Microbial Community- And Metabolite Dynamics of an Anoxic

Jun 11, 2010 - Microbial community structure and dynamics in a membrane bioreactor supplemented with the flame retardant dibromoneopentyl glycol. Mora...
5 downloads 9 Views 411KB Size
Environ. Sci. Technol. 2010, 44, 4884–4890

Microbial Community- And Metabolite Dynamics of an Anoxic Dechlorinating Bioreactor F A R A I M A P H O S A , * ,†,§ H A U K E S M I D T , †,§ W I L L E M M . D E V O S , †,§ A N D ¨ L I N G ‡,§ WILFRED F. M. RO Laboratory of Microbiology, Wageningen University, Molecular Cell Physiology, Faculty of Earth and Life Sciences, VU University Amsterdam, The Netherlands, and NGI Ecogenomics Consortium, Amsterdam, The Netherlands

Received December 22, 2009. Revised manuscript received May 26, 2010. Accepted May 28, 2010.

Monitoring and quantification of organohalide respiring bacteria is essential for optimization of on-site bioremediation of anoxic subsurface sites contaminated with chloroethenes. Molecular monitoring and model simulations were applied to determine degradation performance of an in situ dechlorinating bioreactor and its influence on the contamination plume. Dehalococcoides was the dominant dechlorinating microorganism as revealed by qPCR targeting 16S rRNA- and chloroethene reductive dehalogenase-encoding genes (tceA, vcrA, bvcA). The presence of all three reductive dehalogenases genes indicated coexistence of several distinct organohalide respiring bacterial populations in the bioreactor and groundwater. Mass balancing revealed that main dechlorinating activities were reduction of cis-dichloroethene and vinyl chloride. Analysis of growth kinetics showed that when performance of the bioreactor improved due to especially the addition of molasses, dechlorinating microorganisms were growing close to their maximum growth rate. Once near-complete dehalogenation was achieved, Dehalococcoides only grew slowly and population density did not further increase. The bioreactor influenced dechlorinating populations in the plume with subsequent decrease in chlorinated compound concentrations over time. In the present study, a combination of molecular diagnostics with massbalancing and kinetic modeling improved insight into organohalide respiring bacteria and metabolite dynamics in an in situ dechlorinating bioreactor and showed its utility in monitoring bioremediation.

Introduction Bioremediation of sites contaminated with chlorinated ethenes, such as commonly used industrial solvents tetrachloroethene (PCE) and trichloroethene (TCE) (1, 2), makes use of organohalide respiring bacteria (OHRB) that can anaerobically reduce these toxic compounds to ethene. Key OHRB belong to the genus “Dehalococcoides” (3-7), which reductively dechlorinate PCE to TCE or cis-dichloroethene * Corresponding author. Wageningen University, Laboratory of Microbiology, Dreijenplein 10, 6703HB, Wageningen, The Netherlands. Phone: +31 317 483486; fax: +31 317 483829; email: [email protected]. † Wageningen University. § NGI Ecogenomics Consortium. ‡ VU University Amsterdam. 4884

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 13, 2010

(cDCE), and vinyl chloride (VC) to ethene. Also important are microorganisms from other genera, including Desulfitobacterium, Dehalobacter, Geobacter, and Sulfurospirillum, which are able to dechlorinate PCE to cDCE. Reductive dechlorination of PCE past cDCE, however, has so far been linked exclusively to Dehalococcoides spp. (8-12). In general, reduction of chloroethenes by OHRB is mediated by reductive dehalogenase (RDase) enzymes. RDase-encoding genes have emerged as appropriate diagnostic targets in studying and monitoring bioremediation of contaminated sites (11-17). Quantitative PCR protocols have been developed for detection and monitoring of Dehalococcoides populations and their RDase-encoding genes (17)). Specific targets are tceA, encoding the RDase catalyzing stepwise reductive dehalogenation of TCE to VC in strains 195 (18) and FL2 (5, 19), vcrA, coding for the DCE- and VC- to ethene reducing dehalogenase of strains VS (9) and GT (20), and bvcA that encodes the VC RDase of strain BAV1 (21). Continuing efforts are made in evaluating dechlorination potential of sites and monitoring their in situ bioremediation based on the presence of Dehalococcoides populations, and the abundance and expression of genes that encode RDases associated with the dechlorination of cDCE to VC and ethene (13, 16, 22). Information on growth kinetics and dynamics of Dehalococcoides spp., however, has so far only come from laboratory studies (23, 24). Therefore, the purpose of this study was to (i) describe and quantify key dechlorinating populations using genes that code for 16S rRNA and complementary RDases (tceA, bvcA and vcrA) for an on-site dechlorinating anaerobic bioreactor installed for the in situ bioremediation of a chloroethene-contaminated site, (ii) use mass-balancing and kinetic modeling to quantify the dechlorinating activities of the bioreactor, and (iii) integrate molecular data with mathematical analysis to determine bioreactor performance and degradation dynamics of Dehaloccocoides populations.

Experimental Details Site Description. The study site, located in Ede, The Netherlands, is undergoing in situ bioremediation of a contaminated plume of chlorinated ethenes using an anaerobic bioreactor (Supporting Information (SI) Figure S1-S2). The upflow anaerobic sludge bed bioreactor capacity was 7 m3 and had an inflow rate of 400 L/h. Porous granular material was used to make the sludge bed (SI Figure S2) A detailed description of the site is found in the SI. An anaerobic dechlorinating bioreactor has been installed on site for in situ bioremediation in December 2005. Groundwater was pumped into the bioreactor, and effluent containing a dechlorinating community was discharged back into the plume (SI Figure S1). Inflow rate to the 7 m3 bioreactor is 400 L/h and residence time was about 17.5 h. Molasses (138 g/L) was dosed into the influent stream to the bioreactor in 1:1500 dilution. To enhance bioremediation in the plume 15 m3 of molasses were added as carbon and energy source at 150 infiltration wells, covering 3600 m2 of the pollution plume, in July 2006 and March 2007. Furthermore, to improve degradation of chlorinated compounds beyond cDCE, the bioreactor was bioaugmented in August 2007 by addition of 1.8 m3 IJlst groundwater, which was rich in VC-degrading Dehalococcoides (25) (SI Figure S3). Sample Collection. Samples for physicochemical analysis were obtained monthly from the influent and effluent content (SI Figure S1-S2) and occasionally from plume monitoring wells. On-site analysis was done to determine the influent and effluent temperature, pH, oxygen, and redox potential 10.1021/es903721s

 2010 American Chemical Society

Published on Web 06/11/2010

(mV), and chemical analyses were performed in a certified laboratory (Analytico Milieu B.V., Barneveld, Netherlands) using ISO and EPA protocols (www.analytico.com). These analyses included chlorine hydrocarbons (PCE, TCE, cis-, and trans- isomers of DCE, and VC, di-, tri- and tetrachloromethane, di- and trichloroethane), inorganic chemicals (nitrate and nitrite, sulfate, sulfite and sulfide, iron and manganese), and volatile fatty acids (acetate, propionate, and butyrate). Samples for biomolecular analyses were obtained during a period of 12 months in December 2006 and in May, October, and December 2007 (SI Figure S3). Groundwater samples were taken from monitoring wells (MWI and MW2; SI Figure S1) in December 2006 and December 2007 to assess microbial communities present in the plume. Samples were obtained from monitoring wells at different depths (2-3 m, 5-6 m, 8-9 m, and 11-12 m), after three volumes of the sampled well were pumped out and discarded. The bioreactor was sampled from the influent, reactor content and effluent (SI Figure S2). Care was taken to obtain representative samples from the reactor including the high content of particulate matter. All samples were transported anaerobically to the laboratory at +4 °C. DNA Extraction. Samples (100-150 mL) were filtered through 0.2 µm polycarbonate filters (Millipore BV, Amsterdam, Netherlands) using a Millipore filtering system to collect microbial biomass. Filters were either stored at -20 °C or immediately processed for DNA extraction, whereby filters were cut into small strips and placed in bead beating tubes supplied with the FastDNA Spin Kit for Soil (MP Biomedicals, Solon, OH). Genomic DNA extraction was done according to the manufacturer’s instructions. Quantitative PCR. Quantitative PCR (qPCR) was performed using the iQ SYBR Green Supermix kit and the iQ5 iCycler (BioRad) for 16S rRNA genes of total and dehalogenating bacteria (Dehalococcoides, Desulfitobacterium, Dehalobacter), and Dehalococcoides RDase genes (tceA, bvcA, vcrA) (17, 26, 27). Plasmid constructs with the respective target genes were used as standards for the generation of calibration curves. PCR protocols were followed as previously described (28), and respective gene copy numbers were calculated as copies/mL of water sample. Modeling Approach. Percentage of chlorinated compound transformed and transformation rates of PCE, TCE, cDCE, and VC were calculated by mass-balancing based on the concentration of PCE, TCE, cDCE, and VC measured in the influent and effluent of the reactor. To fully dechlorinate PCE, TCE, cDCE and VC, respectively, 8, 6, 4, and 2 electrons are required. Electron-accepting capacity (EAC) of influent and effluent can thus be calculated as follows: EACx ) 8[PCE]x + 6[TCE]x + 4[cDCE]x + 2[VC]x

(1)

In which [-]x is either the concentration in the influent or effluent (µM). Percentage transformation of EAC in the influent was calculated as follows: % EAC transformation ) 100 ×

EACinfluent - EACeffluent EACinfluent (2)

As TCE, cDCE, and VC are intermediates in organohalide respiration and are being produced in the reactor from their parent molecules, the measured concentrations of these compounds do not directly inform about their transformation (in µM) in the reactor. The transformation was obtained as follows: ∆PCE ) [PCE]influent - [PCE]effluent

(3)

∆TCE ) ∆PCE + [TCE]influent - [TCE]effluent

(4)

∆cDCE ) ∆TCE + [cDCE]influent - [cDCE]effluent

(5)

∆VC ) ∆cDCE + [VC]influent - [VC]effluent

(6)

Rates of transformation were obtained by multiplication of the transformation with the influent inflow rate f (in L/h). The dilution rate D (h-1) of the reactor was calculated as follows: D ) f/V

(7)

where V is volume of the reactor in liters. An estimate of the expected number of organohalide respiring bacteria per mL groundwater was obtained via: X ) (4EAC·Y/B)10-9

(8)

In which Y is growth yield (in g biomass/mol electrons), B is weight (g biomass) per bacterium and 4EAC()EACinfluent - EACeffluent) expressed in µM. To calculate the number of cells per mL, growth yield parameters (Y ) 1.4 g biomass produced per mole of electron accepted, with a range of 0.48-9.4 g/mol (24, 29); and the weight per bacterium (B ) 5.8 × 10-14 g/cell (12)) derived for Dehalococcoides species, were applied. Net consumption of electrons was 15.3 µM, with a range from 8.6 to 19.0 µM). Becker (23) described the growth rate of Dehalococcoides with a Monod-type growth rate equation: TCE cDCE VC PCE + + + Ks,PCE Ks,TCE Ks,cDCE Ks,VC µi ) µmax,i PCE TCE cDCE VC 1+ + + Ks,PCE Ks,TCE Ks,cDCE Ks,VC

(9)

This equation consists of two independent terms, the first corresponds to the maximal growth rate µmax,i and the second comprising metabolite concentrations and their affinity constants (Ks,[PCE], Ks,[TCE], Ks,[cDCE], and Ks,[VC] are the affinity constants for PCE, TCE, cDCE, and VC, respectively). This second term, which we named “saturation index” thus describes the degree to which maximum growth rate was achieved and is independent of the value for µmax,i. The saturation index, which takes a value between 0 (no growth) and 1 (maximum growth rate achieved), was parametrized with published affinity constants and concentrations measured in the effluent (Table 1). To account for the spread in published data we calculated saturation on basis of the minimum, maximum and median values reported in literature. Statistical Analysis. Significance of the observed differences in gene copies number over time for influent, reactor and effluent was analyzed using Student’s Unpaired t-test.

Results Bioreactor Performance. Dechlorination performance of the reactor improved over time (Figure 1). There was an increase in the transformation of chlorinated ethenes (Figure 1A). Calculation of the electron accepting capacity (EAC) of the chlorinated compounds in the influent and effluent of the bioreactor, followed by subsequent determination of the percentage transformation of EAC in the reactor revealed that during the first 5 months only 20% of the EAC present in the influent was consumed (Figure 1B). However, a steady increase of electron acceptor transformation was then observed until after 10 months almost 100% of the inflowing chlorinated compounds were consumed. This high EAC was maintained for 4 months, after which it dropped to 50%. Second, transformation rates of the different chlorinated VOL. 44, NO. 13, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4885

TABLE 1. Constant Values Used to Calculate Saturation of Dehalogenating Capacity microorganism

substrate

Ks

reference

Dehalococcoides ethenogenes PCE TCE cDCE VC

0.54 0.54 0.54 290

23

Dehalobacter restrictus

PCE TCE

0.54 0.54

23

Desulfuromanas michiganensis

PCE TCE

0.54 0.54

23

Desulfitobacterium PCE1

PCE

0.54

23

Dehalococcoides BAV-1 containing consortium

cDCE tDCE VC

8.9 8.5 5.8

8

Dehalococcoides consortia (VS)

TCE cDCE VC

9 3.3 2.6

29

Dehalococcoides consortia (KB-1/VC)

TCE

10

29

Dehalococcoides consortia (Pinellas)

TCE

10.5

29

Point Magu enrichment culture

PCE TCE cDCE VC

3.9 2.8 1.9 602

45

Evanite enrichment culture

PCE TCE cDCE VC

1.6 1.8 1.8 62.6

45

enrichment culture (toluene)

PCE f 0.86 + 0.71 ethene

46

enrichment culture (lactate)

PCE f 0.23 + 1.45 ethene

46

methanogenic dechlorinating consortium

PCE

0.1 ( 0.05

47

methanogenic dechlorinating consortium

PCE TCE cDCE VC

2.8 1.5 3.0 360

48

dechlorinating consortium

PCE TCE cDCE VC

0.11 1.4 3.3 2.6

(29,49)

dechlorinating consortium

PCE TCE cDCE VC

70.7 17.40 11.9 383

50

compounds revealed, by correcting for the internal production of chlorinated intermediates, that transformation rates of VC were relatively high (>0.15 µmol/L/h) in the period April 2007 to November 2007 (Figure 1B), despite the fact that often a net increase of VC in the reactor was observed (Figure 1A). The EAC for VC transformation was comparable to that of cDCE and much higher than those of PCE and TCE. This relates to the fact that in the influent cDCE in general was present at the highest concentration. Rate of VC transformation increased from May 2007 and reached a peak in September 2007, one month after the bioreactor was bioaugmented with IJlst groundwater that contained actively VC-reducing populations (Figure 1B). 4886

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 13, 2010

Abundance and Spatial Distribution of 16S rRNA- and Dehalogenase-Encoding Genes in Bioreactor and Groundwater Samples. The presence of Dehalobacter and Dehalococcoides species 16S rRNA genes and Dehalococcoides RDase genes tceA, vcrA, and bvcA was confirmed in all samples (Figure 2A). Total bacterial 16S rRNA gene copies were generally higher in the effluent (2.0 × 108 ( 4.8 × 107 copies/ mL) than in the influent and reactor, which averaged 6.6 × 106 ( 1.4 × 106 copies/mL and 3.7 × 107 ( 1.0 × 107 copies/ mL, respectively. Dehalobacter species were detected at very low abundance throughout the system at all time points and ranged from 10 to 100 16S rRNA gene copies/mL in the reactor and effluent. Numbers observed in the influent were less than 10 gene copies/mL. Desulfitobacterium species were not detected in any of the reactor samples, including the field samples. Dehalococcoides spp. 16S rRNA gene copies in the reactor were around 103 copies/mL, which was lower than expected since the bioreactor operating conditions were designed so as to optimize for their growth and proliferation. In agreement with the improvement in reactor performance over time (Figure 1) numbers of Dehalococcoides 16S rRNA genes increased from 0.09 × 103 ( 0.03 × 103 in December 2006 to 5.76 × 103 ( 1.72 × 103 copies/mL in December 2007. Dehalococcoides 16S rRNA gene copy numbers and growth varied significantly over time in the reactor compared to the influent and effluent (P ) 0.0066) and were higher in the effluent. Bioaugmentation with IJlst groundwater in August 2007 did not influence significantly the numbers of Dehalococcoides. However, an improvement was observed for the transformation of chlorinated compounds (Figure 1A). An indication on whether OHRB were growing close to their maximum growth rate, was obtained by calculating a saturation index, as described in material and methods. Since Dehalococcoides was the dominant OHRB observed, the saturation index was calculated on basis of its reported affinity constants (23, 30, 31) and concentrations of chlorinated compounds in the reactor. This analysis revealed that in the period that transformation of chlorinated compounds was only 20% and started to increase to nearly 100% (December 2006 to August 2007), OHRB grew close to their maximum growth rates (i.e., the saturation index was near 1) (Figure 3). However, when near 100% transformation of chloroethenes was achieved, the growth rate dropped strongly, to about 10% of the maximum (September 2007 to December 2007). The vcrA gene was numerically dominant (1.0 × 104 copies/mL), compared to bvcA and tceA (Figure 2A). Total numbers of VC RDase-encoding genes, i.e. the sum of vcrA and bvcA, were higher than those observed for Dehalococcoides 16S rRNA genes. There was a significant and gradual increase in vcrA in the bioreactor over time suggesting proliferation of populations that harbor this gene, again in line with the increase in cDCE and VC transformation. Total bacterial and RDase gene numbers transferred to the plume in the effluent were either equal or higher than those in the suspended content of the reactor. In field samples, Dehalococcoides and its catabolic genes were most dominant, whereas Dehalobacter abundance was relatively low (