Report

A Process for Controlling Intracellular Bacterial Infections Induced by Membrane Injury

See allHide authors and affiliations

Science  04 Jun 2004:
Vol. 304, Issue 5676, pp. 1515-1518
DOI: 10.1126/science.1098371

Abstract

Strategies for inhibiting phagolysosome fusion are essential for the intracellular survival and replication of many pathogens. We found that the lysosomal synaptotagmin Syt VII is required for a mechanism that promotes phagolysosomal fusion and limits the intracellular growth of pathogenic bacteria. Syt VII was required for a form of Ca2+-dependent phagolysosome fusion that is analogous to Ca2+-regulated exocytosis of lysosomes, which can be triggered by membrane injury. Bacterial type III secretion systems, which permeabilize membranes and cause Ca2+ influx in mammalian cells, promote lysosomal exocytosis and inhibit intracellular survival in Syt VII +/+ but not –/– cells. Thus, the lysosomal repair response can also protect cells against pathogens that trigger membrane permeabilization.

Synaptotagmins (Syt) are thought to function as transducers of Ca2+ signaling in membrane fusion events, through interactions mediated by their cytosolic C2A and C2B domains (1). The Syt VII isoform is localized on lysosomes, where it regulates Ca2+-triggered lysosomal exocytosis (2). Phagolysosome fusion is also a Ca2+-dependent process (3, 4), so we investigated whether modulating Syt VII function affected the intracellular survival of bacteria that have to evade lysosomes in order to replicate. We found that expression in CHO cells of the soluble Syt VII C2A domain, which inhibits Ca2+-triggered lysosomal exocytosis (2), did not affect invasion by Salmonella enterica serovar Typhimurium, but enhanced its intracellular growth between 1 and 2 hours after it entered the host cell (Fig. 1, A to C).

Fig. 1.

Syt VII function is required for restricting the growth of S. enterica serovar Typhimurium. (A) CHO cells transfected with GFP or Syt VII C2A–GFP were infected with Salmonella for 1 hour, and the number of intracellular bacteria after 1 and 2 hours was determined by inside-outside immunofluorescence. The mean of triplicates ± SD is shown; asterisks indicate significant differences (P < 0.05, t test) between Syt VII +/+ and –/– cells (1 hour, P = 0.034; 2 hours, P = 0.017). (B) CHO cells 2 hours after infection. Blue, 4′,6′-diamidino-2-phenylindole (DAPI) nuclear staining; red, intracellular bacteria; green, extracellular bacteria and cytosolic GFP. (C) CHO cells transfected with GFP or Syt VII C2A–GFP were sorted by fluorescence-activated cell sorting, replated, and infected with Salmonella for 1 hour; the number of intracellular bacteria after 1 and 2 hours was determined by a gentamicin protection assay (GPA). The mean of triplicates ± SD is shown (1 hour, P = 0.233; 2 hours, P = 0.077). (D) MEFs from Syt VII +/+ or deficient –/– embryos were infected with Salmonella for 1 hour, and the number of intracellular bacteria was determined by GPA. The mean of triplicates ± SD is shown; asterisks indicate significant differences between Syt VII +/+ and –/– cells (1 hour, P = 0.253; 2 hours, P = 0.010; 4 hours, P = 0.0001; 6 hours, P = 0.0001). (E) MEFs infected with GFP-expressing Salmonella 6 hours after infection: Bacteria (arrows) replicate in compartments that do not fuse with dextran–Texas red–loaded lysosomes. (F) Bone marrow macrophages from Syt VII +/+ or –/– mice were infected with Salmonella for 1 hour, and the number of intracellular bacteria was determined by GPA. The mean of triplicates ± SD is shown; asterisks indicate significant differences between Syt VII +/+ and –/– cells (1 hour, P = 0.507; 2 hours, P = 0.007; 4 hours, P = 0.0001; 6 hours, P = 0.00001). Fluorescent images on the right show intracellular GFP-expressing Salmonella (arrows) 6 hours after infection.

The involvement of Syt VII in a mechanism that inhibits Salmonella growth was confirmed using primary cells isolated from Syt VII–deficient mice (5). Intracellular growth in the first 6 hours after invasion was enhanced in Syt VII –/– murine embryonic fibroblasts (MEFs) when compared to cells isolated from Syt VII +/+ embryos (Fig. 1D). As previously reported (6, 7), in both Syt VII +/+ and –/– MEFs, Salmonella replicated in vacuoles that did not fuse with lysosomes (Fig. 1E), in spite of containing the lysosomal glycoprotein Lamp-1 (fig. S1) (8). In macrophages from Syt VII +/+ mice, intracellular replication of Salmonella could only be detected approximately 4 hours after infection, which is consistent with previous reports (9). In Syt VII –/– macrophages, however, bacterial growth started shortly after entry and proceeded vigorously with no lag period (Fig. 1F). This increase in intracellular growth was not related to Salmonella-induced macrophage cytotoxicity. Under the growth conditions used, Salmonella induced similar low levels of programmed cell death on both Syt VII +/+ and –/– macrophages and did not cause differential cell loss between 2 and 6 hours after infection (fig. S2). The bacteria replicating intracellularly, as previously reported (7), were in Lamp-1–positive vacuoles devoid of dextran preloaded into lysosomes. Thus, Salmonella evades phagolysosome fusion more efficiently in Syt VII–deficient cells.

We also investigated whether Syt VII deficiency affected the intracellular replication of Listeria monocytogenes. Shortly after invasion, this bacterium escapes from the phagosome and replicates free in the cytosol, where it sequesters host actin as a propelling mechanism (10). Nearly identical intracellular growth curves were observed with Listeria in Syt VII +/+ or –/– macrophages (fig. S3A). In addition, similar numbers of bacteria were found associated with actin tails 2 hours after infection, in both Syt VII +/+ and –/– macrophages (80 and 84%, respectively). Thus, Syt VII deficiency does not affect the process by which Listeria escapes from the phagosome before lysosomal fusion (11) and also does not cause major alterations in components of the cytosol required for actin-based motility.

Lysosomal degradation of endocytosed epidermal growth factor occurred with identical kinetics in Syt VII +/+ and –/– MEFs (fig. S3B). Furthermore, nonpathogenic Escherichia coli was degraded at a similar rate after uptake by Syt VII +/+ or –/– macrophages (fig. S3C). Thus, Syt VII deficiency did not cause any blockage of the normal endocytic traffic into degradative lysosomes. Defects in nitric oxide (NO) production or NADPH oxidase activity also did not provide an explanation for the growth advantage of Salmonella in Syt VII –/– cells: Similar levels of NO were produced in response to lipopolysacccharide and interferon-γ (fig. S3D), and robust superoxide production was visualized in zymozan-containing phagosomes in both Syt VII +/+ and –/– macrophages (fig. S3, G and J).

Syt VII functions in the regulation of rapid Ca2+-dependent fusion of lysosomes with the plasma membrane (2, 5, 12). We investigated whether Syt VII –/– macrophages had a defect in the fusion of lysosomes with recently formed phagosomes, which might still retain plasma membrane properties. Lysosomes were loaded with fluorescent dextran, and green fluorescent protein (GFP)–expressing Salmonella was incubated with the macrophages for 10 or 20 min, followed by antibody staining to distinguish extracellular bacteria and analysis by confocal microscopy (Fig. 2, A and B). The extent to which Salmonella-containing phagosomes colocalized with dextran preloaded into lysosomes was significantly reduced in Syt VII –/– when compared to Syt VII +/+ macrophages (Fig. 2C). These results reconcile previous reports describing early phagolysosome fusion (13) and killing of Salmonella in mouse macrophages (14), with the widely demonstrated capacity of Salmonella to block lysosomal fusion through translocated type III secretion system effectors (7). Syt VII–dependent phagolysosome fusion appears to play an important role in the initial inhibition of intracellular growth observed in wild-type macrophages (Fig. 1F). The fact that intracellular growth is only detected several hours after infection (9) suggests that a fraction of the surviving bacteria, which succeeded in modifying the phagosome to block lysosomal fusion, gradually increase in number.

Fig. 2.

Syt VII promotes phagolysosome fusion. Bone marrow macrophages derived from (A) Syt VII +/+ or (B) –/– mice were incubated with dextran–Texas red to label lysosomes and incubated with GFP-expressing Salmonella for 10 or 20 min before confocal microscopy analysis. Extracellular bacteria were stained with antibodies to Salmonella (blue, arrowheads). Arrows point to phagosomes containing Salmonella that fused with dextran–Texas red–containing lysosomes. (C) The number of phagosomes containing dextran–Texas red was determined visually on confocal Z-stack images. The mean of triplicates ± SD is shown; asterisks indicate significant differences (P < 0.05, t test) between Syt VII +/+ and –/– (10 min, P = 0.005; 20 min, P = 0.040). n = number of phagosomes analyzed. The results are representative of three independent experiments, in which a total of 3167 and 3784 phagosomes were analyzed in Syt VII +/+ or –/– macrophages, respectively.

Syt VII–dependent lysosomal exocytosis promotes membrane resealing, a process triggered by Ca2+ influx through membrane wounds (12). We investigated whether a lysosomal response to membrane injury was involved in the survival advantage of Salmonella in Syt VII–deficient cells. If a Syt VII–dependent form of phagolysosome fusion occurred in response to membrane injury, one prediction is that it should be dependent on extracellular Ca2+. This was confirmed: When wild-type macrophages were infected with Salmonella in medium containing EGTA [a condition that does not inhibit actin-mediated particle uptake (15)], the initial decline in bacterial numbers observed in wild-type macrophages was markedly altered, resembling instead the rapid growth pattern typically observed in Syt VII –/– macrophages (Fig. 3A). These findings are consistent with a requirement for Ca2+ influx from the extracellular medium for the Syt VII–dependent growth inhibition response.

Fig. 3.

Ca2+ influx is required for the Syt VII–dependent killing of intracellular Salmonella, under conditions that induce type III secretion system–dependent lysosomal exocytosis. (A) Syt VII +/+ and –/– macrophages were preincubated or not with 2 mM EGTA and infected with Salmonella for 30 min in the absence (+ calcium) or presence (+ EGTA) of 2 mM EGTA, and the number of intracellular bacteria at various times after infection was determined by GPA. The mean of triplicates ± SD is shown; asterisks indicate significant differences (P < 0.05, t test) between Syt VII +/+ and –/– (+ calcium: 2 hours, P = 0.067; 4 hours, P = 0.040; 6 hours, P = 0.004; + EGTA: 2 hours, P = 0.592; 4 hours, P = 0.472; 6 hours, P = 0.060). (B) Syt VII +/+ and –/– MEFs were exposed to wild-type Salmonella or the type III secretion mutant invA for 20 min at 37°C, followed by live surface immunofluorescent labeling of Lamp-1. Images were acquired under identical exposure conditions; the DAPI stain shows cell nuclei and cell-associated bacteria (arrows) on the same field.

Next, we investigated the possible source of membrane injury responsible for Ca2+ influx during Salmonella infection. Several pathogenic bacteria express specialized secretion systems that mediate direct transport of effector proteins into the cytosol of host cells (16). The type III secretion systems of Gram-negative bacteria have been extensively characterized and are known to transiently permeabilize the plasma membrane of target cells, through assembly of a macromolecular structure referred to as the translocon (17, 18). In Salmonella, the SPI-1 type III secretion system responsible for translocating invasion effectors is known to promote Ca2+ influx (19). We found that wild-type Salmonella exhibits extensive pore-forming activity when exposed to mammalian cells, whereas no activity is seen with the SPI-1 type III secretion mutant invA (19). Pore formation (20) was also detected in cells exposed to the noninvasive, effectorless SB1304 Salmonella mutant (21), confirming that permeabilization is dependent on a functional type III secretion apparatus but not on the introduction of effectors that stimulate membrane ruffling and bacterial entry (fig. S4). In order to determine whether the cell permeabilization induced by the Salmonella SPI-1 secretion system triggered Syt VII–dependent lysosomal exocytosis, we incubated Syt VII +/+ or –/– MEFs with wild-type or invA Salmonella, and stained the cells for detection of surface-exposed Lamp-1 (12). Wild-type Salmonella induced extensive surface exposure of Lamp-1 in Syt VII +/+ MEFs (Fig. 3B). In contrast, a markedly fainter staining was observed under the same conditions in Syt VII –/– MEFs. The invA mutant, which does not express a functional SPI-1 type III secretion system, did not trigger detectable lysosomal exocytosis in either cell type (Fig. 3B).

Recently formed phagosomes are expected to contain high concentrations of Ca2+, similar to those in the extracellular medium. It was therefore of interest to determine whether type III secretion–mediated permeabilization of phagosomes and Ca2+ release would result in bacterial killing, as a consequence of lysosome fusion. A Yersinia pseudotuberculosis strain lacking the YopE and YopH effectors (yopEH mutant), previously shown to promote extensive pore formation in mammalian cells through its well-characterized type III secretion system (17) (fig. S4), was compared with a strain lacking a functional type III secretion apparatus (ysc mutant). Assays for surface exposure of Lamp-1 confirmed that Yersinia behaves similarly to Salmonella, inducing exocytosis of lysosomes in a Syt VII– and type III secretion system–dependent manner (Fig. 4A). Given that these strains of Y. pseudotuberculosis are able to survive and slowly replicate in macrophages (22), we investigated the fate of the yopEH and ysc mutants in Syt VII +/+ or –/– macrophages. The strain expressing a functional type III secretion system, the yopEH mutant, showed an initial decrease in intracellular survival in Syt VII +/+ macrophages (Fig. 4B), similar to what was observed with Salmonella (Fig. 1F). Also similar to what was observed with wild-type Salmonella, intracellular survival of the Yersinia yopEH mutant was enhanced in Syt VII –/– macrophages (Fig. 4B). In contrast, when infections were performed with the ysc mutant lacking a functional type III secretion system, no significant differences were observed in the numbers of viable intracellular bacteria between Syt VII +/+ and –/– macrophages, from 2 to 12 hours after infection (Fig. 4B, lower panel).

Fig. 4.

The Yersinia type III secretion system induces Syt VII–dependent lysosomal exocytosis and killing of intracellular bacteria. (A) Syt VII +/+ and –/– MEFs were exposed to Yersinia yopEH or ysc (type III secretion mutant) for 1 hour at 37°C, followed by live surface labeling with a monoclonal antibody against the lumenal domain of Lamp-1. Images were acquired under identical exposure conditions; the DAPI image shows cell nuclei and cell-associated bacteria (arrows) on the same field. (B) Syt VII +/+ and –/– macrophages were infected with the Yersinia yopEH or ysc type III secretion mutant for 30 min, followed by treatment with gentamicin and isopropyl-β-D-thiogalactopyranoside to induce GFP expression. Numbers of viable intracellular bacteria were determined at different time points; the mean of triplicates ± SD is shown. Asterisks indicate significant differences (P < 0.05, t test) between Syt VII +/+ and –/– (yopEH: 2 hours, P = 0.399; 6 hours, P = 0.015; 12 hours, P = 0.016; ysc: 2 hours, P = 0.975; 6 hours, P = 0.765; 12 hours, P = 0.338).

Thus, membrane injury inflicted by type III secretion systems triggers a Syt VII–dependent process that reduces the number of viable intracellular bacteria. We envision that shortly after invasion, recently formed phagosomes are permeabilized by bacterial type III secretion translocons, triggering Ca2+ influx from the intraphagosomal space into the cytosol, Syt VII–dependent phagolysosome fusion, and bacterial killing (fig. S5). In this scenario, it becomes clear how selection of lysosomes as the intracellular vesicles responsible for membrane repair (12) would have provided an additional evolutionary advantage, by protecting eukaryotic cells from pathogen attack.

Supporting Online Material

www.sciencemag.org/cgi/content/full/304/5676/1515/DC1

Materials and Methods

Figs. S1 to S5

References

References and Notes

View Abstract

Navigate This Article