which binds Zn2+ more tightly. Probably this site for zinc ions does not stabilize ALA-D in a reduced state but is essential for enzyme activity. Similar results were obtained with the renal enzyme and partially purified en- zyme from mouse liver (data not shown). This second site is consistent with ZnA or Zno? described in ALA-D from bovine liver and bacteria (Dent et al. 1990; Jaffe 1995; Spen- cer & Jordan 1994) and apparently is involved in inter ALA bond formation (Jaffe 1995).

120 I ~-

100

80

60

40

20

-0- Control

-0- DlT 10 mM

-A- BALl mM

u 0 ’ 0 25 50 75 1 00

EDTA (pM) Fig. 4. Effect of EDTA on hepatic ALA-D activity, in the presence of 10 mM DL-dithiothreitol or 1 mM BAL. Hepatic ALA-D was pre-incubated with EDTA and/or BAL or DTT (added simul- taneously), for 10 min. before the enzyme reaction was started. The reaction was initiated by adding the substrate. ALA-D specific ac- tivity in the absence of EDTA (100 %) was equal to 124.725.4, 150.2t9.3, and 49.825.2 nmol porphobilinogen formed/hr/mg of protein in the absence of thiol compounds (control), with 10 mM DTT or 1 mM BAL, respectively. Data are presented as mean2S.E.M. (n=4). For abbreviations, see fig. 1.

6-AMINOLEVULINATE DEHYDRATASE INHIBITION BY 2,3-DIMERCAPTOPROPANOL 99

A A

0.1 0 T

0.05

I 0 0.4 1 10 40

BAL (mM)

1.40 -/

0.20 o'40 t k/ . ' ' . " I

0 0.3 0.5 1 4

BAL (mM)

0.1 6

C a, 0.13 .- c

g 0.11

E" g 0.08

$ 0.05

f 0.03

0.00

- 0

0

12

10

8

6

4

2

0

B

0.3 0.5 1 4

BAL (mM)

D k

4/ . . . 0 0.3 0.5 1 4

BAL (mM)

Fig. 5. Effect of BAL on the activity of crude (SI) hepatic ALA-D (A), partially purified ALA-D from kidney (B) or liver (C) of mouse (P55); or purified ALA-D from bovine liver (Sigma Co.) (D). ALA-D, except crude (SI) preparation was reactivated just before assay as described in materials and methods. Incubations were performed in an argon (-0-) or air (-'i-) atmosphere. For the incubations performed in an argon atmosphere, every solution used was previously purged with argon (2 min.) in order to eliminate oxygen, and incubation tubes were submitted to vacuum (5 rnin.) followed by argon (5 rnin.) before the starting of the pre-incubation. Tubes under an argon atmosphere were submitted to cycles of vacuum (5 min.) followed by argon ( 5 min.) during the whole incubation time. ALA-D was pre-incubated with BAL for 10 min. before the enzyme reaction was started. The incubation mixture contained 100 pM DTT. 76 mM phosphate buffer, pH 6.4, and 2.2 mM ALA. ALA-D specific activity is expressed as pmol porphobilinogen formed/h/mg of protein. Data are presented as mean5.E.M. (n=4 to 5 ) . The specific activity of the partially purified enzyme from mouse kidney and liver and of purified enzyme from bovine liver (Sigma Co.) assayed in the presence of 10 mM DTT and 10 pM ZnC12, under an air atmosphere was 0.16t0.01, 1.4320.1 1 and 14.5t0.9 pmol porphobilinogen/hr/mg of protein, respectively. For abbreviations, see fig. I .

The role of 2,3-dimercaptopropanol oxidation on ALA-D in- hibition. As a dithiol compound, 2,3-dimercaptopropanol could un- dergo autoxidation in the conditions used in ALA-D assay and this autoxidation may play a role on ALA-D inhibition by this compound. In order to verify this hypothesis we examined the rate of 2,3-dimercaptopropanol oxidation in the conditions used in ALA-D assay.

The rate of 2,3-dimercaptopropanol oxidation was in- creased (about 30%) by liver (fig. 6) and kidney (data not shown) S, supernatants, when compared to 2,3-dimercapto- propanol oxidation in the absence of tissue supernatants. These results may indicate that oxidized 2,3-dimercaptopro- panol generated in the enzyme assay may be involved in ALA-D inhibition by reduced 2,3-dimercaptopropanol. Oxidized 2,3-dimercaptopropanol significantly inhibited re-

nal (1 mM or higher) and hepatic (1 mM or higher) ALA- D activity (data not shown), however ZnC12 (25 pM) did not reverse this inhibition (data not shown).

Role of reactive oxygen species on ALA-D inhibition by 2,3- dimercaptopropanol. Misra (1974) demonstrated that reactive oxygen species are generated during the oxidation of thiols. Since 2,3-dimer- captopropanol oxidizes in the conditions used in ALA-D assay (fig. 6) , we investigated if reactive oxygen species gen- erated during its oxidation could be involved in ALA-D inhibition by 2,3-dimercaptopropanol.

First we examined the sensitivity of ALA-D to H202 and OH' (hydroxyl radical). The hydroxyl radical was generated by Fenton's reaction (Fe2+ + H202+0H' + OH- + Fe3+), by measuring ALA-D activity, after pre-incubation of the enzyme in the presence of 50 pM Fe2+, 1 mM H202 and 6

100 TATIANA EMANUELLI ET AL.

10.00

a

a 7.50 !! .- c g 5.00

s ' 2.50

0.00 air argon

Incubation atmosphere - EAL BALISl [ZZD EALIP66 BIL1ALA-0

(PURIFIED)

Fig. 6. Effect of different hepatic ALA-D preparations (S,, P55 and a commercial purified product) on the rate of BAL oxidation in an argon or air atmosphere. BAL oxidation was determined in a me- dium containing 76 mM phosphate buffer (pH 6.4). For the incuba- tions performed in an argon atmosphere, every solution used was previously purged with argon (2 min.) in order to eliminate oxygen, and incubation tubes were submitted to vacuum (5 min.) followed by argon (5 rnin.) before the starting of the reaction. Tubes under an argon atmosphere were submitted to cycles of vacuum (5 min.) followed by argon (5 min.) during the whole incubation time. The rate of BAL oxidation is presented as nmol of -SH oxidized/min./ ml. Data are presented as meantS.E.M. (n=4-8). "Significantly dif- ferent from the rate of BAL oxidation in the absence of tissue, re- gardless of the incubation atmosphere. *Significantly different from the rate of BAL oxidation in an air atmosphere, regardless of the hepatic ALA-D preparations. For abbreviations, see fig. 1.

mM ascorbate. Renal and hepatic ALA-D were not inhib- ited by H202 (up to 1 mM) (data not shown). On the other hand, renal and hepatic ALA-D were inhibited (50 and 40%, respectively) by ascorbate, however the presence of Fe2+, H202 or both did not potentiate this inhibition (data not shown), indicating that ALA-D was not susceptible to inhibition by OH' radicals.

According to the mechanism proposed by Misra (1974) for the autooxidation of thiols, H202 might generate OH' radicals by reaction with 02.- or with thiyl radicals. Hence, if OH' radicals were the species responsible for ALA-D in- hibition by 2,3-dimercaptopropanol, the simultaneous ad- dition of H202 and 2,3-dimercaptopropanol would increase the inhibition. We observed that simultaneous addition of H202 and 2,3-dimercaptopropanol did not affect the inhibi- tory effect of 2,3-dimercaptopropanol on renal and hepatic ALA-D (data not shown). These data reinforce our pro- posal that OH' radicals are not responsible for the ALA-D inhibition by 2,3-dimercaptopropanol.

We also investigated the effect of NADH (a 02.- scaven- ger) (Kido & Soda 1984) and 3 15 U/ml superoxide dismu- tase (a 0 2 . - scavenger) (Fridovich 1995) on 2,3-dimercapto- propanol-inhibited ALA-D activity (table 1) and on the rate of 2,3-dimercaptopropanol oxidation (data not shown). None of these scavengers were capable of reducing ALA-D inhibition by 2,3-dimercaptopropanol (table 1). In the case of liver, simultaneous addition of 2,3-dimercaptopropanol

and superoxide dismutase increased the inhibitory potency of 2,3-dimercaptopropanol (table I ) . These compounds did not affect 2,3-dimercaptopropanol oxidation either in the absence or in the presence of tissue supernatants (data not shown).

Effeet of 2,3-dimercuptopropunol on the uctivity of purtially- puriJied ALA-D. The reversal of 2,3-dimercaptopropanol-inhibited ALA-D by the addition of DL-dithiothreitol raised the hypothesis that the removal of zinc from the less tightly binding sites could render the enzyme susceptible to oxidation by some tissue factor present in our crude preparation. In order to investigate this hypothesis we performed a partial purifi- cation of renal and hepatic ALA-D in an attempt to elimin- ate these factors (fig. 5B and C). To reinforce these experi- ments we also investigated the effect of 2,3-dimercaptopro- panol on the activity of commercial ALA-D from bovine liver (Sigma Co.) (fig. 5D). In order to determine a possible role for oxygen on ALA-D inhibition by 2,3-dimercapto- propanol, the effect of 2,3-dimercaptopropanol on the ac- tivity of crude hepatic enzyme (S,), partially purified he- patic, or renal (Pss) ALA-D and on purified commercial ALA-D was determined under an argon or air atmosphere (fig. 5A, B, C and D).

The inhibitory effect of 2,3-dimercaptopropanol on cru- de enzyme preparations was not affected by the removal of oxygen during incubation (fig. 5A). Since partially purified ALA-D stored at -20" and assayed under standard con- ditions (see Materials and Methods for details) presented a specific activity much lower than the maximum activity of that preparation (assay carried out in the presence of 10 mM DL-dithiothreitol and 10 pM ZnC12) (0.6320.05 versus 1.4320.1 1 pmol porphobilinogen/h/mg of protein, for par- tially purified ALA-D from liver) and also that this activity decreased as a function of storage time (data not shown),

Table I Effect of superoxide dismutase (SOD) and NADH on S-aminolevu- h a t e dehydratase (ALA-D) inhibition by BAL.

Kidney Liver

Control 0.1 mM NADH 2 mM NADH SOD BAL BAL+O.l mM NADH BAL+2 mM NADH BAL+SOD

100.0 94.4t4.4 57.02 16.6* 84.62 14.1 59.929.1* 39.725.2* 33.9t5.6* 37.4t5.6*

100.0 102.5t9.9 3 1.0+3.2*" 97.9t3.6 52.323.1* 47.3 t 5.8* 22.4t 3.9*" 33.325.7*"

ALA-D was pre-incubated with 2,3-dimercdptopropanol (BAL) (500 pM), NADH (0.1 or 2.0 mM) and/or SOD (3 15 U/ml) (added simultaneously) for 10 min before the enzyme reaction was started. ALA-D activity is expressed as percentage of control (mean2S.E.M.) (n=5). The control activities of renal and hepatic ALA-D were 73.3 and 228.4 nmol porphobilinogen/hr/mg protein, respectively. *Significantly different from control. "Significantly different from BAL.

6-AMINOLEVULINATE DEHYDRATASE INHIBITION BY 2.3-DIMERCAPTOPROPANOL 101

we performed reactivation of partially purified ALA-D (see Materials and Methods) just before the enzyme assay. In these assays, 100 pM DL-dithiothreitol was added to the incubation medium in order to preserve ALA-D activity. This concentration of DL-dithiothreitol did not alter ALA- D inhibition by 2,3-dimercaptopropanol (fig. 3A and B). The specific activity of partially purified, reactivated ALA- D was similar to the maximum activity of the renal enzyme preparation (fig. 5B) or hepatic enzyme preparation (fig. 5C). As observed for non-purified hepatic ALA-D (fig. 5A), reactivated partially purified renal (fig. 5B) or hepatic en- zymes (fig. 5C) were inhibited by 2,3-dimercaptopropanol (500 pM onwards), regardless of the incubation atmosphere. We also investigated the effect of 2,3-dimercaptopropanol (0-4 mM) on the activity of reactivated, purified commer- cial ALA-D from bovine liver (Sigma Co.) (fig. 5D), incu- bated under an argon or air atmosphere in the presence of 100 pM DL-dithiothreitol and 0.2 pM ZnC12. The behav- iour of this enzyme was similar to that of partially purified ALA-D from mice (fig. 5B and C).

The rate of 2,3-dimercaptopropanol oxidation in the ab- sence or in the presence of tissue preparations was signifi- cantly reduced when this compound was incubated under an argon atmosphere, compared to the assays under an air atmosphere (fig. 6). All of these data suggest that oxygen enhanced 2,3-dimercaptopropanol oxidation in the presence or absence of enzyme preparations but had little effect on ALA-D inhibition by this compound. These results support the hypothesis that ALA-D inhibition by 2,3-dimercaptop- ropanol is not related to the oxidation of this compound.

Discussion

The results on the determination of free zinc ions and on the effect of Zn2+ on ALA-D inhibition by 2,3-dimercapto- propanol demonstrate that this compound chelates Zn2+ under our assay conditions, and that the inhibition of renal and hepatic ALA-D by 2,3-dimercaptopropanol (250-500 pM) can be completely reversed by 25 pM ZnC12. The sta- bility constant reported for the zinc-2,3-dimercaptopropan- 01 complex at pH 3.30-3.78 is approximately (Leuss- ing & Tischer 1961). In more alkaline solutions there is the formation of a soluble Zn(2,3-dimer~aptopropanoI)~ - com- plex, with a greater stability constant (Leussing & Tischer 1961). On the other hand, ALA-D apparently binds Zn2+ with an association constant of (Sommer & Be- yersmann 1984). These results indicate that 2,3-dimercapto- propanol inhibits ALA-D by chelating Zn2+ essential for enzyme activity.

Our results demonstrate that EDTA up to 40 pM and 2,3-dimercaptopropanol up to 1 mM inhibit ALA-D by re- moving zinc ions most likely involved in protecting -SH groups of ALA-D from oxidation, since the inhibition can be reversed by DL-dithiothreitol. On the other hand, ALA- D inhibition by EDTA concentrations higher than 40 pM could not be fully reversed by DL-dithiothreitol, indicating that at these concentrations EDTA was removing zinc ions

more tightly bound to ALA-D, which are not involved in maintaining -SH groups from ALA-D in a reduced state. These results are in agreement with other studies on ALA- D from E. coli (Spencer & Jordan 1994) and from rats (Be- ber et ul. 1998), and indicate a difference in the function of the two metal-ion-binding sites in ALA-D.

The reversal of 2,3-dimercaptopropanol-inhibited ALA- D by the addition of DL-dithiothreitol raised the hypo- theses that the removal of zinc from the site binding zinc relatively weakly could render the enzyme susceptible to oxidation by oxidized 2,3-dimercaptopropanol, atmospheric oxygen or by some tissue factor present in our crude prep- aration.

Although 2,3-dimercaptopropanol oxidizes under the conditions used in the ALA-D assay (pH 6.4, 39"), oxidized 2,3-dimercaptopropanol presents a lower inhibitory potency than the reduced compound. Furthermore the mechanism involved in the inhibition of ALA-D by oxidized 2,3-dimer- captopropanol is different from that involved in ALA-D inhibition by reduced 2,3-dimercaptopropanol, since Zn2+ was ineffective in reversing ALA-D inhibition by the oxid- ized compound. The inhibition of ALA-D by oxidized 2,3- dimercaptopropanol is possibly due to a redox reaction, with reduction of 2,3-dimercaptopropanol at the expense of ALA-D oxidation.

The effect of scavengers of reactive oxygen species on 2,3- dimercaptopropanol-inhibited ALA-D activity revealed that 0 2 . - was not the ultimate ALA-D inhibitor. Further- more, H202 and OH' appeared not to affect ALA-D activ- ity. Hence, although 2,3-dimercaptopropanol oxidizes dur- ing ALA-D assay, neither the oxidized compound nor reac- tive oxygen species generated during thiols oxidation (Misra 1974) seem to be involved in ALA-D inhibition by this com- pound.

The low degree of purification obtained for mice ALA-D leaves some doubt about the efficient removal of prooxidant tissue factors that could contribute to enzyme inhibition. Nevertheless, the use of a commercial purified ALA-D (from bovine liver, Sigma Co.) confirmed the results ob- tained with our partially purified enzyme, indicating that ALA-D inhibition by 2,3-dimercaptopropanol was not in- duced by some oxidant factor present in our crude prepara- tion. The experiments performed with crude (SI), partially purified (P54 and purified (Sigma Co.) ALA-D incubated under natural or argon atmosphere demonstrated that ALA-D inhibition by 2,3-dimercaptopropanol is not sig- nificantly dependent on the presence of oxygen. On the other hand, the rate of 2,3-dimercaptopropanol oxidation in the absence or in the presence of tissue preparations was significantly reduced when the incubation was performed under an argon atmosphere. These results corroborate the hypothesis that 2,3-dimercaptopropanol oxidation is not the main factor underlying ALA-D inhibition by reduced 2,3-dimercaptopropanol.

Our results provide clear evidence of different functions for the two metal ion-binding sites of ALA-D from mouse tissues. We obtained indications that Zn2+ less tightly

102 TATIANA EMANUELLI ET AL.

bound to the metal-ion-binding-site (possibly the B or p metal ion-binding site), is implicated in maintaining -SH groups of ALA-D in the reduced state. Zinc bound to this site is necessary for enzyme activity when ALA-D is assayed in a medium without high concentrations of a thiol-reduc- ing agent. Other work has demonstrated that ALA-D con- taining 4 zindoctamer (zinc bound t o the more stable site) possesses maximum activity (Bevan et al. 1980). Probably these authors did not observe an antioxidant action of Zn2+ due to the presence of high concentrations (5-10 mM) of a thiol-reducing agent (dithiothreitol) in the incubation me- dium.

The present results show that 2,3-dimercaptopropanol re- moves Zn2+ from a more labile site which is compatible with the p- or B-site described by other investigators for the mammalian and bacterial enzyme (Dent et al. 1990; Jaffe 1995; Spencer & Jordan 1994). Furthermore, this site seems to be involved in maintaining sulfhydryl groups in a reduced state. These da t a reveal a biochemical mechanism that may help to explain, a t least in part, the toxic properties of this therapeutic agent used in heavy metal poisoning. The objec- tion to this possibility is the fact that in most tissues, the concentration of ALA-D is much higher than that required for sustaining haem biosynthesis. However the chelation of zinc from other metalloproteins may contribute to the toxic properties of 2,3-dimercaptopropanol.

Acknowledgemen t s This investigation was supported by FIPE (UFSM), FA-

PERGS (grant 92.1001.6 to J.B.T.R.) and C N P q (grant 523387/94-6 to M.E.P. and grant 523761/95-3 t o J.B.T.R.). T. E. is the recipient of graduate fellowship from CNPq/ UFRGS. During par t of this work T.E. was the recipient of a scientific initiation fellowship from C N P q (Programa Institucional UFSM 8/93-7/94). F.A.B. is the recipient of a Scientific Initiation Fellowship (Programa Institucional UFSM 8/95-7/96, CNPq).

References Barbosa, N. B. V., J. B. T. Rocha, G. Zeni, T. Emanuelli, M. C.

Beque & A. L. Braga: Effect of organic forms of selenium on 8- aminolevulinate dehydratase from liver, kidney and brain of adult rats. Toxicol. Appl. Pharmacol. 1998, 149, 243-253.

Barnard, G. E, R. Itoh, L. H. Hohberger & D. Shemin: Mechanism of porphobilinogen synthase -Possible role of essential thiol groups. J. Biol. Chem. 1977, 252, 8965-8974.

Beber, F. A., J. Wollmeister, M. J. K. Brigo, M. C. J. Silva, C. N. Pereira & J. B. T. Rocha: 8-Aminolevulinate dehydratase inhi- bition by ascorbic acid is mediated by an oxidation system existing in the hepatic supernatant. Internat. J. Vir. Nutr. Res. 1998, 68, 181-188.

Bergdahl, I. A., A. Grubb, A. Schiiltz, R. J. Desnick, J. G. Wetmur, S . Sassa & S . Skerfving: Lead binding to 6-aminolevulinic acid dehydratase (ALAD) in human erythrocytes. Pharmacology & Toxicology 1997, 81, 153-158.

Bevan, D. R., I? Bodlaender & D. Shemin: Mechanism of porphobi- linogen synthase. Requirement of Zn2+ for enzyme activity. J. Biol. Chem. 1980, 255, 2030-2035.

Beyersmann. D. & M. Cox: Affinity labelling of 5-aminolevulinate

dehydratase with 2-bromo-3-(5-imidazolyl)propionic acid. Bio- chim. Biophys. Acta 1984, 788, 162-168.

Block, C., R. D. Lohmann & D. Beyersmann: Probing of active site residues of the zinc enzyme 5-aminolevulinate dehydratase by spin and fluorescence labels. Biol. Chem. Hoppe-Seyler 1990,371, 1145-1152.

Bradford, M.: A rapid and sensitive method for quantitation of microgram quantities of protein ut ng the principle of protein- dye binding. Anal. Biochem. 1976,12, 248-254.

Chauhan, S . & M. K. O’Brian: A mutant Bradyrhizobium japonicum delta-aminolevulinic acid dehydratase with an altered metal re- quirement functions in situ for tetrapyrrole synthesis in soybean root nodules. J. Biol. Chem. 1995, 270, 19823-19827.

Chiba, M. & M. Kikuchi: The in vitro effects of zinc and manganese on delta-aminolevulinic acid dehydratase activity inhibited by lead or tin. Toxicol. Appl. Pharmacol. 1984,73, 388-394.

Dent, A. J., D. Beyersmann, C. Block & S . S . Hasnain: Two different zinc sites in bovine 5-aminolevulinate dehydratase distinguished by extended X-ray absorption fine structure. Biochem. 1990, 29, 7822-7828.

Ellman, G. L.: Tissue sulfhydryl groups. Arch. Biochem. Biophys. 1959,82, 70-77.

Emanueh, T., J. B. T. Rocha, M. E. Pereira, L. 0. Porciuncula, V M. Morsch, A. F. Martins & D. 0. Souza: Effect of mercuric chloride intoxication and dimercaprol treatment on 6-aminolevu- h a t e dehydratase from brain, liver and kidney of adult mice. Phurmacology & Toxicology 1996, 79, 138-143.

Fridovich, I.: Superoxide radicals and superoxide dismutases. Annu. Rev. Biochem. 1995, 64, 97-112.

Gibbs, P. N. B., M. G. Gore & P M. Jordan: Investigation of the effect of metal ions on the reactivity of thiol groups in human 5- aminolevulinate dehydratase. Biochem. J. 1985, 225, 573-580.

Gibson, K. D., A. Neuberger & J. J. Scott: The purification and properties of delta-aminolaevulinic acid dehydratase. Biochem J. 1955, 61, 618-629.

Goering. P. L.: Lead protein interactions as a basis for lead toxicity. Neurotoxicology 1993, 14,45-60.

Hasnian. S . S . , E. M. Wardell, C. D. Garner, M. Schlosser & D. Beyersmann: Extended-X-ray-absorption-fine-structure investi- gations of zinc in 5-aminolaevulinate dehydratase. Biochem. J. 1985, 230, 625-633,

Jaffe, E. K.: Porphobilinogen synthase, the first source of heme’s asymmetry. J. Bioenerg. Biomembr. 1995, 27: 169-179.

Kido, T. & K. Soda: Oxidation of anionic nitroalkanes by flavoen- zymes, and participation of superoxide anion in the catalysis. Arch. Biochem. Biophys. 1984, 234, 468-475.

Leussing. D. L. & T. N. Tischer: Mononuclear and polynuclear complex formation by manganese (11) and zinc (11) ions with 2,3- dimercapto-I-propanol: The behaviour of the E, function with mercaptide. J. Amer Chem. 1961, 83, 65-70.

Misra, H. €?: Generation of superoxide free radical during the au- toxidation of thiols. J. Biol. Chem. 1974, 249, 2151-2155.

Nelson, H. M., M. A. Ughes, & €? A. Meredith: Zinc, copper and delta aminolevulinic acid dehydratase in vitro and in vivo. Toxi-

Rocha, J. B. T., A. J. Freitas, M. B. Marques, M. E. Pereira, T. Emanuelli & D. 0. Souza: Effects of methylmercury exposure during the second stage of rapid postnatal brain growth on nega- tive geotaxis and on delta-aminolevulinate dehydratase of suck- ling rats. Brazilian J. Med. Biol. Res. 1993, 26, 1077-1083.

Rocha, J. B. T., M. E. Pereira, T. Emanuelli, R. S . Christofari & D. 0. Souza: Effect of treatment with mercury chloride and lead acetate during the second stage of rapid postnatal brain growth on delta-aminolevulinic acid dehydratase (ALA-D) activity in brain, liver, kidney and blood of suckling rats. Toxicology 1995,

Rodrigues, A. L. S. , J. B. T. Rocha, M. E. Pereira & D. 0. Souza: 6-Aminolevulinic acid dehydratase activity in weaning and adult rats exposed to lead acetate. Bull. Environ. Contam. Toxicol. 1996, 57, 3 M 5 .

cology 1981, 21, 261-266.

100, 27-37.

6-AMINOLEVULINATE DEHYDRATASE INHIBITION BY 2,3-DIMERCAPTOPROPANOL 103

Sakaguchi, S., T. Sakaguchi, I. Nakamura, M. Aminaka, T. Tanaka & Y. Kudo: Effect of beryllium on 6-aminolevulinic acid dehydratase and porphobilinogen deaminase in pregnant mice. Pharmacology & Toxicology 1996,19, 214-216.

Sassa, S.: Delta-aminolevulinic acid dehydratase assay. Enzyme 1982, 28, 133-145.

Shemin, D.: 5-Aminolaevulinic acid dehydratase: structure, func- tion, and mechanism. Phil. Trans. R. Soc. Lond. I976,273B, 109- 115.

Sommer, S. & D. Beyersmann: Zinc and Cadmium in 5-aminolevul-

inic acid dehydratase. Equilibrium, Kinetic, and 13Cd-nmr- studies. J. Inorg. Biochem. 1984, 20, 131-145.

Spencer, I? & I? M. Jordan: Investigation of the nature of the two metal-binding sites in 5-aminolaevulinic acid dehydratase from Escherichia coli. Biochem. J. 1994,300, 373-381.

Tsukamoto, I., T. Yoshinaga & S. Sano: The role of zinc with special reference to the essential thiol groups in delta- aminolevulinic acid dehydratase of bovine liver. Biochem. Biophys. Arm 1979, 570, 167-178.