Chapter19

Geochemical and Biological Controlsover Methylmercury Production and Degradation

in Aquatic Ecosystems

J. M. Benoit l, C. C. Gilmour\ A. Heyea~. R.'P. MasonJ

,

and C. L. Miller'

'Chemistry Department, Wheaton College, Norton, MA 02766JThe Academy ofNatural Sciences, Benedict Estuarine Researeh Centerl

St..Leonard, MD 20685 ''University of Maryland Center for Environmental Science.

Chesapeake Biological Laboratory, Solomon!, MD'20688

It is the goal of this paper to discuss the more salient recentAdvancCll in the understanding of the controls of net CH3Hgformation in Datural systems. The discussion lIighlights thegaps in knowledge and the areas where progress inunderstanding has occurred. In particular, this chapter focuseson recent developments in Hg bioavailability and uptake bymethylating bacteria, on the competing roles of sulfate andsulfide in the control of methylation, and in pathways fordemethylation. The role of sulfide in influencing methylationis discussed in detail. In addition, the impact of otherenvironmental variables such as pH, dissolved organic carbonand temperature on mercury methylation are discussed.Lastly, we provide a synthesis of the variability in themethylation response to Hg inputs across ecosystems. Wesuggest that although methylation is a function of HgconcentratioD, the range of methylation rates acrossecosystems is larger than the range in Hg deposition rates.Overall, we conclude that factors in addition to the amount Hgdeposition playa large role in controlling CH3Hg productionand bioaccumulation in aquatic ecosystems.

262 to 2003 American Chemical Society

263

htlroduction

Mercury (Rg) inputs to the environment have been increased dramaticallysince indU$triali2:ation and a;nthropogenic sources ofHg to the atmosphere nowdomioatt the input (]-3). While inorganic Hg is the major source ofHg·to mostaquatic systems, it is rnethylmercwy (CH:!Hg) thatbioconcentrates in aquaticfood webs and is the source of health advisories worldwide that caution 8.&ainstthe consw:nption of fish containing elevatedCH3Hg (4-7). AlthoughasmaUfraction of the Hg in atmoSpherio depOsition is CHJHg. the dominantsou,rce ofCH3Hg to most aquatic systems is in situ fotrnation, or formation within thewatershed (8-12). ThecUrrentcon:sensus~ based mainly on temperalure-­dependency of Hg methylation and its response to biological subs:trateS (J3-16),is that biological methylation of inorganic Hg to CH3Hg is more important thanabiotic processes in natui'alsystems. Biological methylation Was firstdemonstrated in the late 1960's (I7) and it is now generally accepted thatstdfatereducing bacteria (SRB) are the key Hgmethylai:ors (13.18~2U) although anumber of organisms beSides SRBs have boen shown to produce CH~g in purec.ulture !rqm added Hg(ll) (21).

The role of SIms in methylation has been demonStrated through studiesusing specific: metabolic inhibitors, addition of sulfate,a:nd coincidentmeasurement of $uIfate reduction rate and CH]1Ig production. The addition ofBES. a specific inhibitor of methanogens, was shown to increase Hg methylationwhile molybdate, a spccificin.htbitor ofsulfate reduction, dranultically decreasedCH3Hg production in saltmarsh sediment (13). Since thiselirly study, molybdateinhibition of mercury methylation, and coincident depth-profiles of sulfate·reduction rate and Hg methylation have demonstrated the importanCe of SRBs inestuarine, freshwater "lake. saltmars:il, and Bverg;1ades sediments (19,22·:24).Furthermore, addition of sulfate has been shown to stimulate mercury.methylation in concert with stimUlation of sulfate reduction, most notably duringthe whole lake sulfuric acid addition experiment in Little Rock Lake. WI(18,25,26) and in a series of 'Short and 10000-terin sulfate addition studies infreshwater ecosystems (19,20,24,27). However, while ·snlfatc stimUlates bOthsutfllte reduction and mercury methylation at low sulfate concentrati()ns~ thebuild up of diSSQlved sulfide at high sulfate conccnttationscan inhibit Hgmethylation (28"30). The mechanism of sulfide irihibition of Hg methylation isdiscussed in detaU in this chapter. The sulfate addition experiments suggest thatincreased anno$pheric $ulfuric acid deposition ill this century ("acid rain") mayhave lead 10' enhanced Hg methylation in remote freshwater ecosystems (20.24r

264

Overall, lUi many factors influence both methylation and the reve!se reaction,demethylation, in situ CH)Hg concentration is a complex function of its rate offormation and loss.

Community structure studies, using molecular probes and other techniques,have shown correspondence between the distribution of certain types ofSRB andHg methylation in sediments. and between sulfate reduction rate and Hgmethylation rate (31-34). The primary site of methylation is just below theoxic/anoxic interface, which is often near the sediment surface in aquatic systems(9,16,35-37). It should be noted, however, that CH)Hg can be produced inenvironments where sulfate reduction is low, such as upland soils, where otherbacteria and fungi may be important methylators. However, little work has beendone in these upland environments, as studies have rather focused onenvironments within aquatic ecosystems whoro the CH3Hg produced hill! greatestlikelihood of eJitering the aquatic food chain, and where sulfate reduction is adominant degradation pathway for organic matter in sediments. Even thoughsulfate reduction and Hg methylation are linked, it should be noted that someSRB can methylate Hg while growing fermentatively (38).

One obvious mediator of Hg methylation rate is the concentration ofinorganic Hg substrate, and its chemical form. Although there is a significantrelationship between Hg and CH3Hg across ecosystems. Hg does not appear tobe largest source of variability in CHJ.Hg production among ecosystems. Therelationship between Hg and CH3Hg concentrations in surface lake, river andestuary sediments and wetland soils across many ecosystems is weak but there is,on average. about 1% of tbe total Hg as CH]Hg for the lower concentration«500 nglg) sites (Figure 1), which represent the range in Hg concentration ofnatural, unimpacted environments. The measured concentration at any timepoint is an integration of the impact of all the processes influencing CH3Hg, suchas differing loading rates (39,40) and methylation and demethylation rales (16),which vary spatially and temporarily (with season and temperature), Suchvariation is not accounted for in the data used in this plot. which includepublished and unpublished values from ongoing studies - see Figure caption forreferences. Only data collected and analyzed using trace-metal~free techniqueswere included here and the relationship is geographically biased, and favorscontaminated systems. Additionally, as the data were not normally distributed, alog relationship is plotted (rl = 0.41; p<O.Ol; Fig, 1).

The data in Figure 1 apfear to cluster into two sets, with Hgconcentrations exceeding 500 ng g. having little increased impact on CH]Hgproduction. For ecosystem types, the relationsh~ has been found to besignificant for.estuaries (r = 0.78, p<O.Ol), lakes (r = 0.64, p---O.Ol) and riverR(r2 =0.68. p=O.Ol) but not for wetlands (r=0.29, p>0.05) based on data in Figure

265

1. Overall, within single. rivers or wetlands. or even clusters of simi1~r

ecQsystems, significant relationships ClUl exist, but tberelationships currentlyhaveno predictive power, given the importance of oth~ parameters, discussedbelow, in influencing, m,ethylatipn tate by controlling the availability of Hg to.and activity of, the m~ylatiilg bacteria. and given oUr cuttent level ofunderstanding. EffectiVe regulation of Hg pollution requires the ability to predictthe rehitionship between Hg andCH~Hgam(jng ecosystems, a goal thatresearchers and modelers seem to be slOWly approaching. Detailed investigationsof the mechanisms of CH)Hg formation, degradation, fate and transport arerequired, so that the factors controlling the levels of CH)Hg in fish can beunderstood. Clearly, while many other factors influence Hg ~ethylation. thesupply and availability of Hg is a key parameter;

111 addition to the effect of sulfide, other chemical factors influendngmethylation include the supply of labile carbon (57,58) although the role ofdissolved organic caroon (DOC) is complex. The distribution of methylationactivity is tied to the distribution of biodegradable-organic matter butcomplexation of Hg by DOC may influence He: bioavaiIability. Maximal netmethylation is often observed in surface sediments (15.16) where microbialactivity is greatest due to the input of fresh organic matter. As a result, systemswith high levels of organic matter production, such as wetlands, recently floodedreservoirs, or periodically flooded river Plains. may exhibit extremely high ratesof methylmercury production (10,42,59,60). New rcsearthon Hg complexationwith DOC is highlightul below. Temperature is another important variable (61)as the temperature responses of methylation and ~ylation have beenreported to differ (/6,62). However,sell#onal changes in Hg complexation thataffect methylationaml demethyiatioll differently could account for thc(seobservations. .

Demethylation of CH3Hg can occur via a number of mechanisms. includingmicrobial deIIl(!.thylationand reduction by mer operon-mediated pathways, andby "oxidative demeth)'lation processes" (21,63-7/).1n addition, photochemicalCa3Hg degradation in the wa~ colurnn has been demonstrated (72). The m.er­based pathway is, an inducible detoxific~tion mechanism. while oxidativedemethylation is thought to be a type of Cl metabolism. Recent researchsuggests that oxidativedemethylatiop is the dominant process in u,ncontalIli.n&tedsurftU:e sediments (65.70~71).

It is the goat of this paper to discuss the more salie-nt rt;C:eDt advances in theunderstanding of the controls of net CH3Hgformation in natural systems.Therefore. rather than being a complete review 'of thelitcrature, this- chapter willprovide an in-depth examinlltion of some of the pertinent ~nt papers andcurrent develppments. and will endeavor to highlight the gaps in knowledge andthe ateas where progre1;$ ill un~standing hMo,cc\1tW4.In particular, this-chapter focuses on recent developments in Hg bioavailabiIity and up,take by

~66

Figllre 1. Mercury (Hg) and methylmen:ury (CH;Ig) in near sur/rwe (0-4 em )sediment infreshwater wetlandsfrom: North and South Carolina (41), Ontario,CtJnatfn. (42), Plorida Everglades (37): Marine and Estuarine sedlmen,ts]rom:

cuastaL N. (lnd S. CaroJina (34). The Chesapt!akJ! Bay amLitsEstilaritts (43,44),coastaL F16ri.da (45), cca.stal Texas (46), Slovenia coast (47), cOlUtal Poland(48) coaslti.l Malaysia (48), Anadyr Estuary, Russia (48); lAkes:: Nf!W Jerstry(41), New York State (49).Wisconsin (41,50), Califomia (51), Finland (:.52),PolaiJd(48); Rivets: S. Carolina (41), Wisconsin (53), Nevada (54), Alaska

(55), am"""y (56), Polimd (48).

~

268

methylating bacteria, on the competing rol~ of sulfate and sulfide in the controlof methylation, and in pathways for demethylation. Lastly, we provide asynthesis of the variability in the methylation response to Hg inputs acrossecosystems. We suggest that although methylatio,n is a function of Hgconcentration, fue range of methylation rates across ecosystems is larger than therange in Hg deposition rates, and that fac;tors in addition to Hg deposition playalarge role in controlling CHJHg production and bioaccumulation in aquaticecosystems.

Mercury Speciation and Methylation

Although mercury resistant bacteria possessing the mer operon have theability to actively transport Hg(II}, this operon is not present in SRB thatmethylate mercury (38). It is generally accepted that CH3Hg is produced in anaccidental side reaction of a metabolic pathway involving methylcobalamin (73),although this pathway has only been demonstrated in one SRB. Therefore, it isnot likely that SRB have acquired an active transport for this toxin. A limitednumber of experiments with SRB support this idea (38). For this reason,diffusion across the cell membrane has been proposed as the important uptakemechanism (30,38,74). This hypothesis is consistent with studies that havedemonstrated diffusion of neutral mercury complexes (chloride complexes)across artificial membranes and into diatom cells (75-77).

The diffusion hypothesis is also supported by the relationship betweenmethylation and the distribution of neutral Hg sulfide complexes in sediments. Ithas been noted in many field studies that the rate of Hg methylation, and theCH3Hg concentration in sediments, decrease as the sediment sulfide concen­tration increases (28,29,36,37,74,78,79). A number of mercury complexes existin solution in the presence of dissolved sulfide, including HgSO, Hg(SH)t,Hg(SH)+, HgS;Z· and HgHSz"(80-84) and it is possible that inorganic Hg uptakeby SRB occurs via diffusion of the dissolved neutral Hg complexes, such asHgS°. If so, then the bioavailability of Hg to the bacteria would be a function ofsulfide .levels, as this is the ligand controlling Hg speciation in solution in lowoxygen zones where SRB are active. It has been shown through chemicalcomplexation modeling that the speciation of Hg tends to shift toward chargedcomplexes as sulfide levels increase (74,84,85), decreasing the fraction of Hg asuncharged complexes, and, as a result. the bioavailability of Hg to methylatingbacteria.

The existence of neutrally charged Hg-S complexes, and the notion ofdecreasing bioavailability in the presence of sulfide, was demonstrated in thelaboratory using a surrogate measure of membrane penneability, the octanol­water partition coefficient (K"",; 85). These experiments showed that HgSO and

269

Hg(SHhO, the neutral complexes present under the experimental conditions, hadrelatively higb K"w's such that they would be taken up at rates more thansufficient to account for methylation in both pure cultures and in the field, basedon the estimated permeability through the cell membrane. which is a function ofKow (30,74,86). There is the potential for polysulfide formation in porewatersand the solubility of Hg in the presence of HgS(s) has been shown to bedramatically increased by the complexation of Hg with polysulfide species (87).However. these interactions do not appear to significantly enhance theconcentration of neutral Hg species, as measured by changes in the Kow.suggesting that the dominant polysulfide complexes are charged (e.g., HgSxOH"and Hg(Sxt: 87) and thus unavailable for uptake and methylation by SRB.

It should be noted here that theoretical chemical calculations suggest theneutral species in solution is HOHgSH" rather than HgSO, as it is likely that Hgwould form a more stable linear complex (88). The results of Benoit et al. (30)are not in disagreement with this notion even though HOHgSH" would have alower permeability, because of its larger molecular volume, than RgSo. Therelationship between permeability and Kow is extremely non linear (89) andHOHgSH" would have a permeability that is a factor of five less than that ofHgSo. In the pure culture studies, estimated uptake was greatly in excess ofmethylation rate (Figure 2a) even if the neutral complex is considered to beHOHgSH".

Experiments with pure cultures and other studies indicate that not all the Hgtaken up by the bacteria is methylated as there are other sinks within the cell forHg (30). This is shown illustratively in Figure 2b, where Hgc represents the poolof Hg available for methylation inside the cell. Given uptake of a neutral Hgcomplex ([HgL"°D, the intracellular steady state concentration of Hgc is givenby:

where kM is the methylation rate constant, leo is the diffusion rate, HgLR0 is theconcentration of neutral complex in solution and ks is the rate constant thatincorporates all the other processes that are rendering Hg unavailable formethylation within the cell. The rate of formation of CH)Hg, assuming no lossmechanisms, is then given'by:

If 4t is much greater than ka, then the rate of CH)Hg formation is directlyrelated to the rate of diffusion across the membrane. However, in the oppositecase, the rate of methylation is dependent on both the rate of uptake and the rateat which Hg is being sequestered within the cell. However, there would still be a

270

Fig.2a1!" ,-----------,

Up

, 10HgS' Cone. (pg em'"

_ 1()4

7a 10'...l'l ,0'

l10't:~~~S ,00~ 10-

1 CH3Hg production11)'20.1

Fig.2b

Hg2+ + xS2- +yH<4- = HgSiit2x+y

e.g HgS', HgS.~, HgS,H', Hg(SH),'

HgS" kDHg(SH),' ...-----...

Figure 2. a) Estimated mercury (Hg) uptake rate, assuming pt:lSSive diffusion ojneutral Hgcomplexes - either modeledtzsHg$'or HOHgSFr, and the

simultaneousHg methylation raU inpute cultures ofD. propionicus. b) Modelrepresentation. ofthe assumed process ofHg I,lpfake and methylation within this

organism. The raleofmethylation is ihsignated tis kM; kB is the combined rate of'compe(lng reactionstharse:quester Hgand make it unavailable for methylation

and kv is the uptaJr.e fate. See text for details.

271

linear relationship betw-een the neutral complex uptake rale and methylation rateacross a sulfide gradient, for example. if the: ratc constanl. kB• was relativelyindependent of the environmental conditions outside the ceU. which is likely. Atthe limit, if ka: »~. methylation will not occur to any measurable degree.

Studies with la~or;ll:ory tult~ of Desidfobulbus proprionicus across asulfide gradient have shown that the up:take,rate. as estimated from the dissolvedHg speciation and the estimated permeability of the comple-xes, and methylationratecl1anged in a similar fashion. and both illcrellSed with increasing HgSO

concentration (Figure 4), supporting the above hypothesis (86). These resultsalso suggest that·ks is greater than~ in this case and that a small fraction of theHg taken up is methylated. The relative percent methylated 'Would depend On anumber of factors that would differ bet~n organisms, due fo differences inphysio-Jogy, size and membrane wmposition. We propose tbat differences in Hgpartitioning within cells may partially eJtplain the large differences inmethylation rates among varioussttains ofSRB.

The' hypothesis of neutral complex. bioavailability controlling methylationbegs the question why methylation appears'to becQnfined mostly to SRBs in the·environment Clearly, these organi'sms dominate in the region wnere HgS"dominates the Hg speciation. Why then, in oxiC or s\iboxicenvironments. whereHgCl2°and other neutral Hg complexeseldst. is there is little evidence-ofmethylationt In studies with diatoms, Masonet aI. (76) 'demonstrated that littleof the Hgelt taken up (<10%) reac:hed the cytoplasm of the organisms, withmost Hg being rapidly bound within the ce11ul'ar metiibrane. On the other hand, .CH:;HgCl was less strongly bound within the membrarie and a greater fractionwas found in the cytoplasm of the diatom (631h). The intratellular distribution isrelated to the rate and degrccpf reaction of the accumulated complex withcellular sites. The rate of sequestration (kB), which would detennine where theHg would become bound within a cell, depends ontbe ex;ch~ge reactionbetween the neutral complex and the cellular reaction site (Rl{); for example,HgSO +.RH = HgR+ + SH. The kinetics of this process would be:: to sOme degreedetermined by the reaction mechanism (adjunctive. or disjunctive), but are astrong functiOcn of the J:'eJative magnitude of the equilibri.um COnstants for theaccumulated 'complex. and HgR+. Given that most cellular binding sites for Hg-are likely !bioI groups, the rate of the exchange: r€;8.ctionshould be faster forligCh than for HgSO. as the associated equilibrium conStant (HgCIiO + RH =HgR+ + W + 2en is much greater for HgCli. Furthermore; given the highstability of UgSO, its tate of dissociation will be slower than that of HgCIi.Therefore, in the -presence of HgSo. a higher fraction of the Hg diffusing aeraslithe ollfer membrane is transferred to the site within 'thecyto(>lll$m wheremethylation CWl occurcQUlpaIed to other organisms, beeaus:eof th~ kinetics ofthe intracellular reactions.

272

Thus; it is not sufficient that the pathway for methylation exists within anorganism but alsotliat the Hg can be transported within the cell to the site of thereaction, AS discUSsed below, methyliltion has been linked to the aettyl-CoApathWaY in one bacterium (73) and a simple-cxplanationof why some organismswhich have this pathway do not methylate Hg is that the Hg is being: bound toother intracellular sites before being transported to the site-of methylation. Thus,there are kinetic and biochemical factors that influence the relative degree ofIMthylation between organisms besides Hg bioavailabiIityand Hg content in themedium The kinetics and location of the intercellular reactions are an importantmodifier of the methylation rate and cJearly more studies: on tbe intracellularmechanisms of methylation are needed.

A simple model of Hgpartitioning in sediments was developed tQ explorehow Hg partitioning tQ solids impacl$ Hg 'Concentration and bioavailability insediment pore waters (74). The model developed used adsorption reactions asthe mecbanism controlling porewater concentration. since field measurementsshow that pure cinnabar equibbrium diS$olution dramatically over predicts !heconcentration of Hg in porewateT$. It should also be noted that there is littlefield evidence to support the formation of the pure HgS solid p~ inenvironmental media (9c·92). In the model. water colwnn speciati()D was drivenprimarily by sulfide concentration. The adsorption of Hg was modeled wit~ tWOtypes ofbinding sites. singly or double coordinated thiols, which could representeither an interadion with organic thiol -and/or inorganic sulfide groups in thesolid phase. This model was not only able to predict both the measureddistribution of Hg in sediment porcwaters in two ecosystems (the Patuxent Riverand the Ronda Everglades) but the model-predicted concentration of neutral Hg­S comptexescorre1ated well with the in·Silu CH3Hg concentration (74).

The model am;l the Jaboratory culture studieS (30,74,86) cover thespectitiOnofHg 10 the presence of sulfide but it is known that Hg binds strongly to (jrganicrnatter~ and that dissolved organic carbon (DOC) impacts Hg methylation. Theimpact of organic content on Hg metbylatic)8 appears to be comple;K (16,57,58).While Hg binds strongly to DOC, laboratory complexation studies using DOCisolates from the Florida Evet:glades Stigg~t thar this binding is nor sufficient forHg-OQCcompiexation to be important insysrellls where sulfide is present(93.94). Thus, while DOC has been shown to be the most important complexingligand in .surface waters in the absence. of sulfide (94,95), it is likely Eo beunimportant in Hg complexation in sediment pol'twaters under typical DOCconcentrations and >0.01 $M sulfide (93,94), However, binding of Hg to1)fgani~ m.atter is importailt in the solid phase. Laboratory studies suggestrhat inoxic regions. organic complexation is much more important than binding of Hgto metal oxide phases in all e~ept very low organic matter sediments (96). Ithas been suggested that Fe-Sfonnation scavengeS Hg in anoxic regions· of thesediment (9?)and that Hg binds strongly to pyrite such that, even when only

273

small amounts are present, it is the dominant solid phase binding Hg insediments (90). These Studies focused on regions of low organic COntent and. inc~mtrasl> our sediment sequentialextraetion studies (96) show tha:! Hg isassociated With the organic fraction even in the presence of significant solidsulfide phases (AVS and pyrite); Furthennore, it has been show.l\ that, thesediment particle-dissolved distribution coefficient (K.!) is a strong function oforganic carbon (46). rn the environment" concentrations ofFe. S and C typicallyco~'vary in sediments, and all often correlate with Hg, and it is therefore difficultto ascertain from field data which is the ultimate controlling phase (44.98).Laboratory and field studies t96.99,lOO) suggest that the binding of Hg toorganic matter involv~ interacpon with tiJ.c tbiol groups of the; organicmolecules and thus, in a sense, the complexation of Hg to inorganic sulfidepllases or to organic matter are CQUlparableas both involve: the interactionbf;tween Hg and a reduced _S specieS. "-

'the role of pH needs to be considered as the complexation with sulfide andthiols jnvolv~ acid-ba;se chemistry. An inverse correlation between lake waterpH and mlOrtury in fis:htilisllC5 has, ~nob:;crved in a nwnborof studies (101and rtfererttes therein) stlggesting that pH influences methylation anddemetbyJation in aqmltic etQ$yStems. In some freshwaler studies, methylationwas reduced with decre:asing pH (27,35) while the impact on demethylation wassmall. In other studies; increasing rateS of mercury methylation in epilimneticlake _ waters and at the sediment surface were found with lowered pH(57;102.103). Winfrey and Rudd (35) reviewed potential mechanisms for lowpH effects-on mercury methylation and suggested that changes in mercurybinding could account for the seemingly conflicting results seen' in aU of thesestudies, They pomted -Out that lowering pH may lead to increased association ofmercwy with solid phases, decreased dissolved pore water mercury, and(presumably) to lower availability of Hg(lI) to baderia. The model discussedabove (74) can 00 used M a simple prc:xJictor of lIm impactor pH on Hgmethylation. Considering the reaction of Hg With the solid phase (RSH + Hg2+ "'"RSHg+ +Hi, and the diss.olved speciation, the following overall reaction can bepostu-Iated:

RSHg+ +HzO+HS' =HgSC>+RSH+H~

In the pH range of 7-10 (PKat -7 for H1S and assuming the pK for RSH isaroUnd IO), an increase in pH. at cOIistant'sulfide, will result in an increase inHgSC> relative to RSHg+ ([HS]/[RSH] is essentially constant) and thusmethy1lltion should increase with pH. Below a pH of 7. decreasing pH(increasing IH') leads- to decreasing [HSl; and as a result> HgSO will decreaserelative to RSHg+ with pH Lee., Oiethylation rateshouJd decttuc. This theoreticalcoru.ideration supports: the: notions put forward by Winfrey and Rood (35) and

274

suggests lhat a decrease in pH will lead to a decrew;c in methylation rate insediments because of changes in the concentration of bioavailab1e Hg inporewaters. The magnitude of the effect will depend on the pH range 8$!heimpact of pH is more marked at low pH. Overall, these considerations suggestthat sulfide concentration will have the most significant impact on Hgbioavailability in porewater but tbatother fuctors such as organic matter coh:teilt,pH, temperature and the presence or absence of inorganic sulfide phases all playa role in controlling Hg bioavaiIability to methylating bacteria.

The conflicting influences of ·sulfate and sulfide on the extent of Hgmethylation are welt iUusthUed by the studies in the Florida Everglades (37).Studies oYer four years at eight sites that cover a large gradient in sulfilteandsulfide showed that tbehighest methylation mtes, and the highest %CH)Hg in thesediment. w¢re at sites of intermediate sUlfate-reduction rates and sulfideconcentration (Figure 3). ]n the EvergJades, the north-south trend in sulfateconcentration l~s to a similar trend in sulfate reduction rateandporewatersulfide. As the sulfide concentration decreases, the relative concentration ofpte4icted HgSO concentration Incteases.The peak in methylation rale resultsfrom th~ combination of the increasing availability of Rg. to the SRB coupledwith th~ deCreasing sulfate reduction tate north to south. These results confirmthe importance of bOth Hg speciation. and bacterial activity, in controlling Hgmethylation ratc. Overall, the sites with the high~t HS_ methylation are thosethat also have the highest fish CH3Hg concentrations (l(4), confirming the -directlink betWeen theex.teilt ofHgmethylation and fish CH3Hg levels.

E"lperimentii inwbich Bverglades sediments were incubated with additionalslilfate orstilfide further demonstrated the interplay betWeen bactprial activityand Hg speciation. In cores ·taken from a relatively low sulfate site, addition ofsulfate stimulated methylation, and sulfate reduction. over thataf unamendedcontrol treatments eveR though s-ulfide levels increased slightly (see exampleexperiments in Figure 4-). In these sulfate--limited sediments, the higherinducedbacterial atlivity more than compensated for the slightly lower bioavailability ofHg at the higher sulfide levels. Addition of sulfide alone however resulled ininhibitiOn of methylation. It is- clear from this and o-ther experiments thatinhibition occurs at low lPM sulfide concentrations in Everg4ldessediments.However, high rates of Hg methylation have been demonstrated in highlysulfidic saltmarsh sediments (32). Perhaps the high rates of sulfate·reduetion inthese sediments make up for the very low percentage of dissolved Hg that wouldex,ist as neutral species. For Everglades sites with higher ambient sulfate,addition of,sulfate did not increase methylation but addition ofmore sulfide ledto an irihibition of Hg .methylatiQn.The field data across:s:ites show a decreasein methylation rale when concentration's of·sulfide inCi"etiSe above 10 lPM(Figure l), consistent with the core. incubation data. Overall, the resiilts of thefield and laboratory studies show that the balarice, betweeil sulfate availability,

27~

which controls SRB activity, and sulfide production and accumulation, whichcontrol Hg bioavailablity, are critical in modeling methylation rates. Ongoingmesocosm studies in the Everglades and in a boreal ecosystem Should providemore quantitative equations for Hgcycling models.

Studies in a number of sires have DOW dettl6nstrated a relatively strohgrelationship between the concentration of CHJ{g in ,sediments and theinstantaneous (short~term:) rate of Hg rncthylation. Methylation rates Can beestimated by Hg spike additions, either as a radioacnveor astable isotope,preferably to microbial Communities held relatively intact (18.41,103,105)(Figure 5). The methylation rate constant, kM• is calculated as the amount of newisOtopic CH]Hg formed per unit time, divided by the pool size of substrate.Methylation rate is derived by multiplying kt4 by the total Hg pool size. Use: ofeither custom~synthesized.high specific~activity 203Hg, or the use of individualHgstableisotopes combined with analysis by ICP-MS, allows methylationmeasurements to be done at relatively low, near ambient levels. In addition.when using stable isotopes it is possible to track both the in situ Hg and theadded Hg spike and compare relative methylation rates (106). Furthermore,methylation and de:metbylation can be measured simultaneously in the cores ifdifferent isotopes fife used for Hg and CH]Hg (41,105, 106}.

Figure 5 shows relationships between methylation Tates and ambient CH]ligconcentration. in sets of 1m diameter enclosures 'at four sites across the FloridaBverglades. In these studies, ~C1z was aQd«i to the surface water of theenclosures. and CH]~ production was fonowed o,ver time in surfatesediments. Additionally, 5 em sedimontcores were removed from the enclosuresfor the estimation of instantaneous methylation rat~ using 19!1IgC4, injected intothe cores. The figure sbows the concentration of CHl02HK in sedimenl;S intllespiked enclosures after 51 days, and productioo ofCHl""'1ig in cores after Zhours, both in comparison with in situ sediment CH")Hg concentrations. Shortterm rates. net CH]202Hg production after nearly 2 months. and the in situconcentration of CH]Hg in the enclosureliall provided the same informationabout the relative degree ofmethylation among sites. Measurement ofshort~termgross methylation, hom an exogenous Hg spike, appears to be a good predictorof the relative steady state CH:oHg concentration across sites within a specifICecosystem~ i.e. a good relative measure of the propensity for methylation at eachsite.

The relationship between new CH")Hg production, as measured by shortterm incubation, and in situ CH3Hg concentration remains strong within a, singleecosystems over time. In the Experimental Lakes ~a, Ontario. Canada {ELMwetlands, we have observed a persistent relationship between new productionand in situ concenttation of CHjllg, However, the slope of the line changedseasonally. This was not a purely temperature dependent respOIl$e, as CH,Hgproduction and concentration peaked in the fall and not in the beightofsummer

27.

(,'(SH)~H+ ,S6H palapow)

1/6U 'BH IB4nsN0 "' q "' 0oj - - ci ci

-<5~

'"tIJI-....tIJI-

'"-~ ;1j..,0:: 0 . !Ecr:~:::1(J)~CJ) filH, N

..,:>

-u.

'"zw0 0 0 0 0 0 0 0

'" 0 '" 0 '" 0 '".... '" N 0 .... '" N- - -P::>::>/SSIOWU

aJe; uCl!J::>npa; aJe.lns

0 0 0~

0 00 <Xl '" N-

JllJri 'sPillns

Figure3. Measured sulfate redu{:tum tate, porewater sulfide concentration,percent methylmercury (%CH;;lg), mercury methylation rate arrd modeled

po-rewater HgS" in the upper 4 cm ofFlorida Everglades secJjmentsat 8 ACMEsites. Everglades sites are arrangedfrom left to right by average surface watersulfate concentration (highest cQncenrrations on the left). With the exception .cf

the WCA. 1 site, this represents a north tQsouth transect. numill8from theEverglades Nutrient Rel1l()val Project (ENR) and Water Conservation Area 2A

(F1, 03) in tht/.north,thrQurh Water ConsetvatiOnAreas2B (28S) andJA(3A15), and f6 Taylor Siough in Everglades Natio1t4l Park (TS7! TS9) in thesouth. Dazashown are averagesfro171 thne yearr (l!J95.1P98) ofhi- to'tri-

annualsampling. Metluids are described in ref. 37.

Continued on nextpage

~

278

,., 0.12

'"~ 0.10g" 0.08gJ!l 0,06f!<::

0.040

td>. 0.02£::Q;;;:.

0.00

Measured sulfide concentration

~ ~~

.,.~ ~ "'~ ~~.s, ..,. 'll" " 'I- "' ,'I-

~79

Treatment

Figure 4. Methylmercury (CH]Hg) production in FlOrida Everglades sedimenJi;ores after addition ofeither sulfate (light grey bari) or sUlfide (white bars).

Sediment cores. takenfrom the central area a/the Loxahatchee NaiionalWildlife Refuge (LNWR). were amended with either ;fodium sulfide or sodium

sulfate (at neutral pH), by injection into tlie llJP 4 em ofsediment. Thecalculated concentration oftlie spikes iIi p'qre water after injection is shown on

Ihe bottom axis, basedon measured porosity. After1 hT ofincuba!ion withsulfate or sulfide. mercury mtthylation Was estimated in the cores «sing high _

specific activity lOJHg (11J, 37). Methylation aSsays .were conducied over2 krataJnbient temperature. The fi1IlU measured C()ncentration ofsulfide in sedimentporewaters, three hours afterJke sulfate':or sulfide splices. is shown on 'the topaxis. All measurements were n1J1f/e from triplic.ate cores forea'chtreatment.

Sedimentsspiked with sul.fide sequestered much ofit into tke stJlidphase.Sediments spilced with ~ifate prfjduced measurt:iblepiJrewanr sulfide, via

sulfate reduction, within J hQurs, The lNWR is a Vet)' low-sulfate area withinthe Everglades. In these cores, additionalsulfate stimulatedmethylation event~ugh $ulfirJe levels increased slightly, while the addition ojas little as 5 JIM

sulfldealone inhibited methylation.

280

3

• F1 2 Hours

• 3A15>: .. 2BS

"C.g: •c: 2

1 • • •..:J

~Q. .. •'" •J: 1 .. •~ .. ..~

J: ... ..0 ..•• ..

I, .. .."0

0 1 2 3 4 5

CH,Hg, nglgclw

T U3 51 Days• !lA1' •• F1

1 .. 2BS •>: .~"C

~ •c:

'"~~. .

J: 0.1il~ .. ~

J: • ~ ..0 • ~

• ..• ..

0,01 ••

0.1 1 10

CH,Hg, nglgclw

281

Figure 5. Relationship between in situ methylmercuf)i (CHiilg) conctmtrationand methybnetcury productionftom stable isotope spikes in Everglades surface

sediments: a) Cll/99Hgproduclion!rom 199Hg injected into Evergladessedime:nJ: cores after anincubati6n period 0/2 hrs. YS. in situ CHJfig~' b)

CH/())Hg in enclosuresurjace sediments 51 days after the enclosures wereamended with :NJ2lig Y$. ih.situ native CHJig. These measurements Were madein experimentah. Un dJametenmclosures at four sites across the Everglades, aspqrt afthe ACME project Three enclosures at each site were spiked with 102Hgin. May 2000. with spikeS ranging from halfto two times the equivalent of 1 year

ofatrrJ.ospheric deposition. (see textfor detailS; from Krabbenhoft et ai..unpublished data).

(Figure 6). While we do not know the reason for the seasonal trend. both themeasured methylation and in situ concentration weteresponding in concert tothe same controlling factors.

Overall, the rate of p.roduetiOil of Ca3Hg from. e~gelJ,OUS spikes isg~rally higher than the rate of methylation from in situ Hg pools. Thisprobably reflects the higher bioavailabiUlY of the added Hg ~m.pa,red to that inthe sedim~nt(106). This is illustrated in Figure 7. which shbws sediment GH.lHgleveb in a lake enclosure experiment conducted as part of the METAALICUSproject in the ELl\.. The top panel shows in sini CH]!lg as a percentage of tOtalHg in the sediments in each of the four enclosures throughout the summer. TheeriCl'osures were spiked with ZOOHg either in mid June (imcl.l&:2) o. biweeklythroughout the summer (encL3 & 4). The middle panel shows CH.l~g aspercentage of '2OOHgaccumulated insedimems. The relative percentage of theadded Hg that becomes methylated is' initially much higher for the added Hg.Over time, the percentage decreases and this Is likely a combination of bothreduction in bacterial activity in the fall and a decrease in bioavailability of .theHg oyer time··as itis cycled through the system (methylated. demethylaied andcomplexed to strong binding ligands). With, the caveat that the method of Hgaddition in our spike experiments may not trury reflect reali,ty. these data suggestthat Hg newly deposited to eco$ystems is more available for methylation thanexisting Hg pools. However. more work is needed to further ascertain the crucial'question of the relative importance of newly added versus in situ Hg incontributing tathe Hg £bat is, methylated lind bioaceumulated in aquatic systems.

Thus, W& suggest that 'the in situ CH]Hg concentration across a series ofsites within anecosyste.m can be used to predict which site is likely to be moreactive in terms of methylation, and likely in tenns of bioaccumlilation, all elsebeing equal. However, there is too little information at present to determino thedegree to which these relationships can be used in a quantitatively predictivefashion between ecosystems. Overall, in comparing across systems, the greatestdifficulty is in assessing the pool of Hg available for methylation, which iscrucial to estimating realistic accurate methylation rates. To this point, we havenot been able to measure bioavaHable pools of Hg to bacteria. nor have we beenable to mimic the SpeCiation of in situ Hg: with added Hg. Therefore. short termproduction rertlains more qualitative than quantitative. Both of these questionsare the focus ofongoing .esearcb.

Biological Controls over Methylation

Different organisms clearly have different raiesof Hg methylation, evenamong the SRB, and not all SRB methylate Hg (24,32). A small number of Fe­reducing bacteria. that are phylogenetically similar to methylating SRB, have

283

'""''"

50~ 0 AUiust 18°C

~OIl

40 • SeptuOC •.5 D. June SOC

.a",. 30

".-""" 20.§[ r1 =O.77

10 -'OIl 0_.----·-· 0::t:

" .~ ,- .r=O.42.

0 00 200 400 600 800 1000

·1Ambient CH,Hg (pg g w.w.)

Fii,p'e 6.Native in situ methylmercury (CH:Jlg) concentration and excessCHJ 99.Hg prodUcedfrom J,9Hg in 4 hra, in peat collected in June. August and

September, 2000, from a lakeside, sphagnum wetland (LlJ5J at theExperimental Lakes Area (EUlin northwest Ontario. WorJcwq:s conducted as

part ofthe MErAAUCUSprojecl.

Native Hg ___ Enclosure 1-0--- Enclosure 2--- Enclosure 3--0- Enclosure 4

I'\, ~

~

'--..--r-o06/19 07/03 07/17 07131 08/14 08/28 09/11

10

2

Figur~ 7. Sediment CH)Hg levels in a lake enclosure experiment conducted in£329 in2D!JO as part ofthe' METAAllCUS ptojectinthe Experimental Lakes

Area (ELA). Ontario. Ctlhada. The surface water ofenclosures was spiked withUlOHg eirherin mid·June (encl. 1& 2) arbiweekly throughout the sumnr.er{em!

3&4). Spikes wereequivaJiJitt to I year ofatmospheru: deposition. The t(}ppanelshoW$ native in situ CH]Hg as a percentage a/total Hg in 0-4 em depth

sediments in ell.ok offour enclosures throughout the summer. The middle panelshows eM]200Hg as percentage a/WlHg accumulated in sedj·ments. The bottDmpfJriel shows W41ertemperafure in rheenclosure. A much hig/terpercenrage illthe spike wasfofiJld methylated insediTru!nts than 1Ul1ive Mg. e.specialty within 2

tn()nths oflhe spike.

285

40 200 Hg

o

06/19 07/03 07/17 07/31 08114 '08/28 09111

i~~~,106119 07103 07/17 07/31 08114 08128 09111

figure 7, .continued.,

286

been shown capable of methylating Hg in pure culture (107). While a num~r ofOrganisms other than SRB have been shown to produce CH)Hg in pure culturefrom added Hg(Il) (see ref. 24 fora review) the relative rates of methylation bythese organisms and their role in in sit(L methylation is unknown. FurthermOre,whiles number of SRB that are incomplete organic carbon ox.idizers readilymethylate Hg in culture. (e,g. Des14ljobulbu$ propionic/lS.. (30,86JandDesuljovibrlo desuJfuricans, (lO8)), studies in the environment hav~ suggestedthat the-y may not be the dominant Hg methylators. King et a1. (3,2) showed inpure culture that. of the organisms they tested. Dit8UlfiJba<:;terium methylated Hg-at a substantially greater rate, under the conditions of their experiments, than theother species tested (I:Jesulfobaeter. Desulfococcus. Desu/fovibrioandDe.rulfobulbus). DesK/fobactetiurn is a Complete acetate oxidiier and CH)Hgwas only produced in t!tese cultutes when sulfa1e~reduction wali occurrin~. Thiscontrasts results ofoth~ Who have,shQwn that Desulfobulbus propionicus canmethylate Hgwbilegrowing fennentativeJy (38.86).

King et a1 (32) alsq found that marine sediments amended with acetateproduced mote CH,Hg than sedimenfS amended with lactate; or unamendedcOhtrols. Aqetate;.arncnded slurries were dominated by Desulfobacterium andDesulfobatter. Macalady et aI. (33). using polar lipid fatty acid analysis, alsofound that Desulfobacter·Jike organisms were important Hg methylatorsinliediments of a Hg~CoTltaminated freshwater system., Clear Lake. CaJifornia. Itappears from these results that the organisms capable of complete oxidation ofat;et8tC ar¢ potentially more efficient methylatcirs in the environment.

HoWever, it is clear tbatthere is some aspect of the meChanism of Hgmethylation that allows some bacteria tci methylateHg while others do not. Theability to methylate Hg is riot confined to one pbylogenetic group of sulfate­reducing bacteria but is scattered throughout the phylogenie tree of sulfate·reducing eubacteria (41). Furthermore. phylogenetically similar organisms havedifferingabiHties to methylate Hg -e.g., Delsulfovibrio gigas. D. vulgaris. D.salexigens atld D. thsldjuricans (l(Jstiiariido not methylate Hg but D.desuift.oicans LS (lQ9) and NDl32 (24) do. A pathway for methylation hasbeen demonstrated for only one organism (Desulfovibrio desulfuricans LS).Berman et ai. (J09)sh6Wtd that mercury methylation is an enzymaticallyciitalyzed process in vivo. and suggested, based on the selective inhibition ofmercury methylation in D. desul{uricans LS. thai methylation (s mediated by acobalt porphyrin in this Otganism~ Further work (71,79) led 'the group to proposethat Hg methylation in this organism occurs via transfer of a methyl group frommethyl·tetrahydrofoli;lte toC:obalamin to Hg. The methyl group may originatefrom senncor via the acetyl,CoA synthase pathway.

Mercury methylation by cell elttfacts of D. desulfUricans LS was 6QO..foldhigher compared to free methylcobalamin (73). and thus. it is not merely th-epresence ofcobalanUn that instills the ability to methylate Hg at a significant

287

rate. Indeed, cobalamin is not unique to SRB. Cobalamin and high levels ofacetyl·CoA enzymeS are present in methanogens and acetogens, and indeed. cellexttactsof a metbanogen, have been shown to methylate Hg (llOl. However,these organisms are not thought to play a large role in environmentalmethylation, based on selective inhibitor studies (13,19,22,37). Furthermore, it isnot known whether the corrinoid protein found in strain LS is always present inthe SRB that do methylate Mg. It has beens:uggested that SRBs that methylateHg possess a distinct or highly specific enzyme to catalyze this step. However,the identity of the enzyme responsible for methyl transfer to Hg in mostmethylatots is not known.

Recent studies have shown that of the complete aceiate oxidizers.Di!,rulfocQCCUS multivorons {tbeIl, Desulfosarcina variabilis (3be13) andDesulfobacterlumautotrophicaiu 'all contain the acetyl-CoA pathway andmethylate Hg while Desulfobacler hydrogenophilus does not methylate Hg butdoes bavt the pathway. Similarly, for SRBs that are not complete oxidizers,there is a correspondence between the presence of the acetyl-eoA pathway andthe ability to methylate Hg for Desulfovibrio africanus, D. $ulfurica1l$ LS and D.vulgaris ,(Marburg). However, there are also organisms that methylate Hg thatdo not have this pathway (D~sulfobulbUJprQpionicus (lpr3) and D. propionicus(MUD». Thus the acetyl-eoA pathway cannot be the only mechanism for Hgmethylation in vivo.

Given the reactivity of Hg, it is obvious that the Hg will not be presentinside, cells ,as the free metal ion, Hgu . 'Thus, the transfer of the methyl group toHg likely involves t~ interaction with Hg bound to a ligand. or to an enzyme::,There,area number of mechanisms for methyl transfer within cells, but as the Hgis likely in the +2 state within the bacteria, the methyl group needs to betransferred as a radical or asa carbanion, and this restrit;:ts' the methyl transferprocess to that involving electrophilic attack by Hg(II)on cobalamin (l10). IfHg was in the +!state. then it could directly substitute for Ni(I) in the normaloperation of the aeetyl·CoA pathway i.e., be involved in a nucleophilic attack onthe corrinOid methyl group. This is an intriguing but untested notion.Alternatively, it may be that the Hg is bound to a particular enzyme or thiolgroup in some organisms that places it in the correct location for transfer, or thatsteric hindrance prevents the transfer ofHg in some organisms but not in others.In theacetyl-CoA pathwar.the methyl group is transferred to carbon monoxidedehydrogenase (CODH) and if Hg were bound to the active site of tlle CODH,the transfer of the methyl group directly to the Hgcould occur. These ideas arespeculative, and further studies should focus on identifying the location of Hgwithin the cell during methylation.

288

Bacterial Demethylalion

Microbial de'~radation via the mer operon is the best·stiJdied pathway ofCH)Hg degradation. The operon is widely distributed in nature, ofttJico-existingon transposons that also contain antibiotic resistance genes (111). Varian~on

the mer operon that include the 11U1rB genecoofer "broad spectrum" resistance toa variety of organomercury compounds including methyl- and ethylmercurychloride via organomercurial lyase (2],112).Microbial degradation ofmethylmercury occurs, through the cleavage of thecarbon-metcury bond by theenzyme organomelCllrial IYiUle foDowed by reduction of Hg(Il) by mercuricreductase to yield methane and Hi (21). The physiology and' genetics of m~r·mediated CH:Jlg dCgraQation and merCury resistance have been extensivelyreviewed elsewhen<. (69.111-115). While· the· biochemistry of the cOJIlIIlonlystudied mer operon is fairly well understood, newer studies of the distribution ofthe niet 'operon in the elivirolimentare revealing unexpected polymorphism andgencticdivtlrsity (J16·121).

A$ many as half of the bacteria from Hg:-contaminatedsitcs m~ contain themer oPeron (111). However. another mechanism alSo appears to mediate CH,Hgdegradation. While methane and Htf are the primary products of mer·mediatedHg demethylation, CD,; has also been observed as a major methylmercurydemethylatlon product by Oremlaild and co-workers (64,70,71). These authorssuggested that methylrilerCury degradation can occur through biochemicalpathways used to derive energy from single carbon substrates. and they termedthis process "o~k1ativedemethylation", i.e.:

As a presumptive Cl metabolic pathway. oxidative demethylationis not anactive deto:tification pathway for CH3Hg, unlike mer-mediated dernethylationand Hg reduction. A variety of aerobes and anaerobes (including sulfatereducers and methanogens) have been implicated in carrying out oxidative~methylation, and oxdative demethylation has been observed in freshwater,estuarine and alkaline-hypersaline sediments (64.65.70). However, the identityof the organisms responsible fOf oxidative demethylation in the environmentremains poorly understood, Further, no organism has been isolated that carriesout this pathway.

Pak & Bartha (66) conftitned the ability of two sulfate reducing bacterialstrains and one methanogen strain to demethylate mercury in pure culture. Theyargued that the C02 scen in these studies resulted from oxidation of methanereleased from CH3Hg after cleavage via organomeTcurial lyase !::Iy anaerobicmethanotrophs in the sediments and that CO2 was a secondary product and nottheprinUiry product of dernethylation. However, Marvin-DiPasquale et al. (71)

289

found that the rate of~ production frtim CH;!Hg far exceeded the rate of COzprorlllCtion from CH( in sediments from two of their study ecosystems. underboth aerobic and aiiaerobic conditions.

The, relative importance of mer~modiatedversus oxidative demetbylatioll: ispoorly un4e:rstood (7J). In highly contaminated environments, the mer operon ismore prevalent among the microbial community. and Hg(U) reduction activity isenhanced (1 J1). However, the fate (J,fmicrobial EgO production in theenvironment may not always be proportional to mer transcription (123), Overallrates of microbial activity, the presence of Hg-reducing genes divergent fromcommonly used probes. and the bioavailability of CH)!lg to cells also play awle, In systems that are not highly ct:lntaminated. o'tid:litive demethylatiQoappears tQ dominate, under both aerobic and anaerobic <:dnd~tions. 'the Hgconcenttations. that wo\lld cause a switch from one pathway to the other are olllyloosely defined. Most studies of Hg demethylation via oxidative demethylationhave employed 14C~labe:led CH3Hg and thus only the carbon prodUctS aretrateable. The end-prodllCt of oxidative demethylation haS been presumed to beHg(U), but that b~ not been -confirmed. Demethyliltion studies USing CH3HgCOritaining a specific Stable Hgisotope should help resolve tlIat issue.

Bioluminescent Sensors for Mercury

"Bioreportefs" are genetically engineered microorganisms designed torapidly ass;eQ the biaavailable con<;entrati.Qnof contaminants. or the rate afcontaminant degradation. In these biareportel'S, the bioluminesce:i'tce operon(lux) js inserted as tbesensor component into the biodeglMlition Or resistancepathways of interest. When the pathway 4 expressed. the llU genes are expressede\Jncurrently. A relativeJysimple measurement of Ught production can then beused to as~, for '~Xample, expression ofa mc:ta1 re$is~ce gcme or ahydro.carbon degradation pathway. Clearly, -~ potential advantage ofbioreporters over chemical·measurement lies in the poSsibjlity Of determiningbioavaiiable or bioactive concentrations of the contliIilinant 'of interest.Selifonovaet aI. (124) constructed the -first Hg bioreporter1'. fusing theprolllOtorless lu.x ~run "from Vibrio jistheti into the TIl2I mercury resiStaDt;:eoperon (mer). and using E. coli as the hQSt strain, The orgllni$nl/i sho\Yed semi­quantitative response tOo Hg in contaminated natural waters•.at concentratioi'ls aslow as a few nM, In constructing these first Hg bioreporters. the importance ofunder.'ltanding Hg tr~sport pathways was recognized asb:eing crucial ifbioreporters were to be used to assess Hg bioavailability. The mer 'operonconsists Qf a sequence of genes that encode active Hg transport and Hg reduction(meTA), plus regulatory genc;s(merR and merD). SeljfQuova-et al. oonstrtJcted a'set of threerner-lux fusions with and without ttJ:etranspdrt and reductaSe gen.es

290

(124). Interestingly, light production in response to Hg by strains with andwithout the transPQrt genes was similar. suggesting Hg uptake was occurring bypathways other than mer-based acJive Hg transp'ott However,complic}ningfactors, snchas the potential energy and coutlte.r~ion requirements of Hgtranspottmay cloud interp~tation of these data.

Continued development of this bimeporterderhonstrated that Hg~dependantlight production in this strain was cell density~dependant (125) and dependant onthechetnistryof tbeassay medium (126); The strain used in these studiescontained a mer-luX fusion without Hg transport genes (pRB28). Reduction ofcell density to aboUt las cells per ml (at the high end of the range of celldensities found in n8.tutaI waters) in Hg assays reduced the number ofcompetitive binding sites for Hg, and therefore improved the sensitivity of theassay. into the, pM tange; Dissolved organiq, carbon alsQ decreased thebioavailabilityof Hg to this strain. TOO assay buffer was manipulated to showthat neutnil Hg-CI complexes induced more light production than negativelychargedcomplex:es, suggesting that uptake by this strain, under these conditions,was via diffusion.

Because light production is energy-dependent in these biosensor!, it isnecessary to separate factors that influence cellular activity in general fromfactors that influence Hg bioavailability specifically (127,128). Barkay et al.have used constitutive controls to achieve this goal. constructing an isogenicstrain (pRB27) in which lux expression is constitutive,and therefore Hg­independent (126); Another approach is to construct biose.nsors with only apartial lux operon, so that the aldehyde precursors to lux-mediated lightproduction are not produced by cells. but are supplied ex:ternally (129). Thisreduces the energy requirements of light production. but requires additionalalteration of test media. and potentially affects Hg speciation. In order toexamine Hg bioavailability under a wider range of conditions, KeHyet a1. andScott et at transferred the pRB27 and pR1J28 mer.lra fusions of Barkay et al.into Vibrioanguil1lJrum (130,131). This host strain bas wide salinity tolerance,and is a facultative anaerobe. Refinement of assay conditions also improvedsensitivity to <O.S proal bioavailable Hg L-1

• This strain has been used toexamine the bioavailability of trace level additions of Hg(1D to natural lakewaters; and to examine the bioavailability of Hg in unamended natUTal waters.The percen4tge of ambient total I::Ig in lake and rain water available W!l$ found tobe very low, as was the bioavailabilily of tracer additions of Hg to naturalwaters. Finally, as a first step in understanding Hg bioavailability in the,conditions in whiCh methylation occurs. Golding et al. (132) have wori<ed withE. coli and Vibrio emguil~rum bioreporters und.~ anaerobic conditions. Insome circumstances, Hg(II) uptake by both strains appears to occur via

291

facilitated transport. 'Tbili suggests that Hg uptake by these strains octurs byadifferent meChanism than Hg uptake by methylating SRB, which o.ccurs viadiffusion of neutral5pecies (30.86). Differences in medium content eQuidpotentially account for these differences as. for example. facilitared trllt'isport ofHg bound to amino acids has been shown to occur acrOSS membranes of higherorganisms (13:3).

Genetic engineeriltg u~ng the mer operon has also been applied to Hgbioremediatitm. The advantage of strains 'constructed with the mer operon lies inthe specifitity of the mer-based Uptake and detoxification systems, which areoften unaffected by tile presei:lce of Other metals. For example. a mercJJrybioaceumuIator has been engineered (134), as a potential aid in mercurybiaremedia:tion. Toproduee the bioaccumulator.an E. con strain wasconstructed to e:itpressa: mer-based Hg2+ transport syStem and to overexpress peametallothionine (MT), which protects the cells from Hg tIDticityand aIlows forcontinued accilmu1ation of Hg-MT within cells, Accumulation of Hg by theStrain was not affected by metal chelators such asBDTA and citrate, Organismsthat overexpress organomercurial lyase have also been constructed as potentialaids ill clean-up of organomercury contaminated sites (e~g. 135).

Mercury biosenSOl'Sare a potentially valuable tool for assessing figbioavailability. 'ro date they have been used to demonstrate that Hgcomplexation has a large influence on Hg, bioavailability to thC$e cells., and thatonly a ,small fraction of Hg dissolved in natural waters is generally available foruptake, Also, previously unidentified pathways for Hg transport may need to beconsidered. More studies $lIould lead tQ a fuller understanding of Hgtransp-ortpathways in cells without the mer-bastxl transport system. and allow comparisonof those systemS with the transport systems of methylating and demethylalingmicroorganisms. and the broader spectrum of microorganisms at the bottom ofthe food web, An important issue in understanding the results of Hg biosensorstudies is the role of Hg transport patllways coded by the mer operon and thoseof the host organism. The bioavailability of Hg to methylating organisms isperhaps the key fo mOdeling Hg methylation ral:es~ Bioreporters can potentiallybe used to define that fraction of the ambient Hg pool if it can be shown that Hguptake by rn¢thylatOrs andbioreporters are similar. This should be the focus ofconu[lued research.. However, CH3Flg production itself may be the best"bioreporler" of Hg bioavailability to methylating bacteria. Since CH,1lIgproduction by these microorganisms occurs intracellularly. CH)llg productiondepends on Hg transport .and l1erves as l\. ~OT for Hg biQfI,vai,lllbility. We haveused D. propidnicus in this way to examine uptake of neutral HgS complexes(30,86).

292

References

1. Pacyna. 1M. [n Global and Regional Mercury Cycles: Sources, F~esalid Mass Balances; W. Baeyens~ R. Ebinghaus; O. Vasiliev, Eds.;Kluwer Acadenrlc Publishers, Dordrecht, 1996, pp. 161-178.

2. Mason, R.P.: Fitzgerald, W.F.; Morel, F.M.M. Geochim. Cosmochim.Acta 1994, 58, 3191·3198.

3. Fitzgerald. W.P.; Engstrom; n.R.; Mason, R.P.; Nater, E.A. Enyiron.Sci. TechiloJ. 1998,32,1-7.

4. Lindqvist, 0.; Johannson. K.; AaStrup, M.; Andersson, A.; Bringmark,L; Hovsenius, G.; Hakanson, L.; Melli. M.; Timm, B. Water Air SoilPollut;,1991. Special Issue, Vol. 55.

5. U.S. EPA. Mercury Study Report to Congress. EPA-452JR-97-004, USEPA Office of Air, 1997, Washington,OC.

6; Clarkson, T.W. J. Trace Element Exp. Med. 19.98,11,303·317.7. Mahaffey,K.R. lAMA 1998,280,737·738.8. St. Louis, V.L.; Rudel, I.M.W.; Ken)'. c.A.; :Beaty. KG.; Flett. R.J.;

Roulet,N.T. Environ. Sci. Technol. 1996, 30.2719'::2729.9. Krabbenhoft, D.P.• Benoit. J.M.• Babiaf2, C.L., fturley, J.P., Andren,

A.W. War. Air-Soil Poll. 1995.80.425-433.10. St. Louis:, V.L.; Rudd; J.M"W.; Kelly, C.A.; Beaty, K.G,; Bloom, N.S.;

Flett, R.J. Can. J. Ffsh. Aquat. Sci. 1994,51,1065·1076.1L Driscoll,. C.T.,. Holsapple. J., Schofield, C.L.. Munson. R. Biogeoche.m.,

199840,137-146.12. Waldron, M.e.. Coleman. J.A., Breault.. R.F. Can. J. Fish Aquat. Sci.,

2000,57, 1080-1091.13. Compeau, G.; :Bartha, R. AppL Environ. MicrobioL 1985, 50, 498-$02.14. Berman. M.; Bartha. R. BulL Environ. CQntam, Toxicol. 1986. 36,401 ~

404.15. Ramlal. p.s.; Kelly, CA.; Rood. J.W.M.;Furutani. A Can. J. Fish.

Aquar. Sci. 1992,50,972-979.16. Korthals. B.T.; Winfrey, M.R. AppL En....iron. MicrobioL 1987. 53,

2397-2404.17. JenSl;n, S.; Jernelov, A. Natwe. 1969• .223,753-754.18,GiJmourj C.C;:.; ~iedel; G.S. WQt¢'r Air $oil Pollw-t.1995, 80, 747·756.19. Gilmour. e.e.; Henry, E.A.; Miu::hell, R. Envir¢n. Sti. Technol. 1992,

,6,2281-2287.20. Branfire:un. B.A,; Roulet. N.T.; Kelly, e.A.; Rudd, J.W.M. Globp,l

Biogeochem, Cycles. 1999, 13, 743-750.21. Robinson, J.B.~ Tuavinen. D.H. Microbial. ReV. 1984,48,95-124.22, Chen', y,; Bonibngo, J.-C,J,; Lyons, W, E,; Miller, O,C. Enviton:

Toxicol. Chem. 1997,16.1568-1574.

293

23. King, J.K; Saunders,F.M,; Lee, R.F.~ Jahnke. RA Environ. TaxieoLChem.I999, 18, 13~H3~9.

24, Gilmour, C.C.; HeJuy, B.A. EnvirOn. Poll. 1991,7J, 131-169.25. Watras, C.l. and 21 others. In Mercury PoUution: Intergration and

Synthesis., CJ. Watras and l.W. Huckabee, Eds" Lewis Publishers,Boca Raton, 1994. pp. 153~177.

16. Orban, N.R., Brezonik, P.L., Baker, LA, Sherman, LA. UmnoLOceanogr;, 1994,19, 797-815.

21. Steffan, R.I.; KQrthals; B.T.; Winfrey, M.R. Appl. Environ. MicrobiaL1988,54,2003-2009.

28. Compeau, G.; Bartha. R. Appl. Environ. MiCrobiol. 1987, 53,261·2~5.

29. compeau. G.; Bartf\a, R. BulL Environ. ConJam, Taxieol;, 1983, 31,486493.

39. Benoit, I.M.; MilSon. R.P.; GilmoilI. c.e. AppL Environ. Microbial.2001,67. 51-58.

31. Devereaux. R; Winfrey. M.R; Winfrey, J.; Stahl, D.A. FEMSMicrobial. Eeal. 19!16, 20, 23-31,

32. King. J.K.; Kostka, J.E.; Frisch~, M,E. Appl. Environ, Microbfol.2000. 66(6), 2430-2437.

33. Macalady, J.L. Mack, B.B., Nelson, D.C., Scow. K..M. .4.PP' Environ.MicrobioL,2000, 66, 1419~1488.

34, Kdtska, unpubl35_ Winfrey, M.R; Rood, J.W.M. Environ. ToxicoL Chern. 1990. 9. 853­

869.36. Benoit, 1.M.; Gilrttour. C.C.; Mison, R;P.; Reidel, O.S.; Reidel. G.F.

Biogeochem. 1998, 40,249-265.31. Gilmour, e.C.; Riedel, G.S.; Ederington, M.e.; Bell. J.T;; Benoit.

I.M.; Gill, G.A.; Stordal. M.e. Biogeachem. 1998~ 40,327-345.3R. Helliy, B.A. Ph.D. Thesis, Harvard University, 1992.39. Allan, C.J.; Heyes, A Water Airl1.ndSoil Pollut. 1998. 10~, 513-592.40. Allan. C_J.; Heyes, A.: Roule!, ~T.T.; St. Louis, V.L.; RUOd, J.W,M,

Biogeochem. 2000;, 52, 13-40.41. Heyes, A,; Gilmour, e.c.; Mason. R.P. Unpublished data.42. Heyes. A.; Moore, T~R.; Rudd, J.W.M.; Dugoua. J.I. Can J. Fish.

Aquat. Sci. 2000-. 57, 2211-2122.43. Mason, R.P.; Lawson, N.M; Lawrence, A.L.:Leaner. J.J.; Lee. J.G.;

Sheu, G-R. Mar. Chem, 1999,65. 77-96.44. Mason, R.P.; Lawrence A.L. Env-iron. Toxkol. Chern., 1999. 18; 2438­

2447.

294

45. Kannan, It; Smith, R.I. Jt.. ; Lee. RE; Windom, 1:l.L.; Heitmuller, P.T.;Macauley, I.M.; Sununers, 1.K. 1998. Environ. Contam. ToxiCol.. 1998;34,109-118_

46. Bloom, N.S.; Oill, G.A.; Driscoll, c.; Rood, J.; Mason, R.p~ Environ.Sci. TechooL 1999,33,7-13.

47. Hines. M.E.; HorVat, M,; Fsganeli, J.;Bonzongo. J.e.I.: Batkay.T.;Major. B.B.; Scott. K.J.;~ Baily, E.A;; Waewick, JJ.; Lyons, W.B.EnllirQ't kesearch, 2000, 83,129-139.

48. Kannan, K; Fatandysz. Water Air Soil Pollut. 1998, 103, 129-136.49. Henry, E.A.; Dodge.-Murpily, L.J.; Bighma, a.M.; Klein, S.M.;

Gilmour, C.C. WQltr Air Soil Poll. 1995.80, 489·498.5-0. Watras. CoJ.: Back. R.c.; Ha!vOf$ei\. S.: Hudson. R.J.M.; Mom-son,

K.A.; Wente, S.P. Sci. Tot. Environ. 1998.219, 183-208.51. Suchanek, T.H.; Mullen, L.R.; Lamphere, B.A.; Riclrerson. PJ.;

Woodmansee. C.B.; Siotten. D,G.; Hamer R.I.; Woodward. L.A. WaterAir Soil p()Uur. 1998. 104.77-102.

52. Vetta,M.; Matilaine:n, T. Wtiler,AirSi)ilPo.ll. 1995,80.585-588.53. Hurley, J.P.; CoWell, S.B.; Sbafer, M.M.; l:Inghes. P.E. Environ. Sci~

Technol. 1998.32. 1424-1432.54. BonzQngo~ J.~C.; Heirn., K.J.; Cben, Y.; Lyons. W.B.; Warwick. 1.1.;

:Miller. G.C.; Lech1er, PJ. Environ. ToxJcol. Cheln. 1996. 15, 677~ .683.

55. Hines. unpublished56. Hintelman H;; Wilken. R.D, Yom Wasser 1994, 82. 163-l73.57. Miskimmin, 8.M.; Rudd.. l.W.M.; Kelly, C.A; Can. J. Fish. Aquat. Sci.

1992,49.17-24.58. Miskimmin. B.M. BuU. Environ. Cont!lffL. ToX«:ol. 1991,47,743-750.59. Guimatas, LItO.; Meili. M.; Hylil.nder, L.D.; Silva, E.O.E.: Roulet!

M.; Mauro. J.B;N.; de Lemos, RA. Sci. Total Environ. 2000. 261,99-107.

60. Roulet, M.; Guimaraes. I.R.D.; Lucotte, M. Warer Air and SoilPollution, 2001,128, 41-60.

61. Bodaly,R.A.• Rudd, J.M.W., Fudge, R.J~P., Kelly. c.A. Can, J. Fish., AqUilt. Sci. 1993. 50, 980~9g7.

62. RamlaI, P.S.;Kelly, C.A.;Rudd, J.W.M.; Furutani, A.Can. J. Fish.At]uat: Sci. 19'3,50,972-979.

63. Barkay, T., Liebert, C.• Gillman. M. Appl Environ. Microbial. 1989,196-1202.

64. Oremland. It S.; Culbertson, C.W.; Winfrey, M.R. Appl. Environ.Microbiol.l991.57,130-137.

65. Oremland, R$.; Miller. L.G.; Dowdle, P.; Cannel, T.; Barkay, T. Appl.Environ: MicrobiQl. 1995, tj[, 174$c2753.

295

66. Pak, J(.·R,; Bartha. R. Bull. Environ. Contam. Toxicol.1998, 61, 690~694.

67. Pak, K..R.; Bartha. R Appl. Environ. Microbial. 1998;64, 1013­1017.

68. Spangler, W.J,; Spe.garelli, ],L.; Rose, 1,M.; Miller. H.M. ScieTice.1973. 180. 192-193.

69. Moore. M.J~;_ Distefano, M.D.; Zydowsky, L.D.; Cummings, R.T.;Walsh, C.T. Ace. Chern. Res. 1990; 23; 301-308.

70. Marvjn~Oipasqua1e. M.e.; OremliUld, R,S. En.viron. Sd. Techno!.1998. 32. 2551;-2563.

71. Marvin-DiPasquale. M.; Agee. J.; McGowan, c.; Oremland, R.S.;Thomas. M.; Krabbenhoft, D.; Gilmour. C.C. EnvirOh. Sci. Technol.,2000,34,4908-4916.

72. Sellers. P.; Kelly. C.A.; Rudd, J.W.M.; MacHutehon, AR. Nature.1996. 380,694-697.

73. Choi, S.-C.; Chase; Jr" T; Bartha. R. Appl. Environ. Microbial. 1994,60,4072-4077.

74. Benoit. I.M.; Gilmour, C.C; Mason, R.P. En.viron. SCi. Techno!.1999, 33, 951-957.

75. Gi.Jtkriecht, J.l. J. Membr. Bioi. 1981,61,61-66.76. Mason, R.P.; Reinfelder, l.R.; Morel, F.M.M. Water Air Soil PoUut.

1995, 80, 915-921.71. Mason, k,P,; Reinfelder, l.R; Morel, F.M.M. Environ. Sci, Technol.

1996,30, 1835..J845. '78. Craig, P,]. Moreton, P.A, Mar. Poll. Bull. 1983, 14, 408-411,79. Choi, S-C,; Bartha. R. Bull. EnvimT1, COnlan,.. Toxicol. 1994, 53, 80:$-

812.80. Sch~arzenbac:h. G.; Widmer, M. Chim, Acta. 19"3, 46, 2613-2628.81. Paquette> K.; Helz, G. Water Air Soil Pollut. 1995,80,1053-1056.82. Paquette. K. 1994. Ph.D. Thesis. The University of Maryland, 1,994.8j, Oyrssen, D. Mar. Chern. 1988,24,143-J53.84. Dyrssen. D; Wedborg, M. Water Air Soil Pollut~ 1991, 56, 50'7-52{)'85. Benoit. J.M.; Mason, R.P.; Gilmour. C.c.Environ. To.x;icoL Chern.

1999,18, 213g~2141.86. Benoit, I.M,; Gilmour, C.c.; Mason, R,.P. Environ. SCi. Technol.

2OOl,35, 127-132.87. Jay, J.A.; Morel, F.M.M,; Hemond, H.F. Environ. Sci.. Technol. 2f!'tlO,

34, 219~22(}{}.88. TosseU, J.A I. Phys; Chern. A. 2001, 105,935-94.1.89. Stein, w.n.; Lieb, W.R. TrLJ~pottand OiJJution AcroS,S Celt

Membranes, Harcourt Brace JovauQvith, London,1986.

296

90. l-Iuerta-Diaz. M.A.; Morse, iM. Geoc.hiin. Cosmochim.. Acta. 199Z,56.2681-2702.

91. Bono, A.B. M,S. Dissettalion;McOill University, 1997. 177pp.92. Dmytriw. R., Mucci, A., Lucbtte, M., Pichel, P. 1995. Wat. Air Soil

PqllLft, 1995, 81), 1099-1103.9;3. Benoit, J,M.·; Mason, R.P.; Gilmour. c.C.; Aiken, G.R. Geochim.

Cosmochiin. Acta 2001, 65, 4445-4451.94. Reddy, M.M., Aiken, G.R War. AirSQil Poll. 2000,132,89-104.95. Hudson, RJ.M.; Gherini, S.; Watras, C; Porcella, D. In Mercury as a

Global Pollutant: Towards Integration and SyntheSis. C.J. Watras andl.W. Huckabee, Eds.; Lewis PU1::llishers: BOCa Raton, 1994, pp. 473·526.

96. Miller; C.L.; Mason, R.P. ACS 2221lll Meeting, Chicago, August,2001; ACS abstrac~ 41(2). 514-518.

97. Mikac; N.; Kwokal. z.; May•. K..; Branica. M. Mar. Chern. 1989,28,109-126.

98. Muhaya, a.B.M.; Leemakers, M.; Baeyens. W. Water Air Soil Poll.1997.94.109-123.

99. Xia, K.; Skyllberg, 11.L.; BI~m. W.F.; Bloom, P.R.; Nater, B.A;lIelmke, P.A. 1999, 33, 257·261.

100. . Kim, C.S.; Shaw, S.; Rytuba. J.l.; Gordon,B.B. Jr. ACS 222m!Meeting" Chicago, n, AuguSt 26~3o, 2001, 4, 497~503.

101. Spry, D.l.; Wiener, J.O. Environ. Pollut. 1991,71.'243-304.102. RamlaJ, P.S.; Rood, J.W.M.; Furutarri. A.; XUn. L Can. J. Fish. Aqua!.

Sci. 1'85, 42, 685"692.103. Xun, L.; Campbell, N.E.R.; Rood, l.W.M. Can. J. Fish. Aquat Sci.

1987.44. 75~757.104, Royals, H.E.; Lange, T.R. ACS AbstraCt, 1995 Meeting, Washington.

DC, 1995.105. Hintehnann. H.; Evans. RD.; Villeneuve, I,Y. J. Anal. Atomic Spec

1995.10. 619.62A106. Hintelmann, H.; KeppeI~lones, K.; Evans, RD. Environ; ToxicoL

Ch"". 2000. 19.2204-2211.107. Gilmour, C.C.; Riedel. G.S.• Coates; J.S.; LovIey, D. Abstract, Amer.

Soc. MicrQbiaL, New Orleans, 1996; .108. Choi, S.C.; Bartba, R.App. Environ. Microbiol. 1993, 59:. 290-295.109. Berman, M.: Chase, T.; Bartha, R. Appl. Envir-01l. Microbiol. 1990,

56, 298.300.110. Wood, j,M.; Kennedy, F:S.; ROSen, e.G. Nmure. 1968,220, 173·174.Ill. Liebert. e.A. Hall, RoM.. Summers, AO. Microbia. M()l. Bioi. Rev.,

1999,63,507-522.

297

112. Barkay, T. 2000. In Encyclopedia t>f Microbiology. 2nd edition.Academic Press. San Diego,. pp. 171-181.

113, Stlllltners. A.D. Ann. Rev. Miciobiol. 1986,40. 607-634.114. Foster. TJ. eRe Crit. Rev. Microbial. 19:81,15,117-140.U5. Silver. S" Phung. L.T. Ann. Rev. Microbiol.1996. 50, 753-789.116. Osborn, A.M., Bruce, K.D" Strike,P" Ritchie, D.A.. Appl. Environ.

MicrobiaL, 1993,59.4024-4030.117. Bruce, K.D.Appl. Environ, Micr()bioL 1997, 63,4914-4919.118. Liebert, C.A. Wireman, J.• Smith. T .• Sununers, A.a, Appl. Environ.

Microbial. 1997,63", 1066·1076.119. Ravel. I., DiRuggiero. J.; Rohb. F;T.,HiIl, R.T. J. BacterioL2000, J82,

2345-2349.120. Ravel, I" S~rernpf, B.• Hill. R.T..AppL Environ. Microbial. 1998, 64,

3383-3388.121. Reyes, N.S., Frisber. M.E., Sbbecky. P.A. FEMSMicrobial Beol. 1!)99,

30, 273-284.122. Hines, M.E., Horvat. M.; Faganeli, J., Bbnzongo, I-C.I, Barkay. T .•

Major, KIt. Scott, K1., Bailey. B.A., Warwick. J.1., Lyons, W.B.Enviroti. Res. 2000, 83, 129-139.

1-23. Jeffrey. W.K., Nazaret, S., Barkay. T. Microbia/uoL 1996,32,293­303.

124. Selifonova. O~; Burlage, R.; Barkay, T. Appl. Environ. Microbial.1993.59,3083-3090.

125. Rasmussen, L.D.; Turner, R.R.;Barkay, T. AppL Environ. MicrobiQI.1997,63,3291-32-93.

126. Barkay, T.; Gilman, M.; Turner. R.lt AppL Environ. Microbial. 1997.63,4267-4271.

127. de: Weger. L. A:, Dunbar, P., Mahafee,W.F.; Lugtenberg, B.J.!.;Sayler, 0.5. Appl. Environ. MicTobiol.1991, 57. 3641-3644.

128. Hill. P. J.; Rees, C.B.D.; Winson, M.K.; Stewart, G.S.A.a. BiotechrlOl.Appl. Bfochem. 1993.17,3-14.

129. Virta.M.; Lampinen, J.; Karp, M. Anal. Chern; 1995, 67, 661-669.130. Kelly: C.A.; Scott. K.J.; Holoka. M.~ Rood. l.W.M.ln preparation.13 L Sc-oll; K.L; Rudd. J.W.M.; Kelly, C.A.In preparation.132. Golding, G.R.; Kelly, C.A.~ Sparling, R.; Loewen, P.C.; Rudd,J.W.M.;

Barkay. T. Umnol. Oceanogr. In press.133. Leaner; n. PhD Dissertation, University of Maryland, COllege Park;

2001,182pp.134. Chen, S.; Wilson, D.B. Appl. Environ. Microbiol. 1997, 63. 2442-2445.B5. Horn, I.M.; Brunke, M.; Deckwe', W.; Timmi" KN. Appl. Env"

Microbial. 1994, 60, 357-362.