RpoS Mutation Leads to Prolonged Biolm Mode of Growth and a Higher Fitness in Pseudomonas Aeruginosa Biolms

Pseudomonas aeruginosa is a notorious opportunistic pathogen causing various biolm-related infections. Biolm formation is a unique microbial strategy that allows P. aeruginosa to survive adverse conditions such as antibiotic treatment and human immune responses. In this study, we experimentally evolved P. aeruginosa PAO1 bio ﬁ lms for cyclic treatment in the presence of high dose of imipenem, and enriched hyperbiolm mutants within six cycles in two independent lineages. The competition assay showed the evolved hyperbiolm mutants can outcompete the ancestral strain within biolm by prolonging the biolm mode of growth but not in planktonic cultures. Whole-genome sequencing analysis revealed the hyperbiolm phenotype is caused by point mutations in rpoS gene in all independently evolved mutants and the same mutation was found in P. aeruginosa clinical isolates. We further showed that mutation in rpoS enhanced biolm formation by prolonging the biolm mode of growth and elevating the intracellular c-di-GMP level. Mutation in rpoS increased pyocyanin production and virulence in both P. aeruginosa laboratory strains and clinical isolates. Here, our study revealed that antibiotic treatment of biolm-related P. aeruginosa infections might induce a hyperbiolm phenotype via rpoS mutation, which might partially explain antimicrobial treatment failure of many P. aeruginosa biolm-related infections.


Introduction
Microbial cells undergo rapid evolution under stress to adapt to the environment in nature [1] or inside the host [2]. Certain key mutations on genome will greatly help bacterial populations to gain a competitive advantage in adverse environment [3]. To identify these adaptive traits of microbes, experimental evolution experiments are usually conducted to mimic these diverse environmental conditions for accelerating the emerge of well-tted variants [4]. Employment of next-generation sequencing approaches will facilitate the identi cation of the mutations of experimental adapted bacterial variants and elucidation of the underlying molecular mechanism of the evolved traits [5]. So far, adaptive experimental evolution has been applied to reveal the molecular basis of drug resistance [6], persistence [7], bio lm formation [8] and etc. An important question for experimental evolution is the relevance of laboratory observation to the evolution in natural conditions. For this point, studies have indicated the mutation derived phenotype which observed in laboratory evolution could be found in clinical isolates [9][10][11][12].
Bacterial pathogens are able to form bio lms on both biotic and abiotic surfaces, causing many hospitalacquired and recurrent infections. Bio lms are densely-grown microbial cells embedded in self-secreted Page 3/20 hydrated matrix consisting of polysaccharides, proteins, extracellular DNA and lipids [13]. Previous studies have shown that the bio lm-grown bacteria have distinct phenotypes from planktonic cultures, including gene expression [14,15] and increased antibiotic resistance [16]. The resistance to antimicrobial agents by bio lm cells can be increased up to 1000-fold compared to planktonic cells [17].
Investigating evolution traits of bio lm cells against antibiotics might provide knowledge about bacterial adaptation during chronic infections. P. aeruginosa is a notorious opportunistic pathogen, which causes a variety of infections, including wounds, urinary tract and respiratory infections, and is the leading cause of morbidity and death for patients with cystic brosis (CF) [18]. Infections caused by P. aeruginosa can be very di cult to treat due to its intrinsic resistance to a variety of antibiotics and tends to form bio lms at the site of infection [19,20]. As an important nosocomial pathogen, P. aeruginosa bio lms have been found on various surfaces of indwelling medical devices, including urinary catheters, bone plates, ventricular assist device drivelines and pacemakers [21]. Biofilms provide P. aeruginosa an enormous advantage in clinical infection by protecting bio lm cells from the immune system [22] and tolerance to antimicrobial agents [23,24]. Persisters are subpopulation of isogenic bacteria that tolerance to antibiotics [25], the persister cells were dormant in biofilms and significantly contributes to P. aeruginosa biofilm recalcitrance after the cessation of antibiotic therapy [26]. Consideration the issue of drug tolerance and recalcitrance of P. aeruginosa biofilm-related infections, urgent, novel antibiofilm therapeutics are needed.
The long-term use of antibiotics in the treatment of P. aeruginosa infections in cystic brosis patients is well known to drive emergence of diversi ed drug-resistant variants. Clinical isolates of P. aeruginosa has shown distinct bio lm formation capacity and many of them were strong bio lm producers [27]. However, current adaptive experimental evolution studies of P. aeruginosa are mainly focusing on planktonic cultures [11,28,29]. In this study, we established a clinically relevant model to investigate evolution traits of imipenem-treated bio lm cultures of P. aeruginosa. The set-up of the P. aeruginosa bio lm experimental evolution model in this study is presented in Figure 1A.
We show here that cyclic exposure of P. aeruginosa bio lms to high concentration of imipenem led to emergence of variants with a hyperbio lm phenotype. After 3 and 6 cycles of treatment, the colony forming units (CFU) of bio lm on glass beads were found to increase 1000-fold. The competition assay showed that the evolved hyperbio lm variants can outcompete ancestral strain within bio lms but not in planktonic cultures. Genome sequencing analysis revealed that the hyperbio lm phenotype is caused by single-point mutations in the sigma factor RpoS, which elevated the intracellular c-di-GMP level. Importantly, the mutations on rpoS identified in the in vitro experimental bio lm model occurred in P. aeruginosa clinic isolates with a hyperbio lm phenotype at a substantial rate. Overall, our data show that under imipenem treatment, mutations in rpoS could be selected in P. aeruginosa and subsequently lead to enhanced bio lm formation. Thus, caution should be taken when antimicrobial treatment of P. aeruginosa infections fails in clinical settings, rpoS mutation in cultured isolates may be the root cause and needs to be investigated.

Results
Experimental bio lm evolution selects for P. aeruginosa hyperbio lm mutants In order to examine the evolutionary traits of P. aeruginosa bio lms under antibiotic stress condition, we exposed bio lms of P. aeruginosa PAO1 to different concentrations of imipenem (40,80 and 160 μg/mL), which is 10, 20 and 40 times of the minimum inhibitory concentration (MIC), in a cyclic manner ( Figure  1). Imipenem, a widely used last resort antibiotic, has been chosen as the selective pressure for experimental bio lm evolution owing to its' commonly prescribed for treatment of P. aeruginosa infections [30].
The bio lms of six independent lineages in different concentrations of imipenem treated group initiated from a common ancestor PAO1 strain were formed on the surface of 5 mm glass beads [31] and treated with imipenem for 24 h. Survivor cells on beads were quantitated by CFU counts. Bio lm survivors were collected and reinoculated in fresh LB medium for the 2nd cycle ( Figure 1A). At the rst cycle, the CFU counts on bead of each lineage was between 6.30-6.90 log 10 . After 6 cycles, no hyperbio lm variants observed in 10× and 20× MIC ( Figure S1A and B) treated groups, and two lineages were accumulated hyperbio lm variants in 40× MIC treated group ( Figure S1C). The CFU counts within lineage W1 and W6 bio lms on bead reached 9.82 and 10.01 log 10 in 40× MIC treated group ( Figure 1B). Next, the bio lm formation ability of ancestral, C6W1, C3W6 and C6W6 population were further con rmed by the crystal violet (CV) bio lm assay. Similarly, the CV method revealed that bio lms formed by C6W1, C3W6 and C6W6 population were between 2-and 3-fold higher than the ancestral population ( Figures 1C). In order to track when the hyperbio lm variants have emerged within lineage W1 and W6 population, we picked 6 colonies form each cycle in random and measured the CFU of bio lms on bead. We found that the hyperbio lm variants of lineage W1 and W6 appeared since cycle 5 and cycle 2 and enriched at cycle 6 ( Figure 1D) and cycle 3 ( Figure 1E), respectively. These results indicate that the hyperbio lm variants only accumulated upon higher concentration imipenem treatment, rather than lower concentration of imipenem. The different appearance timing of hyperbio lm variants in two independent linages -possibly as a result of varying evolutionary rate [32]. More lineages might accumulate hyperbio lm variants in 40× MIC treated group if we increase the treatment cycles. Based on these results, we chose colonies C6W1C, C3W6F and C6W6F for further study.
Point mutations in rpoSlead to hyperbio lm phenotype of P. aeruginosa In order to elucidate the genetic mechanisms underlying the hyperbio lm phenotype, we sequenced the C6W1 and C6W6 population, and choose C6W5 population as the negative control. Through comparative genomic analysis, we identi ed only SNPs in one gene, PA3622 (encodes sigma factor RpoS), which was mutated in C6W1 and C6W6 population but not in C6W5 population when compared to the ancestral strain. The sigma factor RpoS is well known as a master regulator that controls the expression of genes involved in stress response and virulence factors production in P. aeruginosa [33,34]. A previous study used transcript profiling to identify that 772 genes were regulated by RpoS in stationary phase and it affects expression of more than 40% quorum-sensing controlled genes [35].
We next re-sequenced of rpoS in C6W1C, C3W6F and C6W6F, and identi ed the nonsynonymous mutations in RpoS of C6W1C (P251L), C3W6F (Q266stop) and C6W6F (Q266stop). To further confirm the causality of rpoS mutations for the hyperbio lm phenotype, we constructed a de novo mutant allele with a SNP on the ancestor PAO1 genome to yielded RpoS P251L and RpoS Q266stop mutant strains. We found that single point mutation in rpoS could produce the hyperbio lm phenotype (Figure 2A). We also tested the bio lm formation ability of ΔrpoS strain and con rmed that knockout rpoS in P. aeruginosa PAO1 indeed increased the bio lm formation ( Figure 3A). Complementing the mutation strains with wild type rpoS reverted the hyperbio lm phenotype to the wild-type level ( Figure 2A).
Protein domain analysis showed that RpoS consist of 4 regions, and region 4 contains a DNA binding domain. P251L located on the end of region 3, knockout this region increased the bio lm formation. Q266stop mutation leads to RpoS lacking region 4, knockout region 4 has the same phenotype of Q266stop mutation. Moreover, we have constructed region 3 and 4, region 2, 3 and 4 and region 1 deletion strains, all of those region deletion mutants produced the hyperbio lm phenotype ( Figure 2B). Together, these results provide definitive evidence that the SNPs identified are necessary and sufficient to cause hyperbio lm phenotype in P. aeruginosa. Moreover, P251L and Q266stop mutations are very likely lead to the inactivation of RpoS.

Hyperbio lm variants outcompete the ancestral strain during bio lm competitions
The convergent emergement of P. aeruginosa hyperbio lm variants from independent lineages suggested a competitive advantage for these variants over the ancestor. We have previously showed that P. aeruginosa cells did not share its EPS with its neighboring cells [36] and thus we hypothesized that the evolved hyperbio lm variants only gain advantage to the ancestor strain under bio lm growth condition. We then tested the competition of hyperbio lm variants with the ancestral strain in both planktonic cultures and bio lms. The uorescently tagged hyperbio lm variants (tagged with mCherry) were mixed with ancestral strain (tagged with GFP) in different ratio and inoculated into the bead-containing 24 well microplate. After 24 h cultivation, the planktonic and bio lm cells were analyzed with ow cytometry. We found that, the planktonic cells of C6W1C was slightly higher than ancestral strain when inoculated at the same ratio ( Figure 3A), while no difference was found between C3W6F and C6W6F with ancestral strain when inoculated at the same ratio ( Figure 3B and C). Interestingly, the competitions in bio lms confirmed that the hyperbio lm variants have a significant and predominant selective advantage against the ancestral strain ( Figure 3A, B and C). Next, we increased the inoculation ratio of ancestor and hyperbio lm variants to 5:1, the proportion of ancestor within bio lm was still much less than hyperbio lm variants ( Figure 3D, E and F). These results indicated that the hyperbio lm variants increased the competition only in bio lms and this phenotype is not related to the change of growth rate, drug resistance and tolerance. Moreover, the competition advantage of hyperbio lm variants is growth model speci c and did not occur in planktonic culture.
Previous studies showed that acquired mutations conferring bene cial traits such as antibiotic resistance will dominate when exposing bio lm bacteria to high concentrations of antibiotic [37]. Therefore, the enrichment of the hyperbio lm variants could have been achieved by accumulation of mutations that conferred resistance to imipenem. We found, however, that the MIC of imipenem for colonies isolated from the evolved lines (C6W1C, C3W6F and C6W6F) was indistinguishable from that for their ancestor ( Figure 3G). One of the most straightforward ways to gain a competitive advantage is increasing the growth rate. To test this point, we measured the growth rates of the evolved hyperbio lm variants and the ancestor in LB medium. Whereas, there is no signi cant difference between the hyperbio lm variants and ancestor ( Figure 3H). Next, we measured the bio lm growth curve of PAO1, ΔrpoS and the complementation strain ΔrpoS/p-rpoS, the CFU of planktonic cultures were measured at the same time.
We found that, the CFU of planktonic culture of all three strains showed no signi cant difference ( Figure  3I). For the bio lm growth, the ΔrpoS strain indeed produce more bio lm than PAO1 after 6 h incubation and developed 2 log of bio lm CFU after 24 h incubation ( Figure 3I). These results indicated that it is not the resistance level and planktonic growth rate select ΔrpoS mutant; whereas ΔrpoS has a prolonged bio lm mode of growth which eventually leads to occupation of the bio lm community.

Mutations in rpoS lead to an elevated intracellular c-di-GMP levels
Quorum-sensing (QS) [38] and c-di-GMP [39] have been well documented to play important roles in P. aeruginosa bio lm formation. To assess whether quorum-sensing and c-di-GMP levels were elevated, we introduced the quorum-sensing and c-di-GMP reporter systems [40][41][42][43] into the variants isolated from bio lm evolution experiments and ancestral strain to determine the relative level of corresponding signaling pathways. We found that the fluorescent signal of P lasB -gfp and P rhlA -gfp in ancestral strain were higher than that of hyperbio lm variants ( Figure S2A and B), while there were no differences in fluorescent signal of P pqsA -gfp between the ancestral and hyperbio lm variants ( Figure S2C). For the fluorescent signal of P cdrA -gfp, the hyperbio lm variants showed 2-fold higher in expression level than the ancestral strain, indicating that the hyperbio lm variants might have elevated intracellular c-di-GMP levels ( Figure S2D). We further showed that the P cdrA -gfp expression level were increased in RpoS P251L , RpoS Q266stop and ΔrpoS strain compared to the PAO1 wild-type strain ( Figure 4A). The second messenger c-di-GMP is a key regulator of P. aeruginosa bio lm formation, which is synthesized from two GTP molecules by diguanylate cyclases (DGC) and is degraded into 5'-phosphoguanylyl-(3'-5') guanosine (pGpG) and/or GMP by phosphodiesterases (PDE) [39]. Till now, 43 DGC and PDE proteins have been identi ed in P. aeruginosa [44].
To investigate the mechanism underlying rpoS point mutation-induced increasing in intracellular c-di-GMP content, we performed transcriptomic analysis of PAO1, RpoS P251L and RpoS Q266stop strains using RNA-sequencing. Samples were collected after 8.5 h culture owing to the P cdrA -gfp fluorescent intensity ( Figure S3) between mutants and wild type PAO1 strain have the biggest difference at this time point. We found that, 15 DGC and PDE proteins were upregulated at least 2-fold in both RpoS P251L and RpoS Q266stop strains compared to PAO1 (Table 1). This result indicated that the c-di-GMP metabolism in rpoS mutant strains were more active than PAO1.
RpoS regulates the expression of small regulatory RNAs rsmY and rsmZ in Legionella pneumophila [45]. Moreover, rsmY/Z participate in the regulation of c-di-GMP production in P. aeruginosa, the c-di-GMP levels were strongly reduced in the rsmY/Z double deletion mutant [46]. Our transcriptomic analysis showed that the expression of rsmY and rsmZ was increased 3.84 and 5.04-fold in RpoS P251L compared to the PAO1 wild-type, respectively. Next, we measured the expression of rsmY/Z in PAO1, RpoS P251L , RpoS Q266stop and ΔrpoS strains using reporter fusions [14]. We found the rsmY/Z expressions were increased in RpoS P251L , RpoS Q266stop and ΔrpoS ( Figure 4B and C), which is consistent with the increased level of c-di-GMP of these mutants. These results showed that the mutation of rpoS has led to the increase in rsmY/Z expression and intracellular c-di-GMP content in P. aeruginosa.

rpoS mutation associated hyperbio lm phenotype in clinical isolates
Our experimental bio lm evolution data has revealed that single-nucleotide mutations on rpoS confer P. aeruginosa hyperbio lm phenotype and produce a pronounced competitive advantage within the bio lm microenvironment. In order to analyze the preference of rpoS mutation, we downloaded 4000 sequences of rpoS from pseudomonas genome database (www.pseudomonas.com). Through comparative analysis, we have identi ed 241 non-synonymous mutations (6.03% of total sequence), 8 insertion or deletion mutations (0.2% of total sequence) and 5 stop coding mutations (0.13% of total sequence) compared to the PAO1 wild-type strain. Among those mutations, 123 mutations were located on the inter-region of rpoS and 131 mutations were within 4 regions ( Figure 5A). Moreover, we have identi ed 2 sequences harbored RpoS P251L mutation. We also analyzed the top 5 mutation sites among 4000 sequences, and found L268Q was the top one with 71 sequences ( Figure 5B).
Since imipenem has been used for clinical treatment of P. aeruginosa infection, we wondered whether rpoS mutation caused hyperbio lm strains exist in clinical isolates. Therefore, we examined the bio lm formation capacity of 288 clinical P. aeruginosa isolates obtained from the patients with culture con rmed P. aeruginosa infections (Table S1). Through quantitative analysis of bio lm formation by measuring crystal violet staining at OD 550 nm and total bacterial growth at OD 600 nm to exclude growth variation, we identi ed 29 hyperbio lm isolates (10.07 % of total isolates) in this collection ( Figure 5C). Next, we target sequenced rpoS of 29 hyperbio lm isolates and con rmed that #16 isolate harbored nonsynonymous mutation in rpoS. Interestingly, #16 isolate, which is isolated form the peritoneal drainage uid, has the same mutation RpoS P251L as our experimental evolved variant C6W6F.

The evolved rpoS variants are hypervirulent
Pyocyanin production is one of the major virulence factors of P. aeruginosa, and plays an important role in P. aeruginosa pathogenesis by causes oxidative stress to the host, induces apoptosis in neutrophils and inhibits phagocytosis of macrophages [47,48]. Previous studies showed that the pyocyanin production was increased in a rpoS-deletion mutant [34]. In order to test the impact of rpoS point mutation on our bio lm evolved variants on pyocyanin production, we compared the production of pyocyanin by P. aeruginosa PAO1 strain, RpoS P251L , RpoS Q266stop and ΔrpoS. As we expected, similar to the ΔrpoS mutant, the P. aeruginosa RpoS P251L and RpoS Q266stop produced higher amounts of pyocyanin than the wild-type PAO1 strain ( Figure 6A). This result suggests that point mutations accumulated in the rpoS gene in P. aeruginosa clinical isolates have similar effect as rpoS gene deletion on its physiology.
Next, we further assessed the impact of evolved rpoS point mutations on virulence using the macrophage cytotoxicity model [49]. The RAW264.7 macrophages were infected with P. aeruginosa PAO1, RpoS P251L , RpoS Q266stop and ΔrpoS, and the release of cytosolic lactate dehydrogenase (LDH) was determined. We found that macrophages infected with RpoS P251L , RpoS Q266stop and ΔrpoS released more LDH compared to P. aeruginosa PAO1 after 4 h infection (Fig 6B), suggesting mutation on rpoS can induce the cell death of macrophage. Altogether, there results suggest that mutations on rpoS can enhance the virulence in P. aeruginosa.

Discussion
Bio lms represent the predominant lifestyle for most microorganisms in nature. Understanding how microorganisms evolve in bio lms can reveal novel insights of adaptive evolution, especially under stress conditions. For example, small colony variants are enriched in P. aeruginosa bio lms after exposure to sodium dodecyl sulfate [50]. The P. aeruginosa hyperbio lm forming variants are often observed from patients who are suffering chronic infections such as CF [51]. We have previously using a planktonic experimental evolution model to demonstrate that oxidative stress drives the evolution of point mutations in the P. aeruginosawspF gene, which lead to increase in intracellular c-di-GMP content and exopolysaccharide synthesis [28]. Here, we performed bio lm experimental evolution to examine the adaptive evolution of P. aeruginosa bio lms during the treatment of imipenem (160 μg/mL), which case can be reached in some situations in clinical settings. The keratitis infection caused by P. aeruginosa usually form corneal biofilms [52]. In a case of the treatment of bacterial keratitis in patients, topical imipenem (50 mg/mL) has been selected as monotherapy for corneal infection [53]. In another study, 1-5 mg/mL imipenem has been used for bacterial keratitis treatment via topical administration [54] . Moreover, we showed that the hyperbio lm variants could be accumulated in bio lms after cycle antibiotic treatment and these variants were able to outcompete the ancestor strain shortly after appearance. Genome sequencing analysis revealed that the adapted P. aeruginosa hyperbio lm variants in different linages shared single-point mutations in the same gene, which encodes the sigma factor RpoS.
The rpoS gene has been previously well characterized in P. aeruginosa for its regulatory role on quorum sensing and virulence. A DNA microarray-based transcriptomic study showed that the expression of rpoS in P. aeruginosa bio lm cells was downregulated compared to the planktonic cells, inactivation of rpoS in P. aeruginosa PAO1 increased bio lm formation in ow-cell reactor [15]. In our study, we demonstrated that rpoS point mutations have similar impact to the rpoS deletion on bio lm formation in P. aeruginosa. Function domain analysis indicates that RpoS contain 4 regions, RpoS as a global regulator, the DNA binding domain was located on region 4. Proline residues are restricts to the rst four positions of an αhelix [55], which plays a special role in the stable of protein structure. RpoS P251L and RpoS Q266stop on P. aeruginosa genome have shown the same phenotype of hyperbio lm, pyocyanin production and virulence as ΔrpoS strain, which means the 251proline to leucine mutation in rpoS might results loss of the function on regulation. Moreover, we showed that mutation on rpoS increased the production of c-di-GMP and pyocyanin, both of those two molecules were play a very important role in bio lm formation [39] and virulence [56] in P. aeruginosa.
Surprisingly, the rpoS mutants were found to outcompete the wild-type PAO1 strain under bio lm mode of growth in a short time frame. The P. aeruginosa small colony variants (e.g. with wspF mutations) are also well known being evolved in P. aeruginosa bio lms, which have even higher bio lm formation capacity than the rpoS mutants. However, these small colony variants often have a lower planktonic growth rate than the P. aeruginosa wild-type strain, and thus can easily be outcompeted by the wild-type in the planktonic phase of growth [28]. Instead, our study showed that the P. aeruginosarpoS mutants are able to outcompete the wild-type PAO1 strain by prolonging the bio lm mode of growth while not compromising its planktonic growth rate. Within the microenvironment of bio lm, PAO1 wild-type and rpoS mutants live with identical niches and also have identical needs, thus they will compete for precisely the same resources. Our observation that rpoS mutants can outcompete the PAO1 could be explained by the competitive exclusion principle [57,58]. As a sigma factor, RpoS controls a wide range of genes under stationary phase of growth, which shared many characters with bio lm mode of growth, such as lack of nutrients and accumulation of waste products. Further studies should be carried to examining the regulatory roles of RpoS on P. aeruginosa physiology and virulence factors under bio lm mode of growth.

Conclusions
Our study showed that imipenem treatment drives rapid evolution of P. aeruginosarpoS de cient mutants within bio lms. We provided evidence that rpoS mutation not only increase P. aeruginosa virulence, but also enhance its intracellular c-di-GMP content. Importantly, the major obstacle for treatment of P. aeruginosa infection in clinical is the formation of bio lms. This study raises the possibility that some clinical P. aeruginosa strains with rpoS mutations could have a selective advantage during imipenem administration, which might have an impact on the antibiotic therapy against P. aeruginosa bio lmassociated infections.

Methods
For details see Supplementary Information.

Bio lm experimental evolution
The experimental evolution of P. aeruginosa PAO1 bio lm was carried out on glass beads formed bio lm [31]. Two autoclaved 5 mm glass beads (Merck KGaA, Darmstadt, Germany) were placed into each well of a 24-well microtiter plate (Nunc, Thermo-Fischer). A LB overnight culture of P. aeruginosa was diluted in LB to approx. 1x10 6 bacteria per mL and dispensed into the bead-containing 24 well microplate (1 mL per well). The microplate was then placed in a moisture box and incubated at 37 ℃ for 24 h at 100 rpm on an orbital shaker. After 24 h, the liquid culture was removed and beads were washed by 0.9 % NaCl for twice to remove loosely attached bacteria. Then transfer one bead into 2 mL microcentrifuge tube containing 1 mL 0.9% NaCl, subjected to 6× 10 s vortex and sonicated in an ultrasonic bath (Worldvicon, Shenzhen, China) at 40 kHz for 5 min. Bacterial suspensions were subsequently serially diluted in 0.9% NaCl before being drop-plated onto lysogeny broth agar plates (Difco). After 24 h of incubation at 37 ℃, the residual biofilm was quantified as CFU/bead. Another bead was transferring to a 24 well microplate contain 1 mL LB with 160 μg/mL imipenem. The microplate was then placed in a moisture box and incubated at 37 ℃ for 24 h without shaking. After 24 h treatment, this bead was washed by 0.9% NaCl for twice and transfer into 2 mL microcentrifuge tube containing 1 mL LB, after vortex and sonicated. 20 μL of bacterial suspensions were subsequently serially diluted in 0.9% NaCl before being drop-plated onto lysogeny broth agar plates (Difco), the rest bacterial suspensions were cultured at 37 ℃ for 24 h at 200 rpm. After 24 h of cultivation, 100 μL P. aeruginosa was diluted in LB to approx. 1x10 6 bacteria per mL and start a new cycle. The rest culture was glycerol stocked at -80 ℃. The CFU/bead increased over 100fold compared to the ancestral strain was de ned as hyperbio lm phenotype variants.

Bio lm competition assay
The bio lm competition assay was carried out on glass beads formed bio lm. The ancestor strain PAO1 and mutants were tagged with gfp and mcherry at the attB site to generate the strain PAO1 attB::gfp and mutant attB::mcherry as previously described [59]. Overnight cultures were adjusted OD 600 to 1.0, cells were mixed 1:1 or 1:5 and con rmed by ow cytometer analysis. The mixed bacteria were diluted in LB to approx.1x10 6 of per mL and dispensed into the bead-containing 24 well microplate (1 mL per well). The microplate was then placed in a moisture box and incubated at 37 ℃ for 24 h at 100 rpm on an orbital shaker. After 24 h treatment, the cells in planktonic and bio lm were analyzed by ow cytometer analysis.
DNA extraction, sequencing, and SNP analysis Genomic DNA of the ancestor and evolved bacterial populations were extracted form glycerol stocked by AxyPerp Bacterial Genomic DNA Miniprep Kit (Corning) and sequenced by Illumina NovaSeq platform. Illumina genomic reads of the isolates were analyzed by CLC Genomics Workbench 20 (Qiagen) using Resequencing analysis module with default parameters for single nucleotide polymorphism (SNP) with P. aeruginosa PAO1 as reference genome.
RNA extraction, sequencing, and transcriptomic analysis Samples were collected at the peak of P cdrA -gfp fluorescence intensity, RNA extraction was performed using the miRNeasy kit (Qiagen) according to the manufacturer's instructions. RNA samples were submitted to Guangdong Magigene Biotechnology Co.,Ltd. (Guangzhou, China) for ribosomal RNA depletion and sequencing. RNA samples were sequenced on an Illumina Hiseq Xten platform and 150 bp paired-end reads were generated.
The quality of raw sequence data was assessed using FastQC (Babraham Bioinformatics). RNA sequence analysis was done using "RNA-seq analysis' module in CLC genomics Workbench 20 (CLC Bio, Aarhus, Denmark) using P. aeruginosa PAO1 reference genome downloaded from NCBI database. Adaptor sequences were removed by adaptor trimming function in CLC. Differential gene expression was analyzed using DESeq2 package in R software.

Statistical analysis
Data are presented as mean ± standard deviation (SD). All other comparisons were made using a oneway analysis of variance (ANOVA) with Student's t test. Analyses were performed using GraphPad Prism v.7 (GraphPad Software). Statistical significance was determined using a P value of 0.05.  were grown on 5 mm glass bead. After 24 h cultivation, one bead was vortex and sonicated for CFU counts, another bead was transferred to a 24 well microplate and treated with 160 μg/mL imipenem.
After 24 h treatment, the surviving cells were grown overnight in fresh medium and start another cycle. (B) Evolution of bio lm bacteria exposed to imipenem resulted in a rapid increase in bio lm bacteria CFU on bead. (C) Crystal violet (CV) staining of bio lms formed by ancestral and hyperbio lm variant strains on PVC plate. Data are presented as the mean±s.d. of ve biological replicates. Significance was determined using a Student's t test: *P < 0.05, **P < 0.01 and ***P <0.001. (D and E) The time frame of emergence of hyperbio lm variants in linage W1 (D) and W6 (E). The bio lm formation by the different colonies was displayed with CFU of bio lm cells on 5 mm glass bead.

Figure 2
Hyperbio lm phenotype is caused by rpoS mutation. rpoS point mutation, full deletion (A) and regions deletion strains (B) were increased the bio lm formation. The bio lm formation by the indicated strains was displayed with CFU of bio lm cells on 5 mm glass bead. Data are presented as the mean±s.d. of four biological replicates. Significance was determined using a Student's t test: ***P <0.001. EV represents the empty vector pHERD20T in this assay.    Pyocyanin production and virulence are increased in rpoS mutants. The production of pyocyanin (A) and cytotoxicity effect against macrophage cells (B) of P. aeruginosa PAO1 wild-type, RpoSP251L, RpoSQ266stop and ΔrpoS. Data are presented as the mean±s.d. of four biological replicates.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. cellandbioscienceSupplementarymaterials.docx