Global systems-level analysis of Hfq and SmpB deletion mutants in Salmonella: implications for virulence and global protein translation

Abstract

Using sample-matched transcriptomics and proteomics measurements it is now possible to begin to understand the impact of post-transcriptional regulatory programs in Enterobacteria. In bacteria post-transcriptional regulation is mediated by relatively few identified RNA-binding protein factors including CsrA, Hfq and SmpB. A mutation in any one of these three genes, csrA, hfq, and smpB, in Salmonella is attenuated for mouse virulence and unable to survive in macrophages. CsrA has a clearly defined specificity based on binding to a specific mRNA sequence to inhibit translation. However, the proteins regulated by Hfq and SmpB are not as clearly defined. Previous work identified proteins regulated by hfq using purification of the RNA-protein complex with direct sequencing of the bound RNAs and found binding to a surprisingly large number of transcripts. In this report we have used global proteomics to directly identify proteins regulated by Hfq or SmpB by comparing protein abundance in the parent and isogenic hfq or smpB mutant. From these same samples we also prepared RNA for microarray analysis to determine if alteration of protein expression was mediated post-transcriptionally. Samples were analyzed from bacteria grown under four different conditions; two laboratory conditions and two that are thought to mimic the intracellular environment. We show that mutants of hfq and smpB directly or indirectly modulate at least 20% and 4% of all possible Salmonella proteins, respectively, with limited correlation between transcription and protein expression. These proteins represent a broad spectrum of Salmonella proteins required for many biological processes including host cell invasion, motility, central metabolism, LPS biosynthesis, two-component regulatory systems, and fatty acid metabolism. Our results represent one of the first global analyses of post-transcriptional regulons in any organism and suggest that regulation at the translational level is widespread and plays an important role in virulence regulation and environmental adaptation for Salmonella.

Figure 1. Comparison of survival in mice.

Five BALB/c mice were i.p. infected with 200 cfu of Salmonella for each strain; the wild type strain, the Δhfq mutant strain and the ΔsmpB mutant strain, and monitored for 21 days. Alive mice were counted and shown in % survival.

Figure 2. Comparison of survival within macrophage cells.

RAW264.7 cells were infected with Salmonella cells grown overnight in LB via MOI of 50. After infection initiation by centrifuging the bacteria onto the cell monolayers at 1,000×g for 5 min, cells were incubated at 37°C with 5% CO₂ for 30 min, treated with gentamicin (100 µg/ml) for 1 h, and incubated in DMEM with gentamicin (20 µg/ml) for the remainder of the test. Macrophage cells were lysed using 1% Triton-X at 1.5 h (A) and 18 h (B) post-infection and the intracellular bacteria were enumerated by serial dilution of cell lysates. The data shown represents the mean and standard deviation of three individual experiments.

Figure 3. Summary of transcriptional and translational control functions of Hfq and SmpB.

Venn diagram showing overlap between transcripts and proteins regulated by Hfq (Panel A), between transcripts and proteins regulated by SmpB (Panel B) and between proteins co-regulated by Hfq and SmpB (Panel C) across all growth conditions.

Figure 4. Measurement of protein levels of pdu genes in ΔsmpB and Δhfq strains grown under LB Stat and acidic minimal media conditions.

Heatmap showing 134 Pdu peptides that represent 11 individual Pdu proteins up-regulated under acidic minimal media (AMM-2) growth condition (Panel A) and 124 Pdu peptides that represent 11 individual Pdu proteins up-regulated under LB Stat growth condition (Panel D) in Δhfq and ΔsmpB strains relative to wild-type. Salmonella strains harboring HA-tag at each Pdu gene were grown in acidic minimal media (AMM-2) (Panel B and C) and under LB stat phase (Panel E and F) as described. Same amount of cell lysates was loaded in each lane and probed by Western blot analysis for the Pdu-encoded proteins and a control protein DnaK. The level of Pdu-encoded protein expressed in each strain was normalized to the DnaK level. Intracellular DnaK level in each lane was set to 100%.

Figure 5. Transcripts of pdu genes remain unchanged in ΔsmpB and Δhfq strains.

Total RNA was isolated from Salmonella cells grown under acidic minimal media (AMM-2) growth condition (Panel A) or LB Stat growth condition (Panel B) and mRNA level of pdu genes was quantified via RT-PCR using gyrB mRNA level for normalization. As controls the mRNA level of smpB and hfq genes were also quantified in a similar fashion. Expression ratios (smpB or hfq/14028s) were averaged from three individual experiments and shown on a logarithmic scale.