Uncovering behavioural diversity amongst high-strength Pseudomonas spp. surfactants at the limit of liquid surface tension reduction

Uncovering behavioural diversity amongst high-strength Pseudomonas spp. surfactants at the limit... Abstract Bacterial biosurfactants have a wide range of biological functions and biotechnological applications. Previous analyses had suggested a limit to their reduction of aqueous liquid surface tensions (γMin), and here we confirm this in an analysis of 25 Pseudomonas spp. strains isolated from soil which produce high-strength surfactants that reduce surface tensions to 25.2 ± 0.1–26.5 ± 0.2 mN m−1 (the surface tension of sterile growth medium and pure water was 52.9 ± 0.4 mN m−1 and 72.1 ± 1.2 mN m−1, respectively). Comparisons of culture supernatants produced using different growth media and semi-purified samples indicate that the limit of 24.2–24.7 mN m−1 is not greatly influenced by culture conditions, pH or NaCl concentrations. We have used foam, emulsion and oil-displacement behavioural assays as a simple and cost-effective proxy for in-depth biochemical characterisation, and these suggest that there is significant structural diversity amongst these surfactants that may reflect different biological functions and offer new biotechnological opportunities. Finally, we obtained a draft genome for the strain producing the highest strength surfactant, and identified a cluster of non-ribosomal protein synthase genes that may produce a cyclic lipopeptide (CLP)-like surfactant. Further investigation of this group of related bacteria recovered from the same site will allow a better understanding of the significance of the great variety of surfactants produced by bacterial communities found in soil and elsewhere. Pseudomonas, surfactant, limit to liquid surface activity, cyclic lipopeptide, non-ribosomal protein synthase INTRODUCTION Biosurfactants produced by bacteria are surface-active agents having a wide range of biological activities including involvement in the solubilisation of hydrophobic substrates, co-ordinated growth and differentiation, cell motility, surface attachment and biofilm development, suppression of competitors and protection from predators, immune modulation and virulence, rotting of plant tissues, causing fungal hyphae swelling and the lysis of oomycete zoospores (Ron and Rosenberg 2001; Abdel-Mawgoud, Lépine and Déziel 2010; Raaijmakers et al.2010). These compounds also have many applications in cosmetic, food, medical, pharmaceutical, oil and bioremediation technology where new high-strength surfactants are constantly in demand as detergents, wetting and foaming agents, emulsifiers and dispersants (Franzetti et al.2010; Marchant and Banat 2012; Gudiña et al.2013; Souza, Vessoni-Penna and de Souza Oliveira 2014; Inès and Dhouha 2015). The activity of surfactants depends on their amphiphilic nature, and a number of different structural classes of surfactants are produced by bacteria (Desai and Banat 1997), including cyclic lipopeptides (CLPs) and rhamnolipids (Abdel-Mawgoud, Lépine and Déziel 2010; Raaijmakers et al.2010). However, the relationship between surfactant activity, biological function (or role) and structural diversity remains poorly understood; we need to separate activity resulting from the fundamental biophysical properties of surfactants from those biological activities that provide the surfactant producer with a selective advantage. For example, very few bacteria would have a selective advantage in lysing erythrocytes, yet this is a common assay for surfactant production (Youssef et al.2004; Afshar et al.2008). Similarly, it is not clear how much of the observed structural diversity amongst surfactants is relevant or redundant. Furthermore, in complex soil or plant-associated communities, where different bacteria are capable of producing a range of surfactants, are these treated as public goods benefiting the whole community or does this represent intracommunity conflict and competition? Our research has focussed on assessing bacterial surfactant strengths and behavioural diversity within the Pseudomonas genus using behavioural assays as a simple and cost-effective proxy for the in-depth biochemical characterisation required to determine structural diversity (Fechtner et al.2017). This genus includes plant and mushroom pathogens that use surfactants to rot tissues as well as many surfactant-producing soil and plant-associated strains found in complex communities where the suppression of the growth of competitors and protection from predators may be particularly important; in addition, surfactants are also required for swarming motility and biofilm maturation (Raaijmakers et al.2010). Recent investigations of high-strength surfactants produced by pseudomonads and other bacteria have suggested that there is a limit (γMin) to the extent surfactants can reduce aqueous liquid surface tension of 24.16–24.24 mN m–1 (Fechtner et al.2011; Mohammed et al.2015) and the biological basis for this is probably the need to minimise self-damage to the producing cells (Fechtner et al.2017). To put this into context, the surface tension of water at 20°C is 72.8 mN m–1 (Vargaftik, Volkov and Voljak 1983), while the sterile media used in these predictions have surface tensions of between 47.0 and 59.6 mN m–1 (Fechtner et al.2011; Mohammed et al.2015). In this work, we want to test the robustness of the prediction by investigating surfactant production amongst a collection of pseudomonads isolated from the same soil community, to determine whether culture and buffer conditions significantly alter liquid surface tension measurements and γMin, and to assess the structural diversity amongst the high-strength surfactants produced by these strains, which may represent a valuable resource for future biotechnological exploitation. MATERIALS AND METHODS Bacterial isolation and cultivation Pseudomonas spp. or Pseudomonas-like strains were isolated from samples taken from bulk soil underlying a section of managed grass lawn at the Dundee Botanic Garden (DBG; Dundee, UK) in February and April 2015. Bacteria were isolated using selective agar (PSA-CFC; Oxoid, UK) spread with soil suspension dilutions and incubated under aerobic conditions for 2–3 days at 20°C–22°C. Colony material resuspended in deionised water (DI) was used to test for surfactant production using the drop collapse assay on petri dish lids as per Persson and Molin (1987), and then confirmed by quantitative tensiometry of modified King's B (KB*; Kuśmierska and Spiers 2016) culture supernatants (see below). Twenty-five surfactant-producing strains (DBG strains 1–25) plus five randomly chosen drop collapse-negative strains (DBG strains c1–c5) were retained for further investigation and stored at –80°C in 15% (v/v) glycerol. Overnight KB* and minimal medium containing 20 mM glucose (M9-Glu; Fechtner et al.2011) cultures incubated with shaking at 28°C were used to prepare samples for testing as required. Strain characterisation and identification Phenotypes were determined using biochemical, growth and behaviour-based assays at 20°C–22°C as per Robertson et al. (2013) (see Supplementary Information for further details), and hierarchical cluster analysis (HCA) was used to group strains on the basis of similarity as per Robertson et al. (2013) and Mohammed et al. (2015). Key strains were further analysed by metabolic profiling using API 20e cards (BioMérieux, Basingstoke, UK) and partial 16S rRNA gene sequencing to determine genus-level identification (see Supplementary Information for further details). Surfactant behaviour and surface tension measurements Twenty-four hour KB* cultures were used to investigate surfactant behaviours using emulsion, foam stability and oil displacement assays at 20°C–22°C as per Coffmann and Garcia (1977), Cooper and Goldenberg (1987) and Morikawa et al. (1993) (see Supplementary Information for further details), and HCA was used to cluster surfactant behaviours on the basis of similarity. For the oil displacement assays (also known as oil spreading assays), mineral oil, vegetable oil, used lubricating oil and diesel were overlaid onto DI water (pH 6), 200 mM NaCl (pH 6) and 50 mM Tris (pH 8) solutions. Surfactants were semi-purified from 24-h KB* cultures by an acid precipitation method adapted from De Souza et al. (2003) and resuspended in DI water to test critical micelle concentrations (CMCs), pH and NaCl surface tension profiles (see Supplementary Information for further details). Quantitative tensiometry of semi-purified surfactant solutions and cell-free 24-h KB* or M9-Glu culture supernatants were performed using a Krüss K100 Mk2 Tensiometer (Krüss GmbH, Hamburg, Germany) at 20°C as per Koza et al. (2009), and mean surface tension measurements are rounded up to one decimal place. In these assays, the surface tension of pure water was 72.1 ± 1.2 mN m–1, and the surface tension of sterile KB* and M9Glu culture media was 52.9 ± 0.4 and 70.7 ± 0.7 mN m–1, respectively. Statistical analyses Experiments were performed with replicates, and means with standard errors (SE) are shown where appropriate. Data were assumed to be normally distributed and were examined using JMP v12 statistical software (SAS Institute Inc., Marlow, UK) with comparisons of means performed using Student's and matched pairs t-tests (t), one-way analysis of variance (ANOVA) (F) models with Tukey–Kramer honest significant difference (HSD) (q*) post hoc tests and correlations (r) examined by multivariate analysis. HCA using the Ward Method with equal weightings was used to investigate similarities between strain phenotypes and surfactant behaviours as per Robertson et al. (2013) and Mohammed et al. (2015). Analyses based on general linear models (GLMs) were used to investigate surface tension and oil displacement data with effects further examined using LSMeans Differences Tukey HSD (Q) tests (see Table S1, Supporting Information for model details, covariates and effects tests). The minimum liquid surface tension (γMin) was determined by individual distribution identification (IDI) as per Fechtner et al. (2011) using mean surface tension data and based on the lowest Anderson-Darling (AD) goodness of fit test value using MINITAB v1.5 statistical software (Minitab Ltd, Coventry, UK). DBG-1 draft genome and identification of possible surfactant synthesis genes The DBG-1 draft genome was determined using the microbial sequencing and strain repository service MicrobesNG (https://microbesng.uk; Birmingham, UK), and trimmed reads and fasta files are available on request. Sequencing was performed on Illumina MiSeq and HiSeq 2500 platforms using 2× 250 bp pair-end reads, and data were put through a standard analysis pipeline for assembly and quality analysis (see https://microbesng.uk for further details). A mean coverage of 42.6× was achieved with 656 944 reads, producing a draft genome of 6860 106 bp comprised of 122 contigs of which the largest was 657 704 bp and a guanine and cytosine (GC) content of 58.9%. A total of 6082 coding sequences (CDS) were predicted within contigs, with an average length of 976 bp and density of 0.89 per kb, and annotations were provided where possible by automated BLAST analyses. A total of 69 tRNA genes were identified, though no rRNA genes were found including the 16S rRNA genes required for species-level identification. Read mapping suggests that this genome is most closely related to the P. fluorescens species, which is consistent with our isolation and selection of the strain as a fluorescent pseudomonad. CDS annotations associated with non-ribosome protein synthases (NRPS) were inspected manually and confirmed by NCBI/NLM BLASTP against non-redundant GenBank CDS translations, PDB, SwissProt, PIR and PRF databases (https://blast.ncbi.nlm.nih.gov/Blast.cgi). The draft DBG-1 genome was also submitted to antiSMASH (https://antismash.secondarymetabolites.org; bacterial version; Weber et al.2015) to further characterise NRPS homologues and predict possible products. RESULTS AND DISCUSSION Isolation of Pseudomonas spp. expressing strong surfactants that significantly lower liquid surface tension We isolated a collection of Pseudomonas spp. or Pseudomonas-like bacteria from soil that produced high-strength surface-active agents or surfactants when incubated in KB* cultures for 24 h, and from a statistically homogeneous Tukey–Kramer HSD group (q* = 3.970; α = 0.05) chose the 25 strains producing the lowest surface tensions for further analysis (DBG-1–25; Fig. 1). Although these showed similar surface tension reducing activity ranging between 25.2 ± 0.1 and 26.5 ± 0.2 mN m–1, a comparison of strain phenotypes by HCA, which also included control strains not producing surfactants under the conditions used here (DBG-c1–5), indicated that most could be differentiated by one or more colony morphology, enzyme and siderophore expression, antibiotics and mercury sensitivity, salt and high temperature tolerance, and motility assays (Fig. 2; see Table S2, Supporting Information, for the ordinal data set), with little evidence of biological replication (i.e. the isolation of the same strain more than once). Further testing of key strains using metabolic profiling and partial 16S rRNA gene sequencing suggests that most are probably Pseudomonas spp. (see Table S3, Supporting Information, for putative identifications). This collection of phenotypically diverse pseudomonads producing high-strength surfactants provided us with an opportunity to test the robustness of earlier predictions of the minimum limit (γMin) to liquid surface tension reduction achieved by bacterial surfactants, and then to examine the degree of structural diversity within a group of high-strength surfactants. Figure 1. View largeDownload slide Identification of a group of a homogeneous group of pseudomonads producing high-strength surfactants. A statistical approach was taken to identify 25 DBG strains producing high-strength surfactants. Differences between mean liquid surface tension measurements of 24-h KB* culture supernatants were determined by a Tukey–Kramer HSD test, and DBG-1–29 were found to form a single homogeneous group (α = 0.05). From these, the first 25 strains (dark grey) were chosen for further analysis. Means ± SE are shown (n = 4), and means not linked by the same letter are significantly different (q* = 3.970, α = 0.05). The ST of sterile KB* was 52.9 ± 0.4 mN m–1 (not shown). Figure 1. View largeDownload slide Identification of a group of a homogeneous group of pseudomonads producing high-strength surfactants. A statistical approach was taken to identify 25 DBG strains producing high-strength surfactants. Differences between mean liquid surface tension measurements of 24-h KB* culture supernatants were determined by a Tukey–Kramer HSD test, and DBG-1–29 were found to form a single homogeneous group (α = 0.05). From these, the first 25 strains (dark grey) were chosen for further analysis. Means ± SE are shown (n = 4), and means not linked by the same letter are significantly different (q* = 3.970, α = 0.05). The ST of sterile KB* was 52.9 ± 0.4 mN m–1 (not shown). Figure 2. View largeDownload slide The DBG pseudomonads producing high-strength surfactants are phenotypically diverse. HCA was used to determine similarities between DBG strain phenotypes using biochemical, growth and behaviour-based assays. Shown here is an HCA constellation plot that clusters similar strains in terminal (short) branches and links strains with greater differences with longer branches. The plot is arbitrarily rooted mid-way along the longest branch (circled) and six major groups (grey arcs) determined automatically. Strains producing surfactants with the highest strength, DBG-1–5, are indicated by black circles, and the remainder of the surfactant-producing strains, DBG-6–25, are indicated by the grey circles. Non–surfactant-producing DBG-c1–c5 control strains are indicated by white circles. Those strains that have been identified as likely Pseudomonas spp. are shown underlined. Figure 2. View largeDownload slide The DBG pseudomonads producing high-strength surfactants are phenotypically diverse. HCA was used to determine similarities between DBG strain phenotypes using biochemical, growth and behaviour-based assays. Shown here is an HCA constellation plot that clusters similar strains in terminal (short) branches and links strains with greater differences with longer branches. The plot is arbitrarily rooted mid-way along the longest branch (circled) and six major groups (grey arcs) determined automatically. Strains producing surfactants with the highest strength, DBG-1–5, are indicated by black circles, and the remainder of the surfactant-producing strains, DBG-6–25, are indicated by the grey circles. Non–surfactant-producing DBG-c1–c5 control strains are indicated by white circles. Those strains that have been identified as likely Pseudomonas spp. are shown underlined. Testing the predicted limit to liquid surface tension reduction We used IDI to predict γMin following the method established by Fechtner et al. (2011). The best fit for the mean surface tension data was found using the three-parameter log-logistic distribution (AD = 0.721) and produced a value of 24.7 mN m–1 for γMin similar to the earlier predictions of around 24.2 mN m–1 determined for a number of different groups of bacteria (Fechtner et al.2011; Mohammed et al.2015). To put this range of predictions into context, it is lower than the standard errors in our measurement of the surface tension of DI water at 20°C, which we undertake as an internal control (72.1 ± 1.2 mN m–1) and less than the change in surface tension of water between 20°C and 25°C (72.8 and 72.0 mN m–1, respectively; Vargaftik, Volkov and Voljak 1983) (in comparison, the surface tension of an 80% (w/w) solution of ethanol at 20°C is 24.3 mN m–1; Vázquez, Alvarez and Navaza 1995). This suggests that these predictions are centring on a common value for γMin; however, differences in pH and solute concentrations in culture media and resuspension buffers might be expected to alter surface tension measurements and γMin in an assay-dependent manner. We decided to explore this further by comparing KB* culture supernatant surface tensions with measurements taken from M9-Glu minimal medium and semi-purified surfactants resuspended in DI water. We obtained a slightly higher γMin prediction from the M9-Glu cell-free culture supernatants of 25.0 mN m–1 using a three-parameter Gamma distribution (AD = 0.917). However, although there was no significant correlation between KB* and M9-Glu strain means (P = 0.27), there were significant differences between pairs (t = –3.7527, P = 0.0003) with seven strains showing more activity and six strains showing less activity in M9-Glu than might have been expected when compared to KB* (see Fig. S1, Supporting Information, for a comparison of KB* and M9-Glu ST means). This suggests that the surfactants produced by some strains were differentially sensitive to the media or final culture supernatant used to determine surface tension measurements. In order to investigate this further, we semi-purified surfactants produced by 14 strains representing the range of surfactant strengths produced by this collection (DBG-1–5, 7, 10, 14–16, 20, 21 and 25). We progressively diluted these samples with DI water to demonstrate that in each case surfactants were produced in KB* cultures above the CMC, and that minor differences in concentration could not explain the differences in surface tension seen between strains or between KB*, M9-Glu and resuspended surfactant surface tension measurements. We then modelled surface tension measurements using a GLM, and found that strain and assay environment were significant effects (GLM model I, P < 0.0001; see Table S1, Supporting Information, for further details), with the resuspended surfactant surface tensions significantly higher than both culture supernatant measurements (by ∼2.5 mN m–1; q* = 2.4973, α = 0.05). Collectively, these findings suggest that the liquid surface tension produced by these surfactants is affected by the environment in which they are measured (e.g. by differences in pH, solute concentrations and the presence of other compounds that may differentially interact with each surfactant), but despite this, variations in surfactant concentrations and environment effects do not explain the γMin limit of 24–25 mN−1 to surface tension-reducing activity. We have proposed that the γMin limit is likely to reflect the extent of self-damage surfactant-producing cells can tolerate (Fechtner et al.2017), as surfactants are known to have a toxic effect and cause the loss of lipopolysaccharide in some bacteria, and more generally damage cell membranes and cause cell lysis in a range of prokaryote and eukaryote cells (Raaijmakers, De Bruijn and De Kock 2006; Franzetti et al.2010; Raaijmakers et al.2010; Inès and Dhouha 2015). Such self-damage is similar in nature to the effect of antibiotics, but whereas protective features such as altered targets and efflux pumps have been identified in many antibiotic-producing or -resistant organisms (Cundliffe 1989), nothing is known about how bacterial cells might protect themselves from damage caused by the surfactants they produce. In contrast to bacterial surfactants, synthetic hydrocarbon surfactants can reduce liquid surface tensions to ∼24 mN m–1, while fluorocarbon and silicon surfactants can reduce surface tensions to as low as 13.7 mN m–1 (Czajka, Hazell and Eastoe 2015). This may represent the physical–chemical limit of liquid surface tension reduction by surfactants, which is a direct consequence of the hydrophobic tail CH3/CH2 ratio per hydrophilic head group of these amphiphilic compounds, with synthetic ‘hedgehog’ surfactants more densely packed with CH3/CH2 groups than linear-chain surfactants (Czajka, Hazell and Eastoe 2015) such as the bacterial CLPs and rhamnolipids with one to two tail chains per head group (Abdel-Mawgoud, Lépine and Déziel 2010; Raaijmakers et al.2010). Significant behavioural variation exists within these high-strength surfactants We have also looked at surfactant behaviours as a proxy for the structural differences found between these compounds, as this is of general interest in determining the diversity of surfactant production and ecological roles these compounds may play within soil communities, as well as potential applications in biotechnology. We used quantitative emulsion, foam stability and oil displacement assays to generate a behavioural data set that we then investigated by HCA to visualise similarities in surfactant behaviours (these assays can all be modified by changing pH, NaCl concentration, temperature, etc. to reveal further behavioural differences; Zhang and Miller 1992; Morikawa, Hirata and Imanaka 2000; Prieto et al.2008; Rocha e Silva et al.2014; Balan, Kumar and Jayalakshmi 2016; Liu et al.2016). Preliminary testing of individual assays by one-way ANOVA found significant differences between surfactant-producing strains as well as between these and the control strains (data not shown). Further HCA of various combinations of assays showed that assay type, oil and aqueous layer conditions resulted in subtly different clustering of surfactants, and in modelling the oil displacement data alone, strain, oil-type and aqueous layer conditions were all found to be significant effects (GLM model II, P ≤ 0.0232; see Table S1, Supporting Information, for further details), and all four oils could be differentiated (Q = 2.57408, α = 0.05), as well as DI water from the Tris aqueous layer conditions (Q = 2.34774, α = 0.05). When the full data set was analysed by HCA, surfactant behaviours were clustered into six major groups of two to seven strains, which suggests that the 25 pseudomonads examined here are likely to be producing six or more structurally distinct surfactants (Fig. 3; see Fig. S2, Supporting Information, for HCA constellation plots based on various combinations of assay data and Table S4, Supporting Information, for the mean data set used in this analysis). However, we note that some HCA terminal (short) branches may not effectively discriminate between some surfactants, as in a preliminary pH and NaCl profiling of semi-purified surfactants produced by DBG-1–4 we could only differentiate between DBG-2 and DBG-4 (GLM model III; see Table S1, Supporting Information, for further details; Q = 2.6077, α = 0.05). Furthermore, in this HCA constellation plot, DBG-20 and DBG-24 clustered with the control strains. While DBG-20 and DBG-24 both express surfactants as assessed by surface tension measurements of culture supernatants, they clearly performed poorly in other assays used to determine surfactant activity and behaviour. We speculate that the surfactants they produce have particularly short hydrophobic tails that limit their ability to interact with hydrocarbons and result in poor oil displacement behaviours. Figure 3. View largeDownload slide Considerable behavioural diversity exists amongst the high-strength surfactants. HCA was used to determine similarities between surfactant behaviours produced by DBG strains using emulsion, foam stability and oil displacement assays. Shown here is an HCA constellation plot that clusters surfactants with similar behaviours in terminal (short) branches and links surfactants with greater differences with longer branches. The plot is arbitrarily rooted mid-way along the longest branch (circled) and the six major groups (grey arcs) determined automatically with the limit being set by the requirement to group all of the non–surfactant-producing control strains together. Strains producing surfactants with the highest strength, DBG-1–5, are indicated by black circles, and the remainder of the surfactant-producing strains, DBG-6–25, are indicated by the grey circles. Non–surfactant-producing DBG-c1–c5 control strains are indicated by white circles. Those strains that have been identified as likely Pseudomonas spp. are shown underlined. Figure 3. View largeDownload slide Considerable behavioural diversity exists amongst the high-strength surfactants. HCA was used to determine similarities between surfactant behaviours produced by DBG strains using emulsion, foam stability and oil displacement assays. Shown here is an HCA constellation plot that clusters surfactants with similar behaviours in terminal (short) branches and links surfactants with greater differences with longer branches. The plot is arbitrarily rooted mid-way along the longest branch (circled) and the six major groups (grey arcs) determined automatically with the limit being set by the requirement to group all of the non–surfactant-producing control strains together. Strains producing surfactants with the highest strength, DBG-1–5, are indicated by black circles, and the remainder of the surfactant-producing strains, DBG-6–25, are indicated by the grey circles. Non–surfactant-producing DBG-c1–c5 control strains are indicated by white circles. Those strains that have been identified as likely Pseudomonas spp. are shown underlined. We have also examined pairwise correlations between the surface tension, oil displacement, emulsion and foam stability data, which supports earlier but limited comparisons of surface tension measurements and oil displacement behaviours for smaller collections of surfactant-producing strains (Yussef et al.2004; Afshar et al.2008). Of the 120 pairwise correlations undertaken here, 52 were significant (P < 0.05) with 39 occurring within oil displacement assays and suggesting that surfactants were responding similarly to the different oils and aqueous layer conditions, and the remaining significant correlations occurring between assays which tested more diverse behaviours (see Table S5, Supporting Information, for pairwise correlations). Further inspection of these correlations could be used to identify those with unexpected behaviours, which might reflect significant structural variations. Finally, a comparison of the HCA grouping of strain phenotypes and surfactant behaviours indicates that some minor clusters are conserved in both comparisons, and this suggests that closely related strain pairs may have conserved surfactant synthesis genes and produce the same compounds (see Table S6, Supporting Information, for HCA grouping of strains). Identification of putative surfactant genes in the DBG-1 draft genome As part of a longer term project, we intend to determine the genome sequences of key pseudomonads and identify potential surfactant biosynthesis genes and predict the chemical nature of the compounds they produce. We now have a draft genome sequence for DBG-1, and our manual inspection of the CDS annotations identified seven NRPS genes in three clusters, which may be involved in the synthesis of a CLP-like surfactant (Table 1). We were initially distracted by these annotations as they suggested that DBG-1 may produce gramicidin or tyrocidine-like antibiotics first described for the gram-negative bacterium Bacillus brevis (Marahier, Nakano and Zuber 1993), though neither of these two antibiotics are reported to have liquid surface tension-reducing activities or otherwise considered to be surfactants. Table 1. Potential NRPS genes involved in surfactant production identified in the DBG-1 draft genome.   CDS  Name  Size (aa)  MW  NRPS homologue  First 50 aa for identification  Similar biosynthetic gene clusters (% similar)  Predicted product  Cluster I  02131  lgrB_1  4106  456 017.62  Gramicidin synthase subunit B  VQELIESVGQLSAKQRKALAVLLKQKGVNLFDIAPVFKRTAEEPLLLSYA  Pyoverdine (20%)  (thr-X-X) +    02132  lgrB_2  2608  284 684.97  Gramicidin synthase subunit B  MSGTMAERIAKRFVGLPLEQRRLFLAKLREDGKDFSLLPLPVSRHDIAPI    (asp-X) +    02133  lgrB_3  1133  124 304.305  Gramicidin synthase subunit B  MNAADAQKLARRFIELPQDKRRLFLAGMAREGIDFAQLPMTACDGIAERD    (X)  Cluster II  03215  tycC_1  5947  650 034.6  Tyrocidine synthase 3  VNVLELLATLKTKDIQLAVTDEQLRVNGNKQALSDPALLAALREHKPALI  Orfamide, putisolvin,  (d-val-gln-d-val-ile-glu) +    03216  grsB  4338  474 976.94  Gramicidin S synthase 2  MHFSELMAAISTRAIRLQQEDEDLVILGSDDALDDALWDSLAAHKAQLLE  Syringomycin, tolaasin  (thr-leu-d-val-ser) +    03217  tycC_2  2136  235 087.06  Tyrocidine synthase 3  MQPTSANVADDVSVPVETFALTAAQRDIWLDQLSRGDSPLYNIGGYVELS  (47–52%)  (d-leu-asp)  Cluster III  04978  lgrB_4  4332  480 759.78  Gramicidin synthase subunit B  MTDAFELPSTLVQALQRRAALTPDRLALRFLAENEEQAVVLSYRELDERA  Pyoverdine (17%)  (X-glu-arg-ala)    CDS  Name  Size (aa)  MW  NRPS homologue  First 50 aa for identification  Similar biosynthetic gene clusters (% similar)  Predicted product  Cluster I  02131  lgrB_1  4106  456 017.62  Gramicidin synthase subunit B  VQELIESVGQLSAKQRKALAVLLKQKGVNLFDIAPVFKRTAEEPLLLSYA  Pyoverdine (20%)  (thr-X-X) +    02132  lgrB_2  2608  284 684.97  Gramicidin synthase subunit B  MSGTMAERIAKRFVGLPLEQRRLFLAKLREDGKDFSLLPLPVSRHDIAPI    (asp-X) +    02133  lgrB_3  1133  124 304.305  Gramicidin synthase subunit B  MNAADAQKLARRFIELPQDKRRLFLAGMAREGIDFAQLPMTACDGIAERD    (X)  Cluster II  03215  tycC_1  5947  650 034.6  Tyrocidine synthase 3  VNVLELLATLKTKDIQLAVTDEQLRVNGNKQALSDPALLAALREHKPALI  Orfamide, putisolvin,  (d-val-gln-d-val-ile-glu) +    03216  grsB  4338  474 976.94  Gramicidin S synthase 2  MHFSELMAAISTRAIRLQQEDEDLVILGSDDALDDALWDSLAAHKAQLLE  Syringomycin, tolaasin  (thr-leu-d-val-ser) +    03217  tycC_2  2136  235 087.06  Tyrocidine synthase 3  MQPTSANVADDVSVPVETFALTAAQRDIWLDQLSRGDSPLYNIGGYVELS  (47–52%)  (d-leu-asp)  Cluster III  04978  lgrB_4  4332  480 759.78  Gramicidin synthase subunit B  MTDAFELPSTLVQALQRRAALTPDRLALRFLAENEEQAVVLSYRELDERA  Pyoverdine (17%)  (X-glu-arg-ala)  CDS numbers, names, sizes (amino acids) and molecular weight (MW), and NRPS homologues are those provided in the first draft genome release, and no cluster crosses contig (node) boundaries. NRPS homology confirmed by independent BLAST analyses. These clusters were also identified by antiSMASH (additional associated genes involved in biosynthesis, transport or regulation are not listed here). Low levels of similarity to known biosynthetic clusters may suggest that the DBG-1 clusters produce different compounds. AntiSMASH predicted products are based on assumed NRPS colinearity without tailoring reactions taken into account (X, unidentified amino or carboxylic acid). View Large We undertook a more sophisticated search using antiSMASH (Weber et al.2015), which identified the same CDSs and predicted that the second cluster CDS might produce a CLP similar to orfamide, putisolvin, syringomycin and tolaasin, all of which are known to be produced by pseudomonads (Raaijmakers et al.2010) (Fig. 4). AntiSMASH also identified the modular structure of these NRPSs, which include adenylation, thiolation and condensation domains responsible for the incorporation of each amino acid in the peptide chain (Strieker, Tanovic and Marahiel 2010). It is noteworthy that NRPS genes contributing to the same CLP are sometimes distributed in clusters across pseudomonad genomes (e.g. P. fluorescens SBW25 and SS101; De Bruijn et al.2007, 2008), and this may also occur in DBG-1. However, it is unclear whether DBG-1 produces a single surfactant, multiple structural analogues of one type or several different surfactant types, as metabolic analyses of surfactant-producing bacteria have revealed considerable complexity within single strains (Raaijmakers et al.2010), and some NRPS are functionally active as monomers (Sieber et al.2002). Further genetic analyses and biochemical characterisation will be required to properly identify the surface-active compounds produced by DBG-1. Figure 4. View largeDownload slide Three NRPS genes in DBG-1 may be involved in surfactant production. Inspection of the Pseudomonas spp. DBG-1 draft genome has identified three clusters of NRPS genes, which may be involved in surfactant production. Shown here is Cluster II containing three NRPS genes (CDS 03215–03217; black) plus genes predicted to have additional biosynthetic and transport-related functions (dark grey) or roles in regulation (light grey), and unrelated genes (white) (A).Within the three NRPS genes, modules consisting of an adenylation domain (white square), thiolation domain (black oval) and condensation domain (grey circle) involved in the elongation of the peptide chain can be identified, and in CDS 03215 two-terminal thioesterase domains are also present (light grey ovals) (B). Predicted gene functions and domain structures shown here are from antiSMASH analysis. Figure 4. View largeDownload slide Three NRPS genes in DBG-1 may be involved in surfactant production. Inspection of the Pseudomonas spp. DBG-1 draft genome has identified three clusters of NRPS genes, which may be involved in surfactant production. Shown here is Cluster II containing three NRPS genes (CDS 03215–03217; black) plus genes predicted to have additional biosynthetic and transport-related functions (dark grey) or roles in regulation (light grey), and unrelated genes (white) (A).Within the three NRPS genes, modules consisting of an adenylation domain (white square), thiolation domain (black oval) and condensation domain (grey circle) involved in the elongation of the peptide chain can be identified, and in CDS 03215 two-terminal thioesterase domains are also present (light grey ovals) (B). Predicted gene functions and domain structures shown here are from antiSMASH analysis. Concluding comment This analysis of soil-isolated Pseudomonas spp. strains producing high-strength surfactants has confirmed earlier predictions of the limit to the reduction of liquid surface tension that bacterial surfactants can achieve in aqueous solutions, and has shown that there is significant behavioural diversity amongst these surface-active compounds. We have begun to investigate the genetic basis of surfactant production in these strains by determining the draft genome sequence for DBG-1 and identifying potential CLP-like surfactant synthesis genes, and have recently submitted a further 10 strains for sequencing to allow further comparison within this group of pseudomonads. SUPPLEMENTARY DATA Supplementary data are available at FEMSLE online. FUNDING We acknowledge the funding for KK received from Umaru Musa Yar’adua University and Tertiary Education Trust Fund (TETFUND) Nigeria. Andrew Spiers is also member of the Scottish Alliance for Geoscience, Environment and Society (SAGES). Conflict of interest. None declared. REFERENCES Abdel-Mawgoud AM, Lépine F, Déziel E. Rhamnolipids: diversity of structures, microbial origins and roles. Appl Microbiol Biot  2010; 86: 1323– 36. Google Scholar CrossRef Search ADS   Afshar S, Lotfabad TB, Roostaazad R et al.   Comparative approach for detection of biosurfactant-producing bacteria isolated from Ahvaz petroleum excavation areas in south of Iran. Ann Microbiol  2008; 58: 555– 60. Google Scholar CrossRef Search ADS   Balan SS, Kumar CG, Jayalakshmi S. Pontifactin, a new lipopeptide biosurfactant produced by a marine Pontibacter korlensis strain SBK–47: purification, characterization and its biological evaluation. Process Biochem  2016; 51: 2198– 207. Google Scholar CrossRef Search ADS   Coffmann CW, Garcia VV. Functional properties and amino acid content of a protein isolate from mung bean flour. J Food Technol  1977; 12: 473– 84. Google Scholar CrossRef Search ADS   Cooper DG, Goldenberg BG. Surface-active agents from two Bacillus species. Appl Environ Microb  1987; 53: 224– 9. Cundliffe E. How antibiotic-producing organisms avoid suicide. Annu Rev Microbiol  1989; 43: 207– 33. Google Scholar CrossRef Search ADS PubMed  Czajka A, Hazell G, Eastoe J. Surfactants at the design limit. Langmuir  2015; 31: 8205− 17. Google Scholar CrossRef Search ADS PubMed  De Bruijn I, de Kock MJ, de Waard P et al.   Massetolide A biosynthesis in Pseudomonas fluorescens. J Bacteriol  2008; 190: 2777– 89. Google Scholar CrossRef Search ADS PubMed  De Bruijn I, de Kock MJ, Yang M et al.   Genome-based discovery, structure prediction and functional analysis of cyclic lipopeptide antibiotics in Pseudomonas species. Mol Microbiol  2007; 63: 417– 28. Google Scholar CrossRef Search ADS PubMed  De Souza JT, de Boer M, de Waard P et al.   Biochemical, genetic, and zoosporicidal properties of cyclic lipopeptide surfactants produced by Pseudomonas fluorescens. Appl Environ Microb  2003; 69: 7161– 72. Google Scholar CrossRef Search ADS   Desai JD, Banat IM. Microbial production of surfactants and their commercial potential. Microbiol Mol Biol Rev  1997; 61: 47– 64. Google Scholar PubMed  Fechtner F, Cameron S, Deeni YY et al.   Limitations of biosurfactant strength produced by bacteria. In: Upton CR (ed). Biosurfactants: Occurrences, Applications and Research . New York, Nova Publishers, 2017, 125– 48. Fechtner J, Koza A, Sterpaio PD et al.   Surfactants expressed by soil pseudomonads alter local soil–water distribution, suggesting a hydrological role for these compounds. FEMS Microbiol Ecol  2011; 78: 50– 8. Google Scholar CrossRef Search ADS PubMed  Franzetti A, Gandolfi I, Bestetti Get al.   (Bio)surfactant and bioremediation, successes and failures. In: Płaza G (ed). Trends Bioremediation Phytoremediation . Thiruvananthapuram, India: Research Signpost, 2010, 145– 56. Gudiña EJ, Rangarajan V, Sen R et al.   Potential therapeutic applications of biosurfactants. Trends Pharmacol Sci  2013; 34: 667– 75. Google Scholar CrossRef Search ADS PubMed  Inès M, Dhouha G. Lipopeptide surfactants: production, recovery and pore forming capacity. Peptides  2015; 71: 100– 12. Google Scholar CrossRef Search ADS PubMed  Koza A, Hallett PD, Moon CD et al.   Characterization of a novel air–liquid interface biofilm of Pseudomonas fluorescens SBW25. Microbiology  2009; 155: 1397– 406. Google Scholar CrossRef Search ADS PubMed  Kuśmierska A, Spiers AJ. New insights into the effects of several environmental parameters on the relative fitness of a numerically dominant class of evolved niche specialist. Int J Evolutionary Biol  2016; 2016, Article ID 4846565, http://dx.doi.org/10.1155/2016/4846565. Liu B, Liu J, Ju M et al.   Purification and characterization of biosurfactant produced by Bacillus licheniformis Y-1 and its application in remediation or petroleum contaminated soil. Marine Poll Bull  2016; 107: 46– 51. Google Scholar CrossRef Search ADS   Marahier MA, Nakano MM, Zuber P. Regulation of peptide antibiotic production in Bacillus. Mol Microbiol  1993; 7: 631– 6. Google Scholar CrossRef Search ADS PubMed  Marchant R, Banat IM. Microbial biosurfactants: challenges and opportunities for future exploitation. Trends Biotechnol  2012; 30: 558– 65. Google Scholar CrossRef Search ADS PubMed  Mohammed IU, Deeni Y, Hapca SM et al.   Predicting the minimum liquid surface tension activity of pseudomonads expressing biosurfactants. Lett Appl Microbiol  2015; 60: 37– 43. Google Scholar CrossRef Search ADS PubMed  Morikawa M, Daido H, Takao T et al.   A new lipopeptide biosurfactant produced by Arthrobacter sp. strain MIS38. J Bacteriol  1993; 175: 6459– 66. Google Scholar CrossRef Search ADS PubMed  Morikawa M, Hirata Y, Imanaka T. A study on the structure-function relationship of lipopeptide biosurfactants. Biochim Biophys Acta  2000; 1488: 211– 8. Google Scholar CrossRef Search ADS PubMed  Persson A, Molin G. Capacity for biosurfactant production of environmental Pseudomonas and Vibrionanaceae growing on carbohydrates. Appl Microbiol Biot  1987; 26: 439– 42. Google Scholar CrossRef Search ADS   Prieto LM, Michelon M, Burkert JFM et al.   The production of rhamnolipid by a Pseudomonas aeruginosa strain isolated from a southern coastal zone in Brazil. Chemosphere  2008; 71: 1781– 5. Google Scholar CrossRef Search ADS PubMed  Raaijmakers JM, De Bruijn I, De Kock MJ. Cyclic lipopeptide production by plant-associated Pseudomonas spp.: diversity, activity, biosynthesis, and regulation. Mol Plant Microbe Interact  2006; 7: 699– 710. Google Scholar CrossRef Search ADS   Raaijmakers JM, De Bruijn I, Nybroe O et al.   Natural functions of lipopeptides from Bacillus and Pseudomonas: more than surfactants and antibiotics. FEMS Microbiol Rev  2010; 34: 1037– 62. Google Scholar CrossRef Search ADS PubMed  Robertson M, Hapca SM, Moshynets O et al.   Air-liquid interface biofilm formation by psychrotrophic pseudomonads recovered from spoilt meat. A Van Leeuw  2013; 103: 251– 9. Google Scholar CrossRef Search ADS   Rocha e Silva NMP, Rufino RD, Luna JM et al.   Screening of Pseudomonas species for biosurfactant production using low-cost substrates. Biocatal Agric Biotechnol  2014; 3: 132– 9. Ron EZ, Rosenberg E. Natural roles of biosurfactants. Environ Microbiol  2001; 3: 229– 36. Google Scholar CrossRef Search ADS PubMed  Sieber SA, Linne U, Hillson NJ et al.   Evidence for a monomeric structure of nonribosomal peptide synthetases. Chem Biol  2002; 9: 997– 1008. Google Scholar CrossRef Search ADS PubMed  Souza EC, Vessoni-Penna TC, de Souza Oliveira RP. Biosurfactant-enhanced hydrocarbon bioremediation: and overview. Int Biodeter Biodegr  2014; 89: 88– 94. Google Scholar CrossRef Search ADS   Strieker M, Tanovic A, Marahiel MA. Nonribosomal peptide synthetases: structures and dynamics. Curr Opin Struc Biol  2010; 20: 234– 40. Google Scholar CrossRef Search ADS   Vargaftik NB, Volkov BN, Voljak LD. International tables of the surface tension of water. J Phys Chem Ref Data  1983; 12: 817– 20. Google Scholar CrossRef Search ADS   Vázquez G, Alvarez E, Navaza JM. Surface tension of alcohol + water from 20 to 50°C. J Chem Eng Data  1995; 40: 611– 4. Google Scholar CrossRef Search ADS   Weber T, Blin K, Duddela S et al.   antiSMASH 3.0 — a comprehensive resource for the genome mining of biosynthetic gene clusters. Nucleic Acids Res  2015; 43: W237– 43. Google Scholar CrossRef Search ADS PubMed  Youssef NH, Duncan KE, Nagle DP et al.   Comparison of methods to detect biosurfactant production by diverse microorganisms. J Microbiol Meth  2004; 56: 339– 47. Google Scholar CrossRef Search ADS   Zhang Y, Miller RM. Enhanced octadecane dispersion and biodegradation by a Pseudomonas rhamnolipid surfactant (biosurfactant). Appl Environ Microb  1992; 58: 3276– 82. © FEMS 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png FEMS Microbiology Letters Oxford University Press

Uncovering behavioural diversity amongst high-strength Pseudomonas spp. surfactants at the limit of liquid surface tension reduction

Loading next page...
 
/lp/ou_press/uncovering-behavioural-diversity-amongst-high-strength-pseudomonas-spp-8MBzg3H1WS
Publisher
Blackwell
Copyright
© FEMS 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
0378-1097
eISSN
1574-6968
D.O.I.
10.1093/femsle/fny008
Publisher site
See Article on Publisher Site

Abstract

Abstract Bacterial biosurfactants have a wide range of biological functions and biotechnological applications. Previous analyses had suggested a limit to their reduction of aqueous liquid surface tensions (γMin), and here we confirm this in an analysis of 25 Pseudomonas spp. strains isolated from soil which produce high-strength surfactants that reduce surface tensions to 25.2 ± 0.1–26.5 ± 0.2 mN m−1 (the surface tension of sterile growth medium and pure water was 52.9 ± 0.4 mN m−1 and 72.1 ± 1.2 mN m−1, respectively). Comparisons of culture supernatants produced using different growth media and semi-purified samples indicate that the limit of 24.2–24.7 mN m−1 is not greatly influenced by culture conditions, pH or NaCl concentrations. We have used foam, emulsion and oil-displacement behavioural assays as a simple and cost-effective proxy for in-depth biochemical characterisation, and these suggest that there is significant structural diversity amongst these surfactants that may reflect different biological functions and offer new biotechnological opportunities. Finally, we obtained a draft genome for the strain producing the highest strength surfactant, and identified a cluster of non-ribosomal protein synthase genes that may produce a cyclic lipopeptide (CLP)-like surfactant. Further investigation of this group of related bacteria recovered from the same site will allow a better understanding of the significance of the great variety of surfactants produced by bacterial communities found in soil and elsewhere. Pseudomonas, surfactant, limit to liquid surface activity, cyclic lipopeptide, non-ribosomal protein synthase INTRODUCTION Biosurfactants produced by bacteria are surface-active agents having a wide range of biological activities including involvement in the solubilisation of hydrophobic substrates, co-ordinated growth and differentiation, cell motility, surface attachment and biofilm development, suppression of competitors and protection from predators, immune modulation and virulence, rotting of plant tissues, causing fungal hyphae swelling and the lysis of oomycete zoospores (Ron and Rosenberg 2001; Abdel-Mawgoud, Lépine and Déziel 2010; Raaijmakers et al.2010). These compounds also have many applications in cosmetic, food, medical, pharmaceutical, oil and bioremediation technology where new high-strength surfactants are constantly in demand as detergents, wetting and foaming agents, emulsifiers and dispersants (Franzetti et al.2010; Marchant and Banat 2012; Gudiña et al.2013; Souza, Vessoni-Penna and de Souza Oliveira 2014; Inès and Dhouha 2015). The activity of surfactants depends on their amphiphilic nature, and a number of different structural classes of surfactants are produced by bacteria (Desai and Banat 1997), including cyclic lipopeptides (CLPs) and rhamnolipids (Abdel-Mawgoud, Lépine and Déziel 2010; Raaijmakers et al.2010). However, the relationship between surfactant activity, biological function (or role) and structural diversity remains poorly understood; we need to separate activity resulting from the fundamental biophysical properties of surfactants from those biological activities that provide the surfactant producer with a selective advantage. For example, very few bacteria would have a selective advantage in lysing erythrocytes, yet this is a common assay for surfactant production (Youssef et al.2004; Afshar et al.2008). Similarly, it is not clear how much of the observed structural diversity amongst surfactants is relevant or redundant. Furthermore, in complex soil or plant-associated communities, where different bacteria are capable of producing a range of surfactants, are these treated as public goods benefiting the whole community or does this represent intracommunity conflict and competition? Our research has focussed on assessing bacterial surfactant strengths and behavioural diversity within the Pseudomonas genus using behavioural assays as a simple and cost-effective proxy for the in-depth biochemical characterisation required to determine structural diversity (Fechtner et al.2017). This genus includes plant and mushroom pathogens that use surfactants to rot tissues as well as many surfactant-producing soil and plant-associated strains found in complex communities where the suppression of the growth of competitors and protection from predators may be particularly important; in addition, surfactants are also required for swarming motility and biofilm maturation (Raaijmakers et al.2010). Recent investigations of high-strength surfactants produced by pseudomonads and other bacteria have suggested that there is a limit (γMin) to the extent surfactants can reduce aqueous liquid surface tension of 24.16–24.24 mN m–1 (Fechtner et al.2011; Mohammed et al.2015) and the biological basis for this is probably the need to minimise self-damage to the producing cells (Fechtner et al.2017). To put this into context, the surface tension of water at 20°C is 72.8 mN m–1 (Vargaftik, Volkov and Voljak 1983), while the sterile media used in these predictions have surface tensions of between 47.0 and 59.6 mN m–1 (Fechtner et al.2011; Mohammed et al.2015). In this work, we want to test the robustness of the prediction by investigating surfactant production amongst a collection of pseudomonads isolated from the same soil community, to determine whether culture and buffer conditions significantly alter liquid surface tension measurements and γMin, and to assess the structural diversity amongst the high-strength surfactants produced by these strains, which may represent a valuable resource for future biotechnological exploitation. MATERIALS AND METHODS Bacterial isolation and cultivation Pseudomonas spp. or Pseudomonas-like strains were isolated from samples taken from bulk soil underlying a section of managed grass lawn at the Dundee Botanic Garden (DBG; Dundee, UK) in February and April 2015. Bacteria were isolated using selective agar (PSA-CFC; Oxoid, UK) spread with soil suspension dilutions and incubated under aerobic conditions for 2–3 days at 20°C–22°C. Colony material resuspended in deionised water (DI) was used to test for surfactant production using the drop collapse assay on petri dish lids as per Persson and Molin (1987), and then confirmed by quantitative tensiometry of modified King's B (KB*; Kuśmierska and Spiers 2016) culture supernatants (see below). Twenty-five surfactant-producing strains (DBG strains 1–25) plus five randomly chosen drop collapse-negative strains (DBG strains c1–c5) were retained for further investigation and stored at –80°C in 15% (v/v) glycerol. Overnight KB* and minimal medium containing 20 mM glucose (M9-Glu; Fechtner et al.2011) cultures incubated with shaking at 28°C were used to prepare samples for testing as required. Strain characterisation and identification Phenotypes were determined using biochemical, growth and behaviour-based assays at 20°C–22°C as per Robertson et al. (2013) (see Supplementary Information for further details), and hierarchical cluster analysis (HCA) was used to group strains on the basis of similarity as per Robertson et al. (2013) and Mohammed et al. (2015). Key strains were further analysed by metabolic profiling using API 20e cards (BioMérieux, Basingstoke, UK) and partial 16S rRNA gene sequencing to determine genus-level identification (see Supplementary Information for further details). Surfactant behaviour and surface tension measurements Twenty-four hour KB* cultures were used to investigate surfactant behaviours using emulsion, foam stability and oil displacement assays at 20°C–22°C as per Coffmann and Garcia (1977), Cooper and Goldenberg (1987) and Morikawa et al. (1993) (see Supplementary Information for further details), and HCA was used to cluster surfactant behaviours on the basis of similarity. For the oil displacement assays (also known as oil spreading assays), mineral oil, vegetable oil, used lubricating oil and diesel were overlaid onto DI water (pH 6), 200 mM NaCl (pH 6) and 50 mM Tris (pH 8) solutions. Surfactants were semi-purified from 24-h KB* cultures by an acid precipitation method adapted from De Souza et al. (2003) and resuspended in DI water to test critical micelle concentrations (CMCs), pH and NaCl surface tension profiles (see Supplementary Information for further details). Quantitative tensiometry of semi-purified surfactant solutions and cell-free 24-h KB* or M9-Glu culture supernatants were performed using a Krüss K100 Mk2 Tensiometer (Krüss GmbH, Hamburg, Germany) at 20°C as per Koza et al. (2009), and mean surface tension measurements are rounded up to one decimal place. In these assays, the surface tension of pure water was 72.1 ± 1.2 mN m–1, and the surface tension of sterile KB* and M9Glu culture media was 52.9 ± 0.4 and 70.7 ± 0.7 mN m–1, respectively. Statistical analyses Experiments were performed with replicates, and means with standard errors (SE) are shown where appropriate. Data were assumed to be normally distributed and were examined using JMP v12 statistical software (SAS Institute Inc., Marlow, UK) with comparisons of means performed using Student's and matched pairs t-tests (t), one-way analysis of variance (ANOVA) (F) models with Tukey–Kramer honest significant difference (HSD) (q*) post hoc tests and correlations (r) examined by multivariate analysis. HCA using the Ward Method with equal weightings was used to investigate similarities between strain phenotypes and surfactant behaviours as per Robertson et al. (2013) and Mohammed et al. (2015). Analyses based on general linear models (GLMs) were used to investigate surface tension and oil displacement data with effects further examined using LSMeans Differences Tukey HSD (Q) tests (see Table S1, Supporting Information for model details, covariates and effects tests). The minimum liquid surface tension (γMin) was determined by individual distribution identification (IDI) as per Fechtner et al. (2011) using mean surface tension data and based on the lowest Anderson-Darling (AD) goodness of fit test value using MINITAB v1.5 statistical software (Minitab Ltd, Coventry, UK). DBG-1 draft genome and identification of possible surfactant synthesis genes The DBG-1 draft genome was determined using the microbial sequencing and strain repository service MicrobesNG (https://microbesng.uk; Birmingham, UK), and trimmed reads and fasta files are available on request. Sequencing was performed on Illumina MiSeq and HiSeq 2500 platforms using 2× 250 bp pair-end reads, and data were put through a standard analysis pipeline for assembly and quality analysis (see https://microbesng.uk for further details). A mean coverage of 42.6× was achieved with 656 944 reads, producing a draft genome of 6860 106 bp comprised of 122 contigs of which the largest was 657 704 bp and a guanine and cytosine (GC) content of 58.9%. A total of 6082 coding sequences (CDS) were predicted within contigs, with an average length of 976 bp and density of 0.89 per kb, and annotations were provided where possible by automated BLAST analyses. A total of 69 tRNA genes were identified, though no rRNA genes were found including the 16S rRNA genes required for species-level identification. Read mapping suggests that this genome is most closely related to the P. fluorescens species, which is consistent with our isolation and selection of the strain as a fluorescent pseudomonad. CDS annotations associated with non-ribosome protein synthases (NRPS) were inspected manually and confirmed by NCBI/NLM BLASTP against non-redundant GenBank CDS translations, PDB, SwissProt, PIR and PRF databases (https://blast.ncbi.nlm.nih.gov/Blast.cgi). The draft DBG-1 genome was also submitted to antiSMASH (https://antismash.secondarymetabolites.org; bacterial version; Weber et al.2015) to further characterise NRPS homologues and predict possible products. RESULTS AND DISCUSSION Isolation of Pseudomonas spp. expressing strong surfactants that significantly lower liquid surface tension We isolated a collection of Pseudomonas spp. or Pseudomonas-like bacteria from soil that produced high-strength surface-active agents or surfactants when incubated in KB* cultures for 24 h, and from a statistically homogeneous Tukey–Kramer HSD group (q* = 3.970; α = 0.05) chose the 25 strains producing the lowest surface tensions for further analysis (DBG-1–25; Fig. 1). Although these showed similar surface tension reducing activity ranging between 25.2 ± 0.1 and 26.5 ± 0.2 mN m–1, a comparison of strain phenotypes by HCA, which also included control strains not producing surfactants under the conditions used here (DBG-c1–5), indicated that most could be differentiated by one or more colony morphology, enzyme and siderophore expression, antibiotics and mercury sensitivity, salt and high temperature tolerance, and motility assays (Fig. 2; see Table S2, Supporting Information, for the ordinal data set), with little evidence of biological replication (i.e. the isolation of the same strain more than once). Further testing of key strains using metabolic profiling and partial 16S rRNA gene sequencing suggests that most are probably Pseudomonas spp. (see Table S3, Supporting Information, for putative identifications). This collection of phenotypically diverse pseudomonads producing high-strength surfactants provided us with an opportunity to test the robustness of earlier predictions of the minimum limit (γMin) to liquid surface tension reduction achieved by bacterial surfactants, and then to examine the degree of structural diversity within a group of high-strength surfactants. Figure 1. View largeDownload slide Identification of a group of a homogeneous group of pseudomonads producing high-strength surfactants. A statistical approach was taken to identify 25 DBG strains producing high-strength surfactants. Differences between mean liquid surface tension measurements of 24-h KB* culture supernatants were determined by a Tukey–Kramer HSD test, and DBG-1–29 were found to form a single homogeneous group (α = 0.05). From these, the first 25 strains (dark grey) were chosen for further analysis. Means ± SE are shown (n = 4), and means not linked by the same letter are significantly different (q* = 3.970, α = 0.05). The ST of sterile KB* was 52.9 ± 0.4 mN m–1 (not shown). Figure 1. View largeDownload slide Identification of a group of a homogeneous group of pseudomonads producing high-strength surfactants. A statistical approach was taken to identify 25 DBG strains producing high-strength surfactants. Differences between mean liquid surface tension measurements of 24-h KB* culture supernatants were determined by a Tukey–Kramer HSD test, and DBG-1–29 were found to form a single homogeneous group (α = 0.05). From these, the first 25 strains (dark grey) were chosen for further analysis. Means ± SE are shown (n = 4), and means not linked by the same letter are significantly different (q* = 3.970, α = 0.05). The ST of sterile KB* was 52.9 ± 0.4 mN m–1 (not shown). Figure 2. View largeDownload slide The DBG pseudomonads producing high-strength surfactants are phenotypically diverse. HCA was used to determine similarities between DBG strain phenotypes using biochemical, growth and behaviour-based assays. Shown here is an HCA constellation plot that clusters similar strains in terminal (short) branches and links strains with greater differences with longer branches. The plot is arbitrarily rooted mid-way along the longest branch (circled) and six major groups (grey arcs) determined automatically. Strains producing surfactants with the highest strength, DBG-1–5, are indicated by black circles, and the remainder of the surfactant-producing strains, DBG-6–25, are indicated by the grey circles. Non–surfactant-producing DBG-c1–c5 control strains are indicated by white circles. Those strains that have been identified as likely Pseudomonas spp. are shown underlined. Figure 2. View largeDownload slide The DBG pseudomonads producing high-strength surfactants are phenotypically diverse. HCA was used to determine similarities between DBG strain phenotypes using biochemical, growth and behaviour-based assays. Shown here is an HCA constellation plot that clusters similar strains in terminal (short) branches and links strains with greater differences with longer branches. The plot is arbitrarily rooted mid-way along the longest branch (circled) and six major groups (grey arcs) determined automatically. Strains producing surfactants with the highest strength, DBG-1–5, are indicated by black circles, and the remainder of the surfactant-producing strains, DBG-6–25, are indicated by the grey circles. Non–surfactant-producing DBG-c1–c5 control strains are indicated by white circles. Those strains that have been identified as likely Pseudomonas spp. are shown underlined. Testing the predicted limit to liquid surface tension reduction We used IDI to predict γMin following the method established by Fechtner et al. (2011). The best fit for the mean surface tension data was found using the three-parameter log-logistic distribution (AD = 0.721) and produced a value of 24.7 mN m–1 for γMin similar to the earlier predictions of around 24.2 mN m–1 determined for a number of different groups of bacteria (Fechtner et al.2011; Mohammed et al.2015). To put this range of predictions into context, it is lower than the standard errors in our measurement of the surface tension of DI water at 20°C, which we undertake as an internal control (72.1 ± 1.2 mN m–1) and less than the change in surface tension of water between 20°C and 25°C (72.8 and 72.0 mN m–1, respectively; Vargaftik, Volkov and Voljak 1983) (in comparison, the surface tension of an 80% (w/w) solution of ethanol at 20°C is 24.3 mN m–1; Vázquez, Alvarez and Navaza 1995). This suggests that these predictions are centring on a common value for γMin; however, differences in pH and solute concentrations in culture media and resuspension buffers might be expected to alter surface tension measurements and γMin in an assay-dependent manner. We decided to explore this further by comparing KB* culture supernatant surface tensions with measurements taken from M9-Glu minimal medium and semi-purified surfactants resuspended in DI water. We obtained a slightly higher γMin prediction from the M9-Glu cell-free culture supernatants of 25.0 mN m–1 using a three-parameter Gamma distribution (AD = 0.917). However, although there was no significant correlation between KB* and M9-Glu strain means (P = 0.27), there were significant differences between pairs (t = –3.7527, P = 0.0003) with seven strains showing more activity and six strains showing less activity in M9-Glu than might have been expected when compared to KB* (see Fig. S1, Supporting Information, for a comparison of KB* and M9-Glu ST means). This suggests that the surfactants produced by some strains were differentially sensitive to the media or final culture supernatant used to determine surface tension measurements. In order to investigate this further, we semi-purified surfactants produced by 14 strains representing the range of surfactant strengths produced by this collection (DBG-1–5, 7, 10, 14–16, 20, 21 and 25). We progressively diluted these samples with DI water to demonstrate that in each case surfactants were produced in KB* cultures above the CMC, and that minor differences in concentration could not explain the differences in surface tension seen between strains or between KB*, M9-Glu and resuspended surfactant surface tension measurements. We then modelled surface tension measurements using a GLM, and found that strain and assay environment were significant effects (GLM model I, P < 0.0001; see Table S1, Supporting Information, for further details), with the resuspended surfactant surface tensions significantly higher than both culture supernatant measurements (by ∼2.5 mN m–1; q* = 2.4973, α = 0.05). Collectively, these findings suggest that the liquid surface tension produced by these surfactants is affected by the environment in which they are measured (e.g. by differences in pH, solute concentrations and the presence of other compounds that may differentially interact with each surfactant), but despite this, variations in surfactant concentrations and environment effects do not explain the γMin limit of 24–25 mN−1 to surface tension-reducing activity. We have proposed that the γMin limit is likely to reflect the extent of self-damage surfactant-producing cells can tolerate (Fechtner et al.2017), as surfactants are known to have a toxic effect and cause the loss of lipopolysaccharide in some bacteria, and more generally damage cell membranes and cause cell lysis in a range of prokaryote and eukaryote cells (Raaijmakers, De Bruijn and De Kock 2006; Franzetti et al.2010; Raaijmakers et al.2010; Inès and Dhouha 2015). Such self-damage is similar in nature to the effect of antibiotics, but whereas protective features such as altered targets and efflux pumps have been identified in many antibiotic-producing or -resistant organisms (Cundliffe 1989), nothing is known about how bacterial cells might protect themselves from damage caused by the surfactants they produce. In contrast to bacterial surfactants, synthetic hydrocarbon surfactants can reduce liquid surface tensions to ∼24 mN m–1, while fluorocarbon and silicon surfactants can reduce surface tensions to as low as 13.7 mN m–1 (Czajka, Hazell and Eastoe 2015). This may represent the physical–chemical limit of liquid surface tension reduction by surfactants, which is a direct consequence of the hydrophobic tail CH3/CH2 ratio per hydrophilic head group of these amphiphilic compounds, with synthetic ‘hedgehog’ surfactants more densely packed with CH3/CH2 groups than linear-chain surfactants (Czajka, Hazell and Eastoe 2015) such as the bacterial CLPs and rhamnolipids with one to two tail chains per head group (Abdel-Mawgoud, Lépine and Déziel 2010; Raaijmakers et al.2010). Significant behavioural variation exists within these high-strength surfactants We have also looked at surfactant behaviours as a proxy for the structural differences found between these compounds, as this is of general interest in determining the diversity of surfactant production and ecological roles these compounds may play within soil communities, as well as potential applications in biotechnology. We used quantitative emulsion, foam stability and oil displacement assays to generate a behavioural data set that we then investigated by HCA to visualise similarities in surfactant behaviours (these assays can all be modified by changing pH, NaCl concentration, temperature, etc. to reveal further behavioural differences; Zhang and Miller 1992; Morikawa, Hirata and Imanaka 2000; Prieto et al.2008; Rocha e Silva et al.2014; Balan, Kumar and Jayalakshmi 2016; Liu et al.2016). Preliminary testing of individual assays by one-way ANOVA found significant differences between surfactant-producing strains as well as between these and the control strains (data not shown). Further HCA of various combinations of assays showed that assay type, oil and aqueous layer conditions resulted in subtly different clustering of surfactants, and in modelling the oil displacement data alone, strain, oil-type and aqueous layer conditions were all found to be significant effects (GLM model II, P ≤ 0.0232; see Table S1, Supporting Information, for further details), and all four oils could be differentiated (Q = 2.57408, α = 0.05), as well as DI water from the Tris aqueous layer conditions (Q = 2.34774, α = 0.05). When the full data set was analysed by HCA, surfactant behaviours were clustered into six major groups of two to seven strains, which suggests that the 25 pseudomonads examined here are likely to be producing six or more structurally distinct surfactants (Fig. 3; see Fig. S2, Supporting Information, for HCA constellation plots based on various combinations of assay data and Table S4, Supporting Information, for the mean data set used in this analysis). However, we note that some HCA terminal (short) branches may not effectively discriminate between some surfactants, as in a preliminary pH and NaCl profiling of semi-purified surfactants produced by DBG-1–4 we could only differentiate between DBG-2 and DBG-4 (GLM model III; see Table S1, Supporting Information, for further details; Q = 2.6077, α = 0.05). Furthermore, in this HCA constellation plot, DBG-20 and DBG-24 clustered with the control strains. While DBG-20 and DBG-24 both express surfactants as assessed by surface tension measurements of culture supernatants, they clearly performed poorly in other assays used to determine surfactant activity and behaviour. We speculate that the surfactants they produce have particularly short hydrophobic tails that limit their ability to interact with hydrocarbons and result in poor oil displacement behaviours. Figure 3. View largeDownload slide Considerable behavioural diversity exists amongst the high-strength surfactants. HCA was used to determine similarities between surfactant behaviours produced by DBG strains using emulsion, foam stability and oil displacement assays. Shown here is an HCA constellation plot that clusters surfactants with similar behaviours in terminal (short) branches and links surfactants with greater differences with longer branches. The plot is arbitrarily rooted mid-way along the longest branch (circled) and the six major groups (grey arcs) determined automatically with the limit being set by the requirement to group all of the non–surfactant-producing control strains together. Strains producing surfactants with the highest strength, DBG-1–5, are indicated by black circles, and the remainder of the surfactant-producing strains, DBG-6–25, are indicated by the grey circles. Non–surfactant-producing DBG-c1–c5 control strains are indicated by white circles. Those strains that have been identified as likely Pseudomonas spp. are shown underlined. Figure 3. View largeDownload slide Considerable behavioural diversity exists amongst the high-strength surfactants. HCA was used to determine similarities between surfactant behaviours produced by DBG strains using emulsion, foam stability and oil displacement assays. Shown here is an HCA constellation plot that clusters surfactants with similar behaviours in terminal (short) branches and links surfactants with greater differences with longer branches. The plot is arbitrarily rooted mid-way along the longest branch (circled) and the six major groups (grey arcs) determined automatically with the limit being set by the requirement to group all of the non–surfactant-producing control strains together. Strains producing surfactants with the highest strength, DBG-1–5, are indicated by black circles, and the remainder of the surfactant-producing strains, DBG-6–25, are indicated by the grey circles. Non–surfactant-producing DBG-c1–c5 control strains are indicated by white circles. Those strains that have been identified as likely Pseudomonas spp. are shown underlined. We have also examined pairwise correlations between the surface tension, oil displacement, emulsion and foam stability data, which supports earlier but limited comparisons of surface tension measurements and oil displacement behaviours for smaller collections of surfactant-producing strains (Yussef et al.2004; Afshar et al.2008). Of the 120 pairwise correlations undertaken here, 52 were significant (P < 0.05) with 39 occurring within oil displacement assays and suggesting that surfactants were responding similarly to the different oils and aqueous layer conditions, and the remaining significant correlations occurring between assays which tested more diverse behaviours (see Table S5, Supporting Information, for pairwise correlations). Further inspection of these correlations could be used to identify those with unexpected behaviours, which might reflect significant structural variations. Finally, a comparison of the HCA grouping of strain phenotypes and surfactant behaviours indicates that some minor clusters are conserved in both comparisons, and this suggests that closely related strain pairs may have conserved surfactant synthesis genes and produce the same compounds (see Table S6, Supporting Information, for HCA grouping of strains). Identification of putative surfactant genes in the DBG-1 draft genome As part of a longer term project, we intend to determine the genome sequences of key pseudomonads and identify potential surfactant biosynthesis genes and predict the chemical nature of the compounds they produce. We now have a draft genome sequence for DBG-1, and our manual inspection of the CDS annotations identified seven NRPS genes in three clusters, which may be involved in the synthesis of a CLP-like surfactant (Table 1). We were initially distracted by these annotations as they suggested that DBG-1 may produce gramicidin or tyrocidine-like antibiotics first described for the gram-negative bacterium Bacillus brevis (Marahier, Nakano and Zuber 1993), though neither of these two antibiotics are reported to have liquid surface tension-reducing activities or otherwise considered to be surfactants. Table 1. Potential NRPS genes involved in surfactant production identified in the DBG-1 draft genome.   CDS  Name  Size (aa)  MW  NRPS homologue  First 50 aa for identification  Similar biosynthetic gene clusters (% similar)  Predicted product  Cluster I  02131  lgrB_1  4106  456 017.62  Gramicidin synthase subunit B  VQELIESVGQLSAKQRKALAVLLKQKGVNLFDIAPVFKRTAEEPLLLSYA  Pyoverdine (20%)  (thr-X-X) +    02132  lgrB_2  2608  284 684.97  Gramicidin synthase subunit B  MSGTMAERIAKRFVGLPLEQRRLFLAKLREDGKDFSLLPLPVSRHDIAPI    (asp-X) +    02133  lgrB_3  1133  124 304.305  Gramicidin synthase subunit B  MNAADAQKLARRFIELPQDKRRLFLAGMAREGIDFAQLPMTACDGIAERD    (X)  Cluster II  03215  tycC_1  5947  650 034.6  Tyrocidine synthase 3  VNVLELLATLKTKDIQLAVTDEQLRVNGNKQALSDPALLAALREHKPALI  Orfamide, putisolvin,  (d-val-gln-d-val-ile-glu) +    03216  grsB  4338  474 976.94  Gramicidin S synthase 2  MHFSELMAAISTRAIRLQQEDEDLVILGSDDALDDALWDSLAAHKAQLLE  Syringomycin, tolaasin  (thr-leu-d-val-ser) +    03217  tycC_2  2136  235 087.06  Tyrocidine synthase 3  MQPTSANVADDVSVPVETFALTAAQRDIWLDQLSRGDSPLYNIGGYVELS  (47–52%)  (d-leu-asp)  Cluster III  04978  lgrB_4  4332  480 759.78  Gramicidin synthase subunit B  MTDAFELPSTLVQALQRRAALTPDRLALRFLAENEEQAVVLSYRELDERA  Pyoverdine (17%)  (X-glu-arg-ala)    CDS  Name  Size (aa)  MW  NRPS homologue  First 50 aa for identification  Similar biosynthetic gene clusters (% similar)  Predicted product  Cluster I  02131  lgrB_1  4106  456 017.62  Gramicidin synthase subunit B  VQELIESVGQLSAKQRKALAVLLKQKGVNLFDIAPVFKRTAEEPLLLSYA  Pyoverdine (20%)  (thr-X-X) +    02132  lgrB_2  2608  284 684.97  Gramicidin synthase subunit B  MSGTMAERIAKRFVGLPLEQRRLFLAKLREDGKDFSLLPLPVSRHDIAPI    (asp-X) +    02133  lgrB_3  1133  124 304.305  Gramicidin synthase subunit B  MNAADAQKLARRFIELPQDKRRLFLAGMAREGIDFAQLPMTACDGIAERD    (X)  Cluster II  03215  tycC_1  5947  650 034.6  Tyrocidine synthase 3  VNVLELLATLKTKDIQLAVTDEQLRVNGNKQALSDPALLAALREHKPALI  Orfamide, putisolvin,  (d-val-gln-d-val-ile-glu) +    03216  grsB  4338  474 976.94  Gramicidin S synthase 2  MHFSELMAAISTRAIRLQQEDEDLVILGSDDALDDALWDSLAAHKAQLLE  Syringomycin, tolaasin  (thr-leu-d-val-ser) +    03217  tycC_2  2136  235 087.06  Tyrocidine synthase 3  MQPTSANVADDVSVPVETFALTAAQRDIWLDQLSRGDSPLYNIGGYVELS  (47–52%)  (d-leu-asp)  Cluster III  04978  lgrB_4  4332  480 759.78  Gramicidin synthase subunit B  MTDAFELPSTLVQALQRRAALTPDRLALRFLAENEEQAVVLSYRELDERA  Pyoverdine (17%)  (X-glu-arg-ala)  CDS numbers, names, sizes (amino acids) and molecular weight (MW), and NRPS homologues are those provided in the first draft genome release, and no cluster crosses contig (node) boundaries. NRPS homology confirmed by independent BLAST analyses. These clusters were also identified by antiSMASH (additional associated genes involved in biosynthesis, transport or regulation are not listed here). Low levels of similarity to known biosynthetic clusters may suggest that the DBG-1 clusters produce different compounds. AntiSMASH predicted products are based on assumed NRPS colinearity without tailoring reactions taken into account (X, unidentified amino or carboxylic acid). View Large We undertook a more sophisticated search using antiSMASH (Weber et al.2015), which identified the same CDSs and predicted that the second cluster CDS might produce a CLP similar to orfamide, putisolvin, syringomycin and tolaasin, all of which are known to be produced by pseudomonads (Raaijmakers et al.2010) (Fig. 4). AntiSMASH also identified the modular structure of these NRPSs, which include adenylation, thiolation and condensation domains responsible for the incorporation of each amino acid in the peptide chain (Strieker, Tanovic and Marahiel 2010). It is noteworthy that NRPS genes contributing to the same CLP are sometimes distributed in clusters across pseudomonad genomes (e.g. P. fluorescens SBW25 and SS101; De Bruijn et al.2007, 2008), and this may also occur in DBG-1. However, it is unclear whether DBG-1 produces a single surfactant, multiple structural analogues of one type or several different surfactant types, as metabolic analyses of surfactant-producing bacteria have revealed considerable complexity within single strains (Raaijmakers et al.2010), and some NRPS are functionally active as monomers (Sieber et al.2002). Further genetic analyses and biochemical characterisation will be required to properly identify the surface-active compounds produced by DBG-1. Figure 4. View largeDownload slide Three NRPS genes in DBG-1 may be involved in surfactant production. Inspection of the Pseudomonas spp. DBG-1 draft genome has identified three clusters of NRPS genes, which may be involved in surfactant production. Shown here is Cluster II containing three NRPS genes (CDS 03215–03217; black) plus genes predicted to have additional biosynthetic and transport-related functions (dark grey) or roles in regulation (light grey), and unrelated genes (white) (A).Within the three NRPS genes, modules consisting of an adenylation domain (white square), thiolation domain (black oval) and condensation domain (grey circle) involved in the elongation of the peptide chain can be identified, and in CDS 03215 two-terminal thioesterase domains are also present (light grey ovals) (B). Predicted gene functions and domain structures shown here are from antiSMASH analysis. Figure 4. View largeDownload slide Three NRPS genes in DBG-1 may be involved in surfactant production. Inspection of the Pseudomonas spp. DBG-1 draft genome has identified three clusters of NRPS genes, which may be involved in surfactant production. Shown here is Cluster II containing three NRPS genes (CDS 03215–03217; black) plus genes predicted to have additional biosynthetic and transport-related functions (dark grey) or roles in regulation (light grey), and unrelated genes (white) (A).Within the three NRPS genes, modules consisting of an adenylation domain (white square), thiolation domain (black oval) and condensation domain (grey circle) involved in the elongation of the peptide chain can be identified, and in CDS 03215 two-terminal thioesterase domains are also present (light grey ovals) (B). Predicted gene functions and domain structures shown here are from antiSMASH analysis. Concluding comment This analysis of soil-isolated Pseudomonas spp. strains producing high-strength surfactants has confirmed earlier predictions of the limit to the reduction of liquid surface tension that bacterial surfactants can achieve in aqueous solutions, and has shown that there is significant behavioural diversity amongst these surface-active compounds. We have begun to investigate the genetic basis of surfactant production in these strains by determining the draft genome sequence for DBG-1 and identifying potential CLP-like surfactant synthesis genes, and have recently submitted a further 10 strains for sequencing to allow further comparison within this group of pseudomonads. SUPPLEMENTARY DATA Supplementary data are available at FEMSLE online. FUNDING We acknowledge the funding for KK received from Umaru Musa Yar’adua University and Tertiary Education Trust Fund (TETFUND) Nigeria. Andrew Spiers is also member of the Scottish Alliance for Geoscience, Environment and Society (SAGES). Conflict of interest. None declared. REFERENCES Abdel-Mawgoud AM, Lépine F, Déziel E. Rhamnolipids: diversity of structures, microbial origins and roles. Appl Microbiol Biot  2010; 86: 1323– 36. Google Scholar CrossRef Search ADS   Afshar S, Lotfabad TB, Roostaazad R et al.   Comparative approach for detection of biosurfactant-producing bacteria isolated from Ahvaz petroleum excavation areas in south of Iran. Ann Microbiol  2008; 58: 555– 60. Google Scholar CrossRef Search ADS   Balan SS, Kumar CG, Jayalakshmi S. Pontifactin, a new lipopeptide biosurfactant produced by a marine Pontibacter korlensis strain SBK–47: purification, characterization and its biological evaluation. Process Biochem  2016; 51: 2198– 207. Google Scholar CrossRef Search ADS   Coffmann CW, Garcia VV. Functional properties and amino acid content of a protein isolate from mung bean flour. J Food Technol  1977; 12: 473– 84. Google Scholar CrossRef Search ADS   Cooper DG, Goldenberg BG. Surface-active agents from two Bacillus species. Appl Environ Microb  1987; 53: 224– 9. Cundliffe E. How antibiotic-producing organisms avoid suicide. Annu Rev Microbiol  1989; 43: 207– 33. Google Scholar CrossRef Search ADS PubMed  Czajka A, Hazell G, Eastoe J. Surfactants at the design limit. Langmuir  2015; 31: 8205− 17. Google Scholar CrossRef Search ADS PubMed  De Bruijn I, de Kock MJ, de Waard P et al.   Massetolide A biosynthesis in Pseudomonas fluorescens. J Bacteriol  2008; 190: 2777– 89. Google Scholar CrossRef Search ADS PubMed  De Bruijn I, de Kock MJ, Yang M et al.   Genome-based discovery, structure prediction and functional analysis of cyclic lipopeptide antibiotics in Pseudomonas species. Mol Microbiol  2007; 63: 417– 28. Google Scholar CrossRef Search ADS PubMed  De Souza JT, de Boer M, de Waard P et al.   Biochemical, genetic, and zoosporicidal properties of cyclic lipopeptide surfactants produced by Pseudomonas fluorescens. Appl Environ Microb  2003; 69: 7161– 72. Google Scholar CrossRef Search ADS   Desai JD, Banat IM. Microbial production of surfactants and their commercial potential. Microbiol Mol Biol Rev  1997; 61: 47– 64. Google Scholar PubMed  Fechtner F, Cameron S, Deeni YY et al.   Limitations of biosurfactant strength produced by bacteria. In: Upton CR (ed). Biosurfactants: Occurrences, Applications and Research . New York, Nova Publishers, 2017, 125– 48. Fechtner J, Koza A, Sterpaio PD et al.   Surfactants expressed by soil pseudomonads alter local soil–water distribution, suggesting a hydrological role for these compounds. FEMS Microbiol Ecol  2011; 78: 50– 8. Google Scholar CrossRef Search ADS PubMed  Franzetti A, Gandolfi I, Bestetti Get al.   (Bio)surfactant and bioremediation, successes and failures. In: Płaza G (ed). Trends Bioremediation Phytoremediation . Thiruvananthapuram, India: Research Signpost, 2010, 145– 56. Gudiña EJ, Rangarajan V, Sen R et al.   Potential therapeutic applications of biosurfactants. Trends Pharmacol Sci  2013; 34: 667– 75. Google Scholar CrossRef Search ADS PubMed  Inès M, Dhouha G. Lipopeptide surfactants: production, recovery and pore forming capacity. Peptides  2015; 71: 100– 12. Google Scholar CrossRef Search ADS PubMed  Koza A, Hallett PD, Moon CD et al.   Characterization of a novel air–liquid interface biofilm of Pseudomonas fluorescens SBW25. Microbiology  2009; 155: 1397– 406. Google Scholar CrossRef Search ADS PubMed  Kuśmierska A, Spiers AJ. New insights into the effects of several environmental parameters on the relative fitness of a numerically dominant class of evolved niche specialist. Int J Evolutionary Biol  2016; 2016, Article ID 4846565, http://dx.doi.org/10.1155/2016/4846565. Liu B, Liu J, Ju M et al.   Purification and characterization of biosurfactant produced by Bacillus licheniformis Y-1 and its application in remediation or petroleum contaminated soil. Marine Poll Bull  2016; 107: 46– 51. Google Scholar CrossRef Search ADS   Marahier MA, Nakano MM, Zuber P. Regulation of peptide antibiotic production in Bacillus. Mol Microbiol  1993; 7: 631– 6. Google Scholar CrossRef Search ADS PubMed  Marchant R, Banat IM. Microbial biosurfactants: challenges and opportunities for future exploitation. Trends Biotechnol  2012; 30: 558– 65. Google Scholar CrossRef Search ADS PubMed  Mohammed IU, Deeni Y, Hapca SM et al.   Predicting the minimum liquid surface tension activity of pseudomonads expressing biosurfactants. Lett Appl Microbiol  2015; 60: 37– 43. Google Scholar CrossRef Search ADS PubMed  Morikawa M, Daido H, Takao T et al.   A new lipopeptide biosurfactant produced by Arthrobacter sp. strain MIS38. J Bacteriol  1993; 175: 6459– 66. Google Scholar CrossRef Search ADS PubMed  Morikawa M, Hirata Y, Imanaka T. A study on the structure-function relationship of lipopeptide biosurfactants. Biochim Biophys Acta  2000; 1488: 211– 8. Google Scholar CrossRef Search ADS PubMed  Persson A, Molin G. Capacity for biosurfactant production of environmental Pseudomonas and Vibrionanaceae growing on carbohydrates. Appl Microbiol Biot  1987; 26: 439– 42. Google Scholar CrossRef Search ADS   Prieto LM, Michelon M, Burkert JFM et al.   The production of rhamnolipid by a Pseudomonas aeruginosa strain isolated from a southern coastal zone in Brazil. Chemosphere  2008; 71: 1781– 5. Google Scholar CrossRef Search ADS PubMed  Raaijmakers JM, De Bruijn I, De Kock MJ. Cyclic lipopeptide production by plant-associated Pseudomonas spp.: diversity, activity, biosynthesis, and regulation. Mol Plant Microbe Interact  2006; 7: 699– 710. Google Scholar CrossRef Search ADS   Raaijmakers JM, De Bruijn I, Nybroe O et al.   Natural functions of lipopeptides from Bacillus and Pseudomonas: more than surfactants and antibiotics. FEMS Microbiol Rev  2010; 34: 1037– 62. Google Scholar CrossRef Search ADS PubMed  Robertson M, Hapca SM, Moshynets O et al.   Air-liquid interface biofilm formation by psychrotrophic pseudomonads recovered from spoilt meat. A Van Leeuw  2013; 103: 251– 9. Google Scholar CrossRef Search ADS   Rocha e Silva NMP, Rufino RD, Luna JM et al.   Screening of Pseudomonas species for biosurfactant production using low-cost substrates. Biocatal Agric Biotechnol  2014; 3: 132– 9. Ron EZ, Rosenberg E. Natural roles of biosurfactants. Environ Microbiol  2001; 3: 229– 36. Google Scholar CrossRef Search ADS PubMed  Sieber SA, Linne U, Hillson NJ et al.   Evidence for a monomeric structure of nonribosomal peptide synthetases. Chem Biol  2002; 9: 997– 1008. Google Scholar CrossRef Search ADS PubMed  Souza EC, Vessoni-Penna TC, de Souza Oliveira RP. Biosurfactant-enhanced hydrocarbon bioremediation: and overview. Int Biodeter Biodegr  2014; 89: 88– 94. Google Scholar CrossRef Search ADS   Strieker M, Tanovic A, Marahiel MA. Nonribosomal peptide synthetases: structures and dynamics. Curr Opin Struc Biol  2010; 20: 234– 40. Google Scholar CrossRef Search ADS   Vargaftik NB, Volkov BN, Voljak LD. International tables of the surface tension of water. J Phys Chem Ref Data  1983; 12: 817– 20. Google Scholar CrossRef Search ADS   Vázquez G, Alvarez E, Navaza JM. Surface tension of alcohol + water from 20 to 50°C. J Chem Eng Data  1995; 40: 611– 4. Google Scholar CrossRef Search ADS   Weber T, Blin K, Duddela S et al.   antiSMASH 3.0 — a comprehensive resource for the genome mining of biosynthetic gene clusters. Nucleic Acids Res  2015; 43: W237– 43. Google Scholar CrossRef Search ADS PubMed  Youssef NH, Duncan KE, Nagle DP et al.   Comparison of methods to detect biosurfactant production by diverse microorganisms. J Microbiol Meth  2004; 56: 339– 47. Google Scholar CrossRef Search ADS   Zhang Y, Miller RM. Enhanced octadecane dispersion and biodegradation by a Pseudomonas rhamnolipid surfactant (biosurfactant). Appl Environ Microb  1992; 58: 3276– 82. © FEMS 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Journal

FEMS Microbiology LettersOxford University Press

Published: Feb 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve Freelancer

DeepDyve Pro

Price
FREE
$49/month

$360/year
Save searches from Google Scholar, PubMed
Create lists to organize your research
Export lists, citations
Read DeepDyve articles
Abstract access only
Unlimited access to over
18 million full-text articles
Print
20 pages/month
PDF Discount
20% off