SEARCH

SEARCH BY CITATION

Summary

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

Under water-limiting conditions Pseudomonas putida produces the exopolysaccharide alginate, which influences biofilm development and facilitates maintaining a hydrated microenvironment. Since alginate is a minor biofilm matrix component it is important to determine whether alginate production occurs by all or a subset of residents, and when and to what extent cells contribute to alginate production. To address these questions we employed stable and unstable fluorescent reporters to measure alginate biosynthesis (algD) operon expression and metabolic activity in vivo quantitatively by flow cytometry and visually by microscopy. Here we report that during growth under water-limiting conditions and when biofilms become dehydrated most residents transiently express the alginate biosynthesis genes leading to distinct spatial patterns as the biofilm ages. Transient alginate gene expression was not a consequence of decreased metabolic activity, since metabolic reporters were still expressed, nor was it likely due to transient cytosolic availability of the alternative sigma factor AlgT, based on qRT-PCR. Our findings also indicate that one or more biofilm attribute, other than alginate, provides protection from desiccation stress. Collectively, our findings suggest that differentiated cells dedicated to alginate production are not part of the P. putida biofilm lifestyle under water-limiting conditions. Alternatively, P. putida biofilm cells may be responding to their own local environment, producing alginate because of the fitness advantage it confers under those particular conditions.


Introduction

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

It is now well accepted that the vast majority of bacterial life in nature is found in surface-bound communities called biofilms rather than as isolated planktonic cells (Davey and O'Toole, 2000). The importance of biofilms in nature is reflected by their prevalence in terrestrial, aquatic, plant and animal ecosystems and their role in many chronic and acute plant and human diseases, antibiotic resistance and soil processes (Holden and Firestone, 1997; Davey and O'Toole, 2000). Biofilms are dense communities, in which cells can share many secreted molecules, including iron-scavenging siderophores, intercellular signal molecules and extracellular polymers. A defining feature of many biofilms is the exopolysaccharides (EPS) or slime that encapsulate the bacteria (Davey and O'Toole, 2000; Watnick and Kolter, 2000). These polymers are typically viewed as a shared resource (West et al., 2007; Xavier and Foster, 2007) that provides a benefit to the biofilm by maintaining its structure (Watnick and Kolter, 2000), facilitate signalling (Kolter and Greenberg, 2006), and protecting residents from environmental stressors (Davey and O'Toole, 2000; Chang et al., 2007). As the community grows, different cell types appear as the original strain diversifies to allow cells to take on different tasks. Although there is increasing evidence that distinct subpopulations contribute to particular developmental processes it is less clear whether polymer production is a task achieved by all or a subset of community members. Significantly, matrix production appears to be accomplished by a specific Bacillus subtilis subpopulation that localizes to distinct regions throughout development of the biofilm (Vlamakis et al., 2008; Lopez et al., 2009). If polymers are produced by a differentiated subset this potentially shared resource could be exploited by non-polymer producing cells under stressful conditions since more available energy could be devoted to survival. Alternatively, polymer production could reflect responses to the immediate local environment because of the fitness advantages the polymer confers upon each individual (Monds and O'Toole, 2009). Thus, it is important know whether polymer production requires cooperation by most or all residents, or is delegated to a dedicated subpopulation.

Biofilms are composed of large numbers of cells, each of which experiences its own microenvironment due to strong physicochemical gradients established by metabolism, diffusion and the physical environment. Biofilm communities are therefore heterogeneous and spatially stratified, so that metabolic activities could occur differentially according to the location of the cell in the biofilm and the cells life history. Numerous studies have demonstrated heterogeneity with regard to cellular activity and metabolism in both flow-through and colony biofilm experimental systems (Wentland et al., 1996; Huang et al., 1998; Werner et al., 2004; Teal et al., 2006; Rani et al., 2007). Exopolymer production is highly regulated and curiously, some bacterial species activate (Quinones et al., 2005; Sakuragi and Kolter, 2007) whereas others terminate (Hammer and Bassler, 2003) polymer secretion after reaching high cell density. When polymer secretion occurs is likely influenced by environmental conditions, particularly if the secreted product is a fitness determinant. For example, the Pseudomonas aeruginosa polysaccharide Psl is important for cell–cell and cell–surface interactions and for maintaining biofilm structure post-attachment (Ma et al., 2006). Constitutive expression of the psl operon by planktonic cells supposedly enables efficient attachment to surfaces and that regulated localized psl operon expression in the centre of microcolonies is required for biofilm differentiation (Overhage et al., 2005).

Many bacteria colonize a variety of habitats that may not be saturated or consistently saturated with water. Survival under water-limited conditions requires unique adaptations distinct from those employed for tolerating high osmolarity or water-replete conditions (Potts, 1994; van de Mortel and Halverson, 2004; Cytryn et al., 2007). One response to this stress by Pseudomonas putida, and other fluorescent pseudomonads, is to produce the EPS alginate, which influences matrix physicochemical properties and biofilm development (Chang et al., 2007). Collectively, these responses may facilitate the maintenance of a hydrated microenvironment, protecting residents from desiccation stress and increasing survival. The importance of alginate in the ecological fitness of Pseudomonads needs to be viewed with the understanding that it is not the primary EPS component under water-limiting conditions and that a relatively small amount is sufficient to protect cells from desiccation stress or modulate biofilm architecture (Chang et al., 2007). At present it is unclear how many biofilm cells produce alginate, and when and to what extent cells produce alginate. It is likely that production is context-dependent, influenced by life histories or when biofilm cells are deprived of water. To address these questions we employed stable and unstable green fluorescent protein (GFP) reporters fused to the promoter for the algD operon, thus called because the first of 12 genes in the operon begins with algD, and qRT-PCR to examine alginate expression by P. putida in vivo. Our findings suggest that alginate production may be a consequence of cells responding to their local environment because of the fitness advantages it confers rather than as a consequence of a differentiated subpopulation devoted to alginate production.

Results

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

To identify individual cells producing alginate we chose to monitor gene expression instead of directly assaying for the production of alginate, which is extracellular, and may not localize near the cells that produce it. Furthermore, aqueous treatments required for alginate localization, such as by immunofluoresence microscopy, disrupts the water status of biofilms formed under water-limiting (desiccation) conditions thereby hindering localization of alginate producing cells. Therefore to most accurately monitor alginate producing cells, we generated transcriptional fusions of the promoter of the alginate biosynthesis and export (algD) operon to genes encoding GFP with or without the AAV peptide tag to shorten protein half-life (Andersen et al., 1998). We also constructed transcriptional fusions of the promoter of the neomycin phosphotrasferase (nptII) gene of Tn5 to genes encoding GFP with or without the AAV peptide tag to be used as a metabolic reporter since previous studies had demonstrated that this promoter was constitutive under a variety of environmental conditions (Miller et al., 2000; Axtell and Beattie, 2002; Chang et al., 2007). We selected this promoter over the growth rate-dependent promoter PrrnB, which is frequently used as a metabolic reporter (Sternberg et al., 1999; Ramos et al., 2000; Pamp et al., 2008), since the fluorescence signal of the short half-life variant was difficult to assess in aged biofilms. We constructed strains harbouring the transcriptional fusions as described in Experimental procedures and Table 1. We assume that cells expressing the short (pPalgD–gfpAAV) or long (pPalgD–gfp) half-life reporters were producing alginate.

Table 1.  Bacterial strains, plasmids and primers used in this study.
P. putida strainsProperties or sequenceSource
mt2Rf-resistant variant of wild-type mt-2 
mt2-HcRedWild-type mt-2 tagged with miniTn7HcRedThis study
mt2 ΔalgDΔalgD-deficient mutantChang et al. (2009)
Plasmids
pPalgDgfpalgD operon promoter cloned into pPROBE-NT, KmRChang et al. (2007)
pPnptIIgfpThe constitutive neomycin phosphotransferase promoter nptII fused to gfp in pPROBE-KT, KmRChang and Halverson (2003)
pPROBE-NTAAVPromoter probe vector with promoterless gfp with the AAV TAG, pBBR1 ori, KmRMiller et al. (2000)
pPROBE-NTPromoter probe vector, pBBR1 ori, KmRMiller et al. (2000)
pTOPO-PalgDpTOPO PCR2.1 with algD operon promoterChang et al. (2007)
pPalgDgfpAAValgD promoter cloned into the MCS of pPROBE-NTAAV, KmRThis study
pPnptIIgfpAAVnptII promoter cloned into the MCS of pPROBE-NTAAV, KmRThis study
pAKNminiTn7HcRed, with the constitutive PA1/04/03 promoter controlling HcRed expressionLambertsen et al. (2004)
Primers
nptII-F5′-GTCAGGCTGTAACAGCT-3′Chang and Halverson (2003)
nptII-R5′-ATCCTGTCTCTTGATCAGATCTTG-3′Chang and Halverson (2003)
algDprm-F5′-TGAAGCCGCGAAGAGGTCAG-3′Chang et al. (2007)
algDprm-R5′-CGCATCGCTATCACCTCATG-3′Chang et al. (2007)
RimM-RT2F5′-GAGCGGTTACGAAATCTGCATCCC-3′; 160 bpThis study
RimM-RT2R5′-CATTACATCGTTCGCACCGGTCTC-3′This study
Alg8-RT2F5′-ATGATCTTCCTCTACGGCGTGTACC F-3′; 218 bpThis study
Alg8-RT2R5′-AGCAGTTCATCGGACATTTCCACC-3′This study
AlgU-RT2F5′-ACGGGCTGAGTTACGAAGACATTG-3′; 109 bpThis study
AlgU-RT2R5′-AACGGCTGCAGGGCTTTATCTATG-3′This study
MucA-RT3F5′-TGAAGCTTTGCAGGAATCGCTGTC-3′; 162 bpThis study
MucA-RT3R5′-TTAGGCAGCAGCAGTTCCTTGTG-3′This study

Analysis of the algD promoter

Since relatively little is known about the P. putida alginate promoter we compared the sequence of the amplified PCR product used to construct our transcriptional fusions with those of other fluorescent Pseudomonads. Sequence analysis indicated that it is comprised of a long untranslated leader sequence located between the algD translational start site and a putative AlgT (σ22) binding site with moderate overall sequence similarities to P. aeruginosa PAO1 but stronger sequence similarities with putative DNA binding sites of various transcriptional factors (Fig. 1) demonstrated to influence alginate expression in P. aeruginosa. Based on blastp (Altschul et al., 1997) analysis, P. putida homologues of AlgR, AmrZ, AlgT and AlgB exhibited 85–93% sequence similarity to PAO1 proteins. Given the similarities to P. aeruginosa it is likely that algD promoter activity in P. putida is regulated similarly.

image

Figure 1. Schematic illustration and clustalw (Thompson et al., 1994) alignment between P. aeruginosa PAO1 (top) and P. putida mt2 (bottom) algD promoter regions. Numbers inside black boxes represent the DNA binding sites relative to the PAO1 transcriptional start site for various regulatory proteins involved in activation of the algD operon (Devault et al., 1991; Mohr et al., 1991; Deretic et al., 1992; Wozniak and Ohman, 1994; Baynham and Wozniak, 1996; Yu et al., 1997). The AmrZ and AlgB DNA binding sites overlap by 11 base pairs. The complete clustalw alignment, including P. syringae pv. syringae, is provided in Fig. S1.

Download figure to PowerPoint

Effect of water stress on biosensor response

We evaluated the effect of water deprivation severity on expression of the pPalgD–gfp and pPalgD–gfpAAV transcriptional fusions. To simulate a matric stress, the water potential of the culture medium was lowered with PEG as described in Experimental procedures. With increasing water limitation there was a corresponding increase in algD promoter activity, particularly at water potentials below −0.5 MPa (Fig. S2). Unlike in other Pseudomonads (Keith and Bender, 1999), neither high osmolarity or the oxidants methyl violgen or hydrogen peroxide induced expression of the alginate reporters, suggesting that water deprivation and not ancillary accumulation of reactive oxygen species (Chang et al., 2009) is the primary inducer of alginate expression in P. putida under our experimental conditions.

Temporal dynamics of alginate expression during biofilm development

Analysis of populations of cells harbouring an empty vector cultivated under water-limiting conditions (light grey peak in Fig. 2A) or the pPalgD–gfpAAV reporter under water-replete conditions (light grey peak in Fig. 2B) revealed that there was little algD promoter activity under these conditions. In contrast, populations of cells harbouring the pPnptII–gfpAAV reporter grown under water-replete (data not shown) and -limiting conditions (dark peaks in Fig. 2A) expressed fluorescence in a biomodal pattern, with most cells highly expressing the metabolic activity reporter. Cells harbouring the pPalgD–gfpAAV reporter grown under water-limiting conditions exhibited an increase in algD promoter activity, revealing uninduced and induced subpopulations (dark peaks in Fig. 2B). From this information and other parameters, gates were established as described in Experimental procedures to identify the proportion of cells expressing the alginate and metabolic reporters, and the extent of induction (fluorescence intensity) by each cell that expressed the reporter. The sizes of the alginate-expressing subpopulations and the extent of alginate induction varied considerably with the age of the biofilm, which is described in more detail below. Cells harbouring the long half-life reporters exhibited similar patterns as those harbouring the short half-life reporters.

image

Figure 2. Representative histograms from flow cytometry analyses illustrating gene expression by 12-h-old biofilm cells harbouring the short half-life (gfpAAV) transcriptional fusions. Cells with the empty vector (negative control; light grey peak in A) grown under water-limiting conditions and the pPalgD–gfpAAV reporter grown under water-replete conditions exhibited a unimodal distribution of fluorescence (light grey peak in B). We considered these as uninduced populations. Cells with the pPnptII–gfpAAV transcriptional fusion (positive control for metabolic activity; dark peaks in A) cultivated under water-limiting conditions exhibited a bimodal distribution of fluorescence, revealing a small uninduced and a large induced populations. Cells with the pPalgD–gfpAAV reporter cultivated under water-limiting conditions exhibited a bimodal distribution of fluorescent cells (dark peaks in B), revealing uninduced and induced populations. Similar profiles were obtained with cells harbouring long half-life transcriptional fusions (data not shown).

Download figure to PowerPoint

Under water-limiting conditions, within 12 h > 85% of biofilm cells expressed the pPalgD–gfpAAV reporter which was followed over time by a precipitous decline (Fig. 3A) during a period of microcolony growth (Chang et al., 2009). The decrease in the proportion of cells expressing the pPalgD–gfpAAV reporter (Fig. 3A) was not due to an increase in the proportion of non-alginate-expressing cells since the relative abundance of long half-life pPalgD–gfp-expressing cells (Fig. S3A) did not change appreciably over time. This suggests that alginate expression is transient and that stress severity influences expression dynamics (Fig. 3A and Fig. S3A). The decline in the pPalgD–gfpAAV-expressing cells is not solely a consequence of a decrease in metabolic activity since the proportion of the population expressing the metabolic activity reporters pPnptII–gfpAAV (Fig. 3B) and pPnptII–gfp (Fig. S3B) decreased more slowly or very little over time, respectively, compared with cells expressing either alginate reporter (Fig. 3A and Fig. S3B).

image

Figure 3. Temporal dynamics of the proportion of P. putida mt2 biofilm cells expressing the algD promoter–gfp transcriptional fusion during growth under water-replete and -limiting conditions. (A) pPalgD–gfpAAV; (B) pPnptII–gfpAAV. 0 MPa Ψ (○); −1.0 MPa Ψ PEG (▿); −1.5 MPa Ψ PEG (□). The flow cytometer was gated as described in Experimental procedures. Data are representative of one of three experiments with similar results. Data points are the mean and error bars the SEM of three replications. Error bars that are not visible are masked by the symbol.

Download figure to PowerPoint

There was a direct relationship between water deprivation severity and the extent of alginate expression (Fig. 4A), at least during early phases of biofilm development (0–24 h). Subsequent to the rapid decline in the proportion of cell expressing the pPalgD–gfpAAV reporter (Fig. 3A) there was a significant, albeit transient, increase in expression (Fig. 4A) by approximately 1% of the biofilm cells (Fig. 3A). Eventually the extent of alginate induction did not differ under the two levels of water deprivation we examined (Fig. 4A and Fig. S4A). The dynamic nature of alginate reporter expression does not appear to be a consequence of fluxes in metabolism since the fluorescence intensity per cell of the metabolic activity reporters (Fig. 4B and Fig. S4B) did not decrease as much over time.

image

Figure 4. Temporal dynamics of the extent of algD promoter–gfp transcriptional fusion induction by P. putida mt2 biofilm cells during growth under water-replete and -limiting conditions. (A) pPalgD–gfpAAV; (B) pPnptII–gfpAAV. 0 MPa Ψ (○); −1.0 MPa Ψ PEG (▿); −1.5 MPa Ψ PEG (□). The fluorometer was gated as described in Experimental procedures. Fluorescence intensity per cell is in arbitrary units. Data are representative of one of three experiments with similar results. Data points are the mean and error bars the SEM of three replications. Error bars that are not visible are masked by the symbol.

Download figure to PowerPoint

qPCR of alginate expression

To establish whether our GFP-based transcriptional fusions accurately reflect alginate expression we verified those findings by quantitative real-time PCR. The alg8 gene is the second gene in the algD operon and was selected since we were unable to design primers for algD that did not result in PCR artefacts. Our qRT-PCR results are in congruence with our algD reporter findings, as evidenced by high levels of expression under water-limiting, but not high-osmolarity conditions and that expression decreased over time (Fig. 5A) in a fashion similar to what we observed with the pPalgD–gfpAAV reporter (Figs 3A and 4A). We also assessed whether the decline in alginate expression was a consequence of decreased abundance in cytosolic sigma factor AlgT. Since AlgT is autoregulatory in P. aeruginosa and free cytosolic AlgT induces expression of the algTmucA-D operon (DeVries and Ohman, 1994; Schurr et al., 1995; Mathee et al., 1997) we reasoned that algT transcript abundance would reflect cytosolic AlgT abundance. Importantly, algT transcript abundance (Fig. 5B) does not change significantly (P > 0.05) over time, suggesting that decreased alginate expression (Figs 3A and 5A) may not be solely a consequence of the lack of free cytosolic AlgT.

image

Figure 5. Differential effect of water limitation and osmolarity on alginate biosynthesis (alg8) and regulatory (algT) gene expression by P. putida mt2 biofilms as measured by qRT-PCR. Water potential was lowered by 1.5 MPa with NaCl or PEG amendments. Solid bars: 6 h; grey bars: 24 h. Error bars indicate the ranges of expression for three to four experiments.

Download figure to PowerPoint

Spatial patterns of alginate expression during biofilm development

Direct microscopic observation of cells corroborated the flow cytometry results. Pseudomonas putida mt2 HcRed (pPalgD–gfp) cells were cultivated under water-replete conditions and flow cytometry revealed that < 1% of the cells in the inoculums expressed the alginate reporter and < 3–5% were dead [propidium iodide (PI)-stained] while microscopy revealed that only 2–3% were found in aggregates of more than two cells. Within 7 h of growth under water-limited conditions we detected at least one cell expressing the pPalgD–gfp reporter in 73.4 ± 1.4% of the 806 microcolonies examined (Fig. S5). Cells that did and did not express pPalgD–gfp coexisted in many microcolonies, although most, if not all, cells in a microcolony expressed the alginate reporter. Interestingly, microcolonies with alginate reporter-expressing cells were comprised of more cells than microcolonies lacking alginate reporter-expressing cells (5–12 versus 2–6 cells). Of the 966 solitary cells examined 6.7 ± 1.5% expressed the pPalgD–gfp transcriptional fusion.

Confocal microscopy was used to visualize the spatial localization of P. putida mt2-HcRed tagged biofilm cells expressing the pPalgD–gfpAAV reporter (Fig. 6) as the biofilm aged. Similar to our quantitative flow cytometry measurements (Fig. 3) the majority of cells expressed the alginate reporter during early phases of biofilm development (Fig. 6A). As biofilms mature, the proportion of alginate-expressing cells decreased (Fig. 3A) and these cells were usually localized in the colony centre surrounded by cells that did not express the alginate reporter (Fig. 6B and C). During the biofilm maintenance phase solitary cells expressing the alginate reporter were observed primarily at the nutritive substratum interface or the colony periphery (Fig. 6C and D).

image

Figure 6. Spatial relationships of P. putida mt2-HcRed (pPalgD–gfpAAV) biofilm cells expressing the alginate reporter when cultivated on media in which the water potential was lowered by 1.5 MPa with PEG amendments. Green fluorescence indicates HcRed (constitutive red fluorescence) tagged cells are expressing the GFP-based alginate reporter. (A): 14 h; (B) 22 h; (C) 48 h; (D) 72 h. Scale bar is 50 µm.

Download figure to PowerPoint

Temporal dynamics of alginate expression by biofilm residents following a drying event

The spatiotemporal dynamics of alginate gene expression may differ in pre-existing biofilms exposed to desiccating conditions from those observed when biofilms form under water-limiting conditions. We explored this possibility by cultivating biofilms on membranes under water-replete conditions for 24 h prior to transferring membranes with intact biofilms to solid media without or with PEG amendments to lower the water potential by 1.5 MPa. Prior to resuming growth following a 6–12 h acclimation period (data not shown), approximately 10% of the biofilm cells expressed the alginate reporter within 3 h after initiating the dehydration event (Fig. 7A). Our findings suggest alginate expression was transient since the proportion of cells expressing the long half-life pPalgD–gfp reporter increased (up to approximately 70% of the population) during the first 24 h after the dehydration shock while only 3–10% of the population actively expressed the short-half life pPalgD–gfpAAV reporter at any particular sampling time (Fig. 7). Given that initially (first 3–24 h) there were fewer alginate-expressing cells (Fig. 7A) with lower levels of induction (approximately twofold; Fig. 8A) within mature biofilms exposed to a drying event than during biofilm formation under water-limiting conditions suggests that one or more biofilm attributes, other than alginate, provides some protection from desiccation stress. A reduction in metabolic activity does not likely explain this difference since the fluorescence intensity per cell of the metabolic activity reporters were similar under both conditions (compare Figs 4B with 8B). We did not observe a similar increase in expression of either reporter in the water-replete control treatments (Fig. S6); however, over time a small proportion of cells highly expressed the alginate reporters (Fig. S7). This could reflect selection for alginate-overexpressing variants or, alternatively, that some cells experience localized alginate inducing conditions.

image

Figure 7. Temporal dynamics of the proportion of P. putida mt2 biofilm cells expressing the algD promoter–gfp transcriptional fusion following a 1.5 MPa water potential dehydration shock. A. pPalgD–gfpAAV (○); pPalgD–gfp (●). B. pPnptII–gfpAAV (□); pPnptII–gfp (inline image). The flow cytometer was gated as described in Experimental procedures. Data points are the mean and error bars the SEM from two experiments each comprised of three replications. Error bars that are not visible are masked by the symbol.

Download figure to PowerPoint

image

Figure 8. Temporal dynamics of the extent of algD promoter–gfp transcriptional fusion induction by P. putida mt2 biofilm cells following a 1.5 MPa water potential dehydration shock. A. pPalgD–gfpAAV (○); pPalgD–gfp (●). B. pPnptII–gfpAAV (□); pPnptII–gfp (inline image). The flow cytometer was gated as described in Experimental procedures. Fluorescence intensity per cell is in arbitrary units. Data points are the mean and error bars the SEM from two experiments each comprised of three replications. Error bars that are not visible are masked by the symbol.

Download figure to PowerPoint

Spatial localization of alginate expression by biofilm residents following a drying event

Spatial patterns of alginate-expressing cells in biofilms exposed to a dehydration shock (Fig. 9) were distinct from those observed during growth under water-limiting conditions (Fig. 6). Cells expressing the alginate reporters were in local areas in the centre of the colony and at the colony periphery within 3 h after the dehydration shock (Fig. 9A and B). Cell growth resumed 6–12 h after the dehydration shock and most cells expressing the alginate reporters were localized at the colony periphery (Fig. 9B and C). By 24 h after the dehydration shock most cells had expressed the alginate reporter, including those in the colony centre (Fig. 9D).

image

Figure 9. Spatial relationships of P. putida mt2-HcRed biofilm cells cultivated under water-replete conditions expressing the algD promoter–gfp transcriptional fusions following exposure to a 1.5 MPa decrease in water potential imposed by PEG amendments. Green fluorescence indicates HcRed (constitutive red fluorescence) tagged cells were expressing the GFP-based alginate biosynthesis reporter. Confocal microscopy images were taken at various times after initiation of the drying event. All overhead (x–y) images are a single scan, except (C) which is comprised of the entire stack. (C) pPalgD–gfpAAV; (A, B and D) pPalgD–gfp; (A) 3 h; (B) 7 h; (C) 18 h; (D) 23 h. Scale bar is 50 µm.

Download figure to PowerPoint

Discussion

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

We previously demonstrated that alginate production by P. putida buffers biofilm cells from drying and influences biofilm architecture when water is limiting (Chang et al., 2007). To further understand alginate production under water-limiting conditions we monitored expression of algD promoter–gfp transcriptional fusions using long and short half-life GFP reporters. Our fluorescent reporters identify cells that initiated a key requirement for alginate production, expression of the transcript for alginate biosynthesis and export. We show that most, if not all, biofilm cells express the alginate biosynthesis and export operon, albeit transiently, suggesting that alginate production is a group effort and not mediated by a subpopulation dedicated to polymer production. Whether cells expressing the algD operon actually secrete alginate is less clear given that alginate export is regulated post-translationally by cyclic-di-GMP (Merighi et al., 2007).

The patterns of alginate expression and metabolic activity were remarkably consistent and correlate specifically with the context in which biofilm cells were deprived of water. We used the constitutive neomycin phosphotransferase (nptII) promoter as an indicator of metabolic activity, which revealed biofilm cells were still capable of expressing GFP when detectable alginate expression ceased (Figs 3, 4, 7 and 8). It is unlikely that oxygen depletion within mature P. putida biofilms limits GFP folding (Werner et al., 2004) and our interpretation of localization of alginate-expressing cells since cells expressing the constitutive metabolic reporter were detected at all depths (data not shown). Alginate expression coincides with microcolony formation and continues through the early phases of biofilm development (Figs 2A and 6A) typically in cells at the colony periphery (Figs 6 and 9). Transient alginate expression could be a consequence of metabolic costs associated with alginate production, feedback regulation indicating sufficient alginate is synthesized to hydrate the cell, or one or more unidentified regulatory networks controlling partitioning of cellular resources during biofilm development.

Our observation that during initial colonization of a water-limited environment, solitary individuals rarely expressed the alginate reporter, that most microcolonies were comprised of at least one alginate-expressing cell, and that microcolonies harbouring alginate-expressing cells were comprised of more cells than microcolonies lacking alginate reporter-expressing cells was intriguing (Fig. S4). Taken together, these observations suggest that microcolony formation may be a prerequisite for alginate gene expression, which could reflect that cells capable of microcolony formation are also capable of alginate expression. Similarly, Monier and Lindow (2003) observed that there was preferential survival of Pseudomonas syringae cells that resided in small aggregates on dry bean leaves, which could be a consequence of quorum-sensing-controlled production of alginate (Quinones et al., 2005) that requires a small quorum size in non-water-saturated environments (Dulla and Lindow, 2008). At present there is no evidence for quorum-sensing control of alginate production by P. putida, but there is clearly a relationship between microcolony formation and alginate gene expression in dry environments.

Regulation of alginate gene expression is controlled by several signal transduction networks integrating environmental and intracellular cues that modulate transcriptional activity of the algD operon. In P. aeruginosa, and likely other Pseudomonads (Fakhr et al., 1999), the hierarchy begins with the alternative sigma factor AlgT, which is liberated from a trans membrane complex of negative regulators in response to regulated proteolysis activated by cell envelope stress (Wood and Ohman, 2009). Transient alginate expression in our study is likely not a consequence of transient availability of free cytosolic AlgT since algT (Fig. 5) and mucA (data not shown) gene expression did not change to the same extent over the time period in which algD promoter activity decreased (Figs 3 and 7). Our alignment between the P. putida algD operon promoter region with those of various Pseudomonads revealed relatively strong sequence similarities with all known DNA binding sites of transcriptional factors characterized in P. aeruginosa (Fig. 1 and Fig. S1). Thus, other environmental or physiological signals modulating the activities of ancillary transcriptional factors, such as those found in P. aeruginosa, may play a role in transient P. putida alginate expression. The tendency for alginate-expressing cells to be localized at the microcolony periphery (Figs 5 and 9) or in low-population-density regions (images not shown) suggests nutrient availability may be important. Alternatively, alginate expression is transient because of some as yet identified feedback signal indicating that the cell is sufficiently hydrated.

Biofilms are often thought to require high levels of cooperation because extracellular polymeric substances are a shared resource produced by one cell that can be used by others. In the absence of substantial bulk water flow alginate will be localized near the cell(s) that synthesize it and sharing of this secreted product will be restricted to neighbouring cells. Our available evidence supports this premise since most if not all biofilm cells expressed alginate. There is ample evidence showing that distinct subpopulations mediate specific biofilm developmental processes (Vlamakis et al., 2008; Lopez et al., 2009). In fully hydrated, flow-through P. aeruginosa biofilm systems various subpopulations have been shown to participate in a sequential process leading to the formation of the initial microcolonies, mushroom-shaped structures (Klausen et al., 2003; Barken et al., 2008), biosurfactant production (Lequette and Greenberg, 2005), and release of DNA by the outer layer of cells in the stalks (Allesen-Holm et al., 2006; Yang et al., 2007). Our lack of evidence for alginate gene expression by a specific subpopulation may reflect the physical constraints of unsaturated biofilms, requiring each resident to encapsulate themselves with this hydrophilic polymer to facilitate survival when deprived of water.

In our studies we assessed whether residents of mature biofilms that formed under water-replete conditions induce alginate expression in response to dehydration stress in a manner similar to cells that form biofilms under water-limiting conditions. We predicted that a layer of cells at the biofilm–air interface would most likely produce alginate since they would be the most water-deprived and because deposition of alginate at the biofilm–air interface would best facilitate biofilm hydration. Our data do not support our prediction. Within hours after exposure to a dehydration shock >10% of the residents express the alginate reporter (Fig. 3A) with most alginate-expressing cells localized to the colony periphery and in foci in the centre of the colony (Fig. 7A); eventually almost all residents expressed alginate. Initially following the dehydration shock the extent of alginate induction was lower than during biofilm formation under water-limiting conditions (compare Fig. 3A with Fig. 8A). This is likely a consequence of the extent of stress the cells experience rather than a consequence of reduced metabolic capabilities since expression of the constitutive metabolic reporters were similar regardless of the context in which the cells were deprived of water (Figs 3B and 8B).

In conclusion, we observed transient alginate gene expression under water-limiting conditions over time in reproducible spatial patterns in a context-dependent fashion. Alginate expression appears to be important for microcolony formation under water-limiting conditions and for pre-existing biofilms to tolerate desiccation stress. Although speculative, our findings suggest that biofilm matrix components other than alginate or the biofilm structure itself confers some protection from dehydration stress; this is a subject of further study in our laboratory. However, this attribute does not diminish the importance of alginate to ecological success. Collectively, our findings suggest that P. putida integrates cues on water abundance into regulatory networks controlling alginate production in a temporally and spatially controlled manner which is influenced by unidentified physiological or environmental cues. In a sense, P. putida cells may be responding to their own local environment, producing alginate because of the fitness advantage it confers under those particular conditions.

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, media and chemicals

Pseudomonas putida strain mt-2 was tagged with miniTn7HcRed in the neutral att site which was verified by PCR as described previously (Lambertsen et al., 2004). Preliminary studies showed that growth rates of P. putida mt-2 with and without the HcRed tag were comparable. Unless indicated otherwise, cells were cultivated on TYE medium, comprised of 2.5 g of tryptone, 1.25 g of yeast extract, 0.7 g of KH2PO4 and 40 ml of Hutner's mineral solution (Smibert and Krieg, 1994) per litre of deionized water. CTYE is comprised of 21-C medium (Halverson and Firestone, 2000) amended with 0.63 g of tryptone and 0.32 g of yeast extract. Media were solidified with 10 g of phytagel gellan gum (Sigma) per litre. The water potential of the medium was lowered with the non-permeating solute polyethylene glycol MW 8000 (PEG) or the permeating solute NaCl to simulate a matric or solute stress, as described previously (Halverson and Firestone, 2000). All fluorescent probes were obtained from Invitrogen.

Construction of PalgD–gfp transcriptional fusions

The upstream promoter region of the P. putida algD gene (PP1288) was PCR amplified from mt-2 genomic DNA with primers AlgDprm-F and AlgDprm-R (Table 1). This 781 bp product comprised of the promoter region and the first 10 bases of the algD gene was then cloned into pTOPO 2.1 (Chang et al., 2007). PCR conditions were as follows: 25 cycles of 94°C, 54°C, 72°C at 30 s each, with a final extension time of 2 min at 72°C. pTOPO-PalgD was digested with XbaI and SacI, and the resulting 870 bp fragment was isolated prior to cloning into pPROBE-NT or -NTAAV containing a promoterless gfp gene (Miller et al., 2000), to generate pPalgD–gfp or pPalgD–gfpAAV. Similarly, the upstream promoter region of the neomycin phosphotransferase nptII gene previously cloned into pPROBE-KT (Chang and Halverson, 2003) was PCR amplified with primers nptII-F and nptII-R that was then cloned into pTOPO (Invitrogen). pTOPO-nptII was digested with XbaI and SacI, and the resulting fragment was isolated prior to cloning into pPROBE-NT or -NTAAV containing a promoterless gfp gene, to generate pPnptII–gfp or pPnptII–gfpAAV. These plasmids were then electroporated into P. putida mt-2 or mt2-HcRed.

Biofilm cultivation and exposure to a drying event

Overnight plate cultures were re-suspended and diluted to an OD600 of 0.1 in liquid media prior to transferring 5 µl aliquots onto TYE plates with or without NaCl or PEG amendments. Biofilms were cultivated at 28°C for various lengths of time and removed from the medium using a sterile cork borer prior to re-suspending cells in sterile water, and sonicating for 5 min in an ultrasonic water bath (Branson) for flow cytometry or fluorometry. For experiments assessing the effect of a drying event on alginate expression, a 100 µl aliquot of a 24-h-old surface-grown culture (OD660 = 0.001) was filtered onto an 80-mm-diameter nylon membrane (MSI, Westboro, MA) and then overlaid onto TYE solid medium. Membranes containing biofilms were incubated for 24 h at 28°C before transferring membranes to a solid medium in which the water potential was lowered by 1.5 MPa with PEG to create a matric shock (Chang et al., 2007). Colonies on the membranes ranged from 0.1 to 0.5 mm in diameter. As a control, membranes containing colonies were also transferred to medium without PEG. After membrane transfer, biofilms were incubated at 28°C for various lengths of time, scrapped off the surface with a sterile spatula, transferred to phosphate buffer diluent, mixed by vortexing, and then the cell suspension was sonicated for 5 min prior to flow cytometry. Alternatively, sections of the membranes with biofilms were transferred to the unsaturated biofilm chamber system for microscopy.

Flow cytometry

Before measuring the samples, we gated the flow cytometer at each sampling period with fresh overnight cultures of cells harbouring pPalgD–gfpAAV, empty vector and pPnptII–gfpAAV grown under water-replete conditions. Bacteria were detected in a dot plot of side-scatter versus Syto 60 red fluorescence, to ensure that all counted particles were cells. To establish gates for GFP-positive cells, the fluorescence intensity of GFP-expressing cells had to be greater than the fluorescence intensity of 99% of a cell population with the same reporter cultivated overnight under water-replete conditions. This approach resulted in similar gates to those derived from cells containing an empty vector. Dual-colour flow cytometry was performed on a BD FACSCanto flow cytometer at the Iowa State University Cell Facility. The 488 nm argon laser was used to excite GFP, or PI or Syto 60 and the 530/30 nm (emission wavelength/bandpass) and 610/20 nm filter sets were used for the detection of GFP and PI or Syto 60 respectively. Propidium iodide was used to measure the proportion of dead cells in the population and only PI-negative cells were included in our analyses. The emission intensities of 10 000–20 000 GFP-positive cells were counted per sample.

Microscopy

We used P. putida mt2-HcRed cells harbouring either pPalgD–gfpAAV or pPalgD–gfp reporters to visualize cells and alginate expression in biofilms over time using a previously described unsaturated biofilm chamber system (Chang and Halverson, 2003). Inocula were prepared by serially diluting 24-h-old plate cultures in TYE broth prior to transferring 0.5 µl aliquots onto a medium-coated coverslip that was then mounted in the biofilm chamber system. Image acquisition was performed with a Nikon C1si confocal laser scanning microscope at the ISU Confocal Microscopy Facility. The confocal microscope was equipped with an argon and NeHe laser and detectors with filter sets for simultaneous monitoring of GFP (excitation 488 nm; emission 517 nm) and HcRed (excitation 590 nm; emission 637 nm). Images were obtained using a dry 20× or 60× oil objectives. Simulated 3D images were generated using C1plus/C1si Control software EZ-C1.

Relationship between microcolony formation and algD–gfp expression

Overnight cultures of P. putida mt2-HcRed (pPalgD–gfp) were cultivated under water-replete conditions, diluted to an OD600 of 0.001 and 2 µl aliquots were transferred onto solid media in which the water potential was lowered by 1.5 MPa with PEG. Biofilms were cultivated for 28°C for 7–8 h in unsaturated biofilm chambers to maintain the desired relative humidity. We used a Nikon E80i epifluorescence microscope with GFP-HYQ (Ex: 450–490; DM: 495; BA: 500–550) and TRITC-HYQ (Ex: 530–560; DM: 570; BA: 590–650) filter cubes for visualizing GFP-expressing and HcRed tagged cells respectively. The experiment was conducted three times comprised of two subsamples per each of three replications. At least 30 microcolonies and 30 solitary cells were counted per subsample.

Quantitative real-time PCR

Pseudomonas putida was cultivated overnight on CTYE solid medium, re-suspended in CTYE broth to an OD600 nm of 0.1 prior to inoculating plates amended with or without NaCl or PEG to lower the water potential by 1.5 MPa. For the 6 h treatments 1.0 ml of the cell suspension was spread onto a 80-mm-diameter nylon membrane (GE) overlaying a plate while for the 24–72 h treatments the cell suspension was diluted 1000-fold prior to inoculating 100 µl onto a quadrant of a nylon membrane. Plates were incubated at 28°C in separate plastic containers to maintain the desired relative humidity. Whole membranes or one to two quadrants were removed and placed into a 2:1 mixture of RNAprotect (Qaigen):CTYE diluent, vortexed, and then sonicated to remove biofilm cells from the membrane. Cells were centrifuged at room temperature and cell pellets were stored at −20°C for up to 3 days. RNA was extracted using the RNeasy Mini Kit (Qiagen) according to manufacturer's directions. DNase digestion was performed twice; an on-column digestion during RNA isolation using Qiagen RNase-free DNase followed by a post RNA isolation using DNA Turbo DNase (Ambion). RNA was quantified with a Nanodrop-1000 (Thermo Scientific) and RNA quality was verified using a Bioanalyser (Agilent). Primers for RT-PCR are described in Table 1. QScript 1-step Sybr Kit (Quanta Biosciences) was used for generation and amplification of cDNA using an Opticon2 (Bio-Rad) thermocycler. Data were normalized to the RimM gene and are expressed relative to water-replete controls.

Acknowledgements

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

We thank Margaret Carter of the ISU Image Analysis and Confocal Microscopy Facility and Shawn Rigby of the ISU Flow Cytometry Facility for their technical assistance. This research was supported by the National Science Foundation under Grant 0446292 and by the Iowa Agriculture and Home Economics Experiment Station.

References

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

Supporting Information

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

Fig. S1. Sequence alignment of the algD promoter region of P. aeruginosa PAO1 (PAO1), P. putida KT2440 (KT2440) and P. syringae FF5 (FF5). AlgR, CRP and AmrZ binding sites are boxed, the AlgB binding site is underlined, and the sigma factor AlgT recognition site is bold and underlined. The algD translational start site is indicated by algD[RIGHTWARDS ARROW]. Identical bases in the alignment between PAO1, KT2440 and FF5 are shaded. Gaps are indicated by dashes. The pairwise alignment between PAO1 and KT2440 was generated using clustalw with a gap open penalty of 15.0 and gap extension of 6.66 using the IUB matrix. FF5 sequence was aligned to the PAO1 and KT2440 sequence using ClustalWProf using the same gap penalties listed above. The sequences selected for alignment were based on the promoter sequence used to construct our alginate reporter systems. Based on blastp analysis at the http://www.pseudomonas.com website, P. putida has homologues to the PAO1 σ22 (PP1427; 92% similarity), AlgR (PP0185; 88% similarity), AmrZ (PP4470; 93% similarity), AlgB (PP0133; 90% similarity) and CRP (PP0424; 85% similarity) transcriptional factors.

Fig. S2. Influence of the severity of water limitation on expression of algD promoter–gfp transcriptional fusion by P. putida mt2 (pPalgD–gfp) biofilm cells. pPalgD–gfp (?); empty vector (?). Biofilm cells were isolated after 24 h of growth, vortexed, and sonicated in an ultrasonic water bath for 5 min prior to measuring GFP expression by fluorometry. Results are expressed relative to water-replete (0 MPa ψ) conditions. Data points are the mean fold induction and error bars the SEM of three replications. Error bars that are not visible are masked by the symbol. Similar results were obtained with P. putida (pPalgD–gfpAAV) on various media, although the fluorescence intensity was usually greater (data not shown).

Fig. S3. Temporal dynamics of the proportion of P. putida cells expressing the pPalgD–gfp (A) and pPnptII–gfp (B) transcriptional fusions during biofilm development. 0 MPa ψ (?); −1.0 MPa ψ PEG (?); −1.5 MPa ψ PEG (?). The flow cytometer was gated as described in Experimental procedures. Data are representative of one of three experiments with similar results. Data points are the mean and error bars the SEM of three replications. Error bars that are not visible are masked by the symbol.

Fig. S4. Temporal dynamics of the extent of pPalgD–gfp (A) and pPnptII–gfp (B) transcriptional fusion expression by P. putida mt2 biofilm cells during growth under water-replete and -limiting conditions. 0 MPa ψ (?); −1.0 MPa ψ PEG (?); −1.5 MPa ψ PEG (?). The flow cytometer was gated as described in Experimental procedures. Fluorescence intensity per cell is in arbitrary units. Data are representative of one of three experiments with similar results. Data points are the mean and error bars the SEM of three replications. Error bars that are not visible are masked by the symbol.

Fig. S5. Relationship between algD promoter–gfp transcriptional fusion expression and microcolony formation by P. putida mt2 HcRed (pPalgD–gfp) cells under water-limiting conditions. We monitored green fluorescence in HcRed-tagged (constitutive red fluorescent) cells or microcolonies by epifluorescence microscopy as described in Experimental procedures. Water potential of the medium was lowered by 1.5 MPa with PEG amendments. Values are the means ± SEM of three experiments.

Fig. S6. Temporal dynamics of the proportion of P. putida mt2 biofilm cells expressing the algD promoter–gfp transcriptional fusion in the absence of dehydration shock. pPalgD–gfpAAV (?); pPalgD–gfp (?); pPnptII–gfpAAV (◊); pPnptII–gfp (?). The flow cytometer was gated as described in the legend to Fig. 2. Data points are the mean and error bars the SEM from two experiments each comprised of three replications. Error bars that are not visible are masked by the symbol.

Fig. S7. Temporal dynamics of the extent of algD promoter–gfp transcriptional fusion induction by P. putida mt2 biofilm cells in the absence of a dehydration shock. pPalgD–gfpAAV (?); pPalgD–gfp (?); pPnptII–gfpAAV (◊); pPnptII–gfp (?). Fluorescence intensity per cell is in arbitrary units. The flow cytometer was gated as described in the legend to Fig. 2. Data points are the mean and error bars the SEM from two experiments each comprised of three replications. Error bars that are not visible are masked by the symbol.

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.

FilenameFormatSizeDescription
EMI_2186_sm_Fig_S1-S7.doc102KSupporting 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.