Computational prediction and experimental validation of Salmonella Typhimurium SopE-mediated fine-tuning of autophagy in intestinal epithelial cells

Abstract

Macroautophagy is a ubiquitous homeostasis and health-promoting recycling process of eukaryotic cells, targeting misfolded proteins, damaged organelles and intracellular infectious agents. Some intracellular pathogens such as Salmonella enterica serovar Typhimurium hijack this process during pathogenesis. Here we investigate potential protein-protein interactions between host transcription factors and secreted effector proteins of Salmonella and their effect on host gene transcription. A systems-level analysis identified Salmonella effector proteins that had the potential to affect core autophagy gene regulation. The effect of a SPI-1 effector protein, SopE, that was predicted to interact with regulatory proteins of the autophagy process, was investigated to validate our approach. We then confirmed experimentally that SopE can directly bind to SP1, a host transcription factor, which modulates the expression of the autophagy gene MAP1LC3B. We also revealed that SopE might have a double role in the modulation of autophagy: Following initial increase of MAP1LC3B transcription triggered by Salmonella infection, subsequent decrease in MAP1LC3B transcription at 6h post-infection was SopE-dependent. SopE also played a role in modulation of the autophagy flux machinery, in particular MAP1LC3B and p62 autophagy proteins, depending on the level of autophagy already taking place. Upon typical infection of epithelial cells, the autophagic flux is increased. However, when autophagy was chemically induced prior to infection, SopE dampened the autophagic flux. The same was also observed when most of the intracellular Salmonella cells were not associated with the SCV (strain lacking sifA) regardless of the autophagy induction status before infection. We demonstrated how regulatory network analysis can be used to better characterise the impact of pathogenic effector proteins, in this case, Salmonella. This study complements previous work in which we had demonstrated that specific pathogen effectors can affect the autophagy process through direct interaction with autophagy proteins. Here we show that effector proteins can also influence the upstream regulation of the process. Such interdisciplinary studies can increase our understanding of the infection process and point out targets important in intestinal epithelial cell defense.

Keywords: Host-microbe interactions; MAP1LC3B; Salmonella Typhimurium; SopE; autophagy; network biology.

Figure 1

Interaction network between selected Salmonella Typhimurium effectors, host cell transcription factors and autophagy core genes. (A) Network analysis of potential interactions between Salmonella and host autophagy. Red ovoid nodes are the selected Salmonella effector proteins. Red edges are PPI predictions. Thin edges were predicted by one of the three methods. Thick edges were predicted by two of the three methods. Yellow triangular nodes are host transcription factors that Salmonella effectors can influence. These are clustered according to the number of Salmonella effectors they are targeted by. Yellow edges reflect transcriptional regulation of core autophagy genes (round green nodes) by the transcription factors. The size of the green nodes is proportional to the number of transcription factors they are connected to. (B) Subnetwork illustrating the potential interaction of the Salmonella effector SopE with transcription factors affecting specific autophagy core genes. The same layout was used here as for the large network.

Figure 2

The Salmonella effector SopE can bind to the human transcription factor SP1, as predicted. (A) Co-immunoprecipitation of GFP-SopE and BirA-Myc-SP1 ectopically expressed in HEK293 cells. (B) SP1 signal intensity ratio after Co-IP with GFP or with GFP-SopE showing enrichment of SP1 when GFP-SopE is used compared with GFP alone, based on two independent measurements.

Figure 3

MAP1LC3B gene expression is increased at the early stages of Salmonella infection but decreased at later stages in a SopE-dependent manner in Salmonella-containing epithelial cells (A) but not in bystander cells (B). MAP1LC3B expression levels expressed as log2-ΔCT so that uninfected cells (ui) are also displayed. Continuous borders = infected cells containing Salmonella (wt or mutant) and dashed borders = bystander epithelial cells not containing Salmonella or uninfected cells. * p<0.05.

Figure 4

SopE dampens the autophagic flux by 6h post infection. LC3 puncta number and p62 dot intensity in HT-29 epithelial cells infected for 6h with either wt Salmonella strain or its ΔsopE gene deletion derivative. (A, B) HT-29 epithelial cells were infected with wt Salmonella and its ΔsopE gene deletion derivative strain as indicated before. (C, D) HT-29 cells were pre-treated with Rapamycin for 11h prior to and during the 6h-long infection (Maximum 17h). LC3 puncta (A, C) and p62 dot intensity (B, D) was quantified from HT-29 cells containing Salmonella. *p=0.05. (E) Micrographs of HT-29 cells pre-treated with Rapamycin showing LC3 puncta (red, left) and p62 dots (red, right), intracellular Salmonella cells (green) and nuclei (blue) illustrating panels (C, D).

Figure 5

Cytosolic Salmonella no longer can dampen the autophagy flux in a SopE-dependent manner. (A, B) Autophagy was pre-induced with Rapamycin in HT-29 epithelial cell monolayers and maintained during the 6h infection with Salmonella, as it exacerbates the impact of SopE on modulating autophagy, making it easier to visualise. LC3 puncta (A) and p62 intensity (B) was quantified from HT-29 cells containing Salmonella. *p=0.05.