Metabolism of octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine by Clostridium bifermentans strain HAW-1 and several other H₂-producing fermentative anaerobic bacteria

Abstract

Several H₂-producing fermentative anaerobic bacteria including Clostridium, Klebsiella and Fusobacteria degraded octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX) (36 μM) to formaldehyde (HCHO) and nitrous oxide (N₂O) with rates ranging from 5 to 190 nmol h⁻¹ g [dry weight] of cells⁻¹. Among these strains, C. bifermentans strain HAW-1 grew and transformed HMX rapidly with the detection of the two key intermediates the mononitroso product and methylenedinitramine. Its cellular extract alone did not seem to degrade HMX appreciably, but degraded much faster in the presence of H₂, NADH or NADPH. The disappearance of HMX was concurrent with the release of nitrite without the formation of the nitroso derivative(s). Results suggest that two types of enzymes were involved in HMX metabolism: one for denitration and the second for reduction to the nitroso derivative(s).

Octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine, Hexahydro-1,3,5-trinitro-1,3,5-triazine, Biotransformation, Clostridium, Fusobacteria, Hydrogenase

1 Introduction

Hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX) and octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX) (Fig. 1) are two widely used explosives that have resulted in contamination of soil and groundwater [1]. They are known to be toxic to various terrestrial and aquatic organisms [2–4], thus necessitating their removal from contaminated environments.

1

Chemical structures of HMX, RDX and detected bacterial transformation products.

Although RDX and HMX have similar chemical structures, both are oligomers of the same structural moiety (CH₂NNO₂)_n (n=3 for RDX and n=4 for HMX, Fig. 1), HMX is often found to be more resistant to biochemical degradation under both aerobic and anaerobic conditions [5–10]. For instance, RDX can be degraded by several anaerobic bacteria including Klebsiella pneumoniae, Serratia marcescens, Morganella morganii, Citrobacter freundii, Escherichia coli[11–13], Clostridium, and Desulfovibrio[14,15], but thus far only one facultative anaerobic bacterium M. morganii has been described to transform HMX to its nitroso derivatives [11]. Although mixed cultures in anaerobic sludge [16], soil [17] and marine sediment [18] are capable of mineralizing HMX, they cannot be used to provide clear insight into its degradation pathway. The present study describes metabolism of HMX by Clostridium bifermentans strain HAW-1 and several other H₂-producing fermentative anaerobes isolated from anaerobic sludge (Clostridium and Klebsiella) and marine sediment (Fusobacteria).

2 Materials and methods

2.1 Microbial isolates and their identification

The sources and identification of the microorganisms used in the present study are listed in Table 1[13–15,18,19]. The major strain of the present study, C. bifermentans strain HAW-1, has been deposited in the American Typical Culture Collection with a number of ATCC BAA791.

1

Microorganisms used in the present study

Isolatesa	Sources	Bacteria matchedb	Similarity (%)c	References
SCZ-1	Anaerobic sludge	Klebsiella pneumoniae	99.5	[13]
HAW-1 (AY604562, 1238b)	Anaerobic sludge	Clostridium bifermentans ATCC 638 (X75906)	99.6	[14,15]
HAW-HC1 (AY604563, 1243b)	Anaerobic sludge	Clostridium butyricum ATCC 43755 (X68176)	99.5	[15]
HAW-G4 (AY604564, 1284b)	Anaerobic sludge	Clostridium celerecrescens DSM 5628 (X71848)	99.6	[15]
HAW-E3 (AY604565, 1311b)	Anaerobic sludge	Clostridium saccharolyticum DSM2544 (Y18185)	99.8	[15]
HAW-EB21 (AY579753, 1266b)	Marine sediment	Propionigenium and Ilyobacter (Fusobacteria)d	92–93	[19]

a

GeneBank Accession Numbers and the length of the 16S rRNA genes of isolates are listed in parentheses for each entry.

b

GeneBank Accession Numbers of their 16S rRNA genes are listed in parentheses.

c

Value was the similarity between the 16S rRNA gene of isolate and that of matched bacterium (Blast method, Genebank).

d

Phylogenetic analyses based on 16S rRNA genes indicated that strain HAW-EB21 fell in the cluster of Fusobacteria with Ilyobacter and Propionigenium as the most closely related bacteria [19].

2.2 Chemicals and media

HMX (99% pure) and [UL-¹⁴C]-HMX (chemical purity, >94%; radiochemical purity, 91%; specific radioactivity, 93.4 μCi mmol⁻¹) and the uniformly ring-labeled [¹⁵N]-HMX (>98% purity) were provided by Defense Research and Development Canada (DRDC), Valcartier, Canada. [¹⁴C]-HCHO (53 mCi mmol⁻¹ specific activity) was from Aldrich, Canada. 4-Nitro-2,4-diazabutanal (NDAB) (99% pure) was provided by R.J. Spanggord from SRI International (Menlo Park, CA, USA). Methylenedinitramine was purchased from the rare chemical department of Aldrich, Oakville, Ont., Canada. All other chemicals used were of reagent grade.

Bacto peptone (1 g l⁻¹) – yeast extract (1 g l⁻¹) – glucose (1 g l⁻¹) medium (YPG) was prepared as described previously [13]. Reinforced Clostridial Medium (RCM), Bacto Brewer Anaerobic Agar and Marine Broth 2216 were purchased from Becton Dickinson (Sparks, MD, USA). The liquid media were sterilized either by autoclaving at 120 °C for 20 min or through filtration using sterile filters (0.22 μm, Millex™ GP, Millipore, Bedford, MA, USA). The dry serum bottles were sterilized by autoclaving at 120 °C for 60 min.

2.3 Anaerobic bacterial growth and biotransformation of HMX

The following anaerobic procedure was used for all biotransformation tests unless otherwise noted. Sterile YPG (for all sludge isolates), RCM (for Clostridium) or Marine Broth 2216 (for the sediment isolate HAW-EB21) medium (19 ml) containing HMX (36 μM) was added to a serum bottle (dry-autoclaved) and degassed using an oil vacuum pump. The headspace was charged with oxygen-free argon after passing the gas through a sterile filter. Sterile sodium sulfide (0.025%) and l-cysteine · HCl (0.025%) were then added to the medium to remove final traces of oxygen. One millilitre of previously grown liquid culture was inoculated to an initial OD_{600 nm} ranging from 0.05 to 0.10, followed by static incubation at 37 °C for anaerobic sludge isolates or 10 °C for the marine sediment isolate. A sterile syringe, flushed with argon and reduced buffer, was used for sampling the liquid culture, and a gas-tight syringe for sampling the headspace for subsequent analyses of N₂O and H₂ (see below). Bacterial growth was monitored by measuring the OD (light path, 1 cm) of the liquid culture medium at 600 nm. Some microcosms were spiked with [UL-¹⁴C]-HMX (0.038 μCi) to measure mineralization (liberated ¹⁴CO₂) using a Tri-Carb 4530 liquid scintillation counter (LSC, model 2100 TR, Packard Instrument Company, Meriden, CT, USA). When ¹⁴CO₂ ceased to form, the microcosm(s) was sacrificed to measure the remaining radioactivity in the culture supernatant and in the biomass.

2.4 Biotransformation of HMX and HCHO by resting cells of strain HAW-1

After growth of strain HAW-1 in 800 ml of YPG medium (in two 500 ml serum bottles) for 20 h (final OD_{600 nm}, 1.04; wet weight, 3 g l⁻¹) under the above conditions, cells were harvested by centrifugation at 9000g (4 °C) for 30 min. The wet cells obtained (wet weight, 2.4 g; dry weight, 0.48 g) were then suspended in a reduced buffer (40 ml) (phosphate, 25 mM; sodium sulfide, 0.025%; l-cysteine · HCl, 0.025%; pH, 7.4). The liquid culture broth or cellular suspensions were handled in anaerobic glove box. Aliquots of cell suspensions (1 ml) were then added to a reduced buffer (4 ml) containing HMX (17 μM) in a serum bottle (20-ml) (biomass, 2.4 g [of dry weight] of cells l⁻¹). The bottles were sealed under argon or H₂ (1 atm) and incubated statically at 37 °C. Reactions were stopped and sacrificed at desired incubation time for analyses of substrate and products. As a comparison, we also conducted biotransformation of RDX (100 μM) by resting cells under the same conditions as used for HMX. Biotransformation of HCHO and [¹⁴C]-HCHO by resting cells was conducted under the same conditions as used for HMX.

2.5 Biotransformation of HMX by crude cellular extract of strain HAW-1

Crude cellular extract was obtained by enzymatic (lysozyme) lysis of wet cells of strain HAW-1 suspended in reduced phosphate buffer (pH 7.4) following the protocol described by Preuss et al. [20]. HMX (17 μM) and RDX (100 μM) transformation by the cellular extract (0–1 mg protein) was carried out at 37 °C in 1 or 2 ml phosphate buffer (20 mM, pH 7.4, in 10-ml sealed vials) containing either 1 mM NADH, 1 mM NADPH or 1 atm of H₂ as electron donors. The amounts of RDX and HMX degraded were measured after 5 and 25 h of incubation, respectively. Hydrogenase activity in cellular extract was assayed in a reduced phosphate buffer (2 ml, pH 7.4, 22 °C) containing 2 mg cellular protein by its ability to reduce the colorless methyl viologen (1 mM) to a blue color (λ 578 nm) in the presence of H₂ (1 atm) [21]. We were not able to quantify hydrogenase activity due to the sensitivity of the resulting reduced methyl viologen product to air. Two negative methyl viologen reduction controls were prepared, one containing cellular extract alone without H₂ and another containing H₂ alone without cellular extract. Protein concentration was determined using the Bicinchoninic Acid Method by following the instruction of the BCA Protein Assay Kit (Pierce, Rockford, IL, USA).

2.6 Analyses HMX, HMX products and H₂

HMX and methylenedinitramine were analyzed by HPLC/UV at 230 nm and by HPLC–MS method as described previously [16,19]. Instrumental conditions for mass analyses of nitroso derivatives of HMX were: capillary column, Zorbax SB-C18 (5 mm ID × 150 mm, Agilent Technology, Mississauga, Ont., Canada); solvent, 20% of aqueous acetonitrile at a flow rate of 12 μl min⁻¹; MS detector, negative electrospray ionization mode [ES(–)] scanned from 40 to 400 Da. Mononitroso HMX (1NO-HMX) yield was estimated by comparing its A_{230 nm} with the A_{230 nm} of HMX removed. The methods for analyses of nitrite (NO₂⁻), formaldehyde (HCHO), methanol (MeOH) and acetate (CH₃COO⁻) were described in earlier report [14]. H₂ was measured using a Gas Chromatograph (Agilent 6890, Wilmington, DE, USA) connected to a Chromosorb 102 column (2 mm × 11 m, Supelco, Bellafonte, Canada) and a thermal conductivity detector (TCD, 150 °C) with argon as a carrier gas. Initial column temperature was maintained at 35 °C for 7.5 min, which was increased to 100 °C at a rate of 75 °C min⁻¹ and then maintained at 100 °C for 6 min. Hydrogen was sampled using a gas tight syringe (500 μl) from the headspace of the microcosm(s). All tests were performed in triplicates unless noted otherwise.

3 Results and discussion

3.1 Biotransformation of HMX in growing and H₂-producing fermentative anaerobes

Six anaerobic bacteria including C. bifermentans, C. saccharolyticum, C. celerecrescens, C. butyricum, K. pneumoniae (from anaerobic sludge) and Fusobacteria (from marine sediment) (Table 1, [13–15,19]), were found to successfully transform HMX (36 μM) under growing conditions (Table 2). All above isolates produced H₂ in the absence and presence of HMX in 20 ml of either yeast extract–peptone–glucose medium (YPG, 0.06–0.43 mmol H₂), Reinforced Clostridial Medium (RCM, 0.22–1.41 mmol H₂), or Marine Broth 2216 (0.05 mmol H₂ for HAW-EB21) (Table 2). Better growth of the clostridial bacteria was found in the RCM medium (0.8–3.2 OD_{600 nm}, 1.0–4.1 g [dry weight] of cells l⁻¹) as compared to that (0.2–0.4 OD_{600 nm}, 0.26–0.52 g [dry weight] of cells l⁻¹) obtained in the YPG medium. Also the RCM medium resulted in faster HMX total removal rate (28–102 nmol day ⁻¹) than the latter one (3.4–30 nmol day⁻¹). Among all isolates incubated in HMX-containing RCM media, strain HAW-1 grew most rapidly (OD_{600 nm} increased from initial 0.05–0.85 after 24 h incubation) and showed the highest specific HMX removal rate (Table 2). In YPG media, although strain HAW-E3 and G4 showed higher HMX removal rates than strain HAW-1 at the log and early stationary phases (Table 2), the two strains had a lag time of 2–3 days before growth [15], causing a delay in HMX transformation.

2

Biotransformation of HMX (36 μM) by growing anaerobic bacteria

Isolates	Bacterial growth, H2 production and HMX removal rates (RateHMX)a
	In YPG mediumb			In RCM mediumc
	Growth (OD600 nm/dry weightd)	H2 (mmol)e	RateHMXf	Growth (OD600 nm/dry weightd)	H2 (mmol)e	RateHMXf
SCZ-1	0.39/0.51	0.06	10	n.d.g	n.d.g	n.d.g
HAW-1	0.28/0.36	0.07	80	0.85/1.1	0.39	190
HAW-E3	0.31/0.40	0.05	150	2.1/2.7	0.22	50
HAW-G4	0.20/0.26	0.16	110	2.7/3.2	0.40	30
HAW-HC1	0.28/0.36	0.43	40	3.2/4.2	1.41	10
HAW-EB21h	1.2/1.6h	0.05	5h	n.d.g	n.d.g	n.d.g

a

The values are the average of duplicate measurements.

b

In 20 ml of yeast extract-peptone(bacto)-glucose medium.

c

In 20 ml of Reinforced Clostridial Medium (containing casein, proteose peptone, beef extract, yeast extract, starch, glucose and sodium acetate).

d

Unit as g [dry weight] of cells l⁻¹.

e

Total H₂ amount (mmoles) produced during growth in the HMX-containing medium.

f

Specific rates (nmolh⁻¹g [dry weight] of cells⁻¹) were measured based on removal of HMX during log and early stationary growth phase.

g

Not determined.

h

Incubated in 20 ml of Marine Broth 2216 (peptone, 5 gl⁻¹, yeast extract, 1 gl⁻¹) at 10 °C.

At the end of the experiment which lasted 27 days, the clostridial isolates transformed 35–92% of HMX in RCM medium as compared to 46–81% in the YPG medium. In YPG or RCM medium controls containing HMX alone, the loss of the energetic chemical was 5% and 18%, respectively. No HMX loss was found in Marine Broth 2216 at 10 °C.

All growing isolates degraded HMX to eventually produce HCHO and N₂O. Strain HAW-1 degraded HMX to N₂O in high yield (74% of total N of HMX removed) (Fig. 2(a)). Using [UL-¹⁴C]-HMX, we obtained 7.8% mineralization (liberated as ¹⁴CO₂) in 30 days, and 9.2% in eight months. Since the radiochemical purity of HMX was approximately 90%, we were unable to confirm whether the little mineralization observed originated from HMX or from the unknown impurity. Earlier we reported poor mineralization (3%) of RDX (radiochemical purity, 96%) with strain HAW-1 [14]. Of the remaining radioactivity in the final media 22% was detected as HCHO, 1.6% associated with biomass and 63% was assigned to unidentified soluble fraction in the aqueous phase.

2

Biotransformation of HMX by (a) growing cells (the OD_{600 nm} values shown on right y axis are 40 times those originally detected values, H₂ produced) in RCM medium and (b) resting cells (biomass, 2.4 g [of dry weight] of cells l⁻¹, 1 atm H₂ added to headspace, 1NO-HMX product was not shown) of C. bifermentans strain HAW-1.

We found that in the case of isolate HAW-1, HMX biotransformed via the formation of the mononitroso derivative octahydro-1-nitroso-3,5,7-trinitro-1,3,5,7- tetrazocine (1NO-HMX) (Fig. 1), identified by its characteristic deprotonated molecular mass ion adduct ([M_NO-HMX-H + NO₂]⁻) at 325.0 Da. Such deprotonated adduct ions have been reported previously for RDX [22]. Only in the case of the sediment isolate strain HAW-EB21 (Table 1), HMX biotransformed to give several nitroso products including 1NO-HMX, dinitroso- (2NO-HMX) and tri-nitroso (3NO-HMX) intermediates (Fig. 1).

Although we were unable to detect nitrite under the growing cell conditions, its absence should not exclude the occurrence of denitration as another route of degradation for HMX. For instance when we incubated NaNO₂ with strain HAW-1 under the same condition the ion disappeared. Earlier we reported that C. bifermentans strain HAW-1 [14] and K. pneumoniae SCZ-1 [13] could degrade RDX through initial denitration and/or reduction to the mononitroso derivative MNX followed by denitration.

3.2 Biotransformation of HMX by resting cells in the presence of H₂

Resting cells (2.4 g [of dry weight] of cells l⁻¹) of strain HAW-1 suspended in reduced phosphate buffer in an atmosphere of argon biotransformed HMX, at a rate of 320 nmol h⁻¹ g [of dry weight] of cells⁻¹. Supplementation of H₂ (1 atm) dramatically increased the rate of HMX biotransformation to 720 nmol h⁻¹ g [of dry weight] of cells⁻¹, suggesting that HMX removal was likely carried out by a reductive hydrogenase enzyme [21,23,24]. This is consistent with earlier studies in mixed anaerobic cultures showing improvement of HMX removal by addition of H₂[5,25].

The disappearance of HMX during incubation with resting cells in the presence or absence of H₂, was accompanied with the formation of NO₂⁻, HCHO, N₂O. In addition we detected the two intermediates 1NO-HMX and methylenedinitramine (MEDINA) (see Fig. 1 for structures). MEDINA is known to undergo spontaneous decomposition in water to produce stoichiometric amounts of N₂O and HCHO [26]. Using ring labeled [¹⁵N]-HMX, we detected N₂O mostly (81%) in the form of ¹⁵N¹⁴NO (m/z 45 Da), indicating that N₂O originated mainly from the nitramine group –N–NO₂ of HMX. For instance, when we incubated NaNO₂ with stain HAW-1 insignificant amounts of N₂O were formed [14]. The concurrent formation of NO₂⁻ with the removal of HMX suggested that in addition to initial reduction to 1NO-HMX, denitration was another significant decomposition route. Denitration likely occurred through one-electron reduction mechanism as proposed earlier for degradation of RDX in K. pneumoniae SCZ-1 isolated from anaerobic sludge [13] and transformation of HMX by a xanthine oxidase from butter milk [27].

After 80 h of incubation of HMX (17 μM) with resting cells in the presence of H₂, we obtained a nitrogen mass balance of 65.4% distributed as follows (of the total N of HMX removed): N₂O (50%), NO-HMX (8.4%), MEDINA (7%) and NO₂⁻ (trace amounts) and a carbon mass balance of 48.4% distributed as follows (of the total C of HMX removed): HCHO (36.5%), MEDINA (3.5%) and NO-HMX (8.4%). We attempted to analyze for other potential HMX products such as methanol, formic acid and acetate, but none was found. For instance, when we incubated [¹⁴C]-HCHO with strain HAW-1 neither ¹⁴C-HCOOH nor ¹⁴C-CH₃COOH was formed. Apparently most of HCHO that might have been produced from HMX either escaped further mineralization or underwent polymerization or interaction with extra cellular components.

Similarly, resting cells of the sediment isolate Fusobacteria HAW-EB21 in the presence of H₂ also transformed HMX to N₂O and HCHO, with detection of MEDINA as a key ring cleavage intermediate. As in the case with growing cells we detected HMX nitroso products 1NO-HMX and 2NO-HMX. None of the nitroso products persisted indefinitely and they all transformed further to apparently produce N₂O and HCHO.

3.3 Biotransformation of HMX in cellular extracts

To determine whether HMX transformation is a protein dependent reductive reaction, we treated the chemical with cellular extract of strain HAW-1 in the presence and absence of the electron-donors H₂, NADH and NADPH. We found that the presence of these electron donors increased the specific removal rate of HMX by a factor ranging from 5- to 10-fold compared to those obtained in their absence (Table 3). We also found that the transformation rates of the energetic chemical were directly proportional to the amount of cellular protein added to the reaction media (Fig. 3). Abiotic reductions of HMX by NAD(P)H or H₂ alone in the absence of cellular extract were negligible (Fig. 3). The present results demonstrate that HMX removal is an enzyme(s)-catalyzed reductive reaction requiring an electron donor.

3

Rates of HMX (17 μM) and RDX (100 μM) biotransformation and yields of metabolites by the cellular extracts of C. bifermentans strain HAW-1

Substrate	Electron donor	Specific removal rate (pmol h−1 mg−1)b	Normalized yields of ring cleavage products (molar)a
			NO2−	MEDINAd	NDABe	N2Oc	HCHO
HMX	NADH	199	1.4 ± 0.1	0.2 ± 0.1	0	1.8	0.8 ± 0.2
HMX	NADPH	247	1.2 ± 0.1	0.2 ± 0.1	0	2.4	NDg
HMX	H2	167	2.1 ± 0.5	0.3 ± 0.2	0	1.6	0.8 ± 0.2
HMX	–f	27	2.1 ± 0.5	0.2 ± 0.1	0	NDg	0.3 ± 0.0
RDX	NADH	14300	0.4 ± 0.1	0.6 ± 0.0	Trace	2.1	0.9 ± 0.1
RDX	NADPH	12100	0.6 ± 0.2	0.4 ± 0.1	Trace	2.2	0.8 ± 0.0
RDX	H2	1500	1.0 ± 0.1	0.7 ± 0.0	Trace	1.2	0.8 ± 0.0
RDX	–f	2000	1.0 ± 0.1	0.5 ± 0.0	Trace	NDg	0.5 ± 0.2

a

Nitroso derivatives were not produced from HMX and RDX in cellular extract reactions.

b

The values are the slopes of the lines (R²=0.99) of Fig. 3 for HMX and of similar lines for RDX (data not shown).

c

The values are the average of duplicate measurements.

d

MEDINA, Methylenedinitramine.

e

NDAB, 4-Nitro-2,4-diazabutanal.

f

Cellular extract alone in buffer.

g

ND, not determined.

3

Dependence of HMX (17 μM) removal on the protein content of cellular extract of strain HAW-1 in the presence of H₂ (1 atm), NADH (1 mM) or NADPH (1 mM) as electron donors (1 ml phosphate buffer, pH 7.4 and 37 °C).

A H₂-uptake hydrogenase activity was detected in the cellular extract of strain HAW-1 as indicated by its ability of using H₂(1 atm) to reduce methyl viologen (1 mM) in a reduced phosphate buffer (2 ml, pH 7.4, 2 mg cellular protein, 10 min) [20,21]. Hydrogenase is an enzyme known to mediate one-electron transfer processes [24], and its detection suggests its involvement in the reductive transformation of HMX with H₂ as an electron donor. However, the cellular extract of strain HAW-1 degraded HMX faster in the presence of NAD(P)H over H₂ as e-donor (Table 3), indicating the involvement of other reductive enzymes that require NAD(P)H as a co-substrate(s) or electron donor.

HMX transformation by cellular extract gave a closely related product distribution (NO₂⁻, N₂O and HCHO) (Table 3) to what was observed with resting cells. Removal of one mole of HMX was accompanied with release of two moles of NO₂⁻, indicating that HMX degraded via initial denitration. The nitroso products that were observed in whole cells reactions were not detected in the cellular extract-catalyzed HMX transformation, suggesting that the enzyme(s) responsible for reduction of HMX to nitroso derivatives, more likely a two-electron nitroreductase [28], were absent from the cellular extract.

3.4 Nitramine metabolism: comparison of HMX with RDX

In general we found that growing cells of clostridial isolates transformed HMX at specific rates (40–150 nmol h⁻¹ g [dry weight] of cells⁻¹) (Table 2) 60–160 times lower than those reported for RDX (2500–24,000 nmol h⁻¹ g [dry weight] of cells⁻¹) [13,15]. In the case of strain HAW-1, resting cells transformed HMX in the presence of extraneous H₂ (headspace contained one atmosphere of H₂) at a rate (320 nmol h⁻¹ g [of dry weight] of cells⁻¹) 15 times lower than that of RDX (4800 nmol h⁻¹ g [of dry weight] of cells⁻¹). The cellular extracts of strain HAW-1 removed HMX at rates 49–72 times lower than those of RDX in the presence of NADH and NADPH as electron donors (Table 3). In addition we found that biomass of Clostridium (0.2–0.4 OD_{600 nm} or 0.26–0.52 g [dry weight] of cells l⁻¹) obtained during growth in YPG medium in the presence of HMX was only one third that obtained in the presence of RDX (0.6–1.1 OD_{600 nm}) in the same medium, indicating that HMX inhibited bacterial growth.

Experimental evidences gathered thus far demonstrate that HMX was more resistant to biotransformation than RDX. This is consistent with previous observations showing that mixed anaerobic microbial population of a methanogenic culture [5], soil [6], anaerobic sludge [16] and marine sediment [18] removed HMX at much lower rates than those observed for RDX. It has been reported that HMX is chemically more stable than RDX [29] and that is possibly caused by the drastic difference in their respective structural conformations. For instance, HMX favors a crown-type conformation while RDX prefers a chair conformation, resembling their cyclooctyl and cyclohexyl compounds, respectively. Cyclohexyl halides were reported to be more reactive toward bimolecular elimination (E₂) of hydrogen halides than their corresponding cyclooctyl ones [30]. Also HMX denitration (E₂ reaction) requires higher energy than that of RDX [31]. This trend in the chemical reactivity between RDX and HMX is in line with their preferential biodegradability as observed in the present study and in earlier reports [5,6,16,18]. The fact that both HMX and RDX are oligomers of the same chemical moiety (CH₂–NNO₂), trimer and tetramers respectively, it is not surprising that both energetic chemicals produced a closely related product distribution (Table 3).

In conclusion, the present H₂-producing fermentative anaerobes from anaerobic sludge and marine sediment cleaved HMX through initial denitration of the chemical or reduction to NO-HMX. Results suggest that two types of enzymes were involved in HMX metabolism: one type enzyme (including hydrogenase) mediating one-electron transfer, brought about denitration as found in the cellular extracts of strain HAW-1 and the second type, nitro reductase mediating two-electron transfer process, led to the reaction of HMX to the nitroso derivative(s) as found in the whole cells of strains HAW-1 and HAW-EB21.

Acknowledgements

We are grateful to Sonia Thiboutot and Guy Ampleman from the Defense Research and Development Canada (DRDC), Valcartier, Que., Canada for providing us with the energetic chemicals. We thank Chantale Beaulieu and Alain Corriveau for technical assistance. Funding was provided by the US Strategic Environmental Research and Development Program (SERDP CU1213) and the Office of Naval Research of US Navy (Award N000140310269).