Migratory properties of naive, effector, and memory CD8(+) T cells

Abstract

It has been proposed that two different antigen-experienced T cell subsets may be distinguishable by their preferential ability to home to lymphoid organs (central memory cells) or nonlymphoid tissues (effector memory/effector cells). We have shown recently that murine antigen-primed CD8(+) T cells cultured in interleukin (IL)-15 (CD8(IL-15)) resemble central memory cells in phenotype and function. In contrast, primed CD8(+) T cells cultured in IL-2 (CD8(IL-2)) become cytotoxic effector cells. Here, the migratory behavior of these two subsets was investigated. Naive, CD8(IL-15) cells and, to a lesser degree, CD8(IL-2) cells localized to T cell areas in the spleen, but only naive and CD8(IL-15) cells homed to lymph nodes (LNs) and Peyer's patches. Intravital microscopy of peripheral LNs revealed that CD8(IL-15) cells, but not CD8(IL-2) cells, rolled and arrested in high endothelial venules (HEVs). Migration of CD8(IL-15) cells to LNs depended on L-selectin and required chemokines that bind CC chemokine receptor (CCR)7. Both antigen-experienced populations, but not naive T cells, responded to inflammatory chemokines and accumulated at sites of inflammation. However, CD8(IL-2) cells were 12 times more efficient in migrating to inflamed peritoneum than CD8(IL-15) cells. Furthermore, CD8(IL-15) cells proliferated rapidly upon reencounter with antigen at sites of inflammation. Thus, central memory-like CD8(IL-15) cells home avidly to lymphoid organs and moderately to sites of inflammation, where they mediate rapid recall responses, whereas CD8(IL-2) effector T cells accumulate in inflamed tissues, but are excluded from most lymphoid organs.

Figure 1

Primed CD8⁺ T cells differentiated with IL-15, but not with IL-2, home to LNs. Splenocytes from T-GFPxP14 mice (n = 4) or from T-GFP mice (n = 2) were stimulated with gp33 peptide or anti-CD3ε, respectively, and cultured for 6–8 d in the presence of IL-2 (black bars) or IL-15 (white bars). 5 × 10⁷ TRITC-labeled cells from either cytokine-treated population were mixed with 5 × 10⁷ fresh T-GFP splenocytes (as an internal standard) and injected intravenously into naive C57BL/6 recipients. The ratio of homed TRITC⁺ cells to GFP⁺ TRITC⁻ cells in different tissues was analyzed by flow cytometry 24 h later (gates were set on live lymphocytes). As cytokine-treated primed T-GFPxP14 and T-GFP cells behaved identically, the results were pooled. (A) Representative two-color dot plots of adoptively transferred T cell subset frequencies in peripheral blood leukocytes (left) and PLNs (right) of two recipient mice. Adoptively transferred CD8^IL-15 cells (top panels) or CD8^IL-2 cells (bottom panels) are both readily detectable in blood, but only CD8^IL-15 cells are found in significant numbers in PLNs. The homing index (i.e., the ratio of TRITC⁺:GFP⁺ events in respective gates divided by the ratio of TRITC⁺:GFP⁺ cells in the mixture before adoptive transfer) is shown in the upper right corner of each panel. (B) Homing indices of CD8^IL-15 cells and CD8^IL-2 cells versus naive T-GFP cells in the lungs and liver. (C) Frequency of homed cells in blood and secondary lymphoid organs. Bars represent the percentage of homed TRITC⁺ cells in gated lymphocyte populations. (D) Homing indices of CD8^IL-15 cells and CD8^IL-2 cells versus naive T-GFP cells in lymphoid organs. *P < 0.05; **P < 0.01; n.s., not significant. Bars represent mean ± SEM of results from six mice.

Figure 2

Localization of homed CD8^IL-15 and CD8^IL-2 cells in PLNs and spleen. Cryostat sections of PLNs (top two rows) and spleen (bottom two rows) 24 h after adoptive transfer of TRITC-labeled CD8^IL-15 cells or CD8^IL-2 cells were stained with anti–B220-FITC (green) and Thy-1.2-biotin followed by Cy5-conjugated streptavidin (blue) to localize B cell follicles and T cell areas, respectively. TRITC-labeled homed donor cells are identified by red fluorescence. Micrographs are representative of organs from three different recipient mice.

Figure 3

L-selectin is essential for CD8^IL-15 cells to home to LNs. Splenocytes from T-GFPxL^−/− mice (white bars) or T-GFP mice (black bars) were stimulated with anti-CD3ε for 2 d and then cultured in the presence of IL-15 (20 ng/ml). After 7 d, 5 × 10⁷ TRITC-labeled CD8^IL-15 cells from each strain were mixed with 5 × 10⁷ freshly isolated T-GFP splenocytes and injected intravenously into C57BL/6 recipients. The frequency of TRITC⁺ cells in blood and secondary lymphoid organs was analyzed 24 h later. The homing index was determined as described in Fig. 1. Bars reflect mean ± SEM from three independent experiments. Homing was compared using the unpaired Student's t test. **P < 0.01; n.s., not significant.

Figure 5

Differential rolling and sticking of CD8^IL-15 and CD8^IL-2 cells in PLN HEVs. CD8^IL-15 cells and CD8^IL-2 cells were labeled with calcein and TRITC, respectively, and successively injected into the right femoral artery of an anesthetized mouse. Interactions of injected cells with HEV in surgically prepared left subiliac LNs were studied by epi-fluorescence intravital microscopy as described (reference 31). (A) Rolling fractions (the percentage of rolling cells in the total flux of cells passing a venule) and (B) sticking fractions (the percentage of cells that arrested for ≥30 s in the total flux) of both T cell populations were analyzed in the same 11 HEVs of 4 animals. Bars represent means ± SEM. Data were compared using the paired Student's t test. *P < 0.005; **P < 0.001.

Figure 5

Differential rolling and sticking of CD8^IL-15 and CD8^IL-2 cells in PLN HEVs. CD8^IL-15 cells and CD8^IL-2 cells were labeled with calcein and TRITC, respectively, and successively injected into the right femoral artery of an anesthetized mouse. Interactions of injected cells with HEV in surgically prepared left subiliac LNs were studied by epi-fluorescence intravital microscopy as described (reference 31). (A) Rolling fractions (the percentage of rolling cells in the total flux of cells passing a venule) and (B) sticking fractions (the percentage of cells that arrested for ≥30 s in the total flux) of both T cell populations were analyzed in the same 11 HEVs of 4 animals. Bars represent means ± SEM. Data were compared using the paired Student's t test. *P < 0.005; **P < 0.001.

Figure 4

Homing of CD8^IL-15 cells to LNs is impaired in plt/plt mice. Splenocytes from T-GFPxP14 mice were stimulated with gp33 peptide and cultured in IL-15 for 6–8 d. TRITC-labeled CD8^IL-15 cells (5 × 10⁷) were mixed with 5 × 10⁷ freshly isolated T-GFP splenocytes and adoptively transferred into either wild-type DDD1-mtv/mtv (white bars) or plt/plt recipients (black bars). The absolute number of homed cells in spleens, PLNs, and MLNs was calculated by multiplying the frequency of fluorescent cells (assessed by flow cytometry) with the total number of resident lymphocytes in each organ. Data are shown as number of homed cells per 10⁶ cells injected. Bars reflect mean ± SEM (n = 5 recipients of each strain). Homing was compared using the unpaired Student's t test. *P < 0.05; **P < 0.01, n.s., not significant.

Figure 6

Both CD8^IL-15 and CD8^IL-2 cells respond to inflammatory chemokines. CD8^IL-2 (•), CD8^IL-15 (□), or naive splenocytes (▾) were loaded into Transwell filter inserts (5 × 10⁵ cells/well), which were placed in 24-well plates containing 600 μl of medium with different concentrations of CCL21, CCL5 (RANTES), or CCL2 (MCP-1). Chemotactic indices were calculated as the ratio of the number of cells that migrated toward a chemokine divided by the number of spontaneously migrating cells recovered from wells that contained medium without chemokine. Data reflect mean ± SEM of at least two duplicate experiments.

Figure 8

CD8^IL-15 cells proliferate in the peritoneal cavity in the presence of peptide antigen. 2.5 × 10⁷ CFSE labeled CD8^IL-15 cells from P14 mice were injected into C57BL/6 mice. 24 h later, peritonitis was induced by intraperitoneal injection of IFA without or with gp33 peptide (50 μg). After 72 h, PEL and LN cells were harvested, stained, and analyzed by FACS^®. (A) Shows the percentage of CFSE⁺ cells in the total CD8⁺ population in each organ. Bars represent mean ± SEM from three animals. Values for each organ were compared using the Student's t test (n.s., not significant; *P = 0.01). (B) Representative FACS^® dotplots of PLNs and PEL from animals receiving intraperitoneal injection of IFA without (top two plots) or with gp33 peptide (bottom two plots). The number in the right corner represents the percentage of CFSE⁺ cells in the CD8⁺ population.

Figure 8

CD8^IL-15 cells proliferate in the peritoneal cavity in the presence of peptide antigen. 2.5 × 10⁷ CFSE labeled CD8^IL-15 cells from P14 mice were injected into C57BL/6 mice. 24 h later, peritonitis was induced by intraperitoneal injection of IFA without or with gp33 peptide (50 μg). After 72 h, PEL and LN cells were harvested, stained, and analyzed by FACS^®. (A) Shows the percentage of CFSE⁺ cells in the total CD8⁺ population in each organ. Bars represent mean ± SEM from three animals. Values for each organ were compared using the Student's t test (n.s., not significant; *P = 0.01). (B) Representative FACS^® dotplots of PLNs and PEL from animals receiving intraperitoneal injection of IFA without (top two plots) or with gp33 peptide (bottom two plots). The number in the right corner represents the percentage of CFSE⁺ cells in the CD8⁺ population.

Figure 7

CD8^IL-2 and CD8^IL-15, but not naive T cells are recruited to sites of inflammation. Peritonitis was induced by injection of IFA into the peritoneal cavity of C57BL/6 mice. 72 h later, one group (n = 6 animals) received 2.5 × 10⁷ TRITC-labeled CD8^IL-15 mixed with 4 × 10⁷ freshly isolated T-GFP splenocytes (A). A second group (n = 8 mice) received an equivalent number (2.5 × 10⁷ cells of each) of TRITC-labeled CD8^IL-2 cells and CFSE-labeled CD8^IL-15 cells (B). Representative two-color dot plots show the frequency of fluorescent CD8^IL-15 versus T-GFP cells (A) and of CD8^IL-2 versus CD8^IL-15 cells (B) in blood, PLNs, and peritoneal exudate leukocytes (PEL). Numbers in the upper right corner of each histogram represent the ratio of CD8^IL-15:naive T-GFP cells (A) or of CD8^IL-2: CD8^IL-15 cells (B). (C) Bars represent the mean ± SEM of the absolute numbers of homed cells recovered from inflamed peritoneal cavities. Statistical differences were assessed using the unpaired Student's t test. *P < 0.05; **P < 0.01.

Figure 7

CD8^IL-2 and CD8^IL-15, but not naive T cells are recruited to sites of inflammation. Peritonitis was induced by injection of IFA into the peritoneal cavity of C57BL/6 mice. 72 h later, one group (n = 6 animals) received 2.5 × 10⁷ TRITC-labeled CD8^IL-15 mixed with 4 × 10⁷ freshly isolated T-GFP splenocytes (A). A second group (n = 8 mice) received an equivalent number (2.5 × 10⁷ cells of each) of TRITC-labeled CD8^IL-2 cells and CFSE-labeled CD8^IL-15 cells (B). Representative two-color dot plots show the frequency of fluorescent CD8^IL-15 versus T-GFP cells (A) and of CD8^IL-2 versus CD8^IL-15 cells (B) in blood, PLNs, and peritoneal exudate leukocytes (PEL). Numbers in the upper right corner of each histogram represent the ratio of CD8^IL-15:naive T-GFP cells (A) or of CD8^IL-2: CD8^IL-15 cells (B). (C) Bars represent the mean ± SEM of the absolute numbers of homed cells recovered from inflamed peritoneal cavities. Statistical differences were assessed using the unpaired Student's t test. *P < 0.05; **P < 0.01.