Fig 2: DCAF1 deficiency leads to p53 protein stabilization.(a,b) The protein expression of p53, p21 and c-Myc detected by immunoblotting (a) and the mRNA expression of p21 and Bax detected by quantitative reverse transcription–PCR (qRT–PCR) (b) in CD4+ naive T cells of different genotypes at indicated time points after being activated with anti-CD3 and anti-CD28. (c,d) CD4+ naive T cells of various genotypes were activated with anti-CD3 and anti-CD28 in the presence of 4-hydroxy-tamoxifen. At indicated time points after activation, p53 and c-Myc protein expression was determined by immunoblotting (c), and p21 and Bax mRNA expression was assessed by qRT–PCR (d). (e,f) The protein expression of p53 by immunoblotting (e) and mRNA expression of p21 and Bax by qRT–PCR (f) in the quiescent and IL-2-re-stimulated effector T cells, which were generated in the presence of 4-hydroxy-tamoxifen as described in Fig. 4e,f. (g) The protein expression of p53 in CD4+ naive T cells of different genotypes at indicated time points after being activated with anti-CD3 and anti-CD28 in the presence and absence of proteasome inhibitor MG132. (h) p53 protein half-life in anti-CD3- and anti-CD28-activated (for 24 h) CD4+ T cells of different genotypes, determined by immunoblotting for p53 in the presence of translation inhibitor cycloheximide (CHX). (i) Human osteosarcoma U2OS cells were transduced (GFP+) by lentiviruses expressing CRISPR only (Control), CRISPR-Dcaf1-1 (sgRNA) or CRISPR-Dcaf1-2 (sgRNA) and then transfected by the plasmids expressing Flag-tagged Mdm2 and HA-tagged Ubiquitin (HA-Ub) as indicated. The poly-ubiquitination of immunoprecipitated (IP) p53 was detected by immunoblotting for HA tag. In this figure, results of immunoblotting are representative of three experiments. For quantitative reverse transcription–PCR assays, means±s.d. of triplicate done in one experiment representative of three are shown (*P<0.05 by Student's t-test). See also Supplementary Figs 4 and 5 and Supplementary Table 2.

Fig 3: Hypoxia Promotes CD8+ T Cell Glycolytic Metabolism in an HIF-1α- but Not HIF-2α-Dependent Fashion(A) qRT-PCR of mRNA levels of Hif1a and Hif2a on magnetically isolated splenic CD8+ T cells before and after activation with αCD3/CD28 for the indicated time points (n = 3). US, unstimulated. Error bars represent SD.(B) Immunoblots showing HIF-1α and HIF-2α expression in T cells collected at the indicated time points after activation. Densitometric analyses are shown at the bottom.(C) Deletion efficiency of Hif1a and Hif2a in genomic DNA from CD8+ lymphocytes isolated from HIF-1αfl/fldlckCRE or HIF-2αfl/fldlckCRE mice (n = 4, error bars represent SD).(D) CFSE (carboxyfluorescein succinimidyl ester) dilution assay 72 hr after in vitro activation (n = 3, error bars represent SD).(E) Proliferation index and percent survival of isolated CD8+ T lymphocytes 4 days after activation (n = 4, error bars represent SD).(F) CD8+ T cells were isolated from spleens and activated with αCD3/CD28 for 48 hr, and then expanded for 3 days in IL-2 and subjected to 21% or 1% O2 for 24 hr. Western blotting was performed on nuclear fractions; densitometric analyses are shown.(G) CD8+ T cells from HIF-1αfl/fldlckCRE (maroon), HIF-2αfl/fldlckCRE (green), and littermate controls (black for HIF-1αfl/fl, gray for HIF-2αfl/fl) were isolated, activated, expanded for 5 days in the presence of IL-2, and cultured for 24 hr under 21% versus 1% O2. qRT-PCR was performed for Hk2, Pdk1, Mct4, and Pgk (n = 3, error bars represent SD).(H) Extracellular acidification rate (ECAR) and oxygen consumption rate (OCR) of CD8+ T cells prepared as in (G) were measured by flux analysis, under basal conditions and after injection (dashed line) of oligomycin and FCCP (stressed) (n = 4 per genotype, error bars represent SEM).(I and J) Supernatants of CD8+ T cells cultured as in (G) were measured for glucose (I) and lactate (J) levels (n = 6, error bars represent SD).(K) Representative flow-cytometry histograms of viability staining and percent survival of CD8+ T cells cultured as in (G) (n = 4). Grouped data were assessed by two-way ANOVA for multiple comparisons with Bonferroni correction.Horizontal lines are the mean values. ∗∗∗∗p < 0.00005, ∗∗∗p < 0.0005, ∗∗p < 0.005, ∗p < 0.05. See also Table S1 and Figure S1.

Fig 4: Hypoxia and HIF-1α, but Not HIF-2α, Support CD8+ T Cell Acquisition of Effector Phenotype(A) CD8+ T cells were isolated from spleens of HIF-1αfl/fldlckCRE (maroon) or HIF-2αfl/fldlckCRE (green) and littermate control (black) mice, and activated with αCD3/CD28 for 48 hr, then expanded for 5 days in IL-2 and subjected to 21% or 1% O2 for 24 hr. Expression of CD44 and CD62L by flow cytometry is shown.(B) Intracellular expression of IFNγ, TNFα, and granzyme B in HIF-1αfl/fldlckCRE (maroon) and HIF-1αfl/fl control littermates (black) by flow cytometry after restimulation (n = 3, error bars represent SD).(C) Intracellular expression of IFNγ, TNFα, and granzyme B in HIF-2αfl/fldlckCRE (green) and HIF-2αfl/fl control littermates (gray) by flow cytometry after restimulation (n = 3, error bars represent SD).(D) Surface expression of costimulatory molecules/checkpoint receptors CD137, OX40, GITR, PD-1, TIM-3, and LAG3 on CD8+ T cells isolated from HIF-1αfl/fldlckCRE (maroon) HIF-1αfl/fl control littermates (black), activated and expanded as in (A) (n = 3, error bars represent SD).(E) Same as in (D) on CD8+ T cells isolated from HIF-2αfl/fldlckCRE (green) or HIF-2αfl/fl control littermate mice (gray) (n = 3, error bars represent SD).(F) Relative mRNA levels of Vegfa (left, n = 3) and amount of VEGF-A in media (right, n = 4, error bars represent SD) on CD8+ T cells for the indicated time points after αCD3/CD28 activation.(G) Amount of VEGF-A in media from CTLs expanded as in (A) and cultured under 21% O2 or 1% O2 for 24 hr (n = 4, error bars represent SD).(H) Relative expression of Vegfa in total LLC tumor lysate and immune populations isolated from subcutaneous Lewis lung carcinoma (LLC) tumors 11 days after implantation (n = 10, error bars represent SD).Grouped data were assessed by two-way ANOVA for multiple comparisons with Bonferroni correction. ∗∗∗∗p < 0.00005, ∗∗∗p < 0.0005, ∗∗p < 0.005, ∗p < 0.05. MFI, mean fluorescence intensity. See also Figure S2.

Fig 5: HIF-1α Is Necessary for Effector CD8+ T Cell Function and Migration(A) CD8+ T cells were isolated from spleens of OT-1 HIF-1αfl/fldlckCRE (maroon) and littermate control (black) mice and activated with cognate peptide for 2 days, then expanded for 5 days in the presence of IL-2 and subjected to 21%, 5%, or 1% O2 for 24 hr. Representative flow-cytometry histograms showing the expression of CD44 and CD62L (n = 3).(B) Expression of intracellular granzyme B and the indicated costimulatory molecules/checkpoint receptors on CD8+ T cells prepared as in (A).(C) Left: diagram outlining the in vivo migration experiment: representative flow-cytometry plots are shown for each pool before and after adoptive cotransfer of HIF-1α mutant (CD45.1+/TdTomato+, maroon) and control (CD45.1+, black) OT-1 T cells into CD45.2+ B16-OVA tumor-bearing WT mice. Right: spleens, lymph nodes (LN), and tumors were collected 48 hr after the adoptive cell cotransfer and relative percentages of migrated cells of the indicated genotypes are shown (error bars represent SD, n = 6).(D) 1 × 106 OT-1 cells were transferred into recipient mice harboring B16-OVA tumors (n = 6): tumor volumes (left) and overall percent survival (right) are shown; statistical analysis by log-rank (Mantel-Cox) test; error bars represent SEM.(E) Tumor growth curves of MC38 tumor cells subcutaneously injected into HIF-1α mutant and littermate controls. Mice were treated with either isotype control antibodies (left, n = 3) or a combination of αPD-1 and αCTLA4 blocking antibodies (right, n = 6), on days 5, 7, and 9 (dashed lines). Statistical analysis was performed by two-way ANOVA with Sidak correction for multiple comparisons. Tumor volumes are shown (mean values ± SEM).∗∗∗∗p < 0.00005, ∗∗p < 0.005, ∗p < 0.05; n.s., not significant. See also Figure S3.