1029
Chem. Res. Toxicol. 1995,8, 1029-1038
Enzymatic Methylation of Arsenic Compounds: Assay, Partial Purification, and Properties of Arsenite Methyltransferase and Monomethylarsonic Acid Methyltransferase of Rabbit Liver Robert Zakharyan,t Yuan Wu,t Gregory M. Bogdan,* and H. Vasken Aposhian*>? Department of Molecular & Cellular Biology and Department of Pharmacology & Toxicology, University of Arizona, Tucson, Arizona 85721 Received March 1, 1995@
A rapid, accurate, in vitro assay utilizing radioactive S-adenosylmethionine (SAM)has been developed for the methylation of arsenite and monomethylarsonate (MMA) by rabbit liver methyltransferases. The assay has been validated by separating, identifying, and measuring the products of the reaction using chloroform extraction, ion exchange chromatography, TLC, or HPLC. The enzymes involved in this pathway, arsenite methyltransferase and MMA methyltransferase, have been purified approximately 2000-fold from rabbit liver. After gel electrophoresis, a single band is obtained with both enzyme activities in it. The pH optima for purified arsenite methyltransferase and monomethylarsonic acid methyltransferase are 8.2 and 8.0, respectively. A thiol, S-adenosylmethionine, and arsenite are required for the partially purified arsenite methyltransferase that catalyzes the synthesis of monomethylarsonate. A different enzyme activity that catalyzes the methylation of monomethylarsonate to dimethylarsinate also requires S A M and a thiol. Even though arsenite methyltransferase and monomethylarsonate methyltransferase have different substrates, pH optima, and saturation concentrations for their substrates, whether the two activities are present on one protein molecule or different protein molecules is still uncertain. Both activities have a molecular mass of 60 kDa a s determined by gel exclusion chromatography. There is no evidence a t the present time for these enzyme activities being on different protein molecules. Neither arsenate, selenate, selenite, or selenide are methylated by the purified enzyme preparations. Results from the use of crude extracts, often called cytosol, to study the properties of these methyltransferases dealing with arsenic species should be viewed with caution since such crude extracts contain inhibiting and other interfering activities. Introduction The form (species) of arsenic influences the type and severity of intoxication (1, 2). For example, arsenate, which has a +5 oxidation state, can replace phosphate in many reactions in the body. It decreases energy stores by inhibiting ATP formation during glycolysis and also by uncoupling oxidative phosphorylation (3). On the other hand, the more toxic arsenite, which has a f 3 oxidation state, has a very high affinity for thiol groups, especially for vicinal thiols on proteins (4). Pyruvate dehydrogenase multienzyme complex, ketoglutarate dehydrogenase, thiolase, and glutathione reductase are examples of the enzymes inhibited by arsenite (5). These enzymes appear to have thiol groups in their active centers (5). In vivo studies suggest that in mammals arsenate must be reduced to arsenite before methylation can occur (2). The biochemical pathways for the biotransformation by fungi of arsenate and arsenite to trimethylarsine have been proposed for many years by Challenger (6)and by Cullen et al. (7). In their proposed biochemical pathways there were two basic steps, oxidative methylation and
* To whom correspondence should be addressed, at the Department of Molecular & Cellular Biology, Life Sciences South, Room 444, University of Arizona, Tucson, AZ 85721. Phone: 520/621-7565; FAX: 520/621-3709; email:
[email protected]. Department of Molecular & Cellular Biology. Department of Pharmacology & Toxicology. @Abstractpublished in Advance ACS Abstracts, August 15, 1995. +
0-
0-
0
I 3 As"I
0-
I
-0- &,m-
0-
0-
ARSBNATB
YH3 CH,
O m Asv-
I
0DIMETHYLARSINATE (DMA)
ARSENITE
?-
O m As"
I
CH,
0MBTHYLARSONATB IMMA)
Figure 1. Inorganic arsenate, arsenite, and methylated metabolites, monomethylarsonate and dimethylarsinate. reduction. S-Adenosylmethioninewas postulated as the methylating agent (8). Examination of arsenic compounds in the urine after administration of arsenate or arsenite to animals or humans has indicated that these arsenic species are biotransformed via methylation (Figure 1)(for reviews, see refs 9-12) in most mammals with the exception, apparently, of the marmoset monkey (13, 14). Methylation of arsenite has been proposed as a detoxification pathway. The idea was based initially on in vivo experiments (15-1 7). Methylation of these inorganic arsenic species (Figure 1) increased their water solubility and their rate of urinary excretion (18, 19). Methylation supposedly de-
0893-228x/95/2708-1029$09.00/00 1995 American Chemical Society
1030 Chem. Res. Toxicol., Vol. 8, No. 8, 1995
creased their toxicity as indicated by increased LD50 (20), a far from exact measurement of their toxicity. The results of other measures of toxicity (21-231, however, are increasingly leading to questions as to whether methylation of inorganic arsenic is unequivocally a detoxification mechanism. To date, however, the biochemical and molecular mechanisms responsible for the methylation of arsenic compounds by the mammalian body have not been clearly characterized. Neither are the identities and properties of the enzymes involved known. The biotransformation of arsenatelarsenite is not a n area free of controversy. Shirachi et al. (24) used subcellular fractions of rat liver for in vitro studies of the methylation of sodium arsenate and suggested that there were two different enzymes, one for methylation of arsenite and the other for methylation of methylarsonate. Buchet and Lauwerys (25, 26) were unable to confirm some of the results of Shirachi et al. (24). Using crude homogenates of rat liver, they studied cofactor requirements, pH, and other parameters of the reactions involved in the conversion of arsenatelarsenite to methylarsonate and dimethylarsinate ( 2 5 , 2 6 ) . Hirata et al. used homogenates prepared from livers or kidneys of rats and mice to study the conversion of arsenate and arsenite to MMAl and DMA (27, 28). Again, GSH or a thiol appeared to be necessary for optimum activity. Lerman and Clarkson reported studies with rat liver slices as well with hepatocytes (29). They suggested that arsenite is methylated by both liver and kidney. Their arsenate studies, however, might have been complicated by minimal uptake by liver slices. Elucidation of these enzyme mechanisms by which arsenite is biotransformed might enable a better understanding, prevention, and treatment of chronic arsenic poisoning, a major environmental health problem in a t least three countries, Chile, Mexico, and Taiwan (3033). For this to be done unambiguously requires rapid, dependable assays and purified preparations of the enzymes that methylate arsenic species. Most of the in vitro work that has been performed in the past has used crude extracts or slices of rat tissues (25-29). The results of such in vitro studies have been contradictory (24,341,perhaps because purified enzymes were not used or the methodology of arsenic analysis used to measure and identify the products of the reactions was either slow and difficult or even, as some have claimed, not reproducible (24, 34). The use of crude extracts for enzyme assays and for the understanding of enzyme mechanisms often yields ambiguous, inconclusive results as well as misunderstandings of what really happens in vitro and in vivo because of the presence of many interfering and/ or antagonizing activities. In addition, the National Research Council (35)has recommended that the rat not be used to study arsenic metabolism. Rabbit liver was chosen as the source of the methyltransferases for the present studies since the metabolism of arsenic compounds by the rabbit appears to be most similar to that by humans (9, 36). One of the major and most effective approaches, historically, for understanding the mechanisms of metaAbbreviations: SAM, S-adenosylmethionine;MMA, monomethylarsonate; DMA, dimethylarsinate; AAS, atomic absorption spectroscopy: DEAE, [(diethylamino)ethylIcellulose;PMSF, phenylmethanesulfonyl fluoride; PAD, periodate-oxidized adenosine; DMPS, 2,3dimercaptopropane-1-sulfonate, Na salt; DMSA, meso-2,3-dimercaptosuccinic acid: DTT, dithiothreitol.
Zakharyan et al.
bolic reactions has been the use of purified enzymes. While absolute purity is often a n unattainable goal, partial purification has yielded preparations free of many antagonistic as well as synergistic factors and reactions. Such purifications increase the possibility of unambiguously understanding the function and mechanism of an enzyme. In the present paper, a rapid assay for the methylation of arsenite and monomethylarsonate using radioactively labeled S-adenosylmethionine (SAM)is presented and validated. Such a rapid uncomplicated assay has made possible the 2000-fold partial purification of these activities by [(diethylamino)ethyllcellulose (DEAE) chromatography and molecular sizing chromatography. The use of these 2000-fold purified enzyme preparations has allowed a study of many of the properties of these enzymes, such as pH optimum, substrate requirements, and molecular mass of the activities, in the absence of many competing and complicating activities.
Experimental Procedures Reagents. Sodium arsenite and sodium arsenate were ACS reagent grade and were purchased from MCB Reagents (Cincinnati, OH). S-[methyL3H]Adenosyl-~-methionine (10.5 or 85 Cii mmol) was purchased from DuPont-NEN (Boston, MA). [l4C1Methylarsonic acid (4.5 mCi/mmol) was purchased from American Radiolabeled Chemicals Inc. (St. Louis, MO). [l4C1Dimethylarsinic acid (11.2 mCi/mmol) was a generous gift from Management Technology (Research Triangle Park, NC). Reduced glutathione was purchased from Sigma Chemical (St. Louis, MO). The sodium 2,3-dimercaptopropane-l-sulfonate was obtained from Hey1 (Berlin, Germany). All other chemicals were analytical reagent grade or of the highest quality obtainable. Animals. New Zealand White male rabbits (2.5 kg body weight) viral antigen-free were purchased from Myrtle's Rabbitry Inc. (Thompson Station, TN) and were acclimated, for 1-2 weeks prior to use, in an environmentally controlled animal facility with a 12 h darMight cycle at 22-24 "C. Animals were provided Harlan Teklad 0533 Rabbit Diet #7008 and water ad libitum. Nine rabbits were used i n these experiments. Incubation Conditions. The following conditions were used to assay arsenite methylation activity of purified fractions IIV: 0.10 M Tris-HC1 buffer (pH 8.01, 3.3 mM GSH, 1.0 mM MgC12, 20 pM sodium arsenite (As3+), 6.5 pmol (0.55 pCi) of (SAM)[specific activcarrier-free [3H]-S-adenosyl-~-methionine ity 85 Ciimmol, DuPont NEN Research Products (Boston, MA)], and enzyme preparation, all in a final volume of 250 pL. Samples were incubated for 60 min at 37 "C and placed on ice to stop methylation activity. When cytosol, fraction I, was assayed, the above conditions were used except t h a t the assay was performed at pH 7.0, using 60 pmol of [3HlSAM, 10.5 Ci/ mmol, at which point maximum methyltransferase activity was found in this fraction. During the purification procedure, assays of arsenite methyltransferase activity were incubated for 60 min. When arsenite methyltransferase was to be compared with MMA transferase activity, all assays were for 1.5 h. One unit of arsenite methyltransferase is defined as the amount of enzyme that will catalyze the formation of 1 pmol of monomethylarsonate in 1h at 37 "C. One unit of monomethylarsonate methyltransferase is the amount of enzyme t h a t will catalyze the formation of 1 pmol of dimethylarsinate i n 1.5 h a t 37 "C. Assays using monomethylarsonic acid (Pfaltz & Bauer Inc., Italy) a s the substrate were performed in the same volume as above containing 0.10 M Tris HC1 (pH 8.01, 1mM MgC12,3 mM GSH, 1 mM DTT (Sigma Co., St. Louis, MO), 1mM MMA, and 6.5 pmol(0.55 pCi) of [3H]SAM(85 Ci/mmol). Reaction mixtures were incubated for 1.5 h at 37 "C and then placed on ice to stop methylation activity. I t should be noted that the MMA concentration in this assay is 50 times greater than the arsenite concentration in the arsenite methyltransferase assay. It is also pertinent to point out t h a t all the results have been confirmed
As Methyltransferases: Assay and Purification at least three times using enzymes prepared at different times and from different rabbits. Standard Extraction Procedure and Assay of Methylated Arsenic Compounds. Reaction mixtures (250 pL) were pipetted into 5 mL polypropylene tubes, and the following were added: 10 pL of 40% KI; 20 pL of 1.5% potassium dichromate; 750 pL of 12 M arsenic-free hydrochloric acid (HC1); and 750 pL of chloroform. Tubes were then capped. The contents were mixed on a vortex for 3 min and centrifuged at 1500g for 3 min. The acidic aqueous (upper) phase contained S A M and was discarded. The chloroform (lower) phase was washed twice with a mixture of 250 pL of water, 5 pL of 40% KI, and 750 pL of 12 M HCl. Samples were mixed on a vortex and centrifuged, and the acidic aqueous (upper) phase was removed after each wash. The methylated arsenic compounds contained in the chloroform phase were back-extracted by adding 1 mL of water, vortexing for 3 min, and centrifuging at 1500g for 5 min. The aqueous (upper) phase (1 mL) was removed, mixed with National Diagnostics (Atanta, GA) Monoflow 3 scintillation cocktail (5 mL), and counted in a Beckman (Fullerton, CA) Model LS 7800 liquid scintillation counter. The above is designated the standard extraction procedure. The final aqueous phase was used also, in some experiments, for separating arsenic compounds by cation exchange chromatography, TLC, or HPLC i n order to confirm further the identity of the products of the enzyme reactions. Confirmation of Methylation Products. Several methods were used to confirm t h e identity of the products resulting from the methylation of arsenite or MMA, Le., MMA and DMA, respectively. (1) HPLC Method. A Phase Sep 25 cm x 4.6 mm C-18 ODS2 reverse-phase column was used to separate MMA and DMA. For MMA, the solvent was 20 mM tetrabutylammonium phosphate, pH 5.3. For DMA the solvent system was the same as t h a t of Hughes et al. (381,with tetrabutylammonium nitrate as the ion-pairing agent in 2-propanol and water. [14C]Monomethylarsonic acid (specific activity 4.5 mCUmmol) and [l4C1dimethylarsinic acid (11.2 mCUmmo1) were used as standards to determine the retention times. The assay mixtures were extracted as described above. Samples, 50 pL, were applied to the column, and MMA or DMA was detected with a n on-line radioisotope detector. (2) Ion Exchange Method. Bio-Rad AG 50W-X4 cation exchange resin (100-200 p m mesh) was packed into a Bio-Rad Poly Prep chromatography column to a 2 mL volume and treated with 0.5 N HC1 (30 mL), sufficient water until pH of effluent was 5.5, 0.5 N NaOH (30 mL), sufficient water until pH of effluent was 5.5, 0.5 N H C l ( 3 0 mL), and 0.05 N HCl(50 mL). Then 1 mL of the final aqueous phase extract of the enzyme reaction mixture from above was applied. The columns were eluted with 6 mL of 0.05 M HC1 to obtain MMA and 10 mL of 0.5 M NaOH to obtain DMA. Fractions (1mL) were mixed with 5 mL of Monoflow 3 cocktail and counted in a Beckman Model LS 7800 liquid scintillation counter. In the case of MMA methyltransferase assays, the samples were counted 12 h after addition of scintillation solvent. (3) TLC Method. The MMA formed during incubations was analyzed further by thin-layer chromatography to confirm its identity. Kodak (Rochester, NY) microcrystalline cellulose plates (20 x 20 cm) without fluorescent indicator were spotted with samples and developed with a butanollacetic acidwater (40:10:50) mobile phase. The solvent front moved about 15 cm. The Rf of a [l4C1MMA standard was compared to that of [14C]MMA which had been incubated with cytosol as above and with the [3Hlproduct formed during incubation with L3H1SAM. Plates were cut into 1 cm strips, which were placed i n vials. Scintillation fluid, 5 mL, was added. Samples were counted for either [3Hlproduct (formed during incubations) or [l4C1(standards) in a Beckman Model LS 7800 liquid scintillation counter. Purification of Methyltransferases. Rabbits (2-3 kg) were euthanized with COz. Their livers were removed rapidly, rinsed i n ice-cold 0.9% saline/l mM EDTA, blotted with filter paper, and cut into 5 g pieces. Pieces were minced and
Chem. Res. Toxicol., Vol. 8, No. 8, 1995 1031 homogenized using a Teflon-coated stainless steel pestle in a glass homogenizer after addition of buffer (1:2 w/v). Homogenization buffer (pH 7.6 at 4 "C) consisted of 10 mM Tris-HC1, 250 mM sucrose, 0.2 mM PMSF, and 0.5 mM GSH. The homogenate was centrifuged at 12000g for 15 min at 4 "C in a Beckman Model J2-21M inductive drive centrifuge to remove unbroken cells, cell membranes, mitochondria, and nuclei. The pellet was discarded, and the supernatant was further centrifuged at 105000g for 90 min at 4 "C in a Beckman Model L265B ultracentrifuge. This second supernatant was designated cytosol (fraction I). Cytosol preparations were either used fresh or stored at -70 "C until needed. The arsenic methyltransferase was purified by DEAE-cellulose, Sephadex G-200, and Sephadex superfine G-100 chromatography. Cytosol (30 mL) was quickly thawed and loaded on a DEAE-cellulose column (1.5 x 10 cm) which had been preequilibrated with 10 mM Tris-HC1 (pH 7.6; 4 "C) at a flow rate of 1 m u m i n . The column was eluted with a gradient of 200 mL of 30 mM Tris-HC1 and 200 mL of 30 mM Tris-HCU0.5 M NaCl (pH 7.6; 4 "C) using a flow rate of 0.3 m u m i n . Fractions (5 mL) were collected at 4 "C and were assayed for each of the arsenic methylation activities. The DEAE-cellulose fractions containing the maximum methylation activity were combined and precipitated by (NH&S04. The precipitate t h a t appeared at 55-80% saturation was dissolved in buffer A (30 mM TrisHCll50 mM NaC1, pH 7.6) at 4 "C and further purified using a Sephadex G-200 (2.5 x 47 cm) column equilibrated with buffer A. The flow rate was 14 mL/h. The active fractions were combined and precipitated by (NH&S04 to 100% saturation, dissolved in 0.5 mL buffer A, and applied to a Sephadex superfine G-100 column (2.5 x 47 cm). Elution was at a rate of 8 mL/h with buffer A a t 4 "C. Fractions (3 mL) were collected. Fractions containing the peak of enzyme activity were used for all studies unless otherwise stated. Other Methods. Protein concentrations of solutions were measured by the method of Bradford (39) using bovine serum albumin standards.
Results Effectiveness of Standard Extraction Procedure. An extraction procedure as described in the Experimental
Procedures section was used to separate radioactive S A M and its degradation products from radioactive MMA and DMA, the putative products of the methyltransferases. This chloroform extraction procedure was found to be effective. Less than 0.1 f 0.02% SD, or 0.06 pmol ( n = 5), of SAM was found to be present in the final aqueous phase which had been back-extracted from chloroform. The effectiveness of the standard extraction procedure for isolating the two methylated arsenic species, MMA and DMA, was also determined. To do this, [l4C1MMA (100-800 pmol) or [l4C1DMA(125-170 pmol) was added as a standard and incubated in the absence of SAM. The amount recovered in the final aqueous phase was 94.7 f 0.02% SD ( n = 4), and 84.6 k 6.7% SD ( n = 5), respectively. Quantitation and Confirmation of MMA and DMA Separation. MMA and DMA were extracted by the standard extraction procedure from reaction mixtures and then separated on cation exchange columns in order to confirm the identity and amount of each formed. [14C]MMA or P4C1DMA was added to reaction mixtures containing cytosol, incubated, chloroform extracted, and separated on the columns. For controls, equivalent amounts of [l4C1MMA or [l4C1DMA were used as standards and separated on the columns. The columns were eluted with 0.05 M HCl to obtain MMA and 0.5 M NaOH to obtain DMA (Figure 2). There was no difference between MMA or DMA elution profiles for incubated
1032 Chem. Res. Toxicol., Vol. 8, No. 8, 1995
Zakharyan et al.
"1lmiAA lOQ0
\\ 5.0
0
t
-&@SI HU+ .
10.0
11.0 0.-
1-~
20.0
pI
DMA
0.0
NaOH
5.0
10.0
15.0
20.0
Elution Volume (mL) Figure 2. Elution profiles of MMA and DMA from cation exchange columns. [14C]MMA or [14C]DMA was added to enzyme (6 mg of fraction I protein) reaction mixtures for arsenite or MMA methyltransferase, respectively, and incubated. The following were added to the reaction mixtures (250 pL) after incubation: 10 pL of 40% KI, 20 p L of 1.5% KLh-z07,750 pL of 12 M HCl, and 750 yL of chloroform. Samples were mixed on a vortex for 3 min and centrifuged at 1500g for 3 min. The acidic aqueous (upper) phase was discarded, and the chloroform (lower) phase was washed twice more with 250 p L of water, 5 p L of 40% KI, and 750 p L of 12 M HC1. The chloroform phase was back-extracted with 1 mL of water, which was placed on columns, eluted, and counted in a scintillation counter (Treated). Standard solutions of MMA or DMA were placed on the columns and eluted (Control).
(treated) or standard (control) samples, thus confirming that the extraction procedure separated MMA and DMA. Formation of MMA after Incubation of Arsenite with Rabbit Liver Cytosol. After verifylng that cation exchange chromatography separated MMA and DMA, arsenite was incubated with r3H]SAM using cytosol as the source of arsenite methyltransferase. The amounts of the methylated metabolites formed were determined (Figure 3A). MMA was the predominant metabolite (more than 90% of the methylated products) formed during incubations under these conditions, namely, using arsenite as the substrate. In order to further confirm the identity of the product of the reaction when arsenite and [3HlSAM were incubated with rabbit liver cytosol, the standard extraction procedure was carried out, and thin-layer chromatography and HPLC were then used to compare the migration of the product to that of the [14C]MMAstandard. The Rf value of the [3H]-methylatedproduct was 0.56 & 0.01 SD and corresponded to that of the [14C]MMAstandard with or without the presence of cytosol: 0.56 f 0.01 SD and 0.57 f 0.01 SD, respectively. Using HPLC, the retention time for the product formed after incubating L3H1SAM with arsenite methyltransferase was 6.01 min f 0.03 SD ( n = 3). It was 5.99 f 0.03 SD ( n = 2) for [14C]MMA standard. This is additional evidence that the methylated product was MMA. Formation of DMA after Incubation of MMA with Rabbit Liver Cytosol. Only DMA was formed when
MMA was the substrate, as shown by ion exchange chromatography (Figure 3B) or by HPLC. With HPLC, the retention time of the [3Hlproduct formed during the incubation of MMA and L3H1SAM with MMA methyltransferase was 5.03 min (5.08-4.97 range) & 0.06 SD. The retention times of standards P4C1DMA and P4C1MMA were 4.69 (4.75-4.66 range) f 0.06 SD and 8.99 f 0.04 SD, respectively. (It is pertinent to emphasize that, for significant amounts of DMA to be formed, a concentration of MMA is needed that is 50-fold greater than the concentration of arsenite needed to form MMA.) Two controls were always performed a t the same time enzyme assays were performed. The blank control was the reaction without addition of enzyme. The other control was the reaction omitting arsenite or MMA in order to monitor potential methylation of other compounds rather than the arsenic-containing substrates. Properties of the Methyltransferases in Crude Liver Cytosol. A number of the properties of these activities in crude homogenates of rabbit liver cytosol were different from those of the 2000-fold highly purified enzyme fractions which have been prepared in this laboratory. Since preliminary work dealing with mechanisms of metabolic reactions are often performed using crude homogenates containing impure enzymes, we include some of the properties of these two methyltransferases in crude homogenates. It is emphasized, however, that the 2000-fold partially purified enzymes are more appropriate for studying the requirements, properties, and mechanisms, both molecular and biochemical, of these enzyme reactions. The pH optima for activity of the enzymes in crude extracts was, for example, different from the optima for the purified enzymes. After validating our assay system, the effect of varying the pH of the reaction on the amount of MMA formed was examined (Figure 4A). Incubations a t final pH 7.3-8.5 were performed in the presence of 0.04 M Tris-HC1 buffer, while those a t pH 5.5-6.8 were done in the presence of 0.05 M phosphate buffer. The optimum pH for incubations of the rabbit liver cytosol fraction was determined to be 6.8, and the total amount of methylated products formed was 4.5 pmol. Other incubation components were altered to determine their effect on arsenite methylation. Varying the concentration of NaCl from 160 to 400 mM had no effect on the amount of methylated metabolites formed (data not shown). Varying the amount of MgClz added to incubations from 0 to 1.6 mM also had no significant effect on arsenite methylation in these crude extracts of rabbit liver (data not shown). Purification of the Methyltransferases. The arsenic methyltransferases were purified approximately 2000-fold from rabbit liver homogenates by DEAEcellulose, ammonium sulfate, Sephadex G-200, and Sephadex superfine G-100 chromatography (Table 1). The methylation activities, using either arsenite or MMA as the substrate, were in the same fraction throughout the purification procedure (Table 1, Figure 5). It is pertinent to point out that more activity was recovered after DEAE chromatography (fraction 11)than had been placed on the column (compare fractions I and 11, Table 1). This type of increased recovery during enzyme purification strongly suggests the presence of an inhibitor of the methyltransferases in the rabbit liver cytosol. Requirements for Methyltransferase Enzymes of Arsenic Species. Using the 2000-fold purified enzymes
As Methyltransferases: Assay and Purification
Chem. Res. Toxicol., Vol. 8, No. 8,1995 1033 1500C
EV f 7
0
I
4000-
4
2000-
3
1000-
0
0
0.7
B C
-
4.0-
-
UNA
0.6 0.5
n 3.0-
a
f
0.4
2
Oe3
a 2.0-
3
8
0.2
$
1.0-
0.1
8.0
l
t OJM NaOH-1
D
0
A
7.0
C
500-
5.0
6.0
FO.OSM H
8 10 12 14 16 18
1000-
-
5.0
6
3000-
LV
-0 . 0
4
1500
B
5e
2
9
0
PH
Figure 4. (A) pH curve for methylation of arsenite by rabbit liver cytosol. Assay conditions are as described in the Experimental Procedures. Fraction I, 6 mg of protein was used. (B) pH curve for purified rabbit liver arsenite methyltransferase activity and monomethylarsonate methyltransferase activity. Assay conditions are as described in the Experimental Procedures. Fraction V (3.7 pg of protein) was used for the arsenite methyltransferase, and fraction V (6 pg of protein) was used for MMA methyltransferase.
(fraction VI, the major requirements for the reactions appear to be the arsenite or monomethylarsonate substrate of the appropriate methyltransferase, S-adenosylmethionine, and GSH or some other thiol (Table 2). In the absence of substrate or enzyme, there was less than 1%of chloroform-extracted radioactivity found as compared to the complete reaction mixture. (1) pH Optimum. The pH optima of purified arsenite methyltransferase and MMA transferase were 8.2 and 8.0, respectively (Figure 4B). It is pertinent to note that this was the pH of the reaction mixture a t the beginning of the reaction, using fraction V, and was not the optimum for the cytosol fraction (Figure 4A).
(2) Substrate Specificity and Concentration. At concentrations of 20, 50, and 100 pM, arsenite is approximately 140, 60, and 25 times more active, respectively, than arsenate as a substrate of arsenite methyltransferase (Figure 8A). When the 2000-fold partially purified fraction V was used as the source of the enzyme, neither selenate, selenite, nor selenide was active as a substrate. In the absence of any arsenic species, reaction mixtures containing 200 pM selenite, selenate, or selenide gave results similar to those found in reactions in which enzyme was omitted. Arsenite appears to saturate arsenite methyltransferase a t about 50 pM (Figure 6A). Substrate saturation of MMA methyltransferase occurred a t about 1000 pM (Figure 6B). The arsenite methyltransferase assay was linear as to amount of protein used from 0 to 1.6 pmol of MMA formed, while the MMA methyltransferase assay was linear to 1.0 pmol of DMA formed (data not shown). (3) Confirmation of the Products of the Reactions. Although the products of these enzyme reactions using crude cytosol (fraction I) and a reaction pH 6.8 have been identified previously by cation exchange chromatography, TLC, and HPLC, the products were again identified using the highly purified fraction V with its pH optimum of 8.2 or 8.0. Under these conditions, when the product of the arsenite methyltransferase reaction was examined by cation exchange chromatography (Figure 3C) and TLC, it was found in the same position as marker MMA. Using cation exchange chromatography (Figure 3D) and HPLC, the product of the MMA methyltransferase reaction was shown to be DMA. (4) Other Properties. The enzymic conversion of arsenite to MMA is relatively rapid (Figure 7A). When this laboratory has followed the synthesis of DMA from MMA, there always has been a consistent, definite, and pronounced lag of 30 min.
1034 Chem. Res. Toxicol., Vol. 8, No. 8, 1995
Zakharyan et al.
Table 1. Purification of Arsenite Methyltransferase (I) and MonomethylarsonicAcid Methyltransferase (11)" act. (pmoVmL) fraction (I) cytosol (11) DEAE-cellulose (111)ammonium sulfate (N) Sephadex G-200 (V) Sephadex G-100
vol (mL) 30 45 3 42 12
protein (mg/mL) 30 1.8 12.7 0.28 0.06
I 1.5 9.5 115.6 5.9 6.2
sp act. (pmoVmg)
I 0.05 5.2 9.1 20.3
I1 0.6 5.42 49.8 1.8 2.5
100
total units (pmol)
I1 0.02 3 3.92 6.4 42
I 45 420 347 249 75
purifn (x-fold)
I1
I
I1
18 243 149 76 30
104 180 400 2000
150 196 320 2100
a Assays of arsenite methvltransferase activity were incubated for 60 min; assays of monomethylarsonic acid methyltransferase activity were incibated for 90 min. "
0
0.25
10.0
20
5 10 15 20 25 30
Fraction
7.5
7.5
5.0
5.0
2.5
2.5
0.0
1
20 22 24 26 28 30
Fraction
0.0 Fraction
Figure 5. (A) DEAE chromatography profile of arsenite methyltransferase and monomethylarsonic acid methyltransferase. (B) G200 chromatography profile of arsenite methyltransferase and monomethylarsonic acid methyltransferase. (C) Sephadex G-100 elution profiles of arsenite methyltransferase and monomethylarsonic acid methyltransferase.
Table 2. Requirements of Methyltransferase Reactions arsenite methyltransferase (As3' MMA) complete" (-1 MgC12 (-1 arsenite (-1 GSH
-
MMA formed (pmol)
(-1 l3H1SAM (-1 enzyme
1.012 0.909 0.010 0.008
