Calreticulin is an endoplasmic reticulum-resident chaperone that is stored in the cytotoxic granules of CTLs and NK cells and is released with granzymes and perforin upon recognition of target cells. To investigate the role of calreticulin in CTL-mediated killing, we generated CTL lines from crt+/+ and crt−/− mice expressing a constitutively active form of calcineurin in the heart. Crt−/− CTLs showed reduced cytotoxic activity toward allogeneic target cells despite normal production, intracellular localization, and activity of granzymes and despite perforin overexpression. Comparable or higher amounts of granzymes were degranulated by crt−/− cells in response to immobilized anti-CD3 Abs, indicating that calreticulin is dispensable for the signal transduction that leads to granule exocytosis. The ability to form conjugates with target cells was affected in the crt−/− CTLs, explaining the observed reduction in cytotoxicity. Conjugate formation and cytotoxicity were completely restored by treatments that facilitate recognition and contact with target cells, a prerequisite for degranulation and killing. Therefore, we conclude that calreticulin is dispensable for the cytolytic activity of granzymes and perforin, but it is required for efficient CTL-target cell interaction and for the formation of the death synapse.
Cytotoxic T lymphocytes and NK cells play a major role in the response mounted by the immune system against virally infected and tumor cells. Killing of target cells by CTLs and NK cells predominantly occurs through two major pathways: apoptosis induced by Fas-Fas ligand interaction and/or cell death induced by the components of the lytic granules (1, 2). Lytic granules are specialized lysosomes containing a pore-forming protein, perforin, and a family of serine proteases collectively named granzymes (3). Upon recognition of target cells, the contents of the lytic granules are released into the synapse formed between CTL and target cells, ultimately leading to destruction of the target (4, 5, 6). The mechanism by which granzymes gain access to the cytosol of target cells is still unknown. Granzyme B can enter target cells through receptor-mediated endocytosis (7, 8), but once inside the cell, it is unable to escape the endocytic compartment to trigger apoptosis without the presence of perforin or other lytic agents (9, 10, 11).
The major role played by perforin in granule-mediated cytotoxicity became evident after studies in perforin knockout mice in which target cells could not be killed by the granule-mediated pathway (12, 13, 14). Initial attempts to purify perforin from the cytotoxic granules of lymphokine-activated killer cells lead to the observation that a 60-kDa protein present in the granules consistently coeluted with perforin (15). This protein was identified as calreticulin, an endoplasmic reticulum (ER)3 chaperone and Ca2+-binding protein (16, 17). Later it was shown that calreticulin interacts with perforin through its P domain, which has lectin-like activity (18) and high Ca2+-binding affinity (19), and, to a lesser extent, through its N domain. Interaction between the two proteins occurs only in the absence of Ca2+ (20). Calreticulin was found to inhibit the lytic activity of preparations of purified perforin on erythrocytes (21, 22). The inhibitory action, however, was not mediated by either the N or P domain, but, rather, by the calreticulin C domain, a highly acidic region that binds Ca2+ with high capacity and low affinity (19). Although the nature of the inhibitory role of calreticulin was not investigated in detail, it was proposed that by binding to the cell membrane, calreticulin could provide protection against osmotic lysis and perforin polymerization within the phospholipid bilayer. It was also speculated that such a mechanism could protect CTLs from their own perforin during the killing of target cells (21). This hypothesis has never been tested, and the possibility that exogenously added calreticulin may block perforin activity in a nonspecific fashion, simply chelating Ca2+ ions in the medium, has never been experimentally excluded. The data on calreticulin interaction with perforin, along with studies indicating that calreticulin expression is up-regulated during T cell activation (23), suggested that the ER chaperone may play a role in degranulation and CTL-induced killing. Unfortunately, the lack of a proper cell model has hampered assessment of calreticulin function in CTLs.
Calreticulin knockout (crt−/−) mice have been generated (24), but they die at the embryonic stage due to impaired heart development, making it impossible to study mature crt−/− CTLs and NK cells. The nature of the cardiac defect observed in crt−/− mice was found to reside in the failure to mobilize ER Ca2+ stores and to activate the serine/threonine phosphatase calcineurin (25). By crossing a transgenic mouse expressing a cardiac-specific, constitutively active form of calcineurin (26) with crt+/− mice, viable crt−/− transgenic mice were obtained. In these animals, normal cardiac development was restored, and embryonic lethality was suppressed (25). Because the expression of active calcineurin occurred only in heart tissue, noncardiac phenotypes specifically due to ablation of the calreticulin gene were not corrected (25). Therefore, this mouse represents an excellent model to study calreticulin functions in systems other than heart. To specifically investigate the role of calreticulin in CTL lytic function, we generated CTL lines from splenocytes derived from these mice. In this study we show that in the absence of calreticulin, CTL cytotoxicity is impaired. Because calreticulin ablation does not seem to affect components of the lytic machinery, the ability of CTLs to degranulate, or their survival during degranulation and attack of target cells, but does affect the efficiency of effector/target conjugate formation, we propose that calreticulin may be involved in the mechanisms underlying target recognition by CTLs and/or formation of the death synapse.
Materials and Methods
Cells and reagents
Mouse mastocytoma P815 cells, mouse lymphocytic leukemia L1210 cells, and mouse lymphoma EL4 cells were obtained from American Type Culture Collection and were maintained in RPMI 1640 medium supplemented with 20 mM HEPES, 50 U/ml penicillin, 50 μg/ml streptomycin, 1 mM sodium pyruvate (Invitrogen Life Technologies), 0.1 mM 2-ME (Sigma-Aldrich), and 10% heat-inactivated FBS (HyClone; RHFM medium). Granzyme A/granzyme B double-knockout mice were generated in our laboratory by crossing granzyme A knockout (27) with granzyme B knockout mice (28) (The Jackson Laboratory) and subsequently inbreeding F1 animals. CTL lines were generated from these mice as described below. The hybridoma 145-2C11 producing anti-CD3ε Ab was obtained from American Type Culture Collection. All reagents were purchased from Sigma-Aldrich unless otherwise specified.
Generation of the crt−/− CTL line
Crt−/− mice expressing a cardiac-specific constitutively active form of calcineurin have been previously described (25). Briefly, homozygote mice bearing the activated calcineurin transgene (FVB/N strain) (26) were crossed with crt+/− mice (C57BL/6J strain) (24). Mice of the resulting progeny were interbred to generate crt+/+ and crt−/− mice expressing activated calcineurin in the heart. CTL lines were generated from these mice by coculture of isolated splenocytes with irradiated (2500 rad) BALB/c splenocytes (1:1 ratio) in RHFM medium supplemented with 80 U/ml rhIL-2 (Chiron) for 3 days. Irradiated stimulators were removed by Ficoll-Paque (Amersham Biosciences) centrifugation, and CTLs were maintained in RHFM medium supplemented with 80 U/ml rhIL-2 for 3 additional days before further stimulation. Subsequent stimulations with irradiated BALB/c splenocytes were performed weekly at a ratio of 1:14 (CTLs:stimulators). CTLs that had been exposed to 4–13 cycles of stimulation were used for the experiments, from 3 to 6 days after stimulation.
CTL cytolytic activity was evaluated as previously described (29), measuring membrane damage (51Cr release) and DNA fragmentation ([3H]thymidine release) induced in target cells. Briefly, allogeneic target cells (106) were labeled with 100 μCi of Na251CrO4 or 1 μCi/ml [3H]thymidine, washed, and incubated at 37°C with effector CTLs at the indicated E:T cell ratio in 96-well plates. Where indicated, Con A was added to the samples at the final concentration of 2 μg/ml. To block perforin activity and to assess the contribution of Fas-mediated killing, CTL were preincubated for 2 h with 500 nM concanamycin A (CMA; Sigma-Aldrich), then added to the target cells in the presence of the inhibitor. 51Cr release in the medium was measured 4 h later using a 1470 Wizard automatic gamma counter (Wallac). DNA fragmentation was measured 2 h after incubation with CTL. Cells were lysed with 1% Triton X-100 in 10 mM Tris and 1 mM EDTA, pH 8.0, and were centrifuged at 10,000 rpm in a microcentrifuge at 4°C. [3H]Thymidine release in the supernatant was measured using an LS7800 beta counter (Beckman Coulter). Maximum [3H]thymidine release was determined by lysis of target cells with 2% SDS in 0.2 M NaOH. The percent membrane lysis and DNA fragmentation were calculated as: [(sample cpm − spontaneous cpm)/(maximum cpm − spontaneous cpm)] × 100. In some experiments CTL were preincubated for 4 h with soluble anti-mouse CD3ε Ab (clone 145-2C11; 1/500 dilution of hybridoma supernatant) before performing the killing assay. Cells were then centrifuged, resuspended in RHFM medium, counted, and incubated with the target cells.
CTL were washed with PBS and fixed in 2% paraformaldehyde in PBS for 15 min at room temperature. After three washes in PBS, cells were permeabilized for 5 min at room temperature with 0.3% saponin in PBS containing 1% FCS. PE-conjugated anti-granzyme B mAb (clone CLB-GB12; PeliCluster) was added to each sample at a final concentration of 5 μg/ml and incubated at room temperature for 30 min. Cells were washed three times with PBS containing 1% FCS and analyzed with a FACScan flow cytometer equipped with CellQuest software (BD Biosciences).
Granzyme B enzymatic assay
Cell lysates were prepared by resuspending CTL in 20 mM Tris-HCl (pH 7.2), 150 mM NaCl, and 1% Triton X-100 at a concentration of 107 cells/ml and incubating for 30 min on ice. Lysates were freeze-thawed once, centrifuged at 10,000 rpm for 7 min to remove cell debris, and stored at −80°C until used. Granzyme B activity in supernatant and total lysates was assayed as previously reported (30). Briefly, aliquots of incubation medium or cell lysates were dispensed into 96-well, flat-bottom tissue culture plates (Nalge Nunc International). Enzymatic reaction was conducted at 37°C in 50 mM HEPES (pH 7.5), 10% (w/v) sucrose, 0.05% (w/v) CHAPS, and 5 mM DTT containing 200 μM acetyl-Ile-Glu-Pro-Asp-paranitroanilide (Kamiya Biomedical) as substrate. Hydrolysis of acetyl-Ile-Glu-Pro-Asp-paranitroanilide was measured at 405 nm using a Multiskan Ascent spectrophotometer (Thermo Lab-system). The absorbance of a blank containing no proteins was subtracted from each sample.
N-α-benzyloxycarbonyl-l-lysine thiobenzyl ester (BLT) esterase activity assay
Serine esterase activity in total cell lysates and medium was measured spectrophotometrically in 96-well, flat-bottom tissue culture plates. Enzymatic reaction was conducted for 30 min at 37°C in PBS containing 0.2 mM BLT and 0.2 mM dithiobis-2-nitrobenzoic acid. Absorbance at 405 nm was read with a Multiskan Ascent spectrophotometer. The absorbance of total cell lysate from granzyme A/granzyme B double-knockout CTLs was subtracted from sample readings. Serine esterase activity was calculated based on the extinction coefficient of dithiobis-2-nitrobenzoic acid (ε412 = 13,600).
CTL were washed in PBS, counted, and resuspended in SDS sample buffer at a concentration of 107 cells/ml. Forty microliters of sample was resolved by electrophoresis on a 12% SDS-polyacrylamide gel under reducing conditions. Proteins were transferred to a polyvinylidene difluoride membrane. The membrane was blocked for 1 h with 5% (w/v) milk/0.1% Tween 20 in TBS, then incubated overnight at 4°C with rabbit anti-perforin Ab (1/1000; Torrey Pines Biolabs). The secondary Ab was HRP-conjugated goat anti-rabbit IgG (Bio-Rad). Visualization of the immunoblot was performed by ECL Plus (Amersham Biosciences).
Cells were washed and resuspended at a concentration of 107 cells/ml in PBS. Fifty microliters of the cell suspension was spotted onto silane-treated Hystobond slides (Marienfeld Laboratory Glassware) and incubated at 37°C for 10 min. PBS was gently withdrawn, and adherent CTL were fixed with 2% paraformaldehyde for 30 min at room temperature. After three washes with PBS, cells were permeabilized with 0.1% Triton X-100 for 3 min and blocked for 1 h with PBS containing 5% donkey serum. Incubation with the primary Abs was performed for 1 h at room temperature. The following Abs were used: mouse monoclonal anti-granzyme B (clone CLB-GB12; 1/50), rabbit anti-perforin (1/50), mouse monoclonal anti-perforin (provided by Dr. G. Griffiths, University College London, London, U.K., and goat anti-lysosomal-associated membrane protein-1), and goat anti-LAMP-1 (1/20; Santa Cruz Biotechnology). Secondary Abs were: donkey anti-rabbit Alexa 555, donkey anti-mouse Alexa 488 and Alexa 594, and donkey anti-goat Alexa 488, all used at 1/400 dilution, except for Alexa 594-conjugated Abs, which were used at 1/200 dilution (Molecular Probes). Slides were mounted with ProLong (Molecular Probes) and analyzed with an LSM510 laser scanning confocal microscope mounted on a Zeiss Axiovert 100M microscope.
Induction of granule exocytosis by immobilized anti-CD3ε Ab
Ninety-six-well, U-bottom plates were coated overnight at 4°C with anti-CD3ε mAb (clone 145-2C11; BD Biosciences) at the indicated dilutions in 0.1 M bicarbonate buffer, pH 9.6. The following day, wells were washed four times with cold PBS and blocked for 30 min at room temperature with 2% FCS in PBS. Blocking solution was then replaced with 100 μl/well CTL suspension in phenol-free RPMI 1640 medium containing 2% FCS (5 × 106 cells/ml). CTL were allowed to degranulate at 37°C for 5 h. After centrifugation at 1200 rpm for 5 min in an Allegra 6R centrifuge (Beckman Coulter), 50 μl of supernatant was transferred to a 96-well, flat-bottom plate for detection of granzyme B, BLT esterase, or lactic dehydrogenase (LDH) enzymatic activity.
LDH released by cells into the medium (50 μl) was measured at room temperature in a reaction volume of 200 μl and in the presence of 0.4 mM sodium pyruvate and 0.2 mM NADH. The kinetics of the enzymatic reaction were monitored spectrophotometrically, and the rate of absorbance decrement at 334 nm over 2 min was used to calculate LDH activity in the samples. Total LDH activity was measured in fresh cell lysates and used to derive the percentage of LDH released into the medium.
Analysis of conjugate formation
CTL were labeled with 2 μM Orange Cell Tracker (Molecular Probes), and P815 cells were labeled with 1 μM Green CellTracker (Molecular Probes) according to the manufacturer’s instructions. Labeled CTLs were left resting for 3 h at 37°C in regular medium or in the presence of soluble anti-CD3ε Abs (1/500), then washed in cold phenol-free HBSS. After resuspension in cold phenol-free HBSS containing 5% FCS, CTL and target cells were mixed at the ratio of 1:3 and centrifuged at 1000 rpm in an Allegra 6R centrifuge (Beckman Coulter) for 3 min at 4°C to facilitate cell-cell contact. Cells were then incubated at 37°C for the indicated time to allow conjugate formation. At the end of the incubation, samples were gently vortexed for 2 s and fixed by addition of an equal volume of 4% paraformaldehyde solution. Analysis of double-labeled cell conjugates was performed with a FACScan flow cytometer.
Data were analyzed by two-tailed Student’s t test for paired samples. A value of p < 0.05 was considered significant.
Crt−/− CTLs exert reduced cytotoxic activity on target cells
Crt+/+ and crt−/− CTL lines were generated from transgenic mice expressing a constitutively active form of calcineurin in the heart. Because the expression of the transgene was under control of the cardiac specific α-myosin H chain promoter, we expected the transgene to be silent in the CTL lines. Western blotting analysis confirmed that the transgene was not expressed at any time in the CTLs (data not shown).
The cytolytic potential of the crt+/+ and crt−/− CTLs was assessed in in vitro killing assays in which the two CTL lines were challenged with allogeneic target cells, the mouse mastocytoma cell line P815 and the lymphocytic leukemia L1210 line (Fig. 1). Compared with wild-type cells, crt−/− CTLs showed reduced cytotoxicity toward target cells at all E:T cell ratios tested (shown in Fig. 1,A for L1210 cells; killing of P815 targets occurred with a similar trend). Both membrane lysis (Fig. 1,B) and DNA fragmentation (Fig. 1 C) in target cells exposed to the CTL attack were measured. In all cases the cytotoxic activity of crt−/− CTLs was significantly reduced with respect to that of wild-type cells. The difference was limited to the perforin-dependent killing mechanism, rather then Fas-mediated killing, because in the presence of CMA (31) cytotoxicity was greatly reduced, with no significant differences between the residual killing activities of the crt+/+ and crt−/− CTLs.
It has been previously demonstrated that the level of granzyme B activity released by CTLs into the medium during incubation with target cells strictly correlates to the levels of degranulation and killing (30). As expected, the amount of granzyme B released by crt−/− CTL during incubation with target cells was 50 ± 23% (p = 0.0032) lower than the amount secreted by wild-type cells. In light of these data, three possible explanations could be considered for the reduced cytotoxicity of crt−/− CTLs: 1) CTL death during the killing assay, 2) lower expression of granzymes and perforin, or 3) inefficient target-induced degranulation.
Crt−/− CTLs are viable during the attack of target cells
Because it has been proposed that calreticulin may have a protective role on CTLs against their own perforin (21), we investigated whether the lower level of killing observed with crt−/− CTLs could be attributed to CTL death during incubation with target cells. Labeling of CTLs with Na251CrO4 allowed us to monitor CTL death during the killing assay. No difference in CTL survival was observed (Fig. 1 D). CTL death was as low as 6–7% for both crt+/+ and crt−/− cells and was comparable to the level of death observed in the absence of target cells (spontaneous death), indicating that CTLs were viable during the assay.
Crt−/− CTLs have normal granzyme content and express high levels of perforin
To test the second hypothesis, we measured granzyme and perforin expression in crt−/− CTLs. As shown in Fig. 2, granzyme B expression, as measured by FACS staining, was not significantly different between crt+/+ and crt−/− cells (Fig. 2,A). Moreover, the enzymatic activity of granzyme B was comparable, suggesting that in the absence of calreticulin, granzyme B can still fold properly without loss of activity (Fig. 2,B). Crt+/+ and crt−/− CTLs also contained the same amount of serine esterase (granzyme A) activity (Fig. 2,C). In contrast, levels of perforin expression were higher in crt−/− CTLs than in wild-type cells (Fig. 2,D). No differences in granzyme B and perforin intracellular localization were evident by confocal microscopy, as assessed by colocalization with lysosomal-associated membrane protein-1, a lysosomal marker also present in lytic granules (32) (Fig. 3).
Crt−/− CTLs degranulate more efficiently than crt+/+ cells when stimulated with immobilized anti-CD3ε Ab
To investigate whether impaired transduction of the signal from the TCR-CD3 complex could account for the reduced cytotoxicity observed in the absence of calreticulin, we triggered activation of the TCR-CD3 complex and exocytosis of lytic granules by exposing CTL to immobilized anti-CD3ε Ab (33, 34). In these experimental conditions, the crt−/− CTLs degranulated and released granzyme B and BLT esterase activity efficiently (Fig. 4). Although at a saturating concentration (10 μg/ml) of immobilized Ab, the release of granzyme B and BLT esterases was comparable in crt+/+ and crt−/− CTLs, at lower Ab concentrations crt−/− cells secreted even more granzyme B than wild-type cells (Fig. 4 A). CTL death during the 5 h of incubation was <7% of the total cell number, as assessed by measuring LDH release into the medium, with no significant difference between cell lines or between control (no Ab) and stimulated samples (data not shown). Therefore, the higher secretion of granzymes by crt−/− cells was due to specific triggering of TCR-CD3 signaling by immobilized anti-CD3 Ab.
Effector/target conjugate formation is impaired in crt−/− CTLs, and cytotoxicity is restored in conditions that favor cell adhesion
Impaired cytotoxicity may result from defective recognition and/or adhesion to target cells. Indeed, we found that crt−/− CTLs were less efficient than wild-type cells in forming conjugates with target cells (Fig. 5,A). If reduced conjugate formation (or stability) was the cause of the low levels of killing observed with crt−/− CTLs, we expected to see increased cytotoxicity after treatments and in conditions that enhance cell adhesion. Therefore, we prestimulated CTLs with soluble anti-CD3 Abs before washing and exposing them to target cells. Soluble anti-CD3 Abs do not induce CTL degranulation (34), but are known to activate adhesion of CTLs to substrates and target cells (35). As expected, antiCD3-stimulated crt−/− CTLs became indistinguishable from wild-type cells with respect to the number of conjugates formed (Fig. 5,B) and the ability to kill target cells (Fig. 6, A and B). The restored cytotoxicity was mainly perforin-dependent, because killing was greatly reduced (from >80 to 17%) by blocking perforin activity with CMA (data not shown). The levels of granzyme B released into the medium correlated with the killing of target cells (Fig. 6 C), further demonstrating that restoration of the crt−/− CTL activity was due to the increased ability to degranulate upon recognition of target cells.
Similar results were obtained when the killing assay was performed in the presence of Con A (Fig. 7,A). Con A is known to induce lectin-dependent, cell-mediated cytotoxicity, a process in which target cells are killed by CTLs in an Ag-independent fashion. This is achieved through the bridging action of Con A that binds to glycoproteins on the surface of both effectors and target cells and at the same time cross-links the TCR on the CTL membrane (36, 37, 38, 39, 40). In the presence of Con A, crt−/− CTL cytotoxicity increased significantly, allowing for the killing of allogeneic targets (P815) at levels close to those observed with wild-type CTLs. Moreover, Ag-independent cytotoxicity toward syngeneic target cells (EL-4) was observed for both crt+/+ and crt−/− CTLs (Fig. 7 B).
Previous studies have identified calreticulin, an ER-resident chaperone and a Ca2+-binding protein, as a component of CTL and NK cell lytic granules (15). Upon activation of CTLs, calreticulin gene expression is up-regulated (23), and the protein is targeted to lytic granules (20) and the plasma membrane (41). Although several studies have reported on functions played by the protein outside the ER (reviewed in Refs.16 and 17), there is no information on its role in the lytic granules and CTL physiology. In this regard, our study provides the first direct evidence of calreticulin involvement in the cytotoxic activity of CTLs. Because the calreticulin knockout is embryonically lethal (24), we generated a CTL line from crt−/− mice that were able to survive due to the ectopic expression of active calcineurin in the heart (25). The transgenic constitutively active form of calcineurin was not expressed in tissues other than heart, allowing for the specific analysis of calreticulin deficiency in CTLs. The crt−/− CTLs were healthy, providing evidence that calreticulin function is dispensable during development and maturation of CTLs. Our study also indicates that calreticulin is not necessary to protect CTLs from lytic activity of stored perforin, a hypothesis that was proposed when the ER chaperone was first identified in the CTL granules in association with perforin (15).
The most striking phenotype of crt−/− CTLs was their markedly reduced cytotoxicity compared with wild-type cells. Levels of killing correlated with levels of granzyme B exocytosis during incubation with target cells, indicating that the nature of the problem could reside in the amount of degranulation occurring in response to the target cells, rather than ineffectiveness of the lytic material per se. Indeed, we found that expression, activity, and intracellular localization of granzymes were comparable in wild-type and crt−/− cells. Interestingly, perforin levels were substantially higher in the crt−/− CTLs. Several hypotheses could be formulated to explain these data. Higher protein content may result, for example, from lack of protein folding quality control in the absence of the ER chaperone calreticulin, with consequent accumulation of misfolded proteins that would otherwise be degraded. Alternatively, higher gene expression or increased protein half-life (or decreased protein turnover) could account for the increased perforin content in the crt−/− CTLs. Whether all the perforin molecules produced by crt−/− CTLs were correctly folded is not known. However, we believe that at least an amount of active perforin comparable with the wild-type content must have been present in the crt−/− CTLs, otherwise we would not have been able to restore cytotoxicity by treatment with aCD3 or Con A. The data on perforin are intriguing, and they suggest that the lack of calreticulin may generate a multifaceted phenotype in CTLs. However, the higher levels of the pore-forming protein in the crt−/− CTLs did not seem to have an impact on CTL viability or their cytolytic function.
It has been proposed that calreticulin may be involved in sorting granule components, perforin in particular, to the lytic granules (20). However, our data argue against such a hypothesis, because intracellular localization of perforin and granzyme B in crt−/− CTLs appeared to be similar to that in wild-type cells.
Calreticulin has been shown to block the lysis of erythrocytes induced by perforin (21, 22). Consequently, it was proposed that calreticulin might bind to the CTL surface upon degranulation, preventing polymerization of perforin within the CTL plasma membrane and consequent lysis of the effector cells. If that hypothesis were correct, one would have expected to observe crt−/− CTL lysis during the killing of target cells and overall reduced target killing. However, when we measured crt−/− CTL lysis during incubation with target cells, we did not find any specific increase in CTL death relative to wild-type cells or with respect to the levels of cell death spontaneously occurring in CTL populations in the absence of targets. Similar results were obtained when CTL lysis was measured during immobilized anti-CD3 Ab-induced degranulation. These data argue against a protective role of calreticulin on CTLs. Other molecules must be involved in CTL protection, such as the recently identified cathepsin B (42).
Because crt−/− CTLs secreted less granzyme B than wild-type CTLs in response to stimulation by target cells, we investigated the hypothesis that aberrant transduction of the signal downstream of the TCR-CD3 complex could result in impaired degranulation. When we triggered granule exocytosis by incubation with immobilized anti-CD3 Abs, thus bypassing the need for recognition and contact with target cells, crt−/− CTLs responded better than wild-type cells, showing a lower threshold for maximal induction of degranulation. Granule exocytosis is triggered in CTLs by initial release of Ca2+ from intracellular stores, followed by sustained influx of extracellular Ca2+ through a store-operated Ca2+ channel (reviewed in Ref.43). In our study calreticulin was dispensable for induction of degranulation when this was triggered by CD3 cross-linking. In line with our observations are data showing that overexpression of calreticulin decreases store-operated Ca2+ influx through a mechanism that seems to be independent from its Ca2+-buffering activity (44).
The remaining possible explanation for the reduced cytotoxic activity of crt−/− CTLs is impaired recognition and/or adhesion to target cells. As a matter of fact, we found that when crt−/− CTLs were mixed with target cells, they formed less (or less stable) effector-target conjugates than wild-type CTLs. To test whether this could be taken into account for the reduction of cytotoxicity in crt−/− CTLs, we used prestimulation with soluble anti-CD3 Abs (35) or addition of Con A (36, 37, 38, 39, 40) as experimental tools to increase the formation of effector:target conjugates during killing assays. Under these conditions, crt−/− CTL granule-mediated cytotoxicity was restored. These results suggest that the lytic machinery of crt−/− CTLs (granzymes and perforin) is functional and has the potential to kill target cells, provided that the contact between effectors and targets is efficiently established.
The effects of calreticulin on recognition and/or binding to target cells are currently under investigation. Target recognition could be affected due to reduced expression of TCR. However, this did not seem to be the case in crt−/− CTLs, because they expressed similar or slightly higher levels of CD3 than wild-type cells. We cannot exclude the possibility that protein misfolding or aberrant post-translational modification in calreticulin-deficient cells may alter the Ag recognition ability of the TCR. Alternatively, other mechanisms involved in the adhesion to target cells may be affected by calreticulin deficiency.
One of these mechanisms involves the family of integrin proteins, specifically LFA-1 (45, 46) and β1 and β3 integrins (47), which have been shown to promote CTL adhesion and TCR-mediated cytotoxicity (for review, see Ref.48). The idea that integrins may be a target of calreticulin deficiency is particularly intriguing because it is supported by studies showing calreticulin involvement in cell adhesion and integrin function in different cell types (49, 50, 51, 52, 53). In addition, anti-CD3 Abs, which restored crt−/− CTL cytotoxicity in our study, are known to enhance LFA-1-dependent adhesion of CTLs to target cells (54). Calreticulin also binds thrombospondin (55, 56, 57), an extracellular matrix protein with general antiadhesive effects (58) that is involved in the regulation of processes such as T cell activation, adhesion, and homeostasis (reviewed in Ref.59). Future studies will address the hypothesis that calreticulin deficiency may alter the integrins and/or thrombospondin-mediated adhesive properties of CTLs.
In conclusion, we have shown that despite functional lytic machinery, cytotoxicity is impaired in calreticulin-deficient CTLs, probably due to inefficient effector-target conjugate formation. Such a deficit could have important implications in the regulation of the immune response in vivo. In light of our data, it is tempting to speculate that calreticulin deficiency could result in reduced or delayed clearance of infected and transformed cells. At the same time, it could severely affect lymphocyte homeostasis and result, as do several genetic defects impeding degranulation of CTLs or inactivating perforin (60, 61), in hemophagocytic-like syndromes or autoimmune diseases. Interestingly, autoantibodies against calreticulin have been found in several autoimmune conditions (62, 63, 64, 65, 66, 67).
The authors have no financial conflict of interest.
We thank Dr. Gillian Griffiths for the gift of the mouse monoclonal anti-perforin Ab. We also thank Drs. Hanna Ostergaard and Ing Swie Goping for helpful discussion.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by grants from the Canadian Institute of Health Research and the National Cancer Institute of Canada. R.C.B. is a Canadian Institute of Health Research Distinguished Scientist, a Medical Scientist of the Alberta Heritage Foundation for Medical Research, a Howard Hughes International Scholar, and a Canada Research Chair. M.M. is a Canadian Institute of Health Research Senior Scientist and a Medical Scientist of the Alberta Heritage Foundation for Medical Research. S.S. was supported by a research fellowship from the Alberta Heritage Foundation for Medical Research.
Abbreviations used in this paper: ER, endoplasmic reticulum; BLT, N-α-benzyloxycarbonyl-l-lysine thiobenzyl ester; CMA, concanamycin A; LDH, lactic dehydrogenase; crt+/+, calreticulin wild type; crt−/−, calreticulin knockout.