Degradation of nitrate ester and nitroaromatic explosives by ...

Biochemical Society Transactions

680

Chu, G. H. and Wallace, K. K. (1992) Biotechnol. Prog. 8, 179- 186 16 Ridgway, T. J. (1996) Ph.D. Thesis, University of Nottingham 17 Wada, S., Ichikawa, H. and Tatsumi, K. (1995). Biotechnol. Bioeng. 45, 304-309 18 Lea, A. G. H. (1995) in Enzymes in Food Processing (Tucker, G. A. and Woods, L. F. J., eds.), 2nd edn., pp. 223-247, Blackie, London 19 Wiseman, H. (1996) Biochem. Soc. Trans. 24, 795-800 20 Ridgway, T. J. and Tucker, G. A. (1997) Biochem. SOC.Trans. 25, 59-63 21 Ridgway, T. J., Tucker, G. A. and Wiseman, H. (1997) J. Endocrinol. 152, S33 22 Beer, A.-J. (1997) The Biologist 44, 296 23 Weiner, J. M. and Lovely, D. R. (1998) Appl. Environ. Microbiol. 64, 775-778 24 van den Brink, J. M., van den Hondel, C. A. M. J. J. and van Gorcom, R. F. M. (1996) Appl. Microbiol. Biotechnol. 46, 360-364

25 Panke, S., Sanchez-Romero, J. M. and de Lorenzo, V. (1998) Appl. Environ. Microbiol. 64, 748-751 26 Gouka, R. J., Punt, P. J. and van den Hondel, C. A. M. J. J. (1997) Appl. Microbiol. Biotechnol. 47,l-11 27 Davies, T. H., Cottingham, P. D. (1994) Water Sci. Technol. 29,227-235 28 Foyer, C. H., Descourvitxes, P. and Kunert, K. J. (1994) Plant Cell Environ. 17, 507-523 29 Daniell, H., Datta, R., Varma, S., Gray, S. and Lee, S.-B. (1998). Nat. Biotechnol. 16, 345-348 30 Steegborn, C. and Skliidal, P. (1996) Biosens. Bioelect. 12, 19-27 31 Wiseman, A. and Lynch, J. M. (1998) in Environmental Biomonitoring: The Biotechnology Ecotoxicology Interface (Lynch, J. M. and Wiseman, A., eds.), pp. 287-292, Cambridge University Press, Cambridge Received 22 June 1998

Degradation of nitrate ester and nitroaromatic explosives by Enterobacter cloucae PB2 A. Basran*, C. E. French*, R. E. Williams*, S. Nicklint and N. C. Bruce*’ *Institute of Biotechnology, University of Cambridge, Tennis Court Road, Cambridge CB2 I QT, U.K. and Evaluation and Research Agency, Fort Halstead, Sevenoaks, Kent TN I 4 7BP, U.K.

t Defence

Introduction

dysfunction and cancer in mammals, whereas

T h e contamination of soil and ground water with explosives represents an extensive environmental and public health problem. T h e production, handling and disposal of high energy explosives are mainly responsible for the presence of these recalcitrant compounds in the environment. T h e most common contaminating explosives are T N T (2,4,6-trinitrotoluene), P E T N (pentaerythritol tetranitrate) and RDX (hexahydro1,3,5-trinitro1,3,5-triazine). Explosives can be broadly classified into three groups, nitroaromatics (e.g. T N T ) , nitramines (e.g. RDX) and nitrate esters (e.g. PETN) (Figure 1). A major concern is the toxic and mutagenic effects of explosives on biological systems. For example, T N T is a highly toxic compound and is known to cause red blood cell abnormalities, liver

RDX affects primarily the central nervous system and the renal and gastrointestinal systems [1,2]. Further reductive metabolism of RDX and T N T

Abbreviations used: PETN, pentaerythritol tetranitrate; TNT, 2,4,6-trinitrotoluene; RDX, hexahydro-l,3,5-trinitro 1,3,5-triazine; FMN, flavin mononucleotide; GTN, glycerol trinitrate; EGDN, ethylene glycol dinitrate; IPTG, isopropyl-P-D-thiogalactoside. To whom correspondence should be addressed.

’

Volume 26

in mammals results in the formation of nitroso and hydroxylaminonitramine derivatives, which are both toxic and mutagenic [3]. T h e scale of the problem of contaminated sites is vast; in the U.S.A. alone, a recent survey showed there to be in excess of 50000 contaminated sites and several hundred thousand leaking underground storage tanks. T h e cost of treating these polluted sites could reach qbUS1.7 x 10” [4]. Unfortunately, explosives are notoriously recalcitrant to natural chemical breakdown and to complete biodegradation by micro-organisms [S]. T h e problems associated with the biodegradation of these compounds originates in the nature of the covalent linkage of the nitro group within the explosive. Nitro-organic compounds are found very rarely in nature, and when present they are generally toxic compounds, such as the antibiotics chloramphenicol and nitrosporin, and the carcinogenic plant-derived aristolochic acids [6]. T h e r e are even fewer examples of naturally

Xenobiotic Pollution and Recovery by Natural Systems

Figure I Chemical structures of some important explosives NITROAROMATIC

NITRAMINE

NITRATEESTERS

68 I

NO, 2.4.6-trinitrotoluene (TNT)

hexahydm1.3S-trinitm1.3.5-triazine(RDX)

glyceroluiniuate(GTN)

0,N-0-H 0,N-0-H,C

C

\c/ / \

CH,-O-NO, CHr0-NO,

pentaerythritoltetranitrate (PETN)

occurring biologically derived organic nitrate esters (C-O-N02), with an insect sex pheromone [7] being the only apparent example and no known biologically synthesized nitramines (N-N02). It is only recently that nitro-organic compounds have been found in the environment, due solely to the synthesis of novel compounds for a wide range of uses from dyes and pesticides to explosives. Because the occurrence of biologically-derived nitro-organic compounds is so rare, and as they have been present in the environment for only a short period of time, it is not surprising that degradative microbial enzyme systems are also very uncommon. Naturally occurring bacterial communities which have been exposed to environmental pollutants offer the most likely source of novel enzymes and degradative pathways, as they are able to adapt rapidly and exploit new niches over relatively short periods of time.

Isolation and characterization of a strain of Enterobactcr cloacae capable of degrading PETN Selective enrichments were performed on samples taken from explosive-contaminated soil, which resulted in the isolation of a bacterial strain, designated PB2, that was able to grow on PETN as the sole source of nitrogen [l]. This organism, a Gram-negative rod, was identified on the basis of physiological characteristics as a strain of E. cloacae by the National Collection of Industrial and Marine Bacteria (NCIMB, Aberdeen, U.K.). Subsequently, the 1 6 s rRNA genes

were amplified by PCR using specifically designed primers [8], cloned and sequenced. Alignments of these sequences with other 16S rRNA sequences in the Genbank and EMBL databases supported the relationship of PB2 to the family Enterobacteriaceae.

Metabolite analysis of E. cloacae PB2 grown on PETN The degradation of PETN by E. cloacae PB2 was followed by growing the organism on the explosive as the sole source of nitrogen. Metabolites that accumulated in the culture supernatant were extracted using ethyl acetate. Three metabolites were separated using T L C and found to have lower Rf values than PETN, indicating that the unknown metabolites were more polar than PETN. Using electron impact MS, the three metabolites were identified as 2,Z-bis [ (nitroxy) methyllpropanedial. This suggests that E. cloacae PB2 removes two nitrite groups from PETN leaving hydroxyl groups, which are subsequently oxidized to aldehyde groups. Crude extracts prepared from E. cloacae PB2 grown on PETN as the sole nitrogen source showed that in the presence of NADPH PETN was found to disappear rapidly with the concomitant production of nitrite. The products of the enzymatic degradation of PETN were isolated by ethyl acetate extraction and analysed by MS, confirming the earlier observation that PETN reductase liberates two nitrite ions from PETN to produce the pentaerythritol trinitrate and pentaerythritol dinitrate [ 13.

I998


Purification and characterization of PETN reductase from crude cell extracts of E. cloacae PB2 682

PETN reductase was purified from cell extracts to apparent homogeneity; analysis by SDS/PAGE revealed a single protein band with a molecular weight of approximately 40 kDa. The native molecular weight was determined to be approximately 40 kDa by gel filtration chromatography, suggesting that the enzyme is active as a monomer. The purified PETN reductase showed an absorption spectrum typical of an oxidized flavoprotein. The flavin was non-covalently bound to PETN reductase and was identified as flavin mononucleotide. Preliminary studies on substrate specificity have shown that PETN reductase is also active against glycerol trinitrate (GTN) and ethylene glycol dinitrate (EGDN). In the case of GTN, it appears that one nitrite is released rapidly from the explosive and at least one more is released at a much lower rate. EGDN is attacked with much lower affinity (the K , for GTN is 0.023 mM compared with 2.4 mM for EGDN), which suggests that at least three nitrite ester groups are required for rapid activity. PETN reductase was also shown to catalyse the NADPH-dependent reduction of 2-cyclohexen-1-one and TNT. The measurement of kinetic parameters such as K , has been hampered for substrates such as PETN and T N T by their very low solubility. Even with these limitations, it is clear that PETN reductase has the highest activity against PETN and GTN.

Cloning of the PETN reductase gene from E. cloacae PB2, and overexpression in Escherichia coli The N-terminal amino acid sequence of PETN reductase was determined by automated Edman degradation, which permitted the synthesis of two degenerate oligonucleotide probes matching the predicted nucleic acid sequence [9]. Southern blot analysis showed that both probes bound to a fragment produced by a NcoIIClaI digest of E. cloaca PB2 genomic DNA. This fragment was cloned into pBluescript SK+ (Stratagene) to give pONR1. The cloned fragment was sequenced and an open reading frame beginning with codons matching the known N-terminal sequence of PETN reductase was identified. The gene encoding PETN reductase (onr) predicted a protein of 364 amino acids with a molecular weight of 39 358 Da excluding the

Volume 26

N-terminal methionine, which was consistent with the size estimated by SDS/PAGE. Escherichia coli JM109 containing the cloned onr gene expressed high levels of PETN reductase ( 3 0 3 0 % of the total soluble protein) when induced with isopropyl-j?-D-thiogalactoside (IPTG) from the lac promoter of pBluescript SK+. In the absence of IPTG, expression was approximately one-half of that observed with IPTG, suggesting that expression is driven both by the lac promoter and the promoter for onr contained in the insert. Recombinant PETN reductase was purified by affinity chromatography using Mimetic Orange 2 A6XL [9]. The purified enzyme appeared homogeneous by SDS/PAGE and the N-terminal 10 amino acid residues was identical to the authentic enzyme from E. cloacae PB2. PETN reductase proved to be relatively stable and showed no detectable loss in activity after several months of storage at -20°C.

PETN reductase and related enzymes The deduced amino acid sequence of PETN reductase was compared with sequences in protein and nucleic acid databases using the BLAST program of the GCG package [lo]. The most similar proteins found were morphinone reductase from Pseudomonas putida M10 [ 11,121 and other members of the old yellow enzyme family of cc/P-barrel flavoprotein oxidoreductases [9]. These include old yellow enzyme of Saccharomyces carlsbergensis and Saccharomyces cerevisiae and homologues from the yeasts Kluyveromyces lactis and Candida albicans, the protozoan Trypanosoma cruzii, and the plants Arabidopsis thaliana, Oryza sativa and Brassica campestris. Other than morphinone reductase, which is believed to be involved in the breakdown of morphine alkaloids, the physiological functions of these enzymes are unknown. PETN reductase and morphinone reductase share 53% sequence identity and 71% sequence similarity as determined by the GAP program of the GCG package. An alignment of the deduced amino acid sequences of these proteins is shown in Figure 2. The physical properties of PETN reductase are consistent with a close relationship with this family: like all of the known members of this group, it is a simple flavoprotein with FMN as a prosthetic group, its subunit M , is approximately 40000, and it is an oxidoreductase that uses a reduced pyridine nucleotide cofactor as an electron donor.


The structure of old yellow enzyme has been shown to be an eight-stranded alp-barrel [3]. The flavin is hydrogen-bonded by sidechains from residues Thr-37, Gln-114, Arg-243 and Arg-348. All of these residues are conserved in both morphinone reductase and PETN reductase (Figure 2). Of the residues that hydrogen-bond with the flavin through the peptide backbone in old yellow enzyme, Gly-324 and Gly-347 are conserved, whereas several others are replaced by conservative substitutions such

as glycine to alanine. Residues Phe-374 and Tyr375, which provide a hydrophobic pocket for the dimethyl-benezene ring of the flavin, are also conserved, although Phe-296 is not. This suggests that all three enzymes have similar structures.

Degradation of TNT by E. cloacae PB2 A considerable amount of effort has been expended on studying the microbial degradation of explosives and in particular TNT, which is

Figure 2 PETN reductase and related enzymes Alignment of sequences of PETN reductase (Onr) [9], morphinone reductase (MorB) [ 121 and old yellow enzyme (OYE I ) [ 181. The alignment was generated using the PILEUP program of the GCG package [ 101. Conserved residues are shown in bold. The positions of a-helices (H I to HE and HE) and p-strands (SA, SB and S I to 58) forming the eight-stranded alp-barrel structure of old yellow enzyme [3] are indicated.

MorB Onr OYEl MorB Onr OYEl

1 1 1

PDT SFSNPGLFTP LQLGSLSLPN RVIMAPLT.. RSRTPDSVPG SAEKLFTP LKVGAVTAPN RVFMAPLTRL RSIEPGDIPT SFVKDFKPQ ALGDTNLFKP IKIGNNELLH RAJIPPLTRM RALHPGNIPN SA SB S1

41 38 49

42 R.LQQIYYGQ RA..SAGLII SEATNISPTA RGYVYTPGIW TDAQEAGWKG 88 39 P.LMGEYYRQ RA..SAGLII SEATQISAQA KGYAGAPGLH SPEQIAAWKK 85 50 RDWAVEWTQ RAQRPGTMII TEGAFISPQA GGYDNAPOVW SEEQMVEWTK 99 H1 s2 H2

MorB 89 WEAVEAKGG RIALQLWHVG RVS.HELVQP DGQQPVAPSA LKAEGAECFV 137 Onr 86 ITAGVEAEDG RIAVQLWHTG RIS.HSSIQP GWAPVSASA LNANTRTSLR 134 OYEl 100 IFNAIHEKKS FVWVQLOWLG WAAFPDNLAR D a L R m A S D . . . . . . .NVF 142 s3 HA sc MorB 138 EFEDGTAGLH PTSTP.RALE TDEIPGIVED YRQAAQRAKR AGFDMVEVHA 186 Onr 135 D.ENGNAIRV DTTTP.RALE LDEIPGIVND FRQAVANARE AGFDLVELHS 182 OYEl 143 MDAEOEAKAK KANNPmLT KDEIKOYIKE YVQAAKNSIA AGADGVEIHS 192 HB SD H3 s4 MorB 187 ANACLPNQFL ATGTNFSTDQ YGGSIENRAR FPLgVVDAVA EVFGPERVGI 236 Onr 183 AHGYLLHQFL SPSSNQRTDQ YGGSvEblRAR LvLgvvDAvC NEWSADRIGI 232 T V EAIGHEKm 242 OYEl 193 ANGaLNQFL DPHSNTRTDE YGGSIENRAR F s5 H4 HC MorB 237 RLTPFLELFG LTDD.EPEAM AF..YLAGEL DRRG... ..L AYLHFNEPDW 278 Onr 233 RVSPIGTFQN VDNGPNEEAD AL..YLIEEL AKRG ..... I AYLEMSETDL 275 OYEl 243 USPYGVFNS MSGGAETGIV AOYAYVAGEL EKRAKAGKRL AFVHLVEPRV 292 H5 S6 MorB 279 ..... IGGDI TYPEGFREQM RQRFKGGLIY CGNYDAGRAQ ARLDDNTADA 323 Onr 276 . . . . .AGQK. PYSEAFRQKV RERFHOVIIG AGAYTAEKAE DLIGKGLIDA 319 OYEl 293 TNPFLTEGEG EYEGGSNDFV YSIWKGPVIR AGNFALHP.. .EVVREEVKD 339 H6 s7 H7 MorB 324 ....VAFGRP FIANPDLPER FRLGAAINEP DPSTFYGGAE VGYTDYPFLD 369 Onr 320 ....VAFGRD YIANPDLVAR LQKKAELNPQ RPESFYGGGA EGYTDYPSL 364 OYEl 340 KRTLIGYGRF FISNPDLVDR LEKGLPLNKY DRDTFYQMSA HGYIDYPTm 389 S8 HD H8 HE MorB 370 NGHDRLG 376 Onr OYEl 390 W K L G W D K K 399

I998

683


684

historically by far the most important military high explosive (for a review, see [13]). T h e oxidative removal of nitro groups by oxygenases has been observed for mono- and dinitro-substituted toluenes; however, the chemistry of T N T is dominated by the highly electron-withdrawing nature of the nitro groups, which makes the aromatic ring highly electron-deficient. Under both aerobic and anaerobic culture conditions the nitro groups of T N T are reduced to produce Z-amino-4,6-dinitrotoluene, 4-arnin0-2~6-dinitrotoluene, diaminotoluene isomers and occasionally triaminotoluene. Polymeric azoxy compounds, tetranitroazoxy-toluene isomers, can be formed by the abiotic condensation of partially reduced nitro groups on two T N T molecules. Anaerobic conditions are required for enzymatic reduction of nitro to amino compounds because the hydroxylamino intermediates are reoxidized in air to the nitroso form. T h e ability to reduce nitro groups is widespread among micro-organisms. Therefore attempts to degrade T N T using natural heterologous cultures often results in the production and accumulation of aminonitrotoluenes. These compounds appear to be highly recalcitrant. Enzymic reduction of aromatic nitro groups is a well-studied phenomenon, catalysed by enzymes known as nitroreductases. A wide variety of enzymes with redox-active metal ions, such as cytochromes, will catalyse a one-electron reduction of the nitro group to a radical which is instantly re-oxidized by oxygen. These enzymes are known as oxygen-sensitive nitroreductases. Oxygen-insensitive nitroreductases have been shown to exist and these are flavoproteins which catalyse successive two-electron reductions of nitro groups to give nitroso, hydroxylamino and amino groups. E. cloacae PB2 was found to grow very slowly on T N T as a sole nitrogen source in mineral media salts with glucose as the sole carbon source [14]. As mentioned previously, PETN reductase does display a low level of activity against T N T with the oxidation of

NADPH. Therefore, this suggests that PETN reductase may allow E. cloacae PB2 to grow on T N T as a sole nitrogen source. T o determine whether PETN reductase might play a role in the degradation of T N T by E. cloacae PB2, PETN reductase was added to reaction mixtures containing T N T and NADPH. T h e reaction mixtures quickly developed an orange colouration with an absorbance maximum at approximately 500nm. No such coloured products were generated in the absence of enzyme, or T N T , or NADPH. Due to the electron-withdrawing nature of the nitro substituents, the aromatic ring of trinitro compounds such as T N T is electrondeficient and is relatively easily reduced to a coloured hydride-Meisenheimer complex [ 151 (Figure 3). Such reduction can be accomplished by weak reducing agents such as sodium octahydrotriborate [ 151. Biological production of the hydride-Meisenheimer complex of T N T has been reported in whole cells of a Mycobacterium sp. able to grow with 4-nitrotoluene as sole nitrogen source [16], and a Pseudomonas sp. able to grow with T N T as a sole nitrogen source [17]; however, the isolation of the enzyme responsible had not been reported. T h e hydride-Meisenheimer complex of T N T has a characteristic absorption spectrum which does not match that observed in the reduction of T N T by PETN reductase. However, repeat experiments have shown that a reducing agent sodium borohydride reduces T N T initially to the hydride-Meisenheimer complex, and subsequently to one or more orange products with absorption spectra similar to that seen with PETN reductase [ 141. Reaction mixtures resulting from the reduction of T N T by PETN reductase/NADPH or sodium borohydride were analysed by HPLC. Products were detected at 260 nm (ultraviolet) and 500 nm (visible). Absorbance spectra of each peak were also measured. Chemical reduction of

Figure 3 Formation of the hydride-Meisenheimer complex of TNT ( I ) TNT; (2) C-3 H--TNT; (3) C-3, C-5 2H--TNT.

Volume 26


T N T resulted in a peak having the characteristic spectrum of the hydride-Meisenheimer complex, and at least five other visible peaks. In enzymic reaction mixtures, two visible peaks appeared, matching the chemical peaks in retention times and absorption spectra. Because chemical reduction experiments suggest that the initial product is the hydride-Meisenheimer complex, with further reduction leading to the other products observed, we speculate that PETN reductase initially reduces T N T to the hydride-Meisenheimer complex, which is rapidly reduced to the dihydride-Meisenheimer complex and then further reduced to the unidentified products. Rapidity of the second reduction step relative to the first one would account for our failure to observe the hydride-Meisenheimer complex itself in enzymic reaction mixtures. Assays of enzymic reaction mixtures using Griess reagents showed that nitrite accumulates [14]. In one reaction mixture left for a period of several days with an enzymic NADPH regeneration system (to regenerate NADPH oxidized by PETN reductase) it was found that 1.0 mol of nitrite had been liberated per mol of T N T initially present. The stoichiometry and timing of nitrite release have yet to be determined. Nitrite may be released by slow non-enzymic breakdown of one or more of the reaction products. As the final reaction products of T N T reduction by PETN reductase contain less nitrogen than TNT and appear to be water-soluble and non-aromatic, they are likely to be less toxic and less recalcitrant than T N T or nitroreductase products of T N T . Therefore, E. cloacae PB2 and recombinant organisms expressing PETN reductase may be useful in the bioremediation of TNT-contaminated soil and water. 1 Binks, P. R., French, C. E., Nicklin, S. and Bruce, N. C. (1996) Appl. Environ. Microbiol. 62, 1214-1219

2 Cleland, W. W. (1970) in The Enzymes, 3rd edn., vol. 2 (Boyer, P. D., ed.), pp. 1-65, Academic Press, New York 3 Fox, K. M. and Karplus, P. A. (1994) Structure 2, 1089- 1105 4 Singleton, I. (1994) J. Chem. Tech. Biotechnol. 59, 9-23 5 Rieger, P. G . and Knackmuss, H. J. (1995) in Biodegradation of Nitroaromatic Compounds (Spain, J., ed.), Plenum Press, New York 6 Gorontzy, T., Drzyzga, O., Kahl, M. W., Brunsnagel, D., Breitung, J., Vonloew, E. and Blotevogel, K. H. (1994) Crit. Rev. Microbiol. 20, 265-284 7 Hall, D. R., Beevor, P. S., Campion, D. G., Chamberlain, D. J., Cork, A., White, R. D., Almestar, A. and Henneberry, T. J. (1992) Tetrahed. Lett. 33, 4811-4814 8 Weisburg, W. G., Barns, S. M., Pelletier, D. A. and Lane, D. J. (1991) J. Bacteriol. 173, 697-703 9 French, C. E., Nicklin, S. and Bruce, N. C. (1996) J. Bacteriol. 178, 6623-6627 10 Genetics Computer Group (1994) Program manual for the Wisconsin package, version 8, 575 Science Drive, Madison, WI 11 French, C. E. and Bruce, N. C. (1994) Biochem. J. 301,97-103 12 French, C. E. and Bruce, N. C. (1995) Biochem. J. 312, 671-678 13 Spain, J. C. (1995) Annu. Rev. Microbiol. 49, 523-555 14 French, C. E., Nicklin, S. and Bruce, N. C. (1998) Appl. Environ. Microbiol. 64, 2864-2868 15 Kaplan, L. A. and Siedle, A. R. (1971) J. Org. Chem. 36,937-939 16 Vorbeck, C., Lenke, H., Fischer, P., Spain, J. and Knackmuss, H.-J. (1998) Appl. Environ. Microbiol. 64, 246-252 17 Haidour, A. and Ramos, J. L. (1996) Environ. Sci. Technol. 30, 2365-2370 18 Stott, K., Saito, K., Thiele, D. J. and Massey, V. (1993) J. Biol. Chem. 268, 6097-6106 Received 16 July 1998

I998

685