Lifestyle transitions and adaptive pathogenesis of Pseudomonas aeruginosa 1

Martina Valentini*#, Diego Gonzalez§, Despoina A.I. Mavridou#, Alain Filloux*# 2

#: MRC Centre for Molecular Microbiology and Infection, Department of Life Sciences, Imperial 3

College London, SW7 2AZ London, United Kingdom; 4

§: Département de Microbiologie Fondamentale, Université de Lausanne, CH-1015 Lausanne, 5

Switzerland. 6

7

*: corresponding authors. E-mail: m.valentini@imperial.ac.uk; a.filloux@imperial.ac.uk 8

9

Abstract 10

Pseudomonas aeruginosa acute and chronic infections are of great concern to human health, 11

especially in hospital settings. It is currently assumed that P. aeruginosa has two antagonistic 12

pathogenic strategies that parallel two different lifestyles; free-living cells are predominantly 13

cytotoxic and induce an acute inflammatory reaction, while biofilm-forming communities cause 14

refractory chronic infections. Recent findings suggest that the planktonic-to-sessile transition is a 15

complex, reversible and overall dynamic differentiation process. Here, we examine how the Gac/Rsm 16

regulatory cascade, a key player in this lifestyle switch, endows P. aeruginosa with both a permissive 17

lifecycle in nature and flexible virulence strategy during infection. 18

Highlights 19

• Pseudomonas aeruginosa causes acute and chronic infections in humans. 20

• Virulence factors are co-regulated with lifestyles. 21

• The Gac/Rsm cascade regulates the lifestyle and infection strategy switch. 22

• The Gac/Rsm action is modulated during biofilm formation. 23

• The Gac/Rsm cascade has a social nature and responds to bacterial competition. 24

Introduction 25

Pseudomonas aeruginosa is an opportunistic human pathogen associated with a wide range of 26

infections affecting, among others, skin, ear, eye, urinary tract, heart, airway and lung tissues [1]. The 27

high frequency of P. aeruginosa strains causing nosocomial infections, the increasing occurrence of 28

multi-drug resistant strains, and the adaptive antimicrobial resistance displayed by the bacterium 29

during chronic infections, cause a severe threat to human health worldwide [2]. P. aeruginosa 30

pathogenesis has been shown to be multifactorial and combinatorial [3]. Some virulence factors have 31

been known for decades, while others have only been identified recently, thanks to whole genome 32

sequencing of environmental and clinical isolates [4]. P. aeruginosa infections can be acute or chronic 33

[5]. Interestingly, the type of infection is independent of the pathogen's genotype, but possibly linked 34

to the host health status and the lifestyle adopted by the bacteria when colonising the host [6]. Acute 35

infections are predominantly associated with bacteria adopting a planktonic lifestyle, while biofilm-36

related functions (attachment, capsule components) play a major role in persistent infections. It has 37

been proposed that a binary regulatory switch controls the transition between the two lifestyles and 38

infection modes of P. aeruginosa [6]. This switch model predicts that biofilm-associated functions 39

and virulence factors of acute infection cannot physiologically co-exist in the bacterium. However, 40

recent studies on the formation of biofilms indicate that the lifestyle switch model might be a 41

simplified and static view of a plastic cell differentiation process, where acute and chronic infectious 42

traits can co-exist within a bacterial population. Here, we review the lifestyle/infection model with an 43

emphasis on the role of the Gac/Rsm global gene regulatory network in ensuring behavioural 44

plasticity in P. aeruginosa. We propose that it is indeed the permissive lifecycle of the bacterium 45

which is key for its success in the environment as well as for its adaptive pathogenesis during host 46

infections. 47

The planktonic-to-sessile switch model 48

The main elements and differences of the two P. aeruginosa lifestyles/infection modes are depicted in 49

Figure 1, and are briefly described below. P. aeruginosa acute lung infection (pneumonia), occurring 50

mainly in immunocompromised or intensive-care patients, is initiated by bacteria binding at the 51

mucosal barrier through two major adhesins, flagella and retractile type IV pili, which trigger a host 52

inflammatory response [1]. Once in contact with the epithelial cells, bacteria cause significant tissue 53

damage by translocating cytotoxic effectors directly into the eukaryotic cells via their type III 54

secretion system (T3SS), and by secreting another set of virulence factors in the extracellular milieu 55

[7]. The T3SS effectors are also involved in the inhibition of phagocytosis and, together with the 56

LasB protease, eventually cause loss of the endothelial barrier integrity. The disruption of this barrier 57

allows the pathogen to grow and spread rapidly within the host, sometimes resulting in septicaemia 58

[7]. Overall, the process of acute infection leads to a massive mobilization of the immune system, 59

starting with the recruitment of neutrophils and macrophages to the lungs. By contrast, bacteria 60

involved in chronic infections are slow growing, less cytotoxic and immunogenic, and can persist in 61

the host for decades without reaching the bloodstream. Chronic infections are common in the lungs of 62

patients with cystic fibrosis (CF), primary ciliary dyskinesia and bronchiectasis [8]. In chronically-63

infected CF lungs, clusters of P. aeruginosa are found encased in a polysaccharide matrix within a 64

thick layer of mucus [9]. Common adaptation routes of P. aeruginosa in the CF lungs include the loss 65

of major determinants of the planktonic cells, like motility or a functional T3SS, and the conversion to 66

a mucoid colony phenotype. This has as a result that CF-adapted lineages are usually avirulent in 67

mouse model of acute infection, but unhampered in their ability to establish chronic infections [10]. 68

Altogether, these observations on the lifestyle of the bacterium led to the suggestion of a binary model 69

of P. aeruginosa pathogenesis, in which cells in the planktonic state are equipped for the aggressive 70

host-invasion strategy seen in acute infections, while biofilm-forming cells are pre-adapted for long-71

term persistence and immune evasion, as observed in the airways of chronically-infected CF patients 72

[6]. 73

The identification and initial mapping of the Gac/Rsm global regulatory network (Box 1) 74

substantiated this model, grounding it into the fundamental decision-making processes of P. 75

aeruginosa [11]. Briefly, it was found that RsmA positively regulates the expression of virulence 76

factors which are important during the planktonic state and acute infection, while it negatively 77

regulates factors associated with biofilm formation and chronic infection. RsmA-induced virulence 78

factors include flagella, rhamnolipids, type IV pili, the T3SS and its effectors, the Xcp type II 79

secretion system (T2SS), exotoxin A and lipase. On the other hand, genes repressed by RsmA encode 80

biofilm matrix components (pel, psl), Quorum Sensing (AHL), the type VI secretion system (T6SS) 81

and secondary metabolites (siderophores, HCN, phenazines) [12]. Moreover, the rsmA mutant showed 82

reduced cytotoxicity for epithelial cells and decreased dissemination efficiency in a mouse model of 83

acute infection, but displayed increased persistence and lung inflammation in a mouse model of 84

chronic infection [13]. Although the molecular cues which determine the choice between the two 85

infection strategies remain elusive, the detection of kin-cell lysis, via RetS and GacS, and the sensing 86

of high calcium levels, via the LadS kinase, have recently been identified as major players in the 87

initiation of P. aeruginosa biofilm formation via the Gac/Rsm cascade [14,15]. 88

The Gac/Rsm cascade as lifestyle modulator 89

Generally, planktonic and sessile bacteria differ in their physiology and behaviour; they have different 90

growth rates, forms of motility, degrees of virulence, propensities for social interactions and antibiotic 91

resistance or tolerance. However, planktonic and biofilm-forming populations of P. aeruginosa are 92

not as significantly distinct as, for example, the infective and dissemination forms of primary 93

pathogens like Legionella pneumophila or Salmonella enterica [16,17]. Several studies have 94

measured the difference in gene expression profiles between P. aeruginosa planktonic and biofilm 95

cultures, but surprisingly little consistence is found across the studies. Estimates of the difference in 96

gene expression between the two lifestyles vary from 1% to 13% of the total genes, and a consensus 97

regarding a putative biofilm-specific regulon has, so far, not been reached [18-21]. To account for 98

this, it is proposed that although changes in cells morphology and physiology always occur during the 99

lifestyle transition, the precise experimental condition or environment in which the transition takes 100

place has a major impact on the process. The transition to a sessile state in P. aeruginosa has recently 101

been described as a progressive differentiation program, which involves several, possibly not clearly 102

delimited stages, generally found during biofilm formation, starting by surface attachment and going 103

on to microcolony formation, biofilm maturation and eventually dispersal or detachment [22]. The 104

bacterial physiology and behaviour continuously change throughout this process. For example, cells 105

undergoing surface-attachment or detaching from biofilms have different gene expression profiles and 106

show increased virulence compared to their planktonic counterparts [23,24]. Bacterial functionalities 107

within a lifestyle are therefore continuously modulated as part of a differentiation program and/or as a 108

response to environmental conditions. Some studies have already offered evidence that P. aeruginosa 109

lifestyles and infection strategies are more flexible than previously thought, as some functions 110

traditionally associated with planktonic cells were found to occur within biofilms and vice-versa. 111

Individual-cell chemotaxis, for example, is not exclusive to flagellated cells, as previously thought, 112

but also plays a role in pili-mediated motility [25]. In addition, Turner and colleagues observed that 113

planktonic/acute- and sessile/chronic-related virulence genes are not mutually exclusive in acute and 114

chronic wound infections in a mouse model. For example, an up-regulation of T3SS, rhamnolipids 115

and pyochelin (siderophore) expression was observed in both chronic and acute infections [26]. On 116

the same lines, P. aeruginosa biofilm components were shown to contribute to acute infections of 117

burn wounds in mice [27], and the T3SS and T6SS genes, which were thought to be antagonistically 118

regulated, were found to be both upregulated during acute murine respiratory infection, along with 119

iron uptake genes, usually associated with chronic infection [28]. 120

A more detailed genetic analysis of the Gac/Rsm cascade has also suggested that the initial 121

binary switch model was not fully capturing the complexity of the regulatory system. The core 122

components of the cascade (Box 1), being post-transcriptional regulators, i.e. RNA-binding protein(s) 123

and small RNAs, have an intrinsic modulatory potential. It is known that post-transcriptional 124

regulation allows both an ON/OFF switch and fine-tuned control of gene expression [29]. RsmA (and 125

its homologue RsmN) shows a wide range of binding affinities for its mRNA targets and high affinity 126

for its titrating small RNAs (RsmY/Z) [30,31]. Therefore, even a subtle variation in small RNA levels 127

could drastically affect the RsmA regulatory output. The regulation of the small RNAs levels can 128

occur either via activation/repression of the Gac/Rsm accessory sensors (Box 1) or via other cross-129

regulatory pathways [12,32,33]. The latter can be global gene regulatory networks, like cyclic di-130

GMP signalling, but also regulatory modules, like sensor kinases (e.g. RetS, LadS and PA1611), two-131

component systems (e.g. BfiSR) and small RNAs (e.g. RsmW) [34-36]. The Gac/Rsm flexibility is 132

particularly important during the planktonic-to-sessile lifestyle transition. The increase in RsmY/Z 133

levels, occurring after the activation of the Gac/Rsm cascade, was initially considered to be the key 134

step initiating biofilm formation by increasing surface attachment [11]. However, RsmY and/or RsmZ 135

levels, and consequently RsmA activity, have more recently been shown to be continuously 136

modulated during the process of biofilm development (Figure 2) [19,23,34,37]. A decrease in RsmZ 137

levels, through activation of the BfiSR two-component-system, ensures the progression of the biofilm 138

development after surface attachment [37]. Moreover, in mature biofilms, the levels of RsmY/Z and 139

RsmW, a newly discovered RsmA-binding small RNA, increase sharply compared to the planktonic 140

state [19,34]. Finally, in dispersed cells, the expression of rsmY and rsmZ is down-regulated compared 141

to both biofilm and planktonic cells [23]. 142

In addition to its modulatory action, it is an open question whether the Gac/Rsm cascade is 143

involved in the generation of physiological heterogeneity in bacterial populations. Phenotypic 144

heterogeneity occurs in biofilm structures, in part because of oxygen and nutrient gradients, and in 145

part because of inter-bacterial communication through secretion of chemical signals in their 146

immediate surroundings [38]. Localised gene expression in P. aeruginosa biofilms has been shown 147

for virulence-factor genes like aprA and phzA1 [39]. The RsmA regulon includes both cooperative 148

(e.g. exopolysaccharide and siderophore production) and competitive (e.g. T6SS, phenazine, 149

pyocyanin) traits [40-42], which could benefit from being expressed heterogenously by specialised 150

subpopulations [41]. 151

To conclude, the architectural complexity of the Gac/Rsm cascade can not easily be reduced 152

to an ON/OFF switch. Elements of this complexity include the multiple, possibly conflicting, inputs 153

which feed into it, the apparent redundancy in the transducing small RNAs (RsmWYZ), and the 154

generalised cross-regulation which links the Gac/Rsm cascade with other networks acting during 155

biofilm development. The impact of these regulatory intricacies on the dynamics of P. aeruginosa 156

pathogenesis and infection strategies remains to be fully evaluated. 157

Conclusions and perspectives 158

P. aeruginosa strains, including lineages which are infectious in humans, are widely distributed in the 159

environment and can be found both in association with non-human eukaryotic hosts and in host-free 160

ecosystems [43]. Environmental strains express typical virulence factors and multidrug resistance 161

determinants, while clinical strains can use oil hydrocarbons as carbon source [44]. Therefore, many 162

of the P. aeruginosa traits and behaviors as a pathogen might have been shaped by selective pressures 163

unrelated to human pathogenesis [45]. It is likely that the continuous cycling between a planktonic 164

and a sessile state, encountered by P. aeruginosa during its evolution, is particularly relevant in 165

changing environments outside the host. Many environmental bacteria, like Caulobacter crescentus, 166

Rhodopseudomonas palustris, and other opportunists like Vibrio cholerae present similar cycles [45-167

47]. Plasticity in lifestyles has been recently shown to contribute to the evolutionary success of V. 168

cholerae under fluctuating conditions [48]. This could also be the case for P. aeruginosa. In most 169

environmental niches, but also in acute and early chronic infections, the aptitude for lifestyle 170

transitions and the behavioural modulation associated with it might be favored [26-28]. By contrast, in 171

relatively stable environments, this flexibility could be lost through drift or counter-selected in favour 172

of niche specialisation. In long-lasting infections, like in CF-lung chronic infections, ecological 173

stability and constant selection against immunogenicity are two important selective forces. 174

Accordingly, over time, P. aeruginosa often looses its behavioural/virulence flexibility and traits 175

exclusively associated with the biofilm-forming lifestyle are fixed [49]. Mutations in global regulators 176

(such as lasR, vfr, rpoN, mucA and retS) are frequently observed in CF isolates, because they affect 177

multiple cell functions or phenotypes simultaneously [50]. However, these mutations are not selected 178

in natural environments, presumably because they are detrimental (evolutionary dead-end) [51]. 179

Further work will be necessary to fully characterize the features and functions of P. aeruginosa 180

phenotypic and virulence-related plasticity during infection and to understand how the Gac/Rsm 181

cascade and other regulatory networks contribute to it. 182

183

Acknowledgements 184

Work in the Filloux laboratory is supported by Biotechnology and Biological Sciences Research 185

Council (BBSRC) Grant numbers BB/N002539/1 and BB/L007959/1 (A.F. and M.V.). D.A.I.M. is 186

supported by the Medical Research Council (MRC) Career Development Award (MR/M009505/1) 187

and D.G. by a Swiss National Science Foundation Postdoc Mobility Fellowship (P300PA_167703). 188

189

References and recommended reading 190

Papers of particular interest, published within the period of review, have been highlighted as: 191

• of special interest 192

•• of outstanding interest 193

1. Gellatly SL, Hancock RE: Pseudomonas aeruginosa: new insights into pathogenesis and host 194 defenses. Pathog Dis 2013, 67:159-173. 195

2. WHO: GLOBAL PRIORITY LIST OF ANTIBIOTIC-RESISTANT BACTERIA TO GUIDE 196 RESEARCH, DISCOVERY, AND DEVELOPMENT OF NEW ANTIBIOTICS. 2017. 197

3. Lee DG, Urbach JM, Wu G, Liberati NT, Feinbaum RL, Miyata S, Diggins LT, He J, Saucier M, 198 Deziel E, et al.: Genomic analysis reveals that Pseudomonas aeruginosa virulence is 199 combinatorial. Genome Biol 2006, 7:R90. 200

4. Huber P, Basso P, Reboud E, Attrée I: Pseudomonas aeruginosa renews its virulence factors. 201 Environmental Microbiology Reports 2016, 8:564-571. 202

5. Klockgether J, Tummler B: Recent advances in understanding Pseudomonas aeruginosa as a 203 pathogen. F1000Res 2017, 6:1261. 204

6. Furukawa S, Kuchma SL, O'Toole GA: Keeping their options open: acute versus persistent 205 infections. J Bacteriol 2006, 188:1211-1217. 206

7. Berube BJ, Rangel SM, Hauser AR: Pseudomonas aeruginosa: breaking down barriers. Current 207 genetics 2016, 62:109-113. 208

8. Lund-Palau H, Turnbull AR, Bush A, Bardin E, Cameron L, Soren O, Wierre-Gore N, Alton EW, 209 Bundy JG, Connett G, et al.: Pseudomonas aeruginosa infection in cystic fibrosis: 210 pathophysiological mechanisms and therapeutic approaches. Expert Rev Respir Med 211 2016, 10:685-697. 212

9. Lam J, Chan R, Lam K, Costerton JW: Production of mucoid microcolonies by Pseudomonas 213 aeruginosa within infected lungs in cystic fibrosis. Infect Immun 1980, 28:546-556. 214

10. Bragonzi A, Paroni M, Nonis A, Cramer N, Montanari S, Rejman J, Di Serio C, Doring G, 215 Tummler B: Pseudomonas aeruginosa microevolution during cystic fibrosis lung infection 216 establishes clones with adapted virulence. Am J Respir Crit Care Med 2009, 180:138-145. 217

11. Goodman AL, Kulasekara B, Rietsch A, Boyd D, Smith RS, Lory S: A signaling network 218 reciprocally regulates genes associated with acute infection and chronic persistence in 219 Pseudomonas aeruginosa. Dev Cell 2004, 7:745-754. 220

12. Mikkelsen H, Sivaneson M, Filloux A: Key two-component regulatory systems that control 221 biofilm formation in Pseudomonas aeruginosa. Environ Microbiol 2011, 13:1666-1681. 222

13. Mulcahy H, O'Callaghan J, O'Grady EP, Adams C, O'Gara F: The posttranscriptional regulator 223 RsmA plays a role in the interaction between Pseudomonas aeruginosa and human 224

airway epithelial cells by positively regulating the type III secretion system. Infect Immun 225 2006, 74:3012-3015. 226

••14. LeRoux M, Kirkpatrick RL, Montauti EI, Tran BQ, Peterson SB, Harding BN, Whitney JC, 227 Russell AB, Traxler B, Goo YA, et al.: Kin cell lysis is a danger signal that activates 228 antibacterial pathways of Pseudomonas aeruginosa. Elife 2015, 4. 229 This study shows how P. aeruginosa detects ecological competition by sensing danger cues 230

released by lysed kin cells. These cues are detected by a dedicated by the Gac/Rsm cascade which 231 regulon can be read as a danger response program. 232 ••15. Broder UN, Jaeger T, Jenal U: LadS is a calcium-responsive kinase that induces acute-to-233

chronic virulence switch in Pseudomonas aeruginosa. Nat Microbiol 2016, 2:16184. 234 By combining genetics and biochemical data, this study demonstrates that calcium induces in 235

P. aeruginosa the activation of the Gac/Rsm regulatory cascade via the LadS sensor. The authors 236 show that the capacity to sense calcium is retained in some P. aeruginosa clinical isolates and they 237 propose calcium serves as a signal in host–pathogen interaction. 238 16. Fabrega A, Vila J: Salmonella enterica serovar Typhimurium skills to succeed in the host: 239

virulence and regulation. Clin Microbiol Rev 2013, 26:308-341. 240 17. Eisenreich W, Heuner K: The life stage-specific pathometabolism of Legionella pneumophila. 241

FEBS Lett 2016, 590:3868-3886. 242 18. Whiteley M, Bangera MG, Bumgarner RE, Parsek MR, Teitzel GM, Lory S, Greenberg EP: Gene 243

expression in Pseudomonas aeruginosa biofilms. Nature 2001, 413:860-864. 244 19. Dotsch A, Eckweiler D, Schniederjans M, Zimmermann A, Jensen V, Scharfe M, Geffers R, 245

Haussler S: The Pseudomonas aeruginosa transcriptome in planktonic cultures and static 246 biofilms using RNA sequencing. PLoS One 2012, 7:e31092. 247

20. Mikkelsen H, Bond NJ, Skindersoe ME, Givskov M, Lilley KS, Welch M: Biofilms and type III 248 secretion are not mutually exclusive in Pseudomonas aeruginosa. Microbiology 2009, 249 155:687-698. 250

21. Hentzer M, Eberl L, Givskov M: Transcriptome analysis of Pseudomonas aeruginosa biofilm 251 development: anaerobic respiration and iron limitation. Biofilms 2005, 2:37-61. 252

22. Monds RD, O'Toole GA: The developmental model of microbial biofilms: ten years of a 253 paradigm up for review. Trends Microbiol 2009, 17:73-87. 254

23. Chua SL, Liu Y, Yam JK, Chen Y, Vejborg RM, Tan BG, Kjelleberg S, Tolker-Nielsen T, 255 Givskov M, Yang L: Dispersed cells represent a distinct stage in the transition from 256 bacterial biofilm to planktonic lifestyles. Nat Commun 2014, 5:4462. 257

24. Siryaporn A, Kuchma SL, O'Toole GA, Gitai Z: Surface attachment induces Pseudomonas 258 aeruginosa virulence. Proc Natl Acad Sci U S A 2014, 111:16860-16865. 259

25. Oliver A, Mulet X, Lopez-Causape C, Juan C: The increasing threat of Pseudomonas 260 aeruginosa high-risk clones. Drug Resist Updat 2015, 21-22:41-59. 261

26. Turner KH, Everett J, Trivedi U, Rumbaugh KP, Whiteley M: Requirements for Pseudomonas 262 aeruginosa acute burn and chronic surgical wound infection. PLoS Genet 2014, 263 10:e1004518. 264

27. Schaber JA, Triffo WJ, Suh SJ, Oliver JW, Hastert MC, Griswold JA, Auer M, Hamood AN, 265 Rumbaugh KP: Pseudomonas aeruginosa forms biofilms in acute infection independent of 266 cell-to-cell signaling. Infect Immun 2007, 75:3715-3721. 267

28. Damron FH, Oglesby-Sherrouse AG, Wilks A, Barbier M: Dual-seq transcriptomics reveals the 268 battle for iron during Pseudomonas aeruginosa acute murine pneumonia. Sci Rep 2016, 269 6:39172. 270

29. Arbel-Goren R, Tal A, Stavans J: Phenotypic noise: effects of post-transcriptional regulatory 271 processes affecting mRNA. Wiley Interdiscip Rev RNA 2014, 5:197-207. 272

30. Vakulskas CA, Potts AH, Babitzke P, Ahmer BM, Romeo T: Regulation of bacterial virulence 273 by Csr (Rsm) systems. Microbiol Mol Biol Rev 2015, 79:193-224. 274

31. Schulmeyer KH, Diaz MR, Bair TB, Sanders W, Gode CJ, Laederach A, Wolfgang MC, Yahr TL: 275 Primary and Secondary Sequence Structure Requirements for Recognition and 276 Discrimination of Target RNAs by Pseudomonas aeruginosa RsmA and RsmF. J 277 Bacteriol 2016, 198:2458-2469. 278

32. Moscoso JA, Jaeger T, Valentini M, Hui K, Jenal U, Filloux A: The diguanylate cyclase SadC is 279 a central player in Gac/Rsm-mediated biofilm formation in Pseudomonas aeruginosa. J 280 Bacteriol 2014, 196:4081-4088. 281

33. Valentini M, Laventie BJ, Moscoso J, Jenal U, Filloux A: The diguanylate cyclase HsbD 282 intersects with the HptB regulatory cascade to control Pseudomonas aeruginosa biofilm 283 and motility. PLoS Genet 2016, 12:e1006354. 284

•34. Miller CL, Romero M, Karna SL, Chen T, Heeb S, Leung KP: RsmW, Pseudomonas 285 aeruginosa small non-coding RsmA-binding RNA upregulated in biofilm versus 286 planktonic growth conditions. BMC Microbiol 2016, 16:155. 287 This study describes a new RsmA-binding sRNA, named RsmW. RsmW levels increase in P. 288

aeruginosa in nutrient-limiting conditions, biofilms, and at higher temperatures, but is not 289 transcriptionally activated by GacA. RsmW indicates the presence of a new uncharacterized pathway 290 that regulates RsmA activity under specific conditions. 291 35. Petrova OE, Sauer K: A novel signaling network essential for regulating Pseudomonas 292

aeruginosa biofilm development. PLoS Pathog 2009, 5:e1000668. 293 36. Francis VI, Stevenson EC, Porter SL: Two-component systems required for virulence in 294

Pseudomonas aeruginosa. FEMS Microbiol Lett 2017, 364. 295 37. Petrova OE, Sauer K: The novel two-component regulatory system BfiSR regulates biofilm 296

development by controlling the small RNA rsmZ through CafA. J Bacteriol 2010, 297 192:5275-5288. 298

38. van Gestel J, Vlamakis H, Kolter R: Division of Labor in Biofilms: the Ecology of Cell 299 Differentiation. Microbiol Spectr 2015, 3:MB-0002-2014. 300

39. Lenz AP, Williamson KS, Pitts B, Stewart PS, Franklin MJ: Localized gene expression in 301 Pseudomonas aeruginosa biofilms. Appl Environ Microbiol 2008, 74:4463-4471. 302

40. LeRoux M, Peterson SB, Mougous JD: Bacterial danger sensing. J Mol Biol 2015, 427:3744-303 3753. 304

41. West SA, Griffin AS, Gardner A, Diggle SP: Social evolution theory for microorganisms. Nat 305 Rev Microbiol 2006, 4:597-607. 306

42. Lapouge K, Schubert M, Allain FH, Haas D: Gac/Rsm signal transduction pathway of gamma-307 proteobacteria: from RNA recognition to regulation of social behaviour. Mol Microbiol 308 2008, 67:241-253. 309

43. Silby MW, Winstanley C, Godfrey SA, Levy SB, Jackson RW: Pseudomonas genomes: diverse 310 and adaptable. FEMS Microbiol Rev 2011, 35:652-680. 311

44. Alonso A, Rojo F, Martinez JL: Environmental and clinical isolates of Pseudomonas 312 aeruginosa show pathogenic and biodegradative properties irrespective of their origin. 313 Environ Microbiol 1999, 1:421-430. 314

45. Brown SP, Cornforth DM, Mideo N: Evolution of virulence in opportunistic pathogens: 315 generalism, plasticity, and control. Trends Microbiol 2012, 20:336-342. 316

46. Collier J: Cell cycle control in Alphaproteobacteria. Curr Opin Microbiol 2016, 30:107-113. 317 47. Pechter KB, Gallagher L, Pyles H, Manoil CS, Harwood CS: Essential Genome of the 318

Metabolically Versatile Alphaproteobacterium Rhodopseudomonas palustris. J Bacteriol 319 2015, 198:867-876. 320

••48. Yan J, Nadell CD, Bassler BL: Environmental fluctuation governs selection for plasticity in 321 biofilm production. ISME J 2017, 11:1569-1577. 322 By performing competition experiment with Vibrio cholerae strains having flexible or rigid 323

biofilm production strategies, the authors show that lifestyle flexibility is, as evolutionary strategy, 324 incompatible with a temporally stable environment. This study provides a direct experimental 325 validation that fluctuating environmental conditions favour lifestyle flexibility in bacteria. 326 •49. Marvig RL, Sommer LM, Jelsbak L, Molin S, Johansen HK: Evolutionary insight from whole-327

genome sequencing of Pseudomonas aeruginosa from cystic fibrosis patients. Future 328 Microbiol 2015, 10:599-611. 329 This is a comprehensive review describing recent findings on P. aeruginosa genetic 330

adaptation to cystic fibrosis lungs. 331

50. Folkesson A, Jelsbak L, Yang L, Johansen HK, Ciofu O, Hoiby N, Molin S: Adaptation of 332 Pseudomonas aeruginosa to the cystic fibrosis airway: an evolutionary perspective. Nat 333 Rev Microbiol 2012, 10:841-851. 334

51. Levin BR, Bull JJ: Short-sighted evolution and the virulence of pathogenic microorganisms. 335 Trends Microbiol 1994, 2:76-81. 336

337

Figures and Boxes 338

339

340

341

Box1: The Gac/Rsm cascade 342

The Gac/Rsm global regulatory network comprises several regulatory components and signal 343

transduction pathways, converging on the post-transcriptional regulator RsmA (associated figure). 344

RsmA belongs to the CsrA/Rsm family of RNA-binding proteins which compete with ribosomes for 345

binding to the 5’ untranslated regions of mRNA targets, and possibly affect their stability [32]. 346

Therefore, the main action of RsmA is to repress the translation of target genes, although it has been 347

reported that sometimes it also exerts a positive control on gene expression, either directly or 348

indirectly. In addition to RsmA, the backbone of the Gac/Rsm signalling pathway is composed of the 349

GacS/GacA two-component system, which in P. aeruginosa induces the transcription of the genes 350

rsmY and rsmZ encoding small RNAs. The latter, are regulatory RNAs which have multiple RsmA-351

binding motifs (GGA motifs), exposed in stem-loops. Upon GacA activation, the small RNAs, 352

abundantly transcribed, sequester RsmA and relieve the target mRNAs from its control. The type of 353

interaction of RsmA with the mRNAs and the resulting levels of RsmY/Z determine the output of the 354

regulatory pathway. The regulatory action of proteins like the sensor kinases RetS, LadS, PA1611 and 355

SagS, the BfiS/R two-component system and the HptB phosphotransfer protein also feed into the core 356

of the Gac/Rsm pathway (grey shaded area) [36]. Finally, additional regulators, like the third RsmA-357

binding small RNA, RsmW, or the RsmN/F RNA-binding protein, have a supportive role in the 358

absence of backbone components [30]. In the box associated figure, the main Gac/Rsm regulatory 359

components are illustrated toghether with their regulatory action ( activation, ‒| repression). Blue is 360

used to indicate the components which, when absent, cause an impairement in biofilm formation, 361

while deletion of genes encoding elements in red lead to hyperbiofilm forming strains. 362

363

364

Figure 1: Schematic representation of the lifestyle switch model in P. aeruginosa. The levels of the 365

small regulatory RNAs, RsmY/Z, are a major determinant for this transition. When the levels of these 366

regulators are low (blue, left), P. aeruginosa cells have a planktonic lifestyle, causing acute infection. 367

Motility, through the expression of type IV pili and flagella, along with the presence of LPS and the 368

release of diffusible virulence factors or T3SS effectors, are crucial in this lifestyle. When the levels 369

of RsmY/Z are high (red/right), the bacterium adopts a sessile lifestyle, causing chronic infection. The 370

extracellular matrix, which encapsulates large numbers of cells (biofilm) and allows productive access 371

to extracellular products and “common goods” (like siderophores, proteases, elastase, rhamnolipids 372

and HCN), and the presence of the T6SS feature prominently in this lifestyle. During the switch from 373

planktonic to biofilm lifestyle, the LPS is known to undergo reversible structure modifications; this 374

represented by the presence of the classical form of the LPS on the left side of the figure (planktonic 375

lifestyle) which is absent on the right side of the figure (biofilm lifestyle). 376

377

378

Figure 2: Schematic representation of biofilm development and of the role of RsmY/Z small 379

regulatory RNAs in modulating this process. The stages of biofilm formation are (i) planktonic, (ii) 380

attachment, (iii) microcolony formation, (iv) macrocolony formation and (v) dispersal. The 381

“reversible attachment” phase indicates weak attachment to the surface while the “irreversible 382

attachment” phase involves mechanisms which allow the bacterium to acquire a stable association 383

with the surface (e.g. use of pili/flagella). Increase of RsmY/Z sRNAs in planktonic cells is cell-384

density dependent and results in increased surface attachment and sessility [11]. The reduction of 385

RsmZ abundance (through BfiSR-CafA) is necessary for biofilm development progression during the 386

transition from reversible to irreversible attachment [35]. In mature biofilms RsmY/Z levels are 387

elevated compared to planktonic cells [19]. Finally, when cells disperse from the biofilm, the 388

expression of RsmY/Z is down-regulated compared to both biofilm and exponentially-growing 389

planktonic cells [23]. The proposed schematic model provides a reference framework for 390

understanding the phenotype plasticity within a lifestyle, and its validity is currently under 391

investigation. 392