Synthesis of Nitric Oxide from the Two Equivalent Guanidino Nitrogens ...

0 downloads 0 Views 138KB Size Report
sourdough, fermenting plant materials, manure, sewage, and the mouths and feces of humans. The GC content of the DNA of this species is 52 to 54 mol%, and ...
JOURNAL OF BACTERIOLOGY, Dec. 1997, p. 7812–7815 0021-9193/97/$04.0010 Copyright © 1997, American Society for Microbiology

Vol. 179, No. 24

Synthesis of Nitric Oxide from the Two Equivalent Guanidino Nitrogens of L-Arginine by Lactobacillus fermentum HIDETOSHI MORITA,1* HIROSHI YOSHIKAWA,1 RYOICHI SAKATA,1 YUKIHARU NAGATA,1 AND HIDEHIKO TANAKA2 School of Veterinary Medicine, Azabu University, Sagamihara 229,1 and Faculty of Agriculture, Okayama University, Okayama 700,2 Japan Received 27 May 1997/Accepted 17 September 1997

Ten strains of Lactobacillus fermentum that differed in origin converted metmyoglobin to nitrosylmyoglobin [a pentacoordinate nitric oxide (NO) complex of Fe(II) myoglobin] in MRS broth at pH 4.3. Of the 10 strains, L. fermentum IFO 3956 possessed the strongest capacity to convert metmyoglobin to nitrosylmyoglobin. This strain synthesizes NO enzymatically from the two equivalent guanidino nitrogens of L-arginine. To our knowledge, this demonstrates for the first time the production of NO synthesized from the guanidino nitrogens of L-arginine by lactic acid bacteria. IFO 3956 may possess a bacterial NO synthase. derived specifically from the two equivalent guanidino nitrogens of L-arginine by stable isotope enrichment experiments.

Nitric oxide (NO) is a secretory product of mammalian cells that plays an important role in the regulation of vascular tone, platelet function, neurotransmission, and host defense mechanisms (13, 20, 22). NO is derived from the guanidino nitrogens of L-arginine by NO synthase (NOS), producing L-citrulline as a coproduct. NOS has been purified from macrophages (8, 9, 19) and bovine aortic endothelial cells (17). Although NO biosynthesis through oxidative (nitrification) or reductive (denitrification) pathways has been well studied in microorganisms (6, 15, 23), the formation of NO from L-arginine has rarely been investigated. Because bacteria generally possess a urea cycle or a deiminase pathway for the metabolism of L-arginine, L-arginine is recomposed to L-ornithine and urea by arginase or to L-citrulline and ammonia by arginine deiminase. Chen and Rosazza (2, 3) recently purified NOS from Nocardia sp., and NOS has been identified in only this one microbial species. The ferrous complex of myoglobin containing one molecule of NO is called nitrosylmyoglobin [NO complexed with Fe(II) myoglobin]. Nitrosylmyoglobin is formed by the reaction of myoglobin with NO generated from nitrite in cured meat products. Arihara et al. (1) screened lactic acid bacteria (22 species, 81 strains) for the ability to convert metmyoglobin to nitrosylmyoglobin on MRS agar plates without nitrite and nitrate addition. Although only Lactobacillus fermentum converted metmyoglobin to nitrosylmyoglobin, it was uncertain how L. fermentum formed NO without addition of nitrite and nitrate to the medium. L. fermentum strains are isolated from yeast, milk products, sourdough, fermenting plant materials, manure, sewage, and the mouths and feces of humans. The GC content of the DNA of this species is 52 to 54 mol%, and this species does not possess L-lactate dehydrogenase. L. fermentum is one of the most important lactic acid bacteria contributing to the formation of healthy intestinal flora and is beneficial to its host (10). In the studies reported here, 10 strains of L. fermentum were screened for the capacity to convert metmyoglobin to nitrosylmyoglobin without addition of nitrite and nitrate. We elucidated the enzymatically synthesized NO in L. fermentum

MATERIALS AND METHODS Nitrogen source and bacterial strains. L-[Guanidino-15N2]arginine (15N enrichment, ;99%; Shoko Co.) was used as a nitrogen source. L. fermentum JCM 1560, JCM 2761, IFO 3071, IFO 3956, IFO 3959, NRIC 1598, NRIC 1736, NRIC 1952, and NRIC 1955 were used in this study. Strains were grown in MRS agar (Oxoid, Unipath Ltd.) at 37°C for 72 h and maintained at 4°C for 3 weeks. One strain was grown in 200 ml of MRS broth (Oxoid, Unipath Ltd.) for 16 h and harvested by centrifugation. The following buffers were used for these experiments. Ten millimolar Tris buffers (TB; pH 7.5) containing 0.1 mM L-arginine and L-[guanidino-15N2]arginine were designated TBA14 and TBA15, respectively. TBA14 and TBA15 containing 0.1 mM NADPH, 80 mM (6R)-5,6,7,8tetrahydrobiopterin (H4B), 10 mM flavin adenine dinucleotide (FAD), and 10 mM flavin mononucleotide (FMN) were designated TBA14C and TBA15C, respectively. Intact cells and the cell extract of L. fermentum were suspended in TBA14, TBA15, TBA14C, and TBA15C and then incubated for 6 h at 37°C. The cells were disrupted for 90 s at 77 K by a Cryo-Press (CP-100W, MacrotechNichion Co., Ltd.). Capacity for metmyoglobin conversion to nitrosylmyoglobin. The strains were examined for the ability to convert metmyoglobin to nitrosylmyoglobin in MRS broth as previously described (1, 12). MRS broth supplemented with 2.0 mg of metmyoglobin (originating from equine heart; Sigma Chemical Co.) per ml was designated MRS-Mb broth. Nitrosylmyoglobin was detected by the method of Okayama and Nagata (14) and by electron spin resonance (ESR) analysis. ESR. Each culture fluid (400 ml) was directly transferred to an ESR tube. ESR spectra were recorded on an ESR spectrometer (JES-TE2X; JEOL Co., Ltd.) under the following conditions: microwave power, 4 mW; modulation frequency and width, 100 kHz and 1.0 mT; temperature, 77 K; measurement time, 8 min. MRS broth without MnSO4 was used for ESR analysis. Analysis of NO22 and NO32. NO22 and NO32 concentrations in the supernatant samples were determined by using a Nitrate/Nitrite Assay Kit (Cayman Chemical Co.) based on the Griess reaction. The final products of NO in vitro are NO22 and NO32. The relative proportions of NO22 and NO32 are variable and cannot be predicted with certainty. Thus, the best index of total NO production was the sum of NO22 and NO32. The total concentration of NO22 and NO32 was measured in a simple two-step process with this kit. The first step was the conversion of NO32 to NO2 utilizing nitrate reductase. The second step was addition of the Griess reagents, which convert NO2 into a deep purple azo compound. Protein determination. Protein was measured by the method of Lowry et al. (11). Bovine serum albumin (fraction V) was used as the standard. 15 N enrichment of NO22 and NO32. 15N enrichment of NO22 and NO32 in cell culture supernatants was measured by gas chromatography-mass spectrometry (GC-MS) after converting benzene to nitrobenzene by the procedure of Tesch et al. (21). Addition of 1 M H2O2 (0.3 ml) and benzene (5 ml) to the cell culture supernatant (1 ml) was followed by the slow addition of concentrated H2SO4 (5 ml). The solution was shaken at room temperature for 10 min. Samples (1 ml) were injected into a Fisons Instruments MD-800 GC-MS system. A Fisons Instruments GC8000 system equipped with an SPB-5 capillary column (30 m by 0.32 mm [inside diameter]; 0.25-mm film thickness) and coupled to a mass spectrometer (Fisons Instruments MD-800) by direct interface at 280°C was used. The following two-step temperature program was applied: (i) an initial

* Corresponding author. Mailing address: School of Veterinary Medicine, Azabu University, 1-17-71 Fuchinobe, Sagamihara 229, Japan. Phone: (427) 54 7111. Fax: (427) 53 3395. E-mail: morita @azabu-u.ac.jp. 7812

VOL. 179, 1997

SYNTHESIS OF NO FROM L-ARGININE BY LACTOBACILLI

7813

TABLE 1. Effect of L-arginine, NADPH, H4B, FAD, and FMN on NO22 and NO32 production by a cell suspension and a cell extract of L. fermentum IFO 3956 incubated for 6 h at 37°C Buffer componenta

Amt of NO22 and NO32 Cell suspension (nmol/mg [dry wt] of cells)

Cell extract (nmol/mg of protein)

7.4 7.6 12.0

35.0 35.3 52.7

TB TBA14 TBA14C a

FIG. 1. ESR spectra of metmyoglobin and a pentacoordinate nitrosylmyoglobin in MRS broth without manganese sulfate at 77 K. A, metmyoglobin (Sigma Chemical Co.); B, nitrosylmyoglobin converted from metmyoglobin by L. fermentum IFO 3956 incubated for 16 h at 37°C. g, g factor (spectroscopic splitting constant).

column temperature of 60°C held for 1 min and then a programmed increase to 120°C at a rate of 8°C/min and (ii) a final column temperature of 250°C reached by means of a temperature increase of 20°C/min. The injector temperature was 250°C, and the inlet pressure was 40 kPa. The nitrobenzene emerged from the column at 6.3 min. The ratio of [15N]nitrobenzene to [14N]nitrobenzene was determined by selected ion recording at m/z 123 and 124 as reported by Green et al. (7).

RESULTS AND DISCUSSION All 10 strains of L. fermentum tested were capable of converting metmyoglobin to red myoglobin derivatives in MRSMb broth. The visible spectra of the acetone-extracted pigment from the MRS-Mb cultures showed four absorption peaks at 395, 476, 535, and 563 nm, which is the typical pattern of nitrosylmyoglobin (14). Each red myoglobin derivative was identified as nitrosylmyoglobin, which is present in cured meat products. Of the 10 strains, L. fermentum IFO 3956 was used in the following experiments because it was found to possess the strongest metmyoglobin conversion capacity. As IFO 3956 converted metmyoglobin to nitrosylmyoglobin, myoglobin was used as a spin-trapping agent for endogenous NO. Duprat et al. (5) reported that in the native form of myoglobin, protoheme is bound to the protein by a proximal histidine; however, upon activation by reaction with NO, the proximal base-to-heme bond is thought by many to be broken, so that the prosthetic group remains in the protein matrix without any covalent or coordinate bond between the protein and the heme. The iron(II) form of myoglobin forms a hexacoordinated NO complex at around neutral pH. At low pH, breaking or stretching of the proximal imidazole-heme bond has been reported for Aplysia limicina nitrosylmyoglobin, suggesting that pH lowering and NO binding act synergistically to weaken the imidazole bond (5). Since it has been shown that the pentacoordinated forms of several heme proteins (e.g., myoglobin, hemoglobin, and horseradish peroxidase) lose the proximal base at sufficiently low pH, they should lose the proximal base more easily after binding NO, a fact confirmed by appearance of the three-line ESR spectrum for a pentacoordinate nitrosylmyoglobin at pHs of ,5 (5, 16). After IFO 3956 was incubated in MRS-Mb broth (pH 6.3) at 37°C for 16 h, the pH of the culture decreased to 4.3. The ESR spectrum of metmyoglobin in MRS broth is shown in Fig. 1A, and no signal can be seen at a g factor of around 2. The ESR spectrum of

For definitions of the abbreviations, see Materials and Methods.

MRS-Mb culture of IFO 3956 showed three g factors with a rhombic symmetry of a g factor of around 2 (Fig. 1B) and a typical pentacoordinate nitrosylmyoglobin. The well-resolved triplet in the g 5 2 region of the spectrum is due to hyperfine splitting of the 14N nuclei of NO with an unpaired electron. Experiments were carried out to determine the amounts of NO22 and NO32 produced by the cell suspension and the cell extract of IFO 3956. Table 1 shows the effects of L-arginine and NADPH, H4B, FAD, and FMN on NO22 and NO32 production by the cell suspension and the cell extract of IFO 3956 incubated for 6 h at 37°C. In mammals and macrophages, L-arginine is found to be a precursor of NO production while NADPH, H4B, FAD, and FMN are required cofactors for NOS activity (13). Although both a cell suspension and a cell extract of IFO 3956 also produced NO22 and NO32 without addition of L-arginine, addition of NADPH, H4B, FAD, and FMN resulted in the release of more NO. We guess the reason why the bacterial cells did not require addition of L-arginine to produce NO22 and NO32 is that the cells took enough Larginine from the MRS broth. To determine which nitrogen atom in L-arginine was converted to NO22 or NO32, the cell suspension and the cell extract of IFO 3956 were incubated in the presence of L[guanidino-15N2]arginine. The NO22 or NO32 in the mixture supernatants was converted to nitrobenzene. The 15N enrichment of the products was determined by GC-MS as shown in Table 2. In this experiment, the general reagent nitrobenzene (Nacalai Tesque, Inc.) was used as the standard and the GC-MS retention time was measured. As shown by the darkened areas in Fig. 2A and B, nitrobenzene was found at approximately 6.3 min. Fig. 2C and D show the result of the measurement using one of the samples listed in Table 2 (cell suspension containing L-[guanidino-15N2]arginine incubated for 3 h) under the same conditions. Although a liquid culture

TABLE 2. GC-MS analysis for 15N enrichment of nitrobenzene derived from 15NO22 and 15NO32 synthesized from L[guanidino-15N2]arginine by a cell suspension and a cell extract of L. fermentum IFO 3956 Buffer componenta

Incubation time (h)

TBA14 1 cell suspension TBA15 1 cell suspension TBA15 1 cell suspension TBA15 1 cell extract TBA15 1 cell extractb TBA15C 1 cell extractb

3 3 6 3 3 3

a

15

N enrichment of nitrobenzene (%)

6.93 7.89 7.71 7.54 7.60 7.69

For definitions of the abbreviations, see Materials and Methods. IFO 3956 precultured in MRS broth containing 0.3% L-[guanidinoN2]arginine for 16 h at 37°C. b

15

7814

MORITA ET AL.

J. BACTERIOL.

FIG. 2. GC-MS of commercial nitrobenzene (selected ion recording at m/z 124 [A] and 123 [B], used as a standard), nitrobenzene (selected ion recording at m/z 124 [C] and 123 [D]), reacted benzene, and NO produced by L. fermentum IFO 3956. rt, retention time (minutes).

system both with and without nitrite was used, NO formation was still observed. Nitrobenzene consists of six carbon, five hydrogen, one nitrogen, and two oxygen atoms. In general, stable isotopes of these elements are known to exist in the natural world at the following ratios: 12C:13C, 98.90:1.10; 1H:2H, 99.985:0.015; 14N: 15 N, 99.634:0.366; 16O:17O, 99.962:0.038. By the measurement method used in this experiment, when more than one of the nitrobenzene components is permuted to its stable isotope, it is detected as nitrobenzene with a stable isotope. When nitrobenzene consists of 12C, 1H, 14N, and 16O, a rate of 92.72% is obtained through the following calculation: 0.9896 3 0.999855 3 0.99634 3 0.997622 3 100. The result obtained by subtracting the result (92.72%) from 100% (7.28%) is the calculated rate, showing that one element in nitrobenzene was permuted into a stable isotope. As shown in Table 2, when we measured nitrobenzene three times as the standard, the average result was 6.94%. All of the data shown later in this text are averages of two measurements. NO in solution has very high reactivity, and most of it turns into NO22 in an aerobic environment. When benzene was added to a solution containing NO22 under the conditions described in Materials and Methods, specific formation of nitrobenzene was observed (19). Assuming that the resulting nitrobenzene was originally from NO produced in the liquid culture, we performed another experiment. First, for IFO 3956, nitrobenzene was prepared by using a cell suspension and the general reagent L-arginine (Nacalai

Tesque, Inc.) was added. The stable isotope percentage contained in the nitrobenzene prepared earlier was 6.93%; this agrees with the stable isotope percentage in the nitrobenzene used as the standard. Next, as shown in Table 2, we performed the experiment under five different conditions with L-[guanidino-15N2]arginine as the substrate; however, since data collected from experiments conducted under equal conditions are considered quantitative, we consider it a separate issue. For the cell suspension, incubation times of 3 and 6 h were used. Very little difference in the amount of nitrobenzene produced was observed, suggesting that the reactivity of the bacterial cells is already lost at 3 h. The (m/z 124)/(m/z 123) increased from 7.71% at 6 h to 7.78% at 3 h, as a result of the increase in 15N originating from L-[guanidino-15N2]arginine. The incubation time for the cell extract was 3 h, and the amount of nitrobenzene produced was about 11 times higher in the cell extract than in the cell suspension. But the (m/z 124)/(m/z 123) was 7.57% in the cell extract and was therefore hardly different from that in the cell suspension. The influence of NADPH, H4B, FAD, and FMN was then examined by using a cell extract. About twice the amount of nitrobenzene was produced when NADPH, H4B, FAD, and FMN were added, showing that the presence of NADPH, H4B, FAD, and FMN is as important as the presence of NOS in mammals. The (m/z 124)/(m/z 123) was 7.60% when NADPH, H4B, FAD, and FMN were added and 7.69% without NADPH, H4B, FAD, and FMN. Therefore, we consider that the significant increase

VOL. 179, 1997

SYNTHESIS OF NO FROM L-ARGININE BY LACTOBACILLI

in the rate of permutation of one element to a stable isotope in nitrobenzene by 0.63 to 0.95% compared to the rate in nature clearly occurred because 15N originating from L-[guanidino15 N2]arginine turned into 15NO and nitrobenzene was thus formed. Although L-guanidino-15N2 (15N enrichment, ;99%) was used as the substrate, the stable isotope permutation rate went up to only 0.63 to 0.95%. In the case of lactic acid bacteria, L-arginine metabolism does not occur in the presence of a glucose source (4, 18) and L-arginine taken in by bacterial cells is generally in peptide form; thus, it is speculated that L-arginine is not likely to be taken in directly by bacterial cells in the form of L-[guanidino-15N2]arginine. Some lactic acid bacteria have an L-arginine metabolic system, the so-called arginine deiminase pathway. In this system, L-arginine is converted into L-citrulline and then to ornithine, producing NH3 and ATP as metabolites. In metabolisms in which NOS is involved, such as those of mammals, NOS is similarly converted into citrulline. Further investigations into the relationship between the arginine deiminase pathway and the presence of NOS in IFO 3956 are necessary. The observed 15N enrichment in nitrobenzene showed that one or both of the two terminal guanidino nitrogens of Larginine were being converted to NO22 or NO32 by the bacterial cells. Iyengar et al. (9) reported that the following compounds were found to be precursors, with the results expressed as a percentage with L-arginine as 100%: L-homoarginine (80%), L-arginine methyl ester (82%), L-arginamide (72%), and the peptide L-arginyl-L-aspartate (84%). D-Arginine, Largininic acid, L-agmatine, L-ornithine, urea, L-citrulline, ammonia, and L-canavinine were among the nonprecursors. They discovered that one or both of the two terminal guanidino nitrogens of L-arginine were being converted to NO22 or NO32 by macrophage NOS (9). From this, we could assume that the putative NOS from L. fermentum might be similar to the macrophage and mammalian NOSs (9, 13, 21) because of NO production from the two terminal guanidino nitrogens of L-arginine and the same cofactor requirements. Chen and Rosazza (2, 3) suggested that the stoichiometry of NO formed versus L-arginine used and the inhibition of Nocardia NOS by NG-nitro-L-arginine, which is a typical reversible inhibitor of mammalian NOSs, were similar to those observed in mammalian NOS systems. As L. fermentum IFO 3956 produced NO without a nitrification or denitrification pathway, IFO 3956 might possess a bacterial NOS and further work is in progress to purify the NOS of L. fermentum. To our knowledge, this demonstrates for the first time the production of NO synthesized from the guanidino nitrogens of L-arginine by lactic acid bacteria. REFERENCES 1. Arihara, K., H. Kushida, Y. Kondo, M. Ito, J. B. Luchansky, and R. G. Cassens. 1993. Conversion of metmyoglobin to bright red myoglobin derivatives by Chromobacterium violaceum, Kurthia sp., and Lactobacillus fermen-

7815

tum JCM 1173. J. Food Sci. 58:38–42. 2. Chen, Y., and J. P. N. Rosazza. 1994. A bacterial nitric oxide synthase from a Nocardia species. Biochem. Biophys. Res. Commun. 203:1251–1258. 3. Chen, Y., and J. P. N. Rosazza. 1995. Purification and characterization of nitric oxide synthase (NOSNoc) from a Nocardia species. J. Bacteriol. 177: 5122–5128. 4. Cunin, R., N. Glansdorff, A. Pierard, and V. Stalon. 1986. Biosynthesis and metabolism of arginine in bacteria. Microbiol. Rev. 50:314–352. 5. Duprat, A. F., T. G. Traylor, G.-Z. Wu, M. Coletta, V. S. Sharma, K. N. Walda, and D. Magde. 1995. Myoglobin-NO at low pH: free four-coordinated heme in the protein pocket. Biochemistry 34:2634–2644. 6. Goretski, J., and T. C. Hollocher. 1990. The kinetic and isotopic competence of nitric oxide as an intermediate in denitrification. J. Biol. Chem. 265:889– 895. 7. Green, L. C., D. A. Wagner, J. Glogowski, P. L. Skipper, J. S. Wishnok, and S. R. Tannenbaum. 1982. Analysis of nitrate, nitrite, and [15N]nitrate in biological fluids. Anal. Biochem. 126:131–138. 8. Hevel, J. M., K. A. White, and M. A. Marletta. 1991. Purification of the inducible murine macrophage nitric oxide synthase. Identification as a flavoprotein. J. Biol. Chem. 266:22789–22791. 9. Iyengar, R., D. J. Stuehr, and M. A. Marletta. 1987. Macrophage synthesis of nitrite, nitrate, and N-nitrosamines: precursors and role of the respiratory burst. Proc. Natl. Acad. Sci. USA 84:6369–6373. 10. Kandler, O., and N. Weiss. 1986. Regular, nonsporing gram-positive rods, p. 1208–1234. In P. H. A. Sneath, N. S. Mair, M. E. Sharpe, and J. G. Holt (ed.), Bergey’s manual of systematic bacteriology, vol. 2. The Williams & Wilkins Co., Baltimore, Md. 11. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265–275. 12. Morita, H., R. Sakata, S. Sonoki, and Y. Nagata. 1994. Metmyoglobin conversion to red myoglobin derivatives and citrate utilization by bacteria obtained from meat products and pickles for curing. Anim. Sci. Technol. 65:1026–1033. 13. Nathan, C. 1992. Nitric oxide as a secretory product of mammalian cells. FASEB J. 6:3051–3064. 14. Okayama, T., and Y. Nagata. 1979. A method for the determination of total heme pigments in meat products. Jpn. J. Zootech. Sci. 50:15–21. 15. Papen, H., R. von Berg, I. Hinkel, B. Thoene, and H. Rennenberg. 1989. Heterotrophic nitrification by Alcaligenes faecalis: NO22, NO32, N2O, and NO production in exponentially growing cultures. Appl. Environ. Microbiol. 55:2068–2072. 16. Pegg, R. B., F. Shahidi, N. J. Gogan, and S. I. DeSilva. 1996. Elucidation of the chemical structure of preformed cooked cured-meat pigment by electron paramagnetic resonance spectroscopy. J. Agric. Food Chem. 44:416–421. 17. Pollock, J. S., U. Forstermann, J. A. Mitchell, T. D. Warner, H. H. Schmidt, M. Nakane, and F. Murad. 1991. Purification and characterization of particulate endothelium-derived relaxing factor synthase from cultured and native bovine aortic endothelial cells. Proc. Natl. Acad. Sci. USA 88:10480– 10484. 18. Poolman, B., A. J. M. Driessen, and W. N. Konings. 1987. Regulation of arginine-ornithine exchange and the arginine deiminase pathway in Streptococcus lactis. J. Bacteriol. 169:5597–5604. 19. Stuehr, D. J., H. J. Cho, N. S. Kwon, M. F. Weise, and C. F. Nathan. 1991. Purification and characterization of the cytokine-induced macrophage nitric oxide synthase: an FAD- and FMN-containing flavoprotein. Proc. Natl. Acad. Sci. USA 88:7773–7777. 20. Stuehr, D. J., and O. W. Griffith. 1992. Mammalian nitric oxide synthases. Adv. Enzymol. Relat. Areas Mol. Biol. 65:287–346. 21. Tesch, J. M., W. R. Rehg, and R. E. Sievers. 1976. Microdeterminations of nitrates and nitrites in saliva, blood, water, and suspended particulates in air by gas chromatography. J. Chromatogr. 126:743–755. 22. Vane, J. R., and R. M. Botting. 1992. Secretory functions of the vascular endothelium. J. Physiol. Pharmacol. 3:195–207. 23. Ye, R. W., B. A. Averill, and J. M. Tiedje. 1994. Denitrification: production and consumption of nitric oxide. Appl. Environ. Microbiol. 60:1053–1058.