Copper accumulation by sulfate-reducing bacterial biolms

C. White, G.M. Gadd *Department of Biological Sciences, University of Dundee, Dundee DD1 4HN, Scotland, UK

Received 26 November 1999; accepted 27 December 1999

Abstract

Sulfate-reducing bacterial biofilms were grown in continuous culture. When exposed to medium containing 20 or 200 WM Cu, biofilmsaccumulated Cu. Energy-dispersive X-ray analysis (EDXA) showed that accumulation of Cu occurred in the form of sulfides while EDXAmapping of Cu and S in biofilm sections indicated that they were not uniformly distributed but located in the surface of the biofilm. Whilethe polymer content of biofilm exposed to 20 WM Cu did not appear to increase relative to control Cu-free biofilms, biofilms exposed to 200WM Cu accumulated carbohydrate and smaller amounts of protein throughout the incubation period. The mechanism of uptake, therefore,appeared to be precipitation of Cu sulfides at the biofilm surface or in the liquid phase followed by entrapment of precipitated Cu sulfide bythe exopolymer-enhanced biofilm. 2000 Federation of European Microbiological Societies. Published by Elsevier Science B.V. Allrights reserved.

Keywords: Sulfate-reducing bacterium; Copper accumulation; Metal^biolm interaction; Bioprecipitation

1. Introduction

The ability of sulfate-reducing bacteria (SRB) to removetoxic metals from waters and waste waters by precipitatinghighly insoluble suldes [1] has bioremediation potentialand is already used in the large-scale biological treatmentfor toxic metals [2,3]. However, current bioreactors, oftenbased on those used in water-treatment technology, arelarge, in contrast to biolm reactors which represent ameans of increasing process intensity thus reducing resi-dence time, working volume and cost. Free-living and bio-lm SRB may interact with metals by precipitation ofmetal suldes [1,4,5] which can be enhanced by metabolicprocesses leading to raised pH [1]. Metals may also under-go biosorption by cell surfaces [3,6,7] and extracellularpolymeric substances (EPS) [8]. EPS comprises a mixtureof polysaccharides, mucopolysaccharides and proteinswhich varies in composition between species and cultureconditions [9] and can take up soluble metals [10] or neparticulates [11,12]. In a previous study, SRB biolms ex-posed to Cd accumulated solid CdS and both the polysac-charide and protein content of the biolms increased si-multaneously [5]. The distribution of CdS was such thatbound Cd occurred mainly at or near the biolm surface

which indicated that precipitated Cd adhered to, but didnot penetrate, the biolm [5]. Although Cu readily formssuldes, its interactions with microorganisms can dier inseveral important respects from those of Cd [13] with itssulde chemistry diering from that of Cd in that it formsmixed Cu(I) and Cu(II) suldes which are less stable thanCdS, especially at low pH [14^16]. The present study wastherefore undertaken in order to explore whether theseaspects of biology and chemistry aect the interactionsbetween Cu and SRB biolms.

2. Materials and methods

2.1. Organisms and maintenance

The mixed sulfate-reducing culture was initially selectedfrom natural sediment samples by chemostat culture [1].No fermentative activity was evident when lactate wasused as a substrate and methanogenesis was not detectedon any substrate [1,17]. The culture was selected for bio-lm growth and biolms were maintained in anaerobicbatch culture in 10 ml volumes of SL10 medium [18] inscrew-topped glass tubes each containing a sterile 5U20mm glass coupon and incubated at 20C for 14 days.Biolm growth was then preferentially subcultured byaseptically transferring the coupon to a further tube ofSL10 medium containing a similar coupon [5].

0378-1097 / 00 / $20.00 2000 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.PII: S 0 3 7 8 - 1 0 9 7 ( 0 0 ) 0 0 0 0 2 - 1

* Corresponding author. Tel. : +44 (1382) 344266;Fax: +44 (1382) 344275; E-mail : g.m.gadd@dundee.ac.uk

FEMSLE 9244 3-2-00

FEMS Microbiology Letters 183 (2000) 313^318

www.fems-microbiology.org

2.2. Experimental culture

Experimental cultures were grown on 10 mm wide Plas-tikard polystyrene coupons (Slaters Ltd, Matlock, UK)suspended in a 1 l Quickt vessel equipped for continu-ous culture [5]. Mixed SRB biolm culture inoculum wasdeveloped in continuous culture in 10 ml syringe barrelsand inoculated into the 1 l culture vessels as describedpreviously [5]. After 48 h batch growth, continuous me-dium ow was started and the biolm was allowed todevelop for 7 days in SL10 medium at 20C and a dilutionrate of 0.2 h31 to produce a mature biolm prior to thestart of the experimental run. Experimental runs were car-ried out under identical conditions except that an appro-priate volume of 100 mM CuSO4 was added to the SL10medium to yield the required nal Cu concentration andNa2S was omitted from both control and metal-containingcultures to avoid premature precipitation of the added Cu.Neither redox poising agents nor metal chelating agentswere present in this medium. CuSO4 stocks were made upin Milli-Q ultrapure water (Millipore UK Ltd, Watford,UK) with the addition of 0.01% (v/vaq) analar H2SO4(Merck, Lutterworth, UK) and stored in acid-rinsed(0.01% (v/vaq) H2SO4) polypropylene bottles.

2.3. Sampling and analysis

Samples were taken by removing entire coupons in apredetermined random sequence. Four 10 mm long sub-samples were removed from the remainder of the couponfor electron microscopy and chemical assays of biolmCu, protein and carbohydrate. This procedure ensuredthat the samples were taken from areas that were sub-merged under a minimum of 10 mm of medium duringgrowth. Cu was assayed by atomic absorption spectropho-tometry (AAS) following digestion in 6 M HNO3 [1,19].Protein was extracted by vortexing the coupon in 1.0 or2.0 ml of 0.5 M NaOH with the addition of approximately0.5 cm3 of 0.5 mm diameter glass beads at 200 rpm for5 min followed by extraction of the resuspended biolmmaterial for 30 min and assaying 100 or 200 Wl of extractby the Bradford method [1]. Carbohydrate was assayedusing the anthrone method [20] after vortexing in distilledwater with glass beads in the same way and removing asuitable volume of suspended biolm. Previous experi-ments had indicated that no further protein or carbohy-drate was extracted with more prolonged treatment (datanot shown). Dissolved sulde was assayed polarographi-cally in whole culture [1].

2.4. Electron microscopy and EDXA analysis

For electron microscopy and EDXA analysis, segmentsof coupon 5U10 mm in size, with attached biolms, wereembedded in 1.5 ml of melted 5% pig-skin gelatin at 50C.They were immersed for 2.5 h to allow inltration and

then cooled to 4C to solidify the gelatin. The couponwas removed and the excess gelatin trimmed o leavingthe complete biolm embedded in gelatin. Approximately5 mm cubes of gelatin containing the embedded biolmwere xed in 2.5% v/vaq glutaraldehyde in 100 mM sodiumphosphate buer, pH 7.0, at 4C overnight and dehy-drated through a 10% incremental ethanol series. The gel-atin had a wax-like consistency following this treatmentand a thinner section (approximately 2 mm thick) was cutusing a hand razor and air-dried prior to sputter coatingwith carbon. Scanning EM and EDXA analysis were car-ried out using a Jeol JSM-35 scanning electron microscopeand EDXA spectra were analyzed using an Apple com-puter equipped with a Link Interface P1445 and software(Link Systems Ltd, High Wycombe, UK).

3. Results

3.1. Copper accumulation by SRB biolm

Biolms accumulated Cu when it was present, the

Fig. 1. Accumulation of (a) Cu, (b) carbohydrate and (c) protein in con-tinuous mixed culture SRB biolms grown with no added Cu (control)(a), 20 WM Cu (b) and 200 WM Cu (E) over a period of 14 days fol-lowing 7 days cultivation with no Cu present in all cases. Each point isthe mean of at least two separate samples and the bars indicate thestandard error of the mean (S.E.M.).

FEMSLE 9244 3-2-00

C. White, G.M. Gadd / FEMS Microbiology Letters 183 (2000) 313^318314

amount being about ten-fold greater at 200 WM than at 20WM (0.047 and 0.489 Wmol cm32 day31 at 20 and 200 WMCu respectively) (Fig. 1, Table 1). Trace metals werepresent in the medium in only 0.1^2.0 WM concentrationsand were undetectable in the biolm with the exception ofFe (73 WM in the medium). Fe accumulated in the biolmin both control and Cu-containing cultures. The amountof Fe accumulated was the same in both the presence andabsence of Cu (0.28 0.07 and 0.26 0.08 Wmol cm32 re-spectively over 21 days) which indicated that there waslittle or no interference between Fe and Cu uptake.

3.2. Protein and carbohydrate content of SRB biolms

Neither the protein nor the carbohydrate content ofcontrol biolms (no added Cu) increased signicantly dur-ing the experiment (Fig. 1) and both protein and polysac-charide increased only slightly in the biolms exposed to20 WM Cu but this was not statistically signicant (Table1). On the other hand, at 200 WM Cu the biolms accu-mulated signicantly larger amounts of EPS, with approx-imately twice the protein and almost ten times the carbo-hydrate content of the controls (Table 1). This indicated asignicant role for carbohydrate in the uptake or retentionof copper. There was a lower increase in protein content,both in absolute terms and as a proportion of controllevels (Fig. 1, Table 1) so its role appears to be of lesserimportance than polysaccharide in relation to Cu uptake.

3.3. Electron microscopy and EDXA analysis

The gelatin sections containing the biolm showed noappreciable distortion and remained at following dryingand carbon coating. The biolm was visible under S.E.M.as an area of etched appearance to one side of the sectionwhich was separated from the smooth surface of the gel-atin by an irregular boundary following the original sur-face of the biolm. This section showed the biolm to bespatially uneven, containing water spaces similar to thoseobserved in other bacterial biolms [21]. The at base ofthe biolm (which was in contact with the support duringgrowth) corresponded to one side of the gelatin section

(Fig. 2a). EDXA spectra of biolms exposed to 200 WMCu showed the presence of large Cu peaks which wereconsiderably smaller or absent in the controls where onlytraces of Cu were present. The Cu-exposed specimens alsoshowed strong peaks for S in addition to Cu, althoughthese were also present (but smaller) in the control speci-mens (Fig. 3). These data are consistent with the deposi-tion of Cu suldes in the biolm. The EDXA spectra werealso used to provide X-ray energy windows for mappingboth Cu and S. In the EDXA map, each dot results fromone X-ray count within the energy window for the elementas the electron beam scans the object and a higher densityof dots in an area of the image indicates a greater concen-tration of the target element, showing its distribution [22].Both Cu and S were found to be present in the greatestconcentrations in the surface layers of the biolm (Fig.2b,c) which indicated that Cu suldes were either precipi-tated in the bulk phase or at the interface with the biolmand were accumulated by entrapment in the biolm.

4. Discussion

Cu appeared to be mainly accumulated by SRB biolmsas solid copper suldes. Suldes were the dominant copperspecies because of the presence of sulde produced by theSRB biolm (in concentrations 1.5^4.5 mM) and very lowsolubilities of CuS and Cu2S (solubility products being1.27U10336 and 2.26U10348 respectively) [16] resultingin precipitation of the suldes. This also occurred whenCu2 was fed to a chemostat containing the free-livingmixed SRB culture [1]. In addition, the biolm also en-trapped precipitated solid Cu suldes and responded totheir presence by accumulation of additional EPS. It wasnot possible to distinguish whether this resulted from addi-tional production or reduced sloughing. Accumulation ofEPS also accompanied Cd uptake by SRB biolms [5] andparticulate accumulation has been observed to have a sim-ilar eect on other bacterial biolms [12]. The higher ac-cumulation of carbohydrate rather than protein with bothCu and Cd suggests that the secretion of extracellularpolysaccharide may be a response to the accumulation

Table 1Copper, EPS and protein accumulation by mixed sulfate-reducing biolm cultures exposed to varying copper concentrations

Run no. Medium Cu concentration(WM)

Biolm Cu(Wmol cm32)

Biolm protein(mg BSA eq. cm32)

Biolm carbohydrate(Wmol glucose eq. cm32)

1 0 0.03 0.66 0.382 20 0.95 0.71 0.743 200 7.82 1.36 3.70LSD 0.73 0.121 0.107

The amounts shown were measured after 14 days incubation at the Cu concentration indicated.Each value is the mean of eight determinations with the least signicant dierence (LSD) being calculated by variance analysis and comprising the resid-ual standard error of the whole data set for the three runs multiplied by an appropriate t value for the required signicance level at 21 degrees of free-dom [23].Mean values diering by more than the LSD shown are distinct at a P6 0.01 signicance level.

FEMSLE 9244 3-2-00

C. White, G.M. Gadd / FEMS Microbiology Letters 183 (2000) 313^318 315

Fig. 2. Scanning electron micrographs and EDXA maps of SRB biolms showing sections of biolms grown with (a) no added Cu and (b) 200 WM Cuembedded in 5% pig-skin gelatin. The biolm is visible as a pale area at the lower part of each frame and its surface is visible as an irregular boundarywith the embedding gelatin which appears darker. The EDXA maps show the distribution of (c) Cu (815^825 keV, 44 722 counts) and (d) S (210^220keV, 664 281 counts) in the biolm section shown in (b). Both Cu and S can be seen to be concentrated in the surface of the biolm although S is alsovisible throughout due to the presence of suldes of metals other than Cu. A control map of the same eld with the energy window covering no ele-mental peaks for relevant elements (800^810 keV, 19 473 counts) (e) shows the absence of topographical emissions. The scale bar shows 100 Wm.

FEMSLE 9244 3-2-00

C. White, G.M. Gadd / FEMS Microbiology Letters 183 (2000) 313^318316

of solid metal suldes in the biolm. While there werequantitative dierences, it is clear that the processes ofCu and Cd accumulation in SRB biolms were fundamen-tally similar, being dominated by sulde precipitation andentrapment of the resulting metal sulde particles by thebiolm EPS. As discussed above, sulde precipitation is anecient means of separating toxic metals from solution,while entrapment in a biolm is also a further means ofreducing the mobility of the metals, which could be ofvalue in bioremediation.

Acknowledgements

G.M.G. gratefully acknowledges nancial support fromthe BBSRC (LINK Programme: Biological treatment ofsoil and water, Ref: 94/BSW 05375). This work was car-ried out as part of the overall project Metal-biolm inter-actions in sulfate-reducing bacterial systems in collabora-tion with British Nuclear Fuels plc, Preston, Lancashireand Westlakes Research Institute, Whitehaven, Cumbria.The authors also wish to thank Martin Kierans of thisDepartment for assistance with electron microscopy andEDXA.

References

[1] White, C. and Gadd, G.M. (1996) Mixed sulphate-reducing bacterialcultures for bioprecipitation of toxic metals: factorial and response-surface analysis of the eects of dilution rate, sulphate and substrateconcentration. Microbiology 142, 2197^2205.

[2] Barnes, L.J., Scheeren, P.J.M. and Buisman, C.J.N. (1994) Microbialremoval of heavy metals and sulphate from contaminated ground-water. In: Emerging Technology for Bioremediation of Metals(Means, J.L. and Hinchee, R.E., Eds.), pp. 38^49. Lewis Publishers,Boca Raton.

[3] Gadd, G.M. and White, C. (1993) Microbial treatment of metal pol-lution - a working biotechnology? Trends Biotechnol. 11, 353^359.

[4] White, C. and Gadd, G.M. (1998) Reduction of metal cations andoxyanions by anaerobic and metal-resistant organisms: chemistry,physiology and potential for the control and bioremediation of toxicmetal pollution. In: Extremophiles : Microbial Life in Extreme Envi-ronments (Grant, W.D. and Horikoshi, T., Eds.,) pp. 233^254. JohnWiley and Sons, New York.

[5] White, C. and Gadd, G.M. (1998) Accumulation and eects of cad-mium on sulphate-reducing bacterial biolms. Microbiology 144,1407^1415.

[6] Gadd, G.M. (1992) Heavy metal pollutants: environmental and bio-technological aspects. In: Encyclopedia of Microbiology (Lederberg,J., Ed.), pp. 351^360. Academic Press, San Diego.

[7] Gadd, G.M. (1992) Molecular biology and biotechnology of micro-bial interactions with organic and inorganic heavy metal compounds.In: Molecular Biology and Biotechnology of Extremophiles (Herbert,R.A. and Sharp, R.J., Eds.), pp. 25^66. Blackie, Glascow.

[8] Beech, I.B. and Cheung, C.W.S. (1995) Interactions of exopolymersproduced by sulphate-reducing bacteria with metal ions. Int. Biodet.Biodeg. 35, 59^72.

[9] Zinkevich, V., Bogdarina, I., Kang, H., Hill, M.A.W., Tapper, R.and Beech, I.B. (1996) Characterization of exopolymers producedby dierent isolates of marine sulphate-reducing bacteria. Int. Biodet.Biodeg. 37, 163^172.

[10] Videla, H.A. (1994) Biocorrosion of nonferrous metal surfaces. In:Biofouling and Biocorrosion in Industrial Water Systems (Geesey,G.G., Lewandowski, Z. and Flemming, H.K., Eds.), pp. 231^243.Lewis Publishers, Boca Raton.

[11] Flemming, H.-K. (1995) Sorption sites in biolms. Water Sci. Tech-nol. 32, 27^33.

[12] Vieira, M.J. and Melo, L.F. (1995) Eect of clay particles on thebehaviour of biolms formed by Pseudomonas uorescens. WaterSci. Technol. 32, 45^52.

[13] Gadd, G.M. (1992) Metals and microorganisms: a problem of de-nition. FEMS Microbiol. Lett. 100, 197^204.

[14] Greenwood, N.N. and Earnshaw, A. (1983) Chemistry of the Ele-ments. Pergamon Press, Oxford.

Fig. 3. EDXA spectra obtained from mixed culture SRB biolms fol-lowing 21 days continuous culture. The spectra were obtained from(a) control cultures grown in the absence of Cu and (b, c) culturesgrown in 20 and 200 WM Cu respectively for 14 days following 7 daysdevelopment in the absence of Cu. The horizontal axis represents X-rayenergy (0^1023 keV) and the vertical axis shows emission counts (totalcounts for each spectrum = 3.1^3.5U105).

FEMSLE 9244 3-2-00

C. White, G.M. Gadd / FEMS Microbiology Letters 183 (2000) 313^318 317

[15] Svehla, G. (1996) Vogels Qualitative Inorganic Analysis. Longman,Singapore.

[16] Chang, J.C. (1993) Solubility product constants. In: CRC Handbookof Chemistry and Physics (Lide, D.R., Ed.), pp. 8^39. CRC Press,Boca Raton.

[17] White, C. and Gadd, G.M. (1996) A comparison of carbon/energyand complex nitrogen sources for bacterial sulphate-reduction: po-tential applications to bioprecipitation of toxic metals as sulphides.J. Ind. Microbiol. 17, 116^123.

[18] Widdel, F. and Pfennig, N. (1981) Studies on dissimilatory sulphate-reducing bacteria that decompose fatty acids. 1. Isolation of newsulphate-reducing bacteria enriched with acetate from saline environ-ments - description of Desulfobacter postgatei gen-nov, sp-nov. Arch.Microbiol. 129, 395^400.

[19] White, C. and Gadd, G.M. (1995) Determination of metals and metaluxes in algae and fungi. Sci. Tot. Environ. 176, 107^115.

[20] Herbert, D., Phipps, P.J. and Strange, R.E. (1971) Chemical analysisof microbial cells. Methods Microbiol. 5, 210^344.

[21] Costerton, J.W., Lewandowski, Z., Caldwell, D.E., Korber, D.R. andLappin-Scott, H. (1995) Microbial biolms. Annu. Rev. Microbiol.49, 711^745.

[22] Morgan, A.J. (1985) Electron Microscopy for Biologists. Oxford Uni-versity Press, Oxford.

[23] Sokal, R.R. and Rohlf, F.J. (1981) Biometry. W.H. Freeman, Ox-ford.

FEMSLE 9244 3-2-00

C. White, G.M. Gadd / FEMS Microbiology Letters 183 (2000) 313^318318