1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Pseudomonas aeruginosa is an opportunistic pathogen ubiquitous to the natural environment but with the capability of moving to the host environment. Long-term infection of the airways of cystic fibrosis patients is associated with extensive genetic adaptation of P. aeruginosa, and we have studied cases of the initial stages of infection in order to characterize the early adaptive processes in the colonizing bacteria. A combination of global gene expression analysis and phenotypic characterization of longitudinal isolates from cystic fibrosis patients revealed well-known characteristics such as conversion to a mucoid phenotype by mucA mutation and increased antibiotic resistance by nfxB mutation. Additionally, upregulation of the atu operon leading to enhanced growth on leucine provides a possible example of metabolic optimization. A detailed investigation of the mucoid phenotype uncovered profound pleiotropic effects on gene expression including reduction of virulence factors and the Rhl quorum sensing system. Accordingly, mucoid isolates displayed a general reduction of virulence in the Caenorhabditis elegans infection model, altogether suggesting that the adaptive success of the mucoid variant extends beyond the benefits of alginate overproduction. In the overall perspective the global phenotype of the adapted variants appears to place them on paths in direction of fully adapted strains residing in long-term chronically infected patients.


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

The bacterium Pseudomonas aeruginosa is a microorganism that is usually found both in aquatic and terrestrial environments. Like other Pseudomonads it is a versatile species with an extended metabolic repertoire allowing it to occupy many different niches. Its genome, one of the largest in the bacterial world with close to 6000 genes, displays the versatility of the organism with its multiple metabolic pathway genes, and the extraordinary large number of regulatory genes (more than 500) suggests that P. aeruginosa has a great competitive capacity to move between different types of environments (Stover et al., 2000). The diversity of gene regulatory activities in the organism ensures adaptation to new environmental conditions. Another reason why P. aeruginosa has attracted considerable interest is connected to its capacity to cause infections in plants, animals and humans (Spencer, 1996; Wolfgang et al., 2003). These infections are associated with the genomic content of a large number of virulence genes, which are important elements in most cases of infections.

One of the best studied human infections caused by P. aeruginosa is airway infections in patients suffering from the genetic disorder cystic fibrosis (CF). Because of a severely reduced mechanical clearing of the airway mucus in these patients they acquire multiple infections of various bacteria and fungi, and a major cause of morbidity and mortality is chronic infection of P. aeruginosa in the lungs (Koch and Hoiby, 1993; Govan and Deretic, 1996). Most patients acquire P. aeruginosa early in their lives and initial colonization is typically caused by environmental strains displaying all the wild-type phenotypes associated with this species (Johansen and Hoiby, 1992; Burns et al., 2001; Koch, 2002). Because the majority of children in a given clinical setting carry unique genotypes ofP. aeruginosa it also seems clear that colonization in most cases is a separate event (Burns et al., 2001; Jelsbak et al., 2007). In many cases the original colonization proceeds into a chronic infection (or alternatively other clones may later establish chronic infections), without any possibility of eradication (Koch, 2002; Jelsbak et al., 2007). During infection the bacteria have to adapt to the new environment comprising the host immune response, antibiotics and a different substrate composition. It is commonly observed that the bacteria gradually loose the function of many genes normally associated with bacterial pathogenicity (Smith et al., 2006; Jelsbak et al., 2007); however, in a chronic infection scenario this does not per se imply a reduction in virulence but likely rather an alteration of virulence (Bragonzi et al., 2009).

Genetic and phenotypic investigations of P. aeruginosa isolates from chronically infected patients have shown that among the many specific traits developing in the bacteria during infection, the mucoid phenotype caused by excessive production of the extracellular polysaccharide alginate is occurring with high frequency (Pedersen, 1992; Martin et al., 1993; Govan and Deretic, 1996). Conversion to a mucoid phenotype is caused by mutations in regulatory genes, mainly in various positions of the mucA gene encoding the anti-sigma factor MucA (Boucher et al., 1997). Binding between MucA and AlgU, the corresponding sigma protein (homologous to the sigma E factor of Escherichia coli) prevents AlgU from acting as a necessary initiation factor required for transcription of the alg operon (encoding the enzymes required for alginate synthesis) (Govan and Deretic, 1996; Boucher et al., 1997; Rowen and Deretic, 2000). Mutations disabling the activity of MucA lead to activation of AlgU, and the resulting elevated transcription of the alg operon leads to alginate overproduction (Rowen and Deretic, 2000). In the CF airways overproduction of alginate has been associated with increased tolerance to the host immune defense (Cabral et al., 1987; Govan and Deretic, 1996; Mathee et al., 1999) and additionally described as a protective extracellular matrix in connection with biofilm development in the infected airways (Høiby et al., 2001; Yang et al., 2008). It is thus assumed that the major impact of the mucA mutations on the infection success of the bacteria is connected to the alginate production.

Pseudomonas aeruginosa in its wild-type configuration is naturally resistant to many antibiotics, and if confronted with antimicrobial agents it will in many cases be able to develop tolerance to increased levels of the compounds (Hancock, 1997; Hancock and Speert, 2000). This capacity to resist antimicrobial treatment is most likely an important reason why P. aeruginosa has become the dominant airway infection problem for CF patients as antibiotic therapy was introduced to these patients. Before antibiotic treatment patients died very young from infections of indigenous pathogens such as Staphylococcus aureus and Haemophilus influenza. The medical success leading to extended life spans of CF patients is coupled to antibiotic treatments with compounds towards which P. aeruginosa is not inherently resistant (Høiby et al., 2005; Hansen et al., 2008). Because antibiotics are used to combat the infections as soon as P. aeruginosa has been diagnosed, it means that persistent colonization and subsequent chronic infection will eventually be associated with development of resistance.

Some of the most common phenotypic traits reported for P. aeruginosa isolates from chronically infected CF patients (‘the chronic infection phenotype’) include slow growth, auxotrophy, alginate overproduction (mucoidy), antibiotic resistance and loss of virulence factors and motility (Mahenthiralingam et al., 1994; Barth and Pitt, 1996; Bagge et al., 2002; Lee et al., 2005; Smith et al., 2006; Jelsbak et al., 2007; Ciofu et al., 2008). Because these phenotypic characteristics have been observed in many different clinical settings it is highly likely that such a phenotypic profile is an end-point result of parallel evolution of the bacteria in many CF patients colonized by many different bacterial genotypes. The actual evolutionary trajectories towards this common phenotypic profile, however, could be quite different in different clinics, patients and colonizing genotypes. In three cases described here the colonizing bacteria are all at the beginning of their colonization history, they all constitute persistent colonization (present for thousands of generations), and none of them have at the present time reached the typical ‘chronic infection phenotype’. The question addressed here is: in each of these three cases – do the bacteria show indications of being on a direct path towards the ‘chronic infection phenotype’, and if so what is the mechanism behind this?


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

From the environment to colonization of human airways

We have collected isolates from the majority of CF children infected with P. aeruginosa associated with the Copenhagen CF clinic during a 4.5 year period (2005–2009). In total 45 CF children were followed from first isolate of P. aeruginosa (or very early in their colonization history). In this communication we distinguish between cases of cured and persistent colonization of the patient airways (cf. Experimental procedures for definitions), and at the conclusion of the study a large group of patients (20 patients) had been cured (one or several times) of P. aeruginosa after a period of treatment with antibiotics, and for 13 patients status was still unknown because of recent colonization within the last 12 months. These cases are not examined further here. Another large group of 13 patients comprises cases of recurrent colonization where the bacteria are not present in airway samples for a period of time (often several months) and then suddenly reappear with the same genotype (to be published). Based mainly on duration of bacterial persistence and frequency of isolation it is from this group of persistent clones we have chosen three cases for further characterization of the bacterial genotypes and phenotypes and for evolutionary changes over time.

The experimental approach has been to combine transcriptional profiling using standardized Affymetrix DNA arrays with phenotypic characterization. The DNA array experiments have been performed under standardized conditions in laboratory media (cf. Experimental procedures), and the results of these provide information about impacts of genetic changes on expression of specific genes, when comparing data sets from isolates harvested from the patients at different time points. Because the Affymetrix gene chips mainly cover the genes present in reference strains of P. aeruginosa such as PAO1, PA14 and PAK, the obtained data only cover expression of genes in the core genome. Parallel phenotypic investigations have been employed to verify the transcription analysis, and to bring about information that may indicate which mutational events may be responsible for the observed changes of gene expression and corresponding phenotype. Gene sequencing was employed in some cases to confirm such mutational events.

Characterization of a transiently colonizing strain of P. aeruginosa

Patient B6 was initially colonized by a strain of P. aeruginosa that persisted for 2 years after which it was no longer isolated from patient samples. The sample time line is shown in Fig. 1. B6-0 is the first isolate of P. aeruginosa in this patient and most likely represents the original environmental strain initiating colonization, while B6-4 is the latest isolate of this particular genotype in the patient. It is not known if this first colonizing clone was removed by the medical treatment, by competition from a new genotype that emerged in the patient airways (Fig. 1), or both. Colony morphology appearance of the two isolates was identical and like that normally observed for wild-type isolates, indicating an environmental origin. The cells are fully motile in agreement with a general wild-type phenotype. Transcriptional profiling was performed through isolation of RNA from cultures growing exponentially in Luria–Broth (LB) medium and harvested at an OD600 of 0.5 followed by analysis using Affymetrix Gene Chips for P. aeruginosa. Global gene expression comparison of isolates B6-0 and B6-4 shows that only six genes had P-values below 0.05, all in the range of a twofold downregulation. Four of these are PA4582-PA4585 of which PA4585 (rtcA) is involved in RNA processing. These results suggest that only a small number of genetic changes occurred during the 2 years of colonization of this particular P. aeruginosa clone.


Figure 1. Time line of P. aeruginosa infection in patients B6, B12 and B38. Black boxes indicate P. aeruginosa-positive samples while stars, for patient B6, indicate samples containing a new genotype only. Unlike patient B6, only one genotype was detected in patients B12 and B38. Arrows indicate sample isolates chosen for DNA microarray analysis. Sample numbers are not completely consecutive as not all P. aeruginosa-positive samples were available for analysis.

Download figure to PowerPoint

Characterization of a persisting non-mucoid strain of P. aeruginosa

In patient B12 a specific clone of P. aeruginosa has persisted for more than 4 years and is still isolated from the patient. The isolate B12-0 does not represent the first acquisition for this patient, as P. aeruginosa was identified in a sample from the patient 4 months prior to sample B12-0. Although we have no stored isolate from this 4 month period it is quite likely that the P. aeruginosa clone was the same as the one described in the subsequent collected samples. DNA microarray analysis was performed on isolates from three samples: B12-0, B12-4 (isolated after 2 years and 10 months) and B12-7 (isolated after 3 and 8 months) (cf. Fig. 1). Gene expression changes are shown in Table 1 and in contrast to the lack of changes observed for the B6 isolates, the analysis of the B12 isolates showed that several genes and operons display changed expression levels in the later isolates compared with the first one. These changed expression levels include upregulation of the MexCD-OprJ multidrug efflux system involved in resistance towards, e.g. fluoroquinolones (Jalal et al., 2000; Jeannot et al., 2008). In agreement with this it was found that the MIC of ciprofloxacin, with which the patient was treated, increased 15-fold from a wild-type level of 0.19 to 3 µg ml−1. The underlying genetic basis is a 13 bp (Δ329–341) frameshift deletion in the MexCD-OprJ operon repressor, nfxB, present in isolate B12-7. The atu operon, also upregulated, is involved in the catabolism of the acyclic monoterpene citronellol, but also partly involved in catabolism of leucine (Aguilar et al., 2006), which may be an important nutrient in the airway mucus. Growth on leucine as carbon source was accordingly significantly improved for isolate B12-7 (doubling time of 4.1 h, SD = ±0.4) compared with isolate B12-4 (doubling time of 9.1 h, SD = ±0.4). A frame-shift deletion (ΔA326) in atuR in B12-7 could explain this. Several genes showed moderately changed expression, among which were genes encoding type III secretion proteins (downregulation). Because isolate B12-0 presumably had a 4 month period of colonization and adaptation prior to the sampling time, some other changes in expression relative to the original environmental colonizer could have occurred during this period. In fact, gene expression in B12-0 in comparison with the reference strain PAO1 revealed up to 10-fold downregulation of wbp genes involved in O-antigen synthesis as well as up to 17-fold downregulation of pil and fim genes involved in motility, in agreement with the finding that B12 isolates are twitching motility deficient.

Table 1.  All gene expression changes between isolates B12-0 and B12-7.a
Locus ID and gene nameDescriptionFold change
  • a.

    Gene expression changes are displayed as discrete fold changes, converted from log (2) fold changes. Negative values correspond to downwards fold change.

  • b.

    Gene expression change present between isolates B12-4 and B12-7, i.e. likely an effect of mutations occurring between these samples.

Antibiotic resistance  
 PA4597 oprJMultidrug efflux outer membrane protein OprJ precursor25
 PA4598 mexDMultidrug efflux transporter MexD53
 PA4599 mexCMultidrug efflux membrane fusion protein MexC precursor86
 PA4600 nfxBTranscriptional regulator NfxB12
Type III secretion  
 PA1694 pscQTranslocation protein in type III secretion−2
 PA1699Conserved hypothetical protein in type III secretion−2
 PA1700Conserved hypothetical protein in type III secretion−2
 PA1707 pcrHRegulatory protein PcrH−2
 PA1708 popBTranslocator protein PopB−3
 PA1709 popDTranslocator outer membrane protein PopD precursor−2
 PA1714 exsDExsD−2
 PA1715 pscBType III export apparatus protein−3
 PA1718 pscEType III export protein PscE−3
 PA3841 exoSExoenzyme S−2
Citronellol/leucine catabolism  
 PA2885 atuRbPutative repressor of atu genes3
 PA2886 atuAbProtein with apparent function in citronellol catabolism15
 PA2887 atuBbPutative dehydrogenase involved in catabolism of citronellol12
 PA2888 atuCbGeranyl-CoA carboxylase, β-subunit23
 PA2889 atuDbPutative citronellyl-CoA dehydrogenase, citronellol catabolism18
 PA2890 atuEbPutative isohexenylglutaconyl-CoA hydratase10
 PA2891 atuFbGeranyl-CoA carboxylase, α-subunit (biotin-containing)3
 PA2892 atuGbGCase, α-subunit (biotin-containing)8
 PA2893 atuHbPutative very long-chain acyl-CoA synthetase2
Exopolysaccharide production  
 PA2232 pslBProbable phosphomannose isomerase−4
 PA2233 pslCProbable glycosyl transferase−3
 PA2234 pslDPslD−2
 PA2235 pslEHypothetical protein−3
 PA2236 pslFHypothetical protein−3
 PA2237 pslGProbable glycosyl hydrolase−3
LPS modification  
 PA3552 arnBArnB−5
 PA3553 arnCArnC−3
 PA3554 arnAArnA−3
 PA3556 arnTArnT−2
 PA3558 arnFArnF−3
 PA0085Conserved hypothetical protein2
 PA0612 ptrBbRepressor, PtrB−2
 PA0637bConserved hypothetical protein−2
 PA0644bHypothetical protein−3
 PA0646bhypothetical protein2
 PA1535Probable acyl-CoA dehydrogenase4
 PA1559Hypothetical protein−4
 PA1560Hypothetical protein−5
 PA2193 hcnAbHydrogen cyanide synthase HcnA−2
 PA2199bProbable dehydrogenase−3
 PA2204bHypothetical protein−10
 PA2639 nuoDbNADH dehydrogenase−2
 PA2655Hypothetical protein2
 PA3115 fimVMotility protein FimV2
 PA3439 folXd-Erythro-7,8-dihydroneopterin triphosphate epimerase−5
 PA3470bHypothetical protein−2
 PA3559Probable nucleotide sugar dehydrogenase−2
 PA3842Probable chaperone−2
 PA3981bConserved hypothetical protein−3
 PA4354bChemotactic transducer PctC−2
 PA4596Probable transcriptional regulator2
 PA4601 morAbMotility regulator3
 PA4773Hypothetical protein−3
 PA4776 pmrAPmrA: two-component regulator system response regulator−2
 PA5298bXanthine phosphoribosyltransferase−3
 Pae_AF241171cds21bProbable two-component system−4
 Pae_tRNA_His_fbtRNA, histidine−2
 Pae_tRNA_Lys_fbtRNA, lysine−2
 Pae_tRNA_Met_sbtRNA, methionine−3
 ig_3874653_3873835bIntergenic region3

Characterization of a persisting mucoid strain of P. aeruginosa

Pseudomonas aeruginosa colonization in patient B38 occurred after a 10 year period of time without P. aeruginosa. The time line of this new infection in B38 is displayed in Fig. 1. The initial colonizing strain was wild-type non-mucoid (same phenotype as for isolates from B6 and B12) but rapidly converted to an alginate overproducing mucoid phenotype. In sample B38-2 five different clonal mucoid lineages could be identified by differences in the respective mucA sequences (Table 2). The five mucoid lineages displayed some phenotypic diversity and could in part also be distinguished based on the plate colony morphotypes of in vitro generated non-mucoid revertants. For the sequenced isolates each colony morphology type matched a specific lineage.

Table 2. mucA mutations of the five different mucoid lineages present in sample B38-2.
StrainmucA mutation
B38-2C-MC insert 251

Sequencing the mucA gene in isolates from later samples indicated that only the type A lineage carrying the T349C mucA allele was continuously present in the patient airways. Investigating colony morphologies of 70 isolates from sample B38-6 revealed that all displayed the morphotype associated with isolates carrying the T349C mucA allele. Thus it is strongly indicated that the type A mucoid lineage has dominated the P. aeruginosa population from sample B38-6 and onwards and most likely is the only persisting mucoid lineage. Consequently, isolates of the type A mucoid lineage were chosen for transcriptome analysis that included strains B38-1NM, B38-6A-M, B38-2A-M and B38-2A-NM*, an isogenic non-mucoid strain having a PA14 wild-type mucA allele. The latter strain was generated in vitro by allelic replacement, cf. Experimental procedures. Inter-strain gene expression comparison enables not only the determination of temporal changes in gene expression between isolates from samples 1, 2 and 6, but also documents that changes between isolates from sample 1 and 2 are related or unrelated to the mucA mutation.

Gene expression in B38-1-NM was considered to be a wild-type reference data set. In B38-2A-M 761 genes had significantly changed expression relative to B38-1-NM with a P-value cut-off of 0.05 and fold change above 2. Most of the gene expression changes were also observed when comparing the isogenic strains B38-2A-M and B38-2A-NM* (668 with a fold change above 2) showing that the genetic basis for most gene expression changes found in B38-2A-M is the mucA mutation. A comparison between strains B38-1NM and B38-2A-NM* disclosed fewer and smaller gene expression changes as only 75 genes showed changed expression (P-value below 0.05 and fold change above 2), while only one gene was significantly changed above twofold in expression in strain B38-6A-M relative to strain B38-2A-M. Selected differentially expressed genes from isolates from samples 1, 2 and 6 are shown in Table 3.

Table 3.  Selected gene expression changes between B38 non-mucoid and mucoid strains.a,b
Locus ID and gene nameDescriptionB38-1-NM vs. B38-2A-NM*B38-1-NM vs. B38-2A-MB38-2A-NM* vs. B38-2A-MB38-2A-NM* vs. B38-2A-M algD
  • a.

    Gene expression changes are displayed as discrete fold changes, converted from log (2) fold changes. Negative values correspond to downwards fold change.

  • b.

    A > or a < denotes a significant upwards or downwards, below twofold, gene expression change respectively (P < 0.05). The presence of no fold change number signifies an insignificant change in gene expression (P > 0.05).

Alginate regulation and biosynthesis     
 PA0762 algUSigma factor 333
 PA0763 mucAAnti-sigma factor 333
 PA5261 algRAlginate biosynthesis regulatory protein 544
 PA5483 algBTwo-component response regulator 332
 PA3540 algDGDP-mannose 6-dehydrogenase 181180 
Virulence factors     
 PA2386 pvdAL-ornithine N5-oxygenase−2369
 PA2396 pvdFPyoverdine synthetase F >49
 PA2426 pvdSSigma factor  65
 PA1905 phzG2Probable pyridoxamine 5′-phosphate oxidase>−3−6−6
 PA4211 phzB1Probable phenazine biosynthesis protein>−4−6−6
 PA4217 phzSFlavin-containing monooxygenase>−5−7−7
 PA2193 hcnAHydrogen cyanide synthase −20−24−27
 PA2194 hcnBHydrogen cyanide synthase>−5−6−5
 PA2195 hcnCHydrogen cyanide synthase −4−6−5
 PA3479 rhlARhamnosyltransferase chain A −5−5−5
 PA4228 pchDPyochelin biosynthesis protein −7−5−5
 PA4230 pchBSalicylate biosynthesis protein −17−10−9
 PA4229 pchCPyochelin biosynthetic protein −7−6−5
Secretion systems     
 PA1713 exsATranscriptional regulator type III secretion>−4−7−5
 PA1718 pscEType III export protein>−12−18−27
 PA1719 pscFType III export protein>−4−6−7
 PA0083Type VI secretion−2−27−13−5
 PA0084Type VI secretion<−18−11−3
 PA0085HcpI, secreted protein<−28−17−6
 PA0652 vfrQuorum sensing regulator<−2−2−2
 PA0905 rsmARegulator of secondary metabolites and quorum sensing 2>>
 PA1003 mvfRQuorum sensing regulator <−2−2
 PA3385 amrZAlginate and motility regulator Z 222
 PA4856 retSRegulator of virulence factors and type III and VI secretion −2<<
Quorum sensing     
 PA3476 rhlIAuto-inducer synthesis protein RhlI −4−4−3
 PA3477 rhlRRhl quorum sensing regulator −2−3−2
 PA1078 flgCFlagellar basal-body rod protein −17−17−21
 PA1092 fliCFlagellin type B −48−44−59
 PA1094 fliDFlagellar capping protein −23−21−23
 PA5042 pilOType 4 fimbrial biogenesis protein −5−5−4
 PA5043 pilNType 4 fimbrial biogenesis protein −5−4−3
 PA5044 pilMType 4 fimbrial biogenesis protein −7−6−5
Osmotic stress     
 PA0059 osmCOsmotically inducible protein>11710
 PA4876 osmEOsmotically inducible lipoprotein 976
 PA1323Putative osmotic stress response21154
 PA1324Putative osmotic stress response>744
 PA2146Putative osmotic stress response 8811
 PA3459Putative osmotic stress response 875
 PA3460Putative osmotic stress response>1085
 PA3461Putative osmotic stress response 555
 PA4880Putative osmotic stress response21263
 PA3584 glpDGlycerol-3-phosphate dehydrogenase −16−20−12
 PA4770 lldPL-lactate permease2−11−23 
Carbohydrate transport     
 PA3186 oprBGlucose/carbohydrate outer membrane porin33><
 PA3187Probable ATP binding component of ABC transporter78 <
 PA3188Probable permease of ABC sugar transporter812><
 PA3189Probable permease of ABC sugar transporter59> 
 PA3190Probable component of ABC sugar transporter45> 

It is evident that most changes in gene expression shown in Table 3 are caused by the mucA mutation, and as expected expression of both mucA and algU were upregulated in the mucoid strain as was expression of the AlgU controlled regulators algR and algB. Among the observed changes in gene expression in the mucA isolate it was striking that expression of several virulence factor genes was downregulated in the mucoid strain suggesting reduced production of phenazines involved in pyocyanin synthesis, hydrogen cyanide, rhamnolipids, type III and type VI secretion proteins, motility factors, chemotaxis proteins and Rhl quorum sensing proteins. These downregulations may not be a direct regulatory effect of AlgU, but rather indirect, as several well-known quorum sensing and virulence factor regulatory genes were also differentially expressed. Other pleiotropic effects of the mucA mutation include the expression upregulation of the genes osmC and osmE induced by osmotic stress in E. coli (Gutierrez and Devedjian, 1991; Gutierrez et al., 1995), and a number of genes shown to be part of the osmotic stress response of PA14 encoding hypothetical proteins or proteins putatively involved in the synthesis of the osmo-protectant N-acetyl glutaminyl glutamine amide (NAGGN) (Aspedon et al., 2006).

Because the most pronounced phenotypic effect of mucA mutations is the constitutive high-level production of extracellular alginate, it was important to clarify the direct and indirect consequences of the synthesis of this polymer. Therefore a B38-A2-M algD mutant strain was constructed, cf. Experimental procedures. The resulting strain produces no alginate, is still mucA deficient and expresses wild-type AlgU. As shown in the transcriptome analysis (Table 3) most of the gene expression changes that could be specifically associated with alginate overproduction are related to metabolism, e.g. downregulation of lactate and glycerol metabolism genes. Otherwise the gene expression profiles of B38-2A-M and the algD mutant strain are quite similar (correlation coefficient factor of 0.95), especially concerning the virulence factors.

Some of the differentially regulated genes unrelated to the mucA mutation include the operon (PA3186-PA3190) involved in transport of glucose, fructose, glycerol and mannitol (Wylie and Worobec, 1995) in addition to an operon involved in glutamate and aspartate transport and metabolism, both of which are upregulated. The only differentially expressed gene with a fold change above 2 when comparing isolates B38-2A-M and B38-6A-M is the mexX gene (2.6-fold), while a fold change of 1.7 is present for mexY. These genes encode part of the MexXY OprM efflux pump system involved in antibiotic resistance towards aminoglycosides, e.g. tobramycin (Aires et al., 1999; Islam et al., 2009). However the fold changes were low and assessing the MIC of tobramycin revealed only a minor increase in MIC from 0.75 for B38-2A-M to 1.5 for B38-6A-M.

Phenotypic consequences of mucA mutations

The gene expression data presented here suggest that the mucoid phenotype is associated with reduced virulence caused by the regulatory consequences of the mucA mutation. Because reduced virulence is one of the characteristics of P. aeruginosa isolates from chronically infected patients it was important to determine if the changed gene expression pattern related to the virulence genes was indeed correlated with a reduced virulence phenotype of mucA strains. Additionally, the generality of a possible reduction of virulence because of mucA mutations should be assessed in different strain pairs of P. aeruginosa, as clinical mucoid isolates have different genomes and carry different specific mucA mutations that could express different associated phenotypes. Thus the phenotypic analysis described in the following is both a control of the transcription analysis described in the previous section and an expanded investigation of the pleiotropy of mucA mutations.

Comparative analysis of mucoid and non-mucoid strains of P. aeruginosa can only be carried out reliably from isogenic pairs of strains. For the B38 isolates allelic replacements of the mutated mucA gene in B38-2A-M with a wild-type allele resulted in non-mucoid B38-2A-NM*, and a replacement of the wild-type algD gene with a mutant algD allele in B38-2A-M resulted in non-mucoid B38-2A-M algD. Similar allelic replacements were constructed for four other different mucoid strains isolated during early infection of CF children, and for the non-mucoid reference strain PAO1 an isogenic mucoid mucA strain, PDO300, was employed. The mucA genes of the mucoid variants of these pairs harbour different mutations. The resulting isogenic mucoid/non-mucoid pairs of strains were analysed phenotypically as described below.

The suggested reduced virulence of mucA mutant strains was confirmed by phenotypic characterization of the strains. Overall virulence activities of the relevant strains were assessed using the Caenorhabditis elegans worm killing assay. It was previously shown that P. aeruginosa isolates from chronically infected patients do not kill the worms, whereas isolates from the environment or from early stages of infection in CF children efficiently killed the worms. The comparison between non-mucoid, mucA mucoid and non-mucoid mucA, algD strains showed that mucA strains independent of the actual production of alginate were less virulent than wild-type strains (Fig. 2) in agreement with the gene expression profile showing reduced virulence factor expression in B38-2A-M and the isogenic algD mutant strain (Table 3). The less pronounced difference observed for the B38 strains is due to rapid reversion to fast growing non-mucoid cells (> 95% for B38-2A-M and about 40% for B38-2A-M algD). However, even for the B38 strains the mucA mutation resulted in reduced killing of the worms.


Figure 2. Survival of C. elegans propagated on isogenic non-mucoid (mucA+), mucoid (mucA-) and algD- mucA- strains as an estimate of virulence. Laboratory-derived (PDO300) and clinically derived (B38, B7 and B11) strains were included. Non-mucoid (mucA+) strains are highly virulent as they cause almost complete killing. Error bars display standard error of the mean.

Download figure to PowerPoint

Among the many virulence factors suggested to be expressed with reduced rates in the mucA strains (based on the transcriptomic data set) we chose to investigate further by phenotypic analysis the following activities: motility, rhamnolipid production and quorum sensing signal (the C4 Rhl signal) production. Table 4 shows that compared with the isogenic non-mucoid strain the mucoid variant B38-2A-M displayed a reduction in all types of motility, rhamnolipid and Rhl quorum sensing signal synthesis. Although B38-2A-M grows substantially slower than its isogenic non-mucoid strain (doubling time 60 vs. 35 min) it is not the cause of reduced virulence as B38-2A-M algD is almost equally slow growing. There was an overall trend of reduction in motility for all investigated mucoid isolates with minor deviations as, e.g. two pair of strains almost retain full swimming motility, and also rhamnolipid production differs between strains. This could indicate that the effect of the mucA mutation is strain background-dependent (e.g. isolate B7 seems more virulent than most others). Nevertheless, the specific virulence-associated phenotypes assayed here support the overall virulence activity measured in the C. elegans assay showing a reduced activity for the mucA strains, which also confirms the transcriptomic data.

Table 4.  Relative phenotypic levels of mucoid and algD strains compared with isogenic non-mucoid strains.a,b
  • a.

    Values are listed in percent and represent the phenotypic level of mucoid or algD strains compared with their isogenic non-mucoid strains.

  • b.

    All values significantly different compared with values of isogenic non-mucoid strains using two-sample t-test, P < 0.05.

  • c.

    Rhamnolipid production was determined in supernatant after 24 h incubation.

  • d.

    C4-AHL quorum sensing signal molecule was determined in supernatant from cultures harvested at an OD600 of 1.

  • e.

    > 90% non-mucoid revertants present at the edge of motility zone.

  • f.

    Low twitching level in isogenic non-mucoid strain.

  • g.

    Minor clumping of isogenic non-mucoid strain at harvest.

  • h.

    Phenotype level is zero for both mucoid and non-mucoid strains.

B38-2A-M algD1453e2635
PDO300 algD24881726
B6-M0f2312Not done
B7-M algD30e826626


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Do the bacteria show indications of being on a path towards a ‘chronic infection phenotype’, and if so which underlying routes and mechanisms are involved?

The first simple answer to the question is that for two of the patients (B12 and B38) genetic changes have led to development of new bacterial phenotypes during the course of colonization, and in both cases the derived prominent phenotypes – antibiotic resistance and mucoidy – are associated with the ‘chronic infection phenotype’. In fact, colonization in these two patients has progressed to a chronic infection. In the former case (B12) the mechanism behind the increased antibiotic resistance of the bacteria is a mutation in nfxB leading to upregulation of expression of the MexCD-OprJ efflux pump (Jalal et al., 2000; Jeannot et al., 2008). In the latter case (B38) the mechanism behind the bacterial mucoidy is upregulation of the alginate biosynthesis enzymes encoded by the alg operon, caused by a mutation in the mucA anti-sigma factor. For the last patient (B6) no obvious phenotypic changes occurring during the 2 year persistence of P. aeruginosa seem related to the ‘chronic infection phenotype’, and persistence can therefore be explained by either introduction of unidentified adaptive mutations or simply the general adaptive repertoire of this organism. Our data do not permit us to suggest any specific mechanisms in this case.

The high resolution of our phenotypic analysis (DNA arrays for transcriptional profiling) allows us, however, to provide more detailed answers to the question. Thus, in addition to the increased ciprofloxacin resistance mediated by the mexCD-OprJ upregulation, the later B12 isolates also show phenotypic changes reducing expression of virulence factors (TTSS). Furthermore, all B12 isolatesare twitching motility deficient and compared with PAO1 expression of genes (wbp) involved in O-antigen synthesis were significantly downregulated, likely causing the LPS modifications frequently found in CF infection (Goldberg and Pier, 1996). Increased expression of the atu operon conferring improved growth on leucine has not previously been reported but probably reflects another less studied but potentially important means of adaptation by metabolic optimization. Sputum of CF patients is known to be rich in amino acids (Barth and Pitt, 1996), and a 12-patient study showed that leucine is the fourth most abundant amino acid in CF sputum present in almost 1 mM concentrations on average (Palmer et al., 2007). Pleiotropic effects of the upregulation of MexCD-OprJ often include reduced virulence and growth rate (Sanchez et al., 2002; Linares et al., 2005; Jeannot et al., 2008), but in our investigations no growth rate differences were observed for B12-4 and B12-7 relative to that of B12-0. Nonetheless, it is clear that the mutations responsible for the changed phenotypes of the B12 isolates – whether pleiotropic or not – shift the colonizing bacteria in direction of a ‘chronic infection phenotype’.

For the B38 isolates, the shift towards the ‘chronic infection phenotype’ is even more apparent. With a single mutation (mucA) the ‘chronic infection phenotype’ is approached covering most of the characteristics, e.g. alginate overproduction and reduction in virulence, motility and growth rate. The transcription data of the B38 isolates show that the mucA mutation has a substantial pleiotropic effect on gene expression. Here we will focus most of our discussion on gene expression changes that have been confirmed by phenotypic analysis (cf. Supporting information for complete data sets). Interestingly, it turned out that several well-known virulence genes in P. aeruginosa showed clear reduction in expression in the mucoid mucA strain compared with an isogenic non-mucoid strain. Previous studies have investigated the pleiotropic effects of mucA mutations or investigated the role of the relevant regulators, e.g. AlgU and AlgR. In a transcriptome analysis of PAO381 – a derivative of PAO1 – and its isogenic mucA- strain very different results compared with this study were obtained (Firoved and Deretic, 2003). Of the 27 differentially expressed genes published by Firoved and Deretic only 3 genes were regulated in the same direction, 9 in the opposite direction and the rest unchanged in the B38 transcriptome analyses. Although Firoved and Deretic (2003) applied a strict cut-off value and only included genes with fold changes above 4 it cannot explain the observed differences. Strain differences and/or the specific conditions of the experiments may explain the observed differences.

Several other publications seem to corroborate the findings of the mucoid phenotype observed in this study. Thus it was found that a mucA mutation resulted in reduced expression of TTSS-related genes, AlgU shown to repress flagellum synthesis, and AlgR found to repress the rhl QS system in biofilms (Wu et al., 2004; Tart et al., 2006; Morici et al., 2007). However, differences have also been observed as mucoid strains produced more hydrogen cyanide than non-mucoid strains (Carterson et al., 2004) using a different experimental setup than ours, which could suggest that hcn regulation through AlgU is dependent on growth conditions. Yet, in non-mucoid strains AlgR was shown to be a hydrogen cyanide repressor (Lizewski et al., 2004).

The reduced expression of virulence factors in the mucA strains suggests that these variants have lost part of their acute infection potential and gained potential for chronic infections. There are several plant and animal models for acute infections with P. aeruginosa, one of which is the C. elegans killing assay (Tan et al., 1999; Thomsen et al., 2006), which was chosen here for assessments of the virulence activity of the mucoid/non-mucoid strain pairs. Previous investigations have shown that P. aeruginosa isolates from the environment or from early stages of infection in CF children in most cases are fully virulent resulting in rapid killing of the nematodes, whereas isolates from older chronically infected patients most often displayed reduced or no virulence (Jelsbak et al., 2007). The results presented here show that mucA strains in general display reduced acute virulence, as measured in the C. elegans assay, thus confirming the in vitro data.

The gene expression consequences of the mucA mutations affecting the AlgU regulon could be further selectively advantageous in the stressful environment of the CF airways. In fact, a subset of genes putatively involved in the osmotic stress response was shown here to be upregulated in the mucA strain (expected to be beneficial in the CF airway mucus). It was furthermore found that a large fraction of cell wall stress-induced genes belong to the AlgU regulon (Wood and Ohman, 2009). In vitro conversion to a mucoid phenotype is known to occur during both oxidative and osmotic stress conditions (Terry et al., 1991; Mathee et al., 1999). In other Gram-negative bacteria the role of AlgU homologues (σE) is also stress-related, e.g. by ensuring homeostasis of lipopolysaccharide and outer membrane porins (Rhodius et al., 2006). It was further proposed (Rhodius et al., 2006) that the σE regulon in general may constitute an adaptation system to facilitate survival in vivo; this seems to be true for P. aeruginosa. The linkage between alginate production, increased oxidative and osmotic stress response and reduction of virulence provides a powerful explanation to why this regulon is repeatedly targeted for activation by MucA mutation in the CF lung environment. Importantly, the effective alteration of the MucA-AlgT regulatory circuitry constitutes the basis for the pleiotropic phenotypic switch by a single mutational event.

At least five different mucoid lineages arose almost simultaneously in patient B38 as evidenced by mucA sequences and while it seems clear that only one lineage persisted, the reason for this is less clear. The presence of additional adaptation specific for this lineage is one explanation whereas another could be the specific pleiotropic effects of the mucA mutation in this lineage. Despite their clonal relationship the five lineages displayed small phenotypic differences in the virulence properties tested here (data not shown). If not because of other genetic differences among the lineages the latter could indicate that different mucA mutations may cause different degrees of pleiotropic effects based on varying interactions with the AlgT or MucB proteins.

The appearance of mucA variants in samples from CF patients is often very dominating for a period after the occurrence of the mutation(s) in the bacterial airways population, but subsequently non-mucoid clones appear and constitute substantial sub-populations in the patients. In most cases of chronically infected patients it has been shown that by far most of these non-mucoid clones still harbour mucA mutations (nearly always the same as the mucoid counterparts) (Ciofu et al., 2001; Jelsbak et al., 2007). In laboratory conditions most of such second-site revertants carry mutations in the algU gene, but this is not the normal cause of non-mucoid phenotypes in the CF patients (Ciofu et al., 2008). Although it is still not clear which mutations are responsible for the phenotypic reversion of mucoidy in these cases, the findings suggest that there is selection for maintenance of the mucA phenotype combined with a wild-type allele of the algU gene, and it is tempting to speculate that one reason for this is the apparent beneficial pleiotropic effects described in this communication. Moreover it could indicate that the most important pleiotropic effect of AlgU activation might not be alginate overproduction, but rather one or a collection of the other aforementioned pleiotropic effects.

The three cases of early colonization of human hosts have shown that P. aeruginosa normally residing in natural environments can move successfully to highly different and indeed stressful environmental settings including the airways of humans and persist for hundreds or even thousands of generations without a significant number of genetic adaptive changes. The normal adaptive repertoire of wild-type strains has some limits, however, and without fitness-increasing genetic alterations the host defence or competition from other strains will result in eventual eradication of the cells. In the event that fitness-improving mutations do occur in the bacterial genome, persistence may be extended. In the present cases, of the three main characteristics of the CF lung environment, host immune system, antibiotics and substrate composition, the B12 isolates seem to possess increased fitness towards all of these while fitness increase in the B38 isolates seems to relate to the immune system and, less pronounced, antibiotics. Meanwhile, the route towards ‘the chronic phenotype’ and fitness increase seems to consist either of the additive effects of several mutations or, for the mucoid variant, by a mutation with profound pleiotropic effects. Identification of pleiotropic adaptive mutations in regulatory genes is in fact a recurrent finding in a number of experimental evolution experiments with microbial populations in the laboratory (Kolter, 1999; Zinser and Kolter, 1999; Elena and Lenski, 2003; Bantinaki et al., 2007), documenting net fitness increases despite highly complex resulting phenotypes. Pseudomonas aeruginosa infections in CF airways may help to further understand how organisms can move between highly different natural environments helped by a low number of mutations with high phenotypic impact.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Bacterial strains and growth conditions

Bacterial strains and clinical isolates used in this study are listed in Table 5. The general isolation procedure and identification of P. aeruginosa from sputum was performed as previously described (Høiby and Frederiksen, 2000). Sputum samples were obtained by expectoration or endo-laryngeal suction, followed by Gram staining. Pseudomonas aeruginosa was isolated on selective plates and identified by conventional biochemical tests. Up to 84 isolates from each sample were selected and stored in microtitre plates at −80°C including all colonies with a distinct morphotype. The genotype of P. aeruginosa isolates was identified by single-nucleotide polymorphism (SNP) typing using AT biochips (Clondiag Chip Technologies, Germany) (Wiehlmann et al., 2007). Unless otherwise noted bacteria were cultured in LB at 37°C. Antimicrobial agents were used where appropriate at the following concentrations: ampicillin at 100 µg ml−1, gentamycin sulfate at 30 µg ml−1 (P. aeruginosa) and 15 µg ml−1 (E. coli), carbenicillin at 200 µg ml−1. Estimation of growth rate on leucine as carbon source was performed essentially as previously described (Martin et al., 1973; Aguilar et al., 2006) using M9 medium containing 0.6% wt/vol l-leucine and supplemented with 0.005% l-isoleucine and l-valine. Growth was performed shaking at 37°C in 250 ml flasks containing 50 ml of medium.

Table 5.  Strains and plasmid used in this study.
Strain or plasmidDescriptionSource or reference
 PAO1Wild-typeHolloway and Morgan (1986)
 PA14Wild-typeRahme and colleagues (1995)
 PDO300PAO1 mucA22 (ΔG430)Mathee and colleagues (1999)
 PDO300 algD-algD-, non-mucoid strain derived from PDO300This study
 E. coli HB101supE44 hsdS20 (rB- mB-) recA13 ara-14 proA2 lacY1 galK2 rpsL20 xyl-5 mtl-1 leuB6 thi-1Kessler and colleagues (1992)
 B3-MMucoid isolate from sample 3 from patient B3 having mucA mutation ΔA368This study
 B3-NM*Non-mucoid strain derived from B3-M. mucA allele replacedThis study
 B6-0Non-mucoid isolate from sample 0 from patient B6This study
 B6-4Non-mucoid isolate from sample 4 from patient B6This study
 B6-MMucoid isolate from sample 6 from patient B6 having mucA mutation C352TThis study
 B6-NM*Non-mucoid strain derived from B6-M. mucA allele replacedThis study
 B7-MMucoid isolate from sample 3 from patient B7 having mucA mutation C424TThis study
 B7-NM*Non-mucoid strain derived from B7-M. mucA allele replacedThis study
 B7-M algDalgD-, non-mucoid strain derived from B7-MThis study
 B11-MMucoid isolate from sample 4 from patient B11 having mucA mutation T382CThis study
 B11-NM*Non-mucoid strain derived from B11-M. mucA allele replacedThis study
 B12-0Non-mucoid isolate from sample 0 from patient B12This study
 B12-4Non-mucoid isolate from sample 4 from patient B12This study
 B12-7Non-mucoid isolate from sample 7 from patient B12This study
 B38-1-NMNon-mucoid isolate from sample 1 from patient B38This study
 B38-2A-MMucoid isolate from sample 2 from patient B38 mucA mutation T349CThis study
 B38-2A-NM*Non-mucoid strain derived from B38-2A-M. mucA allele replacedThis study
 B38-2A-M algD-algD-, non-mucoid strain derived from B38-2A-MThis study
 B38-6A-MMucoid isolate from sample 6 from patient B38 mucA mutation T349CThis study
 pRK600CmR; oriColE1, RK2-Mob+, RK2-Tra+; helper plasmid for triparental matingsde Lorenzo and Timmis (1994)
 pDONR221KmR; Gateway donor vectorRohwer and Edwards (2002)
 pEX18ApGWApR; Gateway destination vectorChoi and Schweizer (2005)
 pPS856ApR, GmR; Gm cassette flanked with FRT sequencesHoang and colleagues (1998)
 pFLP2ApR; source for Flp recombinaseHoang and colleagues (1998)
 pMHR1ApR, GmR; Gm cassette from pPS856 and 1.8 kb algT-mucA-mucB fragment from PA14This study
 pMHR2ApR, GmR; from pEX18ApGW. Gm cassette and FRT sequences from pPS856 flanked with algD gene fragment from B38-2A-MThis study

CF patients

As part of a larger study we investigated P. aeruginosa isolates from the majority of CF children attending the Copenhagen CF clinic from 2005 and onwards having their first acquisition of P. aeruginosa during this period, or for a few patients just before 2004. As part of the general management structure at the CF Center, all patients were monitored on a monthly basis by evaluation of their clinical status, pulmonary function and microbiology of lower-airway secretions. Detection of P. aeruginosa in sputum resulted in treatment with antimicrobial eradication therapy consisting of 3 weeks of oral ciproxin plus 3 months of inhaled colistin. Instead of applying the clinical definitions of intermittent colonization and chronic infection (Høiby et al., 2005), we here distinguish between cases of cured and persistent colonization of the patient airways. Colonization with a specific genotype of P. aeruginosa, which is absent from all consecutive samples collected over a period of at least 1 year, is considered cured, whereas continued colonization for more than a year with the same genotype is considered a persistent airway colonization. Analysis was performed on bacterial isolates from three selected patients (B6, B12 and B38) having persistent P. aeruginosa colonization. Two of the patients (B12 and B38) had chronic P. aeruginosa infections according to the clinical definition based on antibody titre level increases: for B12 the antibody titre level passed 2 after 51 months while reaching 7 after 56 months, and for B38 a level of 2 was passed after 7 months while reaching 21 after 19 months. In contrast the antibody titre level was 1 after 48 months for B6.

mucA allelic replacement

Isogenic non-mucoid strains were generated from mucoid clinical isolates by allelic replacement of the mutated mucA allele with a wild-type P. aeruginosa PA14 mucA allele (the B38-1 mucA sequence is identical to the PA14 mucA sequence). Initially, a 1.8 kb fragment containing the PA14 mucA allele was generated by PCR using primers CCATGGTGCAAGAAGCCCGAGTCTAT and GAATTCCAACTGGGTGAACTGGAAGC containing an NcoI and EcoRI site respectively. The PCR fragment was cloned into the EcoRI and NcoI site of plasmid pEX18ApGW. A PCR fragment containing a gentamycin resistance cassette was amplified from pPS856 using primers CCATGGCGAATTAGCTTCAAAAGCGCTCTGA and GGATCCCGAATTGGGGATCTTGAAGTTCCT and cloned in the NcoI and BamHI sites of pEX18ApGW containing the PA14 mucA PCR fragment, creating pMHR1. Plasmid pMHR1 was introduced into P. aeruginosa clinical isolates by triparental conjugation employing helper strain HB101/pRK600. Single recombinants were selected by screening for gentamycin resistance and verifying carbenicillin resistance, while sacB-mediated sucrose counterselection was employed for plasmid excision leading to allele replacement. To avoid selecting non-mucoid spontaneous revertants it was always verified that sucrose counterselection generated both mucoid plasmid excised and non-mucoid plasmid excised strains. Allellic replacement was verified by mucA sequencing while potential AlgU mutations were excluded by AlgU sequencing.

Generation of unmarked P. aeruginosa algD deletion mutants

algD deletion mutants were generated as described by Choi and Schweizer (2005). Briefly, a gentamycin cassette flanked by B38-2A-M algD sequences were generated using primers up fw TTAACGGAAAGGCCATCAAG, up rev CCCAAACCAAAGATGCTGAT, down fw TCGACCTGGTGAACAAGACC and down rev ATCAGCAGGCTGAGGAACAC. The construct was inserted into pEX18ApGW creating pMHR2 and the plasmid introduced into mucoid strains by triparental mating using the helper strain E. coli HB101/pRK600. Allelic exchange as indicated by a Gmr sucroser Cbs phenotype was verified by PCR using the primers, up fw and down rev. Generation of unmarked deletion mutants was achieved by transforming electro-competent P. aeruginosa with pFLP2 and deletion of the Gmr marker was verified by colony PCR and sequencing.

DNA microarray sample processing

Transcriptomic profiles of clinical isolates were obtained using the Affymetrix P. aeruginosa gene chip. Triplicate experiments were performed for each strain. Bacteria were grown at 37°C in 50 ml of LB medium in a baffled 250 ml Erlenmeyer flask shaking at 240 r.p.m. Cell density starting conditions were 0.01 at OD600 inoculated from an overnight culture grown in LB. Bacteria were harvested in exponential phase at an OD600 of 0.5 and immediately mixed with RNAprotect Bacteria Reagent (Qiagen) and stored at −80°C. For mucoid strains preservation of mucoidy was confirmed by plating the cultures and scoring the colonies for mucoidy. RNA was extracted with RNeasy mini kit (Qiagen) and transcribed to cDNA using SuperScript III Reverse Transcriptase (Invitrogen). Subsequent cDNA purification, fragmentation and labelling were performed based on prokaryotic sample and array processing protocol from Affymetrix (Santa Clara, CA). The labelled cDNA were then hybridized on Affymetrix P. aeruginosa gene chips and stained on the GeneChip® Fluidics Station. The probe arrays were scanned with the GeneChip® Scanner 3000.

Microarray data analysis

The raw data (Tables S1 and S2) were obtained using the Affymetrix GeneChip® Operating System 1.4 (GCOS). Microarray data analysis was performed using the BioConductor package for the R software environment ( Normalization and expression index calculation was done with the RMA function while linear modelling was performed with the LIMMA package and multiple strain-pair contrasts. A P-value < 0.05 and absolute fold change > 2 was applied as cut-off values for each contrast (strain comparison). Transcription factor genes were selected using a P-value threshold < 0.05 only. The fold change was calculated using the difference of average (log2) expression levels of all strain replicate group while the annotations and functional classes were assigned according to the Pseudomonas Genome Database V2 (

Gene sequencing


Acyl-homoserine lactone quantification

Acyl-homoserinelactone production was quantified by high-pressure liquid chromatography and mass spectrometry. ABT medium containing 0.2% glucose and 1.0% cas-aminoacids was used as growth medium while growth conditions were 37°C in 50 ml of medium in a 250 ml Erlenmeyer flask shaking at 240 r.p.m. Cell density starting conditions were 0.01 at OD600 inoculated from an ON culture grown in ABTGC. For mucoid strains preservation of mucoidy was confirmed by plating culture. Bacteria were harvested in late exponential phase at an OD600 of about 1.0 and after centrifuging culture at 12 000 g for 5 min, 1.0 ml of supernatant transferred to an auto-sampler vial along with 50 µl of 3 YY µM D5-C-4 internal standard (hydrogen's in the acyl chain substituted by deuterium) in acetonitrile (ACN) and frozen at −80°C. HPLC-MS/MS was performed on an Agilent (Torrance, CA) 1100 HPLC system controlled by MassLynx V4.1. Samples were separated on a Gemini C6-phenyl 3 µm, 2 mm ID, 50 mm column (Phenomenex, Torrance, CA), using a flow rate of 0.300 ml min−1 at 25°C. A linear water-ACN gradient was used, starting at 2% ACN, going to 40% ACN in 30 s, then 100% ACN in 4.5 min, the 1 min with a flow of 0.500 ml min−1, before reverting to the start conditions for 1 min and holding this for 5 min. Both solvents contained 20 mM formic acid. The HPLC was a coupled Quattro Ultima triple mass spectrometer (Waters, Manchester, United Kingdom) with a Z-spray electrospray ionization source using a flow rate of 700 l h−1 nitrogen at 350°C; hexapole 1 was held at 30 V, and the cone was held at 25 V. Nitrogen was used as collision gas, and the mass spectrometer operated in positive multiple-reaction monitoring mode (dwell time 100 ms). Multiple-reaction monitoring mode were: 0–4 min,(i) C-4 AHL retention time (RT) 2.98 min m/z 172[RIGHTWARDS ARROW]102 @ 15eV (quantifier ion) and 172[RIGHTWARDS ARROW]71 @ 15eV (qualifier), (ii) open lactone-C-4 AHL RT 2.68 min 190[RIGHTWARDS ARROW]120 @ 25eV (quantifier) and 190[RIGHTWARDS ARROW]71 @ 25eV (qualifier) and (iii) D5-C-4 (internal standard) 177[RIGHTWARDS ARROW]102 @ 15eV (same RT as C-4); 4–7 min, (iv) Oxo-C12 AHL RT 5.08 298[RIGHTWARDS ARROW]102 @ 25V (quantifier) and 298[RIGHTWARDS ARROW]197 @ 25eV (qualifier) and (v) open lactone Oxo-C12 AHL RT 4.67 min 298[RIGHTWARDS ARROW]102 @ 25eV (quantifier) and 298[RIGHTWARDS ARROW]197 @ 25eV (qualifier). Calibration was done using D5-C-4 as internal standard (isotope dilution) against standards in ACN. Data were processed in Quanlynx 4.1 (Waters) with s/n ratio of 10 (both transitions per compound), giving detection limits in the 10–30 nM range.

Motility assay

Twitching and swimming motilities were assayed on agar plates containing AB mimimal medium (Hansen et al., 2007) supplemented with 7.4 µM thiamin, 0.5% glucose and 0.5% casaminoacids; 0.3% and 1.5% agar were used for swimming and twitching respectively. Plates were inoculated with nearly equal amounts of biomass as starting conditions based on OD600 measurement of ON culture. The cells were inoculated onto the bottom of the dish for twitching plates and inside the agar for swimming plates. Twitching plates were incubated for 48 h at 37°C while swimming plates were incubated for 24 h at 30°C and subsequently maximum diameter was measured and values background corrected by subtracting values of a motility deficient strains (pilA, fliF). Preservation of mucoidy was tested by streaking on LB plates from motility plates.

Rhamnolipid assay

The concentration of rhamnolipids in culture supernatants was determined by the orcinol method as previously described (Koch et al., 1991; Pamp and Tolker-Nielsen, 2007), with modifications. Briefly, P. aeruginosa strains were grown at 37°C in LB for 24 h, and mucoid strains were tested for preservation of mucoidy by streaking on LB plates. A 0.5 ml of aliquot of culture supernatant was extracted twice with 2 vols of diethyl ether. The ether fractions were pooled, evaporated to dryness, and subsequently 1 ml of orcinol reagent was added and the sample heated at 80°C for 30 min. Orcinol reagent was prepared immediately prior to use and consisted of 7.5 vols of 60% (vol/vol) sulfuric acid and 1 vol. of 1.6% (wt/vol) orcinol in distilled water. After heating, the samples were allowed to cool at room temperature for 15 min, and absorbance (A421) was measured and compared with rhamnose standards. Values were adjusted for differences in final cell density (OD600).

Antibiotic minimum inhibitory concentration

Minimum inhibitory concentrations were estimated by E-test according to the manufacturer's guidelines (AB Biodisk, Solna, Sweden) with minor modifications. Inoculums were prepared from an ON culture diluted to an OD600 of 0.5 of which 100 µl was inoculated and spread with a Drigalski spatula on a predried LB plate. Subsequently E-test strips were carefully placed on the LB plate and MIC values were read after 24 h of incubation.

C. elegans virulence assay

Virulence was assessed in C. elegans as described (Thomsen et al., 2006). A total of 20 µl of overnight culture of each strain was spread onto Nematode Growth Medium (NGM) plates and incubated at 37°C over night. For each strain, about 100 L4 hermaphrodites of the pha-1 (e2123ts) mutant (Schnabel and Schnabel, 1990) were transferred from NGM plates seeded with E. coli OP50 to the plates seeded with P. aeruginosa strains and incubated at 25°C. The plates were scored for live and dead worms every 24 h. Three independent trials were performed for each strain. The E. coli and C. elegans strains used in this work were provided by the Caenorhabditis Genetics Center (University of Minnesota, Minneapolis).


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

The authors wish to thank Dr Anders Folkesson, Department of Systems Biology, DTU, for his advice and assistance in relation to some of the genetic experiments. Also thanks to Dr Hanne Jarmer, Department of Systems Biology, DTU, for initial supervison of DNA array data analysis and Dr Sünje Johanna Pamp, Stanford University School of Medicine, for advice on the rhamnolipid assay. The Lundbeck Foundation and the Danish Council for Independent Research (FNU) supported the work by grants to S.M.


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Aguilar, J.A., Zavala, A.N., Diaz-Perez, C., Cervantes, C., Diaz-Perez, A.L., and Campos-Garcia, J. (2006) The atu and liu clusters are involved in the catabolic pathways for acyclic monoterpenes and leucine in Pseudomonas aeruginosa. Appl Environ Microbiol 72: 20702079.
  • Aires, J.R., Kohler, T., Nikaido, H., and Plesiat, P. (1999) Involvement of an active efflux system in the natural resistance of Pseudomonas aeruginosa to aminoglycosides. Antimicrob Agents Chemother 43: 26242628.
  • Aspedon, A., Palmer, K., and Whiteley, M. (2006) Microarray analysis of the osmotic stress response in Pseudomonas aeruginosa. J Bacteriol 188: 27212725.
  • Bagge, N., Ciofu, O., Hentzer, M., Campbell, J.I., Givskov, M., and Hoiby, N. (2002) Constitutive high expression of chromosomal beta-lactamase in Pseudomonas aeruginosa caused by a new insertion sequence (IS1669) located in ampD. Antimicrob Agents Chemother 46: 34063411.
  • Bantinaki, E., Kassen, R., Knight, C.G., Robinson, Z., Spiers, A.J., and Rainey, P.B. (2007) Adaptive divergence in experimental populations of Pseudomonas fluorescens. III. Mutational origins of wrinkly spreader diversity. Genet 176: 441453.
  • Barth, A.L., and Pitt, T.L. (1996) The high amino-acid content of sputum from cystic fibrosis patients promotes growth of auxotrophic Pseudomonas aeruginosa. J Med Microbiol 45: 110119.
  • Boucher, J.C., Yu, H., Mudd, M.H., and Deretic, V. (1997) Mucoid Pseudomonas aeruginosa in cystic fibrosis: characterization of muc mutations in clinical isolates and analysis of clearance in a mouse model of respiratory infection. Infect Immun 65: 38383846.
  • Bragonzi, A., Paroni, M., Nonis, A., Cramer, N., Montanari, S., Rejman, J., et al. (2009) Pseudomonas aeruginosa microevolution during cystic fibrosis lung infection establishes clones with adapted virulence. Am J Respir Crit Care Med 180: 138145.
  • Burns, J.L., Gibson, R.L., McNamara, S., Yim, D., Emerson, J., Rosenfeld, M., et al. (2001) Longitudinal assessment of Pseudomonas aeruginosa in young children with cystic fibrosis. J Infect Dis 183: 444452.
  • Cabral, D.A., Loh, B.A., and Speert, D.P. (1987) Mucoid Pseudomonas aeruginosa resists nonopsonic phagocytosis by human neutrophils and macrophages. Pediatr Res 22: 429431.
  • Carterson, A.J., Morici, L.A., Jackson, D.W., Frisk, A., Lizewski, S.E., Jupiter, R., et al. (2004) The transcriptional regulator AlgR controls cyanide production in Pseudomonas aeruginosa. J Bacteriol 186: 68376844.
  • Choi, K.H., and Schweizer, H.P. (2005) An improved method for rapid generation of unmarked Pseudomonas aeruginosa deletion mutants. BMC Microbiol 5: 30.
  • Ciofu, O., Fussing, V., Bagge, N., Koch, C., and Hoiby, N. (2001) Characterization of paired mucoid/non-mucoid Pseudomonas aeruginosa isolates from Danish cystic fibrosis patients: antibiotic resistance, beta-lactamase activity and RiboPrinting. J Antimicrob Chemother 48: 391396.
  • Ciofu, O., Lee, B., Johannesson, M., Hermansen, N.O., Meyer, P., and Hoiby, N. (2008) Investigation of the algT operon sequence in mucoid and non-mucoid Pseudomonas aeruginosa isolates from 115 Scandinavian patients with cystic fibrosis and in 88 in vitro non-mucoid revertants. Microbiology 154: 103113.
  • Elena, S.F., and Lenski, R.E. (2003) Evolution experiments with microorganisms: the dynamics and genetic bases of adaptation. Nat Rev Genet 4: 457469.
  • Firoved, A.M., and Deretic, V. (2003) Microarray analysis of global gene expression in mucoid Pseudomonas aeruginosa. J Bacteriol 185: 10711081.
  • Goldberg, J.B., and Pier, G.B. (1996) Pseudomonas aeruginosa lipopolysaccharides and pathogenesis. Trends Microbiol 4: 490494.
  • Govan, J.R., and Deretic, V. (1996) Microbial pathogenesis in cystic fibrosis: mucoid Pseudomonas aeruginosa and Burkholderia cepacia. Microbiol Rev 60: 539574.
  • Gutierrez, C., and Devedjian, J.C. (1991) Osmotic induction of gene osmC expression in Escherichia coli K12. J Mol Biol 220: 959973.
  • Gutierrez, C., Gordia, S., and Bonnassie, S. (1995) Characterization of the osmotically inducible gene osmE of Escherichia coli K-12. Mol Microbiol 16: 553563.
  • Hancock, R.E. (1997) The bacterial outer membrane as a drug barrier. Trends Microbiol 5: 3742.
  • Hancock, R.E., and Speert, D.P. (2000) Antibiotic resistance in Pseudomonas aeruginosa: mechanisms and impact on treatment. Drug Resist Updat 3: 247255.
  • Hansen, C.R., Pressler, T., and Hoiby, N. (2008) Early aggressive eradication therapy for intermittent Pseudomonas aeruginosa airway colonization in cystic fibrosis patients: 15 years experience. J Cyst Fibros 7: 523530.
  • Hansen, S.K., Haagensen, J.A., Gjermansen, M., Jorgensen, T.M., Tolker-Nielsen, T., and Molin, S. (2007) Characterization of a Pseudomonas putida rough variant evolved in a mixed-species biofilm with Acinetobacter sp. strain C6. J Bacteriol 189: 49324943.
  • Hoang, T.T., Karkhoff-Schweizer, R.R., Kutchma, A.J., and Schweizer, H.P. (1998) A broad-host-range Flp-FRT recombination system for site-specific excision of chromosomally-located DNA sequences: application for isolation of unmarked Pseudomonas aeruginosa mutants. Gene 212: 7786.
  • Høiby, N., and Frederiksen, B. (2000) Microbiology of cystic fibrosis. In Cystic Fibrosis, 2nd edn. Hodson, M.E., and Geddes, D.M. (eds). London, UK: Hodder Arnold, pp. 83107.
  • Høiby, N., Krogh Johansen, H., Moser, C., Song, Z., Ciofu, O., and Kharazmi, A. (2001) Pseudomonas aeruginosa and the in vitro and in vivo biofilm mode of growth. Microbes Infect 3: 2335.
  • Høiby, N., Frederiksen, B., and Pressler, T. (2005) Eradication of early Pseudomonas aeruginosa infection. J Cyst Fibros 4 (Suppl. 2): 4954.
  • Holloway, B.W., and Morgan, A.F. (1986) Genome organization in Pseudomonas. Annu Rev Microbiol 40: 79105.
  • Islam, S., Oh, H., Jalal, S., Karpati, F., Ciofu, O., Hoiby, N., and Wretlind, B. (2009) Chromosomal mechanisms of aminoglycoside resistance in Pseudomonas aeruginosa isolates from cystic fibrosis patients. Clin Microbiol Infect 15: 6066.
  • Jalal, S., Ciofu, O., Hoiby, N., Gotoh, N., and Wretlind, B. (2000) Molecular mechanisms of fluoroquinolone resistance in Pseudomonas aeruginosa isolates from cystic fibrosis patients. Antimicrob Agents Chemother 44: 710712.
  • Jeannot, K., Elsen, S., Kohler, T., Attree, I., Van Delden, C., and Plesiat, P. (2008) Resistance and virulence of Pseudomonas aeruginosa clinical strains overproducing the MexCD-OprJ efflux pump. Antimicrob Agents Chemother 52: 24552462.
  • Jelsbak, L., Johansen, H.K., Frost, A.L., Thogersen, R., Thomsen, L.E., Ciofu, O., et al. (2007) Molecular epidemiology and dynamics of Pseudomonas aeruginosa populations in lungs of cystic fibrosis patients. Infect Immun 75: 22142224.
  • Johansen, H.K., and Hoiby, N. (1992) Seasonal onset of initial colonisation and chronic infection with Pseudomonas aeruginosa in patients with cystic fibrosis in Denmark. Thorax 47: 109111.
  • Kessler, B., De Lorenzo, V., and Timmis, K.N. (1992) A general system to integrate lacZ fusions into the chromosomes of gram-negative eubacteria: regulation of the Pm promoter of the TOL plasmid studied with all controlling elements in monocopy. Mol Gen Genet 233: 293301.
  • Koch, A.K., Kappeli, O., Fiechter, A., and Reiser, J. (1991) Hydrocarbon assimilation and biosurfactant production in Pseudomonas aeruginosa mutants. J Bacteriol 173: 42124219.
  • Koch, C. (2002) Early infection and progression of cystic fibrosis lung disease. Pediatr Pulmonol 34: 232236.
  • Koch, C., and Hoiby, N. (1993) Pathogenesis of cystic fibrosis. Lancet 341: 10651069.
  • Kolter, R. (1999) Growth in studying the cessation of growth. J Bacteriol 181: 697699.
  • Lee, B., Haagensen, J.A., Ciofu, O., Andersen, J.B., Hoiby, N., and Molin, S. (2005) Heterogeneity of biofilms formed by nonmucoid Pseudomonas aeruginosa isolates from patients with cystic fibrosis. J Clin Microbiol 43: 52475255.
  • Linares, J.F., Lopez, J.A., Camafeita, E., Albar, J.P., Rojo, F., and Martinez, J.L. (2005) Overexpression of the multidrug efflux pumps MexCD-OprJ and MexEF-OprN is associated with a reduction of type III secretion in Pseudomonas aeruginosa. J Bacteriol 187: 13841391.
  • Lizewski, S.E., Schurr, J.R., Jackson, D.W., Frisk, A., Carterson, A.J., and Schurr, M.J. (2004) Identification of AlgR-regulated genes in Pseudomonas aeruginosa by use of microarray analysis. J Bacteriol 186: 56725684.
  • De Lorenzo, V., and Timmis, K.N. (1994) Analysis and construction of stable phenotypes in gram-negative bacteria with Tn5- and Tn10-derived minitransposons. Methods Enzymol 235: 386405.
  • Mahenthiralingam, E., Campbell, M.E., and Speert, D.P. (1994) Nonmotility and phagocytic resistance of Pseudomonas aeruginosa isolates from chronically colonized patients with cystic fibrosis. Infect Immun 62: 596605.
  • Martin, D.W., Schurr, M.J., Mudd, M.H., Govan, J.R., Holloway, B.W., and Deretic, V. (1993) Mechanism of conversion to mucoidy in Pseudomonas aeruginosa infecting cystic fibrosis patients. Proc Natl Acad Sci USA 90: 83778381.
  • Martin, R.R., Marshall, V.D., Sokatch, J.R., and Unger, L. (1973) Common enzymes of branched-chain amino acid catabolism in Pseudomonas putida. J Bacteriol 115: 198204.
  • Mathee, K., Ciofu, O., Sternberg, C., Lindum, P.W., Campbell, J.I., Jensen, P., et al. (1999) Mucoid conversion of Pseudomonas aeruginosa by hydrogen peroxide: a mechanism for virulence activation in the cystic fibrosis lung. Microbiology 145: 13491357.
  • Morici, L.A., Carterson, A.J., Wagner, V.E., Frisk, A., Schurr, J.R., Honer zu Bentrup, K., et al. (2007) Pseudomonas aeruginosa AlgR represses the Rhl quorum-sensing system in a biofilm-specific manner. J Bacteriol 189: 77527764.
  • Palmer, K.L., Aye, L.M., and Whiteley, M. (2007) Nutritional cues control Pseudomonas aeruginosa multicellular behavior in cystic fibrosis sputum. J Bacteriol 189: 80798087.
  • Pamp, S.J., and Tolker-Nielsen, T. (2007) Multiple roles of biosurfactants in structural biofilm development by Pseudomonas aeruginosa. J Bacteriol 189: 25312539.
  • Pedersen, S.S. (1992) Lung infection with alginate-producing, mucoid Pseudomonas aeruginosa in cystic fibrosis. APMIS Suppl 28: 179.
  • Rahme, L.G., Stevens, E.J., Wolfort, S.F., Shao, J., Tompkins, R.G., and Ausubel, F.M. (1995) Common virulence factors for bacterial pathogenicity in plants and animals. Science 268: 18991902.
  • Rhodius, V.A., Suh, W.C., Nonaka, G., West, J., and Gross, C.A. (2006) Conserved and variable functions of the sigmaE stress response in related genomes. Plos Biol 4: e2.
  • Rohwer, F., and Edwards, R. (2002) The Phage Proteomic Tree: a genome-based taxonomy for phage. J Bacteriol 184: 45294535.
  • Rowen, D.W., and Deretic, V. (2000) Membrane-to-cytosol redistribution of ECF sigma factor AlgU and conversion to mucoidy in Pseudomonas aeruginosa isolates from cystic fibrosis patients. Mol Microbiol 36: 314327.
  • Sanchez, P., Linares, J.F., Ruiz-Diez, B., Campanario, E., Navas, A., Baquero, F., and Martinez, J.L. (2002) Fitness of in vitro selected Pseudomonas aeruginosa nalB and nfxB multidrug resistant mutants. J Antimicrob Chemother 50: 657664.
  • Schnabel, H., and Schnabel, R. (1990) An organ-specific differentiation gene, pha-1, from Caenorhabditis elegans. Science 250: 686688.
  • Smith, E.E., Buckley, D.G., Wu, Z., Saenphimmachak, C., Hoffman, L.R., D'Argenio, D.A., et al. (2006) Genetic adaptation by Pseudomonas aeruginosa to the airways of cystic fibrosis patients. Proc Natl Acad Sci USA 103: 84878492.
  • Spencer, R.C. (1996) An 8 year Microbe Base survey of the epidemiology, frequency and antibiotic susceptibility of Pseudomonas aeruginosa hospital isolates in the United Kingdom. J Antimicrob Chemother 37: 295301.
  • Stover, C.K., Pham, X.Q., Erwin, A.L., Mizoguchi, S.D., Warrener, P., Hickey, M.J., et al. (2000) Complete genome sequence of Pseudomonas aeruginosa PA01, an opportunistic pathogen. Nature 406: 959964.
  • Tan, M.W., Rahme, L.G., Sternberg, J.A., Tompkins, R.G., and Ausubel, F.M. (1999) Pseudomonas aeruginosa killing of Caenorhabditis elegans used to identify P. aeruginosa virulence factors. Proc Natl Acad Sci USA 96: 24082413.
  • Tart, A.H., Blanks, M.J., and Wozniak, D.J. (2006) The AlgT-dependent transcriptional regulator AmrZ (AlgZ) inhibits flagellum biosynthesis in mucoid, nonmotile Pseudomonas aeruginosa cystic fibrosis isolates. J Bacteriol 188: 64836489.
  • Terry, J.M., Pina, S.E., and Mattingly, S.J. (1991) Environmental conditions which influence mucoid conversion Pseudomonas aeruginosa PAO1. Infect Immun 59: 471477.
  • Thomsen, L.E., Slutz, S.S., Tan, M.W., and Ingmer, H. (2006) Caenorhabditis elegans is a model host for Listeria monocytogenes. Appl Environ Microbiol 72: 17001701.
  • Wiehlmann, L., Wagner, G., Cramer, N., Siebert, B., Gudowius, P., Morales, G., et al. (2007) Population structure of Pseudomonas aeruginosa. Proc Natl Acad Sci USA 104: 81018106.
  • Wolfgang, M.C., Kulasekara, B.R., Liang, X., Boyd, D., Wu, K., Yang, Q., et al. (2003) Conservation of genome content and virulence determinants among clinical and environmental isolates of Pseudomonas aeruginosa. Proc Natl Acad Sci USA 100: 84848489.
  • Wood, L.F., and Ohman, D.E. (2009) Use of cell wall stress to characterize sigma 22 (AlgT/U) activation by regulated proteolysis and its regulon in Pseudomonas aeruginosa. Mol Microbiol 72: 183201.
  • Wu, W., Badrane, H., Arora, S., Baker, H.V., and Jin, S. (2004) MucA-mediated coordination of type III secretion and alginate synthesis in Pseudomonas aeruginosa. J Bacteriol 186: 75757585.
  • Wylie, J.L., and Worobec, E.A. (1995) The OprB porin plays a central role in carbohydrate uptake in Pseudomonas aeruginosa. J Bacteriol 177: 30213026.
  • Yang, L., Haagensen, J.A., Jelsbak, L., Johansen, H.K., Sternberg, C., Hoiby, N., and Molin, S. (2008) In situ growth rates and biofilm development of Pseudomonas aeruginosa populations in chronic lung infections. J Bacteriol 190: 27672776.
  • Zinser, E.R., and Kolter, R. (1999) Mutations enhancing amino acid catabolism confer a growth advantage in stationary phase. J Bacteriol 181: 58005807.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Table S1. Gene expression (all) – fold changes.

Table S2. Gene expression (all) – P-values.

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

EMI_2211_sm_tS1.csv465KSupporting info item
EMI_2211_sm_tS2.csv448KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.