Published online before print July 15, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Divisions of Medical Microbiology, IMK, and
Cell Biology, IBK, Linköping University, Sweden;
Drug Discovery Program, H. Lee Moffitt Cancer Center & Research Institute, Department of Oncology, University of South Florida, Tampa; and
Department of Chemistry, Yale University, New Haven, Connecticut
1 Correspondence: Division of Medical Microbiology, IMK, Linköping University, SE-581 85, Linköping, Sweden. E-mail: limzh{at}imk.liu.se
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
70%. Transduction of Tat fusion constructs containing dominant-negative Cdc42 or Rac1 significantly inhibited Salmonella-induced cell death, indicating that the cytotoxicity of Salmonella requires activation of Cdc42 and Rac. In contrast to phagocytosis of opsonized bacteria, invasion by S. typhimurium stimulated Cdc42 and Rac1, regardless of the activities of tyrosine- or PI-3K. Moreover, Salmonella infection activated Akt protein in a tyrosine-kinase or PI-3K-dependent manner, and a reduced expression of Akt by antisense transfection rendered the cells more sensitive to apoptosis induced by opsonized Salmonella. These results indicate that direct activation of Cdc42 and Rac1 by invasive Salmonella is a prerequisite of Salmonella-mediated death of U937 cells, whereas the simultaneous activation of Akt by tyrosine kinase and PI-3K during receptor-mediated phagocytosis protects cells from apoptosis.
Key Words: macrophages • bacterial apoptosis • signal transduction
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Invasive Salmonella strains not only survive inside macrophages, they also induce apoptosis in the phagocytes in vitro and in vivo [6 7 8 9 ]. A functional type III secretion system located in Salmonella pathogenicity islands 1 and 2, designated SPI-1 and SPI-2, is essential for the induction of macrophage apoptosis [10 11 12 13 14 15 16 ]. Moreover, it has been shown that the SPI-1-encoded SipB protein binds directly to and activates caspase-1, which leads to apoptosis and release of proinflammatory cytokines in phagocytes [14 , 17 ]. Also, in mice, activation of caspase-1 and subsequent induction of macrophage apoptosis are necessary for efficient colonization of the lymph nodes, spleen, and liver by S. typhimurium [18 ]. Besides these bacterial proteins, another important factor affecting Salmonella-mediated macrophage apoptosis is the species of the host. For example, apoptosis and intracellular growth of S. typhimurium are much more pronounced in murine than in human macrophages [15 , 19 , 20 ], which may reflect the ability of this pathogen to cause different diseases in mice and human beings. At present, the molecular mechanisms by which S. typhimurium regulates apoptosis in macrophages have been studied more extensively in mice than in humans.
Apoptosis is mediated by a cell-intrinsic suicide program, regulation of which represents the net outcome of the relative balance between simultaneously activated pro- and antiapoptotic pathways within the cells [21 ]. Infection with Salmonella leads to direct activation of caspase-1 by SipB, and it also stimulates a variety of signaling pathways in the host cell, including Rho–GTPases, c-jun NH2-terminal kinase (JNK), p38, protein kinase B/Akt, and Raf-1, which are the key regulators of apoptosis [22 23 24 25 26 ]. In addition, numerous genes that are up-regulated in human macrophages upon infection with wild-type S. typhimurium are known to be involved in the death and/or survival of the phagocytes [27 ]. These findings suggest that the pro- and antiapoptotic signaling pathways activated during Salmonella infection play an important role in determining the fate of macrophages. In that context, in a recent study of Salmonella-infected macrophages, it was found that activation of Raf-1 plays a protective role in antagonizing caspase-1 activation during pathogen-induced apoptosis [24 ].
It has been shown that invasive S. typhimurium can trigger apoptosis in murine macrophages, regardless of their ability to replicate within the cells [7 ], which indicates that the initial interaction between the invasive bacteria and the macrophages generates a signal(s) that is essential for induction of apoptosis. Studying S. typhimurium and human monocytic U937 cells, we found that the bacteria induced apoptosis when they invaded these phagocytes but not when they were internalized through receptor-mediated phagocytosis. Also, we observed that invasion by S. typhimurium and the subsequent induction of apoptosis were mediated primarily by activation of Cdc42 and Rac1, which did not depend on the activities of tyrosine kinases and phosphatidylinositol-3 kinase (PI-3K). In contrast, phagocytic uptake of opsonized Salmonella mutants stimulates Cdc42 and Rac in a manner that did require tyrosine kinases or PI-3K. Moreover, our finding that phagocytosis had no influence on apoptosis appeared to be at least partly a result of the simultaneous activation of survival signal Akt by tyrosine kinases and PI-3K in these cells.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cell culture and treatment with inhibitors
Human monocytic U937 cells were used in all experiments, and they were grown in RPMI-1640 medium supplemented with 10% fetal calf serum, 2 mM L-glutamine, 100 U/ml penicillin, and 100 µg streptomycin/ml (RPMI medium), incubating at 37°C in a humidified air containing 5% CO2. Before infection with S. typhimurium, the cells were washed, resuspended in RPMI medium without the antibiotic, and incubated in tissue-culture plates for 2 h at 37°C.
U937 cells were preincubated at 37°C with 1.5 µM herbimycin A for 3 h or 50 µM genistein for 15 min to inhibit tyrosine kinase activity or with 50 nM wortmannin for 15 min to inhibit PI-3K activity. Likewise, cells were pretreated for 40 h with 5 µM GGTI-298 to inhibit protein geranylgeranylation or with 5 µM FTI-277 to block farnesylation [28 ]. These inhibitors (at the indicated concentrations) were also present throughout the infection period.
Bacteria strains and growth conditions
The invasive wild-type S. typhimurium strain SL1344 and the isogenic mutants hilA- (VV341), prgH- (EE656), and sipC- (EE638) were kindly provided by Dr. Catherine A. Lee (Harvard Medical School, Boston, MA; refs. [29
, 30
]). The bacteria were grown under agitation in 2 ml Luria-Bertani (LB) broth overnight at 37°C. Thereafter, the cultures were diluted 40 times with fresh LB medium and shaken for an additional 3.5 h at 37°C (optical density600: 0.30–0.35). The antibiotics used were kanamycin (50 µg/ml, VV341) and tetracycline (25 µg/ml, EE656 and EE638). The bacteria culture was washed and resuspended in RPMI medium without antibiotics. Opsonized bacteria were provided by using 20% normal human serum, as previous described [31
].
Bacterial infection and invasion assay
U937 cells in multiwell tissue-culture plates were incubated with different strains of S. typhimurium (multiplicity of infection, 20:1) for 30 min at 37°C, washed twice, and then incubated in RPMI medium supplemented with 50 µg/ml gentamicin.
The ability of S. typhimurium to enter U937 cells was determined by the gentamicin-protection assay, as described previously [32 ]. In short, U937 cells seeded on 24-well tissue-culture plates were exposed to bacteria. Thereafter, the cells were washed and incubated with 50 µg/ml gentamicin for 1 h at 37°C and subsequently washed with phosphate-buffered saline (PBS) and lysed in 0.2 ml 1% Triton X-100 in PBS for 10 min. Samples were vigorously mixed with 0.8 ml LB broth, and viable bacteria released from lysed macrophages were quantified by culturing serial dilutions of the mixtures on LB agar plates and then by performing colony-forming unit (CFU) counts. In some cases, the cells were pretreated with inhibitors as described above. To assess the intracellular growth of bacteria, the medium containing 10 µg/ml gentamicin was used.
Flow cytometric analysis of apoptosis
Staining U937 cells with fluorescein isothiocyanate (FITC)-conjugated Annexin V identified early apoptotic changes, according to the protocol provided by the manufacturer. Specific binding of Annexin V was achieved by incubating the cells (5x105) in 60 µl binding buffer saturated with Annexin V for 15 min at 4°C in the dark. To assess the plasma membrane integrity, the cells were simultaneously stained with propidium iodide before analysis. Binding of Annexin V (FL1) and propidium iodide (FL2) to the cells was measured by flow cytometry (FACS Calibur, BD Biosciences, San Jose, CA) using Cell Quest software [33
]. At least 10,000 cells were counted in each sample, and a gate based on forward and side scatters was set to exclude cell debris.
DNA fragmentation assay
U937 cells (106/sample) were lysed, and the genomic DNA was extracted according to the protocol for the apoptotic DNA laddering analysis kit. Gel electrophoresis (1.8% agarose) and ethidium bromide staining analyzed the samples (2 µg DNA/lane). The gel was visually examined under 305 nm UV illumination [33
].
Cell lysis, affinity precipitation, and Western blot analysis
Activation of Rac1 and Cdc42 in U937 cells was determined as described previously [34
, 35
] with minor modifications. A plasmid encoding the p21-binding domain (PBD) of human p21-activated kinase (PAK)1 as a fusion protein with glutathione S-transferase (GST; GST–PBD) was a generous gift from Dr. Gary M. Bokoch (Scripps Research Institue, La Jolla, CA). U937 cells (2x106/sample) in six-well tissue-culture plates were serum-starved for 3 h and were then infected with Salmonella for the indicated periods of time. The cells were washed with ice-cold PBS, lysed by phosphorylation lysis buffer [36
], and cleared by centrifugation. Colorimetric analysis determined protein concentrations of the lysates using DC protein assay reagents (Pierce, Rockford, IL). Equal amounts of cellular proteins were incubated with GST–PBD prebound to glutathione-Sepharose beads at 4°C for 40 min under rotation. The proteins on the beads were eluted with Laemmli buffer [37
], separated by 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and analyzed by Western blotting with Ab against Rac1 and Cdc42. The specificity of this assay was confirmed by adding 1 mM EDTA and 100 µM guanosine 5'-O-(thio)triphosphate to the lysate as a positive control and using GST beads (i.e., beads coupled to GST without PBD of PAK1) as negative control (data not shown).
To prepare for Rac and Cdc42 processing assay, the cells were exposed to medium alone, 5 µM GGTI-298, or 5 µM FTI-277 for 40 h and were then lysed as described above. Equal amounts of cellular proteins were separated by 12.5% SDS-PAGE and analyzed by Western blotting with Ab specific for Rac1 and Cdc42.
Transduction of Tat fusion proteins into U937 cells
Human immunodeficiency virus (HIV) Tat-mediated delivery of fused proteins into cells was performed as described [38
]. In short, cDNAs encoding dominant-negative Rac1 (Rac1N17), Rac2 (Rac2N17), and Cdc42 (Cdc42N17) were cloned into the bacterial expression vector pTat-HA (gift from Dr. Steven F. Dowdy; see ref. [38
]). Expression of Tat fusion proteins in transformed Escherichia coli was induced by isopropylthiogalactoside, and Tat fusion proteins were purified on a Ni-NTA column after sonication of the bacteria in buffer Z [38
]. The eluted proteins were desalted on a PD-10 column into 50 mM Tris-HCl (pH 7.4), frozen in 10% glycerol, and stored at -80°C. The Tat fusion proteins were analyzed by Coomassie blue staining of the gel and by immunoblotting with antibodies against HA, Cdc42, and Rac1 [38
].
U937 cells were pretreated with Tat fusion proteins (200 nM) at 37°C for 45 min and were then infected with Salmonella; Tat fusion proteins were also present in the incubation medium during the infection period to prevent the Tat proteins from diffusing out of the phagocytes.
Assays for Akt activation
Activation of Akt in U937 cells was determined by detection of phosphorylated Akt and by in vitro kinase assay. Cells were serum-starved for 3 h and were then exposed to S. typhimurium for the indicated periods of time. To detect phosphorylation of Akt, the cells were lysed as described above, and equal amounts of cellular proteins were separated by 10% SDS-PAGE and analyzed by Western blotting with a phospho-specific (Ser473), anti-Akt Ab. To confirm that each lane had received the same amount of proteins, the blots were stripped and reprobed with an anti-Akt Ab, which recognized phosphorylated and nonphosphorylated Akt.
Akt kinase assays were performed essentially according to the protocol provided by the manufacturer. After stimulation, U937 cells (4x106/sample) were washed and lysed for 20 min. The lysates were cleared and precipitated with immobilized Akt Ab at 4°C for 2 h under rotation, and the precipitates were washed three times with lysis buffer and twice with kinase buffer. The kinase reaction was allowed to proceed at 30°C for 30 min in kinase buffer supplemented with 200 µM adenosine 5'-triphosphate and 1 µg glycogen synthase kinase-3 (GSK-3)
/ß fusion protein. Adding 3x Laemmli buffer and heating terminated the reaction. Kinase products were subjected to 12% SDS-PAGE and blotted with an antiphospho-GSK-3/ß (Ser21/9) Ab.
To compare the extent of Rac/Cdc42 and Akt activation and thereby, to uncover the relative balance between these signaling proteins, the blots were scanned, and the relative intensity of the bands was quantified using NIH image (v.1.6.1). For comparison, the value in unstimulated cells on each blot was set to 1.
Transfection of U937 cells with antisense RNA expression vectors for Akt
The cDNAs for human Akt1, Akt2, and Akt3, which contain the entire coding sequences, were cloned in its antisense orientation into the mammalian expression vector pCR3; Dr. Hiroyasu Esumi (National Cancer Center Research Institute East, Chiba, Japan; ref. [39
]) kindly provided these constructs. An EndoFree plasmid maxi kit (Qiagen, Valencia, CA) was used to purify the plasmid DNAs. Control vector or antisense RNA expression vectors (25 µg) were transfected into 107 U937 cells by electroporation using a Gene-Pulser (Bio-Rad, Hercules, CA). After the recovery culture of 48 h, transfectants were selected by treatment with 0.5 mg/ml G418. G418-resistant clones were pooled after 3 weeks of selection and were then grown as a mixture that was used within 3 weeks.
Statistical analysis
Data shown represent means ± SD, and the statistical significance was determined by Student’s t-test. P < 0.05 was considered to be statistically significant.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
 View larger version (30K): [in a new window] |
Figure 1. Invasion by S. typhimurium induces apoptosis in U937 cells, which were left untreated (A and E) or were infected with wild-type S. typhimurium SL1344 (B and F) or serum-opsonized hilA- mutants (C and G) for 30 min and were then cultured for 18 h. Necrotic cell death was induced by heating at 55°C for 15 min (D and H). Changes in cell size and granularity were evaluated by flow cytometry with respect to forward-angle (FSC) and side-angle (SSC) light scatter (A–D). A subpopulation of cells infected with the SL1344 strain exhibited decreased cell volume and increased granularity (B). Staining with propidium iodide (FL2) and FITC-conjugated Annexin V (FL1), respectively, determined membrane integrity and phosphatidylserine exposure (E–H). The relative distribution within each population of cells with early (Annexin+propidium iodide-) and late (Annexin+propidium iodide+) manifestations of apoptosis is indicated (%). For clarity, each profile shows only 10% of the analyzed events.
|
 View larger version (32K): [in a new window] |
Figure 2. Invasion by, but not receptor-mediated phagocytosis of, S. typhimurium triggers the death of U937 cells. (A) The cells were left untreated (dashed line) or infected with wild-type S. typhimurium SL1344 ( ) or noninvasive hilA- ( ), prgH- (•), and SipC- ( ) mutants for 30 min, were washed, and were then cultured for the indicated periods of time. The percentage of apoptosis was determined by Annexin V binding (n=4). (B) DNA fragmentation in U937 cells infected with S. typhimurium. The cells were left untreated (lane 1) or infected with SL1344 (lane 2) or hilA- (lane 3), prgH- (lane 4), and SipC- (lane 5) and were then cultured for 18 h. The results are representative of four separate experiments. (C) The number of intracellular bacteria (open bars) was determined at 1 h postinfection by gentamicin-protection assay and expressed as values x105 CFU/106 U937 cells. The percentage of dead cells was determined at 18 h (solid bars). Each value represents the means ± SD of four independent experiments performed in parallel. ops., Opsonized.
|
To ascertain whether the differential effects of invasive and noninvasive Salmonella on apoptosis can be ascribed to varying numbers of intracellular bacteria, we used serum-opsonized bacteria, which can enter macrophages by receptor-mediated phagocytosis [7 , 40 ]. Opsonized, noninvasive mutants did enter the cells more efficiently but had only a minimal effect on apoptosis (shown for opsonized hilA- in Fig. 1C and 1G , and Fig. 2C ). Moreover, exposing the macrophages to serum-opsonized, invasive bacteria resulted in an approximately twofold increase in the number of intracellular bacteria, but it did not magnify apoptosis (32±6% and 34±5% in cells infected with SL1344 and opsonized SL1344, respectively; n=3). It should be noted that 18 h after infection, the number of intracellular bacteria did not differ between macrophages infected with SL1344 strain and those with opsonized hilA- bacteria (data not shown). These results clearly indicate that S. typhimurium entering U937 cells by invasion mechanism induces apoptosis, whereas receptor-mediated phagocytosis of these bacteria had no such effect.
Cdc42 and Rac1, but not tyrosine kinases or PI-3K, are required for Salmonella-induced apoptosis in U937 cells
Salmonella infections are known to activate a variety of host cell signaling pathways, including Rho–GTPases, tyrosine kinase, PI-3K, and Akt [22
23
24
25
, 41
, 42
], which are well-established regulators of apoptosis [21
, 43
44
45
]. Therefore, we investigated which of these signals is involved in Salmonella-induced apoptosis. We found that herbimycin A, a tyrosine kinase inhibitor, did not affect invasion by the bacteria and induction of apoptosis. Moreover, the PI-3K inhibitor wortmannin caused only a marginal decrease in invasion by Salmonella but provoked a slight increase in apoptosis in the infected cells (Fig. 3
). In contrast, both of these inhibitors effectively blocked the phagocytosis of opsonized hilA- bacteria (Fig. 3)
. However, invasion by the SL1344 strain and subsequent apoptosis were markedly curbed by pretreatment with GGTI-298, an inhibitor of protein geranylgeranylation, but not by FTI-277, an inhibitor of protein farnesylation (Fig. 4A
). GGTI-298 also blocked the phagocytosis of opsonized hilA- bacteria (data not shown). These results indicate that a prerequisite of Salmonella-induced apoptosis in U937 cells is that geranylgeranylated proteins, but not tyrosine kinases or PI-3K, are activated.
 View larger version (16K): [in a new window] |
Figure 3. Invasion of U937 cells by S. typhimurium and apoptosis induced by that event do not require the activation of tyrosine kinases or PI-3K. U937 cells were pretreated with an inhibitor of tyrosine kinases (herbimycin A, closed bars), with PI-3K (wortmannin, hatched bars), or with medium alone (open bars) and were then infected with SL1344 strain or serum-opsonized hilA- (ops. hilA–) mutant. The number of intracellular bacteria (A) and the percentage of dead cells (B) were determined as described in Figure 2C
. Each value represents the means ± SD of five separate experiments.
|
 View larger version (34K): [in a new window] |
Figure 4. The activities of Cdc42 and Rac1 are required for Salmonella-induced apoptosis in U937 cells. (A) The cells were pretreated with 0.1% dimethyl sulfoxide (open bars), 5 µM GGTI-298 (hatched bars), or FTI-277 (closed bars) for 40 h and were then infected with wild-type SL1344 bacteria. The number of intracellular bacteria and percentage of dead cells were determined as described in Figure 2C
(n=5). *, Significantly different from the control cells (P<0.01). (B) Cells were pretreated with medium or 5 µM GGTI-298 or FTI-277 for 40 h and were then lysed. Cellular proteins were subjected to Western blot analysis for Rac1 or Cdc42 to demonstrate inhibition of processing, detected as a band shift from the processed (p) to the unprocessed (u) proteins. The illustrated blots are representative of five separate experiments. (C) The cells were preincubated for 45 min at 37°C with the indicated Tat fusion proteins (200 nM) or with medium alone and were then infected with wild-type SL1344 strain for 30 min. Each value represents the means ± SD of four experiments performed in parallel and in duplicate. *, Significantly different from the cells incubated with medium alone (P<0.05); **, significantly different from the cells incubated with Cdc42N17 or Rac1N17 alone (P<0.05).
|
By using HIV Tat-mediated delivery of mutated proteins [38
], we found that transduction of Tat–Cdc42N17 into these phagocytes inhibited Salmonella invasion and subsequent apoptosis by 35%, and the same effect, albeit not as pronounced, was seen for Tat–Rac1N17 (Fig. 4C)
. Furthermore, a combination of Tat–Cdc42N17 and Tat–Rac1N17 inhibited Salmonella-induced cell death by 55% (Fig. 4C)
. Tat–Rac2N17 had no significant effect, which reflects the fact that macrophages primarily express the Rac1 isoform [40
]. The concentrations of Tat fusion proteins used in our study did not affect the viability of uninfected cells (data not shown). Together, these results indicate that activation of Cdc42 and (to a lesser degree) Rac1 is required for Salmonella invasion and subsequent apoptosis in U937 cells.
Invasion by Salmonella and phagocytosis of these bacteria activates Rac and Cdc42 by different mechanisms
Infection with SL1344 or opsonized hilA- bacteria significantly increased the level of activated Rac1 and Cdc42, and the same effect, although not as pronounced, was exerted by lipopolysaccharides (LPS) isolated from Salmonella and nonopsonized hilA- mutant (Fig. 5A
). Neither herbimycin nor wortmannin had an impact on activation of Rac1 and Cdc42 by invasive S. typhimurium. In contrast, stimulation of Rac1 and Cdc42 by opsonized hilA- bacteria was almost completely blocked by herbimycin, and this Rac activation was also largely prevented by wortmannin (Fig. 5B) .
 View larger version (19K): [in a new window] |
Figure 5. Invasion by and receptor-mediated phagocytosis of S. typhimurium activate Rac1 and Cdc42 in U937 cells by different mechanisms. (A) U937 cells were exposed to wild-type SL1344, hilA–, or serum-opsonized hilA- (ops. hilA-) bacteria or to Salmonella LPS (100 ng/ml) for the indicated times and were then lysed. The activated Rac1 and Cdc42 released from the lysed cells were collected by affinity precipitation with PBD–GST followed by Western blot analysis for Rac1 (upper panels) or Cdc42 (lower panels). (B) Cells were preincubated with 1.5 µM herbimycin A (herb.) for 3 h, 50 nM wortmannin (wort.) for 15 min, or medium alone and were subsequently infected with Salmonella. The illustrated blots are representative of five separate experiments.
|
 View larger version (30K): [in a new window] |
Figure 6. Activation of Akt in U937 cells infected with Salmonella. (A) Serum-starved cells were stimulated with the wild-type SL1344, hilA-, or serum-opsonized hilA- (ops. hilA-) bacteria for the indicated times and were then lysed. Equal amounts of cellular proteins were separated by 10% SDS-PAGE and blotted with an antiphospho-Akt Ab (p-Akt; upper panel). To confirm that each lane had received similar amounts of Akt, the blot was reprobed with antitotal-Akt Ab (lower panel). (B) Cells were pretreated with 50 µM genistein (gen.), 50 nM wortmannin (wort.), or medium alone for 15 min and were then infected with Salmonella. (C) Akt was immunoprecipitated, and its activity was determined by an in vitro kinase assay using GSK-3 fusion protein as substrate. Phosphorylation of GSK-3/ß was analyzed by Western blotting with a specific Ab. The illustrated blots are representative of three to five separate experiments.
|
 View larger version (9K): [in a new window] |
Figure 7. The relative ratio between Rac/Cdc42 activity and Akt kinase activation in U937 cells infected with invasive Salmonella SL1344 () or opsonized mutant (•). The blots from Figure 5A
and Figure 6C
were scanned, and the relative intensity of the bands was quantified. Data are given as mean ± SEM and are expressed in relation to unstimulated cells set to one in each experiment. **, Significantly different from the cells infected with opsonized mutant (P<0.01).
|
 View larger version (22K): [in a new window] |
Figure 8. Akt is involved in protecting cells from apoptosis during receptor-mediated phagocytosis of Salmonella. (A) U937 cells were transfected with the control vector PCR3 or with antisense RNA expression vectors for Akt1, Akt2, and Akt3. After selection with G418 for 3 weeks, the transfected cells were left untreated or were infected with S. typhimurium SL1344 or serum-opsonized hilA- (ops. hil A-) bacteria for 30 min and were then cultured for 18 h. The percentage of apoptotic cells was determined as described in Figure 1
, and the results are expressed as means ± SD of three experiments. *, Significantly different from cells transfected with the control vector PCR3 (P<0.05). (B) Western blot analysis of Akt proteins in transfected cells (upper panel). The blot was stripped and reprobed with antiactin Ab to confirm equal loading of proteins in each lane (lower panel).
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Pro- and antiapoptotic-signaling pathways strictly regulate the apoptotic process, and the relative balance between these opposing modulatory routes determines cell fate [21 ]. Infection with S. typhimurium stimulates numerous host cell-signaling pathways, which are known to be involved in cell death and/or survival [22 23 24 25 , 45 , 50 ]. Using Tat-mediated delivery of dominant-negative proteins, we found that Salmonella-induced apoptosis was suppressed in U937 cells transduced with Cdc42N17 and to a lesser extent, Rac1N17. This agrees with the previous finding that Cdc42 and (less prominently) Rac1 are required for invasion by Salmonella and activation of JNK in COS-1 cells [48 ]. Combining Cdc42N17 and Rac1N17 led to an additive inhibitory effect, which implies that the cytotoxic effects of S. typhimurium require activation of Cdc42 and Rac1. This is supported by our observations that inactivation of Cdc42 and Rac by GGTI-298 markedly suppressed the Salmonella-induced apoptosis, and activation of Cdc42 and Rac by invasive Salmonella did not involve tyrosine kinases and PI-3K. Conversely, phagocytosis of opsonized bacteria, which did not induce apoptosis, did require tyrosine kinases and PI-3K for activation of Cdc42 and Rac. These results imply that a survival signal(s) that is coactivated by these protein kinases during phagocytosis protects U937 cells from apoptosis.
Activation of PI-3K/Akt is vital for the survival of human macrophages, and inhibition of Akt induces apoptosis [49 , 50 ]. In that context, we found that infection of U937 cells with opsonized Salmonella caused tyrosine kinase- and PI-3K-dependent increases in the phosphorylation and enzymatic activity of Akt. The enzymatic activity of Akt was significantly higher in the cells after infection with opsonized bacteria. Moreover, decreased Akt protein levels rendered antisense-transfected cells more sensitive to apoptosis induced by opsonized Salmonella. These results indicate that stimulation of Akt antagonizes the simultaneously activated proapoptotic signals and thus protects the cells from apoptosis during phagocytosis. In addition, pathogens (e.g., LAM from Mycobateria) and phagocytosis of apoptotic cells have been shown to inhibit the macrophage apoptosis via the PI-3K/Akt pathway [51 , 52 ]. Thus, Akt can function as general mediators of survival signals in various systems, which is a bacterial-independent pathway.
Macrophages have acquired the ability to recognize conserved bacterial components (e.g., LPS) and to subsequently stimulate multiple cellular signaling pathways [40 , 53 , 54 ]. Therefore, invasive Salmonella can activate macrophage by conserved bacterial components as well as by the invasion process. Presumably, the components trigger pro- and antiapoptotic signals (e.g., Cdc42/Rac and Akt, respectively) simultaneously, whereas invasion selectively activates proapoptotic signals and thereby shifts the balance toward cell death. This mechanism might explain our finding that invasive Salmonella could induce apoptosis in U937 cells, although it activated Akt. This hypothesis is supported by the following three observations: First, the noninvasive hilA- mutant and Salmonella LPS did not induce apoptosis in U937 cells, whereas they did activate Cdc42/Rac and Akt (Figs. 5 and 6) , which is compatible with the mentioned triggering of pro- and antiapoptotic signals by the conserved bacterial constituents [53 , 54 ]. Second, herbimycin and wortmannin inhibited invasive Salmonella-induced activation of Akt but not Cdc42 and Rac, indicating that these signals are stimulated by different mechanisms or bacterial components. Third, in cells with reduced expression or inhibited activity of Akt, we observed a reproducible, moderate increase in the rate of apoptosis after infection with invasive Salmonella (Figs. 3 and 8) . This shows that activation of Akt can also protect host cells from apoptosis after bacteria invasion, although this effect is much less pronounced than that exerted by phagocytosis.
It is not yet known how activation of Rho–GTPases leads to apoptosis. Our data show that Rho–GTPase-dependent invasion is essential for apoptosis but whether Rac1 and Cdc42 trigger apoptosis per se is not proven. Nevertheless, it has been shown that these proteins are essential for the morphological changes in apoptosis and that point mutations engineered to block Rac-induced actin polymerization also reduced its ability to cause apoptosis [43 , 55 ]. Furthermore, constitutively activated Cdc42 has been found to induce apoptosis in T cells via the JNK pathway [45 ]. These results indicate that Cdc42 and Rac not only control the cytoskeleton, but they also serve as additional regulatory components in signaling pathways that trigger apoptosis [43 ]. It is interesting that studies have shown that activation of Cdc42 and Rac by invasive S. typhimurium is essential for invasion of the bacteria and for induction of nuclear responses in host cells [23 , 41 , 48 ].
In studies of murine macrophage cell lines [5 , 7 , 16 , 56 , 57 ], invasive S. typhimurium was found to induce cell death within 2–4 h of infection. However, in our experiments, Salmonella-induced apoptosis occurred in human monocytic U937 cells after 9–18 h of incubation, and the rate of apoptosis was much lower than that seen in murine macrophages. The differential cytotoxicity to human and mouse macrophages may reflect the ability of S. typhimurium to cause disparate diseases in these two vertebrate species: Rapid intracellular growth and induction of macrophage apoptosis can lead to a highly virulent systemic infection in mice [18 19 20 ], whereas in humans, the bacteria do not survive as well in macrophages and also show less cytotoxicity (refs. [19 , 20 ]; and present study).
A recent study using microarray analysis showed that several genes related to cell death were more abundant in U937 cells infected with wild-type Salmonella than in those infected with avirulent mutants, which correlated with the ability of bacteria to kill the macrophages [27 ]. Invasion of macrophages by Salmonella activates Raf-1, which in turn, plays a protective role by antagonizing caspase-1 activation during apoptosis [24 ]. Together with our results, these findings show that selective regulation of host-cell death/survival signals is an important factor in Salmonella-induced cytotoxicity. The exact mechanisms affecting Salmonella-induced apoptosis in U937 cells are not yet clear, but as activation of Cdc42 and Rac by invasive Salmonella is mediated by SopE and SopB, these bacterial proteins are probably involved. However, we noted that inactivation of Cdc42 and Rac by GGTI-298 or dominant-negative mutants did not completely inhibit Salmonella-induced cell death, which indicates that other bacterial-effector proteins also participate in this process. It has been reported that the SipB protein and expression of spv genes in Salmonella are required for Salmonella-induced apoptosis in macrophages [14 , 17 , 20 ], and such a mechanism might also be responsible for the Salmonella-mediated death in U937 cells.
Invasion and receptor-mediated phagocytosis require actin polymerization, but they differ in regard to their underlying regulatory signals and the outcomes to which they lead, such as changes in morphology and the size of vacuole formed [3 , 40 ]. Our findings show that the different signals triggered during bacterial invasion and phagocytosis are also essential for regulation of apoptosis in phagocytic cells and that simultaneous activation of survival signals antagonizes the proapoptotic activities of pathogens. Microbe-induced apoptosis of phagocytes plays an important role in the bacterial pathogenesis, which may represent a pathogenic strategy to eliminate key immune cells and to initiate inflammation [58 , 59 ]. Therefore, in addition to the virulence factors produced by pathogens, factors such as the activation status of host cells and the route of bacterial entry, which might affect the apoptotic pathways, may have a profound effect on cell fate and the subsequent inflammatory responses.
Received December 1, 2002; revised April 17, 2003; accepted April 23, 2003.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2, Shc adapter protein and p38 MAPK J. Immunol. 164,959-965
B, Bad, or caspase activation J. Exp. Med. 194,113-126This article has been cited by other articles:
 |
|
 |
|
  O. S. Shin, L. S. Miller, R. L. Modlin, S. Akira, S. Uematsu, and L. T. Hu Downstream Signals for MyD88-Mediated Phagocytosis of Borrelia burgdorferi Can Be Initiated by TRIF and Are Dependent on PI3K J. Immunol., July 1, 2009; 183(1): 491 - 498. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Tamilselvam and S. Daefler Francisella Targets Cholesterol-Rich Host Cell Membrane Domains for Entry into Macrophages J. Immunol., June 15, 2008; 180(12): 8262 - 8271. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Blomgran, L. Zheng, and O. Stendahl Cathepsin-cleaved Bid promotes apoptosis in human neutrophils via oxidative stress-induced lysosomal membrane permeabilization J. Leukoc. Biol., May 1, 2007; 81(5): 1213 - 1223. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L.-A. H. Allen, J. A. Allgood, X. Han, and L. M. Wittine Phosphoinositide3-kinase regulates actin polymerization during delayed phagocytosis of Helicobacter pylori J. Leukoc. Biol., July 1, 2005; 78(1): 220 - 230. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Ohya, Y. Handa, M. Ogawa, M. Suzuki, and C. Sasakawa IpgB1 Is a Novel Shigella Effector Protein Involved in Bacterial Invasion of Host Cells: ITS ACTIVITY TO PROMOTE MEMBRANE RUFFLING VIA RAC1 AND CDC42 ACTIVATION J. Biol. Chem., June 24, 2005; 280(25): 24022 - 24034. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Blomgran, L. Zheng, and O. Stendahl Uropathogenic Escherichia coli Triggers Oxygen-Dependent Apoptosis in Human Neutrophils through the Cooperative Effect of Type 1 Fimbriae and Lipopolysaccharide Infect. Immun., August 1, 2004; 72(8): 4570 - 4578. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. Yilmaz, T. Jungas, P. Verbeke, and D. M. Ojcius Activation of the Phosphatidylinositol 3-Kinase/Akt Pathway Contributes to Survival of Primary Epithelial Cells Infected with the Periodontal Pathogen Porphyromonas gingivalis Infect. Immun., July 1, 2004; 72(7): 3743 - 3751. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
