SEARCH

SEARCH BY CITATION

Summary

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

A hallmark of the biofilm architecture is the presence of microcolonies. However, little is known about the underlying mechanisms governing microcolony formation. In the pathogen Pseudomonas aeruginosa, microcolony formation is dependent on the two-component regulator MifR, with mifR mutant biofilms exhibiting an overall thin structure lacking microcolonies, and overexpression of mifR resulting in hyper-microcolony formation. Using global transcriptomic and proteomic approaches, we demonstrate that microcolony formation is associated with stressful, oxygen-limiting but electron-rich conditions, as indicated by the activation of stress response mechanisms and anaerobic and fermentative processes, in particular pyruvate fermentation. Inactivation of genes involved in pyruvate utilization including uspK, acnA and ldhA abrogated microcolony formation in a manner similar to mifR inactivation. Moreover, depletion of pyruvate from the growth medium impaired biofilm and microcolony formation, while addition of pyruvate significantly increased microcolony formation. Addition of pyruvate to or expression of mifR in lactate dehydrogenase (ldhA) mutant biofilms did not restore microcolony formation, while addition of pyruvate partly restored microcolony formation in mifR mutant biofilms. In contrast, expression of ldhA in mifR::Mar fully restored microcolony formation by this mutant strain. Our findings indicate the fermentative utilization of pyruvate to be a microcolony-specific adaptation of the P. aeruginosa biofilm environment.


Introduction

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

Pseudomonas aeruginosa is responsible for a wide array of persistent infections including those of non-healing wounds and the lungs of cystic fibrosis sufferers. Within the cystic fibrosis lung, P. aeruginosa forms biofilms, defined as complex surface-attached communities encased in an extracellular matrix composed of exopolysaccharides, DNA and proteins. A hallmark of P. aeruginosa biofilms is the presence of complex multicellular aggregates or microcolonies, the formation of which has been observed both in vitro and in vivo (Lam et al., 1980; Høiby et al., 2001; Sauer et al., 2002; Davey et al., 2003; Garcia-Medina et al., 2005; Sriramulu et al., 2005) and has been associated with DNA release and elevated mutation rates (Allesen-Holm et al., 2006; Conibear et al., 2009). Iron has been demonstrated to control microcolony formation, as P. aeruginosa mutants inactivated in the high-affinity pyoverdine iron acquisition system only form thin unstructured biofilms even when grown in iron-sufficient medium (Banin et al., 2005). While the findings implicated iron as a signal for development of mushroom-like microcolonies, Yang et al. (2007) demonstrated microcolony formation to be favoured in media with low iron availability, with increasing iron concentrations resulting in decreased microcolony formation and DNA release.

Microcolony formation coincides with the formation of steep oxygen and nutrient gradients, with cells located deep within biofilm structures experiencing stressful, growth-limiting conditions (Anderl et al., 2003; Walters et al., 2003; Borriello et al., 2004). Microelectrode analyses have revealed that, while the concentration of oxygen at the surface of the biofilm is similar to that of the bulk fluid, oxygen levels drop rapidly towards the interior of biofilms with the centre of microcolonies being essentially anaerobic (de Beer et al., 1994; Stoodley et al., 1994; 1997; Rasmussen and Lewandowski, 1998; Rani et al., 2007). Additional gradients exist in the biofilm environment with respect to microbial waste products, sulphide and hydrogen ions (pH), which may accumulate within the depths of the biofilm (de Beer et al., 1994; Stoodley et al., 1994; 1997; Rasmussen and Lewandowski, 1998; Rani et al., 2007).

The resident bacterial population has been demonstrated to adjust to these steep gradients through various approaches including the modulation of metabolic rates, dormancy, stress responses and mutation rates (Stewart, 1996; Anderl et al., 2000; Walters et al., 2003; Borriello et al., 2004; Lenz et al., 2008; Pérez-Osorio et al., 2010). Selected metabolic pathways have been associated with biofilm formation, serving both as contributing factors and adaptations to the changing biofilm microenvironment. For instance, availability of amino acids appears to contribute to biofilm formation, as inactivation of the stationary phase sigma factor RpoS has been shown to enhance biofilm formation by alleviating potential amino acid limitation (Shirtliff et al., 2002). Inactivation of the catabolite repression control (Crc) protein, involved in carbon metabolism regulation and control of type IV pili gene expression, resulted in the formation of cellular monolayers devoid of microcolonies typically observed during normal biofilm development (O'Toole et al., 2000). Furthermore, under conditions of oxygen absence or limitation, P. aeruginosa is able to respire nitrate or nitrite through the process of denitrification, which sequentially reduces nitrate to nitrogen gas through the action of four reductases (Carlson et al., 1982; Carlson and Ingraham, 1983; Davies et al., 1989). The activation/upregulation of the components of the denitrification pathway has been repeatedly observed within in vitro biofilms and during persistent P. aeruginosa infections, with sensing and processing of nitrate and other intermediate forms playing an essential role in the establishment, maintenance, resistance and dispersal of biofilms in vitro and in vivo (Hassett et al., 2002; 2009; Borriello et al., 2004; Barraud et al., 2006; Filiatrault et al., 2006; Alvarez-Ortega and Harwood, 2007; Toyofuku et al., 2007; Van Alst et al., 2007; Schobert and Jahn, 2010). Many environments, however, do not have sufficient nitrate present to drive nitrate respiration. Under conditions of oxygen limitation in the absence of nitrate, growth is driven by oxygen respiration, done using high-affinity terminal oxidases including cbb3-1 and cbb3-2 (Comolli and Donohue, 2004; Alvarez-Ortega and Harwood, 2007).

While we have gained much insight into the P. aeruginosa biofilm environment by taking whole-system approaches and by distinguishing top versus bottom layers or oxic versus anoxic zones (Lenz et al., 2008), information regarding the conditions specifically associated with the formation of microcolonies remains limited. Here, we asked whether P. aeruginosa biofilm microcolony formation was governed by specific regulatory mechanisms and metabolic pathways. To address this question, we made use of the distinct phenotypes of the mifR-deficient and -overexpressing strains as a model to study the processes associated with microcolony formation (Fig. 1). The enhancer binding protein MifR is part of a previously identified two-component system MifSR, found to be required for the transition to the latest stages of biofilm maturation by P. aeruginosa PAO1, with mifR mutant biofilms exhibiting an overall thin structure lacking microcolonies (Petrova and Sauer, 2009). Conversely, overexpression of mifR resulted in enhanced microcolony formation when compared to the wild type (Petrova and Sauer, 2009).

figure

Figure 1. The role of MifSR in microcolony formation is strain-independent.

A. Microcolony diameter was assessed from P. aeruginosa PAO1 and PA14 wild type and strains inactivated in or overexpressing mifR following 144 h of growth in flow cells using 1/20-diluted LB medium. Experiments were performed in triplicate with at least five images acquired for each strain per replicate. *Significantly different from the values for the respective wild type strain (P ≤ 0.01).

B. mifR-overexpressing 144-h-old biofilms produce microcolonies that are visible with the unaided eye. Dashed white bar = 1 cm.

C. Brightfield images of P. aeruginosa PAO1, ΔmifR and PAO1/pJN-mifR following 48 and 144 h of growth in flow cells, demonstrating cluster formation by the wild type following 48 h of growth, but lack of microcolony formation following 144 h of growth in mifR::Mar biofilms. Overexpression of mifR correlates with increased microcolony formation. White bars = 100 μm.

D. Inactivation of mifR and mifS in the PA14 parental background correlates with lack of microcolony formation, while mifR overexpression enhances cellular aggregate formation. Insets show cross sections of wild type and mutant cell clusters (< 50 μm) and microcolonies (> 50 μm) near the substratum. Biofilms were grown for 144 h under continuous flowing conditions, stained with BacLight LIVE/DEAD stain, and visualized by CSLM. White bars = 100 μm.

Download figure to PowerPoint

Results

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

To initiate studies on differential biofilm microcolony development (Fig. 1A–C), we first determined whether the role of MifR in microcolony formation was strain-independent. Inactivation of mifR (mifR::Mar) impaired microcolony formation by P. aeruginosa PA14 biofilms in a manner similar to that observed for PAO1 (Fig. 1A and D). Overall, mifR mutant biofilms only demonstrated the presence of cell clusters which we defined as having a diameter of less than 50 μm and a height of less than 30 μm. Cell clusters furthermore differed from microcolonies by their loose, less dense appearance at the substratum (see inset, Fig. 1D). Moreover, inactivation of mifS (mifS::Mar), the cognate sensor kinase of mifR, resulted in a biofilm formation defect comparable to that of the mifR mutant strain (Fig. 1D), further underscoring the role of the TCS MifSR in microcolony formation. Complementation of the mifR mutant (PA14-mifR::Mar/pMJT-mifR) restored microcolony formation and the overall architecture to wild type levels (Fig. 1A and D). In addition, overexpression of mifR resulted in significantly increased microcolony formation in both strains (Fig. 1A), with increased microcolony formation correlating with increased c-di-GMP levels, but reduced phosphodiesterase activity compared to wild type biofilms (Fig. S1A and B, not shown). Furthermore, inactivation and overexpression of mifR in PAO1 correlated with a less than twofold decreased or increased Psl polysaccharide levels respectively. The effect of mifR expression on Psl levels was biofilm-specific (Fig. S1C).

Genome-wide transcriptional profiling of wild type P. aeruginosa and an isogenic mifR mutant

Using differential expression of mifR as a model for variations in biofilm microcolony development (Fig. 1), we next addressed the question of how MifR controls microcolony formation by assessing the full range of genes regulated by MifR in a biofilm by comparing the transcriptomes of P. aeruginosa ΔmifR to wild type biofilms. Among the 5678 genes surveyed, 1500 genes showed significantly altered (either reduced or increased) expression in P. aeruginosa ΔmifR as compared to its isogenic wild type parent (Tables S3 and S4). Selected genes were confirmed by qRT-PCR (Table S5). Consistent with the finding of mifR expression resulting in increased c-di-GMP levels (Fig. S1A), wild type biofilm formation correlated with decreased expression of genes encoding potential phosphodiesterases (Table S3; PA1181, PA3258). No significant difference in psl gene expression between wild type and mifR mutant biofilms was observed. Moreover, many of the genes exhibiting MifR-dependent changes in expression encode proteins involved in metabolic, biosynthetic and energetic processes (Fig. S2A). In parallel with mifR-dependent expression of metabolic genes, we also noted genes encoding membrane proteins as well as those involved in transport of small molecules and secretion/export to have increased expression (Fig. S2A, Table S3). Similarly, genes encoding secreted factors and the type II and III secretion systems were significantly increased in the mifR mutant strain (Fig. S2A, Table S3). Considering that denitrification has been repeatedly observed in in vitro biofilms and during persistent P. aeruginosa infections, it is of interest to note that no consistent changes in the expression of denitrification genes were observed (Fig. S2B). Instead, increased expression of the cbb3-2 oxidase genes (PA1555–PA1557; only two of the three cbb3-1 oxidase genes were differentially expressed, Table 1) in wild type biofilms was detected. The respective genes have been previously reported to increase under microaerophilic conditions (Comolli and Donohue, 2004; Alvarez-Ortega and Harwood, 2007). Moreover, genes involved in energy metabolism including anaerobic metabolism and fermentative pathways using arginine (ArcDABC) and pyruvate (Eschbach et al., 2004) were expressed significantly less in mifR mutant biofilms (Fig. S2C, Tables 1 and S4). Similarly, genes previously associated with growth under oxygen-limiting conditions including those encoding for two universal stress response (USP) proteins (PA1789, UspL and PA4328, UspM) and ANR were significantly downregulated (Tables 1 and S4).

Table 1. Differentially expressed genes in wild type and mifR-deficient biofilms indicating oxygen-limiting conditions in P. aeruginosa biofilmsa
Pyruvate fermentationbOxygen-limiting conditionsArginine fermentation
GeneFold changeGeneFold changeGeneFold change
  1. a

    Respective transcripts have previously been reported to be affected by growth under limited-oxygen or anaerobic conditions (Comolli and Donohue, 2004; Eschbach et al., 2004; Wu et al., 2005; Filiatrault et al., 2006; Alvarez-Ortega and Harwood, 2007; Platt et al., 2008).

  2. b

    Based on Eschbach et al. (2004).

  3. Fold changes refer to ΔmifR versus wild type.

PA0927, ldhA−6.73PA1580, gltA−24.24PA5170, arcD−6.68
PA1400 (PycA)5.12PA0795, prpC−13.64PA5171, arcA−3.38
PA0854, fumC2−9.55PA1562, acnA−39.69PA5172, arcB−3.18
PA5046 (Mdh)−3.38PA1787, acnB−10.05PA5173, arcC−5.28
PA4333 (Fum)−8.33PA2624, idh−5.07  
PA0835 (Pta)−9.38PA2623, icd−4.44  
PA1984 (Ad-DH)−4.52PA1789, uspL−7.24  
PA0836 (AckA)−4.40PA4328, uspM−9.05  
PA5016, aceF−2.55PA1553, ccoQ1−24.7  
  PA1554, ccoN1−8.5  
  PA1555, ccoP2−6.9  
  PA1556, ccoQ2−6.1  
  PA1557, ccoN2−11.1  

Proteomic approach indicates a link between anaerobic conditions, stress and microcolony formation

We furthermore analysed the proteomes of PAO1, ΔmifR and PAO1/pJN-mifR biofilms characterized by wild type, impaired and hyper-microcolony formation, respectively, by 2D/PAGE. Subsequent identification of differentially expressed proteins by LC-MS-MS revealed MifR-dependent alterations in post-transcriptional/translational modification processes (ClpA, ClpB, PA5201) and DNA replication and repair (MutS, ParC and ParE). Protein levels of the DNA repair protein MutS were elevated in mifR mutant biofilms and undetectable upon mifR overexpression (Table 2), indicating MutS to be negatively impacted by MifR and microcolony formation. The findings are in agreement with previous observations indicating that inactivation of mutS correlates with increased biomass accumulation and cellular aggregate formation during biofilm and chronic growth (Conibear et al., 2009). Overall, of the 18 identified proteins, nine were also identified by DNA microarray analysis (Table 2). Consistent with the transcriptome analysis, proteins reduced or absent in ΔmifR biofilms as compared to wild type biofilms were identified as being associated with anaerobic growth. These included one of the two P. aeruginosa aconitases, aconitate hydratase I (AcnA), the two stress response proteins UspK and UspN, as well as outer membrane protein OprE whose concentration has been previously found to increase during anaerobic growth (Yoon et al., 2002; Boes et al., 2006; Schreiber et al., 2006; Alvarez-Ortega and Harwood, 2007) (Table 2). The respective proteins increased in biofilms overexpressing mifR (Table 2). The expression of the gene PA3329 and the level of the respective protein have been previously reported as reduced under oxygen-limiting conditions (Wu et al., 2005; Petrova and Sauer, 2012). Consistent with this observation, PA3329 was detectable in a ΔmifR mutant but absent or not detectable in PAO1 or a strain overexpressing mifR (Table 2). Likewise, CbpD was previously described to be reduced under anaerobic conditions (Wu et al., 2005). Here, CbpD levels were increased threefold in a mifR mutant compared to the wild type. There were no differences in ChpD protein levels between wild type and a strain that overexpressed mifR. The findings suggested wild type biofilm formation correlated with oxygen-limiting conditions.

Table 2. Differentially produced proteins in mifR-deficient and -overexpressing biofilms
Locusa ProteinFold change in 6d biofilmsbAnaerobic growthc
ΔmifRPAO1/pJN-mifR
  1. a

    PA numbers in bold were also detected by DNA microarray analysis (see Tables S3 and S4).

  2. b

    Fold change in protein abundance relative to 6d PAO1 biofilms; 0 indicates that, while detected in PAO1 biofilms, the protein was not detected in biofilms of the indicated strain; dn indicates that protein spot detection is unique for the indicated strain.

  3. c

    Previous reports of proteins or respective transcripts being affected by growth under limited-oxygen or anaerobic conditions; [UPWARDS ARROW] indicates positive effect of anaerobic growth; [DOWNWARDS ARROW] indicates negative effect of anaerobic growth; references are listed in parentheses.

  4. d

    Two protein spots distinct in molecular weight and/or isoelectric point were identified as the indicated protein, suggesting the presence of multiple forms of the protein via post-translational modification events.

PA0291oprEOuter membrane porin OprE0.230.46[UPWARDS ARROW] (Wu et al., 2005)
PA0852cbpDChitin-binding protein CbpD precursor3.31.19[DOWNWARDS ARROW] (Wu et al., 2005)
PA0963aspSAspartyl-tRNA synthetase2.531.14 
PA1156nrdAClass Ia ribonucleotide reductase01.35[UPWARDS ARROW] (Wu et al., 2005; Platt et al., 2008)
PA1562acnAAconitate hydratase 109.18[UPWARDS ARROW] (Platt et al., 2008)
PA1584sdhBSuccinate dehydrogenase (B subunit)dn 
PA1588sucCSuccinyl-CoA synthetase beta chain01.42 
PA2386pvdAL-ornithine N5-oxygenase0.763.356 
PA2620clpAATP-binding protease component ClpA07.97 
PA3309uspKUniversal stress response protein UspKdn[UPWARDS ARROW] (Yoon et al., 2002; Boes et al., 2006; Schreiber et al., 2006; Alvarez-Ortega and Harwood, 2007)
PA3329 Hypothetical proteindn[DOWNWARDS ARROW] (Wu et al., 2005; Filiatrault et al., 2006)
PA3620mutSDNA mismatch repair protein MutS1.930 
PA3687 Phosphoenolpyruvate carboxylasedn 
PA4352uspNUniversal stress response protein UspN0.212[UPWARDS ARROW] (Boes et al., 2006; Schreiber et al., 2006; Alvarez-Ortega and Harwood, 2007)
PA4542 ClpB proteindn 
PA4964parCTopoisomerase IV subunit A5.551.53 
PA4967parETopoisomerase IV subunit B0.491.14 
PA5201 Conserved hypothetical proteind1.30.24 
   0.591.7 

A mifR mutant is defective in growth under anoxic conditions

Considering the large number of metabolic genes and genes involved in anaerobiosis that are significantly less expressed in mifR mutant biofilms, we determined whether the mifR mutant displays any growth defects (not shown). No differences in growth were noted in Lennox Broth (LB) or Vogel-Bonner minimal medium (VBMM) broths when grown with aeration (not shown). To determine whether the observed MifR-dependent changes in the production of anaerobiosis-related proteins are the downstream effect of altered biofilm architecture and biomass accumulation, which have been previously associated with oxygen limitation, growth of PAO1 and mifR mutant under oxygen-limited conditions using soft agar was assessed. Reduced growth of ΔmifR was observed in LB medium under oxygen-limiting conditions, with the reduction in growth being more profound when grown in VBMM than in LB medium (Figs 2A and S3). Oxygen-limiting conditions within the soft agar tubes were confirmed by methylene blue staining and growth of the anaerobic bacterium Clostridium sporogenes (Fig. S3D and E). Addition of arginine or pyruvate to the growth medium partly restored growth (Figs 2A and S3), while addition of nitrate (25 mM) restored growth under the conditions tested (Figs 2B and S3F), indicating that ΔmifR is not defective in denitrification but instead reduced in growth under oxygen-limiting conditions. However, no growth by P. aeruginosa mifR mutant and its parental strain was observed under strictly anaerobic conditions in the absence or presence of pyruvate. Addition of nitrate, however, enabled growth under strictly anaerobic conditions (not shown). The findings corroborate the microarray and proteomic data and indicate that the changes observed are MifR-dependent and not due to differences in the environmental conditions created by the absence or presence of microcolonies.

figure

Figure 2. Growth of the mifR mutant is reduced under conditions of oxygen limitation in the absence of nitrate.

A. PA14 and the mifR mutant grown in soft agar tubes. LB and LB supplemented with 25 mM pyruvate was used as growth medium. An initial inoculum of 106 cells was used and CFU were determined once methylene blue had become colourless.

B. Growth of PA14 and mifR mutant in soft agar tubes grown in the presence of 10 mM nitrate.

C. Sampling depth to determine viability of P. aeruginosa grown in soft agar tubes is indicated by an arrow. Growth was determined by viability counts (CFU ml−1). Error bars indicate standard deviation. Experiments were carried out at least in triplicate.

Download figure to PowerPoint

The universal stress response protein UspK contributes to microcolony formation

Transcriptomic and proteomic studies revealed a decreased expression and protein levels of four universal stress response proteins (UspL, UspK, UspM, UspN) in biofilms lacking mifR. The production of these proteins has been previously shown to be increased upon biofilm formation and conditions of oxygen limitation both in vitro and in vivo (Boes et al., 2006; 2008; Schreiber et al., 2006; Waite et al., 2006; Alvarez-Ortega and Harwood, 2007; Mikkelsen et al., 2009). In particular, UspK has been implicated to play a role upon fermentative growth on pyruvate (Schreiber et al., 2006), while UspN has been linked to survival following anaerobic denitrifying growth (Boes et al., 2006; 2008). As both UspK and UspN positively correlated with MifR levels, we asked whether the proteins play a role in microcolony formation. Biofilms formed by a uspK mutant (uspK::Mar) were composed of sixfold less biofilm biomass and more than twofold reduced thickness when compared to the wild type (Fig. 3, Table 3). Overall, uspK mutant biofilms resembled those of the mifR mutant in that they lacked large, dense microcolonies typically seen in wild type biofilms (Figs 1 and 3). Instead, only loose clusters measuring less than 30 μm in height were detected. Complementation (uspK::Mar/pMJT-uspK) restored biofilm formation to wild type levels (not shown). While expression of mifR in mifR mutant biofilms restored microcolony and biofilm formation to wild type levels (Fig. 1), expression of mifR in uspK mutant biofilms enabled microcolony formation but failed to restore biofilm formation (Fig. 3). While insertional inactivation of uspN (uspN::Mar) correlated with an overall decrease in biofilm biomass accumulation and reduction in biofilm height, uspN mutant biofilms still demonstrated the typical biofilm architecture including the presence of dense microcolonies (Fig. 3, Table 3). This is contrast to inactivation of uspK, which lacked microcolonies (Fig. 3, Table 3).

figure

Figure 3. UspK, but not UspN, is required for microcolony formation. Confocal images of biofilms of the indicated strains acquired following growth for 144 h under continuous flowing conditions using 1/20-diluted LB medium. Insets show cross sections of wild type and mutant cell clusters (< 50 μm) and microcolonies (> 50 μm) near the substratum. Biofilms were stained with BacLight LIVE/DEAD stain. White bars = 100 μm.

Download figure to PowerPoint

Table 3. COMSTAT analysisa of P. aeruginosa wild type and mutant biofilm structure
StrainTotal biomass (μm3/μm2)Substratum coverage (%)Average thickness (μm)Maximum thickness (μm)Roughness coefficient (dimensionless)
  1. a

    comstat analysis was carried out from biofilms grown in triplicate using at least six images per replicate.

  2. * Significantly different from 144-h-old PA14; P ≤ 0.05 as determined by ANOVA.

PA1438.00 (± 17.09)67.05 (± 24.80)35.21 (± 18.59)89.29 (± 29.56)0.84 (± 0.38)
PA14/pJN-ldhA35.97 (± 8.00)86.16 (± 11.11)34.25 (± 7.76)95.88 (± 30.71)0.66 (± 0.20)
mifR10.46 (± 7.87)*29.90 (± 18.19)*9.36 (± 7.35)*48.40 (± 22.50)*1.38 (± 0.35)*
mifR/pMJT-mifR37.16 (± 10.27)85.08 (± 18.91)35.54 (± 12.34)87.03 (± 18.12)0.72 (± 0.28)
mifR/pMJT-uspK6.56 (± 3.07)*30.56 (± 12.86)*5.99 (± 2.98)*42.00 (± 17.93)*1.29 (± 0.40)*
mifR/pMJT-uspN7.16 (± 2.65)*30.45 (± 8.00)*6.35 (± 2.58)*41.00 (± 13.23)*1.22 (± 0.22)*
mifR/pMJT-ldhA41.63 (± 15.55)71.30 (± 19.83)41.79 (± 20.23)90.17 (± 31.72)0.58 (± 0.23)
mifS5.31 (± 4.04)*12.88 (± 6.52)*4.84 (± 3.86)*49.08 (± 17.37)*1.72 (± 0.13)*
uspK8.08 (± 4.01)*29.08 (± 16.19)*7.24 (± 4.02)*37.85 (± 8.73)*1.31 (± 0.29)*
uspK/pMJT-mifR14.68 (± 5.92)*54.27 (± 28.02)13.22 (± 5.74)*62.33 (± 9.26)*1.13 (± 0.30)*
uspN14.71 (± 6.69)*37.57 (± 20.03)*14.00 (± 6.86)*73.82 (± 13.67)*1.29 (± 0.40)*
uspN/pMJT-mifR20.10 (± 7.81)*41.53 (± 25.35)21.61 (± 9.83)*85.67 (± 27.47)1.14 (± 0.39)*
acnA3.14 (± 0.50)*25.63 (± 11.51)*3.398 (± 0.39)*19.37 (± 3.39)*1.06 (± 0.11)*
acnA/pMJT-mifR14.46 (± 8.45)*20.24 (± 7.25)*16.64 (± 12.81)*76.35 (± 21.33)1.36 (± 0.34)*
ldhA10.63 (± 3.61)*49.96 (± 13.47)12.26 (± 6.22)*33.71 (± 15.61)0.90 (± 0.30)
ldhA/pMJT-mifR6.08 (± 4.79)*34.69 (± 25.30)*6.09 (± 5.90)*26.08 (± 19.10)*1.22 (± 0.45)*

Considering that expression of UspK and UspN correlated with MifR levels, we next asked whether the decrease in Usp levels observed in mifR mutant biofilms was the result, rather than the cause, of impaired microcolony formation. We hypothesized that if the microcolony formation defect observed in the mifR mutant were based on the downregulation of Usp systems, complementation with the respective Usp genes/proteins would restore wild type biofilm formation. However, expression of uspK or uspN in the mifR mutant did not enhance biofilm formation or restore microcolony formation, as mifR::Mar/pMJT-uspK and mifR::Mar/pMJT-uspN demonstrated biofilm architecture and cluster formation comparable to mifR mutant biofilms following 144 h of growth (Figs 1 and 3, Table 3). Instead, expression of mifR in the uspN and uspK mutants resulted in significantly enhanced microcolony formation (although without restoring uspN and uspK mutant biofilm phenotypes to wild type levels) (Fig. 3, Table 3). The findings indicated that both UspK and UspN act downstream of MifR (with MifR likely not directly controlling uspK and uspN expression), and are probably not the direct cause of the MifR-dependent phenotypes. Alternatively, it is likely that differences in Usp protein levels are independent of MifR and instead suggest differences in oxygen levels in wild type versus mutant biofilms. Moreover, while our findings suggest no role of UspN in microcolony formation, UspK appeared to be required for microcolony formation.

Extracellular pyruvate promotes biofilm microcolony formation

UspK has been previously shown to be essential for survival on pyruvate under conditions of oxygen limitation (Schreiber et al., 2006), with P. aeruginosa PA14 being able to release pyruvate, using a pyocyanin-dependent mechanism, and subsequently utilize pyruvate during stationary phase (Price-Whelan et al., 2007). As genes associated with pyruvate fermentation are downregulated in mifR mutant biofilms (Tables 3 and S4), we asked whether pyruvate utilization contributes to biofilm formation and the establishment of microcolonies by determining the effect of exogenously added pyruvate on wild type PA14 and mifR mutant biofilms by microscopy. Addition of pyruvate (0.1–1 mM) to the growth medium (diluted LB medium) resulted in enhanced microcolony formation by wild type biofilms with the average diameter of microcolonies increasing 1.6-fold in the presence of 1 mM pyruvate (Fig. 4). It is of interest to note that addition of 1 mM pyruvate did not enhance growth of P. aeruginosa grown planktonically under fully aerated conditions (Fig. S4A). However, the presence of 25 mM pyruvate as the sole carbon source supported some growth by P. aeruginosa (Fig. S4B). While pyruvate did not fully restore microcolony formation in the mifR mutant to wild type levels, addition of pyruvate resulted occasionally in the formation of small microcolonies that structurally resembled those observed in wild type biofilms (Fig. 4). In contrast, addition of nitrate to the growth medium had no apparent effect on the development of biofilm microcolonies (Fig. 4). Similar results were obtained when diluted VBMM medium was used (Fig. S5). Our finding of nitrate not stimulating biofilm formation is in contrast to previous observations. Considering that under the conditions tested, microcolonies did not appear to be fully anoxic as indicated by methylene blue staining (Fig. S6), it is likely that oxygen respiration is still occurring. This is consistent with the finding of increased expression of the cbb3-2 oxidase genes in wild type biofilms (Table 1).

figure

Figure 4. Extracellular pyruvate enhances P. aeruginosa biofilm microcolony formation.

A. PA14, mifR::Mar and ldhA::Mar were grown in 24-well plates in fivefold diluted LB in the presence or absence of 10 mM nitrate or 1 mM pyruvate for 5 days, after which time brightfield images were acquired. White bars = 100 μm.

B. Average diameter of cell clusters (< 50 μm) and microcolonies were determined using ProgRes CapturePro software. Experiments were performed in triplicate with at least 10 images acquired for each strain per replicate.

Download figure to PowerPoint

Depletion of pyruvate impairs microcolony formation

As increased availability of pyruvate coincided with enhanced microcolony formation, when grown in LB or VBMM medium, we next asked whether lack of pyruvate abrogated microcolony formation by enzymatically depleting pyruvate from the growth medium using pyruvate dehydrogenase (PyrDH). Under the conditions tested, the presence of pyruvate dehydrogenase and the appropriate cofactors (CoA, NADH) did not affect growth in broth when grown in microtitre plates, as determined by absorbance (not shown). However, presence of PyrDH during the early stages of biofilm development impaired microcolony formation (Fig. 5), with PyrDH-treated biofilms containing infrequent thin and unstructured clusters not exceeding 20 μm in diameter. In contrast, biofilms treated with heat-inactivated PyrDH or cofactors alone were similar to untreated biofilms and characterized by widespread microcolonies exceeding 150 μm in diameter (Fig. 5).

figure

Figure 5. Depletion of pyruvate impairs microcolony formation by P. aeruginosa.

A. Wild type PA14 biofilms were grown in 24-well plates for 5 days in the presence or absence of 1 mU pyruvate dehydrogenase, 1 μM β-NAD and 1 μM sodium Co-A. Untreated biofilms, as well as biofilms exposed to cofactors only, or those treated with heat-inactivated pyruvate dehydrogenase in the presence of cofactors were used as controls. White bars = 100 μm.

B. Average diameter of cell clusters (< 50 μm) and microcolonies were determined using ProgRes CapturePro software. Experiments were performed in triplicate with at least 10 images acquired for each strain per replicate.

Download figure to PowerPoint

Proteins involved in utilization of pyruvate are essential for microcolony formation

The findings strongly supported a requirement for pyruvate in the process of biofilm microcolony formation regardless of medium used. To determine whether pyruvate needs to be metabolized to enable microcolony formation, we asked whether enzymes involved in pyruvate utilization play a role in microcolony formation. In particular, we focused on NADH-dependent lactate dehydrogenase LdhA which was previously shown to be required for anaerobic pyruvate utilization and long-time survival on pyruvate (Eschbach et al., 2004) (Tables 1 and S4). In addition, we focused on aconitate hydratase AcnA, which plays a role in tricarboxylic acid cycle, glyoxylate bypass and acetyl-CoA assimilation (and only indirectly in pyruvate utilization) (Somerville et al., 1999; Winsor et al., 2009) (Tables 1 and S4). Inactivation of acnA resulted in impaired biofilm formation as indicated by biofilms being characterized by up to 10-fold reduced biofilm biomass and thickness as compared to wild type biofilms (Fig. 6, Table 3). Complementation restored biofilm formation to wild type levels (not shown). Similarly, expression of mifR in the acnA mutant did result in significantly enhanced biofilm and microcolony formation, with parameters including biomass and thickness increasing more than fourfold when compared to the acnA mutant alone (Fig. 6, Table 3). The findings strongly suggested that, while AcnA contributes to biofilm formation, AcnA acts downstream of MifR and is probably not the direct cause of the MifR- or pyruvate-dependent microcolony formation phenotypes. This is in part based on the finding of AcnA being required not only for pyruvate utilization but also for the tricarboxylic acid cycle, glyoxylate bypass and acetyl-CoA assimilation.

figure

Figure 6. AcnA and LdhA are required for microcolony formation.

A. Pathway showing pyruvate utilization. All products of the pyruvate fermentation pathway have been detected by HPLC analysis (Eschbach et al., 2004). The oxidative citric acid cycle enabling anaerobic respiration is highlighted by a grey box. Genes encoding enzymes involved in pyruvate utilization that were expressed in a MifR-dependent manner are underlined. Genes highlighted by a dark grey box (ldhA, acnA) were tested for their role in microcolony formation. Pyc, pyruvate carboxylase; Ldh, fermentative lactate dehydrogenase; Pdh, pyruvate dehydrogenase; Pta, phosphotransacetylase; Ad-DH, aldehyde dehydrogenase; AckA, acetate kinase; Mdh, malate dehydrogenase; Fum, fumarase; Frd, fumarate reductase.

B. Mutants inactivated in acnA and ldhA are impaired in biofilm and microcolony formation. Overexpression of mifR partly restores biofilm formation by the acnA mutant, but not by the ldhA mutant. In contrast, expression of ldhA restores mifR mutant biofilm and microcolony formation to wild type levels, while overexpression of ldhA results in increased microcolony formation. Insets show cross sections of wild type and mutant cell clusters (< 50 μm) and microcolonies (> 50 μm) near the substratum. Biofilms were grown for 144 h under continuous flowing conditions, stained with BacLight LIVE/DEAD stain, and visualized by CSLM. White bars = 100 μm.

C. Larger view of ldhA-expressing strains.

D. Microcolony diameter was assessed for PA14 wild type and mifR mutant strains overexpressing ldhA following 144 h of growth in flow cells using 1/20-diluted LB medium. Experiments were performed in triplicate with at least five images acquired for each strain per replicate. **Significantly different from the values for the respective strain not overexpressing ldhA (P ≤ 0.01).

E. NADH/NAD+ ratios for P. aeruginosa wild type and mutant biofilms inactivated in or overexpressing ldhA. Extraction and quantification of intracellular NADH and NAD+ were carried out as previously described (San et al., 2002; Price-Whelan et al., 2007).

F. NADH/NAD+ ratios obtained for P. aeruginosa wild type and mutant strains grown planktonically. **Significantly different from the values obtained for wild type (P ≤ 0.01).

Download figure to PowerPoint

Inactivation of ldhA resulted in fourfold reduced biofilm biomass accumulation and reduced thickness while complementation restored biofilm formation to wild type levels (Fig. 6B, Table 3). Overall, the architecture of ldhA mutant biofilms was similar to that observed for mifR mutant biofilms and was only composed of small clusters less than 20 μm in diameter. Furthermore, in contrast to the results observed for acnA::Mar, expression of mifR in ldhA mutant biofilms failed to restore biofilm and microcolony formation to wild type levels, with ldhA::Mar/pMJT-mifR demonstrating the same biofilm architecture as the ldhA mutant (Fig. 6B, Table 3). Taken together, these findings strongly support a role for LdhA and pyruvate fermentation in microcolony formation.

Pyruvate promotes microcolony formation in an LdhA-dependent manner

Our findings implicated exogenous pyruvate and lactate dehydrogenase, associated with pyruvate utilization under limited-oxygen conditions, to be required for microcolony formation. To determine whether the two were linked, we asked whether the biofilm-impaired phenotype of the ldhA mutant could be restored by the addition of pyruvate to the growth medium. We reasoned that if the effect of pyruvate on microcolony formation is not related to its fermentative utilization, addition of exogenous pyruvate to ldhA mutant biofilms should enhance biofilm formation and microcolony development, in a manner similar to that observed for the wild type or the mifR mutant. The ldhA mutant, however, did not respond to the addition of pyruvate (grown in diluted LB), with the strain demonstrating similar defects in biofilm formation in the absence and presence of exogenous pyruvate (Fig. 4). Similar results were obtained when VBMM medium was used (Fig. S5).

To further confirm a role of LdhA in microcolony formation, the effect of ldhA expression in biofilms was determined. Expression of ldhA in PA14, while not increasing the overall biofilm biomass, correlated with a significant increase in microcolony formation (Fig. 6B–D). The increase in microcolony formation was comparable to the increase observed upon pyruvate addition. Expression of ldhA in mifR mutant biofilms fully restored the biofilm architecture to wild type levels, resulting in a significant increase in biofilm biomass accumulation and more importantly, the formation of microcolonies exceeding > 150 μm in diameter (Fig. 6B–D). To determine whether LdhA and pyruvate fermentation play a role in redox balancing, NADH/NAD+ levels were determined. ldhA mutant biofilms demonstrated a significantly increased NADH/NAD+ ratio compared to wild type, while overexpression of ldhA significantly decreased the available NADH as indicated by a decreased NADH/NAD+ ratio compared to wild type biofilms (Fig. 6E). The finding suggested an inverse correlation between NADH/NAD+ ratio and microcolony formation. The presence of pyocyanin has been demonstrated to alter NADH/NAD+ ratios (Price-Whelan et al., 2007). To minimize the effect of pyocyanin, NADH/NAD+ ratio measurements were also carried out under planktonic growth conditions. Under the conditions tested, differential expression of ldhA resulted in an overall similar trend with respect to NADH/NAD+ ratios as those obtained under biofilm growth conditions, with ldhA mutant demonstrating increased NADH/NAD+ ratios and overexpression of ldhA correlating with decreased NADH/NAD+ ratios compared to wild type (Fig. 6F). Moreover, inactivation of mifR resulted in significantly increased NADH/NAD+ ratio compared to wild type, while expression of mifR or ldhA in mifR mutants restored the NADH/NAD+ ratio to wild type levels (Fig. 6F). Taken together, the findings suggested a requirement for pyruvate in biofilm microcolony formation, with the observed effects likely being mediated via the pyruvate fermentation pathway, probably as a means of redox balancing.

Discussion

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

In the present study, we investigated the environment and processes associated with the formation of biofilm microcolonies by comparing global transcriptomic and proteomic patterns of biofilms exhibiting different degrees of microcolony production represented by wild type, mifR mutant and mifR-overexpressing biofilms. Our findings are in agreement with previous reports linking biofilms and, in particular, microcolonies with oxygen-limiting growth conditions. Under conditions of oxygen limitation, growth of P. aeruginosa is predominantly driven by denitrification, in which nitrate or nitrite serves as the terminal electron acceptor. While this process is generally accepted as the major survival strategy in the oxygen-limited (but nitrate-rich) biofilm environment, our transcriptomic analysis of wild type and mifR mutant biofilms impaired in microcolony formation did not suggest differences in the expression of denitrification genes. Our findings imply nitrate respiration not to significantly contribute to the survival and growth within microcolonies. This is further supported by our observations of biofilm and large cellular aggregate formation in LB and VBMM media that are either nitrate-poor or lack nitrate (Figs 1 and 4). Instead, our findings strongly suggest a requirement for pyruvate fermentation to enable microcolony formation in environments that are oxygen-limited (and presumably nitrate-limited). P. aeruginosa has recently been shown to be capable of fermentatively utilizing pyruvate for survival under conditions of oxygen limitation in the absence of nitrite and nitrate (Eschbach et al., 2004; Schreiber et al., 2006). The process involves the conversion of pyruvate to lactate, acetate and/or succinate, with experimental evidence suggesting the lactate- and acetate-producing branches of the pathway to be predominating (Fig. 6A). Inactivation of lactate dehydrogenase (LdhA), which converts pyruvate to lactate and regenerates NAD+, severely impairs pyruvate fermentation and compromises survival on pyruvate under conditions of energy (electron) richness (Eschbach et al., 2004). Consistent with a role of pyruvate fermentation as a microcolony-specific adaptation mechanism, an ldhA mutant was impaired in biofilm microcolony formation with exogenous nitrate or pyruvate or overexpression of mifR being unable to restore microcolony formation (Figs 4-6). Previous findings indicated that in the absence of oxygen, pyruvate fermentation in P. aeruginosa does not allow robust growth. Likewise, here we demonstrate that pyruvate does not support growth under strictly anaerobic growth conditions. Moreover, while addition of 0.1–1 mM of pyruvate to 24-well grown P. aeruginosa is not sufficient to increase growth, we demonstrate that higher concentrations of pyruvate are capable of sustaining growth under aerobic and oxygen-limiting (but not anaerobic) conditions. The findings strongly suggest that pyruvate is being used as a means of redox balancing. Consistent with a role of LdhA in regenerating reducing equivalents under oxygen-limiting conditions, expression of ldhA in P. aeruginosa wild type correlated with a significant increase in biofilm biomass accumulation and microcolony formation as well as restoration of the mifR mutant biofilm phenotype to wild type levels (Fig. 6B–D) and decreased NADH/NAD+ ratios (Fig. 6E and F). The results furthermore suggest a requirement of redox balancing to enable microcolony formation in biofilms.

Pyruvate appears to be produced by the resident biofilm bacteria. A model recently proposed by Schobert and Jahn (2010) places P. aeruginosa biofilm cells within different niches, with metabolically active cells exposed to oxygen secreting pyruvate, which then diffuses into the anoxic zones to be utilized by cells residing within these layers. Consistent with this model is the pyruvate-dependent formation of microcolonies observed here, with depletion of extracellular pyruvate impairing biofilm development and abrogating microcolony formation (Fig. 5), while addition of exogenous pyruvate enhances biofilm development by specifically promoting microcolony formation (Figs 4 and S5). Additional observations support a role of pyruvate in microcolony formation. Pyruvate has been demonstrated to be released into the extracellular environment by P. aeruginosa PA14 during stationary phase in a manner dependent on the redox-active phenazine pyocyanin (Price-Whelan et al., 2007). Although phenazine production does not impact the ability of P. aeruginosa PA14 to attach to surfaces, mutants unable to synthesize phenazines including pyocyanin were impaired in microcolony formation (Dietrich et al., 2008; Ramos et al., 2010). Moreover, P. aeruginosa PA14 differs from PAO1 with respect to pyocyanin levels with PA14 producing more pyocyanin than PAO1 (Dietrich et al., 2006) as well as secreting more pyruvate (Price-Whelan et al., 2007). Increased pyocyanin and pyruvate release correlates with our observation of P. aeruginosa PA14 biofilms forming significantly larger microcolonies following 6 days of growth under continuous flowing conditions (Fig. 1) and earlier initiation of microcolony formation compared to PAO1 (Fig. S7).

In addition to oxygen limitation, our findings are in agreement with previous reports linking biofilms and in particular, microcolonies to stressful microenvironments. Indicative of a general change in stress response systems were altered production of multiple proteins involved in protein turnover and processing, transcription and DNA repair, and global stress response (Table 3). Here, we demonstrate a negative correlation between the levels of MifR, microcolony formation, and the levels of the DNA repair protein MutS (Table 2). To our knowledge, this is the first observation of MutS regulation in biofilms at the protein level. In addition, a positive correlation between MifR, microcolony formation and Usp proteins was observed. Usp proteins, which are conserved across all domains of life, have been implicated in the response to hypoxic conditions and establishment of chronic, persistent infections by Mycobacterium tuberculosis (Drumm et al., 2009), stress-mediated adhesion and motility in Escherichia coli (Nachin et al., 2005) and the process of biofilm formation by Porphyromonas gingivalis (Kuramitsu et al., 2005; Chen et al., 2006). In P. aeruginosa, increased expression of all five Usp proteins has been consistently observed under conditions of oxygen limitation, in laboratory-grown biofilms (Yoon et al., 2002; Waite et al., 2006; Alvarez-Ortega and Harwood, 2007), and in the context of in vivo infections (Mashburn et al., 2005; Bielecki et al., 2011). Usp proteins enable anaerobic growth and/or survival of P. aeruginosa, with inactivation of uspN and uspK having been linked to premature cell death during long-term anaerobic existence (Eschbach et al., 2004; Boes et al., 2006; Schreiber et al., 2006). However, UspN and UspK do so via distinct rather than converging mechanisms. UspN is required for survival when cells experience anaerobic energy stress and starvation, with a mutant strain demonstrating significantly reduced cell recovery after prolonged exposure to such conditions (Boes et al., 2006; 2008), while UspK is essential for P. aeruginosa anaerobic survival via pyruvate fermentation (Eschbach et al., 2004; Schreiber et al., 2006). This difference in mechanisms was apparent with respect to microcolony formation (Fig. 3). The small biomass accumulation defect observed following uspN inactivation suggests that subpopulations of biofilm cells experiencing anaerobic energy starvation likely utilize systems such as UspN to promote survival. In contrast, inactivation of uspK resulted in severely reduced microcolonies and biomass (Fig. 3, Table 3), strongly supporting a requirement for UspK in microcolony formation. Given that UspK plays a substantial role during fermentative growth on pyruvate, but is not required for denitrification (Schreiber et al., 2006), these findings strongly underscore the requirement for pyruvate fermentation as a means of addressing oxygen limitation of the biofilm microcolony microenvironment and the need for redox balancing.

In conclusion, our data suggest that bacteria present within microcolonies experience an environment that is distinct from environmental conditions elsewhere in the biofilm. Within microcolonies, P. aeruginosa cells experience low oxygen but energy (electron) -rich conditions and use fermentative processes for survival and growth. In particular, pyruvate is required for microcolony formation by P. aeruginosa with changes in extracellular pyruvate levels positively correlating with average biofilm cellular aggregate sizes. Pyruvate appears to be contributing to growth of biofilm microcolonies via pyruvate fermentation as a means of redox balancing with inactivation of lactate dehydrogenase preventing biofilm development and microcolony formation.

Experimental procedures

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

Bacterial strains, plasmids, media and culture conditions

All bacterial strains and plasmids used in this study are listed in Table S1. P. aeruginosa strains PA14 and PAO1 were used as the parental strains. All planktonic strains were grown in LB or VBMM (Schweizer, 1991) in shake flasks at 220 r.p.m. To assess growth under oxygen-limiting conditions, wild type and mifR mutant strains were grown in solidified LB- or VBMM-containing soft agar (0.1%) in the absence and presence of 25 mM potassium nitrate or 25 mM sodium pyruvate. To do so, 100 μl of overnight culture was inoculated into 10 ml molten medium, gently mixed by rotation for 10 s and the medium subsequently solidified on ice for 5 min. Cultures were incubated undisturbed at 37°C for 24 and 48 h. Methylene blue (Sumitani et al., 2004; Leistikow et al., 2010) and growth of C. sporogenes were used as indicators for oxygen-limiting conditions (Fig. S3). Growth was determined by viability counts (CFU ml−1). Growth in liquid under strictly anaerobic conditions was determined via absorbance following incubation in anaerobic jars.

Strain construction

Complementation of mutant strains or overexpression was accomplished by placing the respective genes under the control of an arabinose-inducible promoter in the pJN105 (Newman and Fuqua, 1999) or pMJT-1 (Kaneko et al., 2007) vectors. Primers used for strain construction are listed in Table S2. P. aeruginosa PA14 transposon mutants were verified by PCR and/or complementation.

Biofilm formation

Initial attachment was measured using the microtitre dish assay system, with crystal violet staining (O'Toole and Kolter, 1998). Biofilms were grown using a once-through continuous flow tube reactor system (1 m long size 14/15 silicone tubing, Masterflex, Cole Parmer) to obtain proteins and RNA. Briefly, biofilms were grown at 22°C for up to 6 days in minimal medium containing glutamate (MMA) (Sauer and Camper, 2001; Sauer et al., 2002) or 20-fold diluted LB medium. A flow rate of 0.2 ml min−1 (for size 14 tubing) or 0.5 ml min−1 (for size 15 tubing) was used. Flow cell-grown biofilms were used to observe the biofilm architecture as previously described (Sauer et al., 2002; Southey-Pillig et al., 2005; Allegrucci et al., 2006; Allegrucci and Sauer, 2007; Petrova and Sauer, 2009). Biofilms were grown at 22°C in 1/20 diluted LB or VBMM in the absence/presence of 0.1% arabinose. Quantitative analysis of confocal scanning laser microscopy (CSLM) images of flow cell-grown biofilms was performed using comstat (Heydorn et al., 2000). Alternatively, biofilms were grown in a 24-well plate system modified from the procedure described by Caiazza and O'Toole (2004) to elucidate the role of pyruvate in microcolony formation. Briefly, overnight cultures were adjusted to an OD600 of 0.05 in fresh 1/5-diluted LB or VBMM and grown at 37°C and 220 r.p.m. in 24-well microtitre plates at a 45° angle, ensuring that the bottom of the wells is at the air–liquid interface, with the medium exchanged every 8 h. Where indicated, 0.1–1 mM pyruvate, 1 mU pyruvate dehydrogenase, cofactors (1 μM β-NAD, 1 μM sodium Co-A) as well as oxygen indicator methylene blue (9 μg ml−1) (Sumitani et al., 2004; Leistikow et al., 2010) were added. At indicated time points, the adherent cells were visualized by brightfield microscopy (Olympus BX60 microscope, Olympus, Melville, NY, USA), with images acquired using a camera from Jenoptik (Jena, Germany). Size determination of cellular aggregates and microcolonies was done using the ProgRes CapturePro software (Jenoptik, Jena, Germany).

Two-dimensional gel electrophoresis and protein identification

2D/PAGE protein separation and gel image analysis was essentially carried out as previously described (Petrova and Sauer, 2009). Identification of proteins exhibiting differential production levels was accomplished via LC-MS/MS as previously described (Petrova and Sauer, 2009).

RNA isolation and preparation for Affymetrix GeneChip analysis

The samples were prepared identically as previously described (Morici et al., 2007). For biofilm growth experiments, three independent replicates of P. aeruginosa strains PAO1 and ΔmifR were grown as biofilms in a flow-through system using minimal medium as described above. RNA isolation and preparation for Affymetrix GeneChip analysis was done identically as previously described (Petrova et al., 2011).

Microarray data analysis

Gene expression signal-level data were computed with the robust GeneChip multi-array average method (GCRMA) (Wu and Irizarry, 2004) using default settings, log-transformed, and normalized to total intensities with GeneSpring® GX 10.0.1 using a guided workflow. To test for differential expression between the different strains grown in a biofilm, the Bayesian adjusted t-statistics was used as implemented in the GeneSpring package. P-values were corrected for multiple testing using the Benjamini and Hochberg's method (Benjamini and Hochberg, 1995) to control the false discovery rate. Only fold change ratios with adjusted P-values below 0.05 were included as statistically significant. Microarray analysis was performed on all three biological replicates. The full description of the array analysis platform and the complete array data have been deposited at the Gene Expression Omnibus website (http://www.ncbi.nlm.nih.gov/projects/geo) under accession number GSE35286.

Quantitative real-time reverse-transcription PCR (qRT-PCR)

qRT-PCR was used to confirm the transcript levels of selected genes using 1 μg of total RNA isolated from P. aeruginosa wild type and mifR mutant biofilms. Isolation of mRNA and cDNA synthesis was carried out as previously described (Basu Roy et al., 2012). qRT-PCR was performed using the Eppendorf Mastercycler® ep realplex (Eppendorf AG, Hamburng, Germany) and the KAPA SYBR FAST qPCR Kit (KAPABIOSYSTEMS, Woburn, MA, USA), with oligonucleotides listed in Table S2.

Extraction and quantification of cyclic di-GMP from P. aeruginosa

Cyclic di-GMP was extracted in triplicate from wild type and mutant strains grown planktonically and as a biofilm by heat and ethanol precipitation followed by detection and quantification by HPLC analysis as previously described (Petrova and Sauer, 2011).

In vitro phosphodiesterase activity assay

Overall phosphodiesterase activity of total cell extracts obtained from planktonic and biofilms cells of P. aeruginosa PAO1, ΔmifR and PAO1/pJN-mifR were determined using the synthetic chromogenic substrate bis(p-nitrophenyl) phosphate (Sigma-Aldrich) essentially as previously described (Bobrov et al., 2005; Kuchma et al., 2007).

Psl polysaccharide dot blot analysis

Psl polysaccharide was extracted from planktonic and 144-h-old biofilm cells essentially as described by Byrd et al. (2009). Quantification of Psl production was done by determining the anti-Psl dot blot spot volume (Irie et al., 2010) using the ImageMaster analysis software (GE Healthcare) (Petrova and Sauer, 2011).

NADH/NAD ratio determination

Extraction and quantification of intracellular NADH and NAD+ from planktonic cells and 6-day-old biofilms were carried out as described by San et al. (2002) and Price-Whelan et al. (2007) using the NADH/NAD+ quantification kit from Abnova (Taipei City, Taiwan).

Statistical analysis

A Student's t-test was performed for pair-wise comparisons of groups, and multivariant analyses were performed using a one-way ANOVA followed by a post-priori test using Sigma Stat software.

Acknowledgement

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

This work was supported by grants from the National Institutes of Health (R01 AI075257, R15 HL073835).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgement
  8. References
  9. Supporting Information
  • Allegrucci, M., and Sauer, K. (2007) Characterization of colony morphology variants isolated from Streptococcus pneumoniae biofilms. J Bacteriol 189: 20302038.
  • Allegrucci, M., Hu, F.Z., Shen, K., Hayes, J., Ehrlich, G.D., Post, J.C., and Sauer, K. (2006) Phenotypic characterization of Streptococcus pneumoniae biofilm development. J Bacteriol 188: 23252335.
  • Allesen-Holm, M., Barken, K.B., Yang, L., Klausen, M., Webb, J.S., Kjelleberg, S., et al. (2006) A characterization of DNA release in Pseudomonas aeruginosa cultures and biofilms. Mol Microbiol 59: 11141128.
  • Alvarez-Ortega, C., and Harwood, C.S. (2007) Responses of Pseudomonas aeruginosa to low oxygen indicate that growth in the cystic fibrosis lung is by aerobic respiration. Mol Microbiol 65: 153165.
  • Anderl, J.N., Franklin, M.J., and Stewart, P.S. (2000) Role of antibiotic penetration limitation in Klebsiella pneumoniae biofilm resistance to ampicillin and ciprofloxacin. Antimicrob Agents Chemother 44: 18181824.
  • Anderl, J.N., Zahller, J., Roe, F., and Stewart, P.S. (2003) Role of nutrient limitation and stationary-phase existence in Klebsiella pneumoniae biofilm resistance to ampicillin and ciprofloxacin. Antimicrob Agents Chemother 47: 12511256.
  • Banin, E., Vasil, M.L., and Greenberg, E.P. (2005) Iron and Pseudomonas aeruginosa biofilm formation. Proc Natl Acad Sci USA 102: 1107611081.
  • Barraud, N., Hassett, D.J., Hwang, S.-H., Rice, S.A., Kjelleberg, S., and Webb, J.S. (2006) Involvement of nitric oxide in biofilm dispersal of Pseudomonas aeruginosa. J Bacteriol 188: 73447353.
  • Basu Roy, A., Petrova, O.E., and Sauer, K. (2012) The phosphodiesterase DipA (PA5017) is essential for Pseudomonas aeruginosa biofilm dispersion. J Bacteriol 194: 29042915.
  • de Beer, D., Stoodley, P., and Lewandowski, Z. (1994) Liquid flow in heterogeneous biofilms. Biotechnol Bioeng 44: 636641.
  • Benjamini, Y., and Hochberg, Y. (1995) Controlling the false discovery rate – a practical and powerful approach to multiple testing. J R Stat Soc Ser B Stat Methodol 57: 289300.
  • Bielecki, P., Puchałka, J., Wos-Oxley, M.L., Loessner, H., Glik, J., Kawecki, M., et al. (2011) In-vivo expression orofiling of Pseudomonas aeruginosa infections reveals niche-specific and strain-independent transcriptional programs. PLoS ONE 6: e24235.
  • Bobrov, A.G., Kirillina, O., and Perry, P.D. (2005) The phosphodiesterase activity of the HmsP EAL domain is required for negative regulation of biofilm formation in Yersinia pestis. FEMS Microbiol Lett 247: 123130.
  • Boes, N., Schreiber, K., Hartig, E., Jaensch, L., and Schobert, M. (2006) The Pseudomonas aeruginosa universal stress protein PA4352 is essential for surviving anaerobic energy stress. J Bacteriol 188: 65296538.
  • Boes, N., Schreiber, K., and Schobert, M. (2008) SpoT-triggered stringent response controls usp gene expression in Pseudomonas aeruginosa. J Bacteriol 190: 71897199.
  • Borriello, G., Werner, E., Roe, F., Kim, A.M., Ehrlich, G.D., and Stewart, P.S. (2004) Oxygen limitation contributes to antibiotic tolerance of Pseudomonas aeruginosa in biofilms. Antimicrob Agents Chemother 48: 26592664.
  • Byrd, M.S., Sadovskaya, I., Vinogradov, E., Lu, H., Sprinkle, A.B., Richardson, S.H., et al. (2009) Genetic and biochemical analyses of the Pseudomonas aeruginosa Psl exopolysaccharide reveal overlapping roles for polysaccharide synthesis enzymes in Psl and LPS production. Mol Microbiol 73: 622638.
  • Caiazza, N.C., and O'Toole, G.A. (2004) SadB is required for the transition from reversible to irreversible attachment during biofilm formation by Pseudomonas aeruginosa PA14. J Bacteriol 186: 44764485.
  • Carlson, C.A., and Ingraham, J.L. (1983) Comparison of denitrification by Pseudomonas stutzeri, Pseudomonas aeruginosa, and Paracoccus denitrificans. Appl Environ Microbiol 45: 12471253.
  • Carlson, C.A., Ferguson, L.P., and Ingraham, J.L. (1982) Properties of dissimilatory nitrate reductase purified from the denitrifier Pseudomonas aeruginosa. J Bacteriol 151: 162171.
  • Chen, W., Honma, K., Sharma, A., and Kuramitsu, H.K. (2006) A universal stress protein of Porphyromonas gingivalis is involved in stress responses and biofilm formation. FEMS Microbiol Lett 264: 1521.
  • Comolli, J.C., and Donohue, T.J. (2004) Differences in two Pseudomonas aeruginosa cbb3 cytochrome oxidases. Mol Microbiol 51: 11931203.
  • Conibear, T.C.R., Collins, S.L., and Webb, J.S. (2009) Role of mutation in Pseudomonas aeruginosa biofilm development. PLoS ONE 4: e6289.
  • Davey, M.E., Caiazza, N.C., and O'Toole, G.A. (2003) Rhamnolipid surfactant production affects biofilm architecture in Pseudomonas aeruginosa PAO1. J Bacteriol 185: 10271036.
  • Davies, K.J.P., Lloyd, D., and Boddy, L. (1989) The effect of oxygen on denitrification in Paracoccus denitrificans and Pseudomonas aeruginosa. J Gen Microbiol 135: 24452451.
  • Dietrich, L.E.P., Price-Whelan, A., Petersen, A., Whiteley, M., and Newman, D.K. (2006) The phenazine pyocyanin is a terminal signalling factor in the quorum sensing network of Pseudomonas aeruginosa. Mol Microbiol 61: 13081321.
  • Dietrich, L.E.P., Teal, T.K., Price-Whelan, A., and Newman, D.K. (2008) Redox-active antibiotics control gene expression and community behavior in divergent bacteria. Science 321: 12031206.
  • Drumm, J.E., Mi, K., Bilder, P., Sun, M., Lim, J., Bielefeldt-Ohmann, H., et al. (2009) Mycobacterium tuberculosis universal stress protein Rv2623 regulates bacillary growth by ATP-binding: requirement for establishing chronic persistent infection. PLoS Pathog 5: e1000460.
  • Eschbach, M., Schreiber, K., Trunk, K., Buer, J., Jahn, D., and Schobert, M. (2004) Long-term anaerobic survival of the opportunistic pathogen Pseudomonas aeruginosa via pyruvate fermentation. J Bacteriol 186: 45964604.
  • Filiatrault, M.J., Picardo, K.F., Ngai, H., Passador, L., and Iglewski, B.H. (2006) Identification of Pseudomonas aeruginosa genes involved in virulence and anaerobic growth. Infect Immun 74: 42374245.
  • Garcia-Medina, R., Dunne, W.M., Singh, P.K., and Brody, S.L. (2005) Pseudomonas aeruginosa acquires biofilm-like properties within airway epithelial cells. Infect Immun 73: 82988305.
  • Hassett, D.J., Cuppoletti, J., Trapnell, B., Lymar, S.V., Rowe, J.J., Yoon, S.S., et al. (2002) Anaerobic metabolism and quorum sensing by Pseudomonas aeruginosa biofilms in chronically infected cystic fibrosis airways: rethinking antibiotic treatment strategies and drug targets. Adv Drug Deliv Rev 54: 14251443.
  • Hassett, D.J., Sutton, M.D., Schurr, M.J., Herr, A.B., Caldwell, C.C., and Matu, J.O. (2009) Pseudomonas aeruginosa hypoxic or anaerobic biofilm infections within cystic fibrosis airways. Trends Microbiol 17: 130138.
  • Heydorn, A., Nielsen, A.T., Hentzer, M., Sternberg, C., Givskov, M., Ersboll, B.K., and Molin, S. (2000) Quantification of biofilm structures by the novel computer program COMSTAT. Microbiology 146: 23952407.
  • 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.
  • Irie, Y., Starkey, M., Edwards, A.N., Wozniak, D.J., Romeo, T., and Parsek, M.R. (2010) Pseudomonas aeruginosa biofilm matrix polysaccharide Psl is regulated transcriptionally by RpoS and post-transcriptionally by RsmA. Mol Microbiol 78: 158172.
  • Kaneko, Y., Thoendel, M., Olakanmi, O., Britigan, B.E., and Singh, P.K. (2007) The transition metal gallium disrupts Pseudomonas aeruginosa iron metabolism and has antimicrobial and antibiofilm activity. J Clin Invest 117: 877888.
  • Kuchma, S.L., Brothers, K.M., Merritt, J.H., Liberati, N.T., Ausubel, F.M., and O'Toole, G.A. (2007) BifA, a c-di-GMP phosphodiesterase, inversely regulates biofilm formation and swarming motility by Pseudomonas aeruginosa PA14. J Bacteriol 189: 81658178.
  • Kuramitsu, H.K., Chen, W., and Ikegami, A. (2005) Biofilm formation by the periodontopathic bacteria Treponema denticola and Porphyromonas gingivalis. J Periodontol 76: 20472051.
  • Lam, J., Chan, R., Lam, K., and Costerton, J.W. (1980) Production of mucoid microcolonies by Pseudomonas aeruginosa within infected lungs in cystic fibrosis. Infect Immun 28: 546556.
  • Leistikow, R.L., Morton, R.A., Bartek, I.L., Frimpong, I., Wagner, K., and Voskuil, M.I. (2010) The Mycobacterium tuberculosis DosR regulon assists in metabolic homeostasis and enables rapid recovery from nonrespiring dormancy. J Bacteriol 192: 16621670.
  • Lenz, A.P., Williamson, K.S., Pitts, B., Stewart, P.S., and Franklin, M.J. (2008) Localized gene expression in Pseudomonas aeruginosa biofilms. Appl Environ Microbiol 74: 44634471.
  • Mashburn, L.M., Jett, A.M., Akins, D.R., and Whiteley, M. (2005) Staphylococcus aureus serves as an iron source for Pseudomonas aeruginosa during in vivo coculture. J Bacteriol 187: 554566.
  • Mikkelsen, H., Bond, N.J., Skindersoe, M.E., Givskov, M., Lilley, K.S., and Welch, M. (2009) Biofilms and type III secretion are not mutually exclusive in Pseudomonas aeruginosa. Microbiology 155: 687698.
  • Morici, L.A., Carterson, A.J., Wagner, V.E., Frisk, A., Schurr, J.R., zu Bentrup, K.H., et al. (2007) Pseudomonas aeruginosa AlgR represses the Rhl quorum-sensing system in a biofilm-specific manner. J Bacteriol 189: 77527764.
  • Nachin, L., Nannmark, U., and Nyström, T. (2005) Differential roles of the universal stress proteins of Escherichia coli in oxidative stress resistance, adhesion, and motility. J Bacteriol 187: 62656272.
  • Newman, J.R., and Fuqua, C. (1999) Broad-host-range expression vectors that carry the arabinose-inducible Escherichia coli araBAD promoter and the araC regulator. Gene 227: 197203.
  • O'Toole, G.A., and Kolter, R. (1998) Initiation of biofilm formation in Pseudomonas fluorescens WCS365 proceeds via multiple, convergent signalling pathways: a genetic analysis. Mol Microbiol 28: 449461.
  • O'Toole, G.A., Gibbs, K.A., Hager, P.W., Phibbs, P.V., Jr, and Kolter, R. (2000) The global carbon metabolism regulator Crc is a component of a signal transduction pathway required for biofilm development by Pseudomonas aeruginosa. J Bacteriol 182: 425431.
  • Pérez-Osorio, A.C., Williamson, K.S., and Franklin, M.J. (2010) Heterogeneous rpoS and rhlR mRNA levels and 16S rRNA/rDNA (rRNA gene) ratios within Pseudomonas aeruginosa biofilms, sampled by laser capture microdissection. J Bacteriol 192: 29913000.
  • Petrova, O.E., and Sauer, K. (2009) A novel signaling network essential for regulating Pseudomonas aeruginosa biofilm development. PLoS Pathog 5: e1000668.
  • Petrova, O.E., and Sauer, K. (2011) SagS contributes to the motile-sessile switch and acts in concert with BfiSR to enable Pseudomonas aeruginosa biofilm formation. J Bacteriol 193: 66146628.
  • Petrova, O.E., and Sauer, K. (2012) Sticky situations: key components that control bacterial surface attachment. J Bacteriol 194: 24132425.
  • Petrova, O.E., Schurr, J.R., Schurr, M.J., and Sauer, K. (2011) The novel Pseudomonas aeruginosa two-component regulator BfmR controls bacteriophage-mediated lysis and DNA release during biofilm development through PhdA. Mol Microbiol 81: 767783.
  • Platt, M.D., Schurr, M.J., Sauer, K., Vazquez, G., Kukavica-Ibrulj, I., Potvin, E., et al. (2008) Proteomic, microarray, and signature-tagged mutagenesis analyses of anaerobic Pseudomonas aeruginosa at pH 6.5, likely representing chronic, late-stage cystic fibrosis airway conditions. J Bacteriol 190: 27392758.
  • Price-Whelan, A., Dietrich, L.E.P., and Newman, D.K. (2007) Pyocyanin alters redox homeostasis and carbon flux through central metabolic pathways in Pseudomonas aeruginosa PA14. J Bacteriol 189: 63726381.
  • Ramos, I., Dietrich, L.E.P., Price-Whelan, A., and Newman, D.K. (2010) Phenazines affect biofilm formation by Pseudomonas aeruginosa in similar ways at various scales. Res Microbiol 161: 187191.
  • Rani, S.A., Pitts, B., Beyenal, H., Veluchamy, R.A., Lewandowski, Z., Davison, W.M., et al. (2007) Spatial patterns of DNA replication, protein synthesis, and oxygen concentration within bacterial biofilms reveal diverse physiological states. J Bacteriol 189: 42234233.
  • Rasmussen, K., and Lewandowski, Z. (1998) Microelectrode measurements of local mass transport rates in heterogeneous biofilms. Biotechnol Bioeng 59: 302309.
  • San, K.-Y., Bennett, G.N., Berríos-Rivera, S.J., Vadali, R.V., Yang, Y.-T., Horton, E., et al. (2002) Metabolic engineering through cofactor manipulation and its effects on metabolic flux redistribution in Escherichia coli. Metab Eng 4: 182192.
  • Sauer, K., and Camper, A.K. (2001) Characterization of phenotypic changes in Pseudomonas putida in response to surface-associated growth. J Bacteriol 183: 65796589.
  • Sauer, K., Camper, A.K., Ehrlich, G.D., Costerton, J.W., and Davies, D.G. (2002) Pseudomonas aeruginosa displays multiple phenotypes during development as a biofilm. J Bacteriol 184: 11401154.
  • Schobert, M., and Jahn, D. (2010) Anaerobic physiology of Pseudomonas aeruginosa in the cystic fibrosis lung. Int J Med Microbiol 300: 549556.
  • Schreiber, K., Boes, N., Eschbach, M., Jaensch, L., Wehland, J., Bjarnsholt, T., et al. (2006) Anaerobic survival of Pseudomonas aeruginosa by pyruvate fermentation requires an Usp-type stress protein. J Bacteriol 188: 659668.
  • Schweizer, H.P. (1991) The agmR gene, an environmentally responsive gene, complements defective glpR, which encodes the putative activator for glycerol metabolism in Pseudomonas aeruginosa. J Bacteriol 173: 67986806.
  • Shirtliff, M.E., Mader, J.T., and Camper, A.K. (2002) Molecular interactions in biofilms. Chem Biol 9: 859871.
  • Somerville, G., Mikoryak, C.A., and Reitzer, L. (1999) Physiological characterization of Pseudomonas aeruginosa during exotoxin A synthesis: glutamate, iron limitation, and aconitase activity. J Bacteriol 181: 10721078.
  • Southey-Pillig, C.J., Davies, D.G., and Sauer, K. (2005) Characterization of temporal protein production in Pseudomonas aeruginosa biofilms. J Bacteriol 187: 81148126.
  • Sriramulu, D.D., Lünsdorf, H., Lam, J.S., and Römling, U. (2005) Microcolony formation: a novel biofilm model of Pseudomonas aeruginosa for the cystic fibrosis lung. J Med Microbiol 54: 667676.
  • Stewart, P.S. (1996) Theoretical aspects of antibiotic diffusion into microbial biofilms. Antimicrob Agents Chemother 40: 25172522.
  • Stoodley, P., deBeer, D., and Lewandowski, Z. (1994) Liquid flow in biofilm systems. Appl Environ Microbiol 60: 27112716.
  • Stoodley, P., Yang, S., Lappin-Scott, H., and Lewandowski, Z. (1997) Relationship between mass transfer coefficient and liquid flow velocity in heterogenous biofilms using microelectrodes and confocal microscopy. Biotechnol Bioeng 56: 681688.
  • Sumitani, M., Takagi, S., Tanamura, Y., and Inoue, H. (2004) Oxygen indicator composed of an organic/inorganic hybrid compound of methylene blue, reductant, surfactant and saponite. Anal Sci 20: 11531157.
  • Toyofuku, M., Nomura, N., Fujii, T., Takaya, N., Maseda, H., Sawada, I., et al. (2007) Quorum sensing regulates denitrification in Pseudomonas aeruginosa PAO1. J Bacteriol 189: 49694972.
  • Van Alst, N.E., Picardo, K.F., Iglewski, B.H., and Haidaris, C.G. (2007) Nitrate sensing and metabolism modulate motility, biofilm formation, and virulence in Pseudomonas aeruginosa. Infect Immun 75: 37803790.
  • Waite, R., Paccanaro, A., Papakonstantinopoulou, A., Hurst, J., Saqi, M., Littler, E., and Curtis, M. (2006) Clustering of Pseudomonas aeruginosa transcriptomes from planktonic cultures, developing and mature biofilms reveals distinct expression profiles. BMC Genomics 7: 162.
  • Walters, M.C. III, Roe, F., Bugnicourt, A., Franklin, M.J., and Stewart, P.S. (2003) Contributions of antibiotic penetration, oxygen limitation, and low metabolic activity to tolerance of Pseudomonas aeruginosa biofilms to ciprofloxacin and tobramycin. Antimicrob Agents Chemother 47: 317323.
  • Winsor, G.L., Van Rossum, T., Lo, R., Khaira, B., Whiteside, M.D., Hancock, R.E.W., and Brinkman, F.S.L. (2009) Pseudomonas genome database: facilitating user-friendly, comprehensive comparisons of microbial genomes. Nucleic Acids Res 37: D483D488.
  • Wu, M., Guina, T., Brittnacher, M., Nguyen, H., Eng, J., and Miller, S.I. (2005) The Pseudomonas aeruginosa proteome during anaerobic growth. J Bacteriol 187: 81858190.
  • Wu, Z., and Irizarry, R.A. (2004) Preprocessing of oligonucleotide array data. Nat Biotechnol 22: 656658.
  • Yang, L., Barken, K.B., Skindersoe, M.E., Christensen, A.B., Givskov, M., and Tolker-Nielsen, T. (2007) Effects of iron on DNA release and biofilm development by Pseudomonas aeruginosa. Microbiology 153: 13181328.
  • Yoon, S.S., Hennigan, R.F., Hilliard, G.M., Ochsner, U.A., Parvatiyar, K., Kamani, M.C., et al. (2002) Pseudomonas aeruginosa anaerobic respiration in biofilms: relationships to cystic fibrosis pathogenesis. Dev Cell 3: 593603.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgement
  8. References
  9. Supporting Information
FilenameFormatSizeDescription
mmi12018-sup-0001-si.pdf1485K

Supporting information

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.