Transition from Cometabolic to Growth-Linked Biodegradation of

It is generally accepted that VC is readily biodegradable under aerobic ... Chemical oxygen demand (COD) was measured with a Hach kit (0−150 mg/l ra...
0 downloads 0 Views 167KB Size
Download PDF
Environ. Sci. Technol. 2001, 35, 4242-4251

Transition from Cometabolic to Growth-Linked Biodegradation of Vinyl Chloride by a Pseudomonas sp. Isolated on Ethene MATTHEW F. VERCE, RICKY L. ULRICH, AND DAVID L. FREEDMAN* Department of Environmental Engineering & Science, Clemson University, Clemson, South Carolina 29634

Pseudomonas aeruginosa strain DL1 was isolated on ethene as a sole carbon and energy source. When ethenegrown DL1 was first exposed to vinyl chloride (VC), the rate of VC consumption was very rapid and then declined sharply, indicative of a cometabolic process. A lack of growth and significant release of soluble products during this interval also indicates that the initial activity on VC was cometabolic. Following the rapid initial rate of VC cometabolism, a slow rate of VC utilization continued. After an extended period of incubation (>40 days), a transition occurred that allowed DL1 to begin using VC as a primary growth substrate, with an observed yield, maximum growth rate, and Monod half saturation coefficient of 0.21 mg of total suspended solids/mg VC, 0.046 d-1, and 1.17 µM VC, respectively, at 22 °C. Acetylene inhibits consumption of ethene and VC by ethene-grown cells, suggesting a monooxygenase is responsible for initiating metabolism of these alkenes. Resting cells grown on ethene cometabolized VC with an observed transformation capacity of 9.1 µmol VC/mg total suspended solids and a transformation yield of 0.22 mol VC/mol ethene. The presence of 40 µM ethene increased the rate and amount of VC cometabolized. However, consumption of higher concentrations of ethene decreased the total amount of VC consumed, and VC inhibited ethene utilization. A kinetic model was developed that describes substrate interactions during batch depletion of ethene and VC for a range of initial concentrations. The results suggest that ethene may stimulate in situ biodegradation of VC either by functioning as a primary substrate to support cometabolism of VC or by selecting for organisms that can utilize VC as a primary substrate.

Introduction Vinyl chloride (VC) is frequently found in groundwaters as an intermediate in the reductive dechlorination of tetrachloroethene, trichloroethene (TCE), and 1,1,1-trichloroethane (1). VC contamination of groundwater is of great concern because it is a known human carcinogen. Ongoing contamination of groundwater with VC is likely, because tetrachloroethene and TCE continue to be among the most heavily used industrial solvents. * Corresponding author phone: (864)656-5566; fax: (864)656-0672; e-mail: [email protected]. 4242

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 21, 2001

It is generally accepted that VC is readily biodegradable under aerobic conditions (2). Laboratory studies have shown VC can be cometabolized by mixed and pure cultures grown on numerous primary substrates, such as ethene (3-6), ethane (3, 4, 7), propene (8), isoprene (9), isopropylbenzene (10), 3-chloropropanol (11), ammonia (12, 13), butane (14), propane (15, 16), and methane (17, 18). These findings support the observation that VC usually does not accumulate in aerobic aquifers (19, 20). Despite the apparent biodegradability of VC under aerobic conditions, relatively few organisms have been isolated that grow on VC as a primary substrate. Organisms isolated to date that can grow on VC include several Mycobacteria (21, 22) and one Pseudomonas sp. (23). Isolating additional strains capable of growth on VC will help determine if this ability is limited to a few microbial species or is widespread in bacteria. Furthermore, characterizing such isolates will aid in the design of engineered bioremediation systems. In this study, ethene was evaluated as a primary substrate to support cometabolism of VC and as a selector for bacteria capable of growth on VC. Ethene was chosen because organisms capable of using it as a sole substrate may be better suited to metabolize VC, due to structural similarities (6). Furthermore, ethene is often associated with VC in contaminated aquifers (19, 20), suggesting ethene can stimulate cometabolic biodegradation of VC in situ. The objectives of this research were to isolate an organism capable of growth on ethene as a primary substrate, evaluate the ability of the isolate to also use VC as a primary substrate, characterize cometabolism of VC when the isolate is grown on ethene, and develop kinetic models that describe VC cometabolism in the presence and absence of ethene.

Materials and Methods Chemicals and Medium. VC gas (99.5%, containing 500 µM): In this case, eq 3 without terms B, C, and E (see eq S3) was fit to this data to determine kinactC (with kS, kC, and KC set equal to values obtained from type I and II experiments). Type IV. Addition of high concentrations of ethene (400 µM) and low concentrations of VC ( 150 µM) to ethene-grown culture: In this case, eqs 2 and 3 without term C (see eqs S6-S7) were fit to determine kenh (with kS, kC, KS, KC, kinactC, KiC, and KiS set equal to values obtained from types I-IV experiments). Type VI. Addition of high concentrations of ethene (400 µM) and VC (>150 µM) to ethene-grown culture: Equations 2 and 3 were fit to this type of data to determine kinactS (with kS, kC, KS, KC, kinactC, KiC, KiS, and kenh set equal to values obtained from types I-V experiments). With type II-VI experiments, no increase in initial biomass occurred, since either no growth substrate was present, the amount used was too low to permit more than a small change in initial biomass concentration, or a high enough concentration of nongrowth substrate was used to prevent any additional growth. Therefore, the initial biomass concentration (Xo) was used in place of the biomass concentration (X) for type II-VI experiments. The stepwise procedure to determine model parameters helps to minimize correlation among parameter estimates, since only one or two are determined with each type of experiment. However, a problem with this approach may occur when all of the terms are combined, since there may be interactions among terms that would be missed in experiments that eliminate or isolate one term for measurement. The approach outlined above minimizes this possibility, since the progression of experiments encompasses increasing levels of interaction among the parameters. Verification is provided in the Results by showing the fit of the completed model on all plots, not just the intermediate fits used to obtain the individual parameters. To provide a consistent source of culture for the kinetic experiments, the isolate was grown in an ethene-fed batch 4244

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 21, 2001

reactor, consisting of a 2.5-L glass reagent bottle capped with a gray butyl rubber septum (30 mm), held in place with a screw cap. The bottle contained 1.56-L of culture. The reactor was operated in a semicontinuous mode, as follows: 200mL of ethene was added; 300-mL of culture was removed as soon as all of the ethene was consumed (approximately 1 day) for kinetic experiments; and after 2 days, 300-mL of fresh medium was added; the pH was adjusted to 7.0 ((0.1); the headspace was purged with oxygen; and more ethene was added. This resulted in an average cell retention time of 10.4 days. Aseptic conditions were maintained during all manipulations. Between feedings, the reactor was stored at room temperature (22 ( 1 °C) on a Gyratory shaker table. After several months of operation in this mode, the biomass concentration in the reactor stabilized at approximately 338 mg of TSS/L (290 mg VSS/L). Effluent samples (25-mL) from the reactor were transferred to serum bottles (70-mL) to perform the six types of kinetic experiments described above. Growth Kinetics. Maximum specific growth rates (µmax) for ethene and VC were calculated as follows:

µmax ) YobskS

(5)

As indicated above, endogenous decay was insignificant for ethene-grown DL1 and therefore was not included in eq 5. Effect of Mass Transfer. The effect of mass transfer on evaluation of kinetic parameters was determined according to Smatlak et al. (34), as described previously (23). The aqueous phase data used to determine Monod parameters was calculated using eq 1, which assumes equilibrium between the gas and aqueous phases. This was tested by including mass transfer in the fit of eqs 2 and 3, using previously determined mass transfer coefficients (23). In all cases, the results with and without mass transfer were indistinguishable after 0.005 days, well before the first data points were measured. Numerical Methods. Fitting of model equations to substrate depletion data was performed numerically using the software “Aquasim 2.0” (35), as previously described (7). Types I and II data were weighted using an inference method (36), to obtain a reasonable fit at low concentrations. For all data types, the initial concentrations of VC and ethene were designated as parameters to be estimated but were constrained to within 5% of their calculated time zero values (based on the known amount of each gas added to the serum bottles). Nucleotide Sequence Accession Number. The complete sequence (1447 bases) of the 16S rRNA gene of strain DL1 was deposited with GenBank under accession no. AF323607.

Results Isolate Identification and Characterization. Strain DL1 is Gram negative, rod shaped, and motile. It possesses a bright yellow color when grown on ethene or VC, which is characteristic of other bacteria grown on ethene (6, 37). Evaluation of DL1 using the BBL CRYSTAL test database match closest to Pseudomonas aeruginosa (99.98% confidence, with no atypical reactions). See Table S1, Supporting Information, for a complete listing of the BBL CRYSTAL reactions. Based on the sequence of strain DL1’s 16S rRNA, the closest match using the GenBank database is to a Pseudomonas aeruginosa sp. The match to P. aeruginosa strain MF1 (23) is nearly perfect. In addition to the homology of their 16S rRNA genes, strains DL1 and MF1 both use VC and ethene as sole substrates (see below). However, RAPD-PCR fingerprinting analysis demonstrates that DL1 and MF1 are genotypically different strains of P. aeruginosa (Figure 1). The differences in banding patterns was reproducible when each strain was grown on VC as a sole substrate or on rich

FIGURE 1. RAPD analysis of strains MF1 and DL1. Lanes 1 and 6 correspond to the XVII 500 bp PCR marker ranging from 10 000-50 bp (Roche, catalog no. 1855646). Lanes 2 and 4 represent amplified chromosomal DNA from MF1 and DL1 cultured in MSM with VC as the sole carbon source. Lanes 3 and 5 represent amplified DNA from MF1 and DL1 subcultured in Luria-Bertani medium (Difco, catalog no. DF0446-17). medium. This is consistent with significant phenotypic differences between DL1 and MF1. DL1 has a bright yellow color when grown on ethene or VC, while MF1 is white when grown on these substrates. Furthermore, Monod growth kinetics of DL1 and MF1 on ethene and VC are quite different (see below). Growth on Ethene, VC, and Other Substrates. Growth of strain DL1 on ethene and VC as sole substrates was indicated by consumption of repeated additions of these substrates (Figure 2a,b), with a concomitant increase in optical density (Figure 2c). Average observed yields (Yobs) are reported in Table 1. The autoclave control results (Figure 2a,b, insets) demonstrate that consumption of ethene and VC in the live bottles was a biotic process, and the gray butyl rubber septa were very effective in retaining both compounds. Nearly all of the COD added as VC was oxidized to CO2 or converted to biomass, based on the small amount of sCOD remaining after VC was consumed (2.6%, Table 2). Stoichiometric recovery of the chlorine in VC as chloride (Table 2) indicates the sCOD products were not chlorinated. Estimates of kS and KS for ethene as a growth substrate (Table 1) were obtained by simultaneously fitting depletion curves from type I experiments (Figure 3). There was typically a short lag in ethene use after it was added. The duration of the lag was minimized by harvesting biomass for kinetic experiments from the batch reactor as soon as it finished its dose of ethene, but the lag was never completely eliminated. Therefore, to obtain an adequate estimate of kS and KS, the first two points in Figure 3 were excluded from the fit. Monod coefficients for VC as a growth substrate (Table 1) were obtained in a similar fashion. Acetylene inhibited ethene and VC use. Two sets of duplicate ethene-grown cultures fed 5% acetylene (v/v of headspace) and 500 µM ethene or VC consumed only 20% of these substrates over a 1 day period. Duplicate cultures not fed acetylene consumed all of the ethene or VC over the same period. Use of ethene and VC resumed after the acetylene was removed by purging with air, indicating acetylene inhibition was reversible for the conditions tested. In addition to ethene and VC, strain DL1 also grows on acetate and glyoxylate after 2 days of incubation and on glycolate and glycerol after 5-20 days. No increase in

FIGURE 2. Growth of strain DL1: (a) ethene as a sole substrate; (b) VC as a sole substrate (results from one of three bottles are shown; others behaved similarly); and (c) correlation of substrate consumed with optical density (triplicate bottles).

TABLE 1. Comparison of P. aeruginosa Strains Capable of Growth on VC and Ethene as Sole Substrates substrate

parameter

DL1a

MF1b

VC

Yobs (mg TSS/mg VC) kS (µmol VC/mg TSS/d) µmax (d-1) KS (µM)

0.21 ( 0.008 3.62 ( 0.172 0.046 1.17 ( 0.154

0.20 ( 0.007 0.41 ( 0.003 0.0048 0.26 ( 0.037

Ethene

Yobs (mg TSS/mg ethene) 0.85 ( 0.110 kS (µmol ethene/mg TSS/d) 2.99 ( 0.087 µmax (d-1) 0.071 KS (µM) 18.8 ( 1.20

0.72 ( 0.076 0.89 ( 0.022 0.018 3.9 ( 0.38

a

This study, ( one standard deviation.

b

From Verce et al. (23).

absorbance occurred in autoclaved controls or uninoculated medium. No growth occurred with chloroacetate or ethylene glycol (14 days of incubation) or with methane, ethane, propane, or butane (100 days of incubation). Transition from Cometabolism to Growth on VC. When pregrown on ethene and then provided with only VC, DL1 rapidly consumed varying concentrations of VC (Figure 4). Low amounts were completely consumed, while use of higher amounts quickly leveled off. At intermediate concentrations (bottle b), the rapid initial use of VC was followed by a slow decline in rate, then a transition to utilization of repeated VC additions. Rapid initial consumption of VC, followed by a lag in VC use, suggests that ethene-grown DL1 cometabolized VC during this interval. The subsequent gradual transition to sustained use of VC suggested it became a growth substrate. This hypothesis was confirmed by using a small amount of VOL. 35, NO. 21, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4245

TABLE 2. Chloride Recovery and sCOD Accumulation during Metabolism and Cometabolism of VC by Strain DL1 VC consumed (µmol/bottle)

culture conditions growth on VC, bottle #1 growth on VC, bottle #2 growth on VC, bottle #3 growth on VC, average cometabolism of VC, set A, bottle #1 cometabolism of VC, set A, bottle #2 cometabolism of VC, set A, bottle #3 cometabolism of VC, set A, average cometabolism of VC, set B, bottle #1 cometabolism of VC, set B, bottle #2 cometabolism of VC, set B, bottle #3 cometabolism of VC, set B, average a Average of duplicate measurements. * 100.

b

1860 1990 2200

Cl(µmol/bottle)a initial final 78.7 76.7 75.8

40.2 40.2 40.2

25.6 25.6 25.6

63.8 63.8 63.2

60.8 63.7 64.1

25.6 25.6 24.6

86.5 88.4 87.3

FIGURE 4. Cometabolism of VC by ethene-grown DL1, followed by transition to VC as a sole substrate (bottle b). Arrow indicates withdrawal of a sample from bottle b to start the growth experiment with VC (Figure 2b,c). culture from bottle b (day 111) to inoculate the cultures shown in Figure 2b,c, which successfully grew on VC as a sole substrate. Two additional experiments were undertaken to demonstrate that the initially rapid degradation of VC by ethenegrown DL1 is a cometabolic process. First, optical density was measured before and after the period of rapid VC utilization. (It was necessary to first dilute the ethene-grown biomass from the semicontinuous batch reactor to an optical density within the range observed during the growth experiments, Figure 2c.) The minor change in optical density (∆OD620 ) 0.011) after consumption of VC (80 µmol/bottle) was statistically equivalent to that in control cultures fed no VC (∆OD620 ) 0.012). Consumption of the same amount of VC by DL1 when it served as a growth substrate (Figure 2c) resulted in a statistically significant increase in optical density. Second, the amount of VC converted to sCOD was measured before and after the period of rapid VC utilization (Table 2). Set A consumed all of the VC added, 22% of which accumulated as sCOD in the medium. Chloride release was slightly below stoichiometric for set A, suggesting that at 9

103 98.6 103 101 94.9 94.9 93.6 94.5 100 98.4 96.1 98.2

sCOD (mg/bottle) initial final 5.74 2.77 2.39

6.95 8.24 8.42

1.15 1.15 1.15

1.87 1.76 1.91

1.15 1.15 1.15

1.65 1.69 1.61

% COD accumulationc 0.81 3.4 3.4 2.6 22 19 24 22 10 11 9.0 10

(Final Cl- - initial Cl-)/(VC consumed) * 100. c (Final sCOD - initial sCOD)/(COD of the VC consumed)

FIGURE 3. Utilization of low concentrations of ethene (type I experiment) to determine kS and KS for ethene. Results from one of three bottles are shown; others behaved similarly. Solid line represents model simulation. First two data points were not included in the fit (see text).

4246

1988 2038 2350

% VC dechlorinationb

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 21, 2001

least a portion of the sCOD consisted of chlorinated products. Set B consumed about one-half of the VC added, 10% of which accumulated as sCOD in the medium. This compares to only 2.6% conversion of VC COD to soluble products when it is being used as a growth substrate. The significantly higher percent conversion of VC to soluble products (rather than CO2 or biomass) immediately following growth on ethene further indicates that the initial, fast uptake of VC is cometabolic in nature, rather than linked to growth. Kinetics of VC Cometabolism. The kinetics of VC cometabolism by ethene-grown DL1 were determined over a range of substrate (ethene) and nongrowth substrate (VC) concentrations. Cometabolism of low concentrations of VC, in the absence of ethene, was adequately described by Monod kinetics (Figure 5a). Estimates of kC and KC (Table 3) were obtained by simultaneously fitting depletion curves from type II experiments (Figure 5a). Reiterative weighting of the data was necessary to capture depletion at low concentrations (inset, Figure 5a). Use of higher concentrations of VC (type III experiments) was characterized by significant inactivation (Figure 5b), which allowed for determination of kinactC (Table 3). The maximum amount of VC consumed was constant and represents the observed transformation capacity (TC,obs), 9.1 (( 0.69) µmol VC/mg TSS. When ethene and VC were fed together to ethene-grown DL1, three distinct kinetic features were observed. First, ethene and VC competitively inhibited each other’s use in a concentration dependent manner. Inhibition of VC use by ethene is apparent in Figure 6, in which high concentrations of ethene and low concentrations of VC resulted in much slower utilization of VC than when VC was added by itself. Similarly, the presence of VC significantly slowed the rate of ethene utilization (Figure 6b). Estimates of KiC and KiS (Table 3) were obtained by simultaneously fitting depletion curves from type IV experiments (Figure 6). Second, the rate of VC cometabolism was weakly stimulated by low concentrations of ethene. This enhancement is described by term B in the kinetics model (eqs 2 and 3). kenh was obtained (Table 3) by simultaneously fitting depletion curves from type V experiments (Figures 7a-c). The presence of a small amount of ethene resulted in an increase in TC,obs (9.58 µmol VC/ mg TSS) (Figure 7d), although this value is within the 95% confidence interval of TC,obs for VC-alone. Third, the maximum amount of VC cometabolized by ethene-grown DL1 decreased in the presence of high concentrations of ethene. Likewise, consumption of large amounts of VC resulted in incomplete use of ethene. kinactS (in term C), which captures this interaction, was obtained (Table 3) by simultaneously fitting depletion curves from type VI experiments (Figure 8).

FIGURE 6. Cometabolism of VC (O) by ethene-grown DL1 in the presence of high concentrations of ethene (4) at (a) low and (b) higher initial VC concentrations (type IV data), to obtain KiC and KiS. Model fits are shown for the open symbols by the solid lines. For comparison, results for VC alone (b) and ethene alone (2) are also shown, along with the corresponding model simulations (dotted lines) for ethene alone and VC alone.

FIGURE 5. Cometabolism of VC alone by ethene-grown DL1: (a) low VC concentrations (type II data), used to obtain kC and KC; (b) high VC concentrations (type III data), used to obtain kinactC. Solid lines represent model simulations.

TABLE 3. Summary of Cometabolic Kinetic Parameters Estimates parameter

value

data type used

kC (µmol VC/mg TSS/d) kenh (L/mg TSS/d) kinactC (L/mg TSS/d) kinactS (L/mg TSS/d) KC (µM VC) KiC (µM VC) KiS (µM ethene)

4.38 ( 0.030 0.251 ( 0.0026 2.82E-3 ( 2.0E-5 1.26E-2 ( 2.9E-4 4.82 ( 0.659 19.0 ( 1.05 3.19 ( 0.226

II V III VI II IV IV

Discussion DL1 is only the second P. aeruginosa strain isolated that is capable of using ethene and VC as sole carbon and energy

sources, with MF1 being the first (23). Strains DL1 and MF1 also appear to be the first Pseudomonas sp. isolated with an ability to grow on ethene. Other ethene-isolates include several strains of Mycobacteria (38, 39) and Xanthobacter (40). Kinetic parameters for DL1 and MF1 are compared in Table 1. These strains have many similar characteristics, despite different strategies employed in their enrichment and isolation. MF1 was isolated from an ethane enrichment culture that was later fed VC as a sole substrate (4), while DL1 was isolated from an ethene enrichment (3). Prior to MF1 and DL1, a few Mycobacteria sp. were the only isolates reported with an ability to use VC as a growth substrate (21, 22). All of these isolates have comparable yields on VC but differ greatly in their Monod kinetics. The maximum specific VC utilization rate for DL1 is about 10 times higher than for MF1 (Table 1), while the rate for L1 is an order of magnitude higher than for DL1 (22). The origin of the inocula may be a factor in these differences; MF1 and DL1 were derived from activated sludge that was not routinely exposed to ethene or VC, while L1 came from soil that had been contaminated for several years with VC. Caution must also be used when comparing kinetic parameters because they are a function of the physiological state of the cells during the experiment as well as the ratio of substrate to biomass COD (41). Monod coefficients for MF1 growing on VC were measured with biomass from a batch fed reactor at steady state, with a cell retention time of 140-d (23), while the values for DL1 were determined with biomass that accumulated VOL. 35, NO. 21, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4247

FIGURE 7. Cometabolism of VC (O) by ethene-grown DL1 in the presence of low concentrations of ethene (4), at varying initial VC concentrations (type V data), to obtain kenh. Model fits are shown for the open symbols by solid lines. For comparison, results for VC alone (b) and ethene alone (2) are also shown, along with the corresponding model simulations (dotted lines) for ethene alone and VC alone. during the growth experiment (Figure 2b,c). Even with these experimental differences, it is apparent that DL1 grows much faster on VC than MF1, but DL1 has a higher half-saturation coefficient for VC. Cometabolism of VC supported by ethene as a primary substrate has been reported previously for mixed (3, 6) and pure cultures (6). Ethene-grown DL1 cometabolizes VC at a slower rate than other organisms. However, TC,obs for DL1 is larger than that reported for most other cultures. (See Tables S2 and S3, Supporting Information for comparisons.) In addition to Tc,obs, the efficiency of cometabolism can be expressed by the transformation yield (Ty ) TC,obs‚Yobs). TY for DL1 is 0.22 mol VC/ mol ethene (0.49 mg VC/ mg ethene), which is similar to previously reported values: 0.27 mol VC/ mol ethene for an ethene enrichment culture (6), 0.013 (18) and 0.2-0.25 mol VC/mole methane (42) for batch methanotrophic cultures, and 0.32 (42) and 0.9 mol VC/mol methane (17) for sediment packed columns pre-enriched with methane. Two features of VC cometabolism by DL1 not captured by previous kinetic models are the slight enhancement of nongrowth substrate degradation at a small So and a decrease in the amount of nongrowth substrate consumed at large So and large Co when significant amounts of growth substrate are also consumed. The ability of the model (eqs 2 and 3) to describe an increase in the amount of VC consumed at a small So (Figure 6) is due to term B, which increases substrate and nongrowth substrate degradation rates by a constant amount (kenh). The subterm makes term B nonzero only at a small So. It also expresses the observation that VC consumption was enhanced primarily by the presence of a small amount of ethene. An alternate form of term B based on ethene consumption (So - S) resulted in much poorer fits, as did leaving this term out entirely and relying only on terms D and E, which encompass competitive inhibition (data not shown). The ability of the model to describe a decrease in the amount of VC consumed at higher initial VC and ethene concentrations (Figure 8) is due to term C, which decreases 4248

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 21, 2001

substrate and nongrowth substrate degradation rates in proportion to ethene consumption. The subterm acts as a switch that makes term C important when both So and Co are relatively large. Term C does not become small enough, however, for the condition of a large So and small Co, causing the model to overpredict growth substrate concentrations at longer times (Figure 6). Substitution of C for Co and S for So in these switching subterms always gave a poorer fit of the model. The empirical switching subterms by themselves describe no kinetic phenomena. They were developed in the absence of a mechanistic understanding of ethene metabolism and VC cometabolism by DL1, which was beyond the scope of this study. Even with these limitations, reasonable fits of the complete model were obtained when ethene and VC were present at the same time (Figures 6-8). The stepwise method of determining kinetic coefficients made it possible to fit for no more than two coefficients from a given type of data. The parameters kinactC, kinactS, and kenh were estimated independently. This limited the extent of correlation among parameters. Sensitivity analysis indicated the parameters KiC and KiS were not correlated. However, significant correlation exists between kS and KS and between kC and KC, based on off diagonal correlation matrix elements greater than 0.9 (43). Difficulty in obtaining unique estimates of Monod parameters has been noted for growth substrates (44) and for cometabolism of chlorinated compounds (32, 45). The effect of ethene consumption (normalized to the amount of biomass present) on the maximum amount of VC cometabolized is summarized in Figure 9, along with the model predictions of TC,obs for a corresponding amount of ethene use. TC,obs increased slightly when a small amount of ethene was used but then decreased by as much as 50% with the highest amount of ethene used. The model follows these trends, although it underestimated TC,obs at the higher amounts of ethene use. To obtain model predictions that matched the experimental amount of ethene consumed, it was necessary to run the model at lower initial concentrations

FIGURE 9. Effect of ethene consumption (normalized by the amount of TSS present) on experimental and model-predicted TC,obs. Initial VC concentrations (Co) are shown for results with ethene consumption.

FIGURE 10. Conceptual model for substrate interactions between VC and ethene by ethene-grown DL1. Cometabolism of VC inhibits ethene metabolism. Low levels of ethene do not compete for monooxygenase activity, and its presence slightly increases total VC consumption. High levels of ethene compete with VC for monooxygenase activity and nonproductive (i.e., cometabolic) consumption of ethene decreases the total amount of VC that can be cometabolized.

FIGURE 8. Cometabolism of VC (O) by ethene-grown DL1 in the presence of high concentrations of ethene (4), at varying initial VC concentrations (type VI data), to obtain kinactS. Model fits are shown for the open symbols by the solid lines. For comparison, results for VC alone (b) and ethene alone (2) are also shown, along with the corresponding model simulations (dotted lines) for ethene alone and VC alone. of VC (C0). Previous studies report a range of growth substrate effects on TC,obs. TCE consumption by a methanotrophic enrichment culture increased linearly with the amount of methane added, whereas TC,obs leveled off with increasing amounts of propane added to a propane-enrichment culture (46). However, TC,obs for toluene and phenol enrichment cultures reached a maximum and then declined at higher concentrations, presumably due to the inhibitory nature of these growth substrates (31). The decrease in TC,obs with an increase in ethene consumed (Figure 9) is not related to growth substrate toxicity, because cultures fed ethene alone showed no apparent inhibition (Figure 8). A similar decrease in Tc,obs for VC was reported for an ethane-grown P. aeruginosa

strain, although it was the presence of ethane, and not its consumption, that caused the effect (7). Substrate interactions between VC and ethene for ethenegrown DL1 are depicted in Figure 10. The involvement of a monooxygenase is supported by acetylene inhibition of ethene and VC utilization. Acetylene is a known inhibitor of ethene monooxygenase (47). Formation of an epoxide intermediate is further suggested by the inability of DL1 to grow on ethane and ethylene glycol, which are other possible intermediates of bacterial ethene metabolism (48). Monooxygenase activity consumes reducing power, which is restored during subsequent oxidation of ethene-epoxide. Cometabolism of VC inhibits ethene metabolism. A low concentration of ethene does not significantly compete with VC for monooxygenase activity, and its presence results in a slight increase in total VC consumption. At sufficiently high concentrations of VC and ethene, competition between them for the monooxygenase increases. Metabolism of ethene is inhibited by VC and/or its metabolites (e.g., VC-epoxide), so that increasing ethene consumption causes a net drain on cell resources. Thus, nonproductive (i.e., cometabolic) consumption of ethene decreases the total amount of VC that can be cometabolized. The results of this study have implications for in situ biodegradation of VC. Previous research has shown that methane is an effective primary substrate in support of VC cometabolism. The TC,obs and TY for ethene-grown DL1 suggest ethene is similarly effective as a primary substrate. Methane and ethene are end products of the anaerobic conditions that promote reductive dechlorination of polychlorinated ethenes. As such, they are found in groundwater at many contaminated sites, possibly eliminating the need for exogenous addition of a primary substrate. Both compounds can promote cometabolism of VC in the aerobic fringe of a VOL. 35, NO. 21, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4249

contaminant plume. However, unlike methane, ethene may also select for organisms that are capable of degrading VC as a sole source of carbon and energy, as was observed with DL1. The relative ease of cometabolic and metabolic biodegradation of VC is consistent with field observations that VC rarely persists under aerobic conditions.

Ty

transformation yield (mass nongrowth substrate/ mass growth substrate)

Vaq

volume of aqueous phase (L)

Vhs

volume of headspace (L)

X

concentration of biomass (mg TSS/L)

Acknowledgments

X0

initial concentration of biomass (mg TSS/L)

Kate Sounders assisted in the experiments with alternate growth substrates. This research was supported in part by a grant from the U.S. Environmental Protection Agency.

Y

true growth yield (mg TSS/µmol substrate)

Yobs

observed growth yield (mg TSS/µmol substrate)

µmax

maximum specific growth rate (d-1)

Supporting Information Available The procedures for culture identification based on 16S rDNA and random amplified polymorphic DNA (RAPD) analysis, BBL crystal results for strain DL1, comparison of VC maximum specific cometabolism rates (kC), comparison of observed transformation capacities for VC (TC,obs), and equations used for estimation of kinetic parameters are provided in the Supplemental Information Section. This material is available free of charge via the Internet at http:// pubs.acs.org.

Nomenclature b

endogenous decay rate (d-1)

C

concentration of nongrowth substrate (µM)

Caq

aqueous concentration of volatile compound (µM)

C0

initial concentration of nongrowth substrate (µM)

HC

dimensionless Henry’s constant (s)

kC

maximum specific nongrowth utilization rate (µmol nongrowth substrate/mg TSS/d)

kenh

specific enhancement coefficient due to presence of growth substrate (L/mg TSS/d)

kinactC

specific inactivation rate constant due to consumption of nongrowth substrate (L/mg TSS/ d)

kinactS

specific inactivation rate constant due to consumption of growth substrate (L/mg TSS/d)

ks

maximum specific growth substrate utilization rate (µmol substrate/mg TSS/d)

Kc

Monod half-saturation coefficient for nongrowth substrate (µM nongrowth substrate)

KiC

inhibition coefficient indicating effect of nongrowth substrate on growth substrate utilization rate (µM nongrowth substrate)

KiS

inhibition coefficient indicating effect of growth substrate on nongrowth substrate utilization rate (µM growth substrate)

Ks

Monod half-saturation coefficient for growth substrate (µM growth substrate)

MTOT

total mass of substrate or nongrowth substrate (µmol)

S

concentration of growth substrate (µM)

So

initial concentration of growth substrate (µM)

t

time (d)

Tc,obs

observed transformation capacity (mass nongrowth substrate/mg TSS)

4250

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 21, 2001

Literature Cited (1) Vogel, T. M.; Criddle, C. S.; McCarty, P. L. Environ. Sci. Technol. 1987, 21, 722-736. (2) Davis, J. W.; Carpenter, C. L. Appl. Environ. Microbiol. 1990, 56, 3878-3880. (3) Freedman, D. L.; Herz, S. D. Wat. Environ. Res. 1996, 68, 320328. (4) Freedman, D. L.; Verce, M. F. Ethene- and ethane-promoted biodegradation of vinyl chloride. In Situ and On-Site Bioremediation; Alleman, B. C., Leeson, A., Eds.; Battelle Press: Columbus, OH; 1997; Vol. 3, pp 255-260. (5) Freedman, D. L.; Danko, A. S.; Verce, M. F. Wat. Sci. Technol. 2001, 43, 333-340. (6) Koziollek, P.; Bryniok, D.; Knackmuss, H. J. Archives Microbiol. 1999, 172, 240-246. (7) Verce, M. F.; Freedman, D. L. Biotechnol. Bioengin. 2001, 71, 274-285. (8) Ensign, S.; Hyman, M.; Arp, D. Appl. Environ. Microbiol. 1992, 58, 3038-3046. (9) Ewers, J.; Freier-Schroder, D.; Knackmuss, H. J. Arch. Microbiol. 1990, 154, 410-413. (10) Dabrock, B.; Riedel, J.; Bertram, J.; Gottschalk, G. Archives Microbiol. 1992, 158, 9-13. (11) Castro, C. E.; Wade, R. S.; Riebeth, D. M.; Bartnicki, E. W.; Belser, N. O. Environ. Toxicol. Chem. 1992, 11, 757-764. (12) Vannelli, T.; Logan, M.; Arciero, D. M.; Hooper, A. B. Appl. Environ. Microbiol. 1990, 56, 1169-1171. (13) Rasche, M. E.; Hyman, M. R.; Arp, D. J. Appl. Environ. Microbiol. 1991, 57, 2986-2994. (14) Hamamura, N.; Page, C.; Long, T.; Semprini, L.; Arp, D. J. Appl. Environ. Microbiol. 1997, 63, 3607-3613. (15) Phelps, T. J.; Malachowsky, K.; Schram, R. M.; White, D. C. Appl. Environ. Microbiol. 1991, 57, 1252-1254. (16) Malachowsky, K. J.; Phelps, T. J.; Teboli, A. B.; Minnikin, D. E.; White, D. C. Appl. Environ. Microbiol. 1994, 60, 542-548. (17) Dolan, M. E.; McCarty, P. L. Environ. Sci. Technol. 1995, 29, 1892-1897. (18) Dolan, M. E.; McCarty, P. L. Environ. Sci. Technol. 1995, 29, 2741-2747. (19) Cox, E.; et al. Intrinsic biodegradation of trichloroethene and trichloroethane in a sequential anaerobic-aerobic aquifer. Intrinsic Bioremediation; Hinchee, R. E., Wilson, J. T., Downey, D. C., Eds.; Battelle Press: Columbus, OH;,1995; pp 223-232. (20) Edwards, E. A.; Cox, E. E. Field and laboratory studies of sequential anaerobic-aerobic chlorinated solvent biodegradation. In Situ and On-Site Bioremediation; Alleman, B. C., Leeson, A., Eds.; Battelle Press: Columbus, OH, 1997; Vol. 3, pp 261265. (21) Hartmans, S.; de Bont, J.; Tramper, J.; Luyben, K. Biotechnol. Lett. 1985, 7, 383-388. (22) Hartmans, S.; de Bont, J. A. M. Appl. Environ. Microbiol. 1992, 58, 1220-1226. (23) Verce, M. F.; Ulrich, R. L.; Freedman, D. L. Appl. Environ. Microbiol. 2000, 66, 3535-3542. (24) Hartmans, S.; Kaptein, A.; Tramper, J.; de Bont, J. A. M. Appl. Microbiol. Biotechnol. 1992, 37, 796-801. (25) Freedman, D. L.; Gossett, J. M. Appl. Environ. Microbiol. 1989, 55, 2144-2151. (26) Gossett, J. M. Environ. Sci. Technol. 1987, 21, 202-208. (27) Freedman, D. L.; Chennupati, S.; Kim, B. Temperature variability of Henry’s Constants and the effect on VOC emissions from Wastewater Treatment Plants; Proceeding of the 66th Annual Water Environment Federation Meeting, Anaheim, CA, 1993.

(28) American Public Health Association. Standard Methods for the Examination of Water and Wastewater, 18th ed.; APHA, AWWA, and WPCF: 1989. (29) Broholm, K.; Christensen, T. H.; Jensen, B. K. Water Res. 1992, 26, 1177-1185. (30) Sae´z, P. B.; Rittmann, B. E. Biodegradation 1993, 4, 3-21. (31) Chang, H.-L.; Alvarez-Cohen, L. Environ. Sci. Technol. 1995, 29, 2357-2367. (32) Chang, W.; Criddle, C. S. Biotechnol. Bioeng. 1997, 56, 492-501. (33) Ely, R. L.; Williamson, K. J.; Guenther, R. B.; Hyman, M. R.; Arp, D. J. Biotechnol. Bioeng. 1995, 46, 218-231. (34) Smatlak, C. R.; Gossett, J. M.; Zinder, S. H. Environ. Sci. Technol. 1996, 30, 2850-2858. (35) Reichert, P. Concepts Underlying a Computer Program for the Identification and Simulation of Aquatic Systems; Report No. CH-8600; Schriftenreihe der EAWAG, Swiss Federal Institute for Environmental Science and Technology (EAWAG): Du ¨ bendorf, Switzerland, 1994. (36) Neter, J.; Kutner, M. H.; Nachtsheim, C. J.; Wasserman, W. Appl. Linear Statistical Models; Irwin: Chicago, 1996. (37) Henry, S. M.; Grbic´-Galic´, D. TCE transformation by mixed and pure groundwater cultures, In Situ Aquifer Restoration of Chlorinated Aliphatics by Methanotrophic Bacteria; EPA Technical Report EPA/600/2-89/033; R. S. Kerr Environmental Research Laboratory: 1989; pp 109-125. (38) de Bont, J. A. M. Antonie van Leeuwenhoek 1976, 42, 59-71.

(39) Habets-Crutzen, A. Q. H.; Brink, L. E. S.; van Ginkel, C. G.; de Bont, J. A. M.; Tramper, J. Appl. Microbiol. Biotechnol. 1984, 20, 245-250. (40) van Ginkel, C.; de Bont, J. Archiv. Microbiol. 1986, 145, 403407. (41) Grady, C. P. L. J.; Smets, B. F.; Barbeau, D. S. Water Res. 1996, 30, 742-748. (42) Anderson, J. E.; McCarty, P. L. Appl. Environ. Microbiol. 1997, 63, 687-693. (43) Beck, J. V.; Arnold, K. J. Parameter Estimation in Engineering and Science; Wiley and Sons: New York, 1977. (44) Robinson, J. A. Adv. Microb. Ecol. 1985, 8, 61-114. (45) Smith, L. H.; Kitanidis, P. K.; McCarty, P. L. Biotechnol. Bioeng. 1997, 53, 320-331. (46) Chang, H.-L.; Alvarez-Cohen, L. Biotechnol. Bioeng. 1995, 45, 440-449. (47) Hartmans, S.; Weber, F. J.; Somhorst, D. P. M.; de Bont, J. A. M. J. Gen. Microbiol. 1991, 137, 2555-2560. (48) Hartmans, S.; de Bont, J. A. M.; Harder, W. FEMS Microbiol. Rev. 1989, 63, 235-264.

Received for review December 29, 2000. Revised manuscript received July 30, 2001. Accepted August 6, 2001. ES002064F

VOL. 35, NO. 21, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4251