Faster cytotoxicity with age: Increased perforin and granzyme levels in cytotoxic CD8+ T cells boost cancer cell elimination

Abstract

A variety of intrinsic and extrinsic factors contribute to the altered efficiency of CTLs in elderly organisms. In particular, the efficacy of antiviral CD8⁺ T cells responses in the elderly has come back into focus since the COVID-19 pandemic outbreak. However, the exact molecular mechanisms leading to alterations in T cell function and the origin of the observed impairments have not been fully explored. Therefore, we investigated whether intrinsic changes affect the cytotoxic ability of CD8⁺ T cells in aging. We focused on the different subpopulations and time-resolved quantification of cytotoxicity during tumor cell elimination. We report a surprising result: Killing kinetics of CD8⁺ T cells from elderly mice are much faster than those of CD8⁺ T cells from adult mice. This is true not only in the total CD8⁺ T cell population but also for their effector (T_EM ) and central memory (T_CM ) T cell subpopulations. TIRF experiments reveal that CD8⁺ T cells from elderly mice possess comparable numbers of fusion events per cell, but significantly increased numbers of cells with granule fusion. Analysis of the cytotoxic granule (CG) content shows significantly increased perforin and granzyme levels and turns CD8⁺ T cells of elderly mice into very efficient killers. This highlights the importance of distinguishing between cell-intrinsic alterations and microenvironmental changes in elderly individuals. Our results also stress the importance of analyzing the dynamics of CTL cytotoxicity against cancer cells because, with a simple endpoint lysis analysis, cytotoxic differences could have easily been overlooked.

Keywords: CD8+ T cells; CTL; aging; cytotoxicity; granzyme; immunosenescence; perforin; tumor immunology.

FIGURE 1

Faster killing kinetics of CD8⁺ T cells from elderly mice. Time‐resolved killing assays with untouched (a) (n = 7–8) and stimulated (d, g) CD8⁺ T cells from adult and elderly mice. P815 (n = 31–35) and EL4 (n = 8–11) tumor cell lines were used as target cells in an effector‐to‐target ratio of 20:1. Box plots represent the average target cell lysis after 60, 120, and 240 min (b, e, h) and the maximum target lysis per 10 min (c, f, i) as a measure of the kinetics. Data are presented as mean ± SEM

FIGURE 2

Central and effector memory CD8⁺ T cells from elderly mice show fast killing kinetics. Time‐resolved killing assays with stimulated and sorted CD8⁺ T cell subtypes T_CM (a) and T_EM (d) from adult (n = 3) and elderly (n = 4) mice and P815 as target cells in an effector‐to‐target ratio of 20:1. Box plots represent the average target cell lysis after 60, 120, and 240 min (b, e) and the maximum target lysis per 10 min (c, f). Data are presented as mean ± SEM

FIGURE 3

Live‐cell imaging confirms rapid target cell lysis by CD8⁺ T cells from elderly mice. Stimulated CD8⁺ T cells were added to calcein labelled P815 cells. Cells were imaged every 2 min with propidium iodide (PI) in the media. (a) Representative overlays with fluorescence of calcein (green), PI (red) and brightfield (grey) after 0, 30, 60, 90 and 120 min are depicted. (b) Quantitative analysis of target cell lysis after CTL contact. (c) Time points of the first CTL‐target cell contact regardless of the cell fate. Bar graphs show values as mean ± SEM, n = 3 mice, each with 60–90 analyzed cells

FIGURE 4

CD8⁺ T cells from adult and elderly mice show similar degranulation ability. Flow cytometry‐based analysis of CD107a (LAMP‐1) mobilization in stimulated CD8⁺ T cells during cytotoxic activity. (a) Representative gating strategy for CD8⁺ T cells from adult (black) and elderly (red) mice immediately after co‐culturing with P815 target cells (effector‐to‐target ratio 1:1). Viable CD8⁺ T cells were defined based on SSC‐A (side scatter area) vs. FSC‐A (forward scatter area) and CD8 surface expression. Expression of CD107a was determined in CD44⁺CD62L⁺ (T_CM) and CD44⁺CD62L⁻ (T_EM) cells. (b) Representative histograms of CD107a surface expression in T_CM and T_EM from adult (black) and elderly (red) mice at different time points after co‐culturing with P815 target cells. (c) Time course of CD107a mobilization analyzed by mean fluorescence intensity (MFI). Data are presented as mean ± SEM, n = 6

FIGURE 5

Similar lytic granule events but increased numbers of CTLs with granule fusion in old age. Stimulated CD8⁺ T cells were transfected with granzyme B‐mcherry construct for 12 h and settled on an anti‐CD3 coated coverslip. Representative SIM images (a) and analysis of GrzB⁺ vesicles per cell (b), n = 6–7. Representative TIRF images showing the individual fusion events of lytic granules are marked by an arrow (c). The numbers on the right of each image represents the number of fusion events in the particular cell i.e., two fusion events by the CD8⁺ T cell of the adult mouse and 4 fusion events by CD8⁺ T cell from the elderly. Bar graphs show the percentage of CD8⁺ T cells with granule fusion (d), and the average number of granule fusion events per cell (e), n = 5. Data are presented as mean ± SEM

FIGURE 6

Increased perforin and granzyme expression in CD8⁺ T cells from elderly mice. Normalized mRNA expression of perforin (Prf) and the granzymes A (GrzA) and B (GrzB) in untouched (a) and stimulated (b) CD8⁺ T cells from adult and elderly mice. Expression levels were normalized to the reference genes hypoxanthine‐phosphoribosyl transferase 1 (HPRT1) and TATA box binding protein (TBP). Data from elderly mice (n = 7–8) are presented as relative fold change to the mRNA levels from adult mice (n = 8–13). (c) Perforin and granzyme B protein expression in stimulated CD8⁺ T cells from adult and elderly mice. Representative western blot and densitometric quantification with GAPDH as reference protein (n = 5–6). (d) Double immunofluorescence staining for granzyme B and CD8 in stimulated CD8⁺ T cells from adult and elderly mice. Representative SIM (Structured Illumination Microscopy) images after staining with Alexa647 anti‐human/mouse granzyme B and FITC anti‐mouse CD8a antibody. (e) Histogram of GrzB mean fluorescence intensity (bin width 200) and statistical analysis (f) of stained CD8⁺ T cells. Data are presented as mean ± SEM, n = 5–7