Initially, we focused on enrichment experiments with 1,2-DCA as the sole electron acceptor (besides CO2 present in these media). Therefore, a bacterial inoculum was transferred from soil that had been polluted with 1,2-DCA since 1970. After nine transfers with H2 as the electron donor and acetate as the carbon source, an unusually high 1,2-DCA dechlorination rate of 1.3 mM/day was achieved in the absence of methanogenesis and acetogenesis (no CO2 consumption). Stoichiometric amounts of ethene and HCl were produced.
We started isolation on agar (3%, wt/wt) media in roll tubes containing H2 as the electron donor and 1,2-DCA as the electron acceptor. About 3% of all transfers of single colonies to liquid media containing H2, 1,2-DCA, and 0.02% (wt/wt) YE resulted in dechlorination activity. Surprisingly, these dechlorinating cultures contained two different types of bacteria: a coccus and a curved rod. The coccus could be isolated from this bibacterial culture and was identified as an Enterococcus casseliflavus strain. This organism showed no dechlorination capacity. Hence, it was assumed that the curved rod was the dechlorinating organism.
A liquid medium dilution series of the bibacterial culture suggested a synergistic interaction between E. casseliflavus and the curved rod. Growth and dechlorination were observed only in dilution cultures containing both types of bacteria. Indeed, media that only contained the curved rod, which was visible under the microscope, did not support growth unless 50% (by volume) sterile filtered supernatant of anaerobically grown E. casseliflavus was added to these cultures. This addition resulted in growth and dechlorination activity, hence allowing isolation and physiological characterization of the curved rod, designated strain DCA1. No contaminating bacteria could be detected after cultivation in roll tubes containing 0.1% YE, 40 mM pyruvate, or 40 mM glucose or after aerobic cultivation on brain heart infusion agar plates.
Formate could replace H2 as an electron donor in the presence of acetate. However, acetate alone did not support 1,2-DCA dechlorination, indicating that it exclusively served as a carbon source. In addition, lactate could be used as electron donor and carbon source, making the addition of acetate unnecessary. Inorganic electron acceptors that sustained growth of strain DCA1 were SO32−, S2O32−, and NO3−. No growth was observed with the electron acceptors SO42−, NO2−, CO2, and fumarate. The supernatant of the E. casseliflavus culture was strictly required for growth of strain DCA1 with any electron acceptor but could not be replaced by YE, amino acids, 100 μM Fe2+, water-soluble vitamins, or volatile fatty acids.
To identify the unknown excretion product, other
Enterococcus strains such as
E. casseliflavus LMG 12901,
Enterococcus faecalis LMG 11207, and
Enterococcus avium LMG 10774 were additionally tested. Interestingly, the supernatant of the first two of these strains allowed growth of strain DCA1, while that of the
E. avium strain did not. It has been reported that both
E. casseliflavus and
E. faecalis produce menaquinones, in contrast to
E. avium (
5). In our tests, indeed, 1 μM vitamin K
1 or vitamin K
2 could replace the
Enterococcus supernatant and supported growth and dechlorination. In addition, a defined mixture of amino acids could replace the YE. Pure cultures of strain DCA1 could be subcultured indefinitely in these completely defined growth media with S
2O
32− or 1,2-DCA as the sole electron acceptor. It would be interesting to investigate if menaquinones are also essential for growing other dechlorinators, such as
Dehalococcoides and
Dehalobacter strains that are difficult to cultivate (
13,
14,
16,
22,
23).
When we added 1,2-DCA, acetate, and 0.01% YE to media containing 1 μM vitamin K2, growth and dechlorination started but rapidly ceased (Fig. ). The addition of an electron donor, such as H2, formate, or lactate, to these media sustained growth and supported dechlorination of up to 6 mM (cumulative) 1,2-DCA. YE was not required for growth and dechlorination but stimulated both and could be replaced by amino acids. Since H2 oxidation and 1,2-DCA reduction cannot be coupled to substrate level phosphorylation, the utilization of these substrates as sole energy sources in defined media indicates energy conservation via dehalorespiration in strain DCA1. The 1,2-DCA dechlorination rate exceeded 350 nmol of chloride released per min per mg of total bacterial protein, while the protein yield was 0.41 ± 0.08 g (mean ± standard deviation; n = 12 cultures) of total bacterial protein per mol of chloride released.
In addition to 1,2-DCA, a defined range of other chlorinated alkanes sustained growth of strain DCA1, including 1,2-D, 1,2-DCB, d-2,3-DCB, l-2,3-DCB, and meso-2,3-DCB. All these substrates (400 μM) were completely dechlorinated to the corresponding alkenes with at least 99.9% molar selectivity. Final concentrations of the substrates dropped below 2 μM. The toxicity limit of 1,2-DCA was found to be 1.5 mM during exponential growth. After six consecutive 5% (by volume) transfers on media containing S2O32− (10 mM) as the sole electron acceptor, strain DCA1 immediately started to dechlorinate vicinal dichloroalkanes when transferred to media containing these compounds. Since preculturing strain DCA1 with S2O32− or dichloroalkanes as the sole utilizable electron source did not influence its subsequent dechlorination behavior, these experiments indicate a constitutive dechlorination activity.
VC and monochloroethane were not used as electron acceptors, nor were they formed during 1,2-DCA dechlorination, indicating that neither of the compounds is a free intermediate of the process. In that sense, strain DCA1 is different from all dehalorespiring bacteria currently isolated because it neither dechlorinates any unsaturated chlorohydrocarbons nor carries out any reductive hydrogenolysis; strain DCA1 exclusively dichloroeliminates its saturated chlorosubstrates. This highly selective, fast, and energy-conserving dichloroelimination has not yet been described for a bacterial isolate in defined media, which suggests that the dichloroelimination mediated by strain DCA1 occurs via a novel type of biochemical reaction mechanism. The defined growth conditions of strain DCA1 and further research on its dehalogenase may allow the characterization of the currently unexplored respiratory dichloroelimination biochemistry.
More highly chlorinated alkanes such as hexa-, penta-, and tetrachloroethanes were not dichloroeliminated. However, 1,1,2-trichloroethane still supported growth producing VC, illustrating the exclusive dichloroelimination mechanism and suggesting that the catalytic center of the dehalogenase might be hindered when more than three chlorine substituents are present on both vicinal carbon atoms. Consistent with the exclusivity for dichloroeliminations, no chlorinated methanes, monochloroalkanes, or nonvicinal dichloroalkanes were dechlorinated. This illustrates a high substrate specialization, which is typical for most dehalorespiring bacteria. All substrates were tested at nontoxic levels. To ensure this lack of toxicity, cultures that did not start dechlorination of the substrates after 14 days were additionally supplemented with 400 μM 1,2-DCA. In all cases, direct conversion of 1,2-DCA and growth with this substrate occurred.
Phase-contrast light microscopy and scanning electron microscopy of the motile gram-positive cells of strain DCA1 revealed curved rods 0.5 to 0.7 μm in diameter and 2 to 5 μm in length. Occasionally, cell lengths exceeding 10 μm were observed. Optimal pH and temperature were 7.2 to 7.8 and 25 to 30°C, respectively. Reductive dechlorination occurred between 12 and 32°C.
Comparative analysis of the almost complete 16S rDNA sequence of strain DCA1 with the sequences currently available from the EMBL database revealed a 96.6% similarity to the closest relative,
Desulfitobacterium hafniense strain DCB-2 (DSM 10664; type strain). Physiologically, strain DCA1 is clearly different from other
Desulfitobacterium species, since chloroethenes and chlorophenols were not converted, while no other
Desulfitobacterium species has been reported to mediate dichloroelimination of vicinal dichloroalkanes. One isolate,
Desulfitobacterium sp. strain Y51, is able to partially dechlorinate hexa-, penta-, and tetrachloroethanes, in addition to partial polychloroethane dechlorination (
26). This anaerobe uses both reductive hydrogenolysis and dichloroelimination reactions, which are slower and which produce chlorinated end products (dichloroethenes). Strain Y51 and strain DCA1 do not have any chlorosubstrate in common and show a 16S rDNA similarity of 96.0%. Strain DCA1 has been deposited in the BCCM/LMG Bacteria Collection (accession no.
P21439) and has been proposed as the type strain of the new species
Desulfitobacterium dichloroeliminans.
Strain DCA1 is not extremely oxygen sensitive, surviving aerobic conditions for at least 24 h, even though anoxic conditions (lower than −180 mV) were essential for 1,2-DCA dechlorination. Vitamin B
12 (40 nM) was required as a growth supplement for significant dechlorination activity of strain DCA1. Propyl iodide (50 μM) completely inhibited dechlorination in the dark; the inhibition could be reversed by illumination. These findings provide strong evidence for the involvement of a corrinoid in the dichloroelimination reaction (
19,
20).
In vivo studies with three different 2,3-DCB stereoisomers as electron acceptors were performed (Fig. ). The
meso stereoisomer was selectively converted to
E-2-butene, while both
d and
l enantiomers produced
Z-2-butene. Interestingly, in the most stable staggered conformational isomers (conformers) of the
meso compound and both
dl compounds, both chlorine atoms are opposite to each other (transantiparallel position). Hence, the product pattern indicates a stereoselective
anti dichloroelimination. The less-stable eclipsed
meso and
dl conformers, rarely observed below 303 K (30°C), would require a
syn dichloroelimination on vicinal adjacent chlorine atoms, resulting in the opposite product formation (except for the highest-energy eclipsed
dl conformers). Moreover, the high dechlorination rate, comparable with those for dehalogenases catalyzing reductive hydrogenolysis reactions, suggests that the most probable conformer is the main substrate of the dehalogenase of strain DCA1. To our knowledge, so far no stereoselective
anti dichloroelimination of vicinal dichloroalkanes in mild aqueous conditions at ambient temperatures has been reported (
3,
4,
6,
10,
24).
The exclusive
anti dichloroelimination of strain DCA1 makes its dechlorination biochemistry distinct from that of all other known dehalorespiring isolates. The evolution of the dehalogenase catalyzing this reaction is enigmatic. Natural production of trace amounts of strain DCA1's chlorosubstrates is not likely to provide sufficient selective evolutionary pressure in a geological time frame (
7). In contrast, the observation that the polyvinyl chloride production intermediate 1,2-DCA is the anthropogenic chlorinated C
2 solvent most massively released to groundwater during the last decades rather suggests a rapid dehalogenase development in response to environmental contamination.
The unraveled nutritional requirements of strain DCA1 allowed cultivation of the organism at 100-liter scale in pure culture. In situ pilot tests showed that the addition of 1 g (dry weight) of cells of strain DCA1 to 1 m3 of groundwater containing 1,2-DCA resulted in a complete detoxification within 1 week. Hence, dichloroelimination of dehalorespiring bacteria can be considered a highly efficient and unique remediation tool.
We presented the isolation, characterization, and growth medium definition of a dehalorespiring bacterium that selectively and completely dichloroeliminates several vicinal dichlorinated alkanes, including 1,2-DCA and 1,2-D. Besides environmental and stereoselective fine-chemical applications, this discovery may contribute to the understanding of respiratory dichloroelimination biochemistry and evolutionary dehalorespiration. The methods described in this report and the key role of menaquinone in the metabolism of strain DCA1 may help to isolate other dehalorespirers that perform a complete dechlorination.