INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Gram-negative bacteria of the genus Shigella cause bacillary dysentery, a potentially fatal diarrheal disease. In the lower gastrointestinal tract, Shigella provokes an acute inflammation that leads to tissue destruction and promotes bacterial invasion (15). Early during infection, Shigella is phagocytosed by resident tissue macrophages. Shigella escapes the phagolysosome and kills macrophages in vitro (18, 43) and in vivo (45) by inducing apoptosis. Apoptotic macrophages release mature interleukin-1 (IL-1) and IL-18, cytokines which trigger the acute inflammation characteristic of dysentery (36, 44). Bacterial virulence factors are encoded on a 220-kb plasmid and include a type III secretion apparatus and the invasion plasmid antigens (Ipa) A to D (17, 25).

The invasin IpaB is secreted by Shigella into the macrophage cytoplasm (37). IpaB is necessary and sufficient to induce apoptosis (6, 42) and binds specifically to caspase-1 (IL-1 converting enzyme; ICE) (19). Caspases are cysteine proteases that cleave after aspartic acid residues. These enzymes are synthesized as zymogens and are activated by proapoptotic stimuli (39). In macrophages infected with Shigella, pro-caspase-1 is proteolytically processed to its mature form which cleaves pro-IL-1 and activates apoptosis (19).

In addition to caspases, other endogenous proteases have been implicated in apoptosis. The eukaryotic 26S proteasome is an N-terminal threonine protease complex with multiple endoproteolytic activities (trypsin-, chymotrypsin-, and caspase-like). The proteasome is required for the cell's general ATP- and ubiquitin-dependent protein turnover, cell cycle progression, and immunosurveillance (3, 31). Moreover, the proteasome participates in apoptotic pathways triggered by growth factor deprivation, radiation, and chemicals (9, 13, 27).

Another high-molecular-weight protease is the cytoplasmic serine protease tripeptidyl peptidase II (TPPII) (40). TPPII is an amino-endopeptidase and exists as a homo-oligomer in the cytoplasm (1, 2, 11) or as a membrane-bound complex (33). Tripeptidyl peptidases are involved in general protein degradation (40), but TPPII also cleaves specific substrates such as neuropeptides (33). In lymphoma cells adapted to proteasome inhibitors, a proteolytic system with TPPII-like activity compensates for the loss of proteasome activity (12), and TPPII substitutes some proteasomal functions (11). Both the proteasome and TPPII share a similar overall structure, as determined by electron microscopy (3, 11, 24).

TPPII and the proteasome differ in their substrate spectra and inhibition patterns. The Streptomyces metabolite lactacystin and its active, more-cell-permeable form, clasto-lactacystin -lactone (lactacystin), inhibit the proteasome (8, 10) and, less efficiently, TPPII (11). Ala-Ala-Phe-chloromethylketone (AAF-cmk) irreversibly blocks purified TPPII but not the proteasome. Conversely, the proteasome inhibitor N-acetyl-Leu-Leu-norleucinal (LLnL) does not inhibit TPPII (11). LLnL blocks the proteasome more potently than N-acetyl-Leu-Leu-methioninal (LLM) (32). Finally, 4-hydroxy-5-iodo-3-nitrophenylacetyl-Leu-Leu-Leu-vinylsulfone (LLL-vs) inhibits the trypsin-, chymotrypsin-, and caspase-like proteasomal activities irreversibly (5). Using these inhibitors, we demonstrate here that TPPII participates in Shigella-induced apoptosis upstream of caspase-1 and in the apoptotic pathways triggered by staurosporine and ATP.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Cells, bacteria, and reagents. Mouse J774 macrophage-like cells, mouse peritoneal macrophages (from B57BL/6 or Swiss Webster mice), and human THP.1 monocytic leukemia cells (American Type Culture Collection [ATCC]) were used in this study. Macrophages were routinely cultured in a humidified atmosphere of 5% CO₂ at 37°C in RPMI 1640 glutamine medium (Mediatech; Fisher Scientific, Pittsburgh, Pa.) supplemented with 10% fetal calf serum (Gibco-BRL, Gaithersburg, Md.), L-glutamine (2 mM), and penicillin and streptomycin (50 µg/ml each). For infections, J774 cells and peritoneal macrophages were cultured overnight. THP.1 cells were differentiated for 3 days in the presence of 160 nM phorbol 12-myristate 13-acetate (PMA). To induce pro-IL-1, the macrophages were treated overnight with 1 µg of lipopolysaccharide (LPS; Shigella serotype 1A; Sigma, St. Louis, Mo.) per ml.

Cytotoxicity assays. Cytotoxicity induced by Shigella (multiplicity of infection [MOI] of 10 to 50; 2 to 3 h) (7), neutralized ATP (5 mM, 30 min) (21), or staurosporine (1 µM, 12 h) was routinely analyzed with the CytoTox96 Kit (Promega, Madison, Wis.) as described earlier (46). The percentage of cytotoxicity was calculated by quantifying the amount of lactate dehydrogenase (LDH) released from dying cells according to the following formula: [(experimental release  spontaneous release)/(total release  spontaneous release)] × 100 (19). LDH release is not specific for apoptosis; however, Shigella (43), ATP (21), and staurosporine (30) represent well-established apoptotic systems. Shigella-induced apoptosis of differentiated THP.1 cells was assayed by using the Cell Death Detection ELISA^PLUS Kit (Boehringer Mannheim, Indianapolis, Ind.) according to the manufacturer's instructions. Prior to the addition of the apoptotic agent, the macrophages were incubated with proteasome inhibitors (2 to 3 h) or caspase inhibitors (1 h) when indicated. The viability of macrophages treated with the inhibitors alone was assayed by LDH release, enzyme-linked immunosorbent assay (ELISA), and oxidation of the tetrazolium salt MTT [3,(4,5-dimethylthiazol-2-yl)2,5-diphenyl-tetrazolium bromide] (16). Within the time frame of the Shigella infections and ATP-induced apoptosis, the inhibitors alone were not cytotoxic for the macrophages, and the toxicity of 5% DMSO was less than 4% (data not shown).

Maturation of caspase-1 and IL-1 in macrophages infected with Shigella. Peritoneal macrophages from B57BL/6 mice were infected with Shigella (MOI of 5) and lysed in situ at the indicated time points, and the maturation of caspase-1 was analyzed by Western blotting using the rabbit anti-mouse caspase-1 antibody R10311 (kindly provided by D. Miller, Merck) as described previously (18, 19). When indicated, the macrophages were preincubated with AAF-cmk (10 µM, 2 h), lactacystin (30 µM, 4 h), or YVAD-cmk (100 µM, 1 h).

Gentamicin protection assay. To discriminate between phagocytosed Shigella and extracellular Shigella, a modification of the gentamicin protection assay was used (26). At 20 min postinfection, J774 macrophages infected with Shigella (MOI of 50) were washed twice with 0.2 ml of phosphate-buffered saline (PBS)-gentamicin (0.1 mg/ml), incubated for another hour in SFM-gentamicin (0.1 mg/ml), and lysed with PBS-Triton X-100 (0.1%). The lysate was diluted and plated on nutrient broth agar (Difco Laboratories, Detroit, Mich.) supplemented with 0.5% sodium chloride, and the CFU were counted. When indicated, the macrophages were treated with either AAF-cmk (1 µM, 1 h), lactacystin (5 µM, 2 h), YVAD-cmk (100 µM, 1 h), zVAD-fmk (100 µM, 1 h), or cytochalasin B (2.5 µg/ml, 1 h) prior to infection.

Detection of actin tails in Shigella-infected macrophages. Shigella-infected J774 macrophages were labeled with fluorescein isothiocyanate (FITC)-conjugated phalloidin and propidium iodide as described for epithelial cells (4). Briefly, at 30 min postinfection (MOI of 20), the macrophages were fixed, treated with RNase (1 mg/ml, 1 h), and stained with phalloidin-FITC (50 µg/ml) and propidium iodide (5 µg/ml) for 45 min. Prior to the infection, some macrophages were incubated with either AAF-cmk (1 µM) or lactacystin (5 µM) for 2 h. The samples were analyzed with a Leica DMRBE confocal microscope equipped with an argon-krypton laser.

Fluorometric assays of TPPII and caspase-1. The hydrolytic activity of TPPII during Shigella infection was measured with the fluorogenic peptide substrate Ala-Ala-Phe-7-amido-4-methyl-coumarin (AAF-amc). J774 cells were cultured overnight and infected with Shigella (MOI of 10) in PBS as described above. At different time points, the cells were lysed with digitonin (final concentration, 20 µg/ml), AAF-amc (100 µM) was added, and the reaction was stopped after 30 min with trifluoroacetic acid (final concentration, 0.2%). Prior to analyzing the hydrolysis of AAF-amc, some macrophages were treated for 2 h with different concentrations of AAF-cmk or lactacystin in PBS. The fluorescence of the leaving group 7-amino-4-methyl-coumarin (amc) was measured at 380-nm excitation and 460-nm emission wavelengths with a MicroMax-2 spectrofluorometer (Instruments SA, Edison, N.J.).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of TPPII and proteasome inhibitors on Shigella-induced macrophage apoptosis. To identify novel components of the apoptotic pathway triggered by S. flexneri, we treated J774 macrophages with various inhibitors of TPPII and the proteasome. The TPPII inhibitor AAF-cmk protected macrophages from the Shigella wild-type strain M90T in a dose-dependent manner by more than 95% at concentrations of as low as 1 µM (Fig. 1A). clasto-Lactacystin -lactone (referred to as "lactacystin") blocked Shigella-induced macrophage cell death by approximately 90% at concentrations of between 1 and 10 µM. In contrast different proteasome inhibitors at concentrations as high as 100 µM had little (LLM and LLL-vs) or no (LLnL) effect on Shigella-induced apoptosis. The peptide inhibitor zFA-fmk, used as a negative control, did not affect Shigella-induced cell death at 100 µM. As expected, the nonvirulent S. flexneri strain BS176 was not cytotoxic. These data indicate that TPPII rather than the proteasome is necessary for Shigella-induced macrophage apoptosis.

View larger version (30K): [in this window] [in a new window]   FIG. 1.   TPPII is involved in Shigella-induced macrophage apoptosis. (A) Mouse J774 macrophages were treated with the TPPII inhibitor AAF-cmk (), the proteasome-TPPII inhibitor clasto-Lactacystin -lactone (lactacystin) (), or the proteasome inhibitors LLnL (), LLM (), LLL-vs (), or with zFA-fmk () and then infected with S. flexneri wild-type strain M90T or the nonvirulent strain BS176 () at an MOI of 10. The cytotoxicity was assayed by measuring the LDH release after 2 h. (B) J774 macrophages were treated with AAF-cmk (5 µM), LLM (100 µM), or the caspase-1 inhibitor YVAD (50 µM) and washed (AAF/w, LLM/w, and YVAD/w) prior to M90T infection. (C) Peritoneal macrophages were preincubated with a 10 µM concentration of the proteasome-TPPII inhibitors indicated or with YVAD-cmk (100 µM) and infected with M90T. (D) J774 macrophages were infected with M90T grown in the presence of 1 µM AAF-cmk (M90T/AAF) or 5 µM lactacystin (M90T/Lacta). In parallel, untreated macrophages (none) or macrophages treated with 1 µM AAF-cmk (AAF) or 5 µM lactacystin (Lacta) were infected with mock-treated M90T or BS176. Means and standard deviations of experiments done in triplicate are shown. Similar results were obtained in at least three (A and D) or two (B and C) independent experiments.

Maturation of caspase-1 and IL-1 in Shigella-infected macrophages is inhibited by AAF-cmk and lactacystin. During a Shigella infection, caspase-1 matures by limited proteolysis from a 45-kDa precursor to the active enzyme composed of 20- and 10-kDa polypeptides (6, 19). To test the effect of the TPPII inhibitors AAF-cmk and lactacystin on the maturation of pro-caspase-1, we preincubated mouse peritoneal macrophages with either compound and infected the cells with Shigella. Lysates of the infected macrophages were prepared at the indicated time points and analyzed by Western blotting with an antibody recognizing the 45-kDa precursor and the 20-kDa processed form of caspase-1. In untreated macrophages, mature caspase-1 was detected as soon as 30 min after infection with the wild-type Shigella strain M90T (Fig. 2A). Macrophages infected with the avirulent strain BS176 did not process pro-caspase-1 even by 4 h postinfection. AAF-cmk (10 µM) and, less potently, lactacystin (30 µM) blocked the processing of pro-caspase-1, indicating that TPPII participates upstream of caspase-1 in the apoptotic pathway. Whereas AAF-cmk completely abolished the maturation of caspase-1 throughout the 4-h infection, lactacystin was less efficient, and at 2 h postinfection the 20-kDa caspase-1 fragment began to appear. As previously shown (19), the caspase-1 inhibitor YVAD-cmk (100 µM) abrogated the processing of pro-caspase-1 during the Shigella infection.

View larger version (42K): [in this window] [in a new window]   FIG. 2.   TPPII is required for the maturation of caspase-1 in Shigella-infected macrophages. (A) Peritoneal macrophages (None) and macrophages treated with AAF-cmk (AAF, 10 µM), lactacystin (Lacta, 30 µM), or YVAD-cmk (YVAD, 100 µM) were infected with virulent (M90T, 0.5 to 4 h) or nonvirulent (BS, 4 h) Shigella strains. Maturation of caspase-1 was analyzed by Western blotting in lysates of infected macrophages. (B) Differentiated, LPS-stimulated THP.1 cells (None) and cells treated with AAF-cmk (AAF, 10 µM), lactacystin (Lacta, 135 µM), or YVAD-cmk (YVAD, 0.5 mM) were infected with virulent (M90T, 0.5 to 4 h) or nonvirulent (BS, 4 h) Shigella strains. Maturation of IL-1 was analyzed by Western blotting in lysates of infected and naive (LPS) macrophages. (C) The specific activity of purified caspase-1 against the fluorogenic peptide substrate YVAD-amc was determined in the absence or in presence of AAF-cmk, lactacystin, or YVAD-cmk at the concentrations indicated. The specific activity is defined as the number of micromoles of amc released per minute per milligram of protein. The experiments were done at least twice, and results similar to those shown were obtained each time.

Shigella is protected within macrophages from gentamicin by AAF-cmk and lactacystin. Intracellular bacteria are protected from the antibiotic gentamicin, which is used to discriminate between intra- and extracellular bacteria (26). As expected, the avirulent S. flexneri strain BS176 was protected within macrophages from gentamicin (Fig. 3). In contrast, gentamicin killed the wild-type strain M90T and, therefore, apoptotic macrophages appear to become permeable to the antibiotic. To test whether blocking apoptosis would result in protection of M90T from gentamicin, we incubated macrophages with TPPII and caspase inhibitors. Interestingly, treatment of macrophages with either AAF-cmk or lactacystin completely protected intracellular wild-type Shigella from gentamicin. The caspase-1 inhibitor YVAD-cmk or the broad-spectrum caspase inhibitor zVAD-fmk, on the other hand, did not protect M90T from gentamicin. However, both caspase inhibitors abolished Shigella-mediated cytotoxicity as determined by LDH release (data not shown).

View larger version (25K): [in this window] [in a new window]   FIG. 3.   Shigella is protected within macrophages from gentamicin by AAF-cmk or lactacystin. J774 macrophages were treated with the TPPII inhibitors AAF-cmk (AAF, 1 µM) or lactacystin (Lacta, 5 µM), with the caspase inhibitors YVAD-cmk or zVAD-fmk (YVAD or zVAD, 100 µM), or with cytochalasin B (CytoB, 2.5 µg/ml), an inhibitor of actin polymerization. After infection with either wild-type (M90T) or avirulent (BS176) S. flexneri and gentamicin treatment, the CFU were determined. Means and standard deviations of experiments done in quadruplicate are shown. Similar results were obtained in three independent experiments.

Bacterial escape from the phagosome is not impaired by AAF-cmk or lactacystin. Since the TPPII inhibitors AAF-cmk and lactacystin did not affect the phagocytic activity of macrophages, we investigated whether these inhibitors impair the escape of Shigella from the macrophage phagosome. Shortly after entry into a host cell, wild-type Shigella escapes the phagosome and moves intracellularly by polymerizing actin, resulting in "actin tails" at a single pole of the bacterium (4, 29). To assay the formation of cytoplasmic actin tails, we labeled infected macrophages with phalloidin-FITC and stained both macrophage nuclei and bacteria with propidium iodide. The wild-type Shigella strain M90T formed bacterium-associated actin tails in untreated macrophages, as well as in macrophages preincubated for 2 h with 1 µM AAF-cmk or 5 µM lactacystin (data not shown). Therefore, AAF-cmk and lactacystin did not prevent the escape of Shigella from the phagosome.

TPPII activity is not altered during Shigella infection of macrophages. To test whether the activity of TPPII is regulated during an infection of macrophages with Shigella, we assayed the hydrolysis of the fluorogenic peptide substrate AAF-amc. TPPII hydrolyzes only peptides with a free amino terminus, and AAF-amc is hydrolyzed by TPPII several orders of magnitude more efficiently than by the proteasome (11). To confirm that the AAF-amc hydrolysis observed was actually due to TPPII activity, we preincubated macrophages with the TPPII inhibitors AAF-cmk and lactacystin and determined the extent of AAF-amc hydrolysis in lysates of macrophages permeabilized with digitonin. AAF-cmk and lactacystin (10 µM each) blocked the hydrolysis of AAF-amc in macrophage lysates to 94 and 57%, respectively (data not shown), suggesting that the activity measured was due to TPPII.

1

1

TPPII is involved in apoptosis induced by ATP and staurosporine. To test if TPPII is specifically involved in Shigella-induced apoptosis or whether it also plays a role in other apoptotic pathways, we analyzed cell death induced by the protein kinase inhibitor staurosporine (30) and ATP (21). At a concentration of 5 µM, the TPPII inhibitors AAF-cmk and lactacystin decreased staurosporine-induced apoptosis in J774 macrophages by 83 and 75%, respectively (Fig. 4A). Similarly, AAF-cmk and lactacystin (10 µM each) inhibited ATP-induced apoptosis in mouse peritoneal macrophages by 75 and 54%, respectively (Fig. 4B). In contrast, the proteasome inhibitors LLnL and LLM (data not shown) did not affect apoptosis induced by either staurosporine or ATP. The induction of apoptosis by staurosporine requires a 12-h incubation period. Within this period, AAF-cmk and LLnL were not cytotoxic to the macrophages. However, lactacystin alone caused 40% cell death (data not shown). In summary, the inhibition pattern obtained for staurosporine- and ATP-induced apoptosis indicates that TPPII is also involved in these apoptotic pathways.

View larger version (17K): [in this window] [in a new window]   FIG. 4.   TPPII is involved in apoptosis induced by staurosporine and ATP. J774 macrophages were incubated with staurosporine (1 µM, 12 h) (A) or LPS-activated mouse peritoneal macrophages were treated with ATP (5 mM, 30 min) (B) in the absence of in the presence of AAF-cmk (), lactacystin (), or LLnL () at the concentrations indicated. The cytotoxicity was determined by measuring the LDH release and is corrected for cell death caused by lactacystin alone. Means and standard deviations of experiments done in triplicate are shown. Similar results were obtained in three (A) or two (B) independent experiments.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

TPPII participates in Shigella-induced macrophage apoptosis. In this report we describe the high-molecular-weight protease TPPII as a previously unidentified component of the apoptotic pathway triggered by S. flexneri in macrophages (Fig. 1). The TPPII inhibitors AAF-cmk and lactacystin, but not the proteasome inhibitor LLnL, potently blocked Shigella-induced apoptosis. This inhibition profile is characteristic for purified TPPII and not the proteasome (11). The proteasome inhibitor LLM but not LLnL protected macrophages to some extent from Shigella-induced apoptosis, which is inversely correlated to the potential of these compounds to block the proteasome (32). In contrast, LLnL inhibited apoptosis more potently than LLM in apoptotic pathways involving the proteasome (14, 34).

Role of TPPII in Shigella-induced apoptosis. TPPII appeared to be upstream of caspase-1 in Shigella-induced apoptosis. The TPPII inhibitors AAF-cmk and lactacystin prevented the maturation of caspase-1 and IL-1 in macrophages infected with Shigella, without inhibiting caspase-1 directly (Fig. 2). By activating caspase-1 and consequently IL-1, TPPII participates in a proinflammatory apoptotic pathway and contributes to an innate immune response. Similarly, lactacystin has been shown to block the activation of caspase-1 and IL-1 in macrophages treated with ATP (34). In this and other studies, however, the authors concluded that the proteasome participates upstream of caspases and mitochondrial events in thymocyte apoptosis (14, 20) and neuronal programmed cell death (34).

TPPII is involved in apoptosis triggered by staurosporine and ATP. TPPII might also participate in the apoptotic pathways triggered by staurosporine and ATP. AAF-cmk and lactacystin, but not proteasome peptide inhibitors, protected macrophages from apoptosis triggered by staurosporine and ATP (Fig. 4). The apoptotic pathways triggered by both ATP (23) and staurosporine (22) do not require caspase-1, suggesting that TPPII might also promote the activation of caspases other than caspase-1.