Biochemical and Biophysical Research Communications 261, 562–566 (1999) Article ID bbrc.1999.1077, available online at http://www.idealibrary.com on
Identification of an Extracellular Catalyst of Carbon Tetrachloride Dehalogenation from Pseudomonas stutzeri Strain KC as Pyridine-2,6-bis(thiocarboxylate) Chang-Ho Lee, Thomas A. Lewis, Andrzej Paszczynski, and Ronald L. Crawford 1 Environmental Biotechnology Institute and Department of Microbiology, Molecular Biology, and Biochemistry, University of Idaho, Moscow, Idaho 83844-1052
Received May 25, 1999
Pseudomonas stutzeri strain KC was originally characterized as having, under iron-limiting conditions, novel carbon tetrachloride (CCl 4) dehalogenation activity, specifically, a net conversion of CCl 4 to CO 2. The exact pathway and reaction mechanisms are unknown, but chloroform is not an intermediate and thiophosgene and phosgene have been identified as intermediates in trapping experiments. Previous work by others using cell-free preparations has shown that cell-free culture supernatants that have been passed through a low-molecular-weight cutoff membrane can confer rapid CCl 4 transformation ability upon cultures of bacteria which otherwise show little or no reactivity toward CCl 4. We used a cell-free assay system to monitor the complete purification of compounds showing CCl 4 degradation activity elaborated by iron-limited cultures of strain KC. Electrospray tandem mass spectroscopy, NMR spectroscopy, and comparisons with synthetic material have identified pyridine-2,6-bis(thiocarboxylate) as a metabolite of strain KC which has CCl 4 transformation activity in the presence of chemical reductants, e.g., titanium[III] citrate or dithiothreitiol, or actively growing bacterial cultures. © 1999 Academic Press
Carbon tetrachloride (CCl 4) is a toxic, carcinogenic compound which has been used as a solvent and degreaser for various purposes, including the manufacture of nuclear weapons (1). Due to improper disposal, CCl 4 is present at many sites as a contaminant of groundwater, posing a potential threat to ecosystems and to human health. Biological transformations of CCl 4 have been studied for many years, both to understand the toxicology of CCl 4 and to identify methods for its destruction in the environment. It is not surprising 1 Corresponding author: Environmental Biotechnology Institute, University of Idaho, Moscow, ID 83844-1052. Fax: (208) 885-5741. E-mail:
[email protected].
0006-291X/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
that the types of transformation identified include reduction of CCl 4 as an initial step; CCl 4 is an oxidized species, the carbon atom having the same formal oxidation state as CO 2. Biological mediators of the reductive transformation of CCl 4 include the tetrapyrroles heme, factor F 430, and cobalamins (2–5). Products arising from reactions of CCl 4 and these cofactors can be explained by the various fates of the trichloromethyl radical (zCCl 3) generated by one-electron reduction of CCl 4, the product spectra depending largely on the chemical reductants used to regenerate the reduced form of the cofactor and their reactivity with trichloromethyl radical (6 – 8). The primary reaction route is the hydrogenolytic removal of chlorine substituents resulting from further reduction of the trichloromethyl radical and protonation of the carbanion (Fig. 1). This route of transformation yields the more reduced chloromethanes in the series chloroform (CHCl 3), dichloromethane, chloromethane, and ultimately methane when a potent reductant such titanium[III] citrate is used (3, 5, 7). The ability to sequentially remove all chlorine atoms of CCl 4 points to versatility in the dehalogenation shown by reduced tetrapyrroles. Activity towards the other chloromethanes, as well as fluorochloromethanes (9), halogenated ethenes (10, 11), and other chloroaliphatic compounds (12, 13), has been demonstrated. The tetrapyrrole cofactors mentioned above, and some obligately anaerobic bacteria and archea (14, 15, 16), all produce CHCl 3 and further dehalogenate it. In contrast, iron-limited cultures of Pseudomonas stutzeri strain KC transform CCl 4 to CO 2 and non-volatile products, while producing only trace amounts of CHCl 3 and showing no activity toward CHCl 3, chlorinated ethanes, or fluorotrichloromethane (17, 18). Though this points out a rather limited spectrum of dehalogenation activity, it also presents an advantage for CCl 4 remediation in that the potential for accumulation of the more reduced and thus more slowly transformed
562
Vol. 261, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Hydrogenolytic dechlorination pathway for carbon tetrachloride. Sequential two-electron reduction process.
CHCl 3 is minimized. We have analyzed products of the CCl 4 transformation by strain KC using radiotracer methodologies and trapping agents (19). These studies identified thiophosgene and phosgene as intermediates of the overall hydrolytic transformation. Our hypothesis for a reaction pathway to explain these results incorporated the prevailing notion of an initial reduction to produce trichloromethyl radical (20). This pathway then predicts the participation of a sulfur species, which reacts to give thiophosgene, as had been postulated for a mineral-based CCl 4 radical-substitution and hydrolysis (21). Alternatively, molecular oxygen may react with the radical to give phosgene (22) (Fig. 2). At the time this reaction pathway was hypothesized, the catalysts of this transformation had not yet been identified, so the sulfur species and reductant(s) were unknown, preventing a definitive characterization of the reaction mechanism. An incomplete understanding of the chemistry also limits practical application of the CCl 4 transformation activity, since potential products and interfering factors, as well as expected transformation capacities, cannot be accurately predicted. Toward this end, we have purified compounds elaborated by iron-limited cultures of strain KC which have CCl 4 transformation activity in the presence of chemical reducing agents or actively metabolizing bacterial cultures. Using bioassays that included actively metabolizing bacteria, Dybas et al. (23) had reported activity in preparations of low molecular weight compounds present in cell-free supernatants from strain KC cultures. We have tested for the presence of such activity using chemical reductants, assuming that the function of live cells was to furnish an appropriate source of reducing equivalents (20). Using this assay we have purified compounds whose CCl 4 transformation activity is consistent with that of live cultures of strain KC with regard to its substrate range and the physiological conditions of its production. Analytical spectroscopy and comparisons with synthetic material have confirmed its structure as pyridine-2,6-bis(thiocarboxylic acid).
MgSO 4, Ca(NO 3) 2, and CuCl 2 were added separately from sterile stock solutions to give 1 mM, 0.1 mM, and 5 nM final concentrations, respectively. Cultures were maintained on tryptic soy agar; single colonies were used to inoculate DRM broth, which was incubated at 30°C with shaking. Actively growing, 1-liter DRM cultures grown in 2-liter Erlenmeyer flasks were used to inoculate large-volume cultures (20 liters) grown in polyethylene carboys at ambient temperature (22–24°C) for 4 days with constant aeration by filtered (0.2-m pore size) compressed air supplied through a 4-mm (i.d.) stainless steel tube inserted to the bottom of the vessel. 13C- and 15N-enriched growth substrates were purchased from Cambridge Isotope Laboratories (Andover, MA). Measurement of CCl 4 dehalogenation activity. CCl 4 transformation activity was assayed by measuring the loss of CCl 4 in incubations of culture supernatant fractions with titanium[III]citrate. Assays and control reactions (volume, 1 ml) were prepared in 10-ml headspace autosampler vials in an anaerobic chamber (Coy, Ann Arbor, MI). Ti[III] citrate was used at concentrations of 0.5 or 2.5 mM. Approximately 30 mmoles CCl 4 was added to each vial with a 25-ml gas-tight syringe (Hamilton, Reno, NV). The reactions were stoppered with Teflon-faced butyl rubber stoppers (the West Co., Phoenixville, PA) immediately after addition of CCl 4 from a stock solution in methanol (0.06% vol/vol), and incubation was carried out at 22–24°C for 2 h to overnight with constant shaking at 200 rpm. Control incubations for determining background losses contained sterile growth medium or deionized water, with and without Ti[III] citrate. CCl 4 and CHCl 3 concentrations remaining in assays and control incubations were quantitated by GC-ECD (18, 20). Bioassays included cells of P. stutzeri ATCC strain 17588 grown in DRM for 48 h at 25°C. These were used by including compounds to be tested, or by resuspending cells in strain KC culture filtrate (filtered through a 0.2-mm pore-size filter) and incubating at 25°C with CCl 4. Strain 17588 showed no significant CCl 4 removal activity as compared to sterile medium. Purification of active compounds. Bacterial cells were removed from large-scale cultures by filtration through a hollow fiber cartridge (H5P100-43, 100 kDa cutoff) using an Amicon ultrafiltration system. The cell-free filtrate was adjusted to pH 2.0 with concentrated HCl and allowed to stand at 4°C overnight to allow formation of a precipitate. The precipitate was collected by centrifugation (9400 3 g, 20 min, 4°C; Beckman JA-10 rotor). After freeze-drying, this preparation gave a yellow-white powder which was redissolved in 1% ammonium hydroxide, and the remaining insoluble material was removed by filtration through a 30-mm-diameter, 0.45-mm (pore size) filter (Millex-25; Millipore, Bedford, MA). This solution was then applied to a 3 3 60 cm column of Bio-Gel P-2 (Bio-Rad Laboratories, Hercules, CA) previously equilibrated and then eluted with deionized water. Fractions were collected using a Frac-100 fraction collector (Pharmacia, Uppsala, Sweden). Those containing CCl 4 transformation activity were pooled and concentrated by freezedrying. Purity was assessed by HPLC using a Columbus C 8 column (15 3 0.46 cm; Phenomenex, Torrance, CA) and 0.2% formic acid:
MATERIALS AND METHODS Bacterial strains and culture conditions. Pseudomonas stutzeri strain KC was originally obtained from C. Criddle (Stanford University). The control strain incapable of CCl 4 dehalogenation was P. stutzeri ATCC 17588. DRM growth medium was used to cultivate strain KC for production of active compounds. The medium, which consists of (per liter) 6 g K 2HPO 4, 2 g Na acetate, 1 g NH 4Cl, and 0.5 g NaNO 3, was adjusted to pH 7.7–7.9 and sterilized by autoclaving.
FIG. 2. Pathways of reductive transformation of carbon tetrachloride involving trichloromethyl radical. Radical-mediated dehalogenation process.
563
Vol. 261, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 3. Electrospray negative MS/MS of biologically produced pyridine-2,6-bis(thiocarboxylate). (A) 12C/ 14N daughters of native 198 Da compound (199-proton); argon used as collision gas for secondary ionization, (B) 13C/ 14N daughters of 205 Da compound (206-proton); proof of seven carbon atoms in structure, (C) 13C/ 14N daughters of 199 Da compound (200-proton); proof of one nitrogen atom in structure. acetonitrile (85:15) at a flow rate of 1 ml/min. A Hewlett Packard 1090 system (Hewlett Packard, Avondale, PA) equipped with a diode array detector and monitoring at 260 and 340 nm was used. Purity was also confirmed by capillary electrophoresis using a Bio-Rad BioFocus 3000 capillary electrophoresis system equipped with a 75 cm 3 75 mm quartz capillary run in 25 mM borate buffer (pH 8.5) containing 25 mM SDS at 25 kV, with detection (absorbance at 270 nm) at the cathode end. Mass spectroscopy. Electrospray ionization tandem MS was performed using a Micromass Quattro II instrument operated in the negative ion detection mode. Samples were directly injected in methanol:water (1:1 by volume) solution by a syringe pump at a rate of 0.3 ml/h. Electron-impact ionization spectra were obtained using a Hewlett Packard 5989A spectrometer. Samples were delivered to the ion source in methanol at 0.4 ml/min using a Hewlett Packard 59980B particle beam interface at 70°C. NMR spectroscopy. 1H FT-NMR spectra were obtained using a Bruker Avance DRX-50 spectrometer operated at 500.13 MHz. 13CFT-NMR spectra were obtained with the same instrument operated at 125.77 MHz. Spectra were obtained in D 2O or DMSO-D 6. All NMR spectra were referenced to internal H 2O and/or DMSO shifts. IR spectroscopy. IR spectra were obtained in KBr pellets on a Perkin-Elmer 1600 series FTIR.
RESULTS Using a chemical assay of CCl 4 transformation activity (20), we monitored purification of compounds
produced by iron-limited cultures of strain KC. Initial experiments determined that activity was reversibly inhibited by acidification to pH 2. Acidification produced a clouding of the culture supernatant; if the clouding was cleared by centrifugation, activity was lost even when the pH was returned to neutral. The precipitate formed at pH 2 was found to contain the activity when dissolved in 1% ammonium hydroxide, so the active compound could be purified from the bulk medium components, which remained soluble at pH 2. Material collected by this procedure from large-volume cultures was further purified by size-exclusion chromatography on a Bio-Gel P2 column with deionized water used as the eluent. Activity was found in two discrete peaks. The earlier and more active peak, with UV absorbance maxima at 270 and 335 nm, was designated the A fraction. The later eluting peak, with less activity per mg and a UV absorbance maximum at 280 nm, was designated the L fraction. Samples of the A fraction were determined to be $99% pure by reversed phase HPLC and capillary electrophoresis. Electrospray tandem MS using negative ion detection mode gave an apparent molecular weight of 198 Da (199-H) 2 for the A fraction. The molecular ion
564
Vol. 261, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS TABLE 1
Analytical Data for Pyridine-2,6-bis(thiocarboxylic Acid) Isolated from Pseudomonas stutzeri KC 13
C and 1H NMR absorption of ammonium salt in DMSO d 6
Position
Carbon shift (ppm)
Proton shift (ppm)
C2, C6 C3, C5 C4 C7, C8
148.595 125.016 145.225 194.43
8.42 (d) 8.63 (t)
showed a propensity to lose fragments totaling 60 and 120 Da (Fig. 3A). To determine the number of carbon and nitrogen atoms present in the molecular ion and lost fragments, preparations were made from cultures of strain KC grown in media containing either 1,2- 13C acetate or 15NH 4SO 4. These gave molecular ions corresponding to the incorporation of seven carbon atoms and one nitrogen atom from the substrates (Fig. 3B, C). 13 C- and 1H FT-NMR gave resonances in the far upfield regions (Table 1). FTIR in KBr gave absorbance maxima at 3085 (CH) 2550 (SH) 1606 (CO) and 850 (CS). Particle beam electron-impact MS operated in positive ion detection mode gave the following m/z (%): 199 (5.56); [M] 1, 166 (28.81); [M- •SH] 1, 139 (100); [M-COS] 1, 138(36.20); [M- •COSH] 1, 110(41.52); [138CO] 1, 105(73.77); [138- •SH] 1, 79(1.49); [ •pyridine] 1, 78(31.28); [ •pyridine-H] 1, 77(58.57); [ •pyridine-2H]. The spectroscopic data were consistent with a symmetrically substituted pyridine with two thiocarboxylate groups in 2,6 position, accounting for the loss of neutral fragments of m/z 60 and in good agreement with data for pyridine-2,6-bis(thiocarboxylic acid) (PDTC, CAS 69945-42-2) (24). PDTC was synthesized by the method of Hildebrand et al. (25), and electrospray/negative ion MS confirmed its identity with the A fraction. This compound was first identified as a bacterial metabolite accumulated by an isolate of Pseudomonas putida grown under iron limitation (26). It has been characterized with regard to its complexes with a variety of transition metal ions (27, 28, 29). Addition of metal salts to PDTC purified from strain KC cultures or
synthetic material and analysis by electrospray MS gave spectra consistent with the expected 2:1 complexes for iron, with evidence for both ferrous and ferric forms, as noted earlier by others (26, 27), as well as cobalt, and a 1:1 complex with copper which includes a chloride ion (not shown). Dithionite-reduced iron complexes of synthetic PDTC, purified active fraction from strain KC cultures, and crude culture filtrates all gave identical visible absorption spectra (max. 687 nm; « 687 5 8.435 3 10 3). A non-reduced ferric complex solution in water showed maxima at 446, 600, and 736 nm. The molar absorption coefficient was « 600 5 2.3731 3 10 3. Structures of PDTC and its iron complexes are shown in Fig. 4. In assays using Ti[III], synthetic PDTC showed the same activity as material purified from strain KC culture supernatant (CCl 4 degradation, but no activity toward CHCl 3, CFCl 3, 1,1,1-trichloroacetic acid). Quantitation of PDTC in cultures by the spectrophotometric method of Budzikiewicz (24) showed a yield of 10 –30 mg/ml, depending on the amount of iron in the medium. Maximal production of PDTC was observed with no added iron; 10 mM iron caused about 70% reduction in PDTC produced. At .25 mM iron PDTC was below our detection limit by the method of Budzikiewicz. Some CCl 4 transformation was seen in the chemical assay using supernatant from cultures grown with 25 mM iron, but none at 100 mM Fe. Measurement of the transformation capacity of synthetic PDTC and crude culture filtrates gave catalytic turnover of CCl 4, with inactivation apparent. All assays showed activity with or without added iron. Synthetic PDTC complexes with both iron and copper (prepared using FeCl 3 and CuCl 2 solutions) gave activity in CCl 4 transformation bioassays or in assays using Ti[III] citrate as reductant. Analysis of the L fraction by the same spectroscopic techniques allowed us to tentatively identify it as pyridine-2-carboxylic-6-thiocarboxylic acid. This compound was noted as an apparent hydrolysis product of PDTC (24), and it may occur as an artifact of the preparation procedure. Alternatively, PDTC biosynthesis may involve pyridine-2-carboxylic-6-thiocarboxylic acid as an
FIG. 4. Structure of Pseudomonas stutzeri strain KC carbon tetrachloride-dehalogenating factor pyridine-2,6-bis(thiocarboxylate) and its iron complexes. Structures were originally described as iron-binding agents for Pseudomonas putida by Budzikiewicz’s group (26, 27). Molecular weight signals of 450 Da and 225 Da were observed for ferric and ferrous complexes, respectively. The apparent mass of the ferrous complex was half that of the ferric complex, because the ferrous complex carried two negative charges (m/z 21 vs. m/z 22). 565
Vol. 261, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
intermediate that may accumulate in the culture medium (24, 25). DISCUSSION Pyridine-2,6-bis(thiocarboxylate) was identified many years ago as a metabolite of a pseudomonad related to iron limitation. It is likely that a biosynthetic system evolved to take advantage of its affinity for essential trace metals. Without data explicitly demonstrating a role in metal acquisition as the physiological function for PDTC, this remains speculative. Our data have shown conclusively that addition of PDTC is sufficient to confer CCl 4 transformation activity upon cultures which otherwise do not display this activity. We have also shown that PDTC biosynthesis is correlated with conditions previously noted for CCl 4 transformation activity by strain KC, i.e., iron limitation. The activity observed with PDTC in chemical assays or in bioassays also shows the same substrate range as that seen with active cultures of strain KC; no reactivity toward CHCl 3, CFCl 3 or 1,1,1-TCA. PDTC also was shown to act as a catalyst with a finite turnover number, indicating inactivation due to CCl 4 transformation. It remains to be determined exactly how PDTC catalyzes CCl 4 dehalogenation and whether metal ions are necessary or not. The redox activity of iron and nickel complexes of PDTC have been studied (27, 29), and it is easy to envision a metal-centered electron transfer to CCl 4 resulting in the loss of chloride ion and formation of trichloromethyl radical. Our preliminary experiments with synthetic PDTC, which should contain only trace amounts of metals, showed CCl 4 removal when assayed with the physiologically relevant sulfhydryl reducing agents Na 2S and dithiothreitol (data not shown). The identification of a sulfur compound as the extracellular catalyst of CCl 4 transformation by strain KC may also lead to an explanation of the characteristic substitution observed previously (19). Further work is in progress to assess the redox activity of unbound PDTC and to determine the products of these simplified reaction mixtures. ACKNOWLEDGMENTS This work was supported by the U.S. Department of Energy NABIR Program under Grant DE-FG03-96ER62273. Dr. Lee was supported by the Korea Science and Engineering Foundation (KOSEF). We thank Dr. Luying Xun (Washington State University, Pullman) for helpful suggestions, and Ms. Connie Bollinger for editorial assistance.
REFERENCES 1. Illman, D. L. (1993) Chem. Eng. News June, 9 –21. 2. Castro, C. E., Wade, R. S., and Belser, N. O. (1985) Biochemistry 24, 204 –210.
3. Krone, U. E., Laufer, K., and Thauer, R. K. (1989) Biochemistry 28, 10061–10065. 4. Wood, J. M., Kennedy, J. S., and Wolfe, R. S. (1968) Biochemistry 7, 1707–1713. 5. Krone, U. E., Thauer, R. K., and Hogenkamp, H. P. C. (1989) Biochemistry 28, 4908 – 4914. 6. Chiu, P.-C., and Reinhard, M. (1995) Environ. Sci. Technol. 29, 595– 603. 7. Lewis, T. A., Morra, M. J., and Brown, P. D. (1995) Environ. Sci. Technol. 30, 292–300. 8. Chiu, P.-C., and Reinhard, M. (1996) Environ. Sci. Technol. 30, 1882–1889. 9. Krone, U. E., and Thauer, R. K. (1991) Biochemistry 30, 2713– 2716. 10. Gantzer, C. J., and Wackett, L. P. (1991) Environ. Sci. Technol. 25, 715–722. 11. Schanke, C. A., and Wackett, L. P. (1992) Environ. Sci. Technol. 26, 830 – 833. 12. Glod, G., Angst, W., Holliger, C., and Schwarzenbach, R. P. (1997) Environ. Sci. Technol. 31, 253–260. 13. Marks, T., Allpress, J. D., and Maule, A. (1989) Appl. Environ. Microbiol. 55, 1258 –1261. 14. Egli, C., Tschan, T., Scholtz, R., Cook, A. M., and Leisinger, T. (1988) Appl. Environ. Microbiol. 54, 2819 –2824. 15. Galli, R., and McCarty, P. L. (1989) Appl. Environ. Microbiol. 55, 837– 844. 16. Mikesell, M. D., and Boyd, S. A. (1990) Appl. Environ. Microbiol. 56, 1198 –1201. 17. Criddle, C. S., DeWitt, J. T., Grbic-Galic, D., and McCarty, P. L. (1990) Appl. Environ. Microbiol. 56, 3240 –3246. 18. Lewis, T. A., and Crawford, R. L. (1993) Appl. Environ. Microbiol. 59, 1635–1641. 19. Lewis, T. A., and Crawford, R. L. (1995) J. Bacteriol. 177, 2204 – 2208. 20. Lewis, T. A., and Crawford, R. L. (1999) in Novel Approaches for Bioremediation of Organic Pollution: Proceedings of the 42nd OHOLO Conference (Fass, R., Flashner, Y., and Reuveny, S., Eds.), pp. 1–11. Plenum Publishing Corporation, New York, NY. 21. Kriegman-King, M. R., and Reinhard, M. (1992) Environ. Sci. Technol. 26, 2198 –2206. 22. Asmus, K.-D., Bahnemann, D., Krischer, K., Lal, M., and Mo¨nig, G. (1985) Life Chem. Rep. 3, 1–15. 23. Dybas, M. J., Tatara, G. M., and Criddle, C. S. (1995) Appl. Environ. Microbiol. 61, 758 –762. 24. Budzikiewicz, H., Hildebrand, U., Ockels, W., Reiche, M., and Taraz, K. (1983) Z. Naturforsch. 516 –520. 25. Hildebrand, U., Ockels, W., Lex, J., and Budzikiewicz, H. (1983) Phosphorus Sulfur 16, 361–364. 26. Ockels, W., Ro¨mer, A., and Budzikiewicz, H. (1978) Tetrahedron Lett. 1978, 3341–3342. 27. Hildebrand, U., Lex, J., Taraz, K., Winkler, S., Ockels, W., and Budzikiewicz, H. (1984) Z. Naturforsch. 39b, 1607–1613. 28. Espinet, P., Lorenzo, C., and Miguel, J. A. (1994) Inorg. Chem. 33, 2052–2055. 29. Kru¨ger, H.-J., and Holm, R. H. (1990) J. Am. Chem. Soc. 112, 2955–2963. 30. Hu¨bner, J., Taraz, K., and Budzikiewicz, H. (1990) Phosphorus Sulfur Silicon 47, 367–374.
566
