Perspective

pubs.acs.org/crt
Terms of Use

Chemical Mechanism of Arsenic Biomethylation

William R. Cullen*
Department of Chemistry, University of British Columbia, Vancouver, British Columbia, Canada V6T 1Z1

ABSTRACT: The bioconversion of inorganic arsenic to methylated metabolites aects the

tissue distribution and retention of arsenic and its actions as a toxicant or carcinogen. Although
enzymes that catalyze the methylation of arsenicals have been identied in all branches of the
tree of life, fundamental questions persist about the chemical processes that underlie reactions
that methylate this metalloid. Here, several reaction schemes for arsenic methylation are
considered to encourage careful consideration of the chemical plausibility of these schemes.

CONTENTS
Signicance of Arsenic Biomethylation 457
containing As(V), and in a transplacental exposure model in the
mouse, methylarsonous acid (MMA(III)) is a carcinogen.37
These ndings and the detection of MMA(III) species in
Chemical Pathways for Arsenic Methylation 458
mammalian urine, especially in human populations exposed to
Conclusions 459
Author Information 460
arsenic in their drinking water, has provoked considerable
Corresponding Author 460 interest in the mechanism of their formation.8,9 Indeed, the
Funding 460 linkage between the biomethylation process and formation of
Notes 460 methylated metabolites containing either As(III) or As(V)
Acknowledgments 460 remains a central question in the study of arsenic as a toxicant
Abbreviations 460 and carcinogen.
References 460 Biomethylation of arsenic to trimethylarsine was conrmed
in fungi by Frederick Challenger and his co-workers in 1933.10
Figure 1 shows the stepwise path involving oxidative addition

SIGNIFICANCE OF ARSENIC BIOMETHYLATION

All living organisms are exposed to arsenic in one form or
followed by the reductive elimination he later proposed for
enzymatically catalyzed methylation.11 Challenger suggested
that active methionine, later identied as S-adenosylmethio-
another and have evolved strategies to cope with this exposure. nine (SAM), was the methyl group donor. Notably, the
One of the most commonly encountered is the reduction of Challenger pathway is analogous to the uncatalyzed oxidative
arsenate, which is adventitiously taken up into a cell via the
phosphate transport system, to arsenite that is subsequently
transported out of the cell with the help of an eux protein.
Because arsenite is more toxic than arsenate, the cell interior
cannot be considered to be detoxied. The same is true for the
cell exterior. Even this process is not straightforward because
the arsenic(III) can be further metabolized to nontoxic
trimethylarsine oxide. This methylation of arsenic is another
process that was once regarded as detoxication because the
widely distributed, stable, and easily analyzed methylarsenic(V)
species are much less toxic than their inorganic precursors.
However, the nding that the less stable, and harder to analyze,
methylarsenic(III) species are more toxic than inorganic arsenic
species has slowly put an end to that belief.1,2
Figure 1. Challenger pathway for biological methylation of arsenic.
The oxidation state of arsenic in inorganic and methylated
species is an important determinant of toxicity in a wide range
of biological systems. For example, arsenicals containing As(III) Received: November 27, 2013
are typically more cytotoxic and genotoxic than are homologues Published: February 11, 2014

2014 American Chemical Society 457 dx.doi.org/10.1021/tx400441h | Chem. Res. Toxicol. 2014, 27, 457461
Chemical Research in Toxicology Perspective

Figure 2. Equations describing the biological methylation of arsenic. Equation 1: methylation by the Challenger pathway with the As(III) species
shown ligated to three RS moieties. Abbreviations: S-adenosylmethione, SAM; S-adenosylhomocysteine, SAH. Equation 2: methylation by the
pathway proposed by Hayakawa and coworkers.22 Equation 3: methylation by the pathway proposed by Wang and co-workers based on studies with
arsenic (+3 oxidation state) methyltransferase (As3mt).31 Equation 4: a modied methylation scheme consistent with the Challenger pathway and
based on studies with arsenite methyltransferase (ArsM).

addition reaction known as the Meyer reaction that is used to the methylation of arsenic by bacteria due to the lack of the
prepare MMA(V) from arsenite and methyl halide.12 Indeed, analytical approaches with the sensitivity required for detection
the Challenger pathway can be fully modeled by using the of the methylated products. However, in modern times such
trimethylsulfonium ion as methyl donor and sulfur dioxide as methylation has been frequently reported; for example,
the reducing agent.13 Michalke and coworkers20 conrmed that anaerobic bacteria,
It should be noted that Challengers proposed pathway was typically those found in sewage digesters, are capable of
largely based on the nding that inorganic arsenic(III) and (V) methylating arsenic, and they report, for example, that
and mono- and dimethylarsenic(V) species were converted to trimethylarsine is the main product from the methogenic
trimethylarsine (TMA) by fungi such as Scopulariopsis archaea Methanobacterium formicicum, Methanosarcina barkeri,
brevicaulis. Challenger was not able to identify any of the and Methanobacterium thermoautotrophicum.

postulated intermediate species of the pathway in the culture
medium due to the lack of analytical tools with the required CHEMICAL PATHWAYS FOR ARSENIC
sensitivity. Hence, he viewed the transformations as occurring METHYLATION
in a fungal black box. Some years later, Cullen and co-workers
identied methylated intermediates in culture media and used Questions about the validity of the Challenger pathway have
deuterium labeling to provide further evidence supporting the prompted speculation about alternative pathways for arsenic
Challenger pathway in fungal and algal cultures.1419 Initially, methylation. Any proposed pathway must be chemically
researchers in this eld assumed that Challengers description plausible; that is, each proposed step in the reaction scheme
should be taken literally with each redox step well delineated must conform to our knowledge of chemical kinetics and
and associated with the release of specic intermediates or nal thermodynamics. In the following paragraphs, I consider the
products. However, it became clear that more than one redox Challenger pathway and several alternatives in these terms.
step could take place before the release of an intermediate or Challengers pathway (Figure 1) makes clear predictions
nal product and that these depended on the organism, the about the reaction in which a methyl group is transferred to an
culture medium, and the composition and concentration of the arsenic atom, about the charge on the methyl group, and about
arsenical substrate.14,15,18,19 For example, arsenate (1 mM) is the oxidation state of the arsenic atom during and after the
rapidly (2 days) reduced to arsenite by the fungus Apotrichum transfer. The pathway is usually written in terms of oxy-species,
humicola (formerly Candida humicola).19 This metabolite is but we can be reasonably sure that AsS bonding plays a major
then slowly (30 days) converted to trimethylarsine oxide role because of the kinetic stability of the AsS bond to
(TMAO) that is released into media along with small amounts hydrolysis (one of the sources of the well-known anity of As
of other methylarsenic(V) species: monomethylarsonic acid for S). Electrons for reduction of the methylarsenic(V) species
(MMA(V)) and dimethylarsinic acid (DMA(V)). In contrast, to methylarsenic(III) probably come from oxidation of two
the unicellular alga Polyphysa peniculus rapidly metabolizes thiols to a disulde as in the real or notional reductive
arsenate to arsenite and DMA, but MMA is absent, and elimination reaction suggested for model systems: R3As(SR) 2
trimethylarsenic species are not found in either the media or R3As: + RS-SR.16 In enzymatically catalyzed reactions,
the cells. Notably, neither DMA nor MMA are metabolized by physiological dithiols such as thioredoxin or glutaredoxin which
this alga.18 Once again, Challenger was not able to demonstrate are reversibly oxidized likely provide these electrons.21
458 dx.doi.org/10.1021/tx400441h | Chem. Res. Toxicol. 2014, 27, 457461
Chemical Research in Toxicology Perspective

Figure 2 summarizes postulated steps in the methylation of prokaryotes.29,30 Wang and co-workers examined methylation
arsenic by the Challenger and alternative pathways. Equation 1 catalyzed by recombinant human As3mt using a physiological
in Figure 2 shows a variant form of Challengers pathway in monothiol (glutathione, cysteine) as the sole reductant.31 They
which the As(III) reactant is written as a tris-thiol derivative suggested that an As(III)-containing substrate initially binds to
such as arsenic tris-glutathione. Here, transfer of the electro- three thiolate residues that are shown as protein-bound S atoms
phile CH3+ from SAM to an As(III) atom yields S- (curved bonds) in the rst step of eq 3, Figure 2. Then,
adenosylhomocysteine (SAH), a neutral species, and a following Hayakawa and co-workers,22 they postulate that the
methylated arsenical containing an As(V) atom. This is a methylated arsenic product remains as As(III) and that a
chemically plausible reaction scheme because there is no protein-bound S, displaced from the arsenic during the
possibility of an unfavorable electrostatic interaction between methylation reaction, binds to the demethylated and doubly
the positive leaving group and the uncharged SAH. charged SAM. This reaction is followed by a novel reduction
Hayakawa and co-workers22 have postulated that arsenic(III) step that does not involve an As(III) intermediate. Here, the
species persist during methylation reactions and that oxidation disulde formed between the displaced S and demethylated
to arsenic(V) species occurs (somehow) after methylation. In SAM is reduced to generate SAH. This reaction is claimed to be
their proposed scheme (Figure 2, eq 2), methylation initially linked to a conformational change in As3mt that releases SAH.
requires the unfavorable release of a negatively charged methyl Unfortunately, the model also fails to address the central
group (CH3) from positively charged SAM, which then problem with such a reaction scheme; namely, how does CH3
displaces an RS-group from As(III). To minimize but not leave a positive center?
eliminate unfavorable electrostatic interactions, the authors A recent paper by Ajees and co-workers32 oers an
suggest that the doubly charged sulfur species formed in this opportunity to examine the two pathways shown in eqs 1
reaction by loss of CH3 from SAM interacts with the thiol and 2 (Figure 2) as they would apply to the methylation of
displaced from arsenic to revert to a singly charged sulfonium arsenite by SAM catalyzed by ArsM. This structural study
species. Hayakawa and co-workers suggest that the methylated provides a model of the enzymes active sites to which inorganic
product of reactions shown in eq 2 is either released as a As(III) and SAM are bound. In a fully charged ArsM molecule,
methylated As(III) species or oxidized during release to the methyl group of SAM is poised to be transferred to the
produce a methylated As(V) species. However, isolation of arsenic which is initially bound to three thiolate-containing
TMAO from cultures of A. humicola described above is an cysteinyl residues (Cys 72, Cys174, and Cys 224). These
unambiguous example of the direct release of an As(V) thiolates can be thought of as equivalent to the protein-bound S
metabolite. In this instance, the postulated precursor As(III) atoms that interact with arsenic as shown in eq 3 (Figure 2).
metabolite, the gas trimethylarsine, is actually produced by the The authors did not write out their reaction scheme but,
fungus only at much higher concentrations (>1 ppm) of following others,22,31 we can use the rst step of eq 3 as a
arsenical substrates.23,24 Therefore, release of TMAO into the framework to keep As in a trivalent oxidation state. Ajees and
media of fungal cultures containing lower concentrations of co-workers predict that a cysteinyl residue, Cys 72, in ArsM is
arsenical is not the result of atmospheric oxidation of the leaving group that forms the SS bond with demethylated
trimethylarsine (half-life in air at 20 C: two days25), which is SAM as shown in eq 3. However, this process is not conrmed
not consistent with the pathway of eq 2. The putative by their structural model, which shows that Cys 72 moves away
methylarsenic(III) precursors are not very stable and would from SAM during the reaction rather than binding to it.
not have been detected in the media in the early studies Finally, eq 4 of Figure 2 gives our interpretation of
because hydride generation under acid conditions was used for enzymatically catalyzed methylation according to the Chal-
the analysis: both MMA(III) and MMA(V) would be detected lenger pathway. In this scheme, there are no problems with
as methylarsine and both DMA(III) and DMA(V) as charge, and it is easy to see that binding another thiolate to the
dimethylarsine. arsonium center either from the protein, there are many
Reductive methylation, is the heart of an alternative cysteinyl residues in ArsM and As3mt, or from exogenous thiol
pathway proposed by Naranmandura and co-workers.26 They reactants would neutralize the charge on arsenic and provide an
suggest that the oxidative methylation and reductive opportunity for reductive elimination of a disulde, leaving the
elimination reactions of the Challenger pathway occur methylarsenic(III) species bound to the protein. The nal
simultaneously so that the real reaction product is an product of eq 4 is also poised to accept another methyl group
arsenic(III) species. In our opinion, this reaction scheme from SAM to yield a DMA(V) species which would be bound
would signicantly reduce the nucleophilicity of the lone to the enzyme by two AsS bonds and would be more likely to
electron pair on arsenic and inhibit the reaction. be released by hydrolysis than the precursor MMA(V) species
Although these alternative pathways are consistent with which is bound by three AsS bonds. This dierence in
evidence that methylated products containing As(III) are often susceptibility to hydrolysis may account for the general
bound to protein targets, they do not account for the preponderance of DMA(V) over MMA(V) in biological
observation that methylated products containing either As(III) systems. The failure of rat and human As3mt to methylate
or As(V) are products of enzymatically catalyzed methylation MMA(V) may be because there is no available reduction path
reactions.27,28 to aord the necessary enzyme bound methylarsenic(III)
intermediates. 29

Other reaction schemes have been suggested based on
evidence from studies that used puried arsenic methyltrans-
ferases. Two such proteins have been identied and their genes CONCLUSIONS
cloned. Arsenic (+3 oxidation state) methyltransferase (As3mt) On the grounds of chemical plausibility, the Challenger
catalyzes arsenic methylation in a wide range of higher pathway with SAM as a donor of CH3+ remains the most
organisms, and arsenic methyltransferase (ArsM) catalyzes rational option to describe the biological methylation of arsenic.
these reactions in Archaea, some eukaryotes, and many We anticipate that future structural and functional studies with
459 dx.doi.org/10.1021/tx400441h | Chem. Res. Toxicol. 2014, 27, 457461
Chemical Research in Toxicology Perspective

As3mt and ArsM will further rene this model. In particular, (8) Le, X. C., Ma, M., Cullen, W. R., Aposhian, H. V., Lu, X., and
although the reaction scheme postulated by Hayakawa and co- Zheng, B. (2000) Determination of monomethylarsonous acid, a key
workers is cited (for example, see Figures 1 and 3 in ref 33), it arsenic methylation intermediate, in human urine. Environ. Health
should be regarded as very speculative, and at least, the Perspect. 108, 10151018.
existence of the sulfonium species with a SS bond required in (9) Valenzuela, O. L., Borja-Aburto, V. H., Garcia-Vargas, G. G.,
Cruz-Gonzalez, M. B., Garcia-Montalvo, E. A., Calderon-Aranda, E. S.,
eq 2 needs to be veried experimentally. The high toxicity of
and Del Razo, L. M. (2005) Urinary trivalent methylated arsenic
the methylarsenic(III) metabolites remains a problem no
species in a population chronically exposed to inorganic arsenic.
matter which methylation pathway is favored. Environ. Health Perspect. 113, 250254.
Finally, one of the groups initially supporting Hayakawas (10) Challenger, F., Higgenbottom, C., and Ellis, L. (1933) The
model now reports34 that the theoretical calculation shows the formation of organo-metalloid compounds by microorganisms. Part I.
methyl group transfer process to be a typical in-line SN2 Trimethylarsine and dimethylethylarsine. J. Chem. Soc., 95101.
nucleophilic substitution reaction in many SAM-dependent (11) Challenger, F. (1945) Biological methylation. Chem. Rev. 36,
methyltransferases. iAs3+ with lone pair can attack the CH3. 315361.

AUTHOR INFORMATION
Corresponding Author
(12) Quick, A. J., and Adams, R. (1922) Aliphatic arsonic and arsinic
acids and aliphatic aromatic arsinic acids. J. Am. Chem. Soc. 44, 805
816.
(13) Antonio, T., Chopra, A. K., Cullen, W. R., and Dolphin, D.
*E-mail: wrc@chem.ubc.ca. (1979) A model for the biological methylation of arsenic. J. Inorg. Nucl.
Funding Chem. 41, 12201225.
I am grateful to the Canadian Water Network for nancial (14) Cullen, W. R., Froese, C. L., Lui, A., McBride, B. C., Patmore, D.
support. J., and Reimer, M. (1977) The aerobic methylation of arsenic by
microorganisms in the presence of L-methionine-methyl-d3. J.
Notes
Organomet. Chem. 139, 6169.
The authors declare no competing nancial interest.

(15) Cullen, W. R., McBride, B. C., and Pickett, A. W. (1979) The

transformation of arsenicals by Candida humicola. Can. J. Microbiol. 25,
ACKNOWLEDGMENTS 12011205.
I thank David J. Thomas, US Environmental Protection (16) Cullen, W. R., McBride, B. C., and Reglinski, J. (1984) The
Agency, and Ken Reimer, Royal Military College, Kingston, reduction of trimethylarsine oxide to trimethylarsine by thiols; a
Canada, for encouragement. mechanistic model for the biological reduction of arsenicals. J. Inorg.

Biochem 21, 4559.

(17) Cullen, W. R., Li, H., Hewitt, G., Reimer, K. J., and Zalunardo,
ABBREVIATIONS N. (1994) The identification of extracellular arsenical metabolites in
MMA(III), methylarsonous acid; MMA(V), methylarsonic the growth medium of microorganisms. Appl. Organomet. Chem. 8,
acid; SAM, S-adenosylmethionine; SAH, S-adenosylhomocys- 303311.
tein; TMA, trimethylarsine; As3mt, arsenic (+3 oxidation state) (18) Cullen, W. R., Pergantis, S., Li, H., and Harrison, L. G. (1994)
methyltransferase; ArsM, arsenite methyltransferase The methylation of arsenate by a marine alga Polyphyses peniculus in

the presence of L-methionine-methyl-d3. Chemosphere 28, 10091019.

REFERENCES (19) Cullen, W. R., Li, H., Pergantis, S. A., Eigendorf, G. K., and
Mosi, A. A. (1995) Arsenic biomethylation by the microorganism
(1) Reimer, K. J., Koch, I., and Cullen, W. R. (2010) Organo- Apiotrichum humicola in the presence of L-methionine-methyl-d3.
arsenicals. Distribution and Transformation in the Environment, in Appl. Organomet. Chem. 9, 507515.
Organometallics in Environment and Toxicology. Metal Ions in Life (20) Michalke, K., Wickenheiser, E. B., Mehring, M., Hirner, A. V.,
Sciences (Sigel, A., Sigel, R., and Sigel, K. O., Eds.) Vol. 7, pp 165229, and Hensel, R. (2000) Production of volatile derivatives of
RSC Publishing, Cambridge U.K. metal(loid)s by microflora involved in anaerobic digestion of sewage
(2) Rosen, B. R. (2002) Biochemistry of arsenic detoxification. FEBS
sludge. Appl. Environ. Microbiol. 66, 27912796.
Lett. 529, 8692.
(21) Waters, S. B., Devesa, V., Del Razo, L. M., Styblo, M., and
(3) Styblo, M., Del Razo, L. M., Vega, L., Germolec, D. R., LeCluyse,
Thomas, D. J. (2004) Endogenous reductants support the catalytic
E. L., Hamilton, G. A., Reed, W., Wang, C., Cullen, W. R., and
function of recombinant rat cyt19, an arsenic methyltransferase. Chem.
Thomas, D. J. (2000) Comparative toxicity of trivalent and
pentavalent inorganic and methylated arsenicals in rat and human Res. Toxicol. 17, 404409.
cells. Arch. Toxicol. 74, 289299. (22) Hayakawa, T., Kobayashi, Y., Cui, X., and Hirano, S. (2005) A
(4) Mass, M. J., Tennant, A., Roop, B. C., Cullen, W. R., Styblo, M., new metabolic pathway of arsenite: arsenicglutathione complexes are
Thomas, D. J., and Kligerman, A. D. (2001) Methylated trivalent substrates for human arsenic methyltransferase Cyt19. Arch. Toxicol.
arsenic species are genotoxic. Chem. Res. Toxicol. 14, 355361. 79, 183191.
(5) Nesnow, S., Roop, B. C., Lambert, G., Kadiiska, M., Mason, R. P., (23) Cox, D. P., and Alexander, M. (1973) Production of
Cullen, W. R., and Mass, M. J. (2002) DNA damage induced by trimethylarsine gas from various arsenic compounds by three sewage
methylated trivalent arsenicals is mediated by reactive oxygen species. fungi. Bull. Environ. Contam. Toxicol. 9, 8488.
Chem. Res. Toxicol. 15, 16271634. (24) Pickett, A. W., McBride, B. C., Cullen, W. R., and Manji, H.
(6) Kligerman, A. D., Doerr, C. L., Tennant, A. H., Harrington-Brock, (1981) The reduction of trimethylarsine oxide by Candida humicola.
K., Allen, J. W., Winkfield, E., Poorman-Allen, P., Kundu, B., Funasaka, Can. J. Microbiol. 27, 773778.
K., Roop, B. C., Mass, M. J., and DeMarini, D. M. (2003) Methylated (25) Jakob, R., Roth, A., Haas, K., Krupp, E. M., Raab, A.,
trivalent arsenicals as candidate ultimate genotoxic forms of arsenic: Smichowski, P., Gomez, D., and Feldmann, J. (2010) Atmospheric
induction of chromosomal mutations but not gene mutations. Environ. stability of arsines and the determination of their oxidative products in
Mol. Mutagen. 42, 192205. atmospheric aerosols (PM10): evidence of the widespread phenomena
(7) Tokar, E. J., Diwan, B. A., Thomas, D. J., and Waalkes, M. P. of biovolatilization of arsenic. J. Environ. Monit. 12, 40916.
(2012) Tumors and proliferative lesions in adult offspring after (26) Naranmandura, H., Suzuki, N., and Suzuki, K. T. (2006)
maternal exposure to methylarsonous acid during gestation in CD1 Trivalent arsenicals are bound to proteins during reductive
mice. Arch. Toxicol. 86, 975982. methylation. Chem. Res. Toxicol. 19, 10101018.

460 dx.doi.org/10.1021/tx400441h | Chem. Res. Toxicol. 2014, 27, 457461

Chemical Research in Toxicology Perspective

(27) Shen, S., Li, X.-F., Cullen, W. R., Weinfeld, M., and Le, X. C.
(2013) Arsenic binding to proteins. Chem. Rev. 113, 77697792.
(28) Ding, L., Saunders, R. J., Drobna, Z., Walton, F., Xun, P.,
Thomas, D. J., and Styblo, M. (2012) Methylation of arsenic by
recombinant human wild-type arsenic (+3 oxidation state) methyl-
transferase and its methionine 287 threonine (M287T) polymorph:
Role of glutathione. Toxicol. Appl. Pharmacol. 264, 121130.
(29) Thomas, D. J., Li, J., Waters, S. B., Xing, W., Adair, B. M.,
Drobna, Z., Devesa, V., and Styblo, M. (2007) Arsenic (+3 oxidation
state) methyltransferase and the methylation of arsenicals. Exp. Biol.
Med. (Maywood, NJ, U.S.) 232, 313.
(30) Thomas, D. J., and Rosen, B. P. (2013) Arsenic Methyltransfer-
ase, Encyclopedia of Metalloproteins (Kretsinger, R. H., Uversky, V. N.,
Permyakov, E. A., Eds.) pp 138143 Springer, New York.
(31) Wang, S., Li, X., Song, X., Geng, Z., Hu, X., and Wang, Z.
(2012) Rapid equilibrium kinetic analysis of arsenite methylation
catalyzed by recombinant human arsenic (+3 oxidation state)
methyltransferase (hAS3MT). J. Biol. Chem. 287, 3879038799.
(32) Ajees, A. A., Marapakala, K., Packianathan, C., Sankaran, B., and
Rosen, B. P. (2012) Structure of an As(III) S-adenosylmethionine
methyltransferase: Insights into the mechanism of arsenic biotrans-
formation. Biochemistry 51, 54765485.
(33) Watanabe, T., and Hirano, S. (2013) Metabolism of arsenic and
its toxicological relevance. Arch. Toxicol. 87, 969979.
(34) Li, X., Geng, Z., Chang, J., Wang, S., Song, X., Hu, X., and Wang,
Z. (2013) Identification of the Third Binding Site of Arsenic in
Human Arsenic (III) Methyltransferase. PLoS One 8, e84231.

461 dx.doi.org/10.1021/tx400441h | Chem. Res. Toxicol. 2014, 27, 457461