Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2019 Mar 20;10(1):1280.
doi: 10.1038/s41467-019-09263-1.

ER Stress-Induced Mediator C/EBP Homologous Protein Thwarts Effector T Cell Activity in Tumors Through T-bet Repression

Affiliations
Free PMC article

ER Stress-Induced Mediator C/EBP Homologous Protein Thwarts Effector T Cell Activity in Tumors Through T-bet Repression

Yu Cao et al. Nat Commun. .
Free PMC article

Erratum in

Abstract

Understanding the intrinsic mediators that render CD8+ T cells dysfunctional in the tumor microenvironment is a requirement to develop more effective cancer immunotherapies. Here, we report that C/EBP homologous protein (Chop), a downstream sensor of severe endoplasmic reticulum (ER) stress, is a major negative regulator of the effector function of tumor-reactive CD8+ T cells. Chop expression is increased in tumor-infiltrating CD8+ T cells, which correlates with poor clinical outcome in ovarian cancer patients. Deletion of Chop in T cells improves spontaneous antitumor CD8+ T cell immunity and boosts the efficacy of T cell-based immunotherapy. Mechanistically, Chop in CD8+ T cells is elevated primarily through the ER stress-associated kinase Perk and a subsequent induction of Atf4; and directly represses the expression of T-bet, a master regulator of effector T cell function. These findings demonstrate the primary role of Chop in tumor-induced CD8+ T cell dysfunction and the therapeutic potential of blocking Chop or ER stress to unleash T cell-mediated antitumor immunity.

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1
Fig. 1
Increased Chop in CD8+ tumor-infiltrating T lymphocytes (TILs) correlates with poor survival in ovarian cancer. a Ddit3 mRNA levels in tumor-associated CD45+ CD8+ T cells (TILs) sorted from subcutaneous 3LL, EL-4, MCA-38, or B16 tumors and CD8+ T cells from the spleens of the same tumor-bearing mice (Tumor bearing) or tumor-free mice (Tumor free). Bar graphs show the mean ± s.e.m. (n = 5 mice/group). b Chop expression in CD8+ TILs from B16 melanoma tumors (left) and 3LL tumors (right), compared with splenic CD8+ T cells from the corresponding tumor-bearing mice. Chop was detected by fluorescence-activated cell sorter and levels indicated by mean fluorescence intensity (MFI). Representative findings from four repeats. c CHOP in CD8+ TILs from ovarian carcinoma patients (Ovarian Ca tumor, n = 18) compared to peripheral blood CD8+ T cells from ovarian carcinoma patients (Ovarian Ca blood, n = 11) or healthy controls (Healthy blood, n = 6). d CHOP levels in autologous CD8+ TILs (Tumor) and peripheral blood CD8+ T cells (Blood) from ovarian carcinoma patients (n = 7). e Representative image (scale 10 μm) showing CHOP expression in CD8+ TILs from ovarian carcinoma patients compared to CD8+ T cells from healthy ovarian tissues. Isotype (red) or CHOP (red, tested by clone 9C8 (left) or polyclonal antibody R-20 (right)), CD8 (green) and DAPI (blue) were detected by confocal microscopy. f Percentage of nuclear CHOP+ cells (clone 9C8) among tumor-associated CD8+ T lymphocytes in a tissue microarray containing advanced ovarian carcinoma tissues (n = 87) vs. normal ovarian tissues (n = 12). g Overall survival in advanced ovarian tumor patients having increased frequency of nuclear CHOP in CD8+ TILs (CHOPhigh) (n = 52) vs. those having low frequency of nuclear CHOP in CD8+ TILs (CHOPlow) (n = 29) (logrank 4.39, p = 0.0361 using Gehan–Breslow–Wilcoxon test; cutoff was established as described in the Methods section). Studies were developed using the anti-CHOP antibody clone 9C8. h Percentage of CD8+ TILs having nuclear CHOP (clone 9C8) in ovarian cancer patients that had optimal (n = 59) vs. suboptimal (n = 23) cytoreductive debulking surgery. Bar graphs represent mean value ± s.e.m., *p < 0.05, **p < 0.01, ***p < 0.001 were calculated using two-tailed unpaired Student’s t test
Fig. 2
Fig. 2
Perk regulates Chop expression in primed CD8+ T cells and CD8+ TILs. a Upper panel: Time-dependent induction of Chop in murine (left) and human (right) T cells primed in vitro. T cells were stimulated with anti-CD3/CD28 and collected at the indicated time points (0–72 h). Lower panel: Densitometry quantitation of immunoblots (n = 3). b Carboxyfluorescein succinimidyl ester-labeled Pmel CD8+ T cells (CD90.1+) were transferred into wild-type mice (CD90.2+), followed or not by vaccination with gp10025–33 plus IFA. On day 4, Chop levels by mean fluorescence intensity (MFI) (left) and percentage of Chop+ cells (right) were established by fluorescence-activated cell sorter (FACS) in gated CD8+ CD90.1+ T cells from lymph nodes of vaccinated vs. non-vaccinated mice (n = 3). c Left: Tauroursodeoxycholic acid (0.5 mM, at time 0) or Thapsigargin (100 nM, after 48 h) were added to CD8+ T cells stimulated with plate-bond anti-CD3/CD28. Protein extracts were collected 72 h post-activation. Right: Quantitation of immunoblots (n = 3). d Unfolded protein response mediators Perk, Ire-1α, and Atf4 in stimulated murine (left) and human (right) T cells (0–72 h). Representative Immunoblot from n = 3. e CD8+ T cells from Eif2ak3flox or Eif2ak3T cell-KO mice were primed with anti-CD3/CD28 for 0–72 h. Upper panel: Chop and Perk detected by immunoblotting; lower panel: Densitometry quantitation of immunoblots (n = 3). f Chop MFI levels in CD8+ TILs from B16-bearing Eif2ak3flox and Eif2ak3T cell-KO mice (n = 3). Chop was detected in viable CD45+CD8+ TILs by FACS (left) and MFI values compiled (right). g Representative reactive oxygen species levels in murine and human DHE-labeled CD8+ T cells activated with anti-CD3/CD28 and treated for 24 h with 5% cell-free ovarian ascites obtained from mice bearing ID8-Defb29/Vegf-a ovarian tumors or 5% primary ascites from patients with ovarian cancer, respectively (n = 4). h Murine CD8+ T cells were primed with anti-CD3/CD28 for 48 h and then cultured with or without ID8-Defb29/Vegf-a cell-free ovarian tumors ascites for 24 h (n = 3). i CD8+ T cells treated as in h were cultured in the presence or the absence of 2 mM L-NAC during the ascites treatment (n = 3). Bar graphs represent mean ± s.e.m., *p < 0.05 was calculated using two-tailed unpaired Student’s t test
Fig. 3
Fig. 3
Chop negatively regulates effector CD8+ T cell activity. a Gene set enrichment analysis was performed to determine the specific enrichment in effector CD8+ T cell gene signature in primed wild-type or Ddit3−/− Pmel CD8+ T cells. b Heatmap showing the expression of selective effector function-related transcripts in primed wild-type vs. Ddit3−/− Pmel CD8+ T cells, as described in the Methods section. c Carboxyfluorescein succinimidyl ester (CFSE)-labeled wild-type or Ddit3−/− CD8+ T cells were stimulated with anti-CD3/CD28 and proliferation tested after 72 h by fluorescence-activated cell sorter (FACS). Left: representative histogram of T cell proliferation (n = 6); right: T cell proliferation rates under different concentrations of anti-CD3/CD28 (µg/ml) (n = 6). d Granzyme B protein (upper panel, long- and short-time exposure) and Gzmb mRNA levels (lower panel) in wild-type or Ddit3−/− CD8+ T cells primed with anti-CD3/CD28 for 72 h (n = 5). e Percentage of IFNγ+ cells in wild-type or Ddit3−/− CD8+ T cells primed as in (d). Right: representative FACS findings; left: merged percentage values from n = 4. f Extracellular acidification rate (ECAR) of wild-type or Ddit3−/− CD8+ T cells primed as in d upon glycolysis stress analysis (n = 3). g Oxygen consumption rate (OCR) of wild-type or Ddit3−/− CD8+ T cells activated as in d after mitochondrial stress analysis (n = 3). h, i Percentage of IFNγ+ cells in CD8+ T cells primed in the presence of Tauroursodeoxycholic acid (n = 4) (h) or from Eif2ak3flox or Eif2ak3T cell-KO mice (n = 5) (i). j In vitro cytotoxicity of wild-type and Ddit3−/− CD8+ Tcells was assessed by measuring EL-4 cell proportion after co-culturing gp10025–33 pre-activated wild-type or Ddit3−/− Pmel CD8+ T cells with EL-4 cells loaded with gp10025–33 or control peptide (high or low CFSE, respectively). Cytotoxicity was evaluated after 24 h of co-culture by FACS. Right: representative FACS result; left: merged results from n = 5. k Representative result of CD44high CD62L effector memory cells in wild-type or Ddit3-null CD8+ T cells primed as in d (n = 10). Bar graphs represent mean ± s.e.m., *p < 0.05, **p < 0.01 were calculated using two-tailed unpaired Student’s t test
Fig. 4
Fig. 4
Chop negatively regulates T-bet expression. a Time-dependent expression (upper panel) and corresponding densitometry quantitation (lower panel) of T-bet in primed wild-type and Ddit3−/− CD8+ T cells. Left: protein level (0–72 h); right: Tbx21 mRNA levels 48 h post-activation. CD8+ T cells were stimulated with plate-bound anti-CD3/CD28 (n = 3). b Tbx21 and Ifng mRNA expression in activated CD8+ T cells infected with control retrovirus (Retro-Ctrl) or Ddit3-expressing retrovirus (Retro-Chop). Cells were primed for 48 h and then infected for additional 48 h in the presence of the stimulating anti-CD3/CD28 antibodies (n = 4). c Ifng, Il12b2, Cbfa3, and Cxcr3 mRNA levels in control vs. Ddit3−/− CD8+ T cells primed as in a (n = 5). d Predicted Chop-binding site in the Tbx21 promoter region (GGGTGCAATCTTC). e Chromatin immunoprecipitation assay for the endogenous binding of Chop to Tbx21 promoter in primed wild-type or Ddit3−/− CD8+ T cells. Chop-binding activity was measured by real-time quantitative PCR, compared with IgG binding activity after normalizing to the activity of anti-H3 (n = 4). f A dual luciferase system composed of 2x-CRE containing Firefly luciferase reporter and the control Renilla luciferase reporter was transfected into 293T cells in combination with Ddit3-expressing or control vectors. n = 4 experimental repeats. g Expression of Chop (left) and T-bet (right) by fluorescence-activated cell sorter (FACS) upon transduction of primed CD8+ T cells with green fluorescent protein (GFP)-coding retroviruses containing control or 8x-CRE sequences. Cells were primed for 48 h and then infected for another 48 h in the presence of the stimulating anti-CD3/CD28 antibodies plus interleukin (IL)-2 (50 U/ml). n = 3 independent repeats. h Interferon-γ (IFNγ) levels in primed CD8+ T cells transduced with: (1) GFP/CD90.1-expressing control virus (Ctrl); (2) Chop/CD90.1-expressing virus and GFP-expressing control virus (Chop); (3) CD90.1-expressing control virus and T-bet/GFP-expressing virus (T-bet); or (4) Chop/CD90.1-expressing virus and T-bet/GFP-expressing virus (Chop/T-bet). Cells were primed for 24 h and then transduced for additional 72 h in the presence of stimulating antibodies plus IL-2 (50 U/ml). Then IFNγ levels were detected by FACS in gated CD90.1+GFP+ cells. Right: Representative FACS result; left: Merged results from three independent experiments. In the bar graphs showing mean ± s.e.m., *p < 0.05, **p < 0.01 were calculated using two-tailed unpaired Student’s t test
Fig. 5
Fig. 5
Ddit3 null T cells exert promoted protective antitumor immunity. a Tumor growth in Ddit3flox (blue) or Ddit3T cell-KO (red) mice bearing B16 tumors. Average kinetics ± s.e.m (n = 15 mice/group). b Left: Representative result of IFNγ+ (upper panel) and TNF-α+ (lower panel), and right: IFNγ+ TNF-α+ in gated CD45+ CD8+ TILs from Ddit3flox or Ddit3T cell-KO mice bearing B16 tumors. Tumors were collected 18 days after B16 tumor injection. (n = 10 tumors/group). c Percentage of IFNγ+ (left), TNF-α+ (middle) and IFNγ+ TNF-α+ (right) CD8+ tumor-infiltrating T lymphocytes (TILs) in B16 tumors isolated from Ddit3flox or Ddit3T cell-KO mice. (n = 10 mice/group). d Frequency of CD44+ CD69+ effector CD8+ TILs in B16 tumors isolated from Ddit3flox or Ddit3T cell-KO mice. (n = 8 tumors/group). e Percentage of KLGR1high CD127low late effector CD8+ TILs in B16 tumors isolated from Ddit3flox or Ddit3T cell-KO mice. (n = 8 tumor/group). f B16 tumor growth in Ddit3flox (blue) or Ddit3T cell-KO (red) mice, with or without 400 µg α-CD8 antibody injection on every third day (green or orange, respectively). Average tumor volume kinetics ± s.e.m in 6 mice/group. g Tumor growth in Eif2ak3flox (blue) or Eif2ak3T cell-KO (red) mice bearing B16 tumors. Average kinetics of tumor volume ± s.e.m in 8 mice/group. In the bar graphs showing mean ± s.e.m., *p < 0.05, **p < 0.01, ***p < 0.001 were calculated using two-tailed unpaired Student’s t test
Fig. 6
Fig. 6
Ddit3 deletion or silencing increases the effect of T cell-based immunotherapy. a Working model of Pmel T cell adoptive transfer. CD90.2+ wild-type mice were injected subcutaneous with B16 tumors, and after 8 days, they received a single dose of cyclophosphamide (CTX). The next day, mice received 1 × 106 primed wild-type or Ddit3−/− or anti-sense oligonucleotide (ASO)-treated CD90.1+CD8+ Pmel T cells. b Tumor growth in B16-bearing mice (black) transferred with wild-type (blue) or Ddit3−/− (red) CD90.1+ CD8+ Pmel T cells. Average kinetics of tumor volume ± s.e.m. in 9 mice/group. c Percentage of IFNγ+ cells in CD90.1+ CD8+ tumor-infiltrating T lymphocytes (TILs) from B16 tumors 5 days after wild-type or Ddit3−/− CD8+ Pmel T cell transfer. (n = 7 tumors/group). d ELISpot for interferon-γ (IFNγ) in the spleens of B16 tumor-bearing mice 5 days after wild-type or Ddit3−/− Pmel T cell transfer exposed to gp10025–33 for 24 h. (n = 6 spleens/group). e T-bet in CD90.1+CD8+ TILs from B16 tumors 5 days after wild-type or Ddit3−/− CD8+ Pmel T cell transfer. Right: representative result; left: merged results from n = 4/group. f Tumor growth in B16-bearing CD90.2+ mice (black) that received wild-type CD90.1+CD8+ Pmel T cells that were non-treated (gray) or pretreated with ASO-Control (blue) or ASO-Ddit3 (red). Average tumor volume kinetics ± s.e.m. of 8 mice/group. g Percentage of IFNγ+ in CD90.1+ CD8+ TILs from B16 tumor-bearing mice transferred with control, ASO-Control, or ASO-Ddit3-pretreated CD90.1+ Pmel T cells. (n = 6 tumors/group). h IFNγ ELISpot in spleens collected from B16 tumor-bearing mice treated with Pmel T cells or ASO-Control or ASO-Ddit3-pretreated Pmel T cells. Spleens were collected 5 days after T cell transfer and exposed 24 h to gp10025–33. (n = 4). Bar graphs show mean value ± s.e.m. and *p < 0.05, ***p < 0.001 were calculated using two-tailed unpaired Student’s t test

Similar articles

See all similar articles

Cited by 6 articles

See all "Cited by" articles

References

    1. Schreiber RD, Old LJ, Smyth MJ. Cancer immunoediting: integrating immunity’s roles in cancer suppression and promotion. Science. 2011;331:1565–1570. doi: 10.1126/science.1203486. - DOI - PubMed
    1. Chen DS, Mellman I. Oncology meets immunology: the cancer-immunity cycle. Immunity. 2013;39:1–10. doi: 10.1016/j.immuni.2013.07.012. - DOI - PubMed
    1. Chen DS, Mellman I. Elements of cancer immunity and the cancer-immune set point. Nature. 2017;541:321–330. doi: 10.1038/nature21349. - DOI - PubMed
    1. Ruter J, et al. Altering regulatory T cell function in cancer immunotherapy: a novel means to boost the efficacy of cancer vaccines. Front. Biosci. (Landmark Ed.) 2009;14:1761–1770. doi: 10.2741/3338. - DOI - PubMed
    1. Nagarsheth N, Wicha MS, Zou W. Chemokines in the cancer microenvironment and their relevance in cancer immunotherapy. Nat. Rev. Immunol. 2017;17:559–572. doi: 10.1038/nri.2017.49. - DOI - PMC - PubMed

Publication types

MeSH terms

Feedback