5&3 . v5 EIEri i g, a I 52:“ W is E w swim» '! .5 «é» 5% 1. {~71 ,. mi a. (’1 M‘ 156311109! “ICHIG‘AN STATE ERSUTY LIBRARIE ”W T"? \I\\\\\\\\\\\\\u\m“111mm um um : Michigan State 31293906299386 i University This is to certify that the dissertation entitled Immunological Studies 0n Dissimilatory Nitrite Reductases in Denitrifying Bacteria presented by Mark Steven Coyne has been accepted towards fulfillment of the requirements for Ph.D. d . Crop and Soil Sciences egree 1n M jor professor Date 8/1749? MS U it an Affirmative Action/Equal Opportunity Institution PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES 'return on or baton due due. DATE DUE DATE DUE DATE DUE i \F W MSU Is An Affirmative AetlorVEqual Opportunity Institmion l x IMMUNOLOGICAL STUDIES ON DISSIMILATORY NITRITE REDUCTASES IN DENITRIFYING BACTERIA BY Mark Steven Coyne . A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements ' for the degree of DOCTOR OF PHILOSOPHY Department of Crop and Soil Sciences 1989 ABSTRACT IMMUNOLOGICAL STUDIES ON DISSIMILATORY NITRITE REDUCTASES IN DENITRIFYING BACTERIA BY Mark Steven Coyne The distribution, divergence, localization, and induction of .dissimilatory nitrite reductases were studied in denitrifying bacteria. Polyclonal antibodies were raised against heme 931 nitrite reductase from Pseudomonas aeruginosa and from Pseudomonas stutzeri, and against Cu nitrite reductase from Achromobacter cycloclastes. Nitrite reductases were identified by Western immunoblot. Diethyldithiocarbamate, which specifically inhibits Cu nitrite reductases, was used to confirm the immunological characterization and determine which type was present in these strains which were nonimunoreactive. Nitrite reductases were characterized in the following groups: 23 taxenomically diverse denitrifiers from culture collections, 100 numerically dominant denitrifiers from geographically diverse environments, and 51 denitrifiers from a culture collection not selected for denitrificatien. Heme £11 nitrite reductases were the most common type found in environmental samples, and appeared more conserved if judged by similiarity in molecular weights and reaction with antisera, The divergence among heme £11 nitrite redUCtases and Cu nitrite reductases was examined by immunodiffusion and immuneblotting techniques. Two groups of heme 9411 type nitrite reductases, exemplified by Pseudomonas aeruginosa and Pseudomonas stutzeri , were observed. Achromobacter cycloclastes Cu nitrite reductase appeared distantly related to the other Cu type nitrite reductases tested. Nitrite reductases were localized in 9511- and Cu-type denitrifiers by immogold labeling. In Achromobacter cycloclastes and Achromobacter xylosoxidans the Cu nitrite reductase was found in the periplasmic space as was the 9511 nitrite reductase of Pseudomonas aeruginosa and Pseudomonas fluorescens. The induction of denitrifying enzyme synthesis and activity was studied in Achromobacter cycloclastes continuous cultures gradually deprived of oxygen. Nitrate and nitrous oxide reductases were synthesized at all oxygen concentrations but full induction only occurred at 2 uM oxygen. Nitrite reductase was neither synthesized nor active until dissolved oxygen was below 2 uM. DEDICATION This is for Sharon, Pepe, and Dad iv ACKOWLEDGEMENTS I would like to thank my advisor Dr. Jim Tiedje and all the members of my graduate committee for their guidance and patience during the PaSt five years. I would especially like to thank all the members of the MSU community: students, graduate students, post does, professorS, and research technicians, who assisted me with my program. I may have done most of the work, but you told me what it meant. TABLE OF CONTENTS Page LIST OF TABLES .................................................... vii LIST OF FIGURES ................................................... ix INTRODUCTION ...................................................... 1 SECTION 1. IMMUNOLOGICAL IDENTIFICATION, DISTRIBUTION, AND DIVERCENCE OF DISSIMILATORY HEME 921 AND NONHEME CU NITRITE REDUCTASES IN DENITRIFYING BACTERIA. ........... 15 INTRODUCTION ........................................ 16 METHODS AND MATERIALS ............................... 19 RESULTS ............................................. 26 DISCUSSION .......................................... 53 SECTION 2. LOCALIZATION OF HEME 92 AND CU NITRITE REDUCTASES IN DENITRIFYING BACTERIA ................... 62 INTRODUCTION ........................................ 63 METHODS AND MATERIALS ............................... 65 RESULTS ............................................. 68‘ DISCUSSION .......................................... 80 SECTION 3. INDUCTION OF N-OXIDE REDUCTASES IN OXYGEN-LIMITED CHEMOSTAT CULTURES OF AQBBQMQBA§I£B_§X§LQ§LA§IE§ ...... 84 INTRODUCTION ........................................ 85 METHODS AND MATERIALS ............................... 86 RESULTS ............................................. 91 DISCUSSION .......................................... 98 APPENDIX.A. PRELIMINARY EFFORTS TO CLONE CU-TYPE NITRITE REDUCTASE ..................................... 103 LIST OF REFERENCES ................................................. 110 vi LIST OF TABLES Table Page INTRODUCTION 1 Selected physical characteristics of Cu-dNirs ................ 6 SECTION 1 1 western immunoblot response of denitrifiers to polyclonal antibodies against Cu- and ggl-dNirs .............. 33 2 Immunochemical classification of nitrite reductase type in denitrifying isolates from sediments, sludges, and the soils of six countries ................................... 34 3 Effect of DDC on nitric and nitrous oxide evolution by denitrifier crude extracts for which nitrite reductase type was classified by immunoblots ........................... 41 4 Effect of DDC on nitric and nitrous oxide evolutibn by denitrifier crude extracts unreactive to antisera against Cu- or ggl-dNirs ..................................... 42 S In 21:12 activity of nitrite reductases in denitrifying bacteria used for immunoblot analysis ....................... 45 6 Measurement of antibody binding to Cu-dNirs in denitrifying cell lysates ................................... 49 7 Measurement of antibody binding to ggl-dNirs in denitrifying cell lysates ................................... 50 8 Distribution of Cu- and gdl-dNirs ........................... 53 vii Table Page SECTION 2 1. Cellular distribution of enzyme activities .................. 69 SECTION 3 1 Effect of different substrates and inhibitors on the protein yield of semianaerobically incubated A. cycloclastes ............................................. 91 2 Denitrification by A. cycloclastes during continuous culture ..................................................... 94 3 In 2113 activity of N-oxide reductases in A. cycloclastes washed cells removed from continuous culture at varying air saturations ................................................. 96 viii Figure LIST OF FIGURES Page SECTION 1 Silver stain of proteins used in rabbit immunizations. Lane A, A. cycloclastes Cu-dNir; B, P. stutzeri ggl-dNir; C, A. cycloclastes N20 reductase; D, P. stutzeri N20 reductase ................................................. 27 Comigration of dNir activity with immunologically recognized proteins in denitrifier crude extracts. A. Nitrite reductase activity stain; B. Western immunoblot with Cu- and gfil-dNir antisera. Lane 1, A. cycloclastes ATCC 21921; 2, P. aeruginosa ATCC 10145. Arrows indicate the location of comigrating proteins .............................. 28 Removal of dNir activity by Cu-dNir antisera. Samples in lanes 1-6 were preincubated with Cu-dNir antisera (1,3,5) or buffer (2,4,6). Lanes 1,2, A. cycloclastes Cu-dNir (1 ug); 3,4, A. cycloclastes ATCC 21921 crude extract; 5,6, P. aeruginosa ATCC 10145 crude extract. Arrows indicate where Cu-dNir activity but not 9d -dNir activity has been lost after preincubation with Cu-dNir antisera ................ 30 Western immunoblots of denitrifier crude extracts separated on 10% SDS-polyacrylamide gels. A. Immunoblotted with A. cycloclastes ATCC 21921 Cu-dNir antisera; B. Immunoblotted with P. aeruginosa ATCC 10145 cd -dNir antisera. Lane 1, A. cycloclastes ATCC 21921 puri ied Cu-dNir (1 ug); 2, A. cycloclastes ATCC 21921 crude extract; 3, P. aeruginosa ATCC 10145 crude extract; 4. P. stutzeri JM300 crude extract; 5, P. stutzeri JM300 purified ggl-dNir (1 ug). Molecular mass indicators are in kDa .................. 31 ix Figure Page 5 Western immunoblot of denitrifier crude extracts separated on 10% SDS-polyacrylamide gels and immunoblotted with Cu-dNir antisera. Lane 1, P. aeruginosa ATCC 10145; 2, #41 A. faecalis; 3, #40 A. faecalis; 4, #15 P. fluorescens; 5, #39 Unknown type 3; 6, #171 A. faecalis; 7, #31 A. faecalis; 8, #3 A. faecalis; 9, #43 A. faecalis; 10, #36 Unknown type 3; 11, #144 A. faecalis; 12, #224 P. stutzeri; 13, #192 Bacillus sp.; 14, #106 A. faecalis; 15, #177 A. faecalis; 16, #191 A. faecalis; 17, #224 P. stutzeri; 18, #167 P. aeruginosa; 19, #62 P. aureofaciens; 20, A. cycloclastes ATCC 21921. Arrows show the location of Cu-dNirs. Strain numbers (#) are from Gamble et a1. (1977) ................................ 39 6 Immunodiffusion analysis with P. aeruginosa ggl-dNir antiserum. Center wells contain undiluted P. aeruginosa ggl-dNir antiserum. Outer wells contain crude extract from each of the ggl-type strains as indicated in the figure ........................................................ 43 7 Western immunoblot of denitrifier crude extracts containing Cu-dNirs incubated with A. cycloclastes Cu-dNir antiserum. Lane 1, A. cycloclastes ATCC 21921; 2, A. xylosoxidans NCIB 11015; 3, P. aureofaciens ATCC 13985; 4, A. faecalis #40; 5, A. faecalis #43; 6, A. eutrophus, #154. Arrows point to Cu-dNirs ............................................. 46 8 Western immunoblot of denitrifier crude extracts containing ggl-dNirs incubated with P. aeruginosa cg -dNir antiserum. Lane 1, P. aeruginosa ATCC 10145; 2, P. f uorescens ATCC 33512; 3, P. stutzeri JM300; 4, Isolate 2.4.E; 5, Isolate 1.6.L. The arrow points to a characteristic protein recognized in P. fluorescens and Isolate 1.6.L. Molecular mass indicator is in kDa ............................ 47 9 Western immunoblot of denitrifier crude extracts containing gfll-dNirs incubated with P. stutzeri gfil-dNir antiserum. Lane.1, P. aeruginosa ATCC 10145; 2, P. fluorescens ATCC 33512; 3, P. stutzeri JM300; 4, Isolate 2.4.E; 5, Isolate 1.6.L. The arrow points to the location of P. stutzeri cfil-dNir ......................................... 48 10 Phylogeny of bacteria which crossreact with antiserum to A. cycloclastes Cu-dNir ....................................... 51 11 Phylogeny of bacteria which crossreact with antiserum to P. aeruginosa ggl-dNir ........................................ 52 Figure Page SECTION 2 1 Electron micrograph of aerobically-grown A. cycloclastes incubated with Cu-dNir antiserum and treated with 15 nm colloidal gold probes (magnification 48,450X) ................ 71 2 Electron micrograph of anaerobically-grown A. cycloclastes incubated with Cu-dNir antiserum and treated with 15 nm colloidal gold probes (magnification 66,300X). Arrow points to periplasmic location of the gold particles ................. 72 3 Electron micrograph of anaerobically-grown A. cycloclastes incubated with gdl-dNir antiserum and treated with 15 nm colloidal gold probes (magnification 48,450X) ................. 73 4 Electron micrograph of aerobically-grown A. xylosoxidans incubated.with Cu-dNir antiserum and treated with 15 nm colloidal gold probes (magnification 48,450X) ................. 74 5 Electron micrograph of anaerobically-grown A. xylosoxidans incubated with Cu-dNir antiserum and treated with 15 nm colloidal gold probes (magnification 48,450X). Arrow points to localization of the gold particles in the periplasmic space ......................................................... 75 6 Electron micrograph of aerobically-grown P. aeruginosa incubated with gfil-dNir antiserum and treated with 15 nm colloidal gold probes (magnification 48,450X) ................. 76 7 Electron micrograph of anaerobically-grown P. aeruginosa incubated with ggl-dNir antiserum and treated with 15 nm colloidal gold probes (magnification 48,450X). Arrow points to localization of the gold particles around the cytoplasmic membrane.. .................................................... 77 8 Electron micrograph of aerobically-grown P. fluorescens incubated with gfil-dNir antiserum and treated with 15 nm colloidal gold probes (magnification 48,450X) ................. 78 9 Electron micrograph of anaerobically-grown P. fluorescens incubated with gfil-dNir antiserum and treated with 15 nm colloidal gold probes (magnification 88,400X). Arrow points to localization of gold particles in the periplasmic space.... 79 10 Anaerobic respiratory chain of P. denitrificans and its proton-translocating properties. Fp, flavoprotein; Fe-S, iron sulfur center; QHZ, ubiquinol; Hy, hydrogenase; NR, nitrate reductase; cyt., cytochrome; N O-R, nitrous oxide reductase; Ox, alternate oxidase. I and I a + IIb are the traditional sites of energy conservation (From F.C. Boogerd, Ph.D. Thesis, Free University of Amsterdam, 1984) ........................... 83 xi Figure Page SECTION 3 1 Synthesis of A. cycloclastes Cu-dNir in the indicated treatments. C - Control, 1 ug of A. cycloclastes Cu-dNir; W - NazWOA; DDC - Diethyldithiocarbamate; -N - No N-oxide terminal electron acceptors ................................... 92 2 Western immunoblot of Cu-dNir synthesis in A. cycloclastes continuous cultures during increasing anaerobiosis. C - Control, 1 ug A. cycloclastes Cu-dNir; FI - Fully induced culture. Values indicate % air saturation. Corresponding dissolved oxygen concentrations are given in Table 3. Marker indicates A. cycloclastes Cu-dNir ............................ 95 3 Western immunoblot of N20 reductase synthesis in A. cycloclastes continuous cultures during increasing anaerobiosis. C - Control, 1 ug A. cycloclastes N 0 reductase; FI - Fully induced culture. Values indicate % air saturation. Corresponding dissolved oxygen concentrations are given in Table 3. Marker indicates N20 reductase ............. 97 APPENDIX A 1 N-terminal sequence of Cu-dNir from Achromobacter cycloclastes ATCC 21921 .................................... .. 104 2 Synthetic oligonucleotides derived from the N-terminal amino acid sequence of A. cycloclastes Cu-dNir. N - A+G+C+T.. 105 3 The N-terminal amino acid sequence of heme 5g -dNir from P. stutzeri JM300. XXX indicates that the amino acid is not known ................................................. 107 4 PCR oligonuleotide primers derived from the N-terminal amino acid sequence of P. stutzeri ggl'dNir. S - G+C; Y - T+C ............................................. 108 5 PCR oligenucleotide primers derived from the N-terminal amino acid sequence of A. cycloclastes Cu-dNir. S - G+C ...... 108 6 PCR oligonucleotides derived from the N-terminal amino acid sequence of P. stutzeri ggl-dNir. H - A+C+T; N - A+G+C+T; R - G+A; V - A+G+C; Y - T+C ..................... 109 xii INTRODUCTION Denitrification is the respiratory reduction of nitrogenous oxides to gaseous products and can be functionally defined by these criterion (Tiedje, 1988): (1) Sensitivity to 02 inhibition; (2) Growth yield proportional to the amount of nitrogenous oxide present; (3) Accumulation of 80% or more of respired NO3' and N02' as N20 or N2; (4) Rapid N03' or N02' dissimilation without NH4+ accumulation; (5) Presence of one of two types of dissimilatory N02' reductase - heme ggl or nonheme Cu. Denitrification is characterized by irreversible Noz' reduction to NO and N20 by the enzyme dissimilatory N02'reductase (dNir). This enzyme is often confused with assimilatory N02“ reductase (Nir) which, in contrast to dNir, is regulated by NH“+ and organic N concentration, and is unaffected by 02 (Cole, 1987). Dissimilatory N02'reductase is also confused with a dissimilatory N02'reductase which leads to N114+ formation (Cole, 1987). A fourth N02' reductase, identified in Propionibacteria by Kaspar (1982), apparently lacks the physiological roles of dissimilatory N02' reductase, and may simply be a detoxification enzyme. Hereafter, by N02' reductase, I mean 1 2 dissimilatory N02' reductase (dNir) not assimilatory N02’ reductase or dissimilatory N02' reduction to NH4+. Some of the most interesting questions in microbial ecology concern mechanisms, like N02' reduction, which enable bacteria to exploit novel environmental niches. Global aspects of denitrification are well studied. Less is known about dNirs in terms of distribution, variability, regulation, and genetics. This thesis addresses these questions. 12153112931231 Denitrification is an ubiquitous environmental process, and its reviews legion (Firestone, 1982; Focht, 1981; Focht and Verstraete, 1977; Hochstein and Tomlinson, 1988; Kuenen and Robertson, 1987; Knowles, 1982; Payne, 1973; Payne, 1981; Stouthammer, 1976, 1988; Tiedje, 1988; Zumft et al., 1987). It is easier to note the groups in which dNir is absent, than those in which it is present (Tiedje, 1988): (1) Obligate anaerobes; (2) Gram-positive organisms other than Bacillus; (3) The Enterebacteriaceae. Nitrite reductase is found in organisms which cover the gamut of physiological groups: general aerobic erganotrophs, oligocarbophiles, fermentors, halophiles, thermophiles, sporeformers, magnetotactic bacteria, dinitrogen-fixers (free and symbiotic), plant and animal pathogens, phototrophs, hydrogen exdizers, sulfur oxidizers, ammonium oxidizers (Tiedje, 1988). The strains for which dNir type is known are few, and fewer still the bacteria from which dNirs have been purified. Two basic types have 3 been found: one contains Cu centers (Cu-dNir), the other homes 9 and g1 (ggl-dNir). Heme cfil-dNirs have been purified from Alcaligenes faecalis (Iwasaki and Matsubara, 1971), Paracoccus (Micrococcus) denitrificans (Newton, 1969), P. aeruglnosa (Gudat et al., 1973; Yamanaka and Okunuki, 1963), P. stutzeri (formerly P. perfectomarina)(Liu et al., 1983), and Thiobaclllus denitrificans (Sawhney and Nicholas, 1978). The enzyme has been detected in Azospirillum brasllense (Lalande and Knowles, 1987), P. halodenltrificans (Grant and Hochstein, 1984), and an Erythrobacter strain (Shioi et al., 1988). Nonheme Gu-dNirs have been purified from Achromobacter cycloclastes (Iwasaki et al., 1975; Rules and Averill, 1988; Kashem et al., 1987), A. xylosoxidans (synonymous with Alcaligenes sp. NCIB 11015, A. denitrlflcans ss xylosoxldans, and P. denitriflcans)(lwasaki et al., 1963), A. faecalis strain S6 (Kakutani et al., 1981), Nitrosomonas europaea (Dispirito et al., 1985; Ritchie and Nicholas, 1974; Miller and Wood, 1983). P. aureofaciens (Zumft et al., 1987), Rhodobacter (Rhodopseudomonas) sphaeroides f.sp. denitrificans (Michalski and Nicholas, 1985, 1988a). It has further been identified in P. denitrificans ATCC 13867 (distinct from A. xylosoxidans)(Radcliffe and Nicholas, 1968; Michalski and Nicholas, 1988b) and Thiosphaera pantotropha (L.A. Robertson, Ph.D. thesis, University of Delft, 1988). To reiterate an observation from Gamble et a1. (1977), it is unfortunate that most physiological and genetic studies on denitrification (particularly dNir) have been done with only a few strains - P. aeruginosa and P. denitrificans to name two - which inaccurately reflect true microbial populations. Nitrite reductase has 4 yet to be purified from P. fluorescens, the numerically dominant denitrifier in soils, sediments, and sludges (Gamble et al., 1977). Heme ggl-dNirs are assumed to be the most common type (Shapleigh and Payne, 1985). This may reflect biased isolation procedures, since the numerically dominant denitrifiers in readily cultured environmental samples are typically pseudomonads (Gamble et al., 1977). In one study, Bacillus sp. were the dominant denitrifiers from soil samples (Abd-el- Malek et al., 1974). Since environmental samples have usually been from agricultural sites, questions remain about soil effects on dNir types. Polyclonal antibodies raised against specific dNir types, could facilitate comparisons of dNirs between strains and between genera. Qualitative immunoblotting procedures don't replace protein purification, but do provide rapid information about the relatedness of dNirs and, in conjunction with SDS-Page, useful information about subunit structure. Kbrner et a1. (1987) used polyclonal antibodies to look at the relatedness of N20 reductases and N02' reductase in P. perfectomarina (now P. stutzeri) isolates. Nitrite reductases from ggl- type denitrifiers were much less related than N20 reductases. Michalski and Nicholas (1988b) immunologically compared NO3', NOZ', and N20 reductases in N03' respiring nondenitrifiers, and denitrifiers which were predominantly of the Cu-dNir type. Cu-dNirs, in the organisms they tested, were very similar in terms of molecular mass and antibody affinity. There was no immunological similarity between 941' and Cu-dNirs. These studies were limited by a narrow focus towards only a few organisms (mostly well-characterized laboratory strains) or one dNir type. Alternately, Shapleigh and Payne (1985) demonstrated that the 5 chelating agent diethyldithiocarbamate (DDC), specifically inhibited Cu- dNir during in 21519 dNir activity assays. They have suggested that dNir types could be characterized by differential DDC inhibition of ggl- and Cu-dNirs. Characteristics The main differences between Cu- and col-dNirs are in molecular mass and metal content. Heme gdl-dNirs have similar molecular weights, typically 120,000. The enzymes are dimers composed of identical subunits. Each subunit contains one home g and one heme d1. Other molecular weights have been reported, ranging from 90,000 to 140,000 (Hochstein and Tomlinson, 1988). A11 gfil-dNirs have cytochrome oxidase activity. This is presumably not their main physiological role. Their affinity for N02" is greater than for 02, their Km for 02 is high, and other oxidases are better competitors for available electrons (Hochstein and Tomlinson, 1988; Sawhney and Nicholas, 1978). The ggl-dNirs isolated from P. stutzeri JM300 (E. Weeg-Aerssens, Ph.D. thesis, Michigan State University, 1987) has a pI of 5.4. The pI of P. aeruginosa ggl-dNir was reported as 6.9 (Barber et al., 1976) and from P. denitrificans as 3.85 (Newton, 1969). There are two forms of Cu-dNir, distinguished by characteristic EPR signals from their Cu centers. One form has type I and type II copper centers, the other form has only type I copper centers. Cu-dNirs are heterogeneous with respect to other parameters: molecular mass, subunit composition, and p1 (Table 1). 6 Table 1. Selected physical characteristics of Cu-dNirs Enzyme source Mass Subunit Cu pI (kDa) Structure types Reference A. cycloclastes 69 36.8 x 2 ISII nd Iwasaki et al., 1972 Liu et al., 1986 A. xylosoxidans 70 37.0 x 2 I 8.4 Iwasaki et al., 1963 Masuko et al., 1984 A. faecalis 86 110 30.0 x 4 I&II 4.5 Kakutani et al., 1981 N. europaea 120 40.0 x 4 I&II 4.63 Miller 8 Wood, 1983 Dispirito et al., 1985 P. aureofaclens 85 40.0 x 2 I 6.05 Zumft et al., 1987 R. sphaeroides 80 37.5 x 1 I nd Sawada et al., 39.5 x 1 II 1970 Michalski & Nicholas, 1985 7 The only Cu-dNir to date with a basic pI is from A. xylosoxidans. Like ggl-dNirs, Cu-dNirs have cytochrome oxidase activity, with the exception of Cu-dNir from P. aureofaciens (Zumft et al., 1987). It isn't clear what importance can be associated with having both type I and type II Cu centers. There is no obvious immunological difference (Michalski and Nicholas, 1988b). Type I Cu is presumably the catalytic site. Type II Cu may be an artifact. It tends to disappear with enzyme purification. Achremobacter cycloclastes Cu-dNir was recently crystallized by Turley et a1. (1988). In contrast to earlier reports from their lab and others, they found only two Cu atoms/protein, and only one type of Cu (Type I) which existed in identical environments. If A. cycloclastes Cu-dNir, the most studied type, serves as a model, one would suspect that the enzyme contains identical subunits each with a Type I Cu center (C. Hulse, Ph.D. thesis, University of Virginia, 1989). Nitrite reductase types are presumably associated with the cytoplasmic membrane in accordance with their role in generating energy through electron transport phosphorylation (Koike and Hattori, 1975a,b). Their exact location, however, is not known. Cu-dNir in R. sphaereides was localized in the periplasm (Sawada and Satoh, 1980). No other Cu- dNirs have been similarly examined. Heme ggl-dNirs may be loosely bound to the cytoplasmic membrane (Sawhney and Nicholas, 1978), they may be soluble (Mancinelli et al., 1986), or both (Zumft et al., 1979). If membrane bound, they have been reported as cytoplasmically (Saraste and Kuronen, 1978) or periplasmically (Wood, 1978) oriented. Dimmers Cu-dNirs appear to be more diverse than ggl-dNirs. DNA sequence comparison of the dNir genes would be the ultimate measure of divergence. Alternately, protein mapping could reveal protein sequence differences in purified proteins (Cleveland et al., 1977). However, without suitably cloned genes, or purified proteins, alternate methods are needed to screen divergence among dNir types. Immunoblot analysis is a quantitative measure of immunological similarity between proteins based on antibody binding (Howe and Hershey, 1981). With saturating concentrations of antibodies, the degree of antibody binding is a quantitative measure of similarity based on common epitopes. Immunoblet distance is a reflection of that similarity. It is the ratio of antibody binding to type strains vs. antibody binding to test strains (Howe and Hershey, 1984). The divergence of dNirs from type strains can then be directly measured in relatively crude protein preparations to determine whether Cu-dNirs are, indeed, more immunologically diverse than col-dNirs; whether there are differences in ggl-dNir antibody binding based on the particular col-dNir used as an antigen; whether the divergence among ggl- and Cu-dNirs is at the same rate. In the same functional class of proteins, sequence evolution goes on at approximately the same rate (Wilson et al., 1977). In a given protein, evolution will depend on the compatibility of sequence change with biochemical function and the likelihood that the organism will survive without that biochemical function. That Cu-dNirs appear to diverge more than heme types may not reflect a longer evolutionary history, only a greater capacity of nonheme Cu-dNirs to accept sequence 9 changes without losing function compared to ggl-dNirs. There are no obligate denitrifiers, and in the presence of 02, dNir synthesis and activity is inhibited (Tiedje, 1988). Consequently, loss of protein activity due to sequence substitution is not a lethal change. The presence of Cu-dNirs in Pseudomonads such as P. aureofaciens and.P. denitrificans is interesting since most Pseudomonas spp. examined so far contain a ggl-dNirs. Bacterial phylogenies based on SS rRNA (Ohkubo et al., 1986), 16S rRNA (Fox et al., 1980), RNA homology (De Vos and De Ley, 1983) or DNA homology (Palleroni et al., 1973) do not predict whether an organism will have Cu- or ggl-dNir. This may indicate insufficient phylogenetic data or perhaps is an indication that mechanisms exist in nature for bacteria to acquire dNir if they have lost this function, and there is a strong selection for denitrification. There is no clear answer for why two enzymes have evolved. Nor is there a clear answer for the existance of both Cu- and col-dNir in the same genus or in closely related genera (Ohkubo et al., 1986). One possiblity is horizontal gene transfer via a plasmid. Nitrite reductase may be on a plasmid in one A. eutrophus strain.(Romermann and Friedrich, 1985). In Pseudomonas sp., therefore, strains which have lost dNir activity - and truncated denitrifiers are common (Jeter and Ingraham, 1981) - may have regained it by plasmid acquisition. Albeit, the dNir they recovered was not the same dNir type they originally possessed. WW Perhaps there is still a positive selection for dNir despite the current 02 rich environment. Denitrification is second only to respiration in terms of the number of gene copies present (J.M. Tiedje, 10 S. Simkins, and P. Groffman, Plant and 8011, in press). Not only do dNirs persist in organisms from anaerobic NO3' free environments (K.S. Jorgenson and J.M. Tiedje, Abstr. Annu. Meet. Am. Soc. Microbiol. 1989, N87, p. 299), but they also persist in organisms from long term, air- dried soils (Smith and Parsons, 1985). It has been suggested that copies of mRNA for synthesis of NO3' and N02’ reductases are '1eng lived' in cells (Michalski and Nicholas, 1985). Smith and Parsons (1988) theorized that the enzymes were stabilized in cells during 02 exposure. Certainly, organisms in transiently anaerobic zones could benefit from denitrification. There are reports of N20 evolution from seemingly well-aerated soils (Bremner and Blackmer, 1978; Smith and Tiedje, 1979). This has been partitioned into three processes based on differential acetylene (Berg et al., 1982) and 02 inhibition (Robertson and Tiedje, 1987): (1) Denitrification; ‘ (2) Nitrification; (3) Other. Denitrification is thought to occur in anaerobic microsites. These develop either from physical restraints on 02 diffusion (Sexstone et al., 1985), or because localized concentrations of available carbon, cause respiratory depletion of 02, and concurrent 002 evolution (Perkin, 1987; Rice et al., 1988). Autotrophic and heteretrophic nitrification have been implicated as another aerobic N20 source. The most commonly cited autotroph, N. europaea, may evolve N20 via true dissimilatory N02' reduction (Path and Focht, 1985). An alternate mechanism postulates that an intermediate from hydroxylamine oxidation dimerizes to hyponitrous acid, and rapidly ll decomposes to form N20 (Bremner and Blackmer, 1978; Yoshida and Alexander, 1970). A heterotrophic, nitrifying, Alcaligenes sp., which also denitrified, produced equimolar amounts of NH4+ and N20 when switched from denitrifying growth to aerobic growth with hydroxylamine (Castignetti and Hollocher, 1982). Nitrous oxide was assumed to formed via a nitroxyl which dimerized to N20. Cu-dNir from A. cycloclastes catalyzes N20 formation from hydroxylamine and N02' (Iwasaki and Matsubara, 1972). Likewise, P. stutzeri ggl-dNir catalyzes nitrosation reactions between hydroxylamine and N02' (Aerssens et al., 1986). The enzyme bound nitrosyl is presumably the reactive species. This reaction, however, was inhibited by 0.1 mM N02'. Other aerobic sources of N20 remain inexplicable. Organisms which dissimilate N03' to NH4+, for example, also release some N20 (Bleakley and Tiedje, 1982; Yeshida and Alexander, 1970). This is apparently the result of NO3' reductase catalyzing the reduction of N02' to N20 (Cole, 1987). Nitrous oxide is also formed by unknown mechanisms of NO3' assimilation by fungi (Yoshida and Alexander, 1970). At least two denitrifying organisms, magnetic Spirillum (Aquaspirillum magnetotacticum and strain M81) and N. europaea, are obligate aerobes (Bazylinski and Blakemore, 1983; Escalante-Semerena et al., 1980; Path and Focht, 1985). Periodic reports of aerobic denitrification not only challenge the dictum that denitrificatien is a strictly anaerobic process (Payne, 1973), but suggest alternate mechanisms for N20 evolution in aerobic soils. Many studies simply reflect inadequate 02 supply in solution (Hochstein et al., 1984). Denitrification in the two most frequently cited organisms, an 12 Alcaligenes sp. from activated sludge (Krul, 1976), and Hyphomicrobium X (Meiberg et al., 1980) consisted solely of N03" reduction. The implication, at least with the Alcaligenes sp., was that dissimilatory NO3' reductase was constitutive. It was still partially repressed by 02, since NO3' reduction was slow until measurable dissolved 02 was gone. Thiosphaera pantotropha, a facultatively anaerobic, facultatively autotrophic, sulfur bacterium isolated from a desulphurizing, denitrifying, effluent treatment system, was reported to denitrify at 90$ air saturation (Robertson and Kuenen, 1983). Part of the evidence for this phenomenon was increased growth in chemostat culture with both NO3' and 02, and instantaneous N2 evolution in anaerobic manometric experiments (Robertson and Kuenen, 1984a). It was proposed that electron flow through the electron transport chain was restricted by a bottleneck which could be overcome by allowing electrons to flow to N-oxide reductases as well as oxygen (Robertson and Kuenen, 1984b). The regulation of NO3' reduction in Alcaligenes spu was partly attributed to inhibition of 02 respiration by NO (Krul, 1976). Parsonage et a1. (1985) also noted that NO, because of its affinity for hemes, inhibited 02 respiration, and allowed N02' reduction in the presence of 02. Genetics The genetics of dNirs are still largely unknown, as is the regulation of their synthesis. Dissimilatery N02' reductase mutants (nirA, D, and E) were isolated from P. aeruginosa by van Hartingsveldt et a1. (1971). They were distinct from assimilatory N02' reductases, and mapped at three separate loci on the chromosome. While attempting to isolate N20 reductase mutants from P. stutzeri by transposon Tn5 mutagenesis, Zumft et a1. (1985) also found mutants defective in N03' 13 and N02' respiration. Three representative nir mutants were selected and studied further (Zumft et al., 1988). One mutant, MK201, overproduced cytochrome £552 but had an in 21:19 functional gfil-dNir. Mutant MK 202 lacked cytochrome gdl-dNir entirely, and had low amounts of cytochrome £552. The last mutant, MK203, synthesized ggl-dNir defective in the heme d1 prosthetic group. All could reduce N0. N0 similar research has been done with organisms which contain an identified Cu-dNir. The only other organism in which denitrification genes have been studied is A. eutrophus. Hogrefe et a1. (1984), while looking for hydrogenase genes, observed mutants with pleiotropic phenotypes which were unable to grow with hydrogen (Hox') and were also unable to either assimilate or dissimilate N03' (Nit'). These mutations mapped on the chromosome, and were given the overall designation Hno'. Later, loss of a megaplasmid, pHG, was correlated with loss of denitrifying ability (Romermann and Friedrich, 1985). Plasmid-free mutants converted some N03' to NOZ', but could not metabolize N02' anaerobically.-Both lithoautotrephy and denitrification could be transferred by plasmid pHG to strains which lacked these functions (Schneider et al., 1988). The chromosomal mutation, with the Hno' phenotype (hydrogen oxidation and denitrification deficient), was cloned from a genomic library of A. eutrophus and used to find the wild-type gene (Romermann et al., 1989). The resulting plasmid, incorporating the wild-type Hno gene, was able to complement Hno' mutants and relieve glutamine auxotrophy in NtrA' mutants of enteric bacteria. This suggested that the hno gene product was functionally similar to the NtrA protein which is a novel sigma factor of RNA polymerase. The hno gene product did not complement Fnr' mutants . l4 It would be tempting to adopt, as a paradigm for denitrifiers, the genetics and regulation associated with anaerobic NO3' metabolism in E. coli. With respect to N03“ reductase, this may be possible, but until further work with the genetics of other N-oxide reductases is done, this paradigm must be viewed with caution. The focus of this thesis eventually became an examination of N02' reductase diversity. The first objective was to examine the distribution of dNirs by immunological and chemical means. The second objective was to compare dNirs of the same type, from among different organisms, to begin identifying physical, and hopefully evolutionary differences. The third objective, was to utilize dNir-specific polyclonal antibodies, and localize dNirs in denitrifiers. The last objective was to examine 02 regulation in the Cu-type denitrifier A. cycloclastes, in the context of aerobic denitrification. SECTION 1 IMMUNOLOGICAL IDENTIFICATION, DISTRIBUTION, AND DIVERCENCE OF DISSIMILATORY HEME CD AND NONHEME CU NITRITE REDUCTASES IN DEN TRIFYING BACTERIA 15 16 INTRODUCTION Dissimilatory N02' reductases (dNirs) are pivotal to the fate of combined nitrogen in the environment. They are the point where nitrogen is routed to dissimilation instead of assimilation (Payne, 1981). Two distinct types of dNirs are known: one contains a Cu center (Cu-dNir), the other hemes c and g1 (ggl-dNir). Both seem to carry out the same physiological reaction. NO is typically produced from N02' during in xitzg enzyme assays, but under some conditions, N20 is also produced (Hochstein and Tomlinson, 1988; Wharton and Weintraub, 1980). Cytochrome gdl-dNir has been identified in Alcaligenes faecalis (Matsubara and Iwasaki, 1971), -Paracoccus (formerly Micrococcus) denitrificans (Newton, 1969), P. halodenitrificans (Grant and Hochstein, 1984), Pseudomonas aeruginosa (Yamanaka and Okunuki, 1963), P. stutzeri (Korner et al., 1987), and Thlobacillus denitrificans (LeGall et al., 1979). Nonheme Cu-dNirs have been isolated from Achromobacter cycloclastes (Iwasaki and Matsubara, 1972; Hulse et al., 1988), A. xylosoxidans (synonymous with Alcaligenes sp. NCIB 11015, A. denitrificans ss. xylosoxidans)(Masuko et al., 1984), A. faecalis strain S6 (Kakutani et al., 1981), Nitrosomonas europaes (Ritchie and Nicholas, 1974), P. aureofaciens (Zumft et al., 1987), Rhodobacter (formerly Rhodopseudomonas) sphaeroides f. sp. denitrificans (Michalski and Nicholas, 1985; Sawada et al., 1978), and Thiosphaera pantotropha (L.A. Robertson, Ph.D. Thesis, University of Delft, 1988). Two forms of Cu- l7 dNir which have different EPR spectra have been observed (Michalski and Nicholas, 1988b). The distribution of the two dNir types in the environment is unknown. Strains which are typically studied in the laboratory, do not reflect the dominant denitrifying populations in nature (Gamble et al., 1977). One way to screen denitrifiers for dNir type is to distinguish between Cu and heme components. Shapleigh and Payne (1985) used diethyldithiocarbamate (DDC), a Cu chelator, to identify Cu-dNirs. Another approach is to characterize the proteins with polyclonal antibodies specific for each dNir type. Polyclonal antibodies raised against both heme ggl- and nonheme Cu-dNirs have facilitated their comparison. Kerner and Zumft (1987) found that ggl-dNirs from.P. aeruginosa and P. fluorescens were not recognized by P. perfectomarina (now P. stutzeri strain Zobell) gdl- dNir antiserum, which did recognize gdl-dNir from a P. stutzeri isolate. Michalski and Nicholas (1988a) showed that polyclonal antibodies specific to Cu-dNir from R. sphaeroides f. sp. denitrificans cross- reacted with Cu-dNirs in two other denitrifying R. sphaeroides strains. They also cross-reacted with Cu-dNirs from A. cycloclastes, A. denitrificans and two Pseudomonas sp. (Michalski and Nicholas, 1988b). This confirmed that dNir type differed among closely related Pseudomonas spp., since Zumft et a1. (1987) had already found Cu-dNir in P. aureofaciens. Bacterial phylogenies based on 16S rRNA sequencing have largely supplanted phylogenies developed from comparing enzyme sequence differences (Weese, 1987). Sequence difference information, based on immunological differences, still reveals diversity and evolutionary 18 change within classes of proteins as was demonstrated for DNA polymerase I (Tafler et al., 1973), glutamine synthetases (Baumann and Baumann, 1970; Tronick et al., 1973), alkaline phosphatases (Cocks and Wilson, 1972), and hydrogenases (Kovacs et al., 1989). Howe and Hershey (1981) have described a method of immunoblotting analysis, by which antibodies can be used to quantitate specific proteins in cell lysates. They used this technique to explore evolutionary divergence of initiation factors in species other than Escherichia coli. The immunological relationships between the species they studied and E. coli were similar to other previously studied proteins (Howe and Hershey, 1984). Immunoblotting analysis may help examine relationships among denitrifiers based on the dNir type they possess, without resorting to individual protein purifications. Rather, a first approximation of diversity can be derived. Because it can be quantitatively employed, a measure of the divergence among different dNir types can be calculated. This would focus attention on the most interesting dNirs to study further. The following research was a joint effort. Collaborators at the University of Virginia, Chuck Hulse and Bruce Averill, provided the purified Cu N02° reductase. Purified ggl N02' reductases from P. aeruginosa and P. stutzeri were provided by Els Weeg-Aerssens and Rick Ye. Dr. Alahari Arunakumari was involved in every step of antisera production and was responsible for much of the immunoscreening for ggl- dNirs. My contribution to the research included antisera production, verification of antisera specificity, all immunoscreening for Cu-dNirs, all rescreening of initially nonimmunogenic strains (for both heme gdl- 19 and Cu-dNirs), identification of denitrifying strains from isolates in the aerobic 2,4-D degrading enrichments, all DDC inhibition assays, and the immunoblot analysis. This section presents the use of polyclonal antibodies in Western immunoblots to identify the distribution and immunological relatedness of 341- and Cu-dNirs in taxonomically diverse denitrifiers, and to determine the predominant dNir type in numerically dominant denitrifiers from the environment. Diethyldithiocarbamate (DDC) was used to confirm the immunological identification and to characterize the dNir type in denitrifiers which did not react with antisera. Immunoblot analyses were used to examine the divergence of Cu- and heme ggl-dNirs. METHODS AND MATERIALS W Denitrifying strains from culture collections, and their sources, are listed in Table 1. Numerically dominant denitrifiers of geographically diverse origin were previously collected and characterized by Gamble et a1. (1977). For immunoblot analysis, A. eutrophus #154, A. faecalis #40, and A. faecalis #43 were obtained from this collection. Strains isolated by W. Holben from aerobic 2,4- dichlorophenoxyacetic acid (2,4-D) degrading soil enrichments were used to examine cultures not selected on the basis of their denitrifying capacity. Of the 250 isolates, 51 were respiratory denitrifiers based on diagnostic criteria summarized by Tiedje (1988) and were tested for dNir type. These isolates were from soils of two Michigan sites (East Lansing and Kellogg Biological Stations), Kansas (Konza Prairie), and 20 Saskatchewan, Canada (W. Holben, personal communication). Isolates 1.6.L and 2.4.E were from this collection. Wane For immunological screening, strains were grown in 16 by 125 mm test tubes sealed with butyl rubber septa. The tubes contained 10 ml medium, an inverted Durham tube, and an initially aerobic headspace. Most strains were grown in tryptic soy broth (TSB)(Difco, Detroit, MI) supplemented with 1.0 g KNO3 or 0.4 g KN02 per liter. Rhizobium spp. and Bradyrhizobium japonicum were grown in yeast extract mannitol broth (YEMB)(Zablotowicz and Focht, 1981) supplemented with 1.0 g KNO3 per liter. Halobacterium species were grown in HYH-2 medium (per liter: 5.0 g yeast-extract, 2.0 g casein, 176.0 g NaCl, 5 g KNO3, 20.0 g MgC12'6H20, 2.0 g KCl, 0.1 g CaClz'ZHZO, 13.0 g HEPES, pH 6.7, L. Hochstein, personal communication). All cultures were incubated at 24 'C and maintained for routine studies on slants or plates of the same medium amended with 1.5% agar. Cultures harvested for enzyme assays were grown in 150 ml TSB in 250 m1 Erlenmeyer flasks sealed with rubber stoppers. Bacterial growth depleted headspace 02 sufficiently to induce denitrification, and all cultures were producing N20 at harvest. Poorly growing strains were .first cultured aerobically in 500 m1 sidearm flasks containing 300 m1 TSB until stationary phase was reached. The flasks were then sealed with rubber stoppers, flushed with sterile argon, and incubated 24 h at 30 ‘C before harvest to induce denitrifying enzymes. Rhizobium spp. and B. japonicum were cultured aerobically in 2 1 flasks with 1 l YEMB until turbid, and were then shifted to anaerobic conditions as described above. 21 Cultures used for immunoblot analysis were grown in 15 g 1'1 Tryptic soy broth (Difco, Detroit, MI) supplemented to 10 mM KNO3 and 1 uM CuSOa'7H20. Growth was at 24 ’C in sealed 500 m1 sidearm flasks filled with 250 ml of media. The headspace was made anaerobic by repeated flushing with sterile argon from a gassing manifold. WWW Polyclonal antibodies against dNirs were raised in New Zealand white rabbits. Purified Cu-dNir from A. cycloclastes ATCC 21921 was provided by Hulse et a1. (1988). Purified gdl-dNirs from P. aeruginosa ATCC 10145 and from P. stutzeri JM300 was provided by E. Weeg-Aerssens and R. Ye. The enzymes were emulsified in 1 m1 of Freund's incomplete adjuvant and injected subcutaneously. Booster injections of enzyme in Freund's incomplete adjuvant were made at 8 weeks and 16 weeks. Blood samples were taken at 2 weeks and on a weekly basis thereafter. Immunization was followed by Ouchterlony double immunodiffusion (Jurd, 1981). Cultures which met one of the following criteria for denitrification were harvested for immunoscreening: (1) Bubble formation in the inverted Durham tube; (2) Accumulation of N20; (3) Removal of N03" and N02' from the media as detected by the diphenylamine test (Tiedje, 1982). An aliquot of the culture was centrifuged, the pellet resuspended in sample buffer (0.0625M Tris-H01 pH 6.8, 2% SDS, 10% glycerol, 5% 2- mercaptoethanol, 0.0013 bromphenel blue), and the cells then lysed in a boiling water bath (Laemmli, 1970). Strains were extracted by phenol (Carlson and Vidaver, 1982) if they lysed poorly under this protocol, or 22 if proteins were not immunologically recognized by antiserum to either dNir type. Proteins from the crude lysates were separated by SDS-PAGE on 10% gels and transferred by electrophoresis onto nitrocellulose (Towbin et al., 1979). All dilutions and incubations for dNir detection were in TBST buffer: 10 mM Tris-HCL (pH 8.0), 150 mM NaCl, 0.05t Tween 20 (Promega western Blot AP System). The nitrocellulose filters were incubated l h with 1% BSA to block nonspecific binding, then incubated l h with a 1:1000 dilution of antiserum. The filters were rinsed in TBST three times for a total of 15 min then incubated with a 1:1000 dilution of alkaline phosphatase-conjugated anti-rabbit goat IgG (Sigma). The filters were rinsed in TBST as before, and the dNirs were visualized by development of purple color following incubation.with nitro blue tetrazolium (NBT) and 5-bromo-4-chloro-3-indoyl-phosphate (BCIP) substrate (Sigma). The criteria for a positive immunological response were: (1) Recognition (color development) of proteins at approximately the same molecular weight as simultaneously recognized purified dNirs, or proteins from denitrifiers of a known dNir type; (2) No recognition when subsequently screened with antiserum from the opposite dNir type. W For gas chromatography assays and native gel analysis, denitrifying cultures were harvested by centrifugation at 4 'C and 10,000 rpm for 10 min. Cells were washed with 50 mM HEPES buffer (pH 7.3), centrifuged as before, and resuspended in 5 m1 of the same buffer. 23 Cultures not immediately used were stored at -70 °C. Resuspended or thawed cultures were passed three times through a French press at 15,000 psi, divided into aliquots and stored at -70 '0. Prior to use, the crude extracts were spun 5 min at 4 'C in a microfuge to remove debris; this supernatant was used in enzyme assays. The protein content in crude extracts was measured by the Lowry method using BSA as a standard (Lowry et al., 1951). For immunoblot analysis, bacteria were grown until all NO3' was removed from the media as judged by the colorimetric diphenylamine assay (Tiedje, 1982). The cells were pelleted at 4 ’C for 10 min at 10,000 rpm, washed in 50 mM HEPES buffer pH 7.3, pelleted as before, and resuspended in 5 m1 of the HEPES buffer. The resuspended cells were disrupted by sonication, cell debris pelleted by centrifugation at 10,000 rpm for 30 min at 4 ’C and the supernatant (crude extract) used in all further assays. Protein concentration was determined by Bicinchoninic acid protein assay reagent (BCA - Pierce) using BSA as a standard. Win Denitrifying cells were ruptured by French press, the cells removed by centrifugation, and the proteins separated by nondenaturing 10% PAGE at 4 ’C and 100 V for 20 h. Sites with NOZ' reducing activity were located by the method of Zumft et a1. (1987). Gels were stained 5 min with 10 mM methyl vielogen in 50 mM HEPES buffer (pH 7.3) with 2 mg ml'1 sodium dithionite. The gels were then immersed in 1 M NaN02 (in HEPES buffer) which caused rapid bleaching where dNirs were located. DDC (10 mM) was incorporated in some of the activity stains to preferentially inhibit Cu-dNir activity (Shapleigh and Payne, 1985). 24 WWM Purified Cu-dNir and extracts prepared by French press of denitrifying A. cycloclastes ATCC 21921 and P. aeruginosa ATCC 10145 were incubated 24 h at 4 'C with an equal volume of either buffer, preimmune serum, Cu-dNir rabbit antiserum, or cdl-dNir rabbit antiserum. Precipitated proteins were removed by centrifugation and the soluble proteins were separated by nondenaturing 10% PAGE at 4 'C and 100 V for 20 h. The gels were stained for dNir activity and dNirs were identified by Western immunoblot. Qumran-m Enzymatic N02' reduction was measured as NO and N20 evolution from crude extracts using NADH/phenazine methosulfate (PMS) as the electron donating system. Assays contained 1 mM NOZ', 0.12 mM PMS, 2 mM NADH, and crude extract (generally 200 ul) in a total volume of 1 m1. All stock solutions were made in HEPES buffer (50 mM, pH 7.3) and the NADH stock was prepared daily (E. weeg-Aerssens, Ph.D. thesis, Michigan State.University, 1987). Assays were started by adding N02' to 8 ml serum'bottles containing the other combined ingredients. The bottles were previously made anaerobic by repeated flushing of the headspace with argon. When DDC was used as an inhibitor, it was added to a concentration of 12.5 mM 20 min prior to the N02' addition and the bottle was incubated on ice (Shapleigh and Payne, 1985). A11 enzyme assays were at 24 '6. Gas formation was monitored on a Perkin-Elmer 910 gas chromatograph equipped with dual columns and 63Ni electron capture detectors for simultaneous analysis of two samples (Kaspar and Tiedje, 1980). Carrier gas was 95% argon, 5% methane. Temperatures were: 25 detector 300 °C, injector 70 °C, and column 30 'C. The columns were 2 m x 2 mm i.d. stainless steel filled with Porapak Q. Carrier gas flow was adjusted to maximize NO and 02 separation. Nitrite reductase activity was measured in the presence and absence of DDC by monitoring the initial rate of NO and N20 evolution over 15 min. Each assay was done in duplicate. No NO or N20 was evolved in the absence of crude extract or in controls with autoclaved crude extract. lmnssiffssicnmlnis Ouchterlony immunodiffusion was done according to Howe and Hershey (1984). Agar gel plates containing 500 mM KCl, 1% agar, and 0.01% sodium azide in 20 mM phosphate buffer pH 6.8 were prepared. Fifty microliters of undiluted antiserum or crude extracts were added where indicated in Figure 6. The plates were allowed to develop at 30 “C for 48 h before examination. Immshlctsnalnis Nitrite reductases in crude extracts were separated from other cell proteins by 10% SDS-PAGE. The crude extracts were prepared for electrophoresis as previously noted. The crude extract added to each well was adjusted so that the amount of dNir activity per sample was equivalent (Table 5). Samples were separated on a Hoefer SE 250 minigel for 90 min at 25 mAmp constant current then transferred by electrophoresis to nitrocellulose filters (Towbin et al., 1979). Filters were incubated overnight at 4 'C in TBST buffer (10 mM Tris pH 8.0; 150 mM Na61; 0.05% Tween 20) containing 5% w/v nonfat dry milk. The filters were then incubated 30 min with a 1:1000 dilution of specific antiserum to Cu-dNir, or to gal-dNir from either P. aeruginosa or P. stutzeri. The filters were washed 3 x 5 min in TBST buffer and 26 incubated l h in TBST buffer containing a 1:7500 dilution of goat anti rabbit IgG conjugated to alkaline phosphatase (Promega). Nitrite reductases were detected by color development on the filters following incubation with nitroblue tetrazolium (NET) and 5- bromo-A-chloro-B-indoyl-phosphate (BCIP) substrate (Sigma). The relative levels of antibody binding to the various dNirs were quantified by whole band analysis using a Visage 110 Image Analyzer. In cases where a doublet of immunologically recognized proteins was present, both bands were quantified, and the sum of those values used in further analysis. RESULTS W Silver stains of the dNirs used for immunization demonstrated their purity (Figure l). Immunodiffusion assays indicated the antisera were specific for the antigens against which they were raised (not shown). Neither type of gdl-dNir antisera crossreacted with crude extracts containing Cu-dNir, nor did they crossreact with crude extracts containing nonhomologous gfil-dNir. In activity-stained, non-denaturing PAGE, reductase activity in crude extracts of A. cycloclastes (Cu-dNir) and P. aeruginosa (ggl-dNir) (Figure 2A) comigrated with immunologically recognized proteins in immunoblots of the same gels (Figure 28). Other proteins with reductase activity were not immunologically recognized. Achromobacter cycloclastes Cu-dNir migrated as two distinct immunologically recognized bands only one of which had reductase activity. Possibly, the more rapidly migrating band represents a degradation product. 27 Figure 1. Silver stain of proteins used in rabbit immunizations. Lane A, A. cycloclastes Cu-dNir; B, P. stutzeri mil-dNir; C, A. cycloclastes N20 reductase; D, P. stutzeri N20 reductase. 28 Figure 2. Comigration of dNir activity with immunologically recognized proteins in denitrifier crude extracts. A. Nitrite reductase activity stain; B. Western immunoblot with Cu- and g; -dNir antisera. Lane 1, A. cycloclastes ATCC 21921; 2, P. aeruginosa ATCC 10145. Arrows indicate the location of comigrating proteins. 29 Bleached regions with reductase activity cut from a duplicate gel of A. cycloclastes and P. aeruginosa crude extracts (indicated by the arrows in Figure 2A) had enzymatic N02' reducing activity and produced 222 and 261 umol N-gas (as a mixture of N0 and N20), respectively, after 90 min incubation in sealed vials. A control from a nonbleached portion of the gel had no enzymatic activity. If DDC, a specific inhibitor of Cu-dNir, was incorporated into activity stains Cu-dNir, but not ggl-dNir activity, was inhibited (not shown). Finally, Cu- and ggl-dNirs were not precipitated by preimmune serum, but activity stains of nondenaturing gels showed specific loss of Cu-dNir activity following incubation with Cu-dNir antiserum (Figure 3). Specific loss of gdl-dNir activity after incubation with gfil-dNir antiserum was observed in similar gels (not shown). WWW Twenty-three strains from culture collections representing 17 taxonomically diverse denitrifying bacteria were screened by Western immunoblot for immmunologic recognition of Cu- and gdl-dNirs (Table 1). The criteria for a positive result were immunological recognition (as described in Methods) after incubation with rabbit antiserum, and nonrecognition following incubation with antiserum from the opposite dNir type (Figure 4A-B). For these strains, antiserum to P. stutzeri gdl-dNir gave similar results to antiserum against P. aeruginosa ggl- dNir. The dNir type was identified for several strains in which dNir had not previously been characterized. Aquaspirillum itersonii, Flavobacterium sp., and P. fluorescens contained.gg1-dNir. Bacillus azotoformans, and C. nephridii contained Cu-dNir. 0f the 23 strains 30 Figure 3. Removal of dNir activity by Cu-dNir antisera. Samples in lanes 1-6 were preincubated with Cu-dNir antisera (1,3,5) or buffer (2,4,6). Lanes 1,2, A. cycloclasces Cu-dNir (1 ug); 3,4, A. cycloclastes ATCC 21921 crude extract; 5.6, P. aeruginosa ATCC 10145 crude extract. Arrows indicate where Cu-dNir but not 9; -dNir has been lost after . preincubation with Cu-dN r antisera. 31 Figure 4. western immunoblots of denitrifier crude extracts separated on 10% SDS-polyacrylamide gels. A. Immunoblotted with A. cycloclastes ATCC 21921 Cu-dNir antisera; B. Immunoblotted with P. aeruginosa ATCC 10145 9d -dNir antisera. Lane 1, A. cycloclastes ATCC 21921 purified Cu-dNir (1 ug); 2, A. cycloclastes ATCC 21921 crude extract; 3, P. aeruginosa ATCC 10145 crude extract; 4. P. stutzeri JM300 crude extract; 5, P. stutzeri JM300 purified gdl-dNir (1 ug). Molecular mass indicators are in kDa. ' 32 (tested by immunological recognition only). 10 contained ggl-dNir and 6 contained Cu-dNir. The dNirs of B. Japonicum, Halobacterium spp., Rhizobium spp., and R. sphaeroides were not recognized by the antisera (Table 1) even though the Halobacterium spp. (L. Hochstein, personal communication) and R. sphaeroides (Michalski and Nicholas, 1985) are ”known to contain Cu-dNirs. .‘ e e ' ‘e "' e“ ‘ ‘ 0 ° 0 g Q - - - . One hundred denitrifying isolates from geographically diverse environmental samples were also screened in Western immunoblots to determine the predominant dNir type in nature. Data are presented in Table 2 . Ninety-one isolates gave a positive immunological response to at least one of the antisera used in the analysis. Most strains (63/100) contained gdlodNir. The gd1.type was present among P. fluorescens and among the denitrifiers most closely related to this species. This group comprises the numerically dominant denitrifiers in soil (Gamble et al., 1977). In cases where isolates were screened with both types of gdl-type dNir antisera, immunological recognition did not always coincide. 1* 1w t... at A». a} an... F». 7.. 3m lea. l.“ 3. DL. 3. P 3. D. P P 0;. Wu Wm pa 1» 33 Table 1. Western immunoblot response of denitrifiers to polyclonal antibodies against Cu- and ggl-dNirs. N02° reductase Strain Source ggl Cu A. cycloclastes ATCC 21921 - + A. xylosoxidans H. Iwasaki - + NCIB 11015 A. itersonii ATCC 11331 + nd A. lipoferum ATCC 29707 + - B. azotoformans ATCC 29788 nd + B. japonicum ATCC 10324 - - C. nephridii ATCC 11425 nd + Flavobacterium sp. ATCC 33514 + nd H. denitrificans 81 L. Hochstein - - Halobacterium sp. D4 L. Hochstein - - Halobacterium sp. Baja 12 L. Hochstein - - P. denitrificans ATCC 19367 + - P. aeruginosa ATCC 10145 + - P. aureofaciens ATCC 13985 - + P. denitrificans ATCC 13867 nd + P. fluorescens ATCC 17822 + - P. fluorescens ATCC 33512 + - P. stutzeri ATCC 11607 + - P. stutzeri JM300 J. Ingraham + - P. stutzeri ATCC 14405 + nd Rh zobium sp. 8A55 D. Focht - - Rhizobium sp. 32Hl D. Focht - - R. sphaeroides f.sp. denitrificans 11106 T. Satoh - - + Positive immunologic response “d Not determined Negative immunologic response 953' 150‘. 34 Table 2. Immunochemical classification of nitrite reductase type in denitrifying isolates from sediments, sludges, and the soils of six countries. Immunoblot b reaction Similarity Species Cu 941 Isolate matrix or -------- DDC No.a Groupa Groupa Pa. Ps. Inhibc 13 1A P. fluorescens II - + nd - 42 1A P. fluorescens II - + - nd 63 1A P. fluorescens II - + nd nd 205 1A P. fluorescens II - + - nd 62 1A P. aureofaciens - + nd nd 70 1A P. fluorescens II - + + nd 72 1A P. fluorescens II - + + nd 47 1A P. fluorescens II - + + nd 64 1A P. fluorescens II - + + nd 44 1A P. fluorescens II - + + nd 53 1A P. fluorescens II - + nd nd 45 1A P. fluorescens II - + + nd 73 . 1A P. fluorescens II - + - nd 79 1A P. fluorescens II - + + nd 84 1A P. type 6 - + + nd 68 1A P. fluorescens II - + + nd 66 1A P. fluorescens II - + + nd 61 1A P. fluorescens II - + + nd 105 1A P. fluorescens II - + nd nd 49 1A P. fluorescens II - + + nd 67 1A P. fluorescens II - + + nd 75 1A P. fluorescens II - + + nd 89 1A P. fluorescens II - + nd nd 82 1A P. type 5 - + + nd 52 1A P. fluorescens II - + + nd 69 1A P. type 5 - + + nd 71 1A P. type 5 - + + nd 74 1A P. type 5 - + + nd 111 1A P. fluorescens - + + . - 85 18 P. type 7 - + + - 183 18 P. fluorescens IV - + - nd 185 lB-C P. fluorescens IV - nd + - 15 lB-C P. fluorescens - + nd nd M.‘ mus (El oil. 0 G 1|. 81‘ Cl \I." ‘III 1: 11111 Table 2 (cont'd) 35 Immunoblot b reaction Similarity Species Cu ggl Isolate matrix or -------- DDC No.a Group8 Group8 Pa Ps. Inhibc 12 10 P. type 2 - + - - 54 10 P. type 2 - + + nd 83 10 P. type 2 - - - - 81 lC-D P. type 2 - + nd nd 110 lC-D P. type 12 - + + nd 55 1C-D P. type 4 - + nd - 78 IC-D P. type 6 - + + nd 58 lC-D P. fluorescens II - + + nd 80 10 P. fluorescens II - + nd - 190 l P. fluorescens - + + nd 206 2 P. fluorescens IV - + - nd 107 2 P. type 11 + - - nd 163 2 P. type 11 + - - + 188 2 P. type 11 + - nd + 143 2 P. type 11 + nd - + 234 2 P. type 11 + - - + 149 2-3 P. type 11 + - - + 98 2 P. fluorescens II + - - nd 133 3 P. type 18 + nd - nd 135 3 P. type 18 + nd - nd 1181 3 P. type 16 + - nd nd 141 3 P. type 16 + - - nd 172 3-4 P. type 11 + - - + 14 3 P. type 1 - + nd nd 156 4 P. aeruginosa - + nd nd 87 4-5 P. type 8 - - - + 103 5A. P. type 10 - - - nd 18 5A. A. faecalis - + nd nd 20 5A, A. fsecalis - + nd nd 4 5A. A. faecalis - + nd nd 30 5A. A. faecalis - + nd nd 43 5A-B A. faecalis + - - + 99 58 A. fsecalis - + - nd 90 58 A. faecalis - + nd nd Table 2 (cont'd) 36 Immunoblot b reaction Similarity Species Cu ggl Isolate matrix or -------- DDC No.a Groupa Groupa Pa Ps Inhibc 104 SC A. faecalis - + - nd 144 SC A. faecalis + - - + 191 SC A. faecalis + - nd nd 102 5 A. faecalis - + nd nd 40 5 A. faecalis + - - + 28 5 A. faecalis + - nd nd 31 5 A. faecalis - + nd nd 41 5 A. faecalis + - - nd 65 5-6 A. faecalis + - + + 36 6 Unknown type 3 - - + nd 151 6 P. type 19 + - - + 153 6 P. type 19 + - - + 179 6-7 P. type 19 + - - + 148 7 A. faecalis + - nd nd 232 7 P. type 25 + - - nd 39 7-8 Unknown type 3 - - - + 221 9 P. stutzeri - + nd nd 224 9-10 P. stutzeri - + - nd 2312 9-10 P. stutzeri - + + - 195 9-10 P. stutzeri - + + nd 108 9-10 Unknown type 3 - - - nd 154 9-10 A. eutrephus + - - + 106 9-10 A. faecalis - + - nd 97 10 P. type 9 - - - nd 177 10 Flavobacterium - + + nd 46 10 Flavobacterium - + nd nd 192 >10 Bacillus - - - +. 193 >10 Bacillus - - - + 199 >10 Unknown type 22 + - nd nd 1£39 >10 Unknown type 21 + - - + 2(34 >10 Unknown type 24 - + nd nd 115 >10 Unknown type 15 - - nd nd 1~771 >10 Ac faecalis + - - + \ ‘posizia 'Segatix 3dFatbe aSpecie. from G isolat level proper in chi 5 Cu - 1 Pa.- Ps.-l °m>c I actix for 1 37 Table 2 (cont'd) + Positive immunologic response ' Negative immunologic response “d Not Determined a Species identification and similarity matrix grouping are from Gamble, et a1. (1977). Approximately one-third of the isolates collected were not identified to the species level or beyond. These isolates, based on the taxonomic properties tested, were given ”type" numbers in the order in which they were isolated. - b Cu - antibodies raised against A. cycloclastes ATCC 21921 Cu-dNir Pa.- antibodies raised against P. aeruginosa ATCC 10145 99 -dNir Ps.- antibodies raised against P. stutzeri JM300 fll'duir ° DDC Inhib: +/- indicates whether inhibition of dNir activity occurred after preincubation with 12.5 mM DDC for 20 min. :esp 33C at 203 inc 30‘; is 38 There were 28 denitrifiers with Cu-dNir among the immunologically reactive isolates. These strains were predominantly A. faecalis and Pseudomonas strains which Gamble et a1. (1977) could not easily characterize to species level (Table 2). One isolate (#65, an Alcaligenes sp.) had a positive immunological response to antisera against Cu-dNir and P. stutzeri ggl-dNir, but not P. aeruginosa ggl- dNir. Only Cu-dNir was detected by DDC analysis (Table 2, Table 3). The response to this particular gdl-dNir antiserum was probably nonspecific. DDC inhibited dNir activity in all denitrifiers determined by immunoblotting to have Cu-dNir (Table 2). Of the nine isolates which gave no immunological response to any of the antisera, DDC analysis indicated that four contained Cu-dNir (#39, #87, #192, #193) and one contained ggl-dNir (#83). The data for some of these strains are included in Table 4. The remaining four strains (#97, #103, #108, #115) could not be checked by DDC analysis because of poor growth under denitrifying conditions. In no case were ggl- and Cu-dNir both present in the same isolate. Matrix group was not an adequate basis on which to predict the dNir type type of an isolate. Seemingly closely related strains within time same matrix group had different types, particularly A. faecalis 1solates. The subunit molecular weight of ggl-dNirs varied less than those (’15 Cu-dNirs. At least four distinct Cu-dNirs were recognized after SDS- IEGKGE, with molecular weight ranging from 35,000 to 91,000 (Figure 5). Fifty-one denitrifying isolates from aerobic 2,4-D degrading soil fillrichments were also used for immunological screening. Forty-seven of t:I'Iese isolates contained ggl-dNir. There was no apparent correlation 39 ' “—7 1 1 7+ . i 2 3 4 5 gla°1ofl12 314 ’10'713'50} i 4 "'"' O- Figure 5. western immunoblot of denitrifier crude extracts separated on 10% SDS-polyacrylamide gels and immunoblotted with Cu-dNir antisera. Lane 1, P. aeruginosa ATCC 10145; 2, #41 A. faecalis; 3, #40 A. faecalis; 4, #15 P. fluorescens; 5, #39 Unknown type 3; 6, #171 A. faecalis; 7, #31 A. faecalis; 8, #3 A. faecalis; 9, #43 A. faecalis; 10, #36 Unknown type 3; 11, #144 A. faecalis; 12, #224 P. stutzeri; 13, #192 Bacillus sp.; 14, #106 A. faecalis; 15, #177 A. faecalis; 16, #191 A. faecalis; 17, #224 P. stutzeri; 18, #167 P. aeruginosa; 19, #62 P. aureofaciens; 20, A. cycloclastes ATCC 21921. Arrows show the location of Cu-dNirs. Strain numbers (#) are from Gamble et a1. (1977). Sitrit éenitr with E fiirs evolut evolu‘. There imam known antis com-p1 main .mtl‘ gene 40 between the ability to denitrify and to degrade 2,4-D or between dNir type and 2,4-D degradation. WM The immunological results were confirmed by DDC inhibition assays. Nitrite reductase activity (measured as NO and N20 evolution) in denitrifiers with gal-dNirs was only inhibited 0 to 14% by preincubation with DDC but was almost completely inhibited in denitrifiers with Cu- dNirs (Table 3). The exception was P. aeruginosa (fill-dNir) in which NO evolution was inhibited. This may be an artifact of the low rate of NO evolution. Nitrite reduction to N20 was completely inhibited by DDC. There was complete agreement between the results of DDC inhibition and immunological characterizations. DDC inhibition was also used to differentiate between the two known dNir types in strains which gave no immunological response to the antisera. With the exception of a pseudomonad type (#83), DDC almost completely inhibited dNir activity in crude extracts of these nonimmunoreactive strains (Table 4). Strains containing Cu-dNir included a Bacillus sp. , B. japonicum, and two Rhizobium species. NO but not N20 was evolved by B. japonicum and Rhizobium strain 32Hl. Since N20 reductase activity is readily lost after cell disruption, the Production of NO but not N20 indicates that Fe catalyzed, conversion of NO to N20 was not a factor in these assays. The complete inhibition of NO production by DDC indicates that the members tested in these two genera have Cu-dNir. Tale 3. Strain/ .d. m ?.typr 1 aer ATCC l flu ATCC P. flu (3111 P. Ste F. str (3231 Cu-tyq 4- Cy. ATCC My NCIB 1- is A- fr 4. re A- fa 5. a2 ATCC 41 Table 3. Effect of DDC on nitric and nitrous oxide evolution by denitrifier crude extracts for which nitrite reductase type was classified by immunoblots. Activity 1 (nmol min' mg protein' ) NO N20 DDC % DDC % Strain/type - + Inhib - + Inhib Ldl'typ. P. type 2 (#12) 0 0.2 0 1.9 2.4 0 P. aeruginosa 0.4 0 100 2.2 1.9 14 ATCC 10145 P. fluorescens 0.7 1.5 0 0 0 «- ATCC 33512 P. fluorescens O 0 - 2.0 2.3 0 (#111) P. stutzeri JM300 11.4 9.6 16 2.0 1.9 5 P. stu zeri 2.2 2.4 0 3.2 3.1 3 (#231 ) Cu-type A. cycloclastes 2.9 0 100 3.3 0 100 ATCC 21921 A. xylosoxidans 9.6 1.8 81 5.0 0.4 92 NCIB 11015 A. faecalis (#40) 1.5 0 100 2.6 0 100 A. faecalis (#43) 0.3 0 100 4.2 0 100 A. faecalis (#65) 0 0 - 2.2 0.1 95 A. faecalis (#171) 1.3 0.1 92 0.1 0 92 3. azotoformans 3.9 0.9 77 1.4 0.2 86 ATCC 29786 0. nephridii 3.2 0 100 2.2 0.1 95 ATCC 11425 P - type 19 (#151) 1.1 0 100 4.7 0 100 P- type 19 (#153) 0.5 0 100 1.5 0.1 93 Unknown type 21 0.5 0 100 0.7 0 100 (#189) \ Iable 4 Strair Sacil 3. ja. ATCC Rhizo ihizo 2. sp in 1 ant: dis. Was and 42 Table 4. Effect of DDC on nitric and nitrous oxide evolution by denitrifier crude extracts unreactive to antisera against Cu- or ggl-dNirs. Act vity (nmol min' mg protein'l) NO N20 DDC % DDC % Strain - + Inhib - + Inhib Bacillus sp. (#192) 0 0 - 0.6 0 100 B . japonicum ATCC 10324 0.8 0 100 0 0 - Rhizobium sp. 8A55 1.5 0 100 0.2 0 100 Rhizobium sp. 32H1 2 9 0 100 0 0 - R. sphaeroides f. sp. denitrificans 3.6 0.2 94 0.2 0 100 Unknown type 3 (#39) 0.3 0 100 0.2 0 100 P. type 2 (#83) 0 0 - 0.2 0.5 0 P. type 8 (#87) 0.7 0 100 0 0 - 1W3 Immunodiffusion analysis of the crude extracts from A. cycloclastes, A. xylosoxidans, A. faecalis, A. eutrophus, and P. aureofaciens was done with antiserum against A. cycloclastes Cu-dNir. Immunodiffusion analysis of the crude extracts from P. aeruginosa, P. {fluorescens, P. stutzeri and two denitrifying soil isolates 1.6.L. and 2..4.E, was done with antiserum against P. aeruginosa gal-dNir. The immunodiffusion analysis for P. aeruginosa gdl-dNir is shown ilJn Figure 6. Precipitin lines which formed in the analysis with Cu-dNir liJntiserum were too faint to interpret, presumably because of <1:1ssimilarities in the Cu-dNirs. When P. aeruginosa ggl-dNir antiserum “'as used, precipitin lines formed.with P. aeruginosa, P. fluorescens, 'Etnd isolate 1.6.L. Spur formation suggested partial identity of P. 07A ”'4 43 P. aeruginosa P. fluorescens 2.4.E P. srutzeri 1.6.L P. aeruginow D stutzeri 2.4.E P. fluorescms ‘-6-L P. oeruginoga 1.6.L 2.4.E P. stutter? P. fluorescens Figure 6. Immunodiffusion analysis with P. aeruginosa ggl-dNir antiserum. Center wells contain undiluted P. aeruginosa £1 -dNir antiserum. Outer wells contain crude extract from each of the gdl-type strains as indicated in the figure. fluoresc of P. fl antigen! aeragim - P. st Imuno‘: antise'. dNir a on dNi antise dNir 1 Bass 1 subun faeca 11mm fluo, fais "as 44 fluorescens and 1.6.L with P. aeruginosa ggl-dNir. The precipitin lines of P. fluorescens and 1.6.L. fused, suggesting identity of the ggl-dNir antigens. Neither P. stutzeri, nor isolate 2.4.E, formed precipitin lines. The pattern of antigenic similarity based on response to P. aeruginosa gdl-dNir was P. aeruginosa > P. fluorescens - 1.6.L > 2.4.E - P. stutzeri. Wis Immunoblot analyses were done on bacterial crude extracts with the antiserum used in the immunodiffusion assays and with P. stutzeri £91- dNir antiserum. Equivalent amounts of dNir protein were analyzed based on dNir activity in crude extracts of each strain (Table 5). Cu-dNir antiserum had less affinity for Cu-dNirs other than A. cycloclastes Cu- dNir (Figure 7). As had been previously noted (Figure 5), the molecular mass of Cu-dNir subunits from the different bacteria varied. The subunits ranged from 36 kDa for A. cycloclastes to 42 kDa for A. faecalis #40. All ggl-dNirs in crude extracts were recognized by antiserum raised against P. aeruginosa gdl-dNir (Figure 8). The recognition by antiserum for gdl-dNir other than P. aeruginosa ggl-dNir varied. Subunit Inolecular masses were similar, but a characteristic doublet of immunologically recognized proteins in crude extracts from P. ‘IFIuorescens and isolate 1.6.L. were observed (Figure 8). When antiserum Iraised against P. stutzeri gfil-dNir was used, the recognition pattern ‘Vas different (Figure 9). Pseudomonas stutzeri cfil-dNir was recognized ‘fhile the other gal-dNirs were not. ‘ 5 Tab" RP? 7.. 150] 45 Table 5. In £1219 activity of nitrite reductases in denitrifying bacteria used for immunoblot analysis. Strain Cu types In 31;:9 activity presenta (nKat mg' protein)b 941 type P. aeruginosa na 1272 P. fluorescens na 1500 P. stutzeri na 3897 Isolate 1.6.L. na 2105 Isolate 2.4.E na 2825 Cu-type A. cycloclastes 6: II 2036 A. xylosoxidans I 2720 A. faecalis #43 nd 947 A. faecalis #43 nd 2158 A. eutrophus #154 nd 2450 P. aureot'aciens I 2814 a na - not applicable Dd - not determined 1) Enzyme activity is expressed in units of Katals (Kat), the conversion of one mol of substrate per sec. One nKat - 10'9 Kat. 46 F 15““ 7. Western inunoblot of denitrifier crude extracts containing Cu-dNir incubated with A. cycloclastes Cu-dNir antiserum. Lane 1, A. cycloclastes ATCC 21921; 2, A. xylosoxidans NCIB 11015; 3, P. aursofaciens ATCC 13985; 4, A. faecalis #40; 5, A. faecalis #43; 6, A. eutrophus, #154. Arrows point to the location of Cu-dNirs. 47 Figure 8. Western immunoblot of denitrifier crude extracts containing gdl-dNirs incubated with P. aeruginosa 91 -dNir antiserum. Lane 1, P. aeruginosa ATCC 10145; 2, P. f uorescens ATCC 33512; 3, P. stutzeri JM300; 4, Isolate 2.4.E; 5, Isolate 1.6.L. The arrow points to a characteristic protein recognized in P. fluorescens and Isolate 1.6.L. Molecular mass indicator is in kDa. 48 Figure 9. Western immunoblot of denitrifier crude extracts containing gdl-dNirs incubated with P. stutzeri gdl-dNir antiserum. Lane 1, P. aeruginosa ATCC 10145; 2, P. fluorescens ATCC 33512; 3, P. stutzeri JM300; 4, Isolate 2.4.E; 5, Isolate 1.6.L. The arrow points to the location of P. stutzeri gdl-dNir. 49 mm The immunoblotting technique is quantitative, since the degree of antibody binding to immunologically recognized proteins is related to the intensity of alkaline phosphatase activity at those proteins. So divergence among dNir types was measured by densitometric scanning of the immunologically recognized bands corresponding to dNirs (Figures 7,8, and 9). When the intensity of antibody binding to antigens against which the anitisera were raised were set to 100%, the values for other immunologically recognized proteins in the crude extracts (with the exception of P. fluorescens) were less (Table 6, Table 7) because they have fewer cross reacting determinants. Table 6. Measurement of antibody binding to Cu-dNirs of denitrifying cell lysates. Strain Area Value (%)a A. cycloclastes ' 100 A. faecalis #43 15 A. faecalis #40 11 A . xylosoxidans 10 A. eutrephus #154 2 P . aureofaciens 0 a Antiserum were raised against A. cycloclastes Cu-dNir. The degree of antibody binding at positions corresponding to Cu-dNirs was measured as described in Methods and Materials. The area value for each strain is shown as a percentage of the A. cycloclastes value which was placed at 100%. 50 Table 7. Measurement of antibody binding to ggl-dNirs of denitrifying cell lysates. Strain Area Value (%)a P. aeruginosa 100 P. fluorescens 138 Isolate 1.6.L 91 Isolate 2.4.E 48 P. stutzeri 48 a Antiserum were raised against P. aeruginosa 9g ~dNir. The degree of antibody at positions corresponding to gdl- irs was measured as described in Methods and Materials. The area value for each strain is shown as a percentage of the P. aeruginosa value which was placed at 100%. The evolutionary distance from the type strains, A. cycloclastes, P. aeruginosa, and P. stutzeri, respectively, can be designated as the immunoblot distance (Hershey and Howe, 1984): Immunoblot distance - 100 x log (Intensity of Species) The immunoblot distance of crude extracts with Cu-dNir is shown in Figure 10. Achromobacter cycloclastes Cu-dNir shares some common epitopes with the other Cu-dNirs examined, but there are clear differences in the amount of antibody binding. Cu-dNirs in three of the strains shared similar numbers of common epitopes to A. cycloclastes Cu- dNir. A fourth strain, A. eutrOpbus, shared even fewer common epitopes. Two patterns appeared when looking at immunoblot distances of crude extracts tested with P. aeruginosa ggl-dNir antiserum (Figure 11). Pseudomonas aeruginosa and P. fluorescens gal-dNirs shared many common epitopes, as did isolate 1.6.L gal-dNir. A second group contained 51 V T A. eutro sass“.H.H.“3.853332.“assessfizs“a.“3.333352%“. 2332.35.33”asses“.“gasses.rm." y we“.snoweuonsezsnsaao“a.“a.“as. . M. 180 was.“asses.“232322.“.“so“ . n43 "8" . .m C V: _r c A. r q — q d 11 J u — 4 -‘q u u d a a q 0 O O O O O O O 0 6 4 2 0 8 5 4. 2 mucoumfi. uoBoczEcE Cu—type Denitrifiers Figure 10. Phylogeny of bacteria which crossreact with antiserum to A. cycloclastes Cu-dNir. 52 180 .- ._-_-._ ___-._. 160~ 140.3 120- 100— 80- 60* Immunoblot Distance 40-« 20— 1\\\\\\\\\“ i\\\\\\\\\\1 2 O — — r . 2.4.E 1.6.L rP. oerug t P. flruor T P. slutz r Heme cd—type Denitrifiers Figure 11. Phylogeny of bacteria which crossreact with antiserum to P. aeruginosa ggl-dNir. 53 P. stutzeri and isolate 2.4.E ggl-dNirs, which shared fewer common epitopes with P. aeruginosa gdl-dNir. The analysis could not be performed with P. stutzeri antiserum because recognition of gdl-dNirs other than P. stutzeri ggl-dNir by this antiserum was weak. DISCUSSION We identified the dNir type in over 90% of more than 150 strains for which dNirs were uncharacterized, by using immunoreaction to at least two antisera, and in many cases, also by DDC inhibition assays. Nitrite reductases which were not recognized by antisera were characterized with DDC inhibition assays. The heme gdl-dNir predominated in the environmental isolates. Data for the three culture groups studied: culture collections, numerically dominant denitrifiers, and aerobic soil isolates are summarized in Table 8. Table 8. Distribution of Cu- and heme ggl-dNirs. Culture group Number N02’ reductase type examined heme gdla Cua Unknown? Culture collections 23 44% 56% 0% Numerically dominant 100 64% 32% 4% isolates Aerobic soil 51 92% 8% 0% enrichmentsc a N02“ reductase type based on immunological reactivity and/or response to DDC. Tested by immunological response only. c Isolates not originally selected for denitrification ability. 54 There are two forms of Cu-dNir which have different EPR spectra depending on the type of Cu they contain, either Type I Cu alone or both Type I and 11 Cu. Immunochemical recognition of Cu-dNirs was independent of whether the enzymes had Type I or Type II Cu, since enzymes containing both types have been characterized in strains recognized by our antiserum (Michalski and Nicholas, 1988b; Ohkubo et a1, 1986; Zumft et al., 1987). Achromobacter xylosoxidans and P. aureofaciens Cu-dNirs contain only Type I Cu. The former was recognized by A. cycloclastes (Type I and II) Cu-dNir antiserum (Figure 7) while the latter was poorly recognized not under the conditions employed in this study (Figure 7). This implies that the type of Cu in Cu-dNirs is largely irrelevant to immunological characterization as has been previously noted (Michalski and Nicholas, 1988b). This would not be surprising if the Cu-containing sites are sequestered in dNirs and have no antigenic role. Antiserum against Cu-dNir from R. sphaeroides prepared by Michalski and Nicholas (1988b) recognized Cu-dNirs in the eight denitrifiers they tested, including A. cycloclastes. Although A. cycloclastes Cu-dNir shares common epitopes with other Cu-dNirs, the number of common epitopes between R. sphaeroides Cu-dNir and other denitrifers may be greater than for A. cycloclastes. This may explain why some Cu-dNirs were not recognized by our antiserum. Another possibility is that these unrecognized strains contain a non Cu, non gdl-type dNir heretofore uncharacterized. The strains we used represented frequently studied laboratory cultures as well as the predominant denitrifiers from soils of four continents, sediments, and sludges (Gamble et al., 1977). Pseudomonas P’!’ P'. er h: 55 spp. are the most prevalent denitrifiers in these environments (Tiedje, 1988) and 73% of the denitrifying Pseudomonas or Pseudomonas type isolates (Tables 1 and 2) contained gdl-dNirs. This suggests that 941- dNirs predominate in nature. They were also the most prevalent dNir among isolates from our study regardless of the geographic origin or environment from which the isolates were collected. This could be an inevitable affect of the isolation methods used to recover the denitrifiers. Bacillus spp., for example, were the predominant population in one study of denitrifying soil isolates from undefined sites (Abd-el-Malek et al., 1974). However, even when selection was not originally made for denitrification, eg. the isolates from 2.4-D degrading soil enrichments, the denitrifiers among these isolates predominantly contained ggl-dNirs. Cu-dNirs, nevertheless, may be significant in denitrifying ecosystems. Cu-dNir has been found in more physiological groups and ecological niches than the gdl-dNir. These groups include chemolithotrophic nitrifiers (Ritchie and Nicholas, 1974), heterotrophic nitrifiers (Kuenen and Robertson, 1987), extreme halophiles (L. Hochstein, personal communication), and phototrophic bacteria (Sawada et al., 1978) as well as pseudomonads (Michalski and Nicholas, 1988b; Zumft et al., 1987). In addition to these groups we observed Cu-dNir in hydrogen oxidizers (A. eutrophus), symbiotic NZ-fixers (B. japonicum), and gram positive spore formers (B. azotoformans). As more genera are examined, ggl-dNirs appear in other taxonomic groups besides Pseudomonas. In addition to P. fluorescens, we observed ggl-dNir in Aquaspirillum, Azospirillum, and Flavobacterium. 56 Cu-dNirs were more heterogeneous than ggl-dNirs. Subunit molecular weights of Cu-dNirs were similar to some, but not all, of those which have been reported: 39,000 (Liu et al., 1986); 37,000 and 39,500 (Michalski and Nicholas, 1985); 40,000 (Zumft et al., 1987). Most were larger (around 40,000) than the molecular weight of the A. cycloclastes standard (35,000)(Hulse et al., 1988). The molecular weight of an immunoreactive protein in Alcaligenes isolate, #171, was much greater than other Cu-dNirs (Figure 5). Cu-dNir from Alcaligenes strain 86 is a tetramer of molecular weight 120,000 which required extensive denaturing at 50 'C before the various oligomers were separated (Kakutani et al., 1981). Isolate #171 may contain a similar enzyme. The variability of Cu- dNirs among denitrifiers may reflect significant differences in Cu-types (Masuko et al., 1984), or simple differences in processing of the enzyme. It may also explain why we were unable to recognize all Cu-dNirs with our antiserum. We showed the utility of using DDC inhibition assays to identify dNir type in a variety of different physiological groups as had been proposed by Shapleigh and Payne (1985). In all cases, DDC inhibited dNir activity in denitrifiers with Cu- but not ggl-dNirs. This verified the results of the immunoscreening. For the nonimmunoreactive isolates, the assays demonstrated, with one exception, that all contained Cu-dNir. The dNir type in virtually all strains used in this study could therefore be identified using a combination of these two methods. With the exception of A. itersonii, all of the genera which were assigned a dNir type were examined with at least two independent tests (immunological or chemical) to establish that designation. 57 DDC inhibition assays and immunoscreening are complementary methods of characterizing dNir types in denitrifying populations. The DDC inhibition assay unambiguously distinguishes Cu- from ggl-dNir. The method is rapid and the analysis simple to perform, but it does not reveal immunological relatedness and heterogeneity among dNirs as antibodies do for the dNirs they recognize. For simple identification of dNir type in a denitrifier, the DDC inhibition assay is preferred because it is independent of dNir recognition. If many isolates are screened, raising antibodies against specific dNirs becomes a reasonable alternative because multiple isolates can be examined simultaneously. The intensity of immunological recognition varies in imunoscreened denitrifier crude extracts. This is undoubtedly due to differences in the amount of dNir present, and the similarity of dNirs in the crude extracts to dNirs against which antiserum were raised. Nonspecific binding may occur, but this can be minimized by adjusting the incubation time and antiserum dilution. Nitrite reductase type may differ among closely related strains - notably Pseudomonas spp. (Table l)(l(6rner et a1. , 1987; Michalski and Nicholas, 1988b; Zumft et al., 1987). Among A. faecalis isolates, 47% contained Cu-dNir and 53% contained gal-dNir (Table 2). ”hi-13 dNir types do not appear to be absolutely conserved within a species, they do appear to be incompatible within the same host. No isolate had more than "‘3 t3)’P¢ present. One can only speculate about evolutionary mechanisms behind the distribution of 9511- and Cu-dNirs among strains. Bacterial phylogenies show that dNir types are widely distributed among many different genera (“0986. 1987) and many different physiological groups (Ti-641°: 1988)‘ 58 Denitrification may have risen in an ancestral photodenitrifier, and like photosynthesis in numerous branches of the purple bacteria, been lost in some strains over time (Betlach, 1982). This would lead to closely related denitrifying (i.e. dNir-containing) and nondenitrifying species. Examples of truncated denitrifiers are numerous (Tiedje, 1988). Differences of dNir type among Pseudomonas and Alcallgenes spp. might be explained by horizontal gene transfer. Nitrite reductase may be plasmid borne in A. eutrophus (Romermann and Friedrich, 1985). Gene transfer provides a plausible route by which dNirs could have spread among pseudomonads and other groups. Since Cu- and sail-dNirs were never observed in the same strain, this could involve dNir acquisition by truncated denitrifiers or replacement of a preexisting dNirs. It is also possible that the two dNir types originated at very different times: the Cu-type very early with the variation resulting from divergence, and the heme ggl-type much later as a variation of “‘78“! respiration. The immunoblot method was used to quantify “valence among dNirs in denitrifying bacteria. This method has the advantage of sensitivity compared to immunodiffusion assays, and Simplicity compared to microcomplement fixation techniques (Howe and Hershey. 1931; 1934). The validity of the results rests on a key assumption that the dNir added from each sample was similar. If the specific activity of din-“differed greatly among strains, the results would be confounded. Assuming,that specific activities are similar, if not identical in some instances, immunoblot analysis is a rapid method.to visually iii-SP“t differences in dNirs from different species. The deNirs examined differed in terms of molecular mass and antigenicity. Achromobacter 59 cycloclastes Cu-dNir was not closely related to any of the other Cu- dNirs based on antibody affinity. Achromobacter xylosoxidans and two A. feecalis species had similar antibody affinities. Alcaligenes eutrephus shared the fewest common epitopes with A. cycloclastes Cu-dNir. Although Michalski and Nicholas (1988b) reported recognition of all Cu-dNirs they used by.R. sphaereides Cu-dNir antiserum, they did not quantify the degree of recognition. Our results suggest that the Cu-dNirs against which antisera are raised have a profound effect on recognition. 1 It was noted (Table 2) that the immunologic response of gdl-dNirs sometimes differed depending on what type of gdl-dNir the antisera were raised against, i.e. P. aeruginosa vs. P. stutzeri. This suggested‘ greater variability in ggl-dNirs than was indicated by comparisons of molecular mass. Antiserum raised against P. aeruginosa ggl-dNir recognized the ggl-dNirs of all strains examined by immunoblot analysis. That P. fluorescens has 138% of the binding of P. aeruginosa, probably reflects slight differences in the amount of enzyme loaded. Three strains, P. aeruginosa, P. fluorescens, and isolate 1.6.L had similar antibody affinity, while the affinities of two other strains, isolate 2.4.E and P. stutzeri were closer to each other than to the first group. Pseudomonas aeruginosa, P. fluorescens, and P. stutzeri are in the same DNA homology group as defined by Palleroni et a1. (1973). Pseudomonas aeruginosa and P. stutzeri also share the same RNA homology group. Consequently, it's surprising that the gdl-dNir they possess seem to differ more than P. aeruginosa and P. fluorescens ggl-dNir. The results confirm earlier observations during screening of many denitrifiers with gdl-dNirs, that the gfil-dNirs also have diverse antibody recognition. in t res] wit} VET flu bee by St] (19 e? 0t 60 There was a characteristic doublet of strongly antigenic bands in P. fluorescens and isolate 1.6.L, which suggests that they are related. Purified ggl-dNirs have all been dimeric, and composed of identical subunits. Michalski and Nicholas (1985) reported that R. sphaeroides Cu- dNir was a dimer composed of non identical subunits. Pseudomonas fluorescens ggl-dNir may also have nonidentical subunits, but has yet to be purified. The higher molecular mass band may represent unprocessed ggl-dNir still attached to a leader region required for its transport and insertion into the cell membrane. The DNA sequence of another denitrifying enzyme, N20 reductase, is consistent with a leader region of approximately 35 amino acid residues (Viebrock and Zumft, 1988). The difference in antigenicity between the two groups of ggl-dNirs in this study - exemplified by P. aeruginosa and P. stutzeri, respectively - was further revealed when the same ggl-dNirs were probed with antiserum to P. stutzeri ggl-dNir (Figure 9). Only P. stutzeri ggl- dNir was strongly recognized under these conditions. Although proteins were recognized in the other strains, only those in isolate 1.6.L and P. fluorescens could be quantified. The immunogenically recognized high molecular mass protein in both of these strains, which had previously been recognized by P. aeruginosa cdl-dNir antiserum, was not recognized by P. stutzeri ggl-dNir antiserum. Kerner and Zumft (1987) noted in immunodiffusion assays, that the gdl-dNir of a P. perfectomarina (now P. stutzeri strain Zobell) only cross reacted with one other ggl-dNir type denitrifier - also P. stutzeri. It appears that there is a suite of epitopes common to P. stutzeri type ggl-dNir that is not found in the other ggl-dNirs. 61 Proteins which differ by more than 40% in amino acid sequence generally fail to show cross reactivity (Cocks and Wilson, 1972). Based on protein sequence differences derived from immunoblot distance (Cocks and Wilson, 1972) other Cu-dNirs differed from A. cycloclastes Cu-dNir by some 20-30 % of the protein sequence. The difference between 31-1" dNirs is less than 10%. To what extent these differences reflect the time for divergence, the conservation of structure, or the diversity of organisms is unclear given the few dNirs which have been examined. In the same functional class of proteins, sequence evolution goes on at approximately the same rate (Wilson et a1. , 1977). In a given protein, evolution will depend on the compatibility of sequence change with biochemical function and the likelihood of organism survival without that biochemical function (Wilson et a1. , 1977). That Cu-dNirs aIlpear to diverge more than fill-dNirs may reflect a greater capacity of 1'lonheme cu-dNirs to accept sequence changes without losing function compared to heme gal-dNirs. Since the basic criterion for the selection of these organisms was their ability to denitrify, it might be more productive to study dNir divergence by looking for dNirs in organisms which have lost the ability to denitrify, NO3' respirers for example, on the asstmption that these organisms may represent truncated denitrifiers which have lost the capacity to make functional dNirs though not the enzymes themselves. SECTION 2 LOCALIZATION 0F CYTOCHROME Q2 AND CU NITRITE REDUCTASES IN DENITRIFYING BACTERIA 62 63 INTRODUCTION Denitrification is a respiratory process in which nitrogenous oxides serve as terminal electron acceptors in the absence of 02 (Firestone, 1982). During the reduction of N02' to N20, energy is conserved by oxidative phosphorylation (Koike and Hattori, l975a,b). (Scansequently, N02' reductases (dNirs) in denitrifying bacteria have been [Dearceived as membrane-bound components of energy generation. ‘lJtnderstanding this role requires information about their cellular ].<:calization. Three approaches have been used: immunochemical detection, Inueasurement of proton translocation, and cellular fractionation. 19erritin-conjugated antibody was used by Saraste and Kuronen (1978) to Tlocalize dNir in Pseudomonas aeruginosa at the inner face of the cytoplasmic membrane. Boogerd et al. (1981) placed dNir from Paracoccus denitrificans in the periplasmic space based on proton translocation studies. This was in contrast to similar studies on the same organism by Kristjansson et a1. (1978) which gave dNir a cytoplasmic orientation. Cellular fractionation approaches have proved more confusing. In P. aeruginosa, dNir was periplasmic, in contrast to the immunochemical results (Wood, 1978). Nitrite reductase was found in the periplasm of P. denitrificans after sphaeroplast formation (Alefounder and Ferguson, 1980), but some dNir activity remained bound to the sphaeroplast membranes. Mancinelli et a1. (1986) found dNir in cytoplasmic and membrane fractions of P. halodenitrificans. Nitrite reductase in the 64 membrane fraction was tightly bound. It was found on the cytoplasmic [ membrane of Thiobacillus denitrificans (Sawhney and Nicholas, 1978), and both the cytoplasmic membrane and cytoplasm of P. perfectomarina (now P. stutzeri strain Zobell)(Zumft and Vega, 1979). Its presence in the cytoplasm was credited to loose binding to the cytoplasmic membrane. A more recent report placed it as a soluble periplasmic protein in P. stutzeri (Mingawa and Zumft, 1988 cited in Heiss et a1. , 1989). All of these cultures contained gal-type dNir. Among organisms with a Cu-type dNir, only Rhodobacter sphaeroides has been examined for its cellular location, and there it was found in the periplasm (Sawada and Satoh, 1980). We recently raised specific antibodies to P. aeruginosa gdl-dNir and Achromobacter cycloclastes Cu-dNir. As an alternative to cellular fractionation or proton translocation studies, we attempted to indirectly visualize dNir location by immunocytochemical localization using colloidal gold probes. This method has been used successfully to study other systems including: restriction endonucleases in Escherichia coli (Kohring et a1. , 1985), component C of the methyl reductase system in methanogens (Ossmer et a1. , 1986), methyl-coenzyme M reductase in Methanobacterium thermosutotrophicum (Aldrich et a1. , 1987), and APS reductase as well as bisulfite reductase in Desulfovibrio spp. (Kramer et al., 1988). In the following study, the localization of 5:511- and Cu-dNirs in denitrifying bacteria are reported based on cellular fractionation and immunocytochemical localization. Immunocytochemical localizations were J - performed with goat anti rabbit-gold complexes (Auroprobe EM GARlS) on v.3 p.- (1‘. 65 resin embedded cells grown either aerobically, or anaerobically with KNO3 as a terminal electron acceptor. The portions of the study for which I was solely responsible were the cellular fractionation assays. Otherwise, the study represents a joint effort. Purified Cu-dNir was provided by C. Hulse of the University of Virginia. Purified mil-dNir was provided by E. Weeg- AAerssens of Michigan State University. Antibody production and cell growth were a joint effort between myself and Dr. Alahari Arunakumari. Electron microscopy and sample preparation were done by Stu Pankratz, Department of Microbiology and Public Health, Michigan State University. METHODS AND MATERIALS W A. cycloclastes ATCC 21921, P. aeruginosa ATCC 10145, and P. fluorescens ATCC 33512 were obtained from the American Type Culture Collection. Achromobacter xylosoxidans NCIB 11015 and P. stutzeri JM300 were obtained from H. Iwasaki and J. Ingraham, respectively. Strains were grown in 15 g l"1 TSB supplemented to 1 uM CuSOa'7H20 and 10 mM K1903. Plates and slants for routine culture contained the same media with 1.5% agar. For cell fractionation, bacterial strains were grown in sealed 500 ml sidearm flasks'with 250 ml of media. The headspace was made anaerobic by. repeated flushing with sterile argon at a gassing manifold, and the flasks were inoculated with 12.5 ml of a late log phase culture of anaerobically-grown cells. Growth was monitored by the optical density at 660 nm during incubation at 30 'C. For immunogold localization, cells were either taken from aerobic slants 66 or plates, or from anaerobic broth cultures which had dissimilated all . NO3' in the media. W The osmotic shock method of Yamamoto et a1. (1987) was used to release periplasmic proteins. Cells in early stationary phase were harvested and washed twice with 10 mM Tris pH 7.5 containing 25% sucrose. Washed cells were resuspended in the same buffer containing 1 mt! EDTA and incubated with shaking for 10 min at 24 'C. The cells were then centrifuged at 7000 rpm for 10 min, the supernatant decanted, and the cells quickly resuspended in ice-cold 10 mM Tris pH 7.5. The Suspension was incubated on ice with periodic shaking for 10 min and a~fterwards centrifuged at 9000 rpm for 10 min. The supernatant was decanted and saved. The pelleted cells were resuspended in 10 mM Tris pH 7 . 5 and sonicated. The periplasmic fraction was defined as those Proteins released into the supernatant after treatment with the cold buffer. The cytoplasmic fraction, containing cytoplasm and cell membranes, was defined as those proteins found in the supernatant after sonication of the pelleted cells. Protein concentrations in both fractions were determined by Bicinchoninic acid protein assay reagent (BCA - Pierce) using BSA as a standard, after pelleting cell debris. W Nitrite reductase activity was measured by gas chromatography as described in Section 1. Acid phosphatase activity was measured by the increase in absorbence at 405 nm in a 2 ml reaction volume containing 0.05% Triton X100, 0.1 ml cell extract, and 4.5 mM p-nitrophenol phosphate in 0.1 M acetic acid buffer pH 4.5 (Absalom, 1986). 67 G1ucose-6-phosphate (G-6-P) dehydrogenase activity was measured by the increase in absorbance at 340 nm. Stock solutions in 50 mM Tris buffer pH 8.9 were prepared for NAD+ (30 mM), (NADP+ 30 mM), and D- glucose-6-phosphate (200 mM). The reaction consisted of 0.2 ml of NAD+ or NADP+ stock solution, 0.2 m1 of G-6-P, cell extract (between 0.05 and 10.2 ml) and Tris buffer to a final volume of 2 m1 (Maurer et al., 1982). Watts Antisera against A. cycloclastes Cu-dNir and P. aeruginosa ggl- (iliir were prepared as described in Section 1. WWW Cells from denitrifying broth cultures and aerobic plate cultures "Uiere pelleted and resuspended in 1% glutaraldehyde, 0.1 M sodium <=¢acodylate pH 7.2 for 40 min then washed three times in 0.1 M sodium eacodylate for a total of 45 min. The cells were dehydrated in an increasing graded series of ethanol washes before embedding in Lowacryl K4M resin overnight. The samples were polymerized at 24 ‘C under a UV lamp for one and a half days. After polymerization, thin sectiOns were cut from the resin with a diamond knife, collected on water, then onto nickel grids and stored at room temperature until use. Grids with resin embedded cells were incubated in TBST buffer (10 mM Tris pH 7.4, 0.5% Tween 80, 2.5% NaCl) containing 1% BSA. The grids were then transferred to 1:200 dilutions of primary antibody (in TBST), preimmune sera, or buffer (depending on the treatment). After 60 min, grids were washed 3 X 10 min in TBST and incubated 60 min with a 1:250 dilution of goat anti-rabbit 15 nm colloidal gold particles (Auroprobe EM GAR15 - Janssen Life Science Products). The grids were then washed 3 X 10 min with TBST before a final counterstain with uranyl acetate. 68 Grids were examined and photographed on a Philips CM-lO electron microscope. RESULTS Mild osmotic shock did not separate periplasmic and cytoplasmic components of denitrifying cells into discrete fractions. In four of the five strains, most dNir activity was in the cytoplasmic fraction after osmotic shock (Table l), as was true of acid phosphatase activity, a marker enzyme for periplasmic proteins. The distribution of dNir activity in cytoplasmic and periplasmic fractions did not always match the distribution of acid phosphatase activity between the same two fractions, although, it was close in P. fluorescens and P. stutzeri (Table 1). The recovery of total dNir and acid phosphatase activities in the periplasmic fractions were A. cycloclastes 27% vs. 3%, A. xylosoxidans l%< vs. 3%, P. aeruginosa 3% vs. 31%, P. fluorescens 27% vs. 18%, and P. stutzeri 88% vs. 83%. These results indicate that mild osmotic shock did not completely release the periplasmic proteins in all strains, as might be expected for different bacterial species. However, in all strains, a cytoplasmic enzyme such as glucose-6-phosphate dehydrogenase was not found in the periplasmic fraction. This indicates that the integrity of cytoplasmic membranes was maintained in those cells which did respond to the osmotic shock (Table 1). These data suggest a periplasmic location for both dNir types. ’d 69 Table 1. Cellular distribution of enzyme activities. Strain Enzyme Fraction NO ' Acid G- 6- P Total type ' Reduc- Phospga- Dehydro- Protein tasea tase genasec (mg ml'l) .A. cycloclastes Cu Cyto 2562 223 115 1.40 ATCC 21921 Peri 924 6 0 0.15 .A. xylosoxidans Cu Cyto 5904 250 13 1.60 NCIB 11015 Peri 14 10 0 0.12 P. aeruginosa ggl Cyto 8795 91 13 1.65 ATCC 10145 Peri 231 40 0 0.22 P. fluorescens ggl Cyto 74 141 44 1.85 ATCC 33512 Peri 28 30 0 0.20 P. stutzeri gdl Cyto 54 10 31 0.90 JM300 Peri 394 50 0 0.39 a Activity measured as nmol N 0 min hi b Activity measured as umol phosphaten8 min c Actiyity1 measured as nmol NAD+ or NADP+ reduced min Cytoplasmic (cyto) and periplasmic (peri) fractions are as defined in the Methods and Materials. Immunocytochemical detection of dNirs using colloidal gold probes was used as a second, indirect, approach to determine their localization in denitrifying cells. Four strains were used, containing either Cu (A. cycloclastes. A. xylosoxidans) or 931 (P. aeruginosa, P. fluorescens) dNirs. The specificity of the primary antiserum for each dNir type was previously established by Western immunoblot assay (see Section 1). When colloidal gold probes were used alone, or in conjunction with preimmune sera, there was no binding of gold particles to anaerobically-grown (induced) cells. Likewise, when aerobically-grown (uninduced) cells were 70 incubated with primary antisera and colloidal gold probes, there was little or no binding of gold particles (Figures 1,4,6, and 8). An anaerobically-grown, Cu-type denitrifier, A. cycloclastes, incubated with gdl-dNir antiserum and colloidal gold probes, had some binding of gold particles, but these mostly appeared on the cell wall and were assumed to be nonspecific (Figure 3). In contrast, when anaerobically-grown A. cycloclastes and A. xylosoxidans were incubated ‘with the Cu-dNir antiserum, followed by colloidal gold probes, numerous gold particles were associated with the periplasmic space of both strains (Figures 2 and 5). This was also the case with anaerobically- grown P. aeruginosa and P. fluorescens incubated with cdl-dNir antiserum and colloidal gold probes (Figures 7 and 9). Few gold particles were found in the cytoplasm indicating that dNirs of both type were not soluble cytoplasmic enzymes in these four strains. Pi 71 Figure 1. Electron micrograph of aerobically-grown A. cycloclastes incubated with Cu-dNir antiserum and treated with 15 nm colloidal gold probes (magnification 48,450X). 72 Figure 2. Electron micrograph of anaerobically-grown A. cycloclastes incubated with Cu-dNir antiserum and treated with 15 nm colloidal gold probes (magnification 66,300X). Arrow points to periplasmic location of the gold particles. 73 Figure 3. Electron micrograph of anaerobically-grown A. cycloclastes incubated with ggl-dNir antiserum and treated with 15 nm colloidal gold probes (magnification 48,450X). 74 Figure 4. Electron micrograph of aerobically-grown A. xylosoxidans incubated with Cu-dNir antiserum and treated with 15 nm colloidal gold probes (magnification 48,450X). 75 Figure 5. Electron micrograph of anaerobically-grown A. xylosoxidans incubated with Cu—dNir antiserum and treated with 15 nm colloidal gold probes (magnification 48,450X). Arrow points to localization of the gold particles in the periplasmic space. 76 Figure 6. Electron micrograph of aerobically-grown P. aeruginosa incubated with cd -dNir antiserum and treated with 15 nm colloidal gold probes (magnification 48,450X). 77 Figure 7. Electron micrograph of anaerobically-grown P. aeruginosa incubated with gdl-dNir antiserum and treated with 15 nm colloidal gold probes (magnification 48,450X). Arrow points to localization of the gold particles around the cytoplasmic membrane. 78 Figure 8. Electron micrograph of aerobically-grown P. fluorescens incubated with gdl-dNir antiserum and treated with 15 nm colloidal gold probes (magnification 48,450X). 79 Figure 9. Electron micrograph of anaerobically-grown P. fluorescens incubated with ggl-dNir antiserum and treated with 15 nm colloidal gold probes (magnification 88,400X). Arrow points to localization of gold particles in the periplasmic space. 80 DISCUSSION Two independent means were used to localize dNirs in denitrifying bacteria: cell fractionation and immunogold labeling. Mild osmotic shock expelled periplasmic proteins such as acid phosphatase without the concomitant expulsion of cytoplasmic proteins such as glucose-6- phosphate dehydrogenase. Nitrite reductases of both ggl- and Cu-type were expelled along with the acid phosphatase. Most of the dNir and acid phosphatase activity remained with the cytoplasmic fraction, which contained cytoplasm and cell membranes. This was expected, since periplasmic proteins will only be lost if the cell wall is grossly disrupted (Alefounder and Ferguson, 1980). Only 57% of the total dNir activity in osmotically shocked R. sphaeroides was recovered in the periplasmic fraction by Sawada and Satoh (1980). Likewise, only 55% of a periplasmic blue Cu protein was recovered from A. fsecalis strain S6 in the osmotic shock procedure used by Yamamoto et al. (1987). An alternative to the osmotic shock procedure was fractionation by sphaeroplast formation. Sphaeroplasts and sphaeroplast supernatants contained 53% and 47%, respectively, of total N02' reductase activity in preparations of P. denitrificans made by Alefounder and Ferguson (1980). A sphaeroplast preparation of R. sphaeroides made by Sawada and Satoh (1980) indicated that sphaeroplast supernatants (i.e. the periplasmic fraction) contained 90% of the dNir activity. However, the sphaeroplast procedures of Sawada and Satoh (1980) and Maagd and Lugtenberg (1986) gave no better results than osmotic shock in preparations of A. cycloclastes which were done as preliminary experiments. A third alternative, expulsion of periplasmic proteins by chloroform treatment 81 (Amos et al., 1984) also gave no better results than the osmotic shock procedure. So, osmotic shock was chosen since it was the easiest method for all strains. It was interesting that the magnitude of dNir and acid phosphatase release into the periplasmic fraction did not always coincide, as might be expected if both are periplasmic proteins. This could reflect differences in the degree of binding between dNir and the cytoplasmic membrane. In P. aeruginosa (Saraste and Kuronen, 1978) and P. perfectomarinus (P. stutzeri) (Cox and Payne, 1973; Zumft and Vega, 1979) the dNirs were described as loosely bound in contrast to P. halodenitrificans in which membrane associated dNir was tightly bound (Mancinelli et al., 1986). With respect to A. cycloclastes, P. fluorescens, and P. stutzeri, the dNirs were more readily expelled from cells by osmotic shock than were acid phosphatases (Table 1). Nitrite reductases in A. xylosoxidans and P. aeruginosa were not expelled as well as acid phosphatases, perhaps indicating tighter association of these dNirs to the cytoplasmic membrane. Immunochemical labelling has been previously used to localize denitrifying enzymes. Most studies with ggl-dNirs have been confined to P. denitrificans and P. aeruginosa. With the exception of R. sphaeroides, Cu-dNirs have not been localized. Our results indicate a specific binding of gold particles only to those cells in which dNir was induced. This binding was not associated with the cytoplasm, but was associated with the cytoplasmic membrane and/or periplasmic space in all the strains examined. These results differ from studies with P. halodenitrificans in which 99% of the dNir activity was associated with the cytoplasm 82 (Mancinelli et a1. , 1986), but do agree with other studies which associate dNirs with the cytoplasmic membrane or periplasmic space (Alefounder and Ferguson, 1980; Sawada andSatoh, 1980; Wood, 1978; Zumft and Vega, 1979). The immunogold localization indicates a periplasmic for the Cu- dNirs of A. cycloclastes, A. xylosoxidans, P. aeruginosa, and P. fluorescens. This was an indirect localization and the dimensions of the antibody-gold complexes have to be considered. The gold particles in the goat, anti rabbit-gold complexes may be up to 20 nm away from the actual location of the dNirs (Kohring et al. , 1985). In cells where plasmolysis has occurred, there is a distinct localization of gold particles between cell wall and cytoplasmic membrane in all four strains. Boogerd et a1. (1981) developed a model based on indirect evidence from alkalinization of external media, to explain proton translocation and electron flow during denitrification. This model required a periplasmic location for dNir to permit external H+ consumption (Figure - 10). The results from our immunogold localization with two Cu-type denitrifiers and two sail-type denitrifiers support this model and a periplasmic location for both dNir types. 3’ PERIFLASMC SHE 2&3H‘4—— ZH'e— ZH’ ‘— ascapcre'fi dehydro- ascoroote oH‘.—/ N05 .JH’ as~,a.r 5H,O¢—/ ”:0 02H."\ ~,.H,a .3 L J Edi} 83 MEMBRANE f—E 2H 4. WW CW. ”40‘ -2&3H‘ 29' Figure 10. Anaerobic respiratory chain of P. denitrificans and its proton-translocating properties. Fp, flavoprotein; Fe-S, iron sulfur center; 0H2, ubiquinol; Hy, hydrogenase; NR, nitrate reductase; cyt. , cytochrome; N O-R, nitrous oxide reductase; Ox, alternate oxidase. I and I a + IIb are the traditional sites of energy conservation (From F.C. Boogerd, Ph.D. Thesis, Free University of Amsterdam, 1984). SECTION 3 INDUCTION or N-OXIDE REDUCTASES m OXYGEN-LIMITED camosm'r CULTURES OF W W 84 85 INTRODUCTION Oxygen represses the synthesis and activity of N-oxide reductases (Payne, 1973). In environmental samples, denitrification generally begins between 3 and 12 uM 02 (Tiedje, 1988), however, some well-aerated environments evolve N20 (Smith and Tiedje, 1979). One frequently invoked source is anaerobic microsites (Parkin, 1987), although, other explanations include chemodenitrification, autotrophic nitrification (Bremner and Blackmer, 1978), heterotrophic nitrification (Castignetti and Hollocher, 1984), and fungal N-metabolism (Yoshida and Alexander, 1970). It is possible, though not yet demonstrated, that N20 evolution from soils comes from aerobic denitrification. Denitrification in Thiosphaera pantotrepha was reported at greater than 90% air saturation (Robertson and Kuenen, 1984a). N-oxide reductases have reportedly been synthesized at relatively high 02 concentrations in Hyphomicrobium X (Meiberg et al., 1980). A.magnetic Spirillum, MS-l, has an absolute requirement for 02 even during denitrifying growth (Escalante-Semerena et al., 1980). Ecologically-oriented denitrification studies typically focus on rates of the entire process, spatial changes in zones supporting denitrification, or temporal changes in pathway intermediates (Firestone and Tiedje, 1979; Rice et al., 1988; Smith and Tiedje, 1979). Few studies have looked for the relationship between specific N-oxide 86 reductases and discrete 02 concentrations which either induce their synthesis or repress their activity. This would be particularly important knowledge for the management of denitrification in any commercial process which seeks to employ denitrifiers as alternatives to obligate aerobes. Nitrate and nitrite reductase were derepressed at 15 and 91 uM 02, respectively, in Thiobacillus denitrificans continuous cultures undergoing a gradual shift to anaerobiosis (Justin and Kelly, 1978). The denitrification products in Paracoccus halodenitrificans continuous cultures undergoing a gradual shift from anaerobiosis to aerobic growth were, sequentially, N2, N20, N02', and N03' (Hochstein et al., 1984). This was interpreted as N-oxide reductase inhibition at different 02 levels in the growing cells. Nitrite still comprised 74% of the N-oxide present when dissolved 02 concentrations were below the detection level of the 02 probe (3 uM). Kerner and Zumft (1989) found discrete 02 levels at which NO3', NOZ', and N20 reductases were derepressed in continuous culture of P. stutzeri strain Zobell. N20 reductase was constitutively expressed, while NO3' and N02' reductases were synthesized at 5.0 and 2.5 mg 02 1'1 (156 and 78 uM 02), respectively. These strains contain gdl-type N02' reductase (ggl-dNir)(Grant and Hochstein, 1984; Kerner et al., 1987; LeGall et al., 1979). No similar studies have been done with organisms containing Cu-type N02' reductase. The purpose of this study was to examine the conditions required for Cu- type N02' reductase synthesis in Achromobacter cycloclastes (Iwasaki and Matsubara, 1972). and to determine the dissolved 02 concentration at which derepression of N02" reductase and other N-oxide reductase synthesis and activity occurs. 87 METHODS AND MATERIALS WWW Achromobacter cycloclastes ATCC 21921 was grown in T88 media (Difco, Detroit MI), 15 g 1'1 TSB, 1 uM CuSOa'7H20, 10 11:14 10103, for routine culture and maintenance. Slants and plates contained 1.5 % agar. Batch culture experiments contained 6 g 1'1 TSB. Chemostat media consisted of 20 mM phosphate buffer pH 6.5 containing: 1 g 1'1 TSB, l g 1'1 NHACl, 0.1 g 1'1 10:03, and 1 uM CuSOa'7H20. Warm}: Continuous culture was in a one liter LH 500 Series fermenter with media outlet set for a 500 ml working volume. An Ingold polarographic 02 electrode, pH electrode, thermistor, heating and cooling rods, sampling column, and the gas in- and outlet tubes penetrated the lid of the chemostat. Temperature was maintained at 30 ’C and stirring speed at 1000 rpm. The high stirring rate and low carbon levels used in the chemostat were intended to insure low cell density and optimal exposure to the measured 02 concentration. Two point calibrations at 0 and 100% air-saturation, were performed by immersing the 02 electrode in water saturated with N2 and atmospheric air, respectively. Gas flow was regulated by flow controllers and adjusted to give the desired air-saturation as recorded by the 02 electrode. Gas flow through the media was 100 ml min'1 and was filter sterilized through a Millex-GS 0.22 uM filter unit (Millipore Corp., Bedford, MA). The dilution rate (D) was 0.05 hr’l. Chemostat experiments were begun by inoculating air-sparged chemostats with 5 ml of late log phase cells grown aerobically in chemostat media which lacked KNO3. Media flow was begun immediately 88 after inoculation, and sampling started when optical density measurements (at 660 nm) were stable. W Nitrous oxide formation during batch culture experiments was monitored on a Shimadzu GC-mini 2 gas chromatograph equipped with a 63Ni electron capture detector. Carrier gas was 95% argon, 5% methane. The temperatures were: detector 300 'C, injector and column 30 'C. The column was 2 m x 2 mm i.d. stainless steel filled with Porapak Q (Waters Assoc. Inc., Framingham, MA). Gas flow was 46 ml min'l. Nitrous oxide formation and disappearance in continuous culture was monitored on a Perkin-Elmer 910 gas chromatograph equipped with dual columns and 63Ni electron capture detectors for simultaneous analysis of two samples (Kaspar and Tiedje, 1980). Conditions were as above except that carrier gas flow was 15 ml min'1 which was adequate to separate NO and 02 peaks. Nitric oxide was obtained from Matheson Corp., as were N20 gas standards. W Procedures for sample preparation and immunoblotting are given in Section 1. Purified N20 reductase from A. cycloclastes was obtained from C. Hulse (C. Hulse, Ph.D. thesis, University of Virginia, 1989). Silver staining indicated that the protein was electrophoretically pure (see Figure 1, Section 1). The emulsified protein was used to immunize a New Zealand white rabbit according to the protocol outlined in Section 1. The antiserum obtained was used without further purification. WW3 Twenty ml of A. cycloclastes was removed from an aerobic batch culture growing without N-oxides and 0.5 m1 aliquots were dispensed into 89 replicate 3.0 ml Venoject tubes (Terumo Medical Corp., Elkton, MD). Treatments were: 10 mM Na2W04 + 1 mM KNO3; 1 mM KNO3; 1 mM Diethyldithiocarbamate (DDC) + 1 mM NaNOz; 1 mM NaNOZ; 0.5 ml of 100% ‘N20; no amendment. The final volume was adjusted to one ml in all cases ‘with sterile batch culture media. The tubes were stoppered and incubated at 30 ‘C for 24 h without evacuating the headspace. After 24 h, the cells in one tube from each treatment were pelleted and washed in 50 mM HEPES buffer pH 7.3. Protein was determined in these samples by the Lowry method (Lowry et al., 1951) with bovine serum albumin as a standard (Sigma). The cells in the remaining tube from each treatment ‘were used for Western immunoblot analysis. Based on the protein analysis, 50 ug total protein from each treatment were loaded into the sample wells for SDS-PAGE. Growth was in continuous culture at D - 0.05 h'l. After one volume change at each air saturation level, the OD (660 nm) was measured, the chemostat headspace was sampled, and a sample of growth medium was removed. The aqueous sample was centrifuged and the supernatant analyzed for N03' by absorbance at 210 nm following HPLC separation (Christensen and Tiedje, 1988) and N02' by absorbance at 543 nm following diazotization (Hanson and Phillips, 1981). To detect the induction of N-oxide reductases, whole cell assays were used. Approximately 40 ml of growth medium was removed, cells harvested by centrifugation at 10,000 rpm for 5 min at 4 'C, washed twice in 20 mM phosphate buffer pH 6.5 containing 10 mM glucose and 1 mg ml'1 chloramphenicol, and resuspended in 10 m1 of the wash medium. The 90 washed cells were kept on ice for the duration of sampling and used within 4 hours. To detect the activity of N-oxide reductases these protocols were used: NO3' reductase - 0.5 ml aliquots of the washed cells were dispensed into replicate serum vials, amended with 0.5 m1 of 1 mM KNO3’, sealed, flushed with argon, and incubated at 25 °C for 1 h. At 10 min intervals, a vial was removed and frozen at -70 °C until N02' accumulation or N03' disappearance could be measured as noted previously; N02' reductase - 0.5 m1 aliquots were dispensed into two replicate 15 ml serum vials, amended with 0.5 m1 of 1 mM NaNOZ, sealed, flushed with argon, amended with 0.1 ml of C232 to block N20 reduction (Yoshinari and Knowles, 1976)(prepared by hydrating calcium carbide in an argon atmosphere), and incubated at 25 'C. At 10 min intervals the vials were shaken and the headspace of each vial was sampled by gas chromatography for N20 production; NO reductase - 0.5 ml aliquots were dispensed into two replicate 15 m1 serum vials, amended with 0.5 ml of wash medium, sealed, flushed with argon, amended with 0.1 ml of 100% NO and 0.1 ml of C2H2, and incubated at 25 'C. At 10 min intervals N20 evolution was monitored by gas chromatography; N20 reductase - 0.5 ml aliquots were dispensed into two replicate 15 ml serum vials, amended with 0.5 ml wash medium, sealed, flushed with argon, and flushed l min with a 30 ppm N20 gas standard. The vials were incubated at 25 'C and at 10 min intervals N20 disappearance was monitored by gas chromatography. BOiled controls were used for all treatments. Additional 0.5 ml aliquots 0f the washed cell suspensions were removed and stored at ~70 'C for analysis of protein and for immunoblotting. 91 RESULTS MW N02' reductase synthesis was detected by Western immunoblot. Synthesis occurred in all treatments except when N-oxides were absent or when DDC was present (Figure 1). Synthesis appeared greatest when NO3' was present. Presumably, N02“ reductase was not synthesized when DDC was present because the cells had little or no growth. One hundred micromolar DDC severely retarded aerobic growth of A. cycloclastes in a preliminary experiment. Total protein yield in the DDC amended treatment was approximately half that of the other treatments (Table 1). Table 1. Effect of different substrates and inhibitors on the protein yield of semianaerobically incubated A. cycloclastes. Incubation Treatment Final Protein Yield (ug ml'l) 10 mM woaz' + 1 mM N03' 344 1 mM N03' 469 1 mM DDC + 1 mM N02‘ 211 1 mM N02' 406 0.5 ml N O 469 No N-oxi 359 Nitrite reductase was synthesized in the presence of 10 mM W042' which prevents formation of functional NO3‘ reductases by competitively inhibiting Mo incorporation into molybdoproteins (Scott et al., 1979). .Assimilatory NO3' reductases should also have been inhibited by the 'Pretein-rich TSB media as well as by the WOA'. So, N02' reductase synthesis does not require prior synthesis of functional NO3' 92 Figure 1. Synthesis of A. cycloclastes Cu-dNir in the indicated treatments. C - Control, 1 ug of A. cycloclastes Cu-dNir; W - Na W0 ; DDC - Diethyldithiocarbamate; -N - No N-oxide terminal electron acceptors. 93 reductases. Without NO3’ reductase, anaerobic growth would be limited as the results in Table 1 indicate. Apparently, sufficient growth occurred on residual 02 in the tubes to permit N02' synthesis (since there was more protein in this treatment than the treatment with DDC, which inhibits growth), or N03' reductase was present in the batch cultures which circumvented W042' inhibition. The 02 concentrations at which N-oxide reductase proteins were synthesized and became active was determined. Dissolved 02 concentrations ranged from approximately 230 to 0 uM. The lowest 02 concentration above 0 (NZ-sparged chemostat) that could reliably be measured by the electrode was 2 uM. The carbon content of the medium (1 g 1'1 TSB) allowed a maximum optical density of 0.19 (Table 2). The optical density in fully anaerobic growth was 55% that of fully aerobic growth. This is consistant with observations that denitrifying growth yield is approximately half that of 02 respiration (Koike and Hattori, 1975a). Trace N20 was present at all 02 concentrations throughout the continuous culture. It was not considered physiologically significant. Trace N02’ was also present at all 02 concentrations and could have been due to low constitutive levels of NO3' reductase activity. Full derepression of NO3' and N027 reductases was not apparent until the measured 02 concentration was 2 uM (Table 2). 94 Table 2. Denitrification by A. cycloclastes during continuous culture. % Air 0 on N03‘8 N02'8 Denitrificationb saturation (ug) (660 nm) (uM) (uM) flux (umol N l‘lh'l) 100 228 0.19 1290 4 nd 50 114 0.20 1190 4 2.5 20 46 0.19 1150 4 3.4 10 23 0.18 990 8 7.4 4 9 0.16 830 4 11.5 1 2 0.14 310 0 24.6 0 0 0.10 ' 0 0 32.4 a Concentration in the chemostat medium. b After one volume change at the various air saturation levels, the effluent gas phase and the growth medium were measured for the indicated N03", NOZ', NO, and N O. No NO and only trace N 0 were observed in the chemostat headspace. The denitrification lux is based on the difference between the total NO '-N in the resevoir medium and sum of NO3'- and N02'-N in the chemostat medium. The activities of individual N-oxide reductases were measured in washed cell suspensions removed from continuous culture at each 'dissolved 02 concentration (Table 3). Nitrate reductase activity was present at all 02 concentrations from complete air-saturation to complete anaerobiosis. Further enzymatic reduction of N02' was expected at the lowest 02 concentrations, since NOz' reductase was active under these conditions. So, for the lowest 02 concentrations, NO3' reductase activity was measured by NO3' disappearance. Nitrite reductase was stringently regulated by 02- No activity was present above 2 uM 02, nor was any enzyme synthesized until 2 uM 02 (as detected by Western immunoblot) (Figure 2). Full induction did not occur until complete anaerobiosis was reached. For comparison, under optimal denitrifying growth conditions in rich medium (15 g l"1 T88, 10 mM KNO3), N03- 95 Figure 2. Western immunoblot of Cu-dNir synthesis in A. cycloclastes continuous cultures during increasing anaerobiosis. C - Control, 1 ug A. cycloclastes Cu-dNir; FI - Fully induced culture. Values indicate % air saturation. Corresponding dissolved oxygen concentrations are given in Table 3. Marker indicates A. cycloclastes Cu-dNir. 96 reductase activity was 3.5 umol-N min'1 mg'l while N02' reductase activity was 5.4 umol-N min'1 mg'l. Table 3. In viyg activity of N-oxide reductases in A. cycloclastes washed cells removed from continuous culture at varying air saturations. % Air N-oxide reductase activity (nmol-N min’l mg'l) Saturation uM 02 N03 ->N02 902 ->N20 no ->N20 N20 ->N2 100 228 6 i o8 o o 0 so 114 10 r o8 o o o 20 45 32 i 18 o o o 10 23 67 r g“ o o 27 r 9 4 9 220 o o 94 r 20 1 2 326b 973c 2235 r 333 151 r 97 o 0 581b 1917 r 190 1784 r 91 31 i 12 a NO3' reductase activity measured by N02' accumulation. b NO3' reductase activity measured by NO3' disappearance. One replicate only. c Replicate lost. Nitric oxide reducing activity was also stringently controlled by dissolved 02 concentrations. No activity was observed until the dissolved 02 concentration was 2 uM. In contrast, low levels of N20 reductase protein were synthesized at all dissolved 02 concentrations (Figure 3) but activity was not detected until the 02 concentration fell to 23 uM. 97 Figure 3. Western immunoblot of N20 reductase synthesis in A. cycloclastes continuous cultures during increasing anaerobiosis. C - Control, 1 ug A. cycloclastes N 0 reductase; FI - Fully induced culture. Values ind cate % air saturation. Corresponding dissolved oxygen concentrations are given in Table 3. Marker indicates N20 reductase. 98 DISCUSSION These results demonstrated the conditions in which N02- reductase was synthesized in A. cycloclastes. They also demonstrated the 02 concentrations at which N-oxide reductases were synthesized or derepressed in A. cycloclastes continuous cultures. Anaerobiosis alone was not sufficient for the synthesis of N02' reductase. The complementary presence of an N-oxide was also required. Nitrate had the greatest effect on enzyme induction. When tungsten was incorporated into one induction treatment to block formation of functional N03" reductase, it had no apparent effect on N02' reductase synthesis. Subsequent experiments demonstrated NO3' reductase activity in aerobic cells, so presynthesized N03" reductase in these treatments probably provided N02‘ for enzyme induction. Korner and Zumft (1989) found that NO3' had the greatest effect on overall enzyme induction in P. stutzeri strain Zobell, but enzyme induction for each terminal reductase was greatest when its specific N-oxide was present. The effect of NO3' on N-oxide reductases other than NO3' reductases was explained as the result of further reduction of N03' to N02' and N20. Pseudomonas denitrificans (Payne, 1973) and Rhodobacter sphaeroides f. sp. danitrificans (Michalski and Nicholas, 1984) reportedly synthesize Cu-type N02' reductase in anaerobic conditions alone. In R. sphaeroides, induction was greater with a terminal N-oxide. The requirement of a complementary N-oxide for enzyme synthesis may be strain specific. Maintenance of some N-oxide reductases anaerobically in the absence of N-oxides could be a competitive advantage. Isolates from NO3'-free freshwater sediments and digested sludge had no lag in 99 denitrification activity when amended with N03- (K.S. Jorgensen and J.M. Tiedje, Abstr. Annu. Meet. Am. Soc. Microbiol. 1989. N87, p. 299). Nitrite reductase was the most stringently regulated N-oxide reductase. No enzyme was synthesized at detectable limits above 2 uM 02, nor was enzyme activity noted above this level. Kerner and Zumft (1989) found N02" reductase expression at S 2.5 mg 02 l'1 (78 uM). The discrepancy between these results could reflect true differences between the strains studied, or differences in 02 experienced by the cells caused by different cell densities, stirring speeds, and gas sparging rates. Christensen and Tiedje (1988) noted that efficient mixing of 02 in solution was a prerequisite for establishing meaningful denitrification thresholds. Thus, with high culture density, and uneven mixing, anaerobic zones resulting from respiration could develop around bacteria or clumps of bacteria, causing them to sense a greater 02 deficit than probes indicate. Nitrite was present at all 02 levels and N03' reductase activity were observed in washed cells from this study and by Kbrner and Zumft (1989). They attributed this to assimilatory NO3' reductase activity even in the presence of ammonium, because their antibody to dissimilatory NO3' reductase did not detect the presence of this protein, and the~antiserum did not cross react with assimilatory NO3’ reductase. I believe this represents low levels of constitutive dissimilatory NO3' reductase activity since, in my study, the activity increased continuously as 02 concentration decreased, and the immunological methods used by Korner and Zumft were insensitive to low protein concentrations. Krul (1976) reported that N03'reductase in a denitrifying Alcaligenes sp. was constitutive. A heterotrophic, .100 nitrifying Alcaligenes sp. also synthesized NO3' reductase in nitrifying, aerobic conditions (Castignetti and Hollocher, 1982). Nitrate reductase activity was affected by growth rate in Hyphomicrobium X (Meiberg et al., 1980). At D - 0.15 h'1 uo3' reductase activity was significant only at low 02 (9.3 uM), while at D - 0.01 h'1 NO3' reductase was present at relatively high 02 (78 uM). In contrast, at D - 0.12 h"1 (Korner and Zumft, 1989) expression of N03' reductase was at 169 uM 02 while I observed NO3' reductase expression at 228 uM 02 with n - 0.05 h'l. Trace N20 was evolved at all 02 concentrations during continuous culture. This was not considered physiologically important, but possible sources other than denitrification are: NO3' reductase acting on N02' (Cole, 1987), an uncharacterized NO3' assimilation pathway (Bleakley and Tiedje, 1982; Yoshida and Alexander, 1970), or heterotrophic nitrification (Castignetti and Hollocher, 1982, 1984; Eylar and Schmidt, 1959; Obaton et al., 1968; Robertson and Kuenen, 1988; Verstraete and Alexander, l972a,b;). Nitrous oxide reductase was synthesized at all dissolved 02 concentrations from air-saturation to full anaerobiosis (Figure 3). However, no activity was apparent until dissolved 02 concentration fell below 23 uM. This may represent synthesis of precuser forms of the enzyme which do not have activity until inserted into the membrane (Cole, 1987). This is consistant with Kerner and Zumft's (1989) results in which P. stutzeri N20 reductase was constitutively synthesized, and for which expression was increased when 02 fell below a discrete level (119 uM oz). 101. Nitrous oxide reductase activity in washed cell assays was approximately 10 times less than N02" reductase activity (Table 3). In direct measurements of the chemostat headspace N20 did not accumulate, despite a rapid flux of NO3' through the system (Table 2). This illustrates a phenomenon noted with other denitrifiers, that exogenously supplied N20 is not reduced as rapidly as internal N20 derived from N02" reduction. Some P. aeruginosa strains are unable to grow well on exogenously supplied N20 as a direct consequence of turnover dependent inactivation of N20 reductase (Snyder et al., 1987). Nitric oxide was never detected during direct measurement of the air-sparged chemostat. If there are separate N02' and NO reductases, they would appear to be similarly regulated by 02. Nitric oxide reducing activity was only observed when 02 concentrations fell to 2 uM. Zafiriou et a1. (1989) found NO evolution from sparged, P. perfectomarina anaerobic batch cultures. As sparging rate increased, the proportion of gaseous N-oxide recovered as NO increased. Nitrite reductase had the highest activity of any of the N-oxide reductases in fully anaerobic conditions (Table 3). Since only low levels of NO3' reductase are synthesized before N02' reductase is induced, it is unlikely that A. cycloclastes faces situations in which N02” accumulates to toxic levels during denitrification. What regulatory model explains differential induction of N-oxide reductases? The rationale is simple, formation of a given electron acceptor is prevented when an alternate terminal electron acceptor that would give a higher energy yield is present (Stouthammer, 1988). While this explains why 02 respiration is preferable to N03’ respiration, it doesn't adequately explain why NO3' respiration should be preferred to 102 N02', NO, or N20 respiration, since oxidative phosphorylation occurs to a similar extent with each reductive step (Gerber et al., 1982; Koike and Hattori, l975b). Nor does it explain why N20 reductase in P. stutzeri and A. cycloclastes is constituitively synthesized. Oxygen controls denitrification at several levels, transcription, translation, and at least with N20 reductase, direct inactivation (Stouthammer, 1988). There is no direct effect of 02 on N03' reductase. Membrane vesicles from Pa. denitrificans reduced 02 and N03' simultaneously (John, 1977; Parsonage et al., 1985). Both ggl- and Cu- type N02" reductases have oxidase activity (Zumft et al., 1987). Each N- oxide reductase could control its own synthesis by autogenous regulation (Stouthammer, 1988). In this model, N-oxide reductase binding with the appropriate N-oxide results in conformational changes that prevent the protein from inhibiting transcription. In the redox control model, the oxidation reduction status of the electron transport chain components, or the terminal N-reductase, regulates synthesis (Robertson and Kuenen, 1984b). The redox state of the electron transport chain could also regulate the assembly of the N- oxide reductases into active, membrane bound forms, thus explaining how synthesis of N-oxide reductases and activity of N-oxide reductases can be temporally separated (Stouthammer, 1988). Terminal electron acceptors compete for available e'. In Thiosphaera pantotropha, aerobic denitrification is explained, in part, by suggesting that e' flow to oxygen as the terminal electron carrier is impeded, thus shunting e' to alternate terminal reductases (Robertson and Kuenen, 1984b). These models still require rigorous proof. APPENDIX APPENDIX A PRELIMINARY EFFORTS TO CLONE CU-TYPE NITRITE REDUCTASE Studies on the physiological genetics of denitrification invariably employ mutants in the reductive steps. We attempted to clone the structural gene for Cu-type N02' reductase (Cu-dNir) with the goal of creating defined mutations in this enzyme. Three basic approaches were used: (1) Probing with synthetic oligonucleotides derived from the N-terminal amino acid sequence of Achromobacter cycloclastes Cu-dNir. (2) Creating expression libraries in the phage vector lambda gtll followed by immunoscreening for expression of Cu-dNir. (3) Generating DNA probes for Cu-dNir using mixed primer Polymerase Chain Reaction (PCR). With respect to the first approach, a cosmid library was constructed in the cosmid vector pWH4 (a gift from Peter Wolk, DOE Plant Research Lab, Michigan State University). Total genomic DNA from A. cycloclastes ATCC 21921, A. xylosoxidans NCIB 11015, P. aeruginosa ATCC 10145, and P. stutzeri JM300 were partially digested with Sau 3A and 103 104 individually ligated into Bam HI digested, alkaline phosphatase treated, pWH4 using methods outlined by Maniatis et a1. (1982). The concatamers formed, were used as substrates for in vitro packaging of bacteriophage particles, which were then introduced into E. coli strain H8101. Based on plasmid minipreps of 10 random kanamycin resistant colonies from each genomic digest, approximately 60% of the clones contained an insert. The clones were stored enmass in 50% glycerol at -70 °C. Purified A. cycloclastes Cu-dNir was obtained from C. Hulse (Hulse et al., 1988) and the N-terminal amino acid sequence determined at the MSU macromolecular sequencing facility. The first 30 amino acids were obtained (Figure l) and two synthetic oligonucleotides, each 17 bases long, were constructed by the sequencing facility from nonoverlapping sections of the N-terminus (Figure 2). Both probes were mixed. Probe #1 contained 256 different species, while Probe #2 contained 1024 different species. 1 2 3 4 5 6 7 8 9 10 ALA PRO VAL ASP ILE SER THR LEU PRO ARG ll 12 l3 14 15 l6 17 18 19 20 VAL LYS VAL ASP LEU VAL LYS PRO PRO PHE 21 22 23 24 25 26 27 28 29 30 VAL ALA ALA ALA ASP GLN VAL ALA LYS THR Figure l. N-terminal sequence of Cu-dNir from A. cycloclastes ATCC 21921. 105 Peptide sequence: MET ALA PRO VAL ASP ILE Probe #1: 5'- ATNTCNACNGGNGCCAT - 3' Peptide sequence: ASP GLN VAL ALA LYS THR Probe #2: 5' - GTNTTNGCNACNTGNTC - 3' Figure 2. Synthetic oligonucleotides derived from the N-terminal amino acid sequence of A. cycloclastes Cu-dNir. N - A+G+C+T. The oligonucleotides were detritylated, 5' end-labeled with 32P- ATP, and used as radioactive probes for colony hybridizations, slot blots, and Southern blots. Although putative hybridization signals were obtained in some colonies, none could be authenticated in subsequent hybridizations. Slot blots were also unsuccessful because of nonspecific hybridization. The apparent lack of sensitivity, and specificity, of the oligonucleotide probes, led to the second approach used in the cloning experiments. An expression library was prepared by Dr. A. Arunakumari from total Eco RI digests of A. cycloclastes. These digests were ligated to the phosphatase treated arms of the phage vector lambda gtll (Promega Corp.) which was packaged, and used to infect E. coli strain Y1090r' (Huynh et a1. , 1985). Production of B-galactosidase fusion proteins was induced by the addition of IPTG, and could then be immunologically screened by Cu-dNir specific antiserum. Self-ligated lambda gtll was identified by blue plaque formation with X-gal, an indicator of functional B-galactosidase activity. Plaques carrying inserts putatively associated with Cu-dNir, were identified by immunoscreening with Cu-dNir antiserum after proteins were 106 adsorbed to nitrocellulose. Immunoscreening protocols and verification of antibody specificity are given in Section 1. Eight immunoreactive plaques were obtained and passed through rescreening. Phage DNA was obtained using Promega lambda-sorb phage adsorbent and DNA isolation protocols. The DNA was exhaustively digested with Eco RI, but no inserts were recovered. The restriction pattern of Ban HI digested phage DNA, compared to the Bam HI restriction pattern of unligated lambda gtll, suggested that phage DNA contained inserts. Subsequent restriction digests with other endonucleases demonstrated that there were at least four unique inserts among the eight phage isolates. None of the isolated phage, which were tested for immunologically recognized B-galactosidase fusion proteins in Western blots, gave a protein of higher molecular mass than B-galactosidase. Immunologically recognized proteins with a slightly higher molecular mass than A. cycloclastes Cu-dNir were observed in some of the phage lysates. The possibility that these represented authentic degradation fragments of B- galactosidase fusion protein bearing Cu-dNir epitopes was raised, but not explored further. The final experimental cloning approach which has been used, is generation of DNA probes using mixed primer polymerase chain reaction (PCR) (Girgis et al., 1988). PCR reactions have been a collaborative effort between myself and Dr. Geoffrey Smith, with assistance provided by Dr. Jim Cole and Dr. W. Holben. An attempt to clone the heme gfil-dNir was included in the PCR approach. The N-terminal amino acid sequence of P. stutzeri JM300 ggl- 107 dNir was previously determined by MSU's macromolecular sequencing facility for R. Ye, and is given in Figure 3. l 2 3 4 5 6 7 8 9 ALA ALA PRO ASP MET XXX ALA GLU GLU 10 ll 12 l3 14 15 16 17 18 LYS ALA ALA LYS LYS ILE TYR PHE GLU 19 20 21 22 23 24 25 26 27 ARG XXX ALA GLY XXX HIS GLY VAL LEU 28 29 30 31 32 33 34 35 LEU LYS GLY ALA THR GLY LYS ASN Figure 3. The N-terminal amino acid sequence of heme gdl-dNir from P. stutzeri JM300. XXX indicates that the amino acid is not known. A search for sequence data on N02' reductase using Sequence Analysis Software (SAS, Genetics Computer Group, Madison, WI) was unsuccessful. No information was available in any of the data bases searched, including: GenBank, EMBL, VecBase, NBRF-Nucleic, and NBRF- Protein. The structural gene for N20 reductase from P. stutzeri has been cloned and sequenced by Viebrock and Zumft (1988) and the codon usage identified. We used this codon usage as a basis on which to generate the PCR primers. In cases where the codon usage was redundant, the most frequently used bases at each wobble position were employed. This reflected a bias towards G and C at the wobble position (Viebrock and Zumft, 1988). Primers were prepared by Genetic Designs Inc. (Houston, TX) and were used without further purification or verification that they represented the desired sequences. Each primer was 5' to 3;, 15 bases long, and from nonoverlapping regions of the N-terminus. Heme A 108 represents the codon choices for amino acids 1-5 (Figure 4). Heme 8 represents the codon choices for amino acids 13-17 (Figure 4). The total length of the anticipated amplified gene product was 51 base pairs. Home A 5' - GCSGCSCCGGAYATG - 3' Heme B 5' - GAAGTAGATTCTTCTT - 3' Figure 4. PCR oligonuleotide primers derived from the N-terminal amino acid sequence of P. stutzeri gdl-dNir. S - G+C; Y - T+C. The codon usage of P. stutzeri N20 reductase was also used to generate primers for Cu-dNir since no information on any proteins from A. cycloclastes were available in the data bases searched. Each primer was 5' to 3', 15 bases long, and from a nonoverlapping region of the N- terminus of A. cycloclastes Cu-dNir. These primers were also synthesized by Genetic Designs Inc. Cu A represents the codon choices for amino acids 5-9 (Figure 5) and Cu B represents the codon choices for amino acids 25-30 (Figure 5). The total length of the anticipated amplification product was 75 base pairs. Cu.A 5' 7 ATCTCCACCCTSCCG - 3' Cu B 5' - GGTCTTSGCSACCTG - 3' Figure 5. PCR oligonucleotide primers derived from the N-terminal amino acid sequence of A. cycloclastes Cu-dNir. S - G+C. 109 PCR was done using the Gene Amp Kit from Perkin Elmer Cetus and thermostable Inn Polymerase in the standard Cetus PCR experimental protocol. Annealing temperatures, and extension time were varied, but no amplified gene products corresponding to the anticipated product length were obtained. It was thought that perhaps the codon choice had made the primers nonspecific by overemphasizing G and C in the wobble position. New primers for P. stutzeri ggl-dNir were obtained from Genetic Designs Inc., which were completely mixed. CD1 was 15 bases long, and represented codon choices for amino acids 31-35 (Figure 6). CD2 was 15 bases long, and represented codon choices for amino acids 13-17 (Figure 6). The anticipated amplification product length was 69 base pairs. CD 1 5' - RTTYTTNCCVGTNGC - 3' CD 2 5' - AARAARATHTAYGAR - 3' Figure 6. PCR oligonucleotides derived from the N-terminal amino acid sequence of P. stutzeri gg -dNir. H - A+C+T; N - A+G+C+T; R - G+A; V - A+G+C; Y - T+C. This ended my actual contribution to the cloning work. Subsequent PCR with the new primers has also failed to give amplified gene products in the desired size range. LIST OF REFERENCES 10. LIST OF REFERENCES . Abd-el-Malek, Y., I. Hosny, and N.F. Emam. 1974. Evaluation of media used for enumeration of denitrifying bacteria. Zentralbl Bakteriol Parasitenkd. Infektionskr. Hyg. Abt. 2 129:415-421. . Absolom, D.R. 1986. Basic methods for the study of phagocytosis. Methods Enzymol. 132:95-182. . Aerssens, E., J.M. Tiedje, and B. Averill. 1986. Isotope labeling studies on the mechanism of N-N bond formation in denitrification. J. Biol. Chem. 261:9652-9656. . Aldrich, H.C., D.B. Beimborn, M. Bokranz, and P. Schonheit. 1987. Immunocytochemical localization of methyl- coenzyme M reductase in Methanobacterium thermoautotrophicum. Arch. Microbiol. 147: 190- 194. . Alefounder, P.R.,‘and S.J. Ferguson. 1980. The location of dissimilatory nitrite reductase and the control of dissimilatory nitrate reductase by oxygen in Paracoccus denitrificans. Biochem. J. 192: 231- 240. . Amos, G. F.-L., C. Prody, and S. Kustu. 1984. Simple, rapid, and quantitative release of periplasmic proteins by chloroform. J. Bact. 160:1181-1183. . Barber, D., S.R. Parr, and C. Greenwood. 1976. Some spectral and steady-state kinetic properties of Pseudomonas cytochrome oxidase. Biochem. J. 157:431-438. . Baumann, L., and P. Baumann. 1978. Studies of relationship among terrestrial Pseudomonas, Alcaligenes, and Enterobacteria by an immunological comparison of glutamine synthetase. Arch. Microbiol. 119:25-30. . Bazylinski, D.A., and R.P. Blakemore. 1983. Denitrification and assmilatory nitrate reduction in Aquaspirillum magnetotacticum. Appl. Envirn. Microbiol. 46:1118-1124. Berg, P., L. Klemedtsson, and T. Rosswall. 1982. Inhibitory effect of low partial pressures of acetylene on nitrification. Soil Biol. Biochem. 14:301-303. 110 11. 12. l3. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 111. Betlach, M.R. 1982. Evolution of bacterial denitrification and denitrifier diversity. Antonie van Leeuwenhoek. J. Microbiol. 48:585-607. Betlach, M.R., and J.M. Tiedje. 1981. Kinetic explanation for accumulation of nitrite, nitric oxide, and nitrous oxide during bacterial denitrification. Appl. Environ. Microbiol. 42:1074-1084. Betlach, M.R., J.M. Tiedje, and R.B. Firestone. 1981. Assimilatory nitrate uptake in Pseudomonas fluorescens studied using nitrogen- 13. Arch. Microbiol. 129:135-140. Bleakley, B.H., and J.M. Tiedje. 1982. Nitrous oxide production by organisms other than nitrifiers or denitrifiers. Appl. Environ. Microbiol. 44:1342-1348. Boogerd, F.C., H.W. VanVerseveld, and.A.H. Stouthamer. 1981. Respiration-driven proton translocation with nitrite and nitrous oxide in Paracoccus denitrificans. Biochem. Biophys. Acta. 638:181-191. Brenner, J.M., and.A.M. Blackmer. 1978. Nitrous oxide: Emission from soils during nitrification of fertilizer nitrogen. Science. 199:295-296. Carlson, R.R., and A.K. Vidaver. 1982. Taxonomy of corynebacterium plant pathogens, including a new pathogen of wheat, based on polyacrylamide gel electrophoresis of cellular proteins. Int. J. Syst. Bacteriol. 32:315-326. Castignetti, D., and T.C. Hollocher. 1982. Nitrogen redox metabolism of a heterotrophic, nitrifying-denitrifying Alcaligenes sp. from soil. Appl. Environ. Microbiol. 44:923-928. Castignetti, D., and T.C. Hollocher. 1984. Heterotrophic nitrification among denitrifiers. Appl. Environ. Microbiol. 47:620-623. Christensen, 8., and J.M. Tiedje. 1988 Oxygen control prevents denitrifiers and barley plant roots from directly competing for nitrate. FEMS Microbiol. Ecol. 53:217-221. Cleveland, D.W., S.G. Fischer, M;W. Kirshener, and U.L. Laemmli. 1977. Peptide mapping by limited proteolysis in sodium dodecyl sulfate and analysis by gel electrophoresis. J. Biol. Chem. 252:1102-1106. Cocks, G.T., and A.C. Wilson. 1972. Enzyme evolution in Enterebacteriaceae. J. Bact. 110:793-802. Cole, J.A. 1987. Assimilatory and dissimilatory reduction of nitrate to ammonia. p. 281-329. In J.A. CCole and S. Ferguson (eds.), The nitrogen and sulfur cycles. Cambridge University Press, Cambridge. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 112 Cox, C.D., and W.J. Payne. 1973. Separation of soluble denitrifying enzymes and cytochromes from Pseudomonas perfectomarinus. Can. J. Microbiol. 19:861-872. DeMaagd, R.A., and B. Lugtenberg. 1986. Fractionation of Rhizobium leguminosarum cells into outer membrane, cytoplasmic membrane, periplasmic, and cytoplasmic components. J. Bact. 167:1083-1085. De Vos, P., and J. De Ley. 1983. Intra- and intergeneric similarities of Pseudomonas and Xanthomonas ribosomal ribonucleic acid cistrons. Int. J. Syst. Bact. 33:487-509. Dispirito, A.A., L.R. Taaffe, J.D. Lipscomb, and A.B. Hooper. 1985. A 'blue' copper oxidase from Nitrosamonas europaea. Biochem. Biophys. Acta. 827:320-326. Escalante-Semerena, J.C., R.P. Blakemore, and R.S. Wolfe. Nitrate dissimilation under microaerophillic conditions by a magnetic Spirillum. Appl. Environ. Microbiol. 40:429-430. Eylar, O.R., and E.L. Schmidt. 1959. A survey of heterotrophic micro-organisms from soil for ability to form nitrite and nitrate. J. Gen. Microbiol. 20:473-481. Firestone, M.K. 1982. Biological denitrification, p. 289-326. In F.J. Stevenson (ed.), Nitrogen in Agricultural Soils. ASA.Inc., Madison. Firestone, M.K., and J.M. Tiedje. 1979. Tempbral change in nitrous oxide and dinitrogen from denitrification following onset of anaerobiosis. Appl. Environ. Microbiol. 38:673-679. Focht, D.D. 1981. Soil denitrification, p. 499-516. In J.M. Lyons (ed.), Genetic engineering of symbiotic nitrogen fixation and conservation of fixed nitrogen. Plenum Press, New York. Focht, D.D., and W. Verstraete. 1977. Biochemical ecology of nitrification and denitrification. Adv. Microb. Ecol. 1:135-214. Fox, G.E., et a1. 1980. The phylogeny of prokaryotes. Science. 209:457-. Gamble, T.N., M.R. Betlach, and.J.M. Tiedje. 1977. Numerically dominant denitrifying bacteria from world soils. Appl. Env. Microbial. 33:926-939. Garber, A.B., D. Castignetti, and T.C. Hollocher. 1982. Proton translocation and proline uptake associated with reduction of nitric oxide by denitrifying Paracocccus denitrificans. Biochem. Biophys. Res. Comm. 107:1504-1507. Girgis, 8.1., M. Alevisaki, P. Denny, L.J.M. Ferrier, and S. Legon. 1988. Generation of DNA probes for peptides with highly degenerate codons using mixed primer PCR. Nuc. Acids. Res. 16:10371. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 113 Grant, M.A., and L.I. Hochstein. 1984. A dissimilatory nitrite reductase in Paracoccus halodenitrificans. Arch. Microbiol. 137:79-84. Gudat, J.C., J. Singh, and D.C. Wharton. 1973. Cytochrome oxidase from Pseudomonas aeruginosa. I. Purification and some properties. Biochem. Biophys. Acta. 292:376-390. Hanson, R.S., and.J.A. Phillips. 1981. Chemical composition, p. 328-364. In P. Gerhardt (ed.), Manual of methods for general bacteriology. ASM, Washington. Heiss, 8., K. Frunzke, and W.G. Zumft. 1989. Formation of the N-N band from nitric oxide by a membrane-bound cytochrome b; complex of nitrate respiring (denitrifying) Pseudomonas stutzeri. J. Bact. 171:3288-3297. Hochstein, L.I., M. Betlach, and G. Kritikos. 1984. The effect of oxygen on denitrification during steady-state growth of Paracoccus halodenitrificans. Arch. Microbiol. 137:74-78. Hochstein, L.I., and G.A. Tomlinson. 1988. The enzymes associated with denitrification. Ann. Rev. Microbiol. 42:231-261. Hogrefe, C., D. Romermann, and B. Friedrich. 1984. Alcaligenes eutrophus hydrogenase genes (Hox). J. Bact. 158:43-48. Howe, 3.0.", and 3.37.13. Hershey. 1981. A sensitive immunoblotting method for measuring protein synthesis initiation factor levels in lysates of Escherichia coli. J. Biol. Chem. 256:12836-12839. Howe, J.G., and J.W.B. Hershey. 1984. The rate of evolutionary divergence of initiation factors IF2 and IF3 in various bacterial species determined quantitatively by immunoblotting. Arch. Microbiol. 140:187-192. Hulse, C.L., J.M. Tiedje, and B.A. Averill. 1988. A spectrophotometric assay for dissimilatory nitrite reductases. Anal. Biochem. 172:420-426. Huynh, T.V., R.A. Young, and'R.W. Davis. 1985. Constructing and screening cDNA libraries in gth and gtll, p.49-78. In D.M. Glover (ed.), DNA cloning: A practical approach, Vol I. IRL Press, Weshington. Iwasaki, H., S. Shidara, H. Suzuki, and T. Mori. 1963. Studies on denitrification. VII. Further purification and properties of denitrifying enzyme. J. Biochem. 53:299-303. Iwasaki, H., and T. Matsubara. 1971. Cytochrome g-557 (551) and cytochrome cg of Alcaligenes faecalis. J. Biochem. 69:847-857. Iwasaki, H., and T. Matsubara. 1972. A nitrite reductase from Achromobacter cycloclastes. J. Biochem. 71:645-652. 52. 53. 54. 55. S6. 57. 58. 59. 60. 61. 62. 63. ‘64. 114 Iwasaki, H., S. Noji, and S. Shidara. 1975. Achromobacter cycloclastes nitrite reductase: The function of copper, amino acid composition, and ESR spectra. J. Biochem. 78:355-361. Jeter, R.M., and J.L. Ingraham. 1981. The denitrifying procaryotes. p. 913-925. In M.P. Starr (ed.) The Procaryotes. Springer-Verlag, New York. John, P. 1977. Aerobic and anaerobic respiration monitored by electrodes. J. Gen. Microbiol. 98:231-238. Jurd, R.D. 1981. Immunoelectrophoresis,p.229-248. In B.D. James and D. Rickwood (eds.), Gel electrophoresis of proteins. IRL Press, Washington, D.C. Justin, P., and D.P. Kelly. 1978. Metabolic changes in Thiobacillus denitrificans accompanying the transition from aerobic to anaerobic growth in continuous chemostat culture. J. Gen. Microbiol. 107:131-137. Kakutani, T., H. Watanabe, K. Arima, and T. Beppu. 1981. Purification and properties of a copper-containing nitrite reductase from a denitrifying bacterium, Alcaligenes faecalis strain S6. J. Biochem. 89:453-461. Kakutani, T., H. Watanabe, K. Arima, and T. Beppu. 1981. A blue protein as an inactivating factor for nitrite reductase in Alcaligenes faecalis strain S6. J. Biochem. 89:463-472. Kashem, M.A., H.B. Dunford, M-Y. Liu, W.J. Payne, and J. LeGall. 1987. Kinetic studies of the copper nitrite reductase from Achromobacter cycloclastes and its interaction with a blue copper protein. Biochem. Biophys. Res. Comm. 145:563-568. Kaspar, H.F. 1982. Nitrite reduction to nitrous oxide by Propionibacteria: Detoxification mechanism. Arch. Microbiol. 133:126-130. Kaspar, H.F., and J.M. Tiedje. 1980. Response of electron capture detector to hydrogen, oxygen, nitrogen, carbon dioxide, nitric oxide, and nitrous oxide. J. Chrom. 193:142-147. Knowles, R. 1982. Denitrification. Microbiological Review; 46:43- 70. Kohring, G.-W., F. Mayer, and H. Mayer. 1985. Immunoelectron microscopic localization of the restriction endonuclease Eta RI in Escherichia coli BS 5. Eur. J. Cell. Biol. 37:1-6. Koike, 1., and A. Hattori. l975a. Growth yield of a denitrifying bacterium, Pseudomonas denitrificans, under aerobic and denitrifying conditions. J. Gen. Microbiol. 88:1-10. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 115 Koike, I., and A. Hattori. l975b. Energy yield of denitrification: An estimate from growth yield in continuous cultures of Pseudomonas denitrificans under nitrate-, nitrite, and nitrous oxideLIimited conditions. J. Gen. Microbiol. 88:11-19. Karner, H., K. Frunzke, K. Dohler, and W.G. Zumft. 1987. Immunochemical patterns of distribution of nitrous oxide reductase and nitrite reductase (cytochrome gg ) among denitrifying pseudomonads. Arch. Microbiol. 148:2 -24. KOrner, H., and W.G. Zumft. 1989. Expression of denitrification enzymes in response to the dissolved oxygen level and respiratory substrate in continuous culture of Pseudomonas stutzeri. Appl. Environ. Microbiol. 55:1670-1676. Kovacs, K.L., L.C. Seefeldt, G. Tigyi, C.M. Doyle, L.E. Mortenson, and D.J. Arp. 1989. Immunological relationship among hydrogenases. J. Bact. 171:430-435. Kramer, D.R., M. Veenhuis, G. Fauque, H.D. Peck Jr., J. LeGall, J. Lampreia, J.J.G. Moura, and T.A. Hansen. 1988. Immunocytochemical localization of APS reductase and bisulfite reductase in three Desulfovibrio species. Arch. Microbiol. 150:296-301. Kristjansson, J.K., B. Walter, and T.C. Hollocher. 1978. Respiration-dependent proton translocation and the transport of nitrate and nitrite in Paracoccus denitrificans and other denitrifying bacteria. Biochemistry. 17:5014-5019. Krul, J.M. 1976. Dissimilatory nitrate and nitrite reduction under aerobic conditions by an aerobically and anaerobically grown Alcaligenes sp. and by activated sludge. J. Appl. Bact. 40:245- 260. Kuenen, J., and L.A. Robertson. 1987. Ecology of nitrification and denitrification. p.16l-218. In J.A. Cole and S. Ferguson (eds.) The Nitrogen and Sulfur Cycles. Cambridge University Press. Laemmli. U.K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 227:680-685. LaLande, R., and R. Knowles. 1987. Cytoplasmic content of Azospirillum brasilense sp7 grown under aerobic and denitrifying conditions. Can. J. Microbiol. 33:151-156. LeGall, J., W.J. Payne, T.V. Morgan, and D. DerVartanian. 1979. On the purification of nitrite reductase from Thiobacillus denitrificans and its reaction with nitrite under reducing conditions. Biochem. Biophys. Res. Comm. 87:355-362. Liu, M-C., W.J. Payne, H.D. Peck, and J. LeGall. 1983. Comparison of cytochromes from anaerobically and aerobically grown cells of Pseudomonas perfectomarinus. J. Bact. 154:278-286. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 116 Liu, M-Y., M-C. Liu, W.J. Payne, and J. LeGall. 1986. Properties and electron transfer specificity of copper proteins from the denitrifier Achromobacter cycloclastes. J. Bact. 166:604-608. Lowry, 0., N.J. Rosebrough, A.L. Farr, and R.J. Randall. 1951. Protein measurement with the folin phenol reagent. J. Biol. Chem. 193:265-285. Mancinelli, R.L., S. Cronin, and L.I. Hochstein. 1986. The purification and properties of a cd-cytochrome nitrite reductase from Paracoccus halodenitrificans. Arch. Microbiol. 145:202-208. Maniatis, T., E.F. Fritsch, and J. Sambrook. 1982. Molecular cloning: A laboratory manual. Cold Spring Harbor Laboratory. Cold Spring Harbor, New York. Masuko,M., H. Iwasaki, T. Sakurai, S. Suzuki, and A. Nakahara. 1984. Characterization of nitrite reductase from a denitrifier, Alcaligenes sp. NCIB 11015. A novel copper protein. J. Biochem. 96:447-454. Matsubara, T., and H. Iwasaki. 1971. Enzymatic steps of dissimilatory nitrite reduction in Alcaligenes faecalis. J. Biochem. 69:859-868. Maurer, P., D. Lessmann, and G. Kurz. 1982. D-glucose-6-phosphate dehydrogenases from Pseudomonas fluorescens. Methods Enzymol. 89:261-270. Meiberg, J.B.M., P.M. Bruinberg, and W. Harder. 1980. Effect of dissolved oxygen tension on the metabolism of methylated amines in Hyphamicrobium X in the absence and presence of nitrate; Evidence for 'aerobic' denitrification. J. Gen. Microbiol. 120:453-463. Michalski, W.P., and D.J.D. Nicholas. 1984. The adaptation of Rhodopseudamanas sphaeroides f. sp. denitrificans for growth under denitrifying conditions. J. Gen. Microbiol. 130:155-165. Michalski, W.P., and D.J.D. Nicholas. 1985. Molecular characterization of copper-containing nitrite reductase from Rhodopseudamanas sphaeroides forms sp. denitrificans. Biochem. Biophys. Acta. 828:130-137. Michalski, W.P., and D.J.D. Nicholas. 1988a. Identification of two new denitrifying strains of Rhodobacter sphaeroides. FEMS Microbiol. Lett. 52:239-244. Michalski, W.P., and D.J.D. Nicholas. 1988b. Immunological patterns of distribution of bacterial denitrifying enzymes. Phytochemisrty. 27:2451-2456. Miller, D.J., and P.M. Wood. 1983. The soluble cytochrome oxidase of Nitrosamanas europaea. J. Gen. Microbiol. 129:1645-1650. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. .117 Newton, N. 1969. The two-haem nitrite reductase of Microccus denitrificans. Biochem. Biophys. Acta. 185:316-331. Obaton, M., N. Amarger, and M. Alexander. 1968. Heterotrophic nitrification by Pseudomonas aeruginosa. Arch Microbiol. 63:122- 132. Ohkubo, 8., H. Iwasaki, H. Hori, and S. Osawa. 1986. Evolutionary relationship of denitrifying bacteria as deduced from 58 rRNA sequences. J. Biochem. 100:1261-1267. Ossmer, R., T. Mund, P.L. Hartzell, U. Konheiser, G.W. Kohring, A. Klein, R.S. Welfe, G. Gottschalk, and F. Mayer. 1986. Immunocytochemical localization of component C of the methyl reductase system in Methanococcus voltae and Methanobacterium thermoautotrophicum. Proc. Nat. Acad. Sci. 83:5789-5792. Palleroni, N.J., R. Kunisawa, R. Contopoulou, and M. Doudoroff. 1973. Nucleic acid homologies in the genus Pseudomonas. Int. J. Syst. Bact. 23:333-339. Parkin, T.B. 1987. Soil microsites as a source of denitrification variability. Soil Sci. Soc. Am. J. 51:1194-1199. Parsonage, D., A.J. Greenfield, and S.J. Ferguson. The high affinity of Paracoccus denitrificans cells for nitrate as an electron acceptor. Analysis of possible mechanisms of nitrate and nitrite movement across the plasma membrane and the basis for inhibition by added nitrite of oxidase activity in permeabilized cellls. Biochem. Biophys. Acta. 807:81-95. Payne, W.J. 1973. Reduction of nitrogenous oxides by microorgansims. Bact. Rev. 37:409-452. Payne, W.J. 1981. Denitrification. Wiley, New York. Payne, W.J., P.S. Riley, and C.D. Cox Jr. 1971. Separate nitrite, nitric oxide, and nitrous oxide reducing fractions from Pseudomonas perfectomarinus. J. Bact. 106:356-361. Path, M., and D.D. Focht. 1985. 15H kinetic analysis of N 0 production by Nitrosamanas europaea: An examination of nitrifier denitrification. Appl. Environ. Microbiol. 49:1134-1141. Radcliffe, D.C., and D.J.D. Nicholas. 1968. Some properties of a nitrite reductase from Pseudomonas denitrificans. Biochem. Biophys. Acta. 153:545-554. Rice, G.W., P.E. Sierzega, J.M. Tiedje, and L.W. Jacobs. 1988. Stimulated denitrification in the microenvironment of a biodegradable organic waste injected into soil. Soil Sci. Soc. Am. J. 52:102-108. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 118 Ritchie, G.A.F., and D.J.D. Nicholas. 1974. The partial characterization of purified nitrite reductase and hydroxylamine oxidase from Nitrosamonas europaea. Biochem. J. 138:471-480. Robertson, G.P., and J.M. Tiedje. 1987. Nitrous oxide sources in aerobic soils: Nitrification, denitrification, and other biological processes. Soil Biol. Biochem. 130:155-165. Robertson, L.A., and J.G. Kuenen. 1983. Thiasphaera pantatropha gen. nov. sp. nov., a facultatively anaerobic, facultatively autotrophic sulfur bacterium. J. Gen. Microbiol. 129:2847-2855. Robertson, LMA., and J.G. Kuenen. 1984a. Aerobic denitrification: a controversy revived. Arch. Microbiol. 139:351-354. Robertson, L.A., and J.G. Kuenen. 1984b. Aerobic denitrification - old wine in new bottles? Antonie van Leeuwenhoek. 50:525-544. Robertson, L.A., and J.G. Kuenen. 1988. Heterotrophic nitrification in Thiosphaera pantotropha: Oxygen uptake and enzyme studies. J. Gen. Microbiol. 134:857-863. Romermann, D., and B. Friedrich. 1985. Denitrification by Alcaligenes eutrophus is plasmid dependent. J. Bact. 162:852-854. Romermann, D., J. Warrelmann, R.A. Bender, and B. Friedrich. 1989. An rpoN-like gene from Alcaligenes eutrophus and Pseudomonas facilis controls expression of diverse metabolic pathways, including hydrogen oxidation. J. Bact. 171:1093-1099. Saraste, M., and T. Kuronen, 1978. Interaction of Pseudomonas cytochrome ggl with the cytoplasmic membrane. Biochem. Biophys. Acta. 513:117-131. Sawada, E., and T. Satoh. 1980. Periplasmic location of dissimilatory nitrate and nitrite reductasee in a denitrifying phototrophic bacterium Rhadopseudomanas sphaeraides farma sp. denitrificans. Plant Cell Physiol. 21:205-210. Sawada, E., T. Satoh, and H. Kitamura. 1978. Purification and properties of a dissimilatory nitrite reductase of a denitrifying phototrophic bacterium. Plant Cell Physiol. 19: 1339-1351. Sawhney, V., and D.J.D. Nicholas. 1978. Sulphide-linked nitrite reductase from Thiobacillus denitrificans with cytochrome oxidase activity: Purification and properties. J. Gen. Microbial. 106:119- 128. Schneider, B., A. Nies, and B. Friedrich. 1988. Transfer and expression of lithoautotraphy and denitrification in a host lacking these metabolic activities. Appl. Environ. Microbial. 54:3173-3176. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 119 Scott, R.H., G.T. Sperl, and J.A. DeMoss. 1979. In vitro incorporation of molybdate into demalybdoproteins in Escherichia coli. J. Bact. 137:719-726. Sexstone, A.J., N.P. Revsbech, T.B. Parkin, and J.M. Tiedje. 1985. Direct measurement of oxygen profiles and denitrification rates in soil aggregates. Soil Sci. Soc. Am. J. 49:645-651. Shapleigh, J.P., and W.J. Payne. 1985. Differentiation of g,g1 cytochrome and copper nitrite reductase production in denitrifiers. FEMS Miro. Lett. 26:275-279. Shioi, Y., M. Doi, H. Arata, and K. Takamiya. 1988. A denitrifying activity in an aerobic photosynthetic bacterium, Erythrobacter sp. strain 0CH.114. Plant Cell Physiol. 29:861-865. Smith, M.S., and L.L. Parsons. 1985. Persistance of denitrifying enzyme activity in dried soils. Appl. Environ. Microbiol. 49:316- 320. Smith, M.S., and.J.M. Tiedje. 1979. Phases of denitrification following oxygen depletion in soil. Soil Biol. Biochem. 11:261- 267. Snyder, W.S., D.A. Bazylinski, and T.C. Hollocher. 1987. Loss of N20 reductase activity as an explanation for poor growth of Pseudomonas aeruginosa on N20. Appl. Environ. Microbiol. 53:2045- 2049. Stouthamer, A.B. 1976. Biochemistry and genetics of nitrate reduction in bacteria. Adv. Microb. Physiol. 14:315-375. Stauthamer, A.B. 1988. Dissimilatory reduction of oxidized nitrogen compounds. p. 245-303. In A.J.B. Zehnder (ed.) Biology of Anaerobic Microorganisms. Wiley, New York Tafler, S.“ , P. Serlow, and L. Levine. 1973. Serological relatedness of bacterial deoxyribonucleic acid polymerases. J. Bact. 113:18-23. Tiedje, J.M. 1982. Denitrification. p.1011-1026. In A.L. Page (ed.) Methods of Soil Analysis: Part 2 - Chemical and Microbiological Properties. ASA Inc., Madison. Tiedje, J.M. 1988. Ecology of denitrification and dissimilatory nitrate reduction to ammonia. p.179-244. In A.J.B. Zehnder (ed.) Biology of Anaerobic Microorganisms. Wiley, New York. Towbin, H., T. Staehelin, and J. Gordon. 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Nat. Acad. Sci. 76:4350-4354. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 12C) Tronick, S.R., J.E. Ciardi, and E.R. Stadtman. 1973. Comparative biochemical and immunological studies of bacterial glutamine synthetases. J. Bact. 115:858-868. Turley, S., E.T. Adman, L.C. Sieker, M.-Y. Liu, W.J. Payne, and J. LeGall. 1988. Crystallization of nitrite reductase from Achromobacter cycloclastes. J. Mol. Biol. 200:417-419. van Hartingsveldt, J., M.G. Marinus, and A.B. Stouthamer. 1971. Mutants of Pseudomonas aeruginosa blocked in nitrate or nitrite dissimilation. Genetics. 67:469-482. Verstraete, W., and M. Alexander. 1972a. Heterotrophic nitrification by Arthrobacter sp. J. Bact. 110:955-961. Verstraete, W., and.M. Alexander, 1972b. Mechanism of nitrification by Arthrobacter sp. J. Bact. 110:962-967. Viebrock, A., and.W.G. Zumft. 1987. Physical mapping of transposan Tn5 insertions defines a gene cluster functional in nitrous oxide respiration by Pseudomonas stutzeri. J. Bact. 169:4577-4580. Viebrock, A., and W.G. Zumft. 1988. Molecular cloning, heterologous expression, and primary structure of the structural gene for the copper enzyme nitrous oxide reductase from denitrifying Pseudomonas stutzeri. J. Bact. 170:4658-4668. Wharton, D.C. and S.T. Weintraub. 1980. Identification of nitric oxide and nitrous oxide as products of nitrite reduction by Pseudomonas cytochrome oxidase (nitrite reductase). Biochem. Biophys. Res. Comm. 97:236-242. Wilson, A.C., S.S. Carlson, and T.J. White. 1977. Biochemical evolution. Ann Rev. Biochem. 46:573-639. Woese, C.R. 1987. Bacterial evolution. Microbiol. Rev. 51:221-271. Wood, P.M. 1978. Periplasmic location of the terminal reductase in nitrite respiration. FEBS Lett. 92:214-218. Yamamato, K., T. Uozumi, and T. Beppu. 1987. The blue copper protein gene of Alcaligenes faecalis 86 directs secretion of blue copper protein from Escherichia coli cells. J. Bact. 169:5648- 5652. Yamanaka, T., and K. Okunuki. 1963. Crystalline Pseudomonas cytochrome oxidase: I. Enzymic properties with special reference to the biological specificity. Biochem. Biophys. Acta. 67:379-393. Yoshida, T., and M. Alexander. 1970. Nitrous oxide formation by Nitrosomonas europaea and heterotrophic microorganisms. Soil Sci. Soc. Am. Proc. 34:880-882. 143. 144. 145. 146. 147. 148. 149. 150. 121. Yoshinari, T., and R. Knowles. 1976. Acetylene inhibition of nitrous oxide reduction by denitrifying bacterium. Biochem. Biophys. Res. Comm. 69:705-710. Zablotowicz, R.M. and D.D. Focht. 1981. Physiological characteristics of cowpea rhizobia: evaluation of symbiotic efficiency in Vigna unguiculata. Appl. Env. Microbial. 41:679-685. Zafiriou, 0.C., Q.S. Manley, and G. Snyder. 1989. Nitric oxide and nitrous oxide reduction by Pseudomonas perfectomarina. J. Biol. Chem. 264:5694-5699. Zumft, W.G., K. Dohler, and H. Korner. 1985. Isolation and characterization of transposan Tn5-induced mutants of Pseudomonas perfectamarina defective in nitrous oxide respiration. J. Bact. 163:918-924. Zumft, W.G., K. Dohler, H. Korner, S. Lochell, A. Viebrock, and K. Frunzke. 1988. Defects in cytochrome ggl-dependent nitrite respiration of transposan T25-induced mutants from Pseudomonas stutzeri. Arch. Microbial. 149:492-498. Zumft, W.G., D.J. Gotzmann, and P.M.H. Kroneck. 1987. Type 1, blue copper proteins constitute a respiratory nitrite-reducing system in Pseudomonas aureafaciens. Eur. J. Biochem. 168:301-307. Zumft, W.G., and J.M. Vega. 1979. Reduction of nitrite to nitrous oxide by a cytoplasmic membrane fraction from the marine denitrifier Pseudomonas perfectomarinus. Biochem. Biophys. Acta. 548:484-499. Zumft, W.G., A. Viebrock, and H. Korner. 1987. Biochemical and physiological aspects of denitrification. p. 245-279. In J.A. Cole and S. Ferguson (eds.) The Nitrogen and Sulfur Cycles. Cambridge University Press. nICHIGnN STATE UNIV. LIBRQRIES “HIWIIIHIIIWIIWHIIWIIWIIWWI 31293006299386