The transcriptional programme of Salmonella enterica serovar Typhimurium reveals a key role for tryptophan metabolism in biofilms

Abstract

Background: Biofilm formation enhances the capacity of pathogenic Salmonella bacteria to survive stresses that are commonly encountered within food processing and during host infection. The persistence of Salmonella within the food chain has become a major health concern, as biofilms can serve as a reservoir for the contamination of food products. While the molecular mechanisms required for the survival of bacteria on surfaces are not fully understood, transcriptional studies of other bacteria have demonstrated that biofilm growth triggers the expression of specific sets of genes, compared with planktonic cells. Until now, most gene expression studies of Salmonella have focused on the effect of infection-relevant stressors on virulence or the comparison of mutant and wild-type bacteria. However little is known about the physiological responses taking place inside a Salmonella biofilm.

Results: We have determined the transcriptomic and proteomic profiles of biofilms of Salmonella enterica serovar Typhimurium. We discovered that 124 detectable proteins were differentially expressed in the biofilm compared with planktonic cells, and that 10% of the S. Typhimurium genome (433 genes) showed a 2-fold or more change in the biofilm compared with planktonic cells. The genes that were significantly up-regulated implicated certain cellular processes in biofilm development including amino acid metabolism, cell motility, global regulation and tolerance to stress. We found that the most highly down-regulated genes in the biofilm were located on Salmonella Pathogenicity Island 2 (SPI2), and that a functional SPI2 secretion system regulator (ssrA) was required for S. Typhimurium biofilm formation. We identified STM0341 as a gene of unknown function that was needed for biofilm growth. Genes involved in tryptophan (trp) biosynthesis and transport were up-regulated in the biofilm. Deletion of trpE led to decreased bacterial attachment and this biofilm defect was restored by exogenous tryptophan or indole.

Conclusions: Biofilm growth of S. Typhimurium causes distinct changes in gene and protein expression. Our results show that aromatic amino acids make an important contribution to biofilm formation and reveal a link between SPI2 expression and surface-associated growth in S. Typhimurium.

Figure 1

Whole genome expression profiling of S. Typhimurium SL1344 flowing biofilms compared to planktonic cells when grown in CFA at 25°C for 72 h. Expression changes of genes belonging to functional groups and pathogenicity islands (numbers in parenthesis refer to the genes assigned to each functional group from the genome of S. Typhimurium LT2). The bars show the percentage of genes belonging to each group that were altered for expression > 2-fold between planktonic and flowing biofilm cells. The blue bars indicate the proportion of genes that are down-regulated and the red bars represent the proportion of up-regulated genes for each group during biofilm growth (P < 0.01).

Figure 2

Biofilm-regulated S. Typhimurium gene expression. (A) A subset of genes up-regulated in the biofilm that encode cell surface structures, motility, global regulation, and oxygen diffusion. Representative examples of the functional categories (Figure 1) are shown. (B) Type III secretion genes. Transcriptomic data from the flowing biofilm system (72 h) was normalised to the planktonic samples and values are shown as fold change on a logarithmic scale (e.g. value of 10 on the Y-axis corresponds to 10-fold up-regulated). All genes were significant at P < 0.01 (* denotes that the csgA gene was significant at P < 0.05).

Figure 3

Biofilm-regulated protein expression. SYPRO^® Ruby stained 2-D gels of total protein extracts from flowing biofilm (A) and planktonic (B) cells of S. Typhimurium grown in CFA medium at 25°C for 72 h. Spots circled in red were excised from the gels and identified by mass spectrometry and peptide mass fingerprinting. (C) Magnification of 2-D gels comparing expression of FlgK (I), DppA (II), and YggE (III), which were all more highly expressed in the biofilm cells than in planktonic cells. (D) Western immunoblot of total protein extracts of mature S. Typhimurium biofilm (B) and planktonic (P) cells both grown in CFA medium for 72 h. Protein extracts were separated on a denaturing SDS-PAGE gel and probed with anti-FliC monoclonal antibody. Densitometric analysis showed that FliC was 3.4 fold induced in the biofilm compared with planktonic cells.

Figure 4

(A) Static biofilm formation of eight targeted gene deletion mutants, compared to attachment of WT SL1344. Following incubation at 25°C for 24 h in CFA medium, the level of biofilm formation is expressed as a percentage of WT SL1344 which had an A_{590 nm} of 0.56 ± 0.04 in this experiment. The ΔtrpE (JH3185) (* P = 0.0001) and ΔSTM0341 (JH3187) (* P = 0.001) mutants showed significantly less attachment to polystyrene than WT SL1344. The mean absorbance values from four wells are shown as a percentage of WT SL1344 and the error bars represent the SD between four technical replicates (n = 3). (B) The attachment of the ΔtrpE (black diamonds) mutant (JH3185) to the bottom surface of a glass flow cell compared with S. Typhimurium SL1344 (back circles). Bacteria were cultured at 25°C in CFA medium in a glass flow cell. Error bars represent the standard deviation between 6 images captured along the length of the flow cell over 24 h (n = 3).

Figure 5

Static biofilm formations. (A) Static biofilm formation of SPI2 and rpoS deletion mutants. Following incubation at 25°C for 48 h in CFA medium, the level of biofilm formation is expressed as a percentage of WT SL1344 which had an A_{590 nm} of 0.15 ± 0.02 in this experiment. The ΔssrA (JH3180) and ΔrpoS (JH3142) mutants showed significantly less attachment to polystrene than WT SL1344 after 24 h (data not shown) and 48 h of growth (* P = 0.001-0.02). Complementation of ΔssrA with a low copy plasmid encoding ssrAB (JH3181) restored the ability of this mutant to form WT biofilm after 48 h (* P = 0.002). The mean absorbance values from four wells are shown as a percentage of WT SL1344 and the error bars represent the SD between 4 technical replicates (n = 2). (B) Static biofilm formation of SL1344 over-expressing SsrAB (p1437-1) in CFA medium (+ 0.1% L-arabinose where indicated) for 24 h at 25°C. The level of biofilm formation is expressed as a percentage of WT SL1344, which had an A_{590 nm} of 0.14 ± 0.01 in this experiment. The induction of SsrAB expression by an arabinose-inducible promoter significantly inhibited attachment (* P = 0.000007) when compared to WT SL1344. No significant difference in attachment was observed in the control strain over-expressing SsrAB in the reverse orientation (p1437-6). The mean absorbance values from six wells are shown and the error bars represent the SD between 4 technical replicates.

Figure 6

The effect of aromatic amino acids on static biofilm formation of wild-type S. Typhimurium. CFA medium was supplemented with increasing concentrations of aromatic amino acids. The level of biofilm formation after 12, 24 and 48 h of growth at 25°C is expressed as a percentage of the biofilm formed after unsupplemented growth in CFA at 48 h, which had an A_{590 nm} of 0.15 ± 0.02 in this experiment. The mean absorbance values from four wells are shown and the error bars represent the SD between 4 technical replicates.

Figure 7

Tryptophan biosynthesis is required for biofilm formation. The effect of the trpE mutation and addition of tryptophan (0.01 mM, 0.1 mM) and indole (0.01 mM, 0.1 mM) on static biofilm growth of ΔtrpE (JH3185) was determined after 24 h of growth in CFA broth at 25°C. The addition of tryptophan and indole significantly (*P = 0.02-0.00003) increased biofilm formation when compared to ΔtrpE grown in CFA alone. The mean absorbance values from four wells are shown and the error bars represent the SD between 4 technical replicates.