Expression and Function of Synaptotagmin VII in CTLs¹

Abs, reagents, and cell lines

The rat mAb recognizing granzyme A (7.1) was a gift from M. Simon (Max Planck Institute, Freiburg, Germany). The affinity-purified rabbit Ab against perforin was purchased from Santa Cruz Biotechnology. Rabbit polyclonal Abs against the NH₂ terminal of Syt VII were generated and affinity purified, as previously described (6). The rat mAb against Lamp-1 (ID4B) was from the Developmental Studies Hybridoma Bank. Secondary Alexa Fluor-488-labeled goat anti-rat IgG, Alexa Fluor-546 goat anti-rabbit, and mouse IgG were purchased from Molecular Probes; secondary HRP-conjugated anti-rabbit and anti-rat IgG Abs were from Kirkegaard & Perry Laboratories. Purified mAb anti-CD3ε (145.2C11) was obtained from BD Pharmingen, and mAbs against talin were from Sigma-Aldrich. The OVA_257–264 peptide (SIINFEKL) was synthesized and purified by the W. M. Keck Biotechnology Resource Center (Yale University).

P815 and EL4 mouse target cell lines (American Type Culture Collection; a gift from M. Starnbach, Harvard Medical School, Boston, MA) were maintained in RPMI 10/5% FBS with penicillin/streptavidin (pen/strep), Na pyruvate, and minimum essential amino acids at 37°C/5% CO₂. Normal rat kidney (NRK) cells were maintained in DMEM/10% with pen/strep.

Mice

Syt VII-deficient mice were generated, as previously described (11), and backcrossed for eight generations into the C57BL/6 background. Seven- to 8-wk-old female B6D2F₁ mice were used as a source of mitomycin C-treated allogeneic stimulators (Charles River Laboratories) (16). For some experiments, Syt VII wild-type (WT) and knockout (KO) mice were crossed to OT-1 transgenic mice (provided by F. Carbone, University of Melbourne, Melbourne, Australia) to generate Syt VII^+/+ OT-1⁺ and Syt VII^−/− OT-1⁺ mice specific for the OVA_257–264 (SIINFEKL) epitope of OVA (17). OT-1⁺ mice were screened by retro-orbital bleeding and flow cytometry after staining with CD8- and TCRVα5-specific mAbs, and by tail DNA PCR screening using the OT-1-specific sequence using primers: 5′-GCAGCAAGTGACAACTATCAGTTG-3′ (forward) and 5′-CAGATC TCAACTGGACCACAG-3′ (reverse). Syt VII^−/− mice were identified by tail PCR screening using primers spanning the Syt VII C2A region, as follows: 5′-CGATGAGTCTGACCGCAGAAC-3′ (forward) and 5′-CCA CCTTACTACCATCTCTGC-3′ (reverse). All mice were maintained at the Yale Animal Resources Center.

Listeria infections

WT (C57BL/6) and Syt VII-deficient animals (8 –10 wk of age) were primed i.v. with 1 × 10⁴ CFU of rOVA-expressing L. monocytogenes strain JJL-OVA (provided by H. Shen, University of Pennsylvania, Philadelphia, PA). Three weeks later, Ag-primed animals received a second challenge with 2 × 10⁵ CFU of JJL-OVA. Five days postchallenge, animals were sacrificed and single-cell suspensions were prepared from the spleens. The number of SIINFEKL-specific CD8 T cells was determined by flow cytometry. Cells were stained with Abs to CD11a and CD8 (both from BD Pharmingen), followed by PE-conjugated control and SIINFEKL peptide-loaded H-2K^b-IgG dimer (BD Pharmingen). Splenic bacterial burden was determined, as previously described (18). In brief, serial dilutions of splenic homogenates were resuspended in PBS containing 0.05% Triton X-100 and plated in triplicate onto brain-heart infusion agar plates and evaluated 48 h after culture at 37°C.

In vitro T cell stimulation

Spleens or mesenteric lymph nodes were removed from Syt VII KO mice and homogenized between frosted glass slides into a single-cell suspensions. RBC were lysed in 1.66% NH₄Cl. CTLs were activated by stimulation of splenocytes from B6D2F₁ mice with mitomycin C in RPMI 10 medium containing 10% FBS, 10% T-Stim medium (BD Discovery Lab-ware), 5% 1 M methyl α-D-glucopyranoside (Sigma-Aldrich), and pen/strep. After 5 days of stimulation, proliferating CTLs were isolated using lymphocyte separation medium (Fisher Scientific) before assays were performed. OVA-specific CTLs were generated in vitro by stimulation of splenocytes with 0.5–1 μg/ml OVA_257–264 peptide (SIINFEKL) in RPMI 10 medium containing 10% FBS and pen/strep. Cells were stimulated for 48 h in the presence of peptide, then washed and cultured in the presence of rIL-2 (20 U/ml) for 24 h before functional assays. CD8⁺ CTLs were purified using the MACS MS CD8⁺ separation columns (Miltenyi Biotec), according to the manufacturer’s instructions.

RT-PCR

mRNA was prepared from homogenized whole spleens or day 5-stimulated CTLs in TRIzol. RT-PCR was performed using primers spanning the Syt VII C2A domain as follows: 5′-CGATGAGTCTGACCGCAGAAC-3′ (forward) and 5′-CCACCTTACTACCATCTCTGC-3′(reverse).

Immunoblotting

A total of 2–3 × 10⁶ CTLs was lysed in lysis buffer with protease inhibitors on ice for 20 min. Cell lysates were run on SDS-PAGE gels, and the protein was transferred to nitrocellulose for 1 h. Membranes were blocked in 5% milk/TBS Tween 20 overnight and probed with anti-Syt VII, anti-perforin, or anti-granzyme A Abs for 1 h, before washing three times with TBS/Tween 20 buffer. Membranes were then incubated in secondary HRP-labeled Abs (anti-rabbit, anti-rat) for 1 h, followed by ECL analysis.

Immunofluorescence

Day 5- and 6-stimulated CTLs were used for immunostaining. Cells in suspension were added to poly-L-lysine-coated coverslips at 37°C for 25 min before fixing with 3.7% formaldehyde at 4°C. After washing with PBS, the cells were quenched with 50 mM ammonium chloride for 15 min at room temperature, and coverslips were washed once and permeabilized with 0.1% Triton X-100 for 2 min before blocking in 3 mM glycine/PBS/1% BSA solution for 30 – 45 min. The cells were then stained with mAbs specific for Lamp-1 or granzyme A, and also for Abs against the NH₂ terminus of Syt VII. Following staining with conjugated secondary Abs, the DNA of cells was stained for 1 min with 4′,6′-diamidino-2-phenylindole (DAPI; Sigma-Aldrich) and diluted to 5 μg/ml in PBS/2% BSA. Coverslips were mounted in Prolong antifade medium (Molecular Probes) and observed under a Zeiss Axiovert 135 epifluorescent microscope.

Granzyme activity assays

Granzyme A activity was assayed by incubation of 950 μl of benzyloxy-carbonyl-L-lysine thiobenzyl ester (BLT) substrate solution (0.20 mM N-α-BLT (Calbiochem), 0.22 mM 5,5′-dithio-bis-2-nitrobenzoic acid (Sigma-Aldrich), PBS, 0.01% Triton X-100) with 50 μl of culture supernatant, or of total cell lysates in 1% Triton X-100 for 60 min at 37°C. The reaction was stopped by adding 1 ml of cold 1 mM PMSF/PBS, and the absorbance was read at 412 nm. β-hexosaminidase was assayed by incubation of 10 μl of supernatant with 50 μl of B-hex substrate (6 mM 4-methyl-umbelliferyl-N-acetyl-β-D-glucosaminide in 0.1 M citric acid, 0.2 M Na₂HPO₄ (pH 4.5), Triton X-100) for 15 min at 37°C. A total of 100 μl of Stop buffer (200 mM Na₂CO₃, 100 mM glycine) was added before reading the fluorescence at excitation 365 nm and emission 450 nm. Percentage of total secretion was calculated as (OD_supernatant/OD_lysate) × 100. All assays were performed in triplicate.

Immunological synapse formation

Day 5- and 7-stimulated CTLs were pulsed for 3 h with 10 kDa of dextran-Texas Red in serum-free medium and washed extensively, before chasing for 3 h with RPMI 10. A total of 5 × 10⁵ dextran-Texas Red-pulsed CTLs was mixed with an equal number of P815 cells in 1 ml of RPMI 10. Cells were centrifuged at 100 × g for 5 min at room temperature and resuspended carefully with a Pasteur pipette. The 100-μl aliquots (1 × 10⁵ cells) were then plated on poly(L-lysine)-coated coverslips and placed at 37°C/5% CO₂ for 20 – 45 min before fixing in 3% paraformaldehyde for 25 min at 4°C. Cells were subsequently permeabilized for 5 min with 0.1% Triton X-100, washed with PBS, and blocked for 1 h with 1% BSA/PBS. Polyclonal mouse anti-talin Abs (Sigma-Aldrich) were added in the presence of 1% BSA for 1 h and washed extensively in 1% BSA/PBS. Alexa Fluor-488 goat anti-mouse secondary Abs were added in the presence of 1% BSA for 30 min and washed in 1% BSA/PBS, and nuclei were stained with DAPI.

Cytotoxicity assays

Syt VII^+/+ or Syt VII^−/− CTLs were mixed with 10⁴ P815 (H2^d) or EL4 (H2^b) target cells at the indicated E:T ratios per well in a final volume of 100 μl in RPMI 10/5% FBS without phenol red. For OT-1 experiments, target cells were coated with or without SIINFEKL peptide (5 μg/ml) for 1 h at 37°C and washed three times with cold medium before mixing with CTLs. After a 4-h incubation at 37°C, cytotoxicity was measured via lactate dehydrogenase activity release from targets, using the CytoTox96 non-radioactive kit, according to the manufacturer’s protocol (Promega). For inhibitor experiments, CTLs were pretreated with 0.25 μM concanamycin A (Alexis Pharmaceuticals) 2 h before being added to targets in the continued presence of the drug. Absorbance was read at 490 nm, and percentage of cytotoxicity was calculated, as follows: (experimental –effector spontaneous – target spontaneous)/(target maximum – target spontaneous) × 100.

FACS analysis

A total of 10⁶ cells was washed with serum-free medium and 1% BSA; stained with CD8 PerCP, CD4 PE, TCRβ FITC, or granzyme A FITC (all from BD Pharmingen) on ice for 20 min in the dark; washed three times with cold PBS; and fixed with 2% paraformaldehyde for 20 min on ice. To evaluate Ag-specific CD8 T cell responses, 10⁶ spleen cells were stained with dimeric K^b (BD Biosciences) loaded with the immunodominant MHC class I-restricted OVA-specific peptide SIINFEKL, as directed by the manufacturer. Samples were then incubated with FITC-labeled anti-mouse IgG to detect dimer staining along with CD11a FITC and CD8 PerCP (all from BD Biosciences) and evaluated on a FACScan flow cytometer.

Splenocyte proliferation assays

A total of 5 × 10⁵ harvested splenocytes was plated into 96-well flat-bottom microtiter plates in 100 μl of RPMI 10 containing soluble anti-CD3 at indicated concentrations. The cultures were incubated for 72 h at 37°C/5% CO₂ before proliferation was measured using a nonradioactive 3-(4,5 dimethylthiazol-2yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) (inner salt) assay (CellTiter96; Promega), according to the manufacturer’s instructions. In brief, 20 μl of the substrate solution was added to wells and incubated at 37°C with 5% CO₂ for 1– 4 h to allow for full colorimetric change to occur. Absorbance 490 nm was read with a Bio-Rad Spectramax ELISA plate reader. All data are shown as means of quadruplicate wells.

Secretion assays

Secretion assays were performed essentially as previously published (19). In brief, round-bottom 96-well plates (Falcon) were coated with 5 μg/ml anti-CD3 Ab (145.2C11) for 30 min at room temperature. Wells were washed three times with 100 μl of RPMI 10 medium before plating 5 × 10⁵ CTLs/50 μl in Ab-coated or uncoated wells. A total of 50 μl of medium was then added to all wells. Total granule release was determined by replacing 10 μl of medium with 10 μl of 1% Triton X-100. Plates were incubated for indicated times at 37°C/5% CO₂ before centrifuging for 5 min at 200 × g at 4°C and collecting the supernatants for granzyme A activity assays.

Syt VII is expressed by activated CTLs and present on lytic granules

Syt VII colocalizes with lysosomal markers, and regulates the fusion of conventional lysosomes with the plasma membrane in several cell types (6 –10, 12, 20, 21). These findings suggested that Syt VII might also be involved in the functional regulation of lytic granules of CTLs, which are known to have lysosomal properties and to be secreted in response to elevations in [Ca²⁺]_i (3, 4).

To examine the role of Syt VII in T cells, the expression of Syt VII was examined in stimulated CTLs from C57BL/6 mice. Western blot analysis using an affinity-purified polyclonal Ab against 65-kDa band on lysates the NH₂ terminus of Syt VII identified a ∼ of NRK cells (as previously described in Ref. 6) and of spleen-derived CD8⁺ CTLs (Fig. 1A, left panel). This 65-kDa band corresponds to the expected size of Syt VIIα, the most abundantly expressed isoform of Syt VII in several tissues (22). Because the lysosomal components of CTL lytic granules are primarily expressed upon activation (23), the expression of Syt VII was then examined in unstimulated and stimulated CD8⁺ T lymphocytes. For this experiment, the T cells were isolated from lymph nodes, to ensure the presence of unstimulated CD8⁺ CTLs (the spleen normally contains higher percentages of Ag-experienced cells). When unstimulated and allogeneic-stimulated lymph node CD8⁺ T cells were analyzed, expression of Syt VII was detected only in the stimulated population (Fig. 1A, right panel). Thus, the expression pattern of Syt VII is similar to that of other secretory lysosome components such as granzyme A and perforin, which are up-regulated only after activation (24).

The subcellular localization of Syt VII in CTLs was then evaluated by immunofluorescence. To identify the secretory lysosomes/lytic granules, the same cells were also labeled with Abs against granzyme A. As shown in Fig. 1B, an almost complete colocalization of granzyme A-positive granules with Syt VII-positive compartments was seen. No reactivity was observed following similar staining with preimmune serum. The Syt VII-positive compartments also colocalized with the late endosome/lysosome marker Lamp-1 (Fig. 1C), further suggesting that Syt VII localizes to secretory lysosomes/lytic granules of CTLs.

Phenotype and function of T cells from Syt VII-deficient mice

To confirm the lack of Syt VII expression in CTLs of Syt VII-deficient mice, as previously shown for fibroblasts, macrophages, and skeletal muscle (10), we isolated total RNA from whole spleens and purified CTLs of WT and Syt VII KO mice. RT-PCR analysis using primers corresponding to the C2A domain of Syt VII identified a 500-bp band in the WT splenocyte population and in purified CTLs, but not in similar preparations from Syt VII KO mice (Fig. 2A).

Compared with their WT littermates, Syt VII-deficient mice are characterized by slightly reduced numbers of splenocytes (5.48 × 10⁷ ± 0.97 × 10⁷ in KO vs 6.49 × 10⁷ ± 1.73 × 10⁷ in WT). Given the observed reduction of total splenocytes, the phenotype and composition of spleen cells in Syt VII-deficient animals were examined. The number of splenic CD4⁺ and CD8⁺ T cells was determined by FACS analysis, after harvesting unstimulated cells or after culturing unstimulated cells for 5 days in the presence of allogeneic stimulator cells. No significant differences were detected in the overall composition of Syt VII KO T cell splenocyte populations, in comparison with WT. After stimulation in culture for 5 days, Syt VII WT and KO splenocytes contained an equivalent fraction of CD4⁺ (9 vs 8%) and CD8⁺ (89.2 vs 89.8%) T cell populations (Fig. 2B). Taken collectively, these results suggest that the expression of Syt VII is not required for the development of CD4 or CD8 T cells in vivo.

We then examined the CD8⁺ T cell populations of WT and KO mice for the expression of two important effectors of lytic granule function, perforin and granzyme A. Western blots were performed on lysates from WT and KO CTLs with Abs to these granule products. Abs against granzyme A or perforin detected bands of 30 and 65 kDa, respectively, in both WT and KO CTLs (Fig. 2C). Thus, both of these secretory products showed no change in expression in response to the absence of Syt VII. Similarly, when the total amount of BLT esterase activity (granzyme A) was determined from Triton X-100 CTL lysates, nearly identical levels were observed (Fig. 2D), eliminating the possibility that Syt VII deficiency affects the production or storage of this cytotoxicity effector molecule. Likewise, when the granule morphology was studied by immunofluorescence staining with anti-granzyme A, no apparent differences in granule size or number were seen in CTLs from Syt VII KO mice, when compared with WT CTLs (data not shown). The fraction of CD8⁺ T cells expressing granzyme A after activation was also similar for WT and KO mice (data not shown), and no differences were detected in the capacity of splenocytes to proliferate following stimulation of allogenic mitomycin C-treated splenocytes (Fig. 2E) or in response to TCR mediated following cross-linking with anti-CD3 Abs (data not shown). Secretion of the cytokines IFN-γ and IL-2 was also comparable between WT and KO splenocytes, following TCR cross-linking stimulation (data not shown).

Syt VII-deficient mice exhibit impaired CD8 effector activity in vivo

To evaluate the effector activity of Syt VII-deficient CTLs in vivo, mice were examined for their ability to resolve an infection with L. monocytogenes. L. monocytogenes is a Gram-positive, facultative intracellular bacterial pathogen. Following infection, the clearance and subsequent immunity to this pathogen are mediated predominantly by Ag-specific CD8 T cells (18). To examine Ag-specific CD8 T cell effector activity, WT and Syt VII KO animals were immunized with PBS or 1 × 10⁴ CFU of OVA-expressing recombinant L. monocytogenes strain JJL-OVA, to prime Ag-specific T cells responses. Three weeks later, animals received a secondary challenge with 2 × 10⁵ CFU of JJL-OVA. Four days following the second challenge, animals were examined for Ag-specific CD8 T cell responses, and for bacterial burden in the spleen. Ag-specific T cell responses to the OVA immunodominant H-2K^b-restricted peptide SIINFEKL were evaluated by flow cytometry. Analysis of CD8 T cells revealed comparable percentages (Fig. 3A) and absolute numbers (Fig. 3B) of Ag-specific CD8⁺ T cells in both primed WT and Syt VII-deficient animals. As expected, no Ag-specific SIINFEKL CD8 T cell responses were observed in either group when primed with PBS alone (Fig. 3, A and B). These results demonstrate that Syt VII-deficient animals successfully generate Ag-specific CD8 T cell responses that are equivalent to littermate controls. However, despite similar numbers of Ag-specific CD8 T cells, primed Syt VII-deficient animals exhibited significantly higher bacterial burden (Fig. 3C), suggesting an impairment in CD8⁺ T cell-mediated pathogen clearance.

Syt VII-deficient CTLs form normal immunological synapses, but are defective in killing target cells

The data described above indicated that several properties of the CD8⁺ T cells of Syt VII-deficient mice appeared to be normal, but the animals were not capable of clearing L. monocytogenes infections as efficiently as the WT controls. To gain insight into the mechanism responsible for this defect, we proceeded to examine the interaction of isolated CTLs with target cells in vitro. We initially determined the capacity of Syt VII WT and Syt VII KO CTLs to form immunological synapses, a process that involves a marked polarization of the lytic granules within a ring containing the cytoskeletal protein talin and other adhesion proteins, at the site of contact between CTLs and target cells (25). To be able to visualize the secretory lysosomes/lytic granules by fluorescence microscopy, we loaded the lysosomal compartments of WT and KO CTLs with 10-kDa dextran-Texas Red by a 3-h pulse-chase incubation at 37°C. These dextran-loaded CTLs were then incubated with P815 target cells, and allowed to adhere to coverslips before fixation and staining with Abs against talin. In conjugates formed between Syt VII WT CTLs and their targets, the expected talin enrichment at the synapse site was detected, in the typical ring morphology seen at the site of tight contact between effector and target. The dextran-labeled granules were tightly polarized at the talin ring site, in direct opposition to the target cell membrane (Fig. 4, A–C). In CTLs derived from Syt VII KO mice, a similar granule polarization at the talin-enriched synapse site was detected (Fig. 4, D–F). In both populations of conjugates, the majority of the secretory lysosomes in the CTLs appeared to be polarized at the target cell contact site. A few dextran-positive vesicles that were more dispersed were also observed, in both WT and KO CTL/target cell conjugates (Fig. 4, B and F). Thus, these results suggest that Syt VII-deficient CTLs are not defective in granule polarization, or in the ability to form an immunological synapse when in contact with a target cell.

As described above, Syt VII is specifically up-regulated upon CTL activation, and it localizes to secretory lysosomes/lytic granules along with the secreted effector molecules responsible for cytotoxicity. Genetic ablation of Syt VII does not affect the composition of lytic granules, the capacity of CTLs to expand in response to activation, and to form immunological synapses containing polarized granules. However, when CTLs lacking Syt VII were examined in OVA_257–264 (SIINFEKL) peptide-specific cytotoxicity assays (after crossing the Syt VII KO mice into the OT1⁺ background), a marked defect in the ability to lyse peptide-coated target cells was observed (Fig. 5A). Similar observations were made in allogeneic T cell cytotoxicity assays. Following a 4-h incubation with P815 target cells, CTLs from WT mice lysed ∼60% killing of targets, at an E:T ratio of 10. In contrast, under the same conditions, the cytotoxicity levels of Syt VII KO CTLs were 30 – 40% reduced in relation to WT controls (Fig. 5B). This cytotoxicity defect was reproducible in several independent experiments (an ANOVA paired t test performed on the results of seven independent cytotoxicity experiments gave p values of 0.006, 0.079, and 0.010 for E:T ratios of 10, 1, and 0.1, respectively, when the cytotoxicity of Syt VII KO CTLs was compared with WT controls). Thus, CTLs from Syt VII-deficient animals exhibit markedly impaired target cytolysis in response to both Ag-specific and allogenic recognition in vitro.

The localization of Syt VII on CTL granules suggested that the reduced cytotoxicity of Syt VII KO CTLs resulted from a defect in the granule-mediated killing pathway. To test this hypothesis, the allogeneic target cell lytic assays were repeated after pretreating the effector CTLs with a specific inhibitor of perforin-mediated lysis, concanamycin A. This inhibitor blocks the vacuolar ATPase-dependent acidification that is required for perforin-mediated cytolysis (26). When perforin-mediated cytotoxicity was inhibited (using the lowest concanamycin A dose with inhibitory effect on the cytotoxicity of WT CTLs), Syt VII WT or KO CTLs showed comparable levels of target cell lysis (Fig. 5C), suggesting that Syt VII is specifically involved in regulation of granule-mediated cy-totoxicity. Comparable levels of granzyme secretion were observed when CTLs of WT or Syt VII KO mice were stimulated by TCR cross-linking with anti-CD3 Abs (Fig. 5D), a stimulus that is known to bypass the requirement for additional stimulatory signals normally required during Ag-triggered responses (27). These results demonstrate that CD8⁺ CTLs of Syt VII-deficient mice, although containing normal levels of secretory products in their lytic granules, are defective in their capacity to kill target cells through the granule-mediated pathway.