SPI-1 of Salmonella enterica serovars is an important virulence factor that has been well-studied in several animal models and in disease outbreaks [53, 54]. SPI-1-mediated infectivity of Salmonella Typhimurium was examined by creation of null mutations in genes that were up-regulated in our previous host-pathogen microarray study (data not shown). Those Salmonella genes, pyruvate formate lyase I (pflB) and bifunctional acetaldehyde-CoA/alcohol dehydrogenase (adhE), play a major role in the pyruvate metabolism of the organism (Figure 1). Here we show that these two genes can also affect virulence and SPI-1 gene expression.
For the HCT-8 intestinal epithelial cell line, internalized mutant Salmonella CFUs were significantly increased over the wild-type at 4 h post-infection and lasted through 48 h post-infection at which time the experiment was terminated (Figure 2). Since the T3SS encoded by SPI-1 is canonically involved in early contact with non-phagocytic cells such as fibroblasts and intestinal epithelia [27, 30, 31], a selected group of SPI-1 genes (Table 1) in mutant Salmonella were measured by qPCR and compared to the wild-type expression. Several SPI-1 genes were up-regulated after HCT-8 cell challenge, from 1 h post-infection lasting to 24 h post-infection when SPI-1 expression in the mutants decreased compared to the wild-type (Table 2). These up-regulated genes included hilC, hilD, rtsA, sipA, - all major genes involved in the early response to infection . Some SPI-1 genes were up-regulated at multiple times post-infection and in both mutants, such as sicA, a gene encoding a complex chaperone protein required for Salmonella entry into host cells [56, 57]. Overall, the results indicated that increased internalized growth of strains of S. Typhimurium deficient for transcripts of the pyruvate pathway in HCT-8 cells may in part be due to a differential regulation of genes in the T3SS.
Additional evidence in supporting this notion included qPCR results on SPI-1 genes in log-phase cultures alone. We found that the major regulators of SPI-1 were up-regulated when compared to the wild-type. These included hilA and hilD in both of the mutant strains, and hilC and rtsA in at least one of the mutant strains (Figure 5). HilA is the major regulator of SPI-1 invasion . The hilA gene encodes an OmpR/ToxR family transcriptional regulator that activates the expression of invasion genes in response to both environmental and genetic factors . Further, the proteins encoded by rtsA, hilC and hilD bind to a DNA region upstream of hilA and induce hilA expression [58–60]. Therefore, these genes could be expected to be induced at an early stage, as they are key modulators of invasive Salmonella infection and hilD appears to bridge communication between SPI-1 and SPI-2 regulation (reviewed in ). These results further indicate the increased virulence of the mutant strains to HCT-8 is occurring via a differential regulation of genes involved in the T3SS. Interestingly, the sicP gene was also up-regulated in both mutant strains measured from cell culture alone. SicP serves as a chaperone for the SPI-1 effector Salmonella protein tyrosine phosphatase (SptP). SptP has been shown to have several functions in the establishment of Salmonella infection, in both epithelial cells and macrophages [61, 62]. While the gene expression of SptP was not examined, the up-regulation of the chaperone sicP transcripts could indicate early formation of SicP-SptP complexes even before contact with a host, as these two genes appear to be transcriptionally coupled . The full significance of the gene expression response of sicP to alterations in the pyruvate metabolism pathway remains to be further explored.
We further examined the effects of the mutant strains on the infectivity of THP-1 macrophage cell line. We observed no significant difference in the internalization of Salmonella mutants over the wild-type strain, whether the invasion assays were using opsonized or non-opsonized bacteria (Figures 2 & 3). Since we were interested in addressing the effects of our mutations on SPI-1-dependent virulence, we chose to examine SPI-1 gene expression using invasion assays with non-opsonized bacteria. Interestingly, while no significant internalization between strains was observed, SPI-1 gene expression was mostly down-regulated in the mutants after a challenge of THP-1 cells (Tables 3 & 4). During the systemic phase of macrophage infection by Salmonella, SPI-1 expression is typically repressed to allow internal replication of the bacterium and suppress excess apoptosis . Therefore, as differentiated THP-1 are phagocytic-type cells and infection is largely established via SPI2-encoded T3SS, low SPI-1 expression of invaded macrophages could therefore be expected based on previous efforts [65–70].
The specific Salmonella knockout genes and the role they may play in its pathogenesis were further examined. Our data supports the overall results of Huang et al.  that a pflB mutation in Salmonella can effect SPI-1 gene expression as measured from cultures alone. In addition to the previous work , we identified multiple SPI-1 genes that are differentially expressed both in cultures and at several time-points post-infection of an in vitro challenge using human HCT-8 cells. Knockout of the pflB gene likely produces an imbalance of metabolites in which the presence mimics the conditions in the host gastrointestinal tract . We further show that another gene of the Salmonella pyruvate pathway (adhE) can have a similar effect on intestinal epithelia, suggesting that gut metabolites in addition to formate and/or other feedback mechanisms due to environmental cues may be at work. As AdhE is a multifunctional enzyme, another possibility for increased invasion of a ΔadhE mutant in addition to an SPI1-mediated virulence mechanism could be formulated. A lack of AdhE appears to stimulate type I fimbrial adhesion [71, 72], important adhesions on the surface of Salmonella that may mediate attachment and internalization of bacteria . Further experimentation would be warranted to determine the exact mechanism and co-factors involved with the pyruvate pathway involvement in Salmonella virulence.
For the generation of mutant strains, each recombination event was confirmed by PCR using primers to the flanking genomic regions. These confirmations, in addition to the evidence provided in this study, give confidence in the specificity of our deletions. By using lambda-red for site-specific mutagenesis, evidence suggests any potential downstream effects are reduced or eliminated with this system over that of random mutagenesis/transposons, including issues of polarity (e.g. ). We do acknowledge however that complementation assays to restore the phenotype could provide additional evidence, but these were not performed here.
We also acknowledge the role the host microbiome plays in Salmonella pathogenicity , particularly since microbiota may contribute to pyruvate pathway metabolites utilized in the establishment of infection . The mammalian gastrointestinal tract contains high levels of short chain fatty acids as the result of the breakdown of foodstuffs by the digestive processes and the action of resident bacteria . We therefore would be highly interested in future in vivo studies to help determine the mechanisms involved in the interplay between the host genetics, the gut environment, and the microbiome on Salmonella pathogenesis when dealing with mutant strains produced through metabolic pathway mutations.