Competitive Inhibition among Chlorinated Ethylenes. Competitive inhibition among chlorinated ethylenes was included in the modeling analysis (9, 10). The reductive dechlorination rate of a CAH was modeled with competitive inhibition kinetics using eq 2

Based on the results of the inhibition experiments, a conceptual inhibition model was developed (Figure 2). The equations for reductive dechlorination by the anaerobic cultures including appropriate inhibition terms are presented in eqs 3-6. The proposed inhibition patterns (Figure 2) and resulting equations assume PCE inhibits TCE transformation, TCE inhibits both c-DCE and VC transformation, and c-DCE inhibits VC transformation. As the inhibition results will show, the less chlorinated ethylene inhibition on more chlorinated ethylenes was very weak and therefore was not included in the model equations.

Figure 2 Proposed inhibition model between chlorinated ethylenes for reductive dechlorination based on the inhibition studies of Figures 4-8. Arrows represent inhibition directions. Dashed arrows indicate very weak inhibition.

Equations 3-6 were combined to generate the series of rate equations that include transformation, production, and partitioning, as shown in eq 1.

Kinetic Model for Biomass Growth. Since VC inhibition experiments lasted several days, the potential growth of culture was included in the model using eq 7. Microbial growth was calculated with respect to the transformation rate of each chlorinated ethylene. The PM culture was assumed to grow on PCE, TCE, and c-DCE. VC transformation to ETH was assumed to be cometabolic, since the measured K_S value for VC transformation of 602 mol/L was in the range of that reported for Dehalococcoides ethenogenes strain 195, which transforms VC cometabolically. Thus, no biomass growth was coupled with kinetic modeling (25). Unlike the PM culture, microbial growth on each dechlorination step was considered for the EV culture (13).

In eq 7, Y is the growth yield of dechlorinating microorganisms, dC/dt is the rate of concentration change of each chlorinated ethylene over time (mol/L/day), and k_d is the decay constant of the dechlorinating microorganisms (day^-1). Values for growth yield, Y (= 0.006 mg of protein/mol of Cl^- dechlorinated), and decay constant, k_d (= 0.024 day^-1), were obtained from the literature (25, 21). For our tests, we do not know the actual biomass, X, of the dechlorinating microorganisms, and whether different dechlorinators are active at different steps of the process. Since k_max values were determined on a total protein basis, the values were normalized to the total mixed culture biomass. The simultaneous solution of the model equations was obtained using STELLA Research 5.0 (High Performance Systems, Lebanon, NH).

Results

Determinations of Kinetic Parameters. The Michaelis-Menten curves for PCE, TCE, c-DCE, and VC transformation by the PM culture determined using the multiequilibration kinetic method are shown in Figure 3. The triplicate determinations show statistically reliable results were obtained for all the chlorinated ethylenes tested. The Michaelis-Menten curves in Figure 3 show good fits for all chlorinated ethylenes. Table 1 presents k_max and K_S values for the PM and EV cultures for each step of the dechlorination process. The standard deviation of k_max and K_S based on triplicate determinations were approximately 11 and 16% of the parameter value, respectively. For comparison, Haston and McCarty (22) reported standard deviations in k_max and K_S of 17% and 60% of the parameter values, respectively, using single equilibrations in multiple reactors.

Figure 3 Initial reductive dechlorination rates of CAHs by the PM culture. The simple multi equilibration method was used to determine kmax and KS values. The results of replicate experiments are represented by the different symbols.

Table 1 also presents the k_max and K_S values obtained by Fennell and Gossett (21) for a mixed culture containing Dehalococcoides ethenogenes strain 195 and those obtained by Haston and McCarty (22), for comparison purposes. The PM and EV cultures have comparable k_max and K_S values for PCE and TCE (Table 1). For both cultures, the TCE k_max values are a factor of 10 greater than those for PCE. The K_S value for PCE of the PM culture is a factor of 2 greater than the EV culture, while the k_max values are essentially equal, indicating that the EV culture transforms PCE more rapidly at concentrations below the K_S values (1.6 M). The k_max and K_S values for c-DCE and VC show significant differences between the two cultures. The k_max for c-DCE of the EV culture was 40% lower than that of the PM culture, while the K_S values were almost the same, indicating slower c-DCE biotransformation by the EV culture.

The k_max value for VC by the EV culture was a factor of 3 greater, while the K_S was a factor of 10 lower than the PM culture, indicating much more rapid VC transformation with the EV culture. This result was consistent with the assumption of potential growth of the EV culture using VC as an electron acceptor as discussed by Pon and et al. (13). Recently, VC-grown dechlorinating cultures (Dehalococcoides sp. strain BAV1 and Dehalococcoides sp. strain VS) were isolated [He et al. (18, 26) and Cupples et al. (19)]. The K_S value (62.6 M) of the EV culture is closer to those of the VC-grown cultures (5.8 M and 2.6 M, respectively) than that of Dehalococcoides ethenogenes strain 195 (290 M) (21), which cometabolically transforms VC. The slower maximum transformation rate and higher K_S value (602 M) for VC with the PM culture are consistent with cometabolic VC transforma tion (21, 25). The transformation characteristics of the PM culture are similar to that of Dehalococcoides ethenogenes strain 195 that has a high K_S value (21).

Competitive Inhibition among Chlorinated Ethylenes. Results of inhibition studies are shown in Figures 4-8. The lines shown in each figure represent model simulations using the series of equations previously described. The model used the independently determined kinetic parameters listed in Table 1 with the initial biomass determined by protein analysis. The measured K_S values were used as putative inhibition constants. Figure 4(a),(b) presents PCE inhibition on TCE transformation. As PCE concentrations increased, rates of c-DCE production decreased for both cultures, indicating that TCE dechlorination was inhibited by PCE. The EV culture was more strongly inhibited by PCE than the PM culture. A PCE concentration approximately 2.5 times higher than that used for the EV culture was required to obtain similar inhibition of the PM culture. The effect of TCE on PCE dechlorination was also evaluated (data not shown). PCE dechlorination was not clearly inhibited by TCE, suggesting there is very weak inhibition of TCE on PCE dechlorination.

Inhibition of TCE on c-DCE dechlorination is presented in Figure 4(c),(d). TCE inhibition of c-DCE transformation was observed with both cultures, with the EV culture being more strongly inhibited. The K_S value for TCE of the EV culture was a factor of 1.5 lower than the PM culture, which was consistent with the greater TCE inhibition with the EV culture. Simulations based on the inhibition model given in eq 5 showed good agreement with the inhibition experimental results. TCE inhibition on VC transformation was also examined, as shown in Figure 4(e),(f) for the PM and EV cultures, respectively. TCE strongly inhibited VC transforma tion to ETH with both cultures. The EV culture more rapidly transformed VC (by a factor of 10) than the PM culture (note the different y axis scales), even though similar protein concentrations were used. This result is consistent with the EV's k_max value being a factor of 3 higher and the K_S value being a factor of approximately 10 lower compared to the PM culture (Table 1). The strong inhibition of TCE on VC transformation is consistent with the much lower K_S value for TCE compared to the K_S values of VC for both the PM and EV cultures. The modeling results using eqs 4-6 provide excellent matches to the experimental data using the independently derived kinetic values given in Table 1, and with the K_S values representing the inhibition constants. Neither culture showed clear VC inhibition on TCE dechlorination (data not shown), which was consistent with the findings of Nielsen and Keasling (27).

VC dechlorination rates to ETH in the presence of c-DCE were measured for a longer time compared to other inhibition experiments, due to the slower rates of VC transformation (Figure 5). The initial protein concentration of the EV culture for this test was a factor of 24 greater than the PM culture. This resulted in a much shorter time for c-DCE and VC transformation, as predicted by the model simulations. ETH production rates for the PM culture increased (in bottles amended with c-DCE), after all c-DCE was essentially transformed to VC (Figure 5(a)-(c)). After complete transformation of c-DCE to VC by the PM culture, ETH production rates were greater in c-DCE amended reactors. The addition and transformation of c-DCE resulted in higher VC concentrations which led to greater rates of ETH production, as was confirmed in model simulations. VC transformations were very well predicted based on the proposed inhibition model. The VC concentrations used here were well below the K_S value for the PM culture. Thus, an increase in VC concentra tion would be expected to result in faster VC transformation, as was observed in the experimental results and model simulations. The simulated increase in biomass concentra tions over the several day-period of the test are presented in Figure 5(d). A greater relative increase in biomass was predicted for the PM culture, which would result in increased rates of ETH production.

	Figure 4 Competitive inhibition studies between CAHs. Lines represent the model simulations. Average initial cell concentrations measured and used in model simulations are 1.5-2.0 (a), 1.3-1.4 (b), 2.0-2.5 (c), 3.9-4.2 (d), 2.2-2.5 (e), and 1.4-1.6 mg/L (f).
	Figure 5 Competitive inhibition studies of c-DCE on VC dechlorination by the PM culture (a, b, c, d) and the EV culture (e, f, g, h). Lines represent model simulations. Cell concentrations from model simulations are shown in (d) and (h).

The inhibition of c-DCE on VC dechlorination by the EV culture is shown in Figure 5(e)-(h). Over the time scale of these tests, the simulated increase in cell mass was less than 3% (Figure 5(h)). Although c-DCE was added as an inhibitor, c-DCE was much more rapidly transformed than VC resulting in transient increases of VC concentrations. Figure 5(e)-(h) shows that VC transformation to ETH was increasingly inhibited as the initial aqueous c-DCE concentration was increased. The lags in ETH production are clearly associated with higher concentrations of c-DCE, with ETH production initiated when c-DCE concentrations were reduced to low values. Unlike the PM culture, after complete transformation of c-DCE to VC, the rate of VC dechlorination to ETH was similar in all cases. Due to the lower K_S value for VC with the EV culture, the increase in VC concentrations had little effect on transformation rates, since the initial VC concentration (290 M) was significantly higher than the K_S value (62.6 M). Also, the higher initial cell mass and short duration of the tests resulted in minimal biomass growth, as predicted by the model simulations. The model simulations did a very good job capturing the temporal changes in c-DCE, VC, and ETH concentrations.

Experiments were performed with both cultures to study PCE inhibition on c-DCE dechlorination (Figure 6). Although the PCE concentration was increased up to 395 M and 491 M for the PM and EV cultures, respectively, essentially no difference was observed in the rate of c-DCE transformation to VC. Simulations were performed with and without PCE inhibition on c-DCE transformation and compared with the experimental results. Simulations without PCE inhibition more closely matched experimental data than those with PCE inhibition, where inhibition was represented by the measured K_S value. The experimental results and model simulations indicate that reductive c-DCE dechlorination is unaffected or very weakly inhibited by PCE. c-DCE dechlorination rates were a factor of about 7 higher with the PM culture, due to the higher initial protein content used, and the higher k_max value of the culture.

Figure 6 Comparison of simulations with and without PCE inhibition and the experimental data for c-DCE dechlorination by the PM culture (a and b) and the EV culture (c and d). Initial c-DCE concentrations were 124 and 107 M with cell concentrations of 5.0 and 1.3 mg/L for the PM and EV cultures, respectively.

Potential VC inhibition on c-DCE dechlorination was also evaluated. Model simulations were performed with and without VC inhibition and were compared to the experimental results (Figure 7). For both cultures, the experimental observations showed no apparent VC inhibition on c-DCE dechlorination, and there was also no clear difference between the different model simulations. The results show that VC had a very weak inhibitory effect on c-DCE transformation for both cultures, even at VC concentrations as high as 290 M, approximately 5 times the K_S value for the EV culture. This result is consistent with the high VC K_S values of the PM and EV cultures of 602 and 62.6 M, respectively, and the low K_S values for c-DCE of 1.9 and 1.8 M. Thus, VC inhibition on c-DCE transformation was shown to be very weak in our proposed model (Figure 2).

Figure 7 Comparison of simulations with and without VC inhibition and the experimental data for c-DCE dechlorination by the PM culture (a and b) and the EV culture (c and d). Initial c-DCE concentrations were 96 and 121 M with cell concentrations of 1.6 and 29 mg/L for the PM and EV cultures, respectively.

Batch kinetic experiments were performed to study c-DCE inhibition on TCE dechlorination (Figure 8). Interestingly, although the K_S values for c-DCE were similar to those of TCE, rates of TCE dechlorination showed no clear inhibition by c-DCE. The model without c-DCE inhibition simulated the experimental data much more closely than that with inhibition. The differences in the simulations for 0 M initial c-DCE with and without inhibition results from c-DCE formed upon TCE transformation. Some inhibition was observed at the highest c-DCE concentration tested, but it was very weak. These results also indicate that substances inhibitory to TCE transformation were not present in the biologically produced c-DCE. These results show that K_S values, when used as inhibition constants, are not representative of less chlorinated ethylene inhibition on more chlorinated ethylene transformation.

Figure 8 Comparison of simulations with and without c-DCE inhibition and the experimental data for TCE dechlorination by the PM culture (a and b) and the EV culture (c and d). Initial TCE concentrations were 21.8 and 21.9 M with cell concentrations of 12 and 10.3 mg/L for the PM and EV cultures, respectively.

Discussion

Several pairs of chlorinated ethylenes (PCE TCE, PCE c-DCE, TCE c-DCE, TCE VC, and c-DCE VC) were investigated for transformation inhibition. The experimental and modeling results indicate that more chlorinated ethylenes inhibited reductive dechlorination of the less chlorinated products, except that PCE did not inhibit c-DCE dechlorination. Less chlorinated ethylenes, however, very weakly or did not inhibit the transformation of more chlorinated ethylenes. The K_S values for VC for the PM and EV cultures were much higher (602 M and 62.6 M, respectively) than the K_S values for the parent compounds. Thus, if inhibition were related to the K_S values, weak inhibition would be expected, as shown in Figure 7 for VC inhibition on c-DCE transformation. However, very weak inhibition of c-DCE on TCE transformation was also observed, contrary to model simulations using K_S values as inhibition constants (Figure 8). Model simulations with inhibition constants much higher than their K_S values would be required to simulate the less chlorinated ethylene inhibition on the more chlorinated ethylene. Cupples et al. (11) reported that VC competitively inhibited c-DCE dechlorination and that the VC inhibition constant was three times higher than the K_S value for VC. The c-DCE inhibition constant on VC transformation, however, was equal to the K_S value for c-DCE. Haston (9) used K_S values for c-DCE inhibition on VC transformation when modeling the results of laboratory studies. In a recent study, Cupples et al. (11) determined a competitive inhibition constant (K_I) for c-DCE inhibition on VC transformation. The K_I value generated for c-DCE was not statistically different than the K_S value of c-DCE (K_S = 3.3 ± 2.2, K_I = 3.6 ± 1.1). Our results were consistent with these studies.

PCE did not inhibit c-DCE transformation. This likely resulted from different microorganisms in the culture or different enzyme systems being responsible for PCE and c-DCE and VC transformation. Reductive dechlorination of c-DCE and VC has been reported to be promoted by different dechlorinators than those capable of reductive dechlorination of PCE and TCE (28). The growth of Dehalococcoides sp. strain VS and Dehalococcoides sp. strain BAV1 are coupled to the dechlorination of c-DCE and VC but not PCE and TCE (18, 19, 26). Dehalococcoides sp. strain FL2 that cometabolically dechlorinates PCE and VC uses only TCE and c-DCE as metabolic electron acceptors (29). Dehalospirillum multivorans, Dehalobacter restrictus strains PER-K23A, and Desulfuromonas sp. can reduce only PCE and TCE in terminal electron accepting processes (14, 30). Another incompletely dechlorinating bacterium (Desulfitobacterium sp. strain PCE1) reductively dechlorinates only PCE to TCE (31). The lack of PCE inhibition on c-DCE transformation is consistent with different microorganisms being responsible for PCE and c-DCE transformation. Different enzyme systems responsible for different dechlorination reactions in the same microorganism, as reported by Magnuson et al. (32) for Dehalococ coides ethenogenes, could also explain the lack of PCE inhibition on c-DCE transformation.

Inhibition of the more chlorinated ethylenes on the less chlorinated ethylenes was represented well using independently determined K_S values as the competitive inhibition constants in model simulations. Experimental results and model simulations both illustrate how low K_S values leading to efficient reductive dechlorination for each chlorinated ethylene can also cause strong inhibition on daughter product dechlorination. Thus, obtaining K_S values is important for modeling both the rates of transformation of parent compounds and their inhibition on the transformation of daughter products.

A simple multiple equilibration method using a single batch reactor was developed to obtain k_max and K_S values for two dechlorinating mixed cultures. The kinetic parameters measured here are comparable to the values reported by Fennell and Gossett (21) and Haston and McCarty (22). The multiple equilibration method provided precise kinetic values and was simpler to use than the single equilibration method in multiple reactors. Compared with values reported by Haston and McCarty (22), the standard deviations of the kinetic parameters reported here tended to be lower. Studies with direct comparisons of the different methods are needed. The multiple equilibration method might also be adapted to obtain inhibition data required to generate Lineweaver-Burk inhibition plots.

A limitation of the current study is that the actual concentrations of the dechlorinating cultures are not known. The reported k_max values are based on mixed culture protein concentrations. The initial protein concentration when used in model simulations yielded good matches to the experimental data collected over short time scales. The results do indicate that stable communities of dechlorinating microorganisms were maintained in the enrichment reactors over the course of these experiments, since protein based k_max values when used in the simulations fit the results of later inhibition experiments well. As described in Yu and Semprini (33), terminal restriction fragment length polymorphism (T-RFLP) analysis showed a predominant peak at 123.6 bp, consistent with the expected terminal fragment lengths (TFLs) for the Dehalococcoides sp. sequences obtained from GenBank. The peak corresponded to 65% of the amplicon DNA indicating that the culture is highly enriched in Dehalococ coides-like microorganisms. Molecular-based analyses, including clone libraries of the mixed cultures and real time PCR (18, 26), are currently being developed to better enumerate the dechlorinating populations in these cultures.

Acknowledgment

The research was supported by a research grant from the U.S. Environmental Protection Agency-sponsored Western Region Hazardous Substance Research Center under agree ment R-828772. This article has not been reviewed by the agency, and no official endorsement should be inferred.