Naive CD8 T cells persist without proliferation in BM chimeric hosts. Lymphocytes from WT mice (a and b) or purified K^bD^b−/− lymphocytes from K^bD^b−/− → WT BM chimeras (c and d) were stained with CFSE and injected i.v. (5 × 10⁶/mouse) into intact WT (a and b) or K^bD^b−/− (c and d) mice and various BM-chimeric recipients. 30 d after transfer, spleens and lymph nodes were harvested and analyzed by flow cytometry on individual mice. (a and c) CFSE dilution of CD8 T cells in the spleen of recipients 30 d after transfer. The outlined area represents CFSE^hiCD44^lo cells analyzed for the survival and numbers above indicate the percentage of cells in CD8 gate. (b and d) Recovery of CFSE^hiCD44^lo donor CD8 (left) and CD4 (right) T cells. Total donor cell numbers in the spleens and lymph nodes are shown. Data are mean ± SD. Results are representative of two independent experiments with at least three mice per group. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

MHC-I–deprivation leads to phenotypic changes in naive CD8 T cells, suggestive of retuning. Lymphocytes from WT mice (a) or purified K^bD^b−/− lymphocytes from K^bD^b−/− → WT BM chimeras (b and c) were injected i.v. (3–5 × 10⁶/mouse) into indicated recipients. (a and b) 30 d after transfer, donor naive CD8 and CD4 T cells in host spleens were detected by CFSE label and a CD44^lo naive T cell marker. Surface expression of IL-7Rα, CD5, and coreceptors (CD8α or CD4) were analyzed. (c) At day 20 after transfer, spleens and lymph nodes were harvested from host mice, and CD8α expression on CD44^loCD8 donor cells (Thy1.1⁺) that stain positive for each Vβ chain was analyzed. Data are presented as relative fluorescence intensity (RFI), indicating the expression which is normalized to the average of mean fluorescence intensity (MFI) in WT → WT hosts (a) or K^bD^b−/− → WT hosts (b and c) defined as 100. Data are mean ± SD. Results are representative of two independent experiments with at least three mice per group. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Increased transcription of CD8 gene after deprived of MHC-I contact. Spleen cells from B6 (a) OT-I (a) and E8_I-CD4 transgenic mice (c) were cultured for 24 h in the absence (control) or presence of anti-K^b and anti-D^b antibodies. In (a), the data show expression levels of CD4 (on CD4 T cells), CD8α and CD8β (on CD8 T cells). Expression of indicated molecules is presented as fluorescence intensity relative to the control cultures. (b) OT-I cells were cultured as in a, except that actinomycin D (AMD) or cycloheximide (CHX) were added to some cultures, as indicated. Expression of indicated molecules is presented as fluorescence intensity relative to control cultures. (c) E8I-CD4 splenocytes were analyzed as in a, except that CD4 levels were monitored on both CD48- T cells (first panel, reflecting endogenous CD4) and on CD48+ T cells (reflecting transgene encoded CD4). For a–c, the data represent mean ± SD. (d) Splenocytes from OT-I Tg mice (Thy1.1) were cultured for 24 h in the presence or absence of anti-K^b and anti-D^b antibodies, and then transferred into B6 mice (n = 3 mice/group). CD8α and CD8β expression levels on donor OT-I cells isolated from lymph nodes 48 h later were determined. This CD8α and CD8β staining on OT-I cells from control cultures was defined as 100 for each time point. Data before transfer is from one sample and data at 48 h after transfer represent mean ± SD. All data are representative of two to three independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

CD8 up-regulation dictates the enhanced functional reactivity of MHC-I–deprived CD8 T cells. (a and b) Purified anti-CD8β antibody (53–5.8; 0.5 mg/ml) was added at indicated titration to OT-I cell (Thy-1.1) suspension. Cells were then incubated with irradiated WT splenocytes pulsed with G4, OVAp or no peptide. After incubation, cells were stained with fluorochrome-conjugated anti-CD8β (53–5.8), and the levels of unblocked CD8 in the peptide-free group are shown as MFI relative to the cells nontreated with the blocking antibody (a). Response of OT-I cells to G4 and OVAp splenocytes is shown (b). Relative response indicates the percentage of CD69^hi cells after the background value (no peptide control) was subtracted, and is shown relative to the mean of the “no antibody” group, which was defined as 100. (c and d) Purified OT-I cells (Thy1.1) were transferred (3–5 × 10⁶/mouse) to WT or K^bD^b−/− mice. 36 h later, host lymph node cells were stimulated with irradiated WT splenocytes pulsed or nonpulsed with G4 and OVAp. Cells from K^bD^b−/− hosts were also stimulated in the presence of blocking anti-CD8β antibody diluted at 1:1,024,200 or 1:51,200, as indicated. (c) The levels of available CD8β were determined by staining the unstimulated group. (d) The capacity of the different cell populations to up-regulate CD69 in response to G4 (left) or OVAp (right) was determined and shown relative to the mean value obtained from OT-I cells from WT hosts (defined as 100). Data represent mean ± SD and are representative of two independent experiments (a and b) or combined results from two independent experiments (c and d). *, P < 0.05; **, P < 0.01; ***, P < 0.001.

MHC-I deprivation causes enhanced CD8 T cell reactivity to low-affinity TCR ligands in vitro. Purified CD8 T cells from K^bD^b−/−Rag^−/−OT-I Tg → WT chimeras were CFSE stained and transferred (3 × 10⁶/mouse) to K^bD^b−/− → K^bD^b−/− or K^bD^b−/− → WT BM chimeric hosts. All experiments were performed 14 d after transfer. (a) Spleens were harvested from recipients and analyzed for the expression of CD5 and CD8 on the surface of donor OT-I cells by flow cytometry. Data are presented as relative fluorescence intensity (RFI) normalized to the average of MFI in K^bD^b−/− → WT hosts. (b–d) Host spleen cells were cultured with irradiated WT splenocytes pulsed with different concentrations of OVAp (0–10⁶ pM; b and d) or indicated peptides (c). Cells were analyzed for the expression of CD69 (b and c) or intracellular TNF (d). (e) Recipients were injected i.v. with WT DCs (10⁶/mouse) pulsed or nonpulsed with OVAp. 18 h after DC injection, donor cells in the spleens (SPL) and lymph nodes (LN) were analyzed for the expression of CD69. The percentage of donor cells with CD69^hi phenotype is shown (b–d). Data represent mean ± SD. Data are representative of two to four independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Impaired survival of CD8 T cells in the complete absence of MHC-I. Purified K^bD^b−/− lymphocytes (Thy1.1) from K^bD^b−/− → WT BM chimeras were transferred (3–5 × 10⁶/mouse) to normal K^bD^b−/− mice and K^bD^b−/− → WT BM chimeras. (a) Detection of donor CD8 T cells by flow cytometry at day 60 after transfer (left and middle). (right) Background events detected by the same method in nontransferred WT mice. Numbers above the outlined area indicate the percentage of cells in the lymphocyte gate. (b and c) Mice were sacrificed at indicated time points and analyzed for the recovery of donor T cells from the spleen and lymph nodes. (b) Absolute number of Thy1.1⁺CD44^loCD8 (left) and Thy1.1⁺CD44^loCD4 (right) donor cells. Half-lives (t_1/2) were calculated from regression lines (Hataye et al., 2006). (c) Combined results of two independent experiments. CD8/CD4 ratio was calculated before (R0) and after (R1) transfer into K^bD^b−/− → WT BM chimeric or normal K^bD^b−/− hosts. Difference of CD8/4 balance in donor cells between experiments was normalized by calculating R1/R0. (d) TCR Vβ analysis of Thy1.1⁺CD44^loCD8 cells at day 20 after transfer. Percentage of cells that stain positive for each Vβ chain in total Thy1.1⁺CD44^loCD8 cells is shown. The bars show the mean ± SD. All data other than c are representative of two independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

CD8 T cell retuning occurs rapidly after MHC-I deprivation. (a-c, e, and f) Purified CD8 T cells from WT OT-I Tg mice (Thy1.1) were transferred (3 × 10⁶/mouse) to WT or K^bD^b−/− mice. 1 d (a–c) or 2 h (e and f) after transfer, host spleen cells or pooled lymph node cells were analyzed for the phenotype of donor OT-I cells (a and e) or were stimulated with irradiated WT splenocytes pulsed with different concentrations of OVAp (0–10⁶ pM; b) or indicated peptides (c and f). (d) Purified CD8 T cells from K^bD^b−/−Rag^−/−OT-I Tg → WT chimeras were CFSE stained and transferred (3 × 10⁶/mouse) to K^bD^b−/− → K^bD^b−/− or K^bD^b−/− → WT BM chimeric hosts. 1 d later, host mice were injected (i.v.) with WT DCs (10⁶/mouse) that were pulsed or nonpulsed with OVAp. 18 h after DC injection, donor cells in the spleens (SPL) and lymph nodes (LN) were analyzed by flow cytometry. The percentage of donor cells with CD69^hi phenotype after stimulation is shown (b-d and f). (g and h) Purified OT-I cells with different congenic markers (Thy1.1⁺Thy1.2⁺ and Thy1.1⁺Thy1.2⁻) were separately transferred (4–5 × 10⁶/mouse) to WT and K^bD^b−/− mice. Cells were CFSE-labeled before transfer in h. 36 h after transfer, lymph nodes cells from both hosts and irradiated B6 spleen cells were mixed at equal number and cultured with or without G4. Cells were analyzed for stained for cell surface CD69 or intracellular pc-Jun at 6 h (g) or for CFSE dye dilution at 48 h (h). In g, the percentage of CD69^hi or pc-Jun⁺ CD8 T cells was determined, and the background value (from control cultures without G4 peptide) was subtracted. The response of the OT-I cells derived from the WT hosts was defined as 100. Data are representative of two independent experiments with three to four mice per group. Data in g represent combined results from three independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001.