JB Accepted Manuscript Posted Online 22 February 2016 J. Bacteriol. doi:10.1128/JB.00055-16 Copyright © 2016, American Society for Microbiology. All Rights Reserved.

1 2

The Protein Network of the Pseudomonas aeruginosa Denitrification

3

Apparatus

4 5

José Manuel Borrero-de Acuñaa*, Manfred Rohdeb, Josef Wissingc, Lothar Jänschc,

6

Max Schoberta, Gabriella Molinarib, Kenneth N. Timmisa†, Martina Jahna and Dieter

7

Jahna#

8 9 10

Institute of Microbiology, Technische Universität Braunschweig, Braunschweig, Germany a; Central

11

Facility for Microscopy, Helmholtz Centre for Infection Research (HZI), Braunschweig, Germanyb;

12

Department Cellular Proteome Research, HZIc and Environmental Microbiology Laboratory, HZId,

13

Braunschweig, Germany

14 15

Running Head: Denitrification Protein Network

16 17

#

Address correspondence to Dieter Jahn, [email protected].

18

†

Current address: Department of Molecular Bacteriology, HZI, Braunschweig, Germany.

19 20 21 22

*

Present address: Universidad Andrés Bello, Biosystems Engineering Laboratory, Center for Bioinformatics and Integrative Biology, Facultad de Ciencias Biológicas, Santiago, Chile

23 24 25

1

26

ABSTRACT

27

Oxidative phosphorylation using multiple component, membrane-associated protein

28

complexes is the most effective way for a cell to generate energy. Here, we

29

systematically

30

denitrification apparatus of the pathogenic bacterium Pseudomonas aeruginosa.

31

During denitrification, nitrate (Nar), nitrite (Nir), nitric-oxide (Nor) and nitrous-oxide

32

(Nos) reductases catalyze the reaction cascade of NO3-  NO2-  NO  N2O  N2.

33

Genetic experiments suggested that the nitric-oxide reductase NorBC and the

34

regulatory protein NosR are the nucleus of the denitrification protein network. We

35

utilized membrane interactomics in combination with electron microscopy co-

36

localization studies to elucidate the corresponding protein-protein interactions. The

37

integral membrane proteins NorC, NorB and NosR form the core assembly platform

38

that binds the nitrate reductase NarGHI and the periplasmic nitrite reductase NirS via

39

its maturation factor NirF. The periplasmic nitrous-oxide reductase, NosZ, is linked

40

via NosR. The nitrate transporter, NarK2, the nitrate regulatory system, NarXL,

41

various nitrite reductase maturation proteins, NirEJMNQ, and the Nos assembly

42

lipoproteins, NosFL, were also found to be attached. A number of proteins associated

43

with energy generation, including electron donating dehydrogenases, the complete

44

ATP synthase, almost all enzymes of the TCA cycle, and the SEC system of protein

45

transport, among many other proteins, were found to interact with the denitrification

46

proteins. This deduced nitrate respirasome is presumably only one part of an

47

extensive cytoplasmic membrane-anchored protein network connecting cytoplasmic,

48

inner membrane and periplasmic proteins, to mediate key activities occurring at the

49

barrier/interface between the cytoplasm and the external environment.

investigated

the

multiple

protein-protein

interactions

of

the

50 2

51

IMPORTANCE

52

Processes of cellular energy generation are catalyzed by large multi-protein enzyme

53

complexes. The molecular basis for the interaction of these complexes is poorly

54

understood. We employed membrane interactomics and electron microscopy to

55

determine the involved protein-protein interactions. The well-investigated enzyme

56

complexes of denitrification of the pathogenic bacterium Pseudomonas aeruginosa

57

served as a model. Denitrification is one essential step of the universal N-cycle and

58

provides the bacterium with an effective alternative to oxygen respiration. This

59

process allows the bacterium to form biofilms, which create low oxygen habitats, and

60

which are a key in infection mechanism. Our results provide new insights into the

61

molecular basis of respiration, as well as opening a new window into infection

62

strategies of this pathogen.

63

INTRODUCTION

64

The most effective mode of biological energy generation is oxygen respiration. During

65

this process electrons are transported from donor molecules like NADH via

66

membrane localized large multi-protein complexes along a redox cascade to the

67

electron acceptor oxygen. The generated free energy is recovered via the formation

68

of a H+/Na+ gradient at the membrane, ultimately driving ATP synthesis by ATP

69

synthases (1). Under conditions of oxygen depletion many microorganisms are

70

capable of replacing oxygen by an alternative electron acceptor including nitrate,

71

sulphate, fumarate or dimethyl sulfoxide (DMSO) (2). Furthermore, the electron

72

acceptors are combined with a whole variety of electron donors by these organisms.

73

In this context, every electron donor and acceptor requires its own complex

74

enzymatic system, either soluble or membrane-associated and often accompanied by 3

75

additional intermediate transfer complexes, like the bc1 complex. The structure and

76

biochemistry of these different enzyme complexes and enzymatic systems are well

77

understood (1). In contrast, much less is known about the dynamic interactions of

78

these protein complexes in the respiratory chains. For humans (3), bovine (4), mouse

79

(5), spinach (6), potato (7) and yeast (8, 9) the existence of supra-complexes has

80

been proposed and proven. Similarily, membrane-localized multienzyme complexes

81

were investigated in great detail for bacterial and plant photosynthesis (10-12). For

82

prokaryotes, only a few investigations on supra-molecular formation are available for

83

the sulfide oxidase-oxygen reductase supercomplex of the bacterium Aquifex

84

aeolicus (13) and the respirasome of Paracoccus denitrificans (14). However, none of

85

the investigations reached the level of protein-protein interaction elucidation.

86

Here, we studied the supermolecular organization of the denitrification machinery of

87

P. aeruginosa. In the environment and during infection, P. aeruginosa forms biofilms,

88

which during anaerobic growth are dependent on anaerobic respiration via

89

denitrification (15). Four different multi-protein complexes receive electrons from

90

various electron donors (Fig. 1) via ubiquinone and intermediate complexes to

91

catalyze the following final reduction steps: NO3-  NO2-  NO  N2O  N2. During

92

the first step of denitrification in various proteobacteria, respiratory nitrate reductase

93

(NarGHI) catalyzes the following reaction: NO3- + 2e- + 2H+  NO2- + H2O. Multiple

94

biochemical and structural data revealed that these quinol-nitrate oxidoreductases

95

receive electrons from ubiquinol via the two heme groups of the membrane spanning

96

small subunit NarI. Electrons get transferred through a linear arrangement of Fe-S

97

clusters of the NarI attached subunit NarH to the active site of the molybdenum

98

cofactor (MoCo) containing subunit NarG. NarH and NarG are directed into the

99

cytoplasm (16-18). The protein NarJ and cardiolipins are involved in the complex 4

100

assembly of the various protein subunits with their corresponding cofactors (19, 20).

101

P. aeruginosa additionally contains the periplasmic nitrate reductase NapAB.

102

Electrons are supplied by the membrane anchored tetraheme cytochrome c NapC.

103

Interestingly, NapABC can compensate for NarGHI deficiency (21). The periplasmic

104

nitrite reductase NirS catalyzes the next reaction as follows: NO2- + 2H+ + e-  NO +

105

H2O. The P. aeruginosa enzyme is a cytochrome cd1 carrying an electron receiving

106

heme c and a catalytic site with heme d1 (3, 8-dioxo-17-acrylate-porphyrindione; 22-

107

24). The two c-type cytochromes NirM and NirC act as electron donors (25). NirN and

108

NirF form a stable complex with the nitrite reductase NirS during enzyme maturation

109

(26). NirF is involved in heme d1 insertion (27). The final step of heme d1

110

biosynthesis, mainly localized in the periplasm, involve the uroporphyrin-III c-

111

methyltransferase NirE (28), siroheme decarboxylase NirDL/NirGH (29) and a novel

112

electron-bifurcating dehydrogenase NirN (30). Nitrate from the environment gets

113

transported into the cytoplasm by the nitrate/nitrite antiporter NarK2, converted by the

114

activity of NarGHI into nitrite, which then returns via NarK2 into the periplasm to

115

serve as substrate for NirS (31). Then, integral cytoplasmic membrane protein nitric

116

oxide reductase NorBC catalyzes the following reaction: 2NO + 2H+ + 2e-  N2O +

117

H2O. The heterodimeric P. aeruginosa enzyme utilizes heme c and heme b for

118

electron delivery to a heme b and non-heme iron in the active site that coordinates

119

the two NO substrates (32, 33). Interestingly, the small subunit C covers most of the

120

surface of the enzyme towards the periplasm (32, 33). The ATP-binding protein NirQ

121

seems to be involved in NorCB maturation (34). For the final step, the periplasmic

122

enzyme nitrous oxide reductase NosZ catalyzes the following reaction: N2O + 2e- +

123

2H+  N2 + H2O. Catalysis of the head-to-tail homodimeric enzyme requires two

124

copper centers (35, 36). The periplasmic domain of the transmembrane iron-sulfur 5

125

flavoprotein NosR is essential for NosZ function (37). The periplasmic lipoprotein

126

NosL anchored to the outer membrane is a copper binding protein involved in NosZ

127

maturation (38, 39). However, the denitrification further relies on and multiple other

128

components of the electron transport chain (Fig. 1, 40, 41, 42, 43).

129

The

130

dehydrogenases (41) and multiple other components of the electron transport chain.

131

Recently, we described the existence of a stable NirS-DnaK-FliC complex in the

132

periplasm of P. aeruginosa involved in flagellum assembly, using a combined

133

membrane and periplasmic proteomics and electron microscopy approach (44).

134

Here, we employed this experimental approach (45-47) to determine the nature of the

135

protein-protein interactions of the denitrification machinery.

136

MATERIALS AND METHODS

137

Construction of the Bait Expression Vectors. Escherichia coli DH10b (Invitrogen,

138

Germany) strain served as cloning host. E. coli was grown in Luria-Bertani medium

139

(LB), at 37ºC and shaking at 200 rpm. Ampicillin and carbenicillin were used at 100

140

µg/mL. P. aeruginosa PAO1 was used as wildtype strain (48) and obtained from the

141

DSMZ, Braunschweig, Germany. The mini-Tn5 luxCDABE transposon mutants of P.

142

aeruginosa nosR, norC, norB and nosZ, respectively, served as hosts for the

143

physiological characterization of the denitrification process (49). Mutants were

144

obtained from an in house mutant collection. P. aeruginosa strains were grown at

145

37ºC in LB medium and 200 rpm shaking. Carbenicillin was supplemented at 250

146

µg/ml. Anaerobic cultures were supplemented with 50 mM KNO3 as described before

147

(50). The E. coli vector pJET1.2 (Thermo Scientific, Germany) was employed as

148

initial cloning vector. The E. coli and P. aeruginosa shuttle vector pUCP20t (51) was

denitrification

process

further

relies

on

electron-donating

primary

6

149

modified by the removal of the BsaI restriction site at the position 3259 of the plasmid

150

via site directed mutagenesis (see Table 1). The resulting plasmid was called pAS40.

151

Due to the multiplicity of constructed plasmids and E. coli as well as P.aeruginosa

152

strains in this study, all used primers for amplification, plasmids and strains are listed

153

in Tables 1 and 2.

154

Nitrite and nitrate consumption rate determination. In order to determine the

155

nitrite reduction activity of the diverse tested P. aeruginosa strains, a nitrate/nitrite

156

colorimetric test was used according to the manufacturer’s guidelines (Roche,

157

Germany).

158

P. aeruginosa interactomic studies using NosR, NorC and NorB as bait

159

proteins.

160

tagged proteins in an appropriate, defined genetic background. For this purpose a

161

plasmid-toolbox based on the Golden-Gate cloning strategy was developed (52). The

162

respective genes were fused at their 3´ end to the corresponding tags in order to

163

avoid interference with signal peptide recognition and protein insertion into the

164

membrane. To mimic natural protein production, the native promoters were used.

165

The nosR, norC and norCB genes including a region between 400 - 500 bp upstream

166

of their start codon were PCR amplified using the primers nosRFw/nosRRv,

167

norCFw/norCRv and norCBFw/norCBRv, respectively. The nosZ gene including the

168

promoter sequence was PCR amplified from the nosRZDFYL operon using the

169

nosZFw/nosZRv and nosPRFw/nosPRRv primers (Table 2). The respective PCR

170

products were cloned into BsaI restriction sites of the carrier vector pJET1.2 (Table

171

1). The second acceptor plasmid encoded the 6 x His tag. The plasmids pJET1.2-

172

nosR, pJET1.2-norC and pJET1.2-norB were united with the pAS40 and subjected to

173

50 cycles of amplification

For our interaction studies, NorB, NorC and NosR were produced as

7

174

via PCR using the individually chosen primers: nosRHis6xFw/nosRHis6xRv;

175

norCHis6xFw/norCHis6xRv;

176

tagIIFw/nosZStrep-tagIIRv, followed by the 5 min restriction with BsaI and 5 min

177

ligation using T4 DNA ligase. Depending on the combination of gene constructs the

178

following final products were generated: pAS40-nosR-His6x, pAS40-norCB-His6x,

179

pAS40-norC-His6x and pAS40-nosZ-Strep-tag II, respectively (Table 1). The

180

corresponding P. aeruginosa transposon mutant strains nosR, norC, norB and nosZ

181

were transformed with the complementing plasmid, respectively.

182

The

183

PAO1norB-pnorB-His6x and PAO1nosZ-pnosZ-Strep-tag II. The expression vector

184

pAS40 was introduced into P. aeruginosa PAO1 wild type and employed as negative

185

control for the interactomic studies (Table 1). P. aeruginosa strains were grown to the

186

exponential growth phase (OD578 = 0.8). The protein complexes were cross-linked in

187

vivo as previously described (44). Formaldehyde, used as the cross-linking agent,

188

was injected into the anaerobic denitrifying culture in the late stationary phase to a

189

final concentration of 0.125% (v/v). The cultures were further incubated by shaking at

190

150 rpm to facilitate cross-linking penetration at 37ºC for 20 additional min. Free

191

aldehyde groups were quenched by adding 135 mM glycine, and the cultures were

192

further incubated at 37ºC for 5 min and shaking at 150 rpm. Then, the cells were

193

harvested at 3,000 x g, for 20 min at 4ºC. The resulting cell pellet was washed twice

194

with phosphate buffered saline (PBS; 137 mM NaCl, 2.7 mM KCl, 4.3 mM Na2HPO4

195

and 1.4 mM KH2PO4, pH 7.3). Interested in the interactome of the outer membrane,

196

the periplasm, the inner membrane and the cytoplasm, the corresponding fractions

197

were separated and isolated at 4°C. For this purpose, washed cells were

198

resuspended in 2/3 times pellet volumes of lysis buffer (50 mM NaH2PO4, 300 mM

resulting

strains

were

norCBHis6xFw/norCBHis6xRv

PAO1nosR-pnosR-His6x,

and

nosZStrep-

PAO1norC-pnorC-His6x,

8

199

NaCl, pH 8.0) with 1 tablet protease inhibitor cocktail (Roche, Germany) per 10 ml

200

buffer. Cells were lysed by using a French press at 1100 psi and 4ºC. Cell debris was

201

removed by centrifugation at 20,000 x g for 30 min at 4ºC. In order to isolate the

202

membrane-bound proteins, lysates were ultra-centrifuged at 100,000 x g for 1 h at

203

4ºC. Subsequently, 200 µg of the membrane proteins of the pellet fraction were

204

dissolved for 1 h at 4ºC in 500 µl of 20 mM potassium buffer (pH 8.0) supplemented

205

with 2% Triton X-100. Cell debris and insolubilized membranes were removed by

206

centrifugation at 20,000 x g for 30 min at 4ºC. Solubilized membrane proteins present

207

in the supernatant were mixed for 1 h at 4ºC with Ni-NTA agarose previously

208

equilibrated with NPI-10 buffer (50 mM NaH2PO4, 300 mM NaCl, 10 mM imidazole;

209

Protino, Macherey-Nagel GmbH, Düren, Germany). The column was washed with

210

imidazole at increasing concentrations (20, 30, 40 and 50 mM in NPI-20, NPI-30,

211

NPI-40 and NPI-50 buffers consecutively). Recombinantly tagged NosR, NorC and

212

NorB proteins, along with their potential interaction partners, were eluted from the

213

column by addition of 300 mM imidazole in NPI-300 buffer. Protein fractions of

214

interest were analyzed by SDS-PAGE. Protein bands were stained with Coomassie

215

Brilliant Blue according to the manufacturer’s instructions. Primary antibodies against

216

His6x-tags were reacted with glutathione-S-transferase (GST)-alkaline phosphatase

217

conjugated secondary antibodies. Stained SDS-PAGE gels and Western blots were

218

visualized and quantified on a GS-800 calibrated densitometer (Biorad).

219

In order to assess whether the detected bound protein interaction partners

220

constituted specific interaction partners or rather were artifacts derived from potential

221

high protein abundance within the membrane fractions, P. aeruginosa PAO1

222

membranes were isolated, and the residing proteins were solubilized as described

223

above. The solubilized membrane and membrane-associated proteins were extracted 9

224

(54). Protein samples of 0.1 - 0.5 mg/ml were analyzed using LC-MS/MS.

225

Experiments were carried out in three independent biological replicates.

226

LC-MS/MS analyses of protein complexes and data interpretation. As described

227

above the PAO1 membrane proteins were solubilized and the remaining detergent

228

was removed by methanol/chloroform protein extraction. Protein samples were

229

loaded onto the LC-MS/MS. Thereby, it was feasible to evaluate interaction partners´

230

enrichment within the co-purifications in comparison to the natural abundance of

231

these proteins in the membrane fraction. The proteins that were significantly enriched

232

in the co-purification in contrast to the controls were defined as specific interactors.

233

Several cytoplasmic and periplasmic proteins were present in the cross-linked

234

samples but did not appear in the control samples indicating their overall low

235

abundance and enrichment via bait binding. The protein abundance index (PAI), as

236

previously introduced (47), was attributed to each protein of the sample mixture and

237

“area values” were deduced. Here, we used the "peak area calculation quantification"

238

method of Proteome Discoverer to determine PAI values. The method integrates the

239

MS signal intensities of each peptide eluting from the chromatographic column over

240

the time resulting in corresponding peptide peak areas. The average of peak areas

241

from the three most abundant distinct peptides defines the individual Protein

242

Abundance Index (PAI; 47, 53). Results were plotted in a box plot (Fig. 5).

243

Ni-NTA agarose eluates from experimental triplicates for the isolation of the

244

crosslinked complexes to the of NosR-His6x, NorC-His6x and NorB-His6x bait proteins,

245

respectively, were subsequently evaluated by LC-MS/MS as previously described

246

(55). Eluted proteins with their attached co-purified interaction partners were

247

identified by quantitative mass spectrometry (AP-QMS) (58). LC-MS/MS analyses 10

248

were performed on a Dionex UltiMate 3000 n-RSLC system connected to an LTQ

249

Orbitrap Velos mass spectrometer (Thermo Scientific, Germany). First eluted protein

250

complexes were digested by trypsin protease treatment with a protein:protease ratio

251

about 50:1 at 37°C overnight. Subsequently, peptides were loaded onto a C18 pre-

252

column (3 µm RP18 beads, Acclaim, 75 μm x 20 mm, Dionex, Thermo Scientific,

253

Germany), washed for 3 min at a flow rate of 6 μl/min. Peptides were separated on a

254

C18 analytical column (3-μm, Acclaim PepMap RSLC, 75 μm x 25 cm, Dionex) at a

255

flow rate of 350 μl/min via a linear 120 min gradient from 100% buffer A (A = 0.1%

256

formic acid in water) to 25% buffer B (B = 99.9% acetonitrile with 0.1% formic acid)

257

followed by a 50 min gradient from 25% buffer A to 80% buffer B. The LC system

258

was operated with the Chromeleon software (version 6.8, Dionex), which was

259

embedded in the Xcalibur software suite (version 2.1, Thermo Scientific, Germany).

260

The effluent from the column was electro-sprayed (Pico Tip Emitter Needles,

261

Thermofischer, Omnilab, Germany) into the mass spectrometer (Orbitrap Velos Pro,

262

Germany). The mass spectrometer was controlled and operated in the data-

263

dependent mode of the Xcalibur software allowing the automatic selection of a

264

maximum of 10 doubly and triply charged peptides and their subsequent

265

fragmentation. Peptide fragmentation was carried out using LTQ settings (min signal

266

2000, isolation width 4, normalized collision energy 35, default charge state 4 and

267

activation time 10 ms). MS/MS raw data files were processed via the Proteome

268

Discoverer program Version 1.4 (Thermo Scientific, Germany) on a Mascot server

269

(V. 2.3.02, Matrix Science) using a P. aeruginosa PA01 database extracted from

270

SwissProt. The following search parameters were used: enzyme, trypsin; maximum

271

missed cleavages, 1; fixed modification: MMTS (C); variable modifications: oxidation

11

272

(M), peptide tolerance, 10 ppm; MS/MS tolerance, 0.4 Da. The results were

273

evaluated and quantitated using Proteome Discoverer 1.4 (54-56).

274

Antibody generation.

275

purified proteins or synthetic peptides representing soluble loops of the membrane

276

proteins of interest. For NosZ antibodies, the protein was recombinantly produced in

277

P. aeruginosa and purified to apparent homogeneity via its Strep-tag II using affinity

278

chromatography (IBA, Germany). Protein eluates were concentrated using Vivaspin

279

15 (10 kDa cutoff) centrifugal concentrators (Sartorius AG, Germany). For the

280

generation of anti-NosR, anti-NarH, anti-NirS and anti-NorC polyclonal antibodies

281

synthetic peptides were used. The amino acid residues 50 - 61; 102 - 116 and 118 -

282

132 and 106 - 130 for NorC and 190 - 203 and 404 - 418 for NosR were selected for

283

this purpose, respectively. These amino acid residues corresponded to medium-

284

length periplasmic loops of these proteins resulting in soluble peptides. For the

285

generation of anti-NarH polyclonal antibodies a mixture of three peptides

286

corresponding to the amino acids: 1 – 276, 277 - 392 and 418 - 513 were selected.

287

Anti-NirS polyclonal antibodies were raised in rabbits with a three peptide mixture

288

corresponding to the amino acid residues between 379 - 392, 526 - 540 and 541 –

289

555, respectively.

290

Double-immunogold labeling based co-localization studies using electron

291

microscopy. P. aeruginosa PAO1 wildtype strain was grown anaerobically in Luria

292

Bertani broth (LB) supplemented with 50 mM nitrate for 8 h. The bacteria were

293

fixed with 1% formaldehyde in the growth medium overnight at 5°C. After

294

centrifugation, the cells were resuspended in Tris-EDTA (ethylenediaminetetraacetic

295

acid) (TE) buffer (10 mM Tris-Cl, 1 mM EDTA, pH 7.5) with the addition of 10 mM

All polyclonal antibodies were raised in rabbits against

12

296

glycine to quench free aldehyde groups. Afterwards, the bacteria were dehydrated

297

and subsequently embedded in Lowicryl K4M resin. After polymerization by ultra-

298

violet (UV)-light treatment, ultrathin sections were cut with a diamond knife,

299

collected onto butvar-coated nickel grids and incubated with the specific antibodies

300

(44). First, sections were incubated with the first round of purified IgG antibodies

301

overnight at 5°C. After washing with PBS, the bound antibodies were rendered

302

visible by adding protein A/G gold-nanoparticles (PA/GG 15 nm in size, 1:75 dilution

303

of the stock solution) for 30 min at room temperature. Sections were washed with

304

PBS containing 0.1% Tween 100 and further incubated with a protein A solution (0.1

305

mg/ml) for 15 min at room temperature. After an additional washing in PBS, sections

306

were incubated for 3 h at room temperature with the second round of a different

307

purified IgG. Antibody dilutions were established at 1:25 for the first IgG and 1:25 for

308

the second charge. After PBS washing, the sections were incubated with PA/GG (10

309

nm in size, 1:200 dilution of the stock solution) for 30 min at room temperature.

310

Subsequently, the sections were washed with PBS containing 0.1% Tween 100, TE-

311

buffer and distilled water. The sections were counter- stained with 4% aqueous

312

uranyl acetate for 1 min. The samples were then examined in a TEM910

313

transmission electron microscope (Carl Zeiss, Germany) at an acceleration voltage of

314

80 kV. Images were recorded digitally at calibrated magnifications with a Slow-Scan

315

CCD-Camera (ProScan, 1024x1024, Germany) with ITEM-Software (Olympus Soft

316

Imaging Solutions, Germany). Contrast and brightness were adjusted with Adobe

317

Photoshop CS4.

318 319

Results and Discussion 13

320

Mutants of P. aeruginosa nitric oxide oxidoreductase genes norBC, and of the

321

regulatory gene nosR, are defective in nitrate and nitrite reduction activity. The

322

selection of the initial bait proteins for our interactomic approach was based on the

323

assumption that the protein network of denitrification should be stably localized, most

324

likely anchored to the membrane. Only NarI, NorB, NorC and NosR were confirmed

325

as integral membrane proteins. In order to investigate the membrane proteins NorBC

326

and NosR for additional structural functions in the overall denitrifying process, the

327

nitrate and nitrite consumption of the corresponding P. aeruginosa mutants were

328

determined. Obtained values were compared to that of mutants of the cytoplasmic

329

nitrate reductase (narG) and periplasmic nitrite reductase (nirS) (Fig. 2A). The norB

330

mutant showed a reduced level of conversion of nitrate to nitrite similar to that of the

331

nitrate reductase narG mutant (Fig. 2A). Anti-NarH antibodies detected intact nitrate

332

reductase in the norB mutant (Fig. 2B) and a cell-free extract of the norB mutant

333

exhibited normal nitrate reductase activity. The norB mutant was also defective in

334

nitrite reduction, with activities similar to those of the nitrite reductase mutant nirS

335

(Fig. 2C). The norC and nosR mutants also showed reduced nitrite reduction activity.

336

The overall nitrite conversion of this mutants was somewhat delayed, indicating an

337

initial defect and a following compensatory role of other components of the system. It

338

was known from previous experiments that NorBC influences NirS formation and

339

stability (57), although the underlying mechanism was unclear. Consistent with this

340

observation, our anti-NirS antibodies failed to detect significant amounts of nitrite

341

reductase in the norB mutant (Fig. 2B), and only residual nitrite reductase activity

342

was found in the cell-free extracts of the norB mutant. Since NorBC and NosR are

343

not enzymatically involved in nitrate and nitrite reduction, and indications for their

344

involvement in narGHIJ and nirS gene expression have not been observed, an 14

345

essential alternative contribution to the overall denitrification process must be

346

deduced. Therefore, we tested NorBC and NosR for their function in denitrification

347

protein network formation.

348

Interactomic studies using P. aeruginosa NosR, NorC and NorB.

349

immunolabeling

350

microscopy, was employed to verify some of the observed protein-protein

351

interactions.

352

For the interatomics approach, protein eluates from the affinity material were

353

trichloroacetic acid (TCA)-precipitated and analyzed by sodium dodecyl sulfate

354

polyacrylamide gel electrophoresis (SDS-PAGE) (Fig. 3). The NosR-His6x eluates

355

contained a protein with a relative molecular mass of approximately 80,000 +/- 5,000

356

as a major component that was identified as NosR-His6x by immunoblot detection

357

(see Fig. 3 lane 5). The SDS-PAGE showed a number of additional protein bands

358

visible above and below NosR-His6x, which represented potential interaction partners

359

(Fig. 3, lane 1). Likewise, the NorC-His6x and NorB-His6x eluates contained prominent

360

proteins, which were identified as NorC-His6x and NorB-His6x, respectively (Fig. 3,

361

lanes 7 & 9). Again, several potential interaction partners were observed (see Fig. 3,

362

lanes 2 & 3). As expected, control reactions with PAO1 carrying the control vector did

363

not result in co-purification of additional proteins (Fig. 3, lanes 4, 6, 8, 10). Similarly,

364

production of a His-tagged cytosolic control protein (porphobilinogen synthase

365

HemB) under identical conditions did not yield the observed bound proteins (not

366

shown). The abundance distribution of potential interaction partners of NosR, NorC

367

and NorB related to denitrification are illustrated in a graph (Fig. 4) representing the

368

concentration distribution of potential interaction partners for the various bait proteins

of

ultrathin

sections,

combined

with

transmission

Double electron

15

369

with NosR, NorC, and NorB determined after co-purification. The percentage of found

370

peptides for certain proteins (Y axis) was plotted against the corresponding area

371

value (X axis). Potential interaction partners involved in denitrification are shown in

372

the left panel (A) while all the other proteins are depicted in the right site (panel B).

373

The various affinity chromatographic eluates were then analyzed by liquid

374

chromatography tandem mass spectrometry (LC-MS/MS). A protein abundance

375

index (PAI) was attributed to each protein of the sample mixture, and an “area value”

376

was calculated, as previously described (47). The percentage of the overall potential

377

interaction partners was plotted against the corresponding area value range.

378

Standard deviations were calculated from biological triplicates and integrated into a

379

box plot (Fig. 5). Most of the potential interaction partners were detected at the area

380

range of 106 - 107, and high affinity interaction partners were in the area value range

381

of 108 - 109. Abundant proteins in an average bacterial cell, such as chaperones and

382

translation elongation factors, were found within an area value of approximately 106.

383

Therefore, proteins within higher area values were considered specific and

384

interaction partners, and those below a 106 area value were not further considered in

385

this study. The candidate proteins were separated into the categories of

386

“denitrification related proteins” and “other proteins”, as shown in Tables 3 and S1.

387

This approach does not distinguish between proteins directly bound to the preys and

388

indirectly bound proteins cross-linked and attached to the whole protein complex via

389

already associated proteins.

390

The nitric-oxide reductase NorBC represents one major assembly platform

391

protein for the denitrification protein network.

392

heterodimeric nitric-oxide reductase have revealed that subunit B forms a tight

Crystal structures of the

16

393

complex with its second subunit C. In agreement, our interatomics results showed the

394

strongest interaction of NorB with NorC and vice versa, which provided a proof of

395

principle of the approach (Table 3). Overall, the interactions of the majority of the

396

detected proteins with NorC were significantly stronger than those with NorB with

397

only a few exceptions. Most likely, this difference is due the NorBC structural

398

arrangement where most of the NorC subunit is exposed towards the periplasm.

399

Alternatively, differences in the solubilities of NorB and NorC in the experiments

400

might explain these results. The differences in the affinity of NorB (bait) for NorC

401

(prey) of 10 and NorC (bait) for NorB (prey) of 435 support this assumption (Table 3).

402

Nitrate reductase NarGHI is interacting with the NorBC platform. Cytoplasmic

403

nitrate reductase subunits NarG and NarH were found to be strongly bound to NorB

404

and NorC (Table 3). No interaction was detected for the membrane subunit NarI.

405

Moreover, the nitrate/nitrite antiporter NarK2 interacts with NorC (Table 3). In nitrate

406

respiration, nitrate has to be transported into the cytoplasm and cytotoxic nitrite has

407

to be promptly returned to the periplasm to minimize damage of cytoplasmic

408

components, so integration of NarK2 into the protein network is not unexpected (31).

409

Interestingly, no detectable interaction of the various tested proteins with the

410

periplasmic nitrate reductase NapAB was observed.

411

Integration of the nitrite reductase NirS, its maturation apparatus and azurin to

412

NorBC. The strongest interaction of NorBC (bait) for an isolated prey protein was

413

detected for the periplasmic, membrane-attached lipoprotein NirF, the last enzyme in

414

heme d1 synthesis and a cofactor of nitrite reductases (26). For the final insertion of

415

heme d1 into NirS, the system uses NirN, a structural homologue of NirS without

416

nitrite reductase activity. NirS, NirF and NirN form a stable complex during nitrite 17

417

reductase maturation (26). In agreement, protein interactions of NorB and NorC with

418

NirN were detected in our study (Table 3). Furthermore, strong interactions were

419

measured between NorC and the C-type cytochrome NirM, which serves as an

420

electron donor for the nitrite reductase (Table 3), (25). Additionally, an interaction of

421

NorC with the ATP-binding protein NirQ was observed (Table 3). This is consistent

422

with the observation that deletion of the nirQ gene resulted in the simultaneous loss

423

of nitrite and NO reduction in P. aeruginosa (59). Posttranslational enhancement of

424

nitrite and nitric oxide reductase activity was observed for P. aeruginosa NirQ activity,

425

similar to the posttranslational activation of 1, 5-bisphosphate carboxylase/oxygenase

426

by the NirQ homologue (59). By analogy, NirQ might act as a nitrite and nitric-oxide

427

reductase-specific assembly factor.

428

Azurin was also found to attach to the protein network via NorC. Azurin, like

429

cytochrome c, mediates the transfer of electrons from the cytochrome bc1 complex to

430

various terminal reductases (41). It is known that cytochrome C551 NirM, the electron

431

donor to nitrite reductase NirS, forms a tight complex with azurin (60).

432

Surprisingly, the nitrite reductase NirS itself showed stronger interaction with NorB

433

than with NorC (Table 3), especially when taking the relative observed interaction

434

strength of both proteins into account. Nevertheless, our immuno-electron

435

microscopy approach confirmed the NirS-NorC interaction (Fig. 6). Thus, the nitrite

436

reductase NirS is bound to the denitrification super-complex via NorB, while the

437

electron-donor system NirM and the enzyme maturation machinery NirN – NirF –

438

NirQ, interacting with NirS, are bound via NorC.

439

Nitrous-oxide reductase NosZ and the assembly protein NosL are integrated

440

into the protein network via NosR. Originally, NosR was discovered as a 18

441

component necessary for the expression of the nitrous oxide reductase (61). NosR

442

from P. stutzeri is a polytopic membrane protein required for the formation of nitrous

443

oxide reductase NosZ. However, NosZ purified from a nosR-deficient background

444

revealed nitrous oxide reductase activity in vitro (62).

445

NosR is an iron-sulfur flavoprotein with its flavin co-factor directed towards the

446

periplasm and redox centers towards the cytoplasm. The exact biochemical function

447

of the protein is currently unknown. NosR was found to be strongly attached to the

448

NorBC platform via NorC, which, in turn, had strong interactions with the nitrate and

449

nitrite reductase systems. In agreement with these results, tagged NosR bait indeed

450

“captured” NorBC as prey (Table 3). Moreover, strong interactions of NosR to nitrate

451

reductase with NarG and NarH, as well as to NirE, NirQ, NirS and NirM of the nitrite

452

reductase, were observed, indicating a tightly bound complex at the NorBC-NosR

453

platform. Most importantly, further interactions were detected with the nitrous oxide

454

reductase NosZ (Table 3), whereas NosZ exhibited only weak interactions with the

455

NorBC proteins. Consequently, NosZ is bound to the denitrification protein network

456

via NosR. The nitrous oxide reductase NosZ is a homodimeric metalloprotein

457

containing two unusual copper centers (39). These centers require additional proteins

458

for maturation. Multiple proteins encoded by the nosDFYLtatE operon were proposed

459

to be involved in this process (64). We detected strong protein interactions of NosL

460

with NorC and NosR. Achromobacter cycloclastes NosL was identified as a copper

461

binding protein (39). The protein contains a putative lipobox, a signal peptide that

462

targets the protein for translocation through the inner membrane, followed by

463

cleavage of the signal leader sequence and amino-acylation of the N-terminal

464

cysteine to diacylglycerol. NosL is most likely shuttled through the periplasm where

465

the attached diacylglycerol anchor tethers the protein to the outer membrane. 19

466

Immunogold labeling for in vivo co-localization of NosR and the four

467

reductases of the denitrification pathway in P. aeruginosa. Antibodies specific for

468

NarH, NirS, NorC, NosZ and NosR were raised and used to determine whether these

469

proteins co-localize in P. aeruginosa. Double immunogold labeling was performed

470

with all possible combinations of the two proteins, one protein of the pair being

471

labeled with nanogold particles 15 nm in size and the other of the pair with particles

472

10 nm in size. As observed in Fig. 6, all proteins seem to co-localize to a certain

473

extent. As can be

474

deduced from quantitative data from the double labeling studies (Table 4), all of the

475

pairs involving proteins localized within the cytoplasmic membrane and/or in the

476

periplasm showed a higher percentage of co-localization, well above 1, when

477

compared with pairs of proteins separated by the cytoplasmic membrane, with an

478

average of co-localization less than 1. Thus, co-localization of the pairs NosZ-NirS,

479

NorC-NirS, NosZ-NorC and NosZ-NosR was more frequent than the pairs NarH-NirS,

480

NarH-NosZ, NarH-NorC, NosR-NirS, NarH-NosR and NorC-NosR.

481

Thus, the results of the immunogold co-localisation experiments are consistent with

482

the conclusions of the interatomics experiments. One possibility we considered was

483

that protein pairs more functionally distant in the denitrification process, i.e. regarding

484

the reduction steps that they perform, such as the nitrate reductase and the nitrous-

485

oxide reductase, might also be more physically separated from one another than

486

from their functional neighbor enzymes. Similarly, proteins involved in intermediate

487

steps, such as NirS and NorCB, might be separated from proteins mediating

488

reductions at the beginning or end of the catalytic chain by an intermediate distance.

489

With this notion in mind, the co-localization experiments suggest that the first two 20

490

enzymes of the pathway, Nar and Nir, are attached to the third, Nor, while the fourth

491

protein, Nos is mainly attached to NosR. Interestingly, the distance between the gold

492

particles of NarH and NosZ correspond approximately to the sum of the sizes (in nm)

493

of the NarH-NorC and NosZ-NorC pairs.

494

Electron donating primary dehydrogenases, ATPse, TCA cycle enzyme and

495

proteins of protein translocating SEC are part of the protein network amongst

496

others. In addition to the proteins involved in the denitrification pathway, other

497

proteins of the electron transport chain were identified to be bound to the NorBC –

498

NosR platform (Table A1). These included electron-donating NADH dehydrogenase

499

Nuo

500

dehydrogenase (PA4771), succinate dehydrogenase (PA1582-1584), D-amino acid

501

dehydratase (PA3357) and malate: quinone oxidoreductase (PA3452). Two of these

502

enzymes participate in the TCA cycle. Similarly, succinyl – CoA synthetase

503

(PA1588), isocitrate dehydrogenase (PA2624), citrate synthase (PA1580), 2-oxo-

504

glutorate dehydrogenase (PA1585) and enzymes of the pyruvate knot, including

505

various subunits of pyruvate dehydrogenases (PA5015), pyruvate kinase (PA4329),

506

phosphopyruvate hydratase (PA3035), acetyl – CoA carboxylase (PA3112, PA3639,

507

PA4848) and phosphoenolpyruvate synthase (PA1770), were also found to be

508

bound. The cofactor-synthesizing enzymes of heme, cytochrome c, iron sulfur

509

cluster, the molybdenum cofactor and ubiquinone were also part of the super-

510

complex (Table A1). Additionally, multiple subunits of the F0F1 ATP synthase

511

(PA5553-5560) were part of the mega-complex. Furthermore, various subunits of

512

chemotactic complexes CheA (PA1458), CheZ (PA1457), ChpA (PA0413) and PctC

513

(PA4307) and the aerotaxis receptor AER (PA1561) were found to be bound to

514

NorBC – NosR. Surprisingly, multiple subunits of the nucleotide-metabolizing

(PA2638-2644),

proline

dehydrogenase

PutA

(PA0782),

L-lactate

21

515

ribonucleotide reductases NrdJ (PA5496/97) and NrdAB (PA1156, PA1155) and of

516

the iron-storage protein bacterioferritin (PA3531, PA4235) were also found to be

517

strongly attached to the denitrification platform. FtsH (PA4751), FtsA (PA4408), FtsZ

518

(PA4407), ZipA (PA1528), MniD (PA3244) and MreB (PA4481), which are all

519

involved in cell division, were mainly found attached to NorC.

520

Model for a nitrate respirasome of P. aeruginosa. Currently, respirasome research

521

- the investigation of mega-complexes involved in respiration - focuses almost

522

exclusively on the mitochondrial electron transport chain of eukaryotes (3-10). Here,

523

we describe a highly structured, well-organized membrane-attached protein-protein

524

network in bacteria involved in anaerobic energy generation (Fig. 7). From the

525

obtained data one can deduce the presence of a dynamic mega-complex mediating

526

diverse membrane-associated and periplasmic functions of the cell envelope,

527

including inter alia electron donor systems, cofactor-synthesizing complexes, ATPase

528

activity, flagellum assembly and motility, cell wall and outer membrane formation and

529

integrity, cell division, chemotaxis, iron sulfur protein formation, stress responses and

530

cellular defenses to noxious agents and biotic attacks, DNA and RNA metabolism

531

and signal transduction and transport processes, many of which are central to the

532

pathogenic lifestyle of P. aeruginosa (Fig. 8). Overall, a complex network of

533

interacting proteins, functionally connecting central physiological functions of the

534

bacterial cytoplasmic cell membrane was detected. Similar complex and dynamic

535

protein structures might provide the basis for the coordination of membrane

536

biochemistry in other bacteria.

537 538

ACKNOWLEDGMENTS 22

539

This work was generously supported by ERC grant IPBSL, awarded to KT and

540

Ricardo

541

(Forschergruppe PROTRAIN) to MJ and DJ. We thank BioMed Proofreading LLC for

542

the support. We profoundly thank Simone Virus for her outstanding technical support.

Amils,

and

funding

by

the

Deutsche

Forschungsgemeinschaft

543 544 545

23

546

REFERENCES

547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596

1.

2. 3. 4. 5.

6. 7. 8. 9. 10. 11. 12. 13.

14.

15.

16.

Jahn, M., and D. Jahn (ed.). 2012. Electron transfer reactions and oxidative phosphorylation, Michal G. & Schomburg D. (eds.); In Biochemical pathways: an atlas of biochemistry and molecular biology. 2nd ed. Wiley, New York. Tielen, P., M. Schobert, E. Härtig, and D. Jahn (ed.). 2012. Anaerobic regulatory networks in bacteria. In Fillaux A.A.M.(Ed) Bacterial regulatory networks. Horizon Academic Press, Norwich UK. Moreno-Lastres, D., F. Fontanesi, I. Garcia-Consuegra, M. A. Martin, J. Arenas, A. Barrientos, and C. Ugalde. 2012. Mitochondrial complex I plays an essential role in human respirasome assembly. Cell Metab 15:324-335. Dudkina, N. V., M. Kudryashev, H. Stahlberg, and E. J. Boekema. 2011. Interaction of complexes I, III, and IV within the bovine respirasome by single particle cryoelectron tomography. Proc Natl Acad Sci U S A 108:15196-15200. Lapuente-Brun, E., R. Moreno-Loshuertos, R. Acin-Perez, A. Latorre-Pellicer, C. Colas, E. Balsa, E. Perales-Clemente, P. M. Quiros, E. Calvo, M. A. RodriguezHernandez, P. Navas, R. Cruz, A. Carracedo, C. Lopez-Otin, A. Perez-Martos, P. Fernandez-Silva, E. Fernandez-Vizarra, and J. A. Enriquez. 2013. Supercomplex assembly determines electron flux in the mitochondrial electron transport chain. Science 340:1567-1570. Krause, F., N. H. Reifschneider, D. Vocke, H. Seelert, S. Rexroth, and N. A. Dencher. 2004. "Respirasome"-like supercomplexes in green leaf mitochondria of spinach. J Biol Chem 279:48369-48375. Bultema, J. B., H. P. Braun, E. J. Boekema, and R. Kouril. 2009. Megacomplex organization of the oxidative phosphorylation system by structural analysis of respiratory supercomplexes from potato. Biochim Biophys Acta 1787:60-67. Marques, I., N. A. Dencher, A. Videira, and F. Krause. 2007. Supramolecular organization of the respiratory chain in Neurospora crassa mitochondria. Eukaryot Cell 6:2391-2405. Schagger, H., and K. Pfeiffer. 2000. Supercomplexes in the respiratory chains of yeast and mammalian mitochondria. EMBO J 19:1777-1783. Dudkina, N. V., I. M. Folea, and E. J. Boekema. 2015. Towards structural and functional characterization of photosynthetic and mitochondrial supercomplexes. Micron 72:39-51. Liu, L. N. 2015. Distribution and dynamics of electron transport complexes in cyanobacterial thylakoid membranes. Biochim Biophys Acta. Niederman, R. A. 2015. Development and dynamics of the photosynthetic apparatus in purple phototrophic bacteria. Biochim Biophys Acta. Prunetti, L., P. Infossi, M. Brugna, C. Ebel, M. T. Giudici-Orticoni, and M. Guiral. 2010. New functional sulfide oxidase-oxygen reductase supercomplex in the membrane of the hyperthermophilic bacterium Aquifex aeolicus. J Biol Chem 285:41815-41826. Stroh, A., O. Anderka, K. Pfeiffer, T. Yagi, M. Finel, B. Ludwig, and H. Schagger. 2004. Assembly of respiratory complexes I, III, and IV into NADH oxidase supercomplex stabilizes complex I in Paracoccus denitrificans. J Biol Chem 279:5000-5007. Worlitzsch, D., R. Tarran, M. Ulrich, U. Schwab, A. Cekici, K. C. Meyer, P. Birrer, G. Bellon, J. Berger, T. Weiss, K. Botzenhart, J. R. Yankaskas, S. Randell, R. C. Boucher, and G. Doring. 2002. Effects of reduced mucus oxygen concentration in airway Pseudomonas infections of cystic fibrosis patients. J Clin Invest 109:317-325. Einsle, O., and P. M. Kroneck. 2004. Structural basis of denitrification. Biol Chem 385:875-883.

24

597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650

17. 18.

19. 20. 21.

22.

23. 24. 25. 26.

27. 28.

29.

30.

31.

Bertero, M. G., R. A. Rothery, M. Palak, C. Hou, D. Lim, F. Blasco, J. H. Weiner, and N. C. Strynadka. 2003. Insights into the respiratory electron transfer pathway from the structure of nitrate reductase A. Nat Struct Biol 10:681-687. Dias, J. M., M. E. Than, A. Humm, R. Huber, G. P. Bourenkov, H. D. Bartunik, S. Bursakov, J. Calvete, J. Caldeira, C. Carneiro, J. J. Moura, I. Moura, and M. J. Romao. 1999. Crystal structure of the first dissimilatory nitrate reductase at 1.9 A solved by MAD methods. Structure 7:65-79. Lanciano, P., A. Vergnes, S. Grimaldi, B. Guigliarelli, and A. Magalon. 2007. Biogenesis of a respiratory complex is orchestrated by a single accessory protein. J Biol Chem 282:17468-17474. Arias-Cartin, R., S. Grimaldi, P. Arnoux, B. Guigliarelli, and A. Magalon. 2012. Cardiolipin binding in bacterial respiratory complexes: structural and functional implications. Biochim Biophys Acta 1817:1937-1949. Van Alst, N. E., L. A. Sherrill, B. H. Iglewski, and C. G. Haidaris. 2009. Compensatory periplasmic nitrate reductase activity supports anaerobic growth of Pseudomonas aeruginosa PAO1 in the absence of membrane nitrate reductase. Can J Microbiol 55:1133-1144. Nurizzo, D., M. C. Silvestrini, M. Mathieu, F. Cutruzzola, D. Bourgeois, V. Fulop, J. Hajdu, M. Brunori, M. Tegoni, and C. Cambillau. 1997. N-terminal arm exchange is observed in the 2.15 A crystal structure of oxidized nitrite reductase from Pseudomonas aeruginosa. Structure 5:1157-1171. Rinaldo, S., G. Giardina, N. Castiglione, V. Stelitano, and F. Cutruzzola. 2011. The catalytic mechanism of Pseudomonas aeruginosa cd1 nitrite reductase. Biochem Soc Trans 39:195-200. Farver, O., M. Brunori, F. Cutruzzola, S. Rinaldo, S. Wherland, and I. Pecht. 2009. Intramolecular electron transfer in Pseudomonas aeruginosa cd(1) nitrite reductase: thermodynamics and kinetics. Biophys J 96:2849-2856. Hasegawa, N., H. Arai, and Y. Igarashi. 2001. Two c-type cytochromes, NirM and NirC, encoded in the nir gene cluster of Pseudomonas aeruginosa act as electron donors for nitrite reductase. Biochem Biophys Res Commun 288:1223-1230. Nicke, T., T. Schnitzer, K. Munch, J. Adamczack, K. Haufschildt, S. Buchmeier, M. Kucklick, U. Felgentrager, L. Jansch, K. Riedel, and G. Layer. 2013. Maturation of the cytochrome cd1 nitrite reductase NirS from Pseudomonas aeruginosa requires transient interactions between the three proteins NirS, NirN and NirF. Biosci Rep 33. Bali, S., M. J. Warren, and S. J. Ferguson. 2010. NirF is a periplasmic protein that binds d(1) heme as part of its essential role in d(1) heme biogenesis. FEBS J 277:4944-4955. Storbeck, S., S. Saha, J. Krausze, B. U. Klink, D. W. Heinz, and G. Layer. 2011. Crystal structure of the heme d1 biosynthesis enzyme NirE in complex with its substrate reveals new insights into the catalytic mechanism of S-adenosyl-Lmethionine-dependent uroporphyrinogen III methyltransferases. J Biol Chem 286:26754-26767. Haufschildt, K., S. Schmelz, T. M. Kriegler, A. Neumann, J. Streif, H. Arai, D. W. Heinz, and G. Layer. 2014. The crystal structure of siroheme decarboxylase in complex with iron-uroporphyrin III reveals two essential histidine residues. J Mol Biol 426:3272-3286. Adamczack, J., M. Hoffmann, U. Papke, K. Haufschildt, T. Nicke, M. Broring, M. Sezer, R. Weimar, U. Kuhlmann, P. Hildebrandt, and G. Layer. 2014. NirN Protein from Pseudomonas aeruginosa is a Novel Electron-bifurcating Dehydrogenase Catalyzing the Last Step of Heme d1 Biosynthesis. J Biol Chem 289:30753-30762. Sharma, V., C. E. Noriega, and J. J. Rowe. 2006. Involvement of NarK1 and NarK2 proteins in transport of nitrate and nitrite in the denitrifying bacterium Pseudomonas aeruginosa PAO1. Appl Environ Microbiol 72:695-701. 25

651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703

32. 33. 34. 35. 36. 37. 38.

39. 40. 41. 42. 43. 44.

45. 46. 47. 48.

49.

Hino, T., Y. Matsumoto, S. Nagano, H. Sugimoto, Y. Fukumori, T. Murata, S. Iwata, and Y. Shiro. 2010. Structural basis of biological N2O generation by bacterial nitric oxide reductase. Science 330:1666-1670. Shiro, Y. 2012. Structure and function of bacterial nitric oxide reductases: nitric oxide reductase, anaerobic enzymes. Biochim Biophys Acta 1817:1907-1913. Hayashi, N. R., H. Arai, T. Kodama, and Y. Igarashi. 1998. The nirQ gene, which is required for denitrification of Pseudomonas aeruginosa, can activate the RubisCO from Pseudomonas hydrogenothermophila. Biochim Biophys Acta 1381:347-350. Wust, A., L. Schneider, A. Pomowski, W. G. Zumft, P. M. Kroneck, and O. Einsle. 2012. Nature's way of handling a greenhouse gas: the copper-sulfur cluster of purple nitrous oxide reductase. Biol Chem 393:1067-1077. Pomowski, A., W. G. Zumft, P. M. Kroneck, and O. Einsle. 2011. N2O binding at a [4Cu:2S] copper-sulphur cluster in nitrous oxide reductase. Nature 477:234-237. Wunsch, P., and W. G. Zumft. 2005. Functional domains of NosR, a novel transmembrane iron-sulfur flavoprotein necessary for nitrous oxide respiration. J Bacteriol 187:1992-2001. Taubner, L. M., M. A. McGuirl, D. M. Dooley, and V. Copie. 2004. 1H, 13C, 15N backbone and sidechain resonance assignments of apo-NosL, a novel copper(I) binding protein from the nitrous oxide reductase gene cluster of Achromobacter cycloclastes. J Biomol NMR 29:211-212. McGuirl, M. A., J. A. Bollinger, N. Cosper, R. A. Scott, and D. M. Dooley. 2001. Expression, purification, and characterization of NosL, a novel Cu(I) protein of the nitrous oxide reductase (nos) gene cluster. J Biol Inorg Chem 6:189-195. Dell'acqua, S., I. Moura, J. J. Moura, and S. R. Pauleta. 2011. The electron transfer complex between nitrous oxide reductase and its electron donors. J Biol Inorg Chem 16:1241-1254. Williams, H. D., J. E. Zlosnik, and B. Ryall. 2007. Oxygen, cyanide and energy generation in the cystic fibrosis pathogen Pseudomonas aeruginosa. Adv Microb Physiol 52:1-71. Ugidos, A., G. Morales, E. Rial, H. D. Williams, and F. Rojo. 2008. The coordinate regulation of multiple terminal oxidases by the Pseudomonas putida ANR global regulator. Environ Microbiol 10:1690-1702. Kawakami, T., M. Kuroki, M. Ishii, Y. Igarashi, and H. Arai. 2010. Differential expression of multiple terminal oxidases for aerobic respiration in Pseudomonas aeruginosa. Environ Microbiol 12:1399-1412. Borrero-de Acuna, J. M., A. Bielecka, S. Haussler, M. Schobert, M. Jahn, C. Wittmann, D. Jahn, and I. Poblete-Castro. 2014. Production of medium chain length polyhydroxyalkanoate in metabolic flux optimized Pseudomonas putida. Microb Cell Fact 13:88. Lee, P. Y., K. H. Bae, C. W. Kho, S. Kang, H. Lee do, S. Cho, S. C. Lee, B. C. Park, and S. G. Park. 2008. Interactome analysis of yeast glutathione peroxidase 3. J Microbiol Biotechnol 18:1364-1367. Leitner, A., T. Walzthoeni, A. Kahraman, F. Herzog, O. Rinner, M. Beck, and R. Aebersold. 2010. Probing native protein structures by chemical cross-linking, mass spectrometry, and bioinformatics. Molecular & Cellular Proteomics 9:1634-1649. Ong, S. E., and M. Mann. 2005. Mass spectrometry-based proteomics turns quantitative. Nat Chem Biol 1:252-262. Klockgether, J., A. Munder, J. Neugebauer, C. F. Davenport, F. Stanke, K. D. Larbig, S. Heeb, U. Schock, T. M. Pohl, L. Wiehlmann, and B. Tummler. 2010. Genome diversity of Pseudomonas aeruginosa PAO1 laboratory strains. J Bacteriol 192:1113-1121. Lewenza, S., R. K. Falsafi, G. Winsor, W. J. Gooderham, J. B. McPhee, F. S. Brinkman, and R. E. Hancock. 2005. Construction of a mini-Tn5-luxCDABE mutant 26

704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748

50.

51. 52. 53.

54. 55.

56. 57. 58.

59. 60. 61. 62.

library in Pseudomonas aeruginosa PAO1: a tool for identifying differentially regulated genes. Genome Res 15:583-589. Benkert, B., N. Quack, K. Schreiber, L. Jaensch, D. Jahn, and M. Schobert. 2008. Nitrate-responsive NarX-NarL represses arginine-mediated induction of the Pseudomonas aeruginosa arginine fermentation arcDABC operon. Microbiology 154:3053-3060. Schweizer, H. P., T. Klassen, and T. Hoang. 1996. Improved methods for gene analysis and expression in Pseudomonas spp. Engler, C., R. Gruetzner, R. Kandzia, and S. Marillonnet. 2009. Golden gate shuffling: a one-pot DNA shuffling method based on type IIs restriction enzymes. PLoS One 4:e5553. Lambert, J. P., G. Ivosev, A. L. Couzens, B. Larsen, M. Taipale, Z. Y. Lin, Q. Zhong, S. Lindquist, M. Vidal, R. Aebersold, T. Pawson, R. Bonner, S. Tate, and A. C. Gingras. 2013. Mapping differential interactomes by affinity purification coupled with data-independent mass spectrometry acquisition. Nat Methods 10:1239-1245. Wessel, D., and U. I. Flugge. 1984. A method for the quantitative recovery of protein in dilute solution in the presence of detergents and lipids. Anal Biochem 138:141-143. Duvel, J., D. Bertinetti, S. Moller, F. Schwede, M. Morr, J. Wissing, L. Radamm, B. Zimmermann, H. G. Genieser, L. Jansch, F. W. Herberg, and S. Haussler. 2012. A chemical proteomics approach to identify c-di-GMP binding proteins in Pseudomonas aeruginosa. J Microbiol Methods 88:229-236. Kaake, R. M., X. Wang, and L. Huang. 2010. Profiling of protein interaction networks of protein complexes using affinity purification and quantitative mass spectrometry. Mol Cell Proteomics 9:1650-1665. Zumft, W. G., C. Braun, and H. Cuypers. 1994. Nitric oxide reductase from Pseudomonas stutzeri. Primary structure and gene organization of a novel bacterial cytochrome bc complex. Eur J Biochem 219:481-490. Howden, A. J., V. Geoghegan, K. Katsch, G. Efstathiou, B. Bhushan, O. Boutureira, B. Thomas, D. C. Trudgian, B. M. Kessler, and D. C. Dieterich. 2013. QuaNCAT: quantitating proteome dynamics in primary cells. Nature methods 10:343346. Hayashi, N. R., H. Arai, T. Kodama, and Y. Igarashi. 1998. The nirQ gene, which is required for denitrification of Pseudomonas aeruginosa, can activate the RubisCO from Pseudomonas hydrogenothermophila. Biochim Biophys Acta 1381:347-350. Santini, S., A. R. Bizzarri, T. Yamada, C. W. Beattie, and S. Cannistraro. 2014. Binding of azurin to cytochrome c 551 as investigated by surface plasmon resonance and fluorescence. J Mol Recognit 27:124-130. Cuypers, H., A. Viebrock-Sambale, and W. G. Zumft. 1992. NosR, a membranebound regulatory component necessary for expression of nitrous oxide reductase in denitrifying Pseudomonas stutzeri. J Bacteriol 174:5332-5339. Honisch, U., and W. G. Zumft. 2003. Operon structure and regulation of the nos gene region of Pseudomonas stutzeri, encoding an ABC-Type ATPase for maturation of nitrous oxide reductase. J Bacteriol 185:1895-1902.

749 750

27

751

FIGURE LEGENDS

752 753

Figure 1. Proposed respiratory chains of P. aeruginosa based on the model by Williams, H.D. et al. 2007 (41).

754 755 756 757 758 759 760 761 762 763 764

Figure 2. In vivo and in vitro nitrate and nitrite conversion of P. aeruginosa denitrification mutants. A) In vivo nitrate consumption by wildtype (orange) and mutants of narG (brown), norB (red), norC (green), nosZ (lilac) and nosR (blue) is shown in the left panel. In vitro nitrate utilization and nitrite production of wildtype (red) and mutants of norB (orange) and narG (brown) were determined with cell-free extracts from the corresponding strains. B) The cell-free extract of the indicated strains were tested for the presence of the nitrate reductase subunit NarH (left) and the nitrite reductase NirS (right) using Western blot analyses. C) In vivo nitrite conversion of wildtype (orange) and mutants of nosZ (lilac), norC (green), norB (red), nosR (blue) and nirS (black) is shown in the left panel. In vitro nitrite utilization by wildtype (orange) and the norB (red) and nirS (black) mutant strains was determined using cell-free extracts from the corresponding strains (right panel).

765 766 767 768 769 770 771 772 773 774

Figure 3. Interactomics of the NosR, NorC and NorB membrane proteins with multiple potential interaction partners. Tagged bait proteins were produced as fusion proteins in P. aeruginosa, chemically cross-linked to interaction partners and affinity-purified. Fractions were resolved on an SDS-PAGE gel and visualized by Coomassie Brilliant Blue staining. The copurification fractions for NosR (lane 1), NorB (lane 2), NorC (lane 3) and the control reaction (PA01-PAS40) are shown as indicated (lanes 6, 8, 10). Western-blot analyses showed the presence of NosR-His6x fusion protein (lane 5), NorB-His6x fusion protein (lane 7) and the NorC-His6x fusion protein (lane 9). MW = molecular weight expressed in kDa. The recombinant strains PAO1nosR-pnosR-His6x, PAO1norC-pnorC-His6x and PAO1norB-pnorBHis6x were subjected to these interactomic studies.

775 776 777 778 779 780 781 782

Figure 4. Abundance distribution of potential interaction partners of NosR, NorC and NorB non- or related to denitrification. Graph representing the concentration distribution of potential interaction partners for the various bait proteins with NosR (green), NorC (blue) and NorB (red) were determined after co-purification. The percentage of found peptides for certain proteins (Y axis) was plotted against the corresponding area value (X axis). B) Potential interaction partners involved in denitrification are shown in the left panel (A) while all the other proteins are depicted in the right site (B). Data of this plot were obtained from triplicate experiments with corresponding average and standard deviation.

783 784 785 786 787 788 789 790

Figure 5. Potential interaction partners for NorB, NorC and NosR involved in denitrification. Box-Plot represents interaction partners found for NorC (blue), NorB (red) and NosR (green) plotted against the area value. Only interaction partners found in all replicates for each protein are depicted here. Boxes height encloses the difference between the first and the third quartiles of the values enclosing the 50% of the data. Lines extending vertically (whiskers) denote maximal and minimal values encountered. The median (second quartile) is presented as a black band within the boxes. The area value (Y axis) is given in logarithmic scale. The R program was used to generate this plot.

791 792 793 794 795 796

Figure 6. Double immunogold labeling of NosR, NarH, NirS, NorC and NosZ to determine their in vivo co-localization in the P. aeruginosa cell. The name of the proteins immunogold-labeled and tested for co-localization are shown above each picture. Numbers for each protein represent the gold-particle size employed for each labeling. Bar sizes correspond to 200 nm. Arrows indicate proteins as co-localized only if the distance between them was less than 25 nm.

797 798

Figure 7. Denitrification super-complex model proposed in this study. The figure is based on the interactomics and immunogold labeling data. A tight complex of the nitrate 28

799 800

(Nar), nitrite (Nir), NO (Nor) and N2O reductase enzymes with their corresponding assembly and cofactor proteins was observed.

801 802

Figure 8. Schematic model of the respirasome relying on the NorCB-NosR supracomplex. C = cytoplasm; IM = inner membrane; P = periplasm and OM = outer membrane.

803

29

804

Tables

805

Table 1. List of bacterial strains and plasmids used in this study. Plasmids

Relevant characteristics

Source / reference

r

pJET 1.2

Ap , Carrier vector, toxin expressed Thermo Scientific, upon self-ligation Darmstadt, Germany

pAS40

pJET 1.2-nosR

Cbr, expression vector for P. Laboratory aeruginosa. Based on pUCP20t collection without BsaI. (GGTCTC  GCTCTC) Apr, pJET 1.2, carrying nosR This study

pJET 1.2-norC

Apr, pJET 1.2, carrying norC

pJET 1.2-norCB pJET 1.2-nosPR pJET 1.2-nosZ pAS40-Strep-tag II (Z) pAS40-His6x (R,C,B) pAS40-nosR-His6x pAS40-norC-His6x pAS40-norCB-His6x pAS40-nosZ-Strep-tag II

r

Ap , pJET 1.2, carrying norCB

This study This study

r

Ap , pJET 1.2, carrying nos 500 bp This study upstream region (includes promoter) Apr, pJET 1.2, carrying nosZ This study Cbr, pAS40 carrying the Strep-tag II suitable for nosZ cloning Cbr, pAS40 carrying the His6x tag suitable for nosR, norC and norB cloning, respectively Cbr, pAS40 carrying nosR His6x tagged Cbr, pAS40 carrying norC His6x tagged Cbr, pAS40 carrying norCB with His6x tagged norB Cbr, pAS40 carrying nosZ Strep-tag II tagged

This study This study This study This study This study This study

Bacterial strains E. coli DH10b

Parental strain

Invitrogen, Darmstadt, Germany

DH10b pJET1.2-nosZ

DH10b carrying the P. aeruginosa nosZ DH10b carrying the nosZ promoter

This study

DH10b carrying the P. aeruginosa nosR DH10b carrying the P. aeruginosa norC DH10b carrying the P. aeruginosa norCB DH10b harboring the expression vector pAS40-nosZ DH10b harboring the expression vector pAS40-nosR

This study

DH10b pJET1.2-nosZPR DH10b pJET1.2-nosR DH10b pJET1.2-norC DH10b pJET1.2-norCB DH10b pAS40-nosZ DH10b pAS40-nosR

This study

This study This study This study This study

30

DH10b pAS40-norC

DH10b harboring the expression vector pAS40-norC DH10b harboring the expression vector pAS40-norCB

This study

PAO1 wild type

Parental strain

PAO1 pAS40

PAO1 wild type carrying pAS40

DSMZ, Braunschweig, Germany This study

PAO1nosR

PAO1nosR::mini-Tn5 luxCDABE

(49)

DH10b pAS40-norCB

This study

P. aeruginosa

PAO1norC

PAO1norC::mini-Tn5 luxCDABE

(49)

PAO1norB

PAO1norB::mini-Tn5 luxCDABE

(49)

PAO1nosZ

PAO1nosZ::mini-Tn5 luxCDABE

(49)

PAO1nosR –pnosRHis6x

PAO1nosR harboring nosR under This study the control of the nos promoter and encoding a His6x-tag PAO1norC harboring norC under the This study control of the nor promoter and encoding a His6x PAO1norB harboring norCB under This study the control of the nor promoter and encoding a His6x at the norB gene PAO1nosZ harboring nosZ under the This study control of the nos promoter and encoding a Strep-tag II

PAO1norC-pnorC-His6x PAO1norB-pnorB-His6x PAO1nosZ-pnosZ-Streptag II 806 807 808

Antibiotic resistance: Apr = Ampicillin resistance; Cbr = Carbenicillin resistance. (R), (C) and (B) indicate that the plasmid was used for the cloning of nosR, norC and norB His6x fusion genes, respectively. (Z) indicates that the plasmid is suitable for nosZ fusion gene cloning.

809 810

31

811

Table 2. List of primers used in this study. Primers nosRFw

CACCGGTCTCGGGTGTATCTCACCTGCGGCAA

nosRRv

CCTCGGTCTCTGAGGTTCTCCGCGGGTATCCG

nosRHis6xFw

norCFw

AATTGGTGTGAGACCAGATCTGATATCGGTCTCACCTCCATCATCAT CATCATCATTGA AGCTTCAATGATGATGATGATGATGGAGGTGAGACCGATATCAGAT CTGGTCTCACACC GCCAGGTCTCGCTGATGCTCTACTGGCACATGGTC

norCRv

CAAGGGTCTCTACCCTCCTTGTTCGGCGGCCA

norCHis6xFw

norCBFw

AATTCTGATGAGACCAGATCTGATATCGGTCTCAGGGTCATCATCAT CATCATCATTGA AGCTTCAATGATGATGATGATGATGACCCTGAGACCGATATCAGATC TGGTCTCATCAG GCCAGGTCTCGCTGATGCTCTACTGGCACATGGTC

norCBRv

CAAGGGTCTCTGGCGGCCGCCTTGCCGCGCCG

norCBHis6xFw

nosZFw

AATTCTGATGAGACCAGATCTGATATCGGTCTCACGCCCATCATCAT CATCATCATTGA AGCTTCAATGATGATGATGATGATGGGCGTGAGACCGATATCAGAT CTGGTCTCATCAG TACCGGTCTCTGGTAACCTCTGACCTGGCTCC

nosZRv

GGCTGGTCTCTAGCCTTTTCCACCAGCATCCG

nosPRFw nosPRRv

GATAGGTCTCGTATCTCACCTGCGGCAACGAC GGTAGGTCTCGTACCGGCGAAAAAAGGGACGC

nosZStrep-tag IIFw nosZStrep-tag IIRv

AATTTATCCGAGACCAGATCTGATATCGGTCTCCGGCTTGGAGCCA CCCGCAGTTCGAAAAATGA AGCTTCATTTTTCGAACTGCGGGTGGCTCCAAGCCGGAGACCGATA TCAGATCTGGTCTCGGATA

nosRHis6xRv

norCHis6xRv

norCBHis6xRv

812 813 814

Sequence (5´-3´)

All primers were purchased from Invitrogen, Darmstadt, Germany. Within the primers sequences of the BsaI restriction site is underlined, the 4 first nucleotides and the 4 last of the amplicon sequence are colored in red. The BglII-EcoRV restriction sequence is colored in green.

815 816

32

817 818 819 820

Table 3. Determination of NorB/C, NosR interaction partners from the denitrification process via binding, cross-linking, affinity co-purification and proteomic identification. The LC-MS/MS data reveal the enrichment of these proteins during the co-purification given as area values (fold enrichment) in comparison to the negative control. PA Nº PA4292 PA0527 PA3875 PA3874 PA3873 PA3876 PA3879 PA3878 PA0510 PA0516 PA0511 PA0514 PA0518 PA0509 PA0520 PA0519 PA0524 PA0523 PA0525 PA3393 PA3394 PA3396 PA3391 PA3392

Interaction Partners Azurin Transcriptional regulator Dnr Respiratory nitrate reductase alpha chain NarG Respiratory nitrate reductase beta chain NarH Respiratory nitrate reductase delta chain NarJ Nitrite extrusion protein NarK2 Two-component response regulator NarL Two-component sensor NarX Uroporphyrin-III cmethyltransferase NirE Heme d1 biosynthesis protein NirF Heme d1 biosynthesis protein NirJ Heme d1 biosynthesis protein NirL Cytochrome c551 NirM Putative c-type cytochrome NirN Regulatory protein NirQ Nitrite reductase NirS Nitric-oxide reductase subunit B NorB Nitric-oxide reductase subunit C NorC Probable denitrification protein NorD NosD protein NosF protein NosL protein Regulatory protein NosR Nitrous oxide reductase NosZ

NorB ND ∞ 0.98

NorC 0.6 ND 7.08

1.26

12.06 ++ 3.41 ++

ND

∞

ND

ND ∞

3.11 ND

ND ND

ND ∞

∞ ∞

ND ND

18.89

163.94

14.42

ND

ND

∞

∞

ND

ND

0.11 0.96

4.17 2.27

1.00 ND

1.75 2.17

10.54 2.26 1.97 +++ 0.98 ++ 435.73 ND

10.61

NosR ND ND 2.46

ND ++

ND

15.73

∞ ∞ 2.82 1.91 0.77

ND

NarH

NosZ

-

++

++

+++

++

+++

++ ++

+++ -

3.81

ND ND ∞ 68.91 18.65 52.88 ++ 2.61 +++ 7.11 +++

821 822 823 824 825

Each number represents the average area of the protein in co-purifications over the average area of the protein in the control fraction. 0 = protein present in the control fraction but not in the co-purification. ∞ = protein found present in the cross-linked, co-purified fraction but not in the control fraction. GI accession numbers for the corresponding genes are shown in 33

826 827 828 829

brackets. ++ = specific interaction with the tagged bait protein in the immunogold labeling experiments. +++ = highly specific interaction with the tagged bait protein in the immunogold labeling experiments. Underlined values correspond to the proteins with a strong interaction in the co-purification experiments. The proteins in the table are ordered alphabetically.

830 831 832

Table 4: Quantification of in vivo protein co-localizations using double immunogoldlabeling studies and electron microscopy Letters Interaction (referring to Figure 6)

Average in 20 cells

Total N° Average co-localization in co-localization 20 cells in 20 cells

PAG 15 PAG 10 NarH NirS 3 1.95 3 0.15 B NarH-NorC NarH NorC 7.5 4.8 14 0.7 C NorC - NirS NorC NirS 8.35 3.65 25 1.25 D NosZ-NirS NosZ NirS 8.6 4.6 25 1.25 E NosR - NorC NosR NorC 9.95 2.7 17 0.85 F NosZ-NosR NosZ NosR 7.45 4.4 20 1.7 G NosZ-NarH NosZ NarH 8.25 3 12 0.6 H NarH - NosR NarH NosR 6.05 3.15 12 0.6 I NosZ - NorC NosZ NorC 10.5 7.2 36 1.8 J NosR-NirS NosR NirS 8.05 3.55 22 1.1 For 20 longitudinal sections with approximately the same area, large and small gold-particles as well as co-localization events in those cells were counted. The average co-localization value was then calculated for the 20 cells. A

833 834 835

Average in 20 cells

NarH-NirS

836 837

34