CTLs polarize cytotoxic granules toward the area of contact with recognizing CD8⁺ T cells in the absence of TCR specificity. (A) H-2^d/b (F1) CTLs were labeled with LysoTracker Red for the detection of granules and targeted with 2C CD8⁺ T cells for 1 hour at 37°C. Cells were fixed, stained, and visualized with an LSM 510 Laser Scanning Confocal Microscope (Carl Zeiss) and 100× PLAN Apochromat objectives having a numerical aperture of 1.4. The F1 CTL appears on the left of each panel and stains positive for both H-2D^d (blue, merge) and H-2D^b (green, merge). The 2C cell appears on the right and stains positive only for H-2D^b. Granules appear in red in the merged image. Scale bar = 5 μm. (B) Quantitation of CTL granule polarization. Conjugates of H-2^d/b (F1) or H-2^d CTLs with 2C CD8⁺ T cells were prepared as in panel A and were evaluated for the presence of cytotoxic granule polarization. CTLs were scored as polarized when granules clearly accumulated in the area of contact. “%Polarization” refers to the percentage of conjugates in which the CTL is found polarizing its granules toward the contact area with the recognizing T cell. Values shown are the calculated means ± SD of 3 independent experiments, n > 70 in each.

CTL stopping, lengthy conjugate duration and sustained granule polarization precede death of the recognizing cell. Lysis of recognizing T cells by targeted CTLs as demonstrated by live cell video microscopy. Target H-2^d/b (F1) CTLs were loaded with LysoTracker Red, which accumulates in granules (orange in appearance), and added to calcein-loaded 2C CD8⁺ T cells (green). Cells were then subject to live video microscopy with a Deltavision Restoration microscope (Applied Precision Instruments) using a MicroMax 5 MHz cooled CCD camera (Roper Scientific). Image sequences of the time-lapse recording were processed with SoftWoRx (Applied Precision). (A) Wide field microscopy demonstrating killing of recognizing T cells. F1 CTLs and 2C CD8⁺ cells were stained as described. After 2 hours of incubation, marked depletion of the recognizing 2C CD8⁺ cells was observed. (B) Snapshots depict the sequence of events occurring in the targeted CTL, culminating in lysis of the recognizing T cell. Time elapsed from the beginning of stable cell-cell contact is indicated. Top row: the CTL shifts its granules toward the contact area. Bottom row: sequential lysis of the 2 2C cells may be observed. Scale bar = 10μm. (C) The effect of specific recognition of CTL on conjugate duration and lysis. Calcein-loaded 2C CD8⁺ T cells were placed in culture with target H-2^d/b CTLs or with unrecognizable H-2^s CTLs and were then monitored by video microscopy. The durations of CTL-2C contacts were recorded and plotted. Data shown were acquired from several microscopic fields and are from a single experiment representative of 3 experiments.

CTLs are induced to survive and proliferate when targeted. (A) Effect of targeting on CTL viability. H-2^d/b (F1) CTLs or H-2^s CTLs were placed in culture with 2C CD8⁺ T cells at a CTL-to-2C ratio of 1:1 or 1:10, or were incubated without 2C cells. In parallel, CTLs were placed alone in untraversable transwells and cultured in the surrounding media of respective CTLs interacting with 2C cells at both 1:1 and 1:10 CTL-to-2C ratios. After 72 hours of incubation, viable CTL counts were obtained by enumerating the number of 7AAD-excluding CTLs as determined by FACS. Values are presented as the number of viable CTLs per 10⁵ CTLs cultured at time zero. (B) Effect of targeting on CTL proliferation. F1 CTLS were loaded with CFSE and placed in culture or in untraversable transwells as described in panel A. After 72 hours of incubation the percentage of divided CTLs was determined by CFSE dilution. (C) Specific TCR recognition and integrin-mediated adhesion are required to induce CTL survival. F1 CTLs were incubated as described in A with either 2C CD8⁺, 2C CD8⁻, OT-I CD8⁺, or wild-type C57B/6 CD8⁺ cells. Only the recognizing 2C CD8⁺ cells induced significant CTL survival and proliferation. The observed CTL response was inhibited on addition of 20ug/mL anti–LFA-1 antibody. Data in panels A-C are from a single experiment representative of 3 independent experiments. Error bars represent SD from triplicate wells. **P < .01, ***P < .001.

Targeted CTL activation is CD8 dependent. (A) CD8 mediates Src activation loop phosphorylation. H-2^d CTLs were targeted with 2C CD8⁺ cells for 90 minutes, and Src phosphorylation was evaluated by FACS. The role of the CTL CD8 molecule in phosphorylation was determined by applying a blocking antibody specific for the 2C MHC-I α3 domain (H-2^b). (B) Contribution of the CTL CD8 molecule to killing. Depletion of targeting 2C CD8⁺ cells was determined in the presence or absence of the blocking antibody from A. In A and B nonspecific CTL activity was controlled by H-2^s CTLs. (C) Relevance of CD8 in adhesion. CTLs were targeted in the presence or absence of the blocking antibody from panel A, and adhesion was assessed by counting CTL-2C conjugates per total CTLs in several microscopic fields. (D) Polarization dependence on CD8. The effect of blocking CD8 ligation on CTL granule polarization toward conjugated 2C cells was evaluated by confocal microscopy. (E) CD8 contributes to CTL survival and proliferation. CD8⁺ and CD8⁻ 2C CTLs (H-2^b) were loaded with SIINFEKL peptide and targeted with OT-1 CD8⁺ cells. After 72 hours, survival and proliferation of the two CTL populations were evaluated by viable cell count and CFSE dilution. (F) ERK phosphorylation depends on CTL CD8 level. H-2^d CTLs were targeted with 2C CD8⁺ cells for 90 minutes. The relationship between CD8 levels and targeting-induced ERK phosphorylation was determined by FACS analysis. In panels A-F, single experiments representative of 3 experiments are shown. SD is from triplicate wells/microscopic fields. *P < .05, **P < .01, ***P < .001; n > 50 (C); n > 100 (D).

CTLs kill recognizing CD8+ T cells in rapid, granule-mediated fashion. (A) Targeted CTL cytotoxic response. H-2^d CTLs (directed against a third party) were targeted with 2C CD8⁺ T cells loaded with calcein, a probe that is lost on cell death. After 1, 2, 5, 10, 15, and 20 hours of incubation at 3 CTL-to-2C ratios (1:1, 1:2, and 1:5), depletion of calcein-positive 1B2⁺CD8⁺ 2C T cells was evaluated by FACS. 2C CD8⁺ T cells were incubated with identically treated H-2^s CTLs as control. (B) Depletion of recognizing T cells by targeted H-2^d/b (F1). F1 CTLs were targeted with 2C cells prepared as in A at a 1:1 ratio for 5 hours. Depletion was determined as in panel A and compared with the depletion induced by H-2^d CTLs (anti–third-party) CTLs. (C) Dependence on intracellular Ca²⁺. CTLs were pretreated with BAPTA-AM and targeted with 2C CD8⁺ T cells. After 5 hours, 2C cell depletion was evaluated as in A and compared with the depletion induced by untreated CTLs. (D) Specific granzyme release associated with CTL activity. H-2^d/b (F1) CTLs were targeted with 2C CD8⁺ T cells or incubated with the specific stimulator cells against which they had been generated for 4 hours, and granzyme release was evaluated by the BLT-esterase assay. 2C cells had no detectable granzyme secretion. Similarly treated H-2^s CTLs did not release granzyme. Error bars in panels A-D represent SD from triplicate samples. Data shown are single experiments representative of 3 experiments. ***P < .001.

Targeted CTL activity is perforin dependent. (A) Perforin and granzyme content of anti–third-party H-2^d CTLs. Wild-type H-2^d CTLs (top row) and perforin^−/− CTLs (bottom row) were fixed and stained intracellularly for perforin (green) and granzyme B (red). 4,6-diamidino-2-phenylindole (DAPI) staining of nuclei may be observed in merge (blue). Scale bar = 15μm. (B) The role of perforin in targeted CTL cytotoxicity. Perforin ^−/− H-2^d CTLs were targeted with 2C CD8⁺ T cells. After 5 hours, 2C cell depletion was evaluated and compared with the depletion induced by wild-type CTLs, as well as by nonspecific CTLs. Error bars represent SD from triplicate samples. Data shown are single experiments representative of 3 experiments. **P = .0015. (C) The role of the granule-mediated cytotoxic pathway in the induction of tolerance toward MHC-mismatched bone marrow by targeted CTLs. In an established graft rejection model,⁶ supralethally irradiated C3H mice were reconstituted with host T cells and transplanted with allogeneic (H-2^d), fully mismatched, T cell–depleted bone marrow either alone (n = 13) or supplemented with donor-type CTLs (H-2^d, n = 6). For evaluation of the role of the granule-mediated pathway, H-2^d perforin ^−/− CTLs were used (n = 17), and their capacity to induce tolerance was compared with that of perforin-competent wild-type CTLs. Data shown are from 2 experiments combined.

CTL targeting induces Src Kinase signaling leading to cytotoxicity and Erk phosphorylation. (A) Src mediates the cytotoxic response of targeted CTLs. H-2^d CTLs preincubated with PP2/PP3 for 1 hour were targeted with calcein-loaded 2C CD8⁺ cells (1:1 ratio). After 2 and 5 hours, 2C cell depletion was evaluated and compared with untreated CTLs. H-2^s CTLs served as the control for nonspecific activity. (B) PP2 does not affect conjugate formation. Cells were fixed, differentially stained, and imaged 1 hour into culture. Conjugation was assessed counting conjugated CTLs per total CTLs in several microscopic fields. Error bars represent SD from triplicate datasets, n > 40 in each. (C) Lck is responsible for mediating the CTL response. H-2^d CTLs were transfected with Lck siRNA. After 60 hours, CTLs were targeted with 2C CD8⁺ cells, and 2C cell depletion was compared with depletion induced by untreated CTLs or CTLs transfected with non-target siRNA. (D) Src activation-loop phosphorylation. H-2^d CTLs were targeted with 2C CD8⁺ cells for 90 minutes and evaluated by FACS for activating Src kinase phosphorylation (Tyr416). H-2^s CTLs controlled nonspecific phosphorylation. The effect of PP2 is shown. (E) Erk phosphorylation in targeted CTLs. CTLs were targeted as in panel D and evaluated for ERK phosphorylation. The effects of Erk inhibitor U0126 and Src inhibitor PP2 are shown. (F) Overlay of FACS plots representative of the Erk phosphorylation shown in panel E. Solid dark gray, targeted CTLs; black line, untargeted CTLs. Panels A-F: Data from single experiment representative of 3 experiments. Error bars represent SD from triplicate wells. **P < .01, ***P < .001.