The translocated Salmonella effector proteins SseF and SseG interact and are required to establish an intracellular replication niche

Abstract

The facultative intracellular pathogen Salmonella enterica causes a variety of diseases, including gastroenteritis and typhoid fever. Inside epithelial cells, Salmonella replicates in vacuoles, which localize in the perinuclear area in close proximity to the Golgi apparatus. Among the effector proteins translocated by the Salmonella pathogenicity island 2-encoded type III secretion system, SifA and SseG have been shown necessary but not sufficient to ensure the intracellular positioning of Salmonella vacuoles. Hence, we have investigated the involvement of other secreted effector proteins in this process. Here we show that SseF interacts functionally and physically with SseG but not SifA and is also required for the perinuclear localization of Salmonella vacuoles. The observations show that the intracellular positioning of Salmonella vacuoles is a complex phenomenon resulting from the combined action of several effector proteins.

FIG. 1.

Identification of new effectors required for Golgi localization of SCVs in HeLa cells. (A and B) HeLa cells were infected with GFP-expressing wild-type (wt) Salmonella or mutants lacking different effectors and fixed at 8 h postinfection. (A) Cells were immunostained for Lamp1 and the Golgi marker Giantin and scored by immunofluorescence for the intracellular positioning of bacteria. Only bacteria enclosed in a Lamp1-positive compartment that were either completely or partially surrounded by the Golgi marker were counted as being Golgi associated. Bacterial clusters that were found adjacent to the Golgi but did not fulfill the above criteria were counted as nonassociated. At least 50 host cells, corresponding to more than 100 bacteria, were scored blind in each experiment. Statistical analysis for comparison of wild-type Salmonella and mutant strains indicated a significant difference for the ssaV, sseF, and sifA strains (P < 0.001), whereas no significant difference was observed in comparison of the wt to other mutants or the complemented sseF and sifA strains (P > 0.05). (B) Confocal immunofluorescence images of the dispersed distribution of the sseF mutant in contrast with the Golgi localization of the wt strain. The cell shape is marked. Bars, 10 μm. (C) Intracellular replication of Salmonella strains. Values indicate the fold increase, calculated as the ratio of intracellular bacteria between 2 and 16 h after invasion. Statistical analysis indicated a significant difference between mutant strains and wild-type Salmonella (P < 0.001), whereas no significant difference was observed between mutant strains. (D) Complementation of the intracellular positioning of the sseF and sifA mutants to the Golgi region by ectopic expression of myc-SseF and myc-SifA, respectively. (A, C, and D) Standard deviations of the means are shown and correspond to three independent experiments.

FIG. 2.

Role of membrane-targeting domains and cellular localization of ectopically expressed effectors. (A) Protein lysates of myc-tagged proteins expressed in HeLa cells were partitioned using Triton X-114. Aliquots of each fraction were analyzed by Western blotting after SDS-PAGE. S, soluble phase; TX114, membrane-enriched phase. For controls, the partitioning of GFP, Lamp1, and Rab7 was analyzed. (B) Confocal immunofluorescence analysis of HeLa cells transfected with plasmids expressing myc-tagged effectors (two upper panels) and cotransfection of myc- and HA-tagged effectors (lower panel). The cells were fixed 24 h posttransfection and labeled using appropriate antibodies. Tagged SseF and SseG colocalized with each other as well as with the Golgi marker TGN46. Bars, 10 μm. (C) Schematic diagram showing that the region of SseF homologous to the Golgi-targeting domain of SseG is not sufficient for Golgi localization of the ectopically expressed myc-tagged protein. HR, hydrophobic region.

FIG. 3.

Interaction of SseF and SseG. (A) Competitive index analysis of Salmonella mutant strains. The competitive indices of wild type (wt) versus sseG (n = 3) and of wt versus sseF (n = 6) are not significantly different (P = 0.71) but are significantly (P = 0.0012) or very significantly (P = 0.0006) lower than 1, respectively. The competitive indices of sseG versus sseF sseG (n = 3; P = 0.30) and sseF versus sseF sseG (n = 4; P = 0.12) are not significantly different from 1. These data indicate that the two genes are functionally linked. (B) GST-SseF bound to beads was incubated with extracts of HeLa cells expressing myc-tagged Salmonella effectors. Total lysates and proteins bound to washed beads were analyzed by Western blotting using an anti-myc antibody. (C) HeLa cells were cotransfected with plasmids coding for myc- and HA-tagged effector proteins as indicated, lysed, and immunoprecipitated (IP) using an anti-myc antibody coupled to Sepharose beads. HA-tagged proteins present in lysates (upper panels) and coimmunoprecipitated with myc-tagged effectors (lower panel) were analyzed by Western blotting (WB). The lysate membrane was reanalyzed for myc-tagged proteins (encircled bands, middle panel). Controls indicated the effective binding of myc-tagged proteins to the beads (data not shown). The top and bottom bands seen after anti-myc immunoprecipitation (lower panel) correspond to mouse IgG heavy and light chains.

FIG. 4.

Interaction of translocated SseF and SseG. HeLa cells were infected with an S. enterica serovar Typhimurium sseF sseG double mutant harboring plasmid p2888 (sseF-HA and sseG-M45) (C). As controls, double mutants harboring plasmids p2643 (sseF-HA) (A) or p2788 (sseG-M45) (B) were used. The tagged proteins were detected in extracts of bacteria grown under phosphate starvation, inducing ssrAB-regulated proteins in vitro (left panel). SseG-M45 and SseF-HA were detected in the postnuclear supernatant (PNS) of HeLa cells infected for 12 h (middle panel). The signal for SseF-HA was intensified by successive incubations of the blotting membrane with a rabbit anti-rat and a goat anti-rabbit antibody, both coupled to peroxidase. Tagged proteins were subsequently immunoprecipitated (IP) from the PNS using anti-HA or anti-M45 antibodies coupled to the Sepharose beads. Immunoprecipitated proteins were analyzed by Western blotting (WB, right panel). The fractions of translocated SseF detected in the coimmunoprecipitation or PNS differ in electrophoretic mobility (apparent molecular masses, 33.7, 32.5, and 30.5 kDa) compared to SseF-HA secreted in vitro (apparent molecular mass, 31.6 kDa). Expression and secretion of SseJ were controlled using Salmonella wild type, the double mutant strain expressing sseF-HA and sseG-M45, and an sseJ mutant. The absence of translocated SseJ with sseF-HA and sseG-M45 was used to check the specificity of the immunoprecipitations.