HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 35, 32339-32347, August 30, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/35/32339    most recent M205153200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Bidère, N. Articles by Senik, A. Search for Related Content PubMed PubMed Citation Articles by Bidère, N. Articles by Senik, A. Social Bookmarking                      What's this?

Selective Inhibition of Dipeptidyl Peptidase I, Not Caspases, Prevents the Partial Processing of Procaspase-3 in CD3-activated Human CD8⁺ T Lymphocytes^*

Nicolas Bidère^§, Marie Briet, Antoine Dürrbach, Céline Dumont, Jérôme Feldmann^¶, Bernard Charpentier, Geneviève de Saint-Basile^¶, and Anna Senik

From the  Laboratoire de Greffes d'Epithéliums et Régulation de l'Activation Lymphocytaire, Unité INSERM 542, Hôpital Paul Brousse, 94807 Villejuif, France and the ^¶ Laboratoire de Développement Normal et Pathologique du Système Immunitaire, Unité INSERM 429, Hôpital Necker-Enfants Malades, 75015 Paris, France

Received for publication, May 24, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Activation of primary human T cells by anti-CD3 and interleukin-2 resulted in partial processing of procaspase-3 in activated nonapoptotic (_m^high) CD8⁺ T cells but not in CD4⁺ T cells. Apical caspases-8 and -9 were not activated, and Bid was not processed to truncated Bid. Boc-D.fmk, a broad spectrum caspase inhibitor, did not prevent this process, whereas GF.dmk, a selective inhibitor of dipeptidyl peptidase I, was effective. Dipeptidyl peptidase I is required for the activation of granule-associated serine proteases. It is enriched in the cytolytic granules of cytotoxic lymphocytes, where it promotes the proteolytic activation of progranzymes A and B. Inhibition of granzyme B (GrB)-like serine proteases by Z-AAD.cmk prevented partial processing of procapase-3, whereas inhibition of GrA activity by D-FPR.cmk had no effect. Specific inhibitors of other lysosomal proteases such as cathepsins B, L, and D did not interfere in this event. Patients with Chediak-Higashi syndrome or with perforin deficiency also displayed partial processing of procaspase-3, excluding the involvement of granule exocytosis for the delivery of the serine protease in cause. The p20/p12 processing pattern of procaspase-3 in our model points to GrB, the sole serine protease with aspase activity. Small amounts of GrB were indeed exported from cytolytic granules to the cytosol of a significant fraction of GrB-positive cells.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Caspase-3 belongs to a family of cytosolic cysteine proteases that are synthesized as proenzymes and are converted during apoptosis into mature enzymes that can cleave crucial death substrates immediately downstream from Asp residues. Caspases are thus processed to form active heterodimeric enzymes by cleavage between their large and small subunits and by the removal of their prodomain (reviewed in Ref. 1). The proteolytic maturation of procaspase-3 is not only performed by initiator caspases-8 and -9, but also by granzyme (Gr)¹ B, the most abundant serine protease contained in the lytic granules of cytotoxic lymphocytes (CTL). GrB cleaves after Asp residues, preferentially at the IEXDX sequence, as shown in vitro (2). Accordingly, once introduced into a target cell, GrB is able to cleave procaspase-3 (3) between the large and small subunits, at the IETDS activation site, which is also recognized by caspase-8 (4, 5).

We have previously shown that stimulation of primary T lymphocytes through the CD2 receptor leads to partial processing of procaspases-3 and -7, as defined by cleavage between the large and small subunits, without removal of the prodomain (6). This was true for cells with a high mitochondrial membrane potential (_m) and no surface phosphatidylserine residues, two well known indicators of cell viability. Also, poly(ADP-ribose)polymerase, an indicator substrate for these caspases, was not cleaved. This implies that although the cleaved caspases were potentially active at this stage of processing (7), they were inhibited by an endogenous inhibitor, probably the X-linked IAP (inhibitor of apoptosis) protein (8). Concentrating on caspase-3, we obtained similar results after CD3 stimulation of primary T cells in the presence of IL-2. This prompted us to identify the protease(s) in cause.

We found that partial procaspase-3 cleavage preferentially occurred in a portion of nonapoptotic (_m^high) activated CD8⁺ but not CD4⁺ T lymphocytes. It was not prevented by Boc-D.fmk, a broad spectrum caspase inhibitor, but it was prevented by GF.dmk, a selective inhibitor of DPPI (also called cathepsin (Cat) C). The data presented in this study collectively point to a granule-associated serine protease whose activation is strictly dependent on DPPI activity and that exhibits a specific p20/p12 pattern of aspase activity toward procaspase-3, likely GrB. The involvement of granule exocytosis was excluded because patients with Chediak-Higashi syndrome (CHS) or with perforin deficiency also displayed partial processing of procaspase-3. Small amounts of GrB were indeed exported from cytolytic granules to the cytosol of a significant fraction of GrB-positive cells, suggesting an internal way of action.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Antibodies-- Anti-CD3 (OKT3) and anti-CD2 (T11₁ and D66 mAb) were given by Dr. A. Bernard (U343, Nice, France). Anti-CD95 mAb (CH-11) was from Immugenex Corp. (Los Angeles, CA). In confocal microscopy studies, mAb directed against perforin (G9), against granzyme A (MOPC-21), against LAMP-2 (H4B4), and against cytochrome c (6H2.B4) were from PharMingen (Becton Dickinson, Le Pont de Claix, France). The biotin-conjugated anti-GrB mAb (CLB-GB-11) was from Tebu (Le Perray-en-Yvelines, France). Rabbit anti-AIF was a gift from Dr. S. Susin (Institut Pasteur, France). Secondary reagents were from Caltag (Burlingame, CA). Anti-CD8-phycoerythrin (B-H7) and anti-CD4-PE (B-5F) were from Diaclone (Besançon, France).

Synthetic Inhibitors and Enzymatic Substrates-- Z-VAD.fmk (benzyloxycarbonyl-Val-Ala-Asp(OMe)-fluoromethylketone), Boc-D.fmk (Boc-Asp(OMe)-fluoromethylketone), Z-FA.fmk (benzyloxycarbonyl-Phe-Ala-fluoromethylketone), GF-dmk (Gly-Phe-diazomethylketone), GP.dmk (Gly-Pro-diazomethylketone), and D-FPR.cmk (D-Phe-Pro-Arg.chloromethylketone) were from Enzyme Systems Products (Livermore, CA). CA-074-Me, Z-FF.fmk (Z-Phe-Phe.fluoromethylketone), and Z-AAD.cmk (Z-Ala-Ala-Asp-chloromethylketone) were from Calbiochem (France Biochem, Meudon, France). The DPPI substrate, Gly-Phe-Naphtylamide (GF-NA); the cathepsin (Cat) B substrate, Z-Arg-Arg-NA (z-RR.NA); the Cat L substrate, Phe-Arg-NA (FR-NA); the granzyme A substrate, N-benzyl-oxycarbonyl-L-lysine thiobenzyl ester (BLT ester); and the -hexosaminidase substrate, p-nitrophenyl N-acetyl--D-glucosaminide, were purchased from Sigma.

T Lymphocyte Isolation and Culture Conditions-- Peripheral blood leukocytes were isolated from blood bank leukophoresis packs obtained from healthy volunteers (Etablissement Français du Sang). Adherent cells were removed by incubation on plastic dishes and passage over nylon wool columns. T lymphocytes were stimulated for 3 or 4 days with 0.25 µg/ml OKT3 plus 100 units/ml IL-2. CD4⁺ and CD8⁺ T cells were then negatively separated by immunomagnetic selection using anti-CD8- or anti-CD4-coated magnetic beads (Miltenyi Biotec, Auburn, CA). Alternatively, CD8⁺ T cells were negatively selected by the "panning" technique using anti-CD4 and anti-CD16 mAb (from Diaclone)-coated Petri dishes. Discontinuous density Percoll gradients (Amersham Biosciences) were used to separate large activated T cells (in the F2 fraction) from shrunken cells (in the F5 fraction) as described previously (9). The CHS patient (unique patient number 3 in Ref. 10) presents an homozygous 1-bp deletion in the CHS1 gene leading to a frameshift and a subsequent early truncated protein. In the two familial hemophagocytic lymphohistiocytosis patients, intracellular perforin expression was undetectable in their lymphocyte cytotoxic granules, and cytotoxic activity was either absent (unique patient number 92) or severely impaired (unique patient number 27) as a result of mutation in the perforin gene (11).

Cell Death Induction and Flow Cytometric Analysis of _m-- The inner mitochondrial transmembrane potential (_m) and the percentage of dead cells were measured by cytofluorometry after incubating the cells with DiOC₆ and propidium iodide (PI) respectively, as described previously (9). Cells with complete _m loss were obtained by a 10-min incubation with 5 µM carbonyl cyanide m-chlorophenyl hydrazone (Sigma).

Immunoblots-- Pellets of 5-10 × 10⁵ cells were directly resuspended in Laemmli buffer containing 4% sodium dodecyl sulfate and 2--mercaptoethanol and boiled for 5 min to avoid post-lysis processing of procaspase-3 by GrB contained in CTL (12). The membranes were probed with rabbit sera anti-caspase-3 (PharMingen), anti-caspase-9 (Cayman Chemicals, Spi-Bio, Massy, France), anti-Bid (given by Dr. X. Wang, Howard Hughes Medical Institute, Dallas, TX), and anti-actin (Sigma). We also used mAb against caspase-8 (5F7, Upstate Biotechnology, Euromedex, Souffelmeyersheim, France), against GrB (B18.1, Alexis Corporation, Coger, Paris, France), against Lamp-1 (mAb 25, Transduction Laboratories, Becton Dickinson), against poly(ADP-ribose)polymerase (C2-10, PharMingen), and against perforin (P1-8, Kamyia Biomedical Co., Seattle, WA). GrA and Cat B were probed with specific polyclonal goat IgG (Santa Cruz, Tebu). When necessary, the blots were stripped by using a Western blot recycling kit (Chemicon, Euromedex). Immunoblots were developed using enhanced chemiluminescence reagents (ECL kit; Amersham Biosciences) after incubation with horseradish peroxidase-coupled secondary reagents.

Confocal Microcopy Analysis-- The cells were fixed with 3% paraformaldehyde in phosphate-buffered saline for 30 min at 4 °C, washed with phosphate-buffered saline, and permeabilized with 0.05% Triton X-100 for 5 min at room temperature. After three washings, staining was performed as described previously (9). The cells were examined using a confocal laser scanning microscope (Leica).

Subcellular Fractionation-- 15 × 10⁶ cells were washed twice in phosphate-buffered saline and resuspended for 10 min at 4 °C in 300 µl of homogenization buffer consisting of 10 mM triethanolamine, 1 mM EDTA, 250 mM sucrose, 10 mM acetic acid, and a mixture of protease inhibitors (from Roche Molecular Biochemicals), pH 7.4. The cells were then subjected to 60 pestle strokes of a glass Dounce homogenizer (Merck). Unbroken cells and nuclei were pelleted by centrifugation at 760 × g for 10 min at 4 °C. The resulting supernatants were centrifuged at 10,000 × g for 15 min to eliminate heavy membrane pellets and yield S10 supernatants. The latter were centrifuged over a sucrose cushion (2.5 M) at 225,000 × g for 1 h to separate cytosolic (S225) and vesicular fractions. The protein concentration of S225 fractions was determined using the micro BCA kit (Pierce).

Enzyme Assays-- Cat C activity was assayed by hydrolysis of the chromogenic GF-NA substrate and absorbance at 340 nm essentially as described (13). Cat B and Cat L activities were assayed by hydrolysis of their respective specific substrates z-RR-NA and FR-NA (14). GrA tryptase activity was measured by the hydrolysis of BLT ester according to the protocol of Ref. 15. Hexosaminidase activity was assayed according to the protocol of Ref. 16. The percentage of hexosaminidase secretion was calculated according to the formula: (OD_supernatant/OD_lysate) × 100.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Partial Processing of Procaspase-3 Is Preferentially Initiated in Nonapoptotic Primary CD8⁺ T Cells after CD3 + IL-2 Stimulation-- Large activated T cells, generated by stimulating purified T cells for 4 days with soluble anti-CD3 and IL-2, were isolated upon discontinuous density Percoll gradients as large cells sedimenting in the F2 fraction. These cells were negatively separated into CD4⁺ and CD8⁺ T cell subsets by use of magnetic immunobeads. This procedure eliminates dead cells (PI-positive, in the low buoyant density F1 fraction) and cells undergoing apoptosis and thus already shrinking (below the F2 fraction) (9). This enabled us to compare cells at the same stage of activation in terms of cell size and proliferation as assessed by light scatter properties in flow cytometry and [ ³H]TdR incorporation (Fig. 1A, panel a). Most of the cells (~90%) in each subset were strongly labeled with the fluorescent lipophilic DiOC₆ (3) probe, an indicator of _m (Fig. 1A, panel b). In these conditions, the caspase-3 proenzyme had been partially processed in the lysates of CD8⁺ T cells, but not in those of CD4⁺ T cells, into a doublet of 20- and 22-kDa protein species, as recognized in Western blotting by an antibody directed against the large subunit (Fig. 1A, panel c). The caspase-3 proenzyme consisted of a p31-p33 doublet that was best detected at very short autoradiography times (Fig. 1B). The p20-p22 doublet corresponds to isoforms of the large subunit, still associated with their respective prodomains after initial cleavage at the IETDS site (17). As in CD2-activated T lymphocytes (6), the partial processing of procaspase-3 in CD3 activated T lymphocytes did not affect the integrity of poly(ADP-ribose)polymerase (Fig. 1C).

View larger version (40K): [in this window] [in a new window]   Fig. 1.   Partial processing of caspase-3 proenzyme by cleavage between the large and the small subunits preferentially occurs in CD8+ nonapoptotic T cells after CD3 + IL-2 stimulation. A, purified T cell populations were stimulated for 4 days with soluble OKT3 (250 ng/ml) and 100 units/ml IL-2, fractionated upon discontinuous Percoll gradients to yield large F2 cells, and then negatively selected on magnetic immunobeads into CD4+ and CD8+ T cell subsets (85% purity). Panel a, forward side scatter analysis of the two subpopulations and of resting T cells (numbers refer to the mean forward scatter). [ 3H]TdR incorporation in the CD4+ and CD8+ T cell subsets was 88,000 and 110,000 cpm, respectively, versus 1,500 cpm in resting T cells. Panel b, the cells were doubly stained with PI to detect dead cells and with DIOC6 (3) to evaluate m. The numbers are the percentages of DIOC6(3)high/PI-negative (viable) cells. The protoionophore carbonyl cyanide m-chlorophenyl hydrazone (m-ClCCP) was used for cytofluorometry settings. Panel c, 5 × 105 CD4+ and CD8+ (F2) T cells were lysed and analyzed by Western blot with anti-caspase-3. The blot was stripped and reprobed with an anti-actin. This experiment is representative of six others. B, visualization of the procaspase-3 p33/p31 doublet at short autoradiography time. C, immunoblot analysis of poly(ADP-ribose)polymerase in CD4+ and CD8+ (F2) T cell lysates. Control apoptotic cells consisted of T cells exposed to 500 nM staurosporine for 2 h.

We followed the kinetics of the appearance of the p20-p22 doublet using whole CD8⁺ T cell populations isolated by the panning technique on successive days of the culture period. The p20-p22 doublet was first detected on day 3 of the stimulation period. On day 4, the amount of the doublet either remained stable (Fig. 2A) or increased substantially (Fig. 2B). In the latter case, the mature 17-kDa form (p17) of the large subunit also appeared on day 4. The production of p17 is essentially dependent on autocatalysis or on caspase-3-like activity targeted at the ESMDS site (amino acids 25-29) (18). Given that large activated CD8⁺ T cells display the p20-p22 doublet of caspase-3 without the p17 form (Fig. 1), we assumed that p17 would only appear in shrunken (apoptotic) cells present in the CD8⁺ T cell population. The large F2 (91% _m^high) and the shrunken F5 (55% _m^high) cells of this population were isolated on density Percoll gradients, and we found that such was indeed the case (Fig. 2C). We thus defined experimental conditions suitable for studying the activated CD8⁺ T lymphocytes that display only partial processing of the caspase-3 proenzyme, limited at the IETDS cleavage site; such cells are preferentially found on day 3 of the stimulation period in whole CD8⁺ T cell populations or even later when using large (F2) CD8⁺ T cells isolated on density Percoll gradients.

View larger version (33K): [in this window] [in a new window]   Fig. 2.   The activation status of caspases-3, -8, and -9 and Bid in nonapoptotic, activated CD8+ T cells versus those committed to apoptosis. CD8+ T cells were isolated from a CD3-activated whole T cell populations after removal of dead (F1) cells on discontinuous density Percoll gradients and after elimination of CD4+ and CD16+ cells by the panning technique. The activation status of caspases-3, -8, and -9 in the lysates of 5 × 105 purified CD8+ T cells was then examined by Western blot. A, kinetics of procaspase-3 processing showing that partial processing to the p20 doublet starts on day 3 of stimulation and may be maintained as such until day 4 in the absence of caspases-8 and -9 processing. B, kinetic analysis of another CD8+ T cell preparation showing that procaspase-3 cleavage can proceed on day 4 to the p17 form. This event coincides with caspases-8 and -9 activation, as well as Bid processing to tBid. 1 × 106 cells were required for the visualization of the various forms of tBid. C, isolation upon density Percoll gradients of F2 and F5 cells contained in a whole CD8+ T cell population stimulated for 4 days, and immunoblot analysis of caspase-3. 91% of F2 cells and only 55% of F5 cells were mhigh. D, activated (F2) CD8+ T cells display a punctiform (mitochondrial) immunostaining pattern of cytochrome c (cyt. c) and AIF as visualized by confocal immunofluorescence microscopy. Apoptotic control consisted of the same cells exposed for 2 h to 500 nM staurosporine.

The activation status of apical caspases-8 and -9 was analyzed in a similar manner, as was the status of Bid, a cytosolic death promoter and member of the Bcl-2 family. Truncated Bid (tBid), which can be generated from Bid by the proteolytic activity of caspase-8, GrB, and lysosomal extracts (see "Discussion"), can induce a conformational change of Bax, allowing it to be inserted and oligomerized at the outer mitochondrial membrane, which in turn leads to the release of cytochrome c (19, 20). When the caspase-3 proenzyme was only partially cleaved into the p20-p22 doublet (day 3 of the stimulation period), the caspase-8 and -9 proenzymes were not processed, and Bid was not degraded proteolytically in activated CD8⁺ T cells (Fig. 2B). In contrast, the presence on day 4 of the p17 subunit of caspase-3 coincided with the appearance of the p41/43 and p18 cleavage products (large subunit) of caspase-8 and of the p37 cleavage product (large subunit) of caspase-9. In these conditions, the Bid proform coexisted with the 15-, 14-, and 12-kDa species of tBid. Consistent with the apparent integrity of Bid and procaspases-8 and -9 in nonapoptotic (F2) CD8⁺ T cells, cytochrome c was exclusively located in mitochondria, as shown by the punctate immunostaining pattern observed by laser scanner confocal microscopy (Fig. 2D). AIF, another apoptogenic factor normally sequestered in the intermembrane space of mitochondria (21), also displayed a punctate immunostaining pattern that perfectly matched that of cytochrome c. Therefore, permeabilization of the outer membrane of mitochondria probably did not occur in T cells displaying partial caspase-3 processing.

The Partial Processing of Procaspase-3 into the p20-p22 Doublet Is Caspase-independent-- Cell-permeable, broad spectrum caspase inhibitors Z-VAD.fmk and Boc-D.fmk and control peptide Z-FA.fmk (common inhibitor of Cat B and L) were added to CD3-stimulated T cell cultures at the beginning of the culture period to evaluate their effect on partial procaspase-3 processing. Preliminary experiments (Fig. 3A) confirmed previous reports (17, 22) that Z-VAD.fmk inhibits T cell proliferation induced by soluble anti-CD3 + IL-2 in a dose-dependent manner (90% inhibition exerted by 100 µM on day 4). Surprisingly, Z-FA.fmk exerted almost the same inhibitory effect. The mechanisms by which these peptide inhibitors inhibit T cell proliferation have not been established (see "Discussion"). In contrast, T cell proliferation was only moderately affected by 100 µM Boc-D.fmk, excluding the involvement of the methyl ketone group. Lower doses of Z-VAD.fmk were added to minimize its anti-proliferative effect (25 µM on days 0, 2, and 3). However, Z-VAD.fmk still inhibited ~40% of T cell proliferation, and this was accompanied by enhanced background cell death (Fig. 3B). In contrast, Boc-D.fmk did not affect cell proliferation, cell viability, or cell size, and it blocked anti-CD95-induced apoptosis of activated human peripheral T lymphocytes (Fig. 3B). We have previously shown that this inhibitor can also prevent the internucleosomal DNA fragmentation induced by apoptotic stimuli that do not rely on death receptors (6, 9). As shown in Fig. 3C, Boc-D.fmk did not prevent the generation of the p20-p22 doublet of caspase-3, although when the same cells were exposed to staurosporine, it did prevent the full maturation of the large subunit to p17, a caspase-dependent step. Therefore, the p20-p22 doublet appeared to be caused by the action of an upstream protease insensitive to caspase inhibitors.

View larger version (27K): [in this window] [in a new window]   Fig. 3.   Partial processing of procaspase-3 in activated CD8+ T cells is caspase-independent. A, Boc-D.fmk has only a moderate anti-proliferative effect. The peptide inhibitors were added to T cells at the beginning of culture. [3H]TdR incorporation was measured on day 4 of the stimulation period. The results are representative of six experiments. DMSO, dimethyl sulfoxide. B, Boc-D.fmk is as effective as Z-VAD.fmk in preventing Fas-mediated apoptosis without inducing enhanced background cytotoxicity. 25 µM of z-VAD.fmk or Boc-D.fmk was added to T cell cultures on days 0, 2, and 3 of the stimulation period. [3H]TdR incorporation was measured on day 4 (values are the means ± S.D. of triplicate determinations), and cell death induced by anti-Fas (2 µg/ml) was assessed by trypan blue exclusion and cell morphology. C, partial processing of procaspase-3 is caspase-independent. CD8+ T cells were isolated as in Fig. 2 from whole T cell cultures exposed to Boc-D.fmk (25 µM added at days 0, 2, and 3). Immunoblot analysis of caspase-3 processing was performed with the lysates of 5 × 105 cells. The blot was stripped and reprobed with an anti-actin to monitor protein loading. The apoptotic signal consisted of 500 nM staurosporine delivered for 2 h. This experiment is representative of five others.

GF.dmk, a Peptide Inhibitor of DPPI Activity, Prevents the Partial Processing of Procaspase-3-- Because the p20-p22 doublet of caspase-3 was seen in activated CD8⁺ but not CD4⁺ T cells, we hypothesized that a protease contained in the cytolytic granules of CTL might be the protease causing the partial processing of procaspase-3 in our system. We first used GF.dmk, a synthetic peptide that readily enters the cells and that specifically and irreversibly inhibits the activity of DPPI, a lysosomal thiol protease with dipeptidyl aminopeptidase activity that is required for post-transational processing and activation of many myeloid and lymphoid granule-associated serine proteases (13, 23-26). Enriched in cytolytic granules of CTL (27) and coordinately expressed with GrA during CD8⁺ T cell development and differentiation (28), DPPI is the sole protease that performs in vivo the proteolytic activation of the proenzyme forms of GrA and GrB (29). T lymphocytes were therefore stimulated for 3 days in the continuous presence of graded doses of GF.dmk. Chronic exposure of CD8⁺ T cells to 10 µM GF.dmk almost totally prevented the cleavage of procaspase-3 between the large and small subunits, whereas the control peptide (GP.dmk) was ineffective (Fig. 4A). The concentration of GF.dmk used was indeed sufficient to inhibit totally the enzymatic hydrolysis of GF.NA (the DPPI substrate) by CD8⁺ T cells lysates. It was also sufficient to prevent the occurrence of BLT esterase activity in these lysates, indicative of GrA activity. The same concentration of GF.dmk failed to block the enzymatic activities of two other lysosomal thiol proteases, Cat B and Cat L, which could in contrast be inhibited by similar doses of their respective specific inhibitors CA-074-Me (30) and Z-FF.dmk (14). As expected, caspase activity was not affected in cells subjected to apoptotic stimuli (not shown). GF.dmk and GP.dmk were not toxic and did not impair the proliferation of CD8⁺ T cells (Fig. 4C) or CD4⁺ T cells (not shown).

View larger version (37K): [in this window] [in a new window]   Fig. 4.   A granule-associated serine protease is the agent causing partial cleavage of procaspase-3 in CD8+ T cells. A, inhibition of DPPI activity by GF.dmk prevents partial processing of procaspase-3. The indicated concentrations of GF.dmk and of control GP.dmk (an inhibitor of DPPIV activity) were added to cell cultures each day during a 3-day stimulation period. CD8+ T cells were then isolated as in Fig. 2. DPPI and BLT esterase activities in the treated cells were assayed by hydrolysis of the chromogenic substrates GF-NA and BLT ester. T cells incubated with 10 µM GF.dmk or GP.dmk had intact Cat B and Cat L activities, as determined by hydrolysis of the Z-RR-NA and FR-NA chromogenic substrates, respectively. Cat B and Cat L activities were otherwise totally inhibited by their respective specific inhibitors, CA-074-Me and z-FF.fmk (added to T cells for the last 16 h of the stimulation period). Similar results were obtained in four other independent experiments. B, inhibition of GrB-like but not GrA activity prevents partial processing of procaspase-3. Z-AAD.cmk and D-FPR.cmk, inhibitors of GrB-like serine proteases and GrA activities, respectively, were added for the last 16 h of culture. Western blot analysis of caspase-3 was performed in CD8+ T cells isolated after a 3-day stimulation period as described for Fig. 2. After thorough washing, the cells were lysed and assayed for BLT esterase activity; D-FPR.cmk (12.5 µM) had effectively entered the cells and inhibited GrA activity (representative of three experiments). C, GF.dmk and Z-AAD.cmk do not inhibit CD8+ T cell proliferation. [3H]TdR incorporation was estimated after a 3-day culture in the presence of the peptide inhibitors (10 µM). D, specific inhibition of Cat B and Cat D activities does not affect the partial processing of procaspase-3. The inhibitors were present during the last 16 h of the 3-day stimulation period.

To better characterize the granule-associated serine protease(s) targetted by DPPI in our system, we performed inhibitor studies with the peptide chloromethylketones Z-AAD.cmk and D-FPR.cmk that can directly inhibit mature GrB and GrA activity, respectively (reviewed in Ref. 31). These compounds were added to T cell cultures on day 2 of the stimulation period, i.e. 1 day before the initiation of procaspase-3 processing. On day 3, the partial processing of procaspase-3 was strongly attenuated by 12.5 µM of Z-AAD.cmk (Fig. 4B). The same concentration of D-FPR.cmk was ineffective, although it inhibited the BLT esterase activity of GrA in CD8⁺ T cells. Neither of these two compounds affected T cell viability or proliferation at the concentrations used (Fig. 4C). When tested in the same conditions, neither CA-074-Me (Cat B inhibitor) or Z-FF.fmk (Cat L inhibitor) conferred any protection against procaspase-3 cleavage (Fig. 4D). The same was true for pepstatin A, an inhibitor of Cat D (tested at the highest nontoxic dose, 25 µM) (Fig. 4D). Collectively, these data strongly suggested that the protease causing partial processing of procaspase-3 in CD8⁺ T lymphocytes was a granule-associated serine protease colocalized with DPPI in cytolytic granules, likely GrB.

The Active Granule-associated Serine Protease Is Not Exocytosed-- We used CD8⁺T lymphocytes from a patient suffering from Chediak-Higashi syndrome, a genetic disorder caused by a defect in the lysosomal trafficking regulator called CHS1 (32). CTL from CHS patients exhibit giant lytic granules that are unable to release their cytotoxic proteins, consistent with the general impairment of lysosome secretion affecting hematopoietic cells (33). Western blot analysis showed that CD3-activated CD8⁺ T cells from the CHS patient studied predominantly exhibited the p20 subunit of pro-caspase-3 with little mature p17 (Fig. 5A, panel a). No -hexosaminidase activity was detected in the supernatants of these cells following an exocytosis triggering signal (Fig. 5A, panel b). Instead, -hexosaminidase, a lysosomal resident enzyme, was released from normal activated CD8⁺ T cells subjected to degranulation. We also tested CD8⁺T lymphocytes from a pediatric patient with familial hemophagocytic lymphohistiocytosis, displaying perforin deficiency (34). Again, Western blotting showed that CD8⁺ T cells exhibited a partial procaspase-3 processing pattern, indicating that the p20-p22 doublet observed was generated in a perforin-independent manner. All of these experiments strongly suggested that the active granule-associated serine protease was not introduced into the cytosol of CD8⁺ T cells through the granule exocytosis pathway.

View larger version (29K): [in this window] [in a new window]   Fig. 5.   GrB-like action is unlikely to result from lytic granule exocytosis. A, activated CD8+ T cells from a CHS patient also display partial processing of procaspase-3. Panel a, 3-day activated CD8+T cells were prepared as for Fig. 2. Expression of GrB, perforin, and actin is also shown. Panel b, when exposed for 1 h to a CD2 degranulation signal (1 µg/ml D66+T111), normal CD8+ T cells degranulate and secrete -hexosaminidase, whereas CD8+ T cells from the CHS patient do not degranulate. B, activation status of caspase-3 in activated CD8+ T cells from a perforin-deficient patient.

GrB Is Translocated from Lytic Granules to the Cytosol in a Portion of Activated CD8⁺ T Cells-- We thus examined the localization of several proteins normally sequestered in cytolytic granules, namely perforin, GrA, GrB, and Cat B. About 30% of activated CD8⁺ T cells were positive for GrB, as assessed by cytofluorometry, using phycoerythrin-anti-CD8 and fluorescein isothiocyanate-anti-GrB. Activated CD8⁺ T cells (F2 cells) were examined by confocal laser scanning microscopy after double labeling with antibodies directed against GrB and either perforin, Lamp-2, or GrA. Perforin and GrA displayed a punctate immunostaining pattern consistent with their localization in lytic granules (Fig. 6A). Lamp-2, a lysosomal membrane glycoprotein, also displayed a bright punctate immunostaining pattern with some additional faint and diffuse staining. GrB was colocalized with the three other molecules inside punctate structures, but in 18% of GrB-positive cells (of 165 scored) there was evidence of diffuse GrB in the cytosol.

View larger version (33K): [in this window] [in a new window]   Fig. 6.   GrB translocates from cytolytic granules to cytosol. A, analysis by confocal immunofluorescence microscopy of the subcellular localization of GrB, perforin, Lamp-2, and GrA in activated CD8+ T cells. In some cells, GrB is not colocalized with the other molecules and displays instead a diffuse distribution pattern (arrowheads). B, GrB is specifically released into the cytosol. Panel a, S225 cytosolic extracts (70 µg) and cell lysates were prepared from 4-day activated (F2) CD8+ and CD4+ T cells. Immunoblots were probed with anti-GrB + anti-perforin, stripped, and successively reprobed with anti-GrA, with anti-Cat B + anti-Lamp-1 and with anti-actin. Panel b, -hexosaminidase activity was measured in cytosolic extracts to ensure the integrity of the lysosomal compartment.

We performed subcellular fractionation experiments and Western blot analysis to further examine whether GrB or other soluble compounds from the lytic granules (i.e. perforin, GrA, and Cat B) were present in the cytosolic fractions of activated CD8⁺ T cells (Fig. 6B). To detect contaminating lysosomes and to ensure that the fractionation procedure did not disrupt the lysosomes, the cytosolic fractions were also tested for Lamp-1 glycoprotein expression (Fig. 6B, panel a) and for -hexosaminidase activity (Fig. 6B, panel b). Activated CD4⁺ T cells were used as internal controls. GrB was the only granule protein to be detected in significant quantities in the cytosolic fractions of CD8⁺ T cells. The other proteins were found in the cell lysates (Fig. 6B, panel a). As for CD8⁺ T cells, significant amounts of GrA (and perforin) were detected in CD4⁺ T cell lysates but not in the corresponding cytosolic fractions. The fact that neither Cat B nor GrA were released in significant amounts implies that the mechanism of GrB release from lytic granules to cytosol does not involve the generalized rupture of lysosomal membranes at the stage of T cell activation examined.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

In this study, we demonstrated that inhibition of DPPI, an enzyme required for the activation of myeloid and lymphoid granule-associated serine proteases, prevents the partial processing of procaspase-3 proenzyme that occurs in CD3-activated human CD8⁺ T lymphocytes. We favor the notion that the serine protease responsible for this type of processing is a GrB-like protease, likely GrB itself, and that this protease is exported from lytic granules to the cytosol by an internal mechanism. These conclusions are based on four lines of argument: 1) Boc-D.fmk, a broad spectrum inhibitor of caspase activity devoid of nonspecific side effects, did not prevent the partial processing of the caspase-3 proenzyme. In contrast, GF.dmk, a specific and irreversible inhibitor of DPPI, was effective, as well as Z-AAD.cmk, which directly inhibits GrB-like serine proteases. 2) In CD4⁺ T cells costimulated in culture with CD8⁺ T cells, the caspase-3 proenzyme remained intact. 3) CD8⁺ T cells from CHS and perforin-deficient patients had the same fate as normal CD8⁺ T cells, despite the impairment of the perforin/granzyme-based exocytosis pathway. 4) Small amounts of GrB (but not of GrA or Cat B) were found in the cytosol of nonapoptotic activated CD8⁺ T cells (_m^high, with cytochrome c and AIF displaying a mitochondrial distribution pattern) in at least 18% of activated CD8⁺ T cells expressing GrB.

In a stimulation system quite similar to ours, others observed that full blown activation of caspases-3, -6, -7, and -8 occurred in actively proliferating T cell populations, even when they were depleted by the sorting of dying annexin-V-positive cells (17). In contrast, by selecting _m^high CD8⁺ T cells, we only observed partial processing of procaspase-3. We have shown that phosphatidylserine exposure (detected by annexin-V) is a rather late apoptotic event that occurs downstream of _m loss, concomitantly with caspase activation (9). Thus, the use of CD8⁺ T differently engaged in the commitment phase to apoptosis may explain these discrepancies. Although we did not examine the executioner caspase-7, our previous results indicated that it was also partially processed in _m^high activated T cells, at a cleavage site compatible with that of apical caspases and GrB (6) and hence probably subjected to the same protease as procaspase-3. In agreement with previous studies (17, 22), we found that Z-VAD.fmk inhibited both caspase activation and T cell proliferation. However, Z-FA.fmk, a frequently used Cat B and L inhibitor, had the same inhibitory effect when added at the very beginning of the culture period. In fact, Z-VAD.fmk efficiently blocks Cat B, which is a lysosomal housekeeping cysteine protease (35). It may also facilitate the death of activated T cells located at the G₂/M phase of the cell cycle (36), a phenomenon that possibly occurred at background levels in our system. As to Z-FA.fmk, it can behave in certain systems as a potent inhibitor of NF-B gene expression (37) and hence affect T cell proliferation in our experimental conditions. In contrast, Boc-D.fmk had little effect on T cell proliferation (Fig. 3) or GrB expression (not shown). Nevertheless, Boc-D.fmk prevented the caspase-dependent maturation of the large subunit of caspase-3 to its p17 form. These data assessed the specific caspase blocking capacity of Boc-D.fmk in our system. This part of our work not only questioned the notion that caspase activity is required for the initiation of T cell proliferation (17, 22) but also indicated that a noncaspase protease was responsible for the partial procaspase-3 processing occurring in activated CD8⁺ T cells.

Selective inhibition of DPPI activity by GF.dmk prevented this type of processing. DPPI is a thiol protease with aminodipeptidase activity required for the activation of many granule-associated serine proteases. In T lymphocytes, it is predominantly located in the cytolytic granules of CTL (27, 28), where it performs the proteolytic maturation/activation of proGrA and proGrB (13, 23, 24, 29). We therefore assumed that a granule-associated serine protease dependent on DPPI activity was in charge of procaspase-3 processing. Cytolytic granules from human cytotoxic T lymphocytes express four serine proteases called granzymes: GrB, which cleaves after Asp residues (Aspase); GrA and tryptase 2, which cleave after Arg or Lys; and GrH, a chymase that cleaves after Phe (reviewed in Ref. 38). To date, only GrA and GrB have been reported to be activated by DPPI. The partial cleavage pattern of procaspase-3 in CD8⁺ T cells perfectly matched the IETDS activation site recognized by GrB, suggesting that the protease in cause was GrB. The serine protease inhibitor Z-AAD.cmk, whose preferred substrate at low concentrations is GrB (31), also inhibited the processing of procaspase-3 at 12.5 µM, corroborating this conclusion.

The likely involvement of GrB in the partial processing of procaspase-3 raises two questions: 1) Why was the caspase-3 proenzyme initially targetted? 2) Why was Bid not cleaved to its pro-apoptotic tBid fragment? Procaspase-3 is in fact much more vulnerable in vitro to GrB than to the upstream caspase-9 (7), caspase-8, and caspase-10 (5). In cells treated with GrB and defective adenovirus, caspase-3 is activated before caspases-8 and -9 (39). Thus, it is conceivable that small amounts of GrB, delivered to the cytosol of activated CD8⁺ T cells, could preferentially interact with procaspase-3. Cytochrome c release, one major consequence of tBid translocation to the outer membrane of mitochondria, was not observed either. Bid can be cleaved at different sites by caspase-8 (40, 41), GrB (42-44), and lysosomal extracts (45). Bid is a better substrate for granzyme B than caspases-3 and -8 by more than 10-fold (44). However, it has been reported that tBid protein generated by caspase-8 activity in intact cells does not accumulate but is instead very rapidly degraded by the ubiquitin proteolytic system (46). In this view, it is possible that small amounts of tBid generated in activated CD8⁺ T cells might also be degraded in this way.

The partial processing of procaspase-3 also seen in patients with CHS or with perforin deficiency precluded the involvement of granule exocytosis for the delivery of the responsible granule serine protease. Part of GrB was indeed found into the cytosol, outside the cytolytic granules of a significant proportion (~20%) of GrB-positive _m^high CD8⁺ T cells. The fact that neither Cat B nor GrA was released in significant amounts implies that this mechanism does not involve the generalized rupture of lysosomal membranes at the stage of T cell activation examined. There is increasing evidence that some lysosomal proteases are specifically relocated from lysosomes to the cytosol, acting as death mediators in several models of apoptosis (reviewed in Ref. 47). This is the case for Cat D (48, 49) and for Cat B (50-52). Our results indicate for the first time that GrB may, in certain circumstances, be relocated from lytic granules to cytosol. The partial processing of procaspase-3, seen at day 3 of the culture period, is likely an early apoptotic event, held in check by endogenous IAPs. Further accumulation of GrB into the cytosol might irrevocably lead to cell death. This possibility is currently being examined.

	ACKNOWLEDGEMENTS

We thank Dr. Olivier Déas, Jérôme Estaquier, and Christophe Baron for helpful discussions.

	FOOTNOTES

^* This work was supported by the CNRS and INSERM and by grants from the Association pour la Recherche pour le Cancer and the Hôpital Universitaire de Bicêtre, Faculté de Médecine, Paris Sud.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^§ Supported by a grant from the Ministère de la Recherche et de la Technologie.

To whom correspondence should be addressed: Laboratoire de Greffes d'Epithéliums et Régulation de l'Activation Lymphocytaire, Unité INSERM 542, Hôpital Paul Brousse, 14 avenue Paul Vaillant-Couturier, Bâtiment Lavoisier, 94807 Villejuif, France. E-mail: asenik@infobiogen.fr.

Published, JBC Papers in Press, June 21, 2002, DOI 10.1074/jbc.M205153200

	ABBREVIATIONS

The abbreviations used are: Gr, granzyme; Cat, cathepsin; CHS, Chediak-Higashi syndrome; CTL, cytotoxic T lymphocytes; DPPI, dipeptidyl peptidase I; IL, interleukin; mAb, monoclonal antibody; PI, propidium iodide; tBid, truncated Bid.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				X. Que, J. C. Engel, D. Ferguson, A. Wunderlich, S. Tomavo, and S. L. Reed Cathepsin Cs Are Key for the Intracellular Survival of the Protozoan Parasite, Toxoplasma gondii J. Biol. Chem., February 16, 2007; 282(7): 4994 - 5003. [Abstract] [Full Text] [PDF]

M. Laforge, N. Bidere, S. Carmona, A. Devocelle, B. Charpentier, and A. Senik Apoptotic Death Concurrent with CD3 Stimulation in Primary Hum