Oxygen Consumption by Desulfovibrio Strains with and without Polyglucose

Abstract

The kinetics of oxygen reduction by Desulfovibrio salexigens Mast1 and the role of polyglucose in this activity were examined and compared with those of strains of D. desulfuricans and D. gigas. Oxidation rates were highest at air saturation (up to 40 nmol of O₂ min⁻¹ mg of protein⁻¹) and declined with decreasing oxygen concentrations. Studies with cell extracts (CE) indicated that NADH oxidase was entirely responsible for the oxygen reduction in strain Mast1. In D. desulfuricans CSN, at least three independent systems appeared to reduce oxygen. Two were active at all oxygen concentrations (NADH oxidase and NADPH oxidase), and one was maximally active at less than 10 μM oxygen. In contrast to D. gigas and D. salexigens strains, the D. desulfuricans strains also contained NADH peroxidase and NADPH peroxidase activities and did not accumulate polyglucose under nonlimiting growth conditions. At air saturation, initial activities of the oxidases and peroxidases of cells harvested at the end of the log phase were on the order of 20 to 140 nmol of O₂ min⁻¹ mg of protein⁻¹. In all strains, these enzymes were relatively stable but were susceptible to inactivation as soon as substrates were added to the assay mixture. Under those conditions, all oxidation activity disappeared after ca. 1 h of incubation. The same finding was observed with whole cells of D. desulfuricans CSN and D. desulfuricans ATCC 27774, but inactivation was less pronounced with cells of D. salexigens Mast1. It appeared that the presence of polyglucose in the whole cells retarded the process of inactivation of NADH oxidase, but this property was lost in crude CE. In spite of the effect of polyglucose on the oxidative potential, oxygen-dependent growth of D. salexigens Mast1 could be demonstrated neither in batch nor in continuous culture.

There have been only a few studies on the presence of polysaccharides in sulfate-reducing bacteria (SRB). Stams et al. (32) observed the accumulation of polyglucose in several Desulfovibrio species and Desulfobulbus propionicus. In Desulfovibrio vulgaris Hildenborough and D. baculatus HL21, polyglucose was produced when growth was limited by Fe²⁺ or NH₄⁺. In D. gigas (32) and D. salexigens Mast1 (35), polyglucose accumulated in high quantities under nonlimiting growth conditions. Both organisms were able to convert polyglucose anoxically and with oxygen as an electron acceptor (29, 35).

Various SRB are aerotolerant to some degree (6, 10, 13, 28), and even after prolonged exposure to oxygen many species can resume anoxic growth. Most of them contain superoxide dismutase, and catalase has been detected in some of them (1, 2, 13, 14). Little is known about the enzymes involved in oxygen consumption in SRB. In D. gigas, an oxygen reduction chain consisting of an NADH oxidase (NADH rubredoxin oxidoreductase) and a rubredoxin oxygen oxidoreductase has been described (4, 5). NADH oxidase activities have also been observed in D. desulfuricans NCIB 8301 (1) and D. vulgaris Hildenborough (3). However, Hardy and Hamilton (13) observed oxygen reduction activities in several D. vulgaris strains but were unable to detect any NADH oxidase activity. In D. desulfuricans CSN, maximum oxygen consumption rates were observed below 10 μM dissolved oxygen (1, 8, 19). It was found that oxygen reduction in this organism takes places in the periplasm and is linked to cytochrome c₃ (5a), as Postgate (26) already had proposed for another D. desulfuricans strain.

H₂, various organic compounds, and inorganic sulfur compounds all have been identified as possible substrates coupled to oxygen reduction (7, 22, 35). Although D. gigas and D. desulfuricans CSN produces ATP under oxic conditions, the coupling of ATP formation to oxygen reduction has been observed only in the latter organism (8, 29). However, truly oxygen-dependent growth has never been demonstrated for these bacteria. In our opinion, this fact includes the recently reported oxygen-dependent growth of D. vulgaris Hildenborough (18), in which an approximate 50% linear increase in cell density was observed. This observation is most likely explained by growth at the expense of thiosulfate, produced by chemical reduction of oxygen by hydrogen sulfide, as was concluded much earlier for the growth of D. vulgaris DSM 2119 in oxygen sulfide gradient tubes (6).

In a recent paper (35), we reported that D. salexigens Mast1 oxidized substrates with oxygen only as long as the cells contained polyglucose. It was therefore hypothesized that D. salexigens Mast1, having been isolated from the oxic-anoxic layer of a microbial mat, was dependent on polyglucose to survive during oxic periods (35). We report here the presence of NADH oxidase activity in D. salexigens Mast1 and in several other Desulfovibrio strains. The NADH oxidases in all of these strains were prone to inactivation as soon as they catalyzed the oxidation of NADH. We further show that the presence of polyglucose in cells of D. salexigens Mast1 prolonged the activity of NADH oxidase.

MATERIALS AND METHODS

Microorganisms.

The following strains were used: D. salexigens Mast1 (from the top layer of a marine microbial mat, Paleohori Bay, Isle of Milos, Greece; isolated from anoxic batch enrichment cultures on alanine [35]), D. salexigens Mast2 (same origin as strain Mast1; isolated from anoxic continuous enrichment cultures on alanine), D. salexigens DSM 2638 (obtained from the Deutsche Sammlung von Mikroorganismen, Braunschweig, Germany), D. gigas NCIMB 9332, D. desulfuricans BH, and D. desulfuricans ATCC 27774 (kindly provided by T. Hansen), and D. desulfuricans CSN (kindly provided by H. Cypionka, Oldenburg, Germany).

Cultivation.

The D. salexigens strains were cultivated on basal morpholinepropanesulfonic acid-buffered MPM medium (35) supplemented with various substrates (see Results). All other Desulfovibrio strains were grown in a bicarbonate-buffered medium (25) supplemented with trace elements (24), vitamin solution (1 ml liter⁻¹) (15), 10% yeast extract solution (10 ml liter⁻¹), and 1.25% Na₂S · 9H₂O solution (10 ml liter⁻¹). The carbon and energy source was either lactate or pyruvate (20 mM), unless stated otherwise. Screw-cap tubes and rubber-stoppered serum bottles (300 to 500 ml) were used for batch cultures and prepared as described by Van Niel et al. (35). When necessary, air was injected with syringes through a sterile filter (0.2 μm) to the headspace. Anoxic chemostat cultures of D. salexigens Mast1 were grown on 20 mM alanine at a dilution rate of 0.03 h⁻¹ and a temperature of 30°C; the pH was kept at 7.4 by automatic titration with 1 N NaOH. The chemostat had a working volume of 750 ml, and contents were stirred by means of a magnetic stirrer; the headspace of the culture vessel was flushed continuously with N₂-CO₂ gas (80:20, vol/vol) freed of O₂ traces by passage over hot copper turnings. Studies of the possible growth of D. salexigens Mast1 under oxic conditions in continuous culture were done with two coupled chemostats. The first one was kept strictly anoxic as described above. The cells were grown on a mixture of 10 mM alanine and 2 mM glucose, and N₂-CO₂ gas was sparged through the culture to strip the sulfide. The effluent of the first fermentor was pumped continuously into the second fermentor, which was kept under oxic conditions. The oxygen concentration in this fermentor was varied by changing the ratio of air and N₂-CO₂ gas. The gas was sparged through the culture, and the oxygen concentration in the liquid was measured continuously with an oxygen electrode. Additional medium (containing 10 mM alanine) was pumped into the second fermentor. Both fermentors were run at the same dilution rate (0.03 h⁻¹).

Preparation of whole-cell suspensions and CEs.

At the end of the exponential growth phase, batch cultures (200 to 500 ml) were harvested by centrifugation for 10 min at 25,000 × g and 4°C. After being washed with Tris-HCl buffer (50 mM, pH 7.6), the pellet was resuspended in the same buffer to a final protein concentration of 0.5 to 1 g liter⁻¹ for experiments with intact cells. For cell extract (CE) preparation, the pellet was resuspended in 0.5 to 2 ml of the same buffer. The cells were broken by two successive passages through a French pressure cell. Then, part of the crude extract was centrifuged for 30 min at 18,000 rpm to obtain the soluble CE fraction. The membrane fraction was washed once with Tris-HCl buffer and finally resuspended in 1 ml of this buffer. The cell suspensions and CEs were stored on ice until use.

Enzyme assays.

The rate of oxidation of NADH and NADPH in the presence of oxygen (oxidase activity) was measured spectrophotometrically at 340 nm (ɛ = 6,220 M⁻¹ cm⁻¹). The rate of peroxidation of NADH and NADPH was also measured at 340 nm but under anoxic conditions; for this assay, the reaction was started by the addition of 2 to 4 mM H₂O₂. The assay mixture contained either crude or soluble CE diluted in Tris-HCl buffer (50 mM, pH 7.6) (final volume, 1 ml). NADH and NADPH were added at concentrations of 50 to 500 μM.

Determination of oxygen consumption rates and catalase activity.

The oxygen-dependent respiration kinetics of whole cells and in CEs were determined polarographically with a biological oxygen monitor (BOM) (Yellow Springs Instruments Co., Yellow Springs, Ohio). The oxygen concentration was monitored on a paper recorder. The oxygen uptake rates were calculated from the slope of the tangents of the oxygen concentration versus time at various intervals. Catalase activity was determined by adding 2.2 mM H₂O₂ (final concentration) to an anoxic cell suspension in the BOM. The initial oxygen production rate was taken as a measure of catalase activity. The rate was corrected for the rate of chemical decomposition of H₂O₂ in the absence of cells.

Analytical techniques.

Cells were counted microscopically with a Bürker-Türk counting chamber after appropriate dilution of culture samples in basal salt medium containing 1% formaldehyde. Polyglucose in washed cells was determined as glucose equivalents after hydrolysis in 2 N H₂SO₄ for 20 min at 120°C with glucose oxidase (Boehringer GmbH, Mannheim, Germany). Polyglucose was not removed from the cells prior to this determination. Ammonium and sulfide were measured colorimetrically according to Richterich (27) and Trüper and Schlegel (33), respectively. Oxygen concentrations in the input and output gases were analyzed with a gas chromatograph (Pye Unicam 104) equipped with a katharometer (thermal conductivity detector) and a Poropack Q (Waters Associates Inc., Milford, Mass.) 100/120-mesh column (4 mm by 1.2 m) as described by Gerritse et al. (11). Protein in intact cells and in CEs was measured according to the microbiuret method (12) and the Bradford method, respectively, with bovine serum albumin as a standard.

RESULTS

Growth in the presence of oxygen.

The possibility of the growth of D. salexigens Mast1 at the expense of oxygen was studied with cultures that were pregrown under strictly anoxic conditions in both batch and continuous cultures. During the mid-exponential phase of batch cultures on 20 mM alanine, aliquots were distributed over four bottles, of which three were anoxic and one contained air. To two anoxic bottles sterile air was added to provide the cultures with oxygen concentrations of 2 and 5% in the headspace. All bottles were thoroughly shaken during the experiment. At regular intervals, growth and activity were recorded as cell counts and the formation of ammonium resulting from the conversion of alanine. After 20 h, the cells had divided only once at 2% oxygen and alanine was partly consumed (Table 1). At 5% oxygen, only a fraction of alanine was degraded, and at 21% oxygen, no activity could be detected. Since the metabolism of polyglucose in cells of D. salexigens Mast1 was shown to allow oxygen reduction activity for a longer time (35), oxygen-dependent growth was also attempted with cells with a high polyglucose content. For this purpose, the organism was pregrown on a mixture of alanine (10 mM) and glucose (2 mM) in an anoxic chemostat. The effluent of this culture was pumped continuously into a second chemostat that was kept under oxic conditions and was supplemented with a fresh supply of alanine. In spite of the fact that the polyglucose in the cells was partly converted (from 0.8 to 0.5 g of polyglucose g of protein⁻¹), no significant growth, measured as protein, could be detected at oxygen concentrations ranging from 1 to 10% in the gas phase.

TABLE 1.

Growth and alanine degradation in batch cultures of D. salexigens Mast1 with 20 mM alanine after 20 h of exposure to several concentrations of oxygen^a

O2 (%)	Alanine (mM)	Increased cell count (108 cells ml−1)
0	11.7	8.8
2	4.4	2.2
5	0.4	0
21	0	0

^aEach culture was started with 2.2 × 10⁸ cells ml⁻¹. 

Oxygen-dependent respiration.

To study the kinetics of oxygen consumption of D. salexigens Mast1, batch cultures were grown anoxically on pyruvate. The cells were harvested at the end of the exponential growth phase. Washed cell suspensions were aerated and examined for oxygen consumption in the presence of pyruvate. The oxygen reduction rate was highest at air saturation and declined with decreasing oxygen concentrations (Fig. 1A). A similar response was found in the absence of external substrates and with other substrates, such as glycerol and alanine. Identical experiments with other D. salexigens strains, D. gigas, and several D. desulfuricans strains showed similar kinetics and specific oxidation rates (Table 2). Since in an earlier study by Dilling and Cypionka (8) it was shown that highest oxidation rates for several SRB, including D. desulfuricans CSN, occurred below 10 μM dissolved oxygen, we also checked the oxygen uptake activity of D. salexigens Mast1, D. gigas, and D. desulfuricans CSN at low oxygen concentrations. Washed cell suspensions were kept anoxic. D. salexigens Mast1 was given alanine and successive pulses of 20 μM H₂O₂. Because of high catalase activity (Table 3), the oxygen concentration increased quickly, and the initial oxygen uptake rate was determined after each pulse. The oxygen uptake rate increased with increasing oxygen concentration (Fig. 1B). The same result was obtained with anoxic cell suspensions of D. gigas that were sparged with air for only a few seconds. However, when anoxic cell suspensions of D. desulfuricans CSN were sparged with air for a few seconds, the rate of oxidation of lactate increased with decreasing oxygen concentration. Highest oxidation rates were obtained with 0 to 10 μM oxygen (Fig. 1C), confirming the observations of Dilling and Cypionka (8). In the present experiments, the oxygen concentrations were kept below 100 μM.

FIG. 1.

(A) Oxygen uptake rate as a function of oxygen concentration in a fully aerated cell suspension of D. salexigens Mast1 in the presence of 4 mM pyruvate. (B) Oxygen uptake rate in a cell suspension of D. salexigens Mast1 in the presence of 4 mM alanine. The cell suspension was initially anoxic. The oxygen concentration was increased by adding successive pulses of 20 μM H₂O₂. (C) Lactate-dependent oxygen uptake rate in a cell suspension of D. desulfuricans CSN at a low initial concentration of oxygen. (D) Oxygen uptake rate in a fully aerated CE of D. salexigens Mast1 in the presence of 1 mmol of NADH. The rates are expressed in nanomoles of O₂ minute⁻¹ milligram of protein⁻¹.

TABLE 2.

Initial specific oxygen uptake rates in air-saturated suspensions of whole cells and CEs of several Desulfovibrio strains^a

Organism	Oxygen uptake rate (nmol of O2 min−1 mg of protein−1) in:
	Whole cells		CEs + NADH
	Without substrate	With substrate
D. salexigens Mast1	12	40 (Pyruvate)	68
D. salexigens Mast2	10	25 (Pyruvate)	21
D. salexigens DSM 2638	8	20 (Pyruvate)	ND
D. gigas NCIMB 9332	10	38 (Lactate)	28
D. desulfuricans ATCC 27774	14	24 (Lactate)	30
D. desulfuricans CSN	4	40 (Lactate)	ND
D. desulfuricans BH	4	16 (Lactate)	ND

^aUptake rates for whole cells were measured in the presence or absence (endogenous conditions) of the indicated substrate (4 mM). NADH (1 mM) was the substrate for CEs. No oxygen consumption was present in CEs without NADH. ND, not determined. 

TABLE 3.

Maximum specific activities of several enzymes involved in the reduction of O₂ and H₂O₂ in soluble CEs of several Desulfovibrio strains pregrown in anaerobic batch cultures^a

Organism	Polyglucose (g g of protein−1)	Activity (nmol of NADH or NADPH oxidized min−1 mg of protein−1) of:				Catalase activity (nmol of O2 produced min−1 mg of protein−1)
		NADH oxidase	NADPH oxidase	NADH peroxidase	NADPH peroxidase
D. salexigens Mast1	0.5	140	0	0	0	770
D. salexigens Mast2	0.1	40	0	0	0	460
D. salexigens DSM 2638	0.3	40	0	0	0	450
D. gigas NCIMB 9332	0.5	59	0	69	0	4,000b
D. desulfuricans ATCC 27774	0.0	75	45	120	140	280
D. desulfuricans CSN	0.0	29	25	35	38	0
D. desulfuricans BH	0.0	89	24	40	25	0

^aO₂ reduction was measured in air-saturated conditions, and H₂O₂ reduction was measured in both air-saturated and anaerobic conditions. 

^bFrom Bell and LeGall (2). 

Oxygen reduction by CEs.

CEs from anoxic cultures of D. salexigens Mast1 grown on pyruvate or alanine showed oxygen consumption with pyruvate and NADH or with alanine and NADH, respectively. NADH oxidation activity decreased with decreasing oxygen concentration (Fig. 1D). Similar kinetics of NADH oxidation with respect to the oxygen concentration were found with the other strains. All rates were comparable to the substrate-dependent oxygen uptake rates for whole-cell suspensions (Table 2).

All strains were further examined for the presence of polyglucose as a storage polymer and for the initial specific activities of oxidases, peroxidases, and catalase (Table 3). It appeared that the strains could be divided into two groups. One group accumulated polyglucose and contained NADH oxidase activity only (except for D. gigas, which also contained NADH peroxidase activity). The other group did not accumulate polyglucose and contained activities of both NADH and NADPH oxidases and NADH and NADPH peroxidases. All of the organisms of the former group contained catalase activity, while from the latter group only D. desulfuricans ATCC 27774 contained catalase activity.

Measuring the initial NADH consumption rates in air-saturated CEs at different NADH concentrations revealed Michaelis-Menten kinetics for NADH oxidation. The affinity constants for D. salexigens Mast1 and D. gigas were 29.6 and 5.7 μM NADH, respectively.

No oxygen reduction activity was found in the membrane fractions of any of the strains tested, and the activities were identical in both crude and soluble CEs.

Enzyme inactivation.

When we were studying the kinetics of NADH oxidases, it became apparent that the activities declined in the presence of oxygen, although all substrates were present in excess. This finding was observed for CEs of all strains investigated. For example, the decline in oxygen consumption activity of D. desulfuricans CSN in the presence of 1 mM NADH was monitored with the BOM (Fig. 2A). After ca. 1 h, the oxygen concentration had declined from 180 to 13 μM. The activity could be restored neither by aeration nor by additional NADH. It appeared that the enzyme had become inactive. The addition of catalase or mannitol (50 mM) at the beginning of the assay did not influence the rate or extent of inactivation.

FIG. 2.

(A) Time course of the inactivation of NADH oxidase activity in a CE of D. desulfuricans CSN in the presence of 1 mM NADH. Arrows: 1, activity after reaeration of the CE; 2, activity after the addition of more NADH. (B) Time course of the inactivation of oxygen uptake in D. desulfuricans ATCC 27774 under endogenous conditions (•) and in the presence of 4 mM lactate (○). The rates are expressed in nanomoles of O₂ minute⁻¹ milligram of protein⁻¹.

Inactivation kinetics for oxygen reduction may be described as follows (21):

1

where r(O₂) is residual activity, V_max is maximum activity, S_O2 is the concentration of O₂, K_O2 is the saturation coefficient for oxygen, t is time, and k is the inactivation constant (hour⁻¹). Applying this formula to the NADH oxidase of D. desulfuricans CSN resulted in a value for k of 6.84 h⁻¹ (r, 0.989) (Fig. 2A). In similar experiments with soluble and crude CEs of D. salexigens Mast1, values of 0.6 to 1.2 h⁻¹ were found for k. Inactivation was also found with NADPH oxidase, NADH peroxidase, and NADPH peroxidase (results not shown). The addition of flavin adenine dinucleotide (FAD) (300 μmol) at the start of the assay stimulated maximum NADH oxidation activity by a factor of 1.6 to 2.0 but did not affect the process of inactivation. Product inhibition by NAD⁺ was less than 10% at concentrations in the range present in the assay and therefore could be excluded.

Inactivation was not due to instability of the enzyme itself because storage of CEs under oxic conditions in the absence of oxidizable substrate at 0°C overnight resulted in a loss of initial activity of less than 10% (results not shown).

Inactivation of oxygen consumption also occurred with washed cell suspensions. For D. desulfuricans ATCC 27774, the inactivation constants were similar with (k, 4.83 h⁻¹; r, 0.998) and without (k, 4.74 h⁻¹; r, 0.997) 4 mM lactate in the medium (Fig. 2B). It should be emphasized that with lactate, the oxidation rate was 10 times higher than under endogenous conditions, and hence about 10 times more oxygen was consumed in total. For D. salexigens Mast1, the k values ranged from 0.1 to 0.6 h⁻¹. The values depended on the polyglucose content and not on the presence of external substrates (Fig. 3). The decline in the specific oxygen consumption rate for cell suspensions was clearly related to the decline in the specific NADH oxidase activity as measured in cells of cultures continuously exposed to oxygen (Fig. 3).

FIG. 3.

Inactivation of NADH oxidase activity in cells of D. salexigens Mast1 exposed to 10% oxygen in closed bottles that initially contained 0.019 mg of polyglucose mg of protein⁻¹ (100% activity is 80 nmol of NADH min⁻¹ mg of protein⁻¹) (○) and 0.3 mg of polyglucose mg of protein⁻¹ (100% activity is 43 nmol of NADH min⁻¹ mg of protein⁻¹) (•) and inactivation of pyruvate-dependent oxygen uptake by cells of D. salexigens Mast1 exposed to oxygen in the BOM containing 0.034 mg of polyglucose mg of protein⁻¹ (100% activity is 33 nmol of O₂ min⁻¹ mg of protein⁻¹) (□) and 0.04 mg of polyglucose mg of protein⁻¹ (100% activity is 38 nmol of O₂ min⁻¹ mg of protein⁻¹) (▪).

DISCUSSION

Despite its oxygen consumption with external substrates and intracellularly accumulated polyglucose, D. salexigens Mast1 showed no significant oxygen-dependent growth in either batch or continuous cultures. Only with 2% oxygen in the gas phase was this strain able to double its cell numbers (Table 1), a result similar to what had been found for two D. desulfuricans strains (Essex and CSN) and Desulfobacterium autotrophicum DSM (22). In a continuous culture, D. desulfuricans NCIB 8301 cell numbers increased when oxygen was introduced, with a maximum at 1.5% oxygen (1). The disadvantage of introducing oxygen to a single-chemostat system is the risk of washout. This problem was overcome with the two-chemostat system used in this study. The second, oxic chemostat was continuously supplied with active cells containing high quantities of polyglucose. However, even under these conditions, no oxygen-dependent growth of D. salexigens Mast1 was seen.

The strains of D. salexigens, D. desulfuricans, and D. gigas used in this study immediately oxidized various substrates with oxygen upon aeration. Cell suspensions of all strains showed highest oxidation rates at air saturation, but the rates were about 100-fold lower than those observed with aerobic microorganisms (20). The oxidation rates declined with decreasing oxygen concentrations (Fig. 1A). This result is in contrast to that for several D. desulfuricans strains (among them strain CSN) whose highest oxygen uptake activity occurred at oxygen concentrations between 0 and 10 μM (1, 8). This contrast was only apparent because the exposure of anoxic cell suspensions of D. desulfuricans CSN to low oxygen concentrations (Fig. 1C) confirmed the findings of Dilling and Cypionka (8). This finding was not observed with D. salexigens Mast1 (Fig. 1B) and D. gigas. It is likely that at least three independent systems reduce oxygen in D. desulfuricans CSN. One system is active only at low oxygen concentrations and is inhibited by high oxygen concentrations. The other two systems are active at all oxygen concentrations. The former system is probably active in the periplasm and linked to cytochrome c₃ (5a). The latter consists of NADH and NADPH oxidases, as was further indicated by the facts that the oxidation activities of NADH and NADPH were found in soluble CEs at all oxygen concentrations and declined with decreasing oxygen concentrations (Table 3 and Fig. 1D) and that the activities were on the same order as the oxygen consumption rates for whole cells. NADH oxidase activity was present in all strains examined (Tables 2 and 3), and all responded similarly to changes in oxygen concentrations. The initial NADH oxidase specific activities for both O₂ (Table 2) and NADH (Table 3) revealed that the NADH/O₂ ratios were 2 for strain Mast1, strain Mast2, and D. gigas and closer to 1 for D. desulfuricans CSN. These results suggest that the former three organisms contain an NADH oxidase that reduces O₂ directly to H₂O but that the NADH oxidase of the latter organism reduces O₂ to H₂O₂. H₂O₂ can be further reduced to H₂O in this organism because it possesses NADH peroxidase. All of the D. desulfuricans strains examined here contained both NADH and NADPH oxidases and peroxidases (Table 3). Also, these strains did not accumulate polyglucose during growth with various substrates. These two properties clearly distinguish them from the D. salexigens strains and D. gigas.

In all cases, the NADH oxidase activity declined rapidly (ca. 1 h) upon addition of the substrates. The same result was seen with NADPH oxidase, NADH peroxidase, and NADPH peroxidase and with oxygen consumption by whole cells of D. desulfuricans strains (Fig. 2). During oxidation reactions with oxygen, it is possible that reactive intermediates which damage the enzyme are formed. Hence, it is expected that the more and the faster that oxidation is taking place, the more extensive will be the damage and thus inactivation. However, for D. desulfuricans ATCC 27774 (Fig. 2B), oxygen consumption activity in both the presence and the absence of lactate was inactivated at the same rate, even though the oxidation rate with lactate was 10 times higher than that under endogenous conditions and 10 times more oxygen was consumed. Therefore, it is believed that inactivation in this organism is an intrinsic property of the enzymes and their substrates rather than that it is caused by products or by-products. Inactivation only happened when both oxygen and oxidizible substrates were present. This result suggests that the configuration of the enzyme-substrate complex is prone to denaturation. One of the best-studied NADH oxidases is that of Streptococcus faecalis. The NADH oxidase of D. gigas (5) and that of S. faecalis (30) are quite similar in that both contain FAD and no metals and in that they share sensitivity to sulfhydryl agents. Hoskins et al. (16) found that the instability of the NADH oxidase of S. faecalis was due to the gradual removal of FAD. Although the addition of FAD to the CEs used in our study increased the rate of oxidation of NADH, it had no effect on the rate of inactivation.

With whole cells of D. salexigens Mast1 it was found that polyglucose prevented the rapid inactivation of NADH oxidase (Fig. 3). The inactivation constant was taken as a measure of the rate of inactivation. From the limited data available, a tentative relationship between the polyglucose content and the inactivation constant could be visualized (Fig. 4); it seemed that even low polyglucose contents caused substantial deceleration of the rate of inactivation of NADH oxidase. For this organism, the highest inactivation constants were found with CEs and were the same for crude and soluble CEs. Most of the polyglucose (96 to 98%) was detected in the cell debris after centrifugation of the crude extract. This result indicates that the protective property of polyglucose was lost once the cells were broken up.

FIG. 4.

Inactivation constant (k) of NADH oxidase of D. salexigens Mast1 as a function of polyglucose content.

We previously showed that D. salexigens Mast1 could oxidize pyruvate with oxygen only as long as polyglucose was present in the cell (35), and we could only hypothesize about an explanation for this phenomenon. Based on our present findings, it seems likely that the observed dependency originates from protection against inactivation of NADH oxidase by polyglucose. The actual mechanism still remains to be clarified. It is possible that polyglucose somehow stabilizes the enzyme-substrate complex. Interactions between polysaccharides and enzymes have been described before for other organisms (e.g., 9, 17, 23) and it was noted that such interactions altered enzymatic activity, prevented inhibition, or protected against denaturation. However, to our knowledge, no examples of protection against oxygen damage have been reported.

D. salexigens Mast1 was isolated from the oxic-anoxic interface of a microbial mat exposed to cycles of dark-anoxic and light-oxic conditions. The results of this study suggest that the organism uses polyglucose during oxic periods, in which the environment can become supersaturated with oxygen, to keep the cells in a viable state. Polyglucose is replenished during dark periods. This mechanism for surviving unfavorable periods has already been described for other microorganisms in microbial mats, such as cyanobacteria (31) and purple sulfur bacteria (34). It might explain why Desulfovibrio strains were found in large quantities in the top layers of a microbial mat (28). The results also suggest that aerotolerant SRB which do not accumulate polyglucose cannot survive high oxygen concentrations for a long time and hence prefer regions with permanently low or limiting oxygen concentrations. This situation has indeed been found for D. desulfuricans CSN (22).

Our study has revealed that polyglucose not only is used as a carbon and energy source but also protects against the rapid inactivation of NADH oxidases. Further studies should be undertaken to reveal how widely spread such a protective mechanism is among anaerobic microbes occasionally exposed to molecular oxygen.