Figure 1.
Primary KC from B6 mouse pups were cultured in multi-well plates until 70% confluent, and then loaded with SIINFEKL peptide. Antigen-specific CD8+ T cells from EGFP+ B6 mice were isolated and co-cultured with target cells in the presence of cell-permeable caspase-3 indicator dye and imaged by time-lapse fluorescence microscopy or by confocal microscopy every 12 minutes for 30 h. Cell death was noted by colour change, morphology, and behaviour.
Figure 2.
Imaging of CTL and KC targets.
A. Series of still confocal microscopic images from Movie S1 of a single field over 30 hours. CD8+ T cells (green) attach to KC after 7 h of co-culture and cause apoptosis (red) at 21 h. Overlay images were from brightfield and fluorescence channels, acquired using a ×25 objective lens, mean intensity projection of a Z-stack of 5 confocal images of 2 µm each. (Bar, 50 µm) B. Analysis of epifluorescent imaging of CTL and KC co-cultures by software spot detection. Merged views of brightfield, rhodamine filter and GFP filter are shown in each frame at time 0 (i) and at 18 h (ii–iv). i. At t = 0, a KC and several smaller green T cells can be seen. ii. Spot detection without correction, using parameters: Red+, size 7 µm. iii. Spot detection after manual correction for adjacent spots. iv. Spot detection using parameters: Green+, size 7 µm (green); red channel+, size 7 µm (red); and including Red+, size 15 µm (purple).
Figure 3.
CD8+ T cells require direct prolonged contact with target cells to kill KC.
Primary KC loaded with ovalbumin peptide were co-cultured with CD8+ T cells from EGFP+OT-1 mice and imaged over a 30-hour period in the presence of a cell-permeable indicator dye that fluoresces red upon activation of caspase-3. A. The number of caspase-3 positive keratinocytes in each 5 hour time period per experiment was determined by standardized objective analysis of movies by specialized imaging software. N = 6, error bars are SEM. B. In primary KC culture, contaminating fibroblasts are readily distinguished by their morphology. These also uptake peptide and present as targets for CTL. KC and fibroblast survival in co-culture with OT-1 cells; (Pooled results N = 5, error bars are SD). C. Addition of caspase-3 inhibitor, Z-DVED-FMK, an inhibitor of caspase-3, resulted in abrogation of CTL-mediated apoptosis. (Pooled results N = 5, error bars are SD). D. T cell death, following a similar pattern to KC death, is abrogated by the addition of Z-DVED-FMK (Pooled results N = 3, error bars are SD).
Figure 4.
CD8+ T cell activation is sufficient to promote enhanced kinematics.
CTL from OT-1 mice (OT-1), or from B6 mice either immunized with OVA and stimulated with SIINFEKL (Ova-stim), or stimulated with concavalin A (ConA), or not stimulated (Naïve), were added to KC cultures loaded with SIINFEKL and imaged for 30 h. A. CTL were isolated by FACS based on size and granularity and by EGFP and CD8 expression, and stained for CD44. Percentage of isolated cells that are CD44+ are shown. B. Percentage of total KC that were killed after 30 hours of co-culture with different groups of CD8+ T cells. KC killing was determined by red fluorescence change indicating activation of caspase-3. C. Vector displacement and velocity (D) of the effector cells added to antigen-presenting KC over 30 hours of imaging. (A–D N = 5, error bars SEM) (n.s. not significant, *p<0.05; N.S. no significant difference between indicated groups, *p<0.05 between indicated groups ANOVA.).
Figure 5.
KC killing is dependent on antigen presentation and T cell activity.
A. Antigen-specific CTL were co-cultured with B6 KC loaded with increasing concentrations of SIINFEKL peptide. Percentage of total KC killed during of imaging 30 hours was measured by fluormetric detection of caspase-3 activation. (N = 5, error bars SEM). B. Average duration of attachments of activated T cells, either antigen specific Ova-stim CTL or non-specifically activated ConA CTL, to targets with or without peptide. (N = 6, error bars SEM). C. Velocities of antigen-specific CTL added to target KC, with and without peptide loading, were calculated from 30 h tracks. Values were binned 25 nm.s−1 and averages of 5 experiments are shown. Error bars are SEM. D. Examination of the velocities of the effector T cells (from (A)) in response to different concentrations of peptide loaded onto target cells. Linear regression line shown. E. Average displacements of CTL in the presence or absence of cognate antigen. (N = 5, error bars SEM) F. Example of actual tracks of CTL cultured with target cells, with and without peptide, translated to a common starting point. (*p<0.05, *p<0.05 between indicated groups ANOVA.).
Figure 6.
Kinematics of activated CD8+ T cells correlate with their cytotoxicity.
CD8+ T cells were generated by prolonged stimulation (PS) of splenocytes from OVA-immunised mice with irradiated feeder cells pulsed with 0.1 mg.ml−1 peptide, or they underwent short stimulation (SS) by overnight culture with 1 mg.ml−1 of peptide; or they were not stimulated (NS) in vitro after immunization. These were added to cultured KC targets presenting SIINFEKL and imaged. A. T cells were isolated by FACS based on size, granularity, EGFP and CD8 expression. These CD8+ cells were then stained for CD44 and CD62L expression. B. Percentage of KC targets that were killed by each group of CTL over a 30 hour time-course is shown. PS T cells were also added to KC without peptide to show non-specific killing. Death rate of KC in monoculture is also shown. C. The time for each group of T cells to reach 50% of their total target cell death is shown. (N = 3 independent experiments) D. Examples of the movement tracks taken by CTL among the KC cells colour-coded for displacement and velocity. Scale bars, 50 µm. E. Survival of PS, NS and SS CTL over 30 hours of co-culture with antigen-loaded KC was determined from the number of GFP+ cells remaining as imaging progressed, as a percentage of the total number of GFP+ cells at time 0. Pooled data 5 experiments, error bars are SD. F. & G. Displacement and velocity of effector cell groups in culture with targets with and without antigen. (B, C, F, G N = 5, error bars SEM) H. Example of frequency plot of velocity values for the different populations of CTL. (*p<0.05 between indicated groups, n.s. not significant, ANOVA).
Figure 7.
Velocity of a CTL directly correlates with its ability to kill.
CTL were isolated from OVA-immunised mice and underwent short stimulation with antigen in vitro before adding to KC monolayers loaded with SIINFEKL and imaged for 30 h. Movies were examined in 5-hour time intervals to analyse target cell death in that time frame, and its killer CTL identified. If more than one CTL was attached to a dying KC, both were analysed. The average track velocity for each identified CTL in each time interval was determined. CTL which were not moving, were excluded. Pooled results N = 3 independent experiments. Mean and SD are shown.
Figure 8.
Co-culture conditions affect CD8+ T cell survival.
CTL were isolated from OVA-immunised mice and were stimulated for a short time in vitro before adding to KC monolayers loaded with SIINFEKL and imaging for 20 hours. CTL death was determined by size (<10 µm), colour change (EGFP to red indicating loss of EGFP expression and caspase-3 activation) and morphology. A. T cell death in response to KC targets with 1 mg.ml−1 SIINFEKL and without peptide. (N = 5, error bars SEM, *p<0.05 between indicated groups) B. T cells were added to KCs loaded with increasing concentrations of SIINFEKL and CTL death over 30 hours was measured. (N = 5) C. CTL were added to KC monolayers loaded with 10 mg.ml−1 SIINFEKL. Still images from the beginning of co-culture (upper panel) and at 10 hours (lower panel) were taken by epifluorescent time-lapse microscopy.