Article
Open access
Published: 01 October 2025

Proteotoxic stress response drives T cell exhaustion and immune evasion

Nature (2025)Cite this article

Subjects

Abstract

Chronic infections and cancer cause T cell dysfunction known as exhaustion. This cell state is caused by persistent antigen exposure, suboptimal co-stimulation and a plethora of hostile factors that dampen protective immunity and limit the efficacy of immunotherapies^1,2,3,4. The mechanisms that underlie T cell exhaustion remain poorly understood. Here we analyse the proteome of CD8⁺ exhausted T (T_ex) cells across multiple states of exhaustion in the context of both chronic viral infections and cancer. We show that there is a non-stochastic pathway-specific discordance between mRNA and protein dynamics between T effector (T_eff) and T_ex cells. We identify a distinct proteotoxic stress response (PSR) in T_ex cells, which we term T_ex-PSR. Contrary to canonical stress responses that induce a reduction in protein synthesis^5,6, T_ex-PSR involves an increase in global translation activity and an upregulation of specialized chaperone proteins. T_ex-PSR is further characterized by the accumulation of protein aggregates and stress granules and an increase in autophagy-dominant protein catabolism. We establish that disruption of proteostasis alone can convert T_eff cells to T_ex cells, and we link T_ex-PSR mechanistically to persistent AKT signalling. Finally, disruption of T_ex-PSR-associated chaperones in CD8⁺ T cells improves cancer immunotherapy in preclinical models. Moreover, a high T_ex-PSR in T cells from patients with cancer confers poor responses to clinical immunotherapy. Collectively, our findings indicate that T_ex-PSR is a hallmark and a mechanistic driver of T cell exhaustion, which raises the possibility of targeting proteostasis pathways as an approach for cancer immunotherapy.

Main

T cell exhaustion represents a hypofunctional state characterized by reduced effector function and increased inhibitory receptor expression that arises from persistent antigen exposure and a hostile microenvironment⁷. T_ex cells observed in cancer fail to eliminate malignant cells, and this limitation mediates a key mechanism of resistance to immunotherapies^1,2,3. The exhaustion program generates a heterogeneous T_ex cell population. Progenitor T_ex (T_prog) cells retain stemness and self-renewal capacity that respond to immune checkpoint blockade (ICB) therapies and differentiate into intermediate T (T_int) cells with cytolytic capacity^8,9,10. Conversely, terminal T_ex (T_tex) cells accumulate over time and respond poorly to ICB therapies^8,11,12,13. T cell exhaustion also limits the efficacy of chimeric antigen receptor (CAR) T cell therapy against solid tumours^14,15,16,17. Consequently, a better understanding of T cell exhaustion is essential to overcome the limitations of current immunotherapies.

Although transcriptomic profiling has provided insights into T_ex cell biology, mRNA abundance is not always a faithful proxy of protein expression across various organisms^{18,19,20,21,22}. Previous studies have revealed poor mRNA–protein correlation in T cells regardless of functional status^23,24,25 and the importance of post-transcriptional regulation in T cell differentiation and function²⁶. In this context, a high-resolution proteomic map of T_ex cells would be valuable. In this study, we define the proteomic landscape of T_ex cells across various settings, including an in vitro exhaustion model, in vivo chronic lymphocytic choriomeningitis virus (LCMV) infection in mice, and colon tumour and bladder tumour mouse models. We demonstrate that there is pathway-specific discordance between transcript and protein levels.

We also elucidate the intricate layers of protein-level regulation pertaining to a PSR that is specific to T_ex cells. We show that PSR in T_ex cells shares similarities to unfolded protein responses and integrated stress responses. However, a marked difference is that PSR in T_ex cells is characterized by increased global protein synthesis. This T_ex cell-associated PSR signature, which we term T_ex-PSR, is further marked by the selective activation of chaperone proteins such as gp96 (also known as GRP94; encoded by Hsp90b1) and BiP (encoded by Hspa5) and the accumulation of protein aggregates that are predominantly driven by sustained activation of the AKT pathway. The introduction of misfolded proteins alone could convert T_eff cells to T_ex cells, which demonstrates the causality of dysregulated proteostasis in T cell exhaustion. Finally, we demonstrate that T_ex-PSR is also a hallmark of human T_ex cells in cancer and that it may contribute to resistance to cancer immunotherapy.

Discordance between RNA and protein levels

To determine whether gene expression levels reflect protein expression levels in T cells, we used an established in vitro exhaustion model that induces T cell exhaustion through repeated T cell receptor (TCR) stimulation^27,28. We then performed parallel RNA sequencing and quantitative proteomics by mass spectrometry (MS) (Extended Data Fig. 1a). Compared with acutely activated T cells (T_eff cells), chronically stimulated cells (T_ex cells) exhibited impaired survival, proliferation and cytokine production, which were accompanied by increased expression levels of exhaustion markers, including PD1 and TIM3 (Extended Data Fig. 1b,c). To improve detection sensitivity, quantification and reproducibility, we applied the chromatogram library approach for all proteomic data collection (Extended Data Fig. 1d). The expression dynamics of key activation, exhaustion and stemness markers of T cells, such as CD25, PD1, CD39 and TCF1, obtained from the MS results aligned well with data obtained by spectral flow cytometry (Extended Data Fig. 1e).

We then ascertained whether transcript levels are a reliable surrogate for protein expression levels in T cells. In both T_eff and T_ex cells, mRNA and protein expression levels were weakly correlated, as indicated by Pearson’s correlation coefficients of 0.31 and 0.38, respectively (Extended Data Fig. 2a). Moreover, the degree of correlation between RNA and protein levels did not seem to be stochastic but were functionally related (Extended Data Fig. 2b,c). A group of proteins that exhibited comparable mRNA and protein expression levels in >300 cancer cell lines²⁹, termed ‘housekeeping’ in this context, aligned well with RNA expression levels in T cells (Extended Data Fig. 2b). Proteins involved in the regulation of TCR signalling, cell death and cytokine responses exhibited a similarly strong correlation with RNA expression levels, with correlation coefficient values of around 0.7. By comparison, transcription factors (TFs) exhibited a moderate correlation between mRNA and protein expression levels, with some TFs showing detectable changes exclusively at the protein level, including FOXO1 and T-bet (Extended Data Fig. 2c,d). Furthermore, the levels of proteins associated with metabolic processes, post-transcriptional regulation and epigenetic regulation aligned poorly with mRNA levels (Extended Data Fig. 2c). The three major metabolic pathways—glycolysis, oxidative phosphorylation and fatty acid metabolism—showed discrepancies between RNA and protein levels, with most of the changes in mRNA levels not reflected at the protein level (Extended Data Fig. 2e). These results underscore the importance of directly defining the proteome rather than inferring it from the transcriptome.

PSR and specialized chaperone enrichment

We next generated a kinetic proteomic landscape of T cells during T cell exhaustion by leveraging the LCMV infection model. Antigen-specific CD8⁺ T cell subpopulations were isolated after acute (Armstrong strain) or chronic LCMV (clone 13 strain) infection and analysed (Fig. 1a). The following endogenous CD8⁺ T cell subpopulations specific for viral antigens (gp33 and gp276) were sorted for MS analysis at days 8 and 30 after infection: short-lived effector T cells (SLECs), memory precursor effector cells (MPECs) and central memory (T_CM) and effector memory (T_EM) T cells from Armstrong infection; and SLAMF6⁺CX3CR1^– T_prog, CX3CR1⁺ T_int and SLAMF6^–CX3CR1^– T_tex cells from clone 13 infection (Extended Data Fig. 3a). We applied a previously defined transcriptomic signature for T cell exhaustion³⁰ to analyse our proteomic data. Notably, in addition to increased expression of protein signatures associated with exhaustion and pro-apoptosis, T_tex cells exhibited a marked enrichment of proteins in the PSR pathway (Fig. 1b). We applied the same gene signatures to a single-cell RNA sequencing (scRNA-seq) dataset of mouse gp33⁺CD8⁺ T cells after LCMV clone 13 infection³¹. The upregulation of PSR in T_ex cells was readily discerned at the protein but not the transcript level (Extended Data Fig. 3b). We also generated a proteomic dataset of transgenic T cells against a D^b-restricted gp33 epitope of LCMV (called P14) at days 8, 15 and 30 after infection³² (Extended Data Fig. 3c). PSR was similarly upregulated in P14 cells after clone 13 infection but not Armstrong infection (Extended Data Fig. 3d).

**Fig. 1: PSR is triggered in T_ex cells with a dynamic expression of chaperone proteins.**

To further investigate proteome dynamics and PSR in T cell exhaustion, we performed proteomic analyses of T_eff and T_ex cells generated in vitro, which enabled us to achieve increased proteome coverage. T cells were collected on days 2, 4, 6 and 8 after initial activation to track proteomic dynamics during T cell activation and exhaustion as described above (Extended Data Fig. 1a). T cells exhibited distinct protein expression dynamics depending on the differentiation state (Fig. 1c). The expression dynamics of key activation, exhaustion and stemness markers, such as TCF1, SLAMF6, PD1 and TIM3, obtained by MS analyses aligned well with T cell states (Extended Data Fig. 4a). Overall, eight distinct clusters of proteins were identified (Fig. 1c and Extended Data Fig. 4b). Notably, proteins in cluster 6 were specifically upregulated in T_ex cells, which typically expressed exhaustion markers such as TIM3, CD39 and LAG3 (Fig. 1c), as well as molecules involved in protein transport, modification and quality control (Fig. 1d). In particular, proteins involved in the endoplasmic reticulum (ER) stress response were significantly increased as T cells became more exhausted (Fig. 1e). An increase in the expression of proteins integral to translation, transport and quality control in T_ex cells compared with T_eff cells indicates that there is an induction of a distinct PSR during the T cell exhaustion process (Fig. 1e).

We next performed a proteomic study of T cells isolated ex vivo from the tumour milieu of mouse models of MC38 colon cancer and MB49 bladder cancer (Fig. 1f). Antigen-experienced CD44^hiCD8⁺ T cells were sorted by flow cytometry into T_prog, T_int and T_tex cells^8,11 (Fig. 1g). In the MC38 tumour model, the T_tex population exhibited upregulated proteins associated with the ER stress response and proteins associated with autophagy and transport (Fig. 1h). The upregulation of ER stress proteins was also observed in T cells isolated from MB49 bladder tumours (Fig. 1h).

Chaperone proteins are crucial for ensuring protein quality control. Notably, we observed a heterogeneous expression of chaperones (Fig. 1i). On the basis of their expression patterns, we categorized them into three groups: quiescence, activation and PSR chaperones (Fig. 1i). The quiescence-related chaperones showed the highest expression in naive T cells. The second group of chaperones, compromising cytosolic HSP90α and HSP90β, TRiC complex subunits (TCP1α–TCP1θ) and mitochondrial chaperone HSP75 (also known as TRAP1), were induced by TCR stimulation but reduced in T_ex cells, which implicated a link with T cell activation. Conversely, proteins in the third group, including several ER chaperones such as BiP and gp96, were overexpressed in T_ex cells, which indicated that they may have specialized roles in T cell exhaustion. These exhaustion-associated chaperones were significantly upregulated in the T_tex cell population from both MC38 and MB49 tumours (Fig. 1j). We performed additional analyses of proteomes of LCMV-specific T cells and tracked the expression kinetics of these chaperones during chronic infection (Extended Data Fig. 3c). Consistently, BiP, gp96 and HSPA13 were upregulated in P14 cells from mice infected with clone 13, whereas the expression of TRiC complex subunits was downregulated (Extended Data Fig. 3e). Together, these findings indicate that the activation of PSR and the associated upregulation of PSR chaperones are a common hallmark of T cell exhaustion.

Dysregulated proteostasis in T_ex cells

Our proteomic data indicated that PSR is activated in T_ex cells. We next examined whether proteostasis in T_ex cells is disrupted. We stained T_eff and T_ex cells with the fluorescent dye NIAD-4 or its derivative CRANAD-2, which bind to amyloid-like structures enriched in misfolded proteins and aggregates³³. T_ex cells showed a significant accumulation of protein aggregates compared with T_eff cells (Fig. 2a,b). In tumour-infiltrating T cells (TILs) isolated from MC38 and MB49 tumours, protein aggregates progressively increased as T cells became more exhausted (Fig. 2c,d). This result suggests that the protein quality control system is significantly impaired in T_ex cells.

**Fig. 2: Chronic TCR stimulation disrupts proteostasis during T cell exhaustion.**

The activation of PSR, the upregulation of specialized chaperone proteins and the accumulation of protein aggregates in T_ex cells led us to initially predict that T cell exhaustion is associated with attenuated translation to prevent further protein overload. However, this prediction was incorrect based on the following considerations. First, proteins with internal ribosome entry sites (IRES) that are not subject to EIF2α-mediated translation attenuation should be enriched relative to proteins without IRES in T_ex cells³⁴. However, this was not the case (Extended Data Fig. 5a). Second, we did not observe enrichment of proteins that rely on the alternative translation initiation factor EIF2A owing to the presence of an unconventional 5′ upstream open reading frame³⁵ (Extended Data Fig. 5b). Third, proteins with concurrent upregulation at RNA and protein levels in T_ex cells were overrepresented in translation initiation, elongation and stress-response pathways (Extended Data Fig. 5c). Fourth, levels of the translation repressor PDCD4 (refs. ^36,37) were significantly reduced in T_ex cells both in vitro and in vivo (Extended Data Fig. 5d). Fifth, our proteomic analysis revealed an upregulation of translation but not transcriptional machinery in T_ex cells (Extended Data Fig. 5e–g). Proteins involved in translation were upregulated in T_ex cells induced by chronic LCMV clone 13 infection (Extended Data Fig. 6a). Moreover, the translation initiation factors EIF2D and EIF4G3 exhibited sustained high expression in P14 cells from mice with chronic infection but were reduced over time in mice with acute infection (Extended Data Fig. 6b).

These findings prompted us to directly assess translation rates by measuring the incorporation of l-homopropargylglycine (HPG), a methionine analogue, into newly synthesized proteins. Notably, T_ex cells exhibited a significantly increased rate of protein synthesis compared with acutely stimulated T_eff cells (Fig. 2e). We also measured translation changes in vivo in different TIL subsets by injecting O-propargyl-puromycin (OPP) into mice with MC38 tumours to label elongating polypeptide chains during active translation³⁸. T_tex cells showed significantly higher OPP incorporation than T_prog and T_int cells (Fig. 2f). We validated this result by isolating TILs into single-cell suspensions followed by ex vivo HPG translation assays. This experiment was performed to further exclude the possibility that different spatial distributions of T cell subpopulations caused inconsistent access to OPP. Consistently, T_tex cells showed increased protein translation rates in both MC38 and MB49 tumour models (Extended Data Fig. 6c,d).

Next, we characterized important subcellular events typically associated with PSR. Stress granules (SGs) are dynamic, reversible protein and RNA granules that form under cellular stress and are evident during T cell activation^39,40. The formation of SGs in T_ex cells was increased, as evidenced by both analyses of morphology and expression of the SG marker G3BP1 (Extended Data Fig. 7a–c). We further explored the functional roles of SGs in T_ex cells. Disruption of SGs by knocking out G3bp1 resulted in an increased production of the cytokines IFNγ and TNF (Extended Data Fig. 7d,e). However, loss of G3bp1 significantly compromised the survival of chronically stimulated T_ex cells (Extended Data Fig. 7f).

T_ex cells also exhibited increased proteasome and lysosome activity (Extended Data Fig. 7g,h). We therefore investigated the kinetics of global protein catabolism in T_ex cells. Cells were labelled with HPG for 30 min, then changed to regular culture conditions without HPG. Although nascent proteins were produced in high levels, they were rapidly degraded in T_ex cells (Extended Data Fig. 7i). This protein catabolism in T_ex cells was driven largely by autophagy (Extended Data Fig. 7j).

Taken together, these results show that T_ex cells have a distinct non-canonical PSR, which we term T_ex-PSR. T_ex-PSR is characterized by the induction of PSR and the formation of SGs, protein aggregate accumulation and increased protein catabolism, coupled paradoxically by enhanced global protein synthesis.

Molecular definition of protein aggregates

We next asked what proteins were prone to aggregation in T_ex cells. To that end, we used native gel electrophoresis to analyse the migration pattern of proteomes (Fig. 2g). On the basis of proteome characterization, we focused on the following three differentiation states of T cells: T_eff cells (day 4 acutely activated T cells); early T_ex cells (day 4 chronically activated T cells); and late T_ex cells (day 8 chronically activated T cells) (Extended Data Fig. 4c,d). Both early and late T_ex cells showed upregulation of ER stress responses compared with T_eff cells (Extended Data Fig. 4e,f). To define aggregation-prone proteins, T_eff, early and late T_ex cells were first lysed with mild lysis buffer to maintain their native conformation followed by high-speed centrifugation to remove nucleoli and other insoluble materials. Supernatants were then subjected to native PAGE electrophoresis. Proteins that migrated below 140 kDa (low molecular weight (LMW) species) and above (high molecular weight (HMW) species) from the gel were then defined by MS. We detected and quantified 3,889 proteins, and 2,878 of these proteins (74%) shifted from LMW to HMW species in T_ex cells. This finding indicated that there was a large-scale level of protein aggregation in the exhaustion state (Fig. 2h). We next examined whether proteins associated with specific pathways were preferentially aggregated in T_ex cells. However, there was no such preference, which suggested that protein aggregation occurred globally (Fig. 2i). Still, effector molecules, including granzyme B, granzyme C and perforin, in T_ex cells showed significant enrichment in HMW species compared with T_eff cells (Fig. 2j). The T_ex-PSR chaperones gp96 and BiP also showed a trend in moving towards HMW species. Notably, AKT1 did not demonstrate signs of aggregation in T_ex cells (Fig. 2j). We also resuspended and profiled proteins in insoluble material through additional harsh lysis (Extended Data Fig. 8a). We observed an enrichment of granzyme B in T_ex cells compared with T_eff cells (Extended Data Fig. 8b). The inhibitory receptor PD1, despite being a transmembrane protein, was retained in the soluble fraction (Extended Data Fig. 8c).

Next, we used immunoblotting to validate the aggregation state of some of the identified proteins (Fig. 2k). In T_ex cells, granzyme B and perforin showed a migration pattern above 242 kDa, which was in contrast to T_eff cells (Fig. 2l,m). The T_ex-PSR chaperone gp96 also formed more abundant and distinct HMW species in T_ex cells, which was not the case with cytosolic HSP90α (Fig. 2l,m). The aggregation-prone proteins in T_ex cells were also enriched in the insoluble fraction (Fig. 2n). Taken together, these data indicate that protein quality control is severely compromised in T_ex cells, which showed a tendency of protein aggregation at the global level.

Misfolded proteins drive exhaustion

An important question is whether protein aggregation is the cause or a consequence of T cell exhaustion. We induced protein aggregation in T cells using two approaches and then programmed these cells under non-exhaustion conditions. First, we treated T_eff cells with the l-proline analogue l-azetidine-2-carboxylic acid (AZC) (Fig. 3a) to cause protein misfolding through the four-membered ring^41,42. l-Proline was not depleted in the culture medium to avoid metabolic disruption. After 6 days of treatment of T cells with AZC, apparent protein aggregation was observed (Fig. 3b). Despite culturing T cells in the non-exhaustion conditions, AZC treatment caused them to develop into the T_ex cell state, with a significantly increased PD1⁺TIM3⁺ population and impaired cytokine production (Fig. 3c,d). As expected, T_ex cells were vulnerable to AZC treatment, with significantly more cell death (data not shown). We also analysed AZC-treated cells in the tumour environment. OT-1 cells were activated and transiently pulsed with AZC for 3 days before adoptive transfer into mice with B16-OVA tumours (Fig. 3e). AZC-treated OT-1 cells showed reduced numbers in the tumours and an increased SLAMF6^–TIM3⁺ T_tex population (Fig. 3f,g). Second, we genetically induced the expression of an aggregation-prone and functionally inert protein into acutely activated T cells by retroviral transduction (Fig. 3h). Cystic fibrosis transmembrane conductance regulator (CFTR) is an ion channel protein expressed primarily in epithelial cells, with low levels of expression in T lymphocytes⁴³. Deletion of phenylalanine at the 508th position (CFTR(ΔF508)) in human CFTR leads to protein misfolding and ER retention^44,45. Transduction of the folding-deficient mutant CFTR(ΔF508) resulted in the intracellular accumulation of CFTR aggregates in mouse CD8⁺ T cells (Fig. 3i). Moreover, overexpression of CFTR(ΔF508) induced bona fide T cell exhaustion without repetitive TCR stimulation (Fig. 3j). CFTR(ΔF508) also increased CD39⁺ populations in acutely activated human CD8⁺ T cells (Fig. 3k). Together, these results suggest that protein aggregation has a causal role in T cell exhaustion.

**Fig. 3: Protein misfolding promotes T cell exhaustion without chronic TCR stimulation.**

Sustained AKT activity causes exhaustion

We next aimed to elucidate the upstream signalling hub that is responsible for mediating T_ex-PSR and promoting T cell exhaustion. We examined the expression level of components of key signalling pathways defined in the Gene Ontology database in our T cell proteome dataset. The AKT pathway was specifically upregulated in the late T_ex cell population (Fig. 4a). Flow cytometry analysis further showed that AKT phosphorylation was enhanced in T_ex cells (Fig. 4b). Chronic AKT signalling was also observed in T_ex cells isolated from MC38 tumours (Fig. 4c).

**Fig. 4: Sustained AKT signalling induces T_ex-PSR and underlies T cell exhaustion.**

AKT often operates in the same signalling axis as PI3K and mTOR and has an important role in mediating T cell proliferation, survival and function^46,47. Low-dose treatment with the AKT inhibitor MK2206 (0.2 μM and 1 μM) significantly increased the SLAMF6⁺TIM3^– population and cytokine production without impairing cell viability or proliferation (Fig. 4d,e and Extended Data Fig. 9a–d,f). However, treatment with the mTOR inhibitor rapamycin or the PI3K inhibitor LY294002 at any dose did not block T cell exhaustion, as indicated by the comparable proportion of SLAMF6⁺TIM3^– populations to the untreated group (Extended Data Fig. 9c). PI3K and mTOR inhibition also did not rescue cytokine production in T_ex cells (Extended Data Fig. 9d,f). All inhibitors directed cells to differentiate into the CD44⁺CD62L⁺ population, which validated that the dose levels used had pharmacological activities in chronically stimulated T cells (Extended Data Fig. 9e). We therefore focused on AKT in subsequent studies. Moderate attenuation of AKT signalling was sufficient to reduce the protein synthesis rate in T_ex cells and to reduce protein aggregation and T_ex-PSR chaperone gp96 expression (Fig. 4f–h). These results strongly suggest that AKT has a key role in driving T_ex-PSR and T cell exhaustion.

To further determine whether AKT signalling is the upstream driver of T_ex-PSR and T cell exhaustion, we expressed myristoylated AKT (myrAKT), a constitutively active form of AKT, in T cells^48,49 (Fig. 4i). MyrAKT expression converted T_eff cells into T_ex cells with a significant induction of SLAMF6^–TIM3⁺ terminal exhausted phenotype under non-exhaustion conditions (Fig. 4j). This constitutively active AKT also upregulated protein translation, increased protein aggregation and induced the expression of the PSR chaperone gp96, which are all hallmarks of T_ex-PSR (Fig. 4k–m). To assess its functional impact, we transduced myrAKT into activated OT-1 cells and transferred them into mice with B16-OVA tumours. MyrAKT OT-1 T cells were no longer effective in controlling the growth of B16-OVA tumours in mice compared with wild-type (WT) OT-1 cells (Fig. 4n). MyrAKT-expressing T cells in the tumour microenvironment showed less tumour infiltration and more exhausted phenotypes compared with control T cells (Fig. 4o,p). We conclude that sustained AKT activation drives T_ex-PSR and T cell exhaustion.

T_ex-PSR chaperones underlie exhaustion

We next asked whether the T cell exhaustion program can be altered through the manipulation of T_ex-PSR chaperones. We initially selected ten genes that encode the following proteins that showed increased expression in T_ex cells and represent diverse molecular functions: ADAM8; annexin A2; cathepsin D (which is associated with cell death); ACADL (a fatty acid metabolic enzyme); the cytotoxic granzymes granzyme C and granzyme A; the temperature-sensitive channel protein TRPV2; and the three chaperone proteins BiP, gp96 and ERO1A (Extended Data Fig. 10a). The roles of these proteins on T cell exhaustion have not been previously defined. We knocked out these genes individually by CRISPR–Cas9 after T cell activation and then performed repetitive TCR stimulation. Gene deletion was confirmed by PCR with reverse transcription (RT–PCR) or flow cytometry (Extended Data Fig. 10b,c). Deleting each of the three chaperone genes Hspa5, Hsp90b1 and Ero1a with single guide RNAs (sgRNAs sgHspa5, sgHsp90b1 and sgEro1a, respectively) significantly enhanced cytokine production, whereas individual knockout of the other seven genes had minimal effects (Fig. 5a and Extended Data Fig. 10d). Cells deficient in BiP, ERO1A or gp96 also showed increased SLAMF6 expression, along with reduced levels of TIM3 and CD39 (Extended Data Fig. 10e–g). Although chaperone proteins are responsible for facilitating protein folding, knocking out Hsp90b1 resulted in the most significant reduction in protein aggregation, which indicated its potentially pivotal role in mediating T_ex-PSR-associated protein aggregate formation (Fig. 5b).

**Fig. 5: Targeting T_ex-PSR chaperones prevents T cell exhaustion and enhances cancer immunotherapy.**

To validate whether Hsp90b1 deletion has the same effect in counteracting T cell exhaustion in vivo, we generated CD8⁺-specific Hsp90b1 knockout (KO: E8i-cre-Hsp90b1^flox/flox) and knockdown (Het: E8i-cre-Hsp90b1^flox/WT, with 50% reduction in gp96 levels) mouse models, which were subjected to chronic LCMV clone 13 infection (Fig. 5c,d). Thirty days after infection, gp96 expression was upregulated in antigen-experienced CD8⁺ T cells from WT mice, with T_tex cells demonstrating the highest level of expression (Extended Data Fig. 10h). Hsp90b1 deletion resulted in a significant expansion of total and antigen-specific CD8⁺ T cells in spleens (Fig. 5e). There was clear evidence of reprogramming of T cells after Hsp90b1 deletion (Fig. 5f,g), with enrichment of TCF1⁺CX3CR1^– progenitor cells and CX3CR1⁺ intermediate populations, along with reduced expression of TIM3 and CD39 (Fig. 5h,i).

Similarly, we assessed whether deleting chaperone Hspa5 or Ero1a in P14 CD8⁺ T cells improves their antitumour effect in a gp33-expressing MB49 tumour model (Fig. 5j). Mice receiving Hspa5 or Ero1a KO P14 cells showed significantly better tumour control compared with those that received WT P14 cells (Fig. 5k,l). We also analysed the impact of BiP deficiency on P14 cells in the tumour microenvironment 5 days after adoptive transfer. BiP-null P14 T cells were more enriched in the cytolytic CX3CR1⁺ population and showed improved stemness, as indicated by increased TCF1 expression, compared with WT cells (Fig. 5m,n). These data collectively suggest that targeting T_ex-PSR chaperones offer a potential approach for improving adoptive T cell transfer therapy for cancer.

Clinical relevance of T_ex-PSR in cancer

Finally, we investigated the human relevance of T_ex-PSR in cancer. We performed pan-cancer CD8⁺ T cell analyses with publicly available single-cell transcriptomic datasets that encompass 17 cancer types (Extended Data Fig. 11a). These T cells were isolated directly from samples taken from patients with cancer and did not have any other in vitro manipulations. To determine whether T_ex-PSR is also a feature of tumour-associated T_ex cells in humans, we generated a T_ex-PSR signature based on our in vivo and in vitro proteomic data (Extended Data Fig. 11b). The T_ex-PSR signature consisted of genes associated with proteostasis regulation that are concurrently upregulated in T_ex cells at the mRNA and protein levels (Extended Data Fig. 11c). The expression level of mRNAs encoding T_ex-PSR signature proteins was the highest in T_tex cells compared with all other T cell subsets from the tumour samples (Extended Data Fig. 11d). We also performed pseudotime trajectory analysis based on RNA velocity on pan-cancer CD8⁺ T cells (Extended Data Fig. 11e). We observed two opposing differentiation trajectories: effector and exhaustion. The T_ex-PSR gene signature appeared early in the exhaustion trajectory, which increased proportionally as T cells become progressively exhausted (Extended Data Fig. 11f). By contrast, T_ex-PSR was significantly reduced during the effector trajectory. We analysed the T_ex-PSR score in CD8⁺ T cells from patients with liver cancer. Patients with a lower T_ex-PSR signature in CD8⁺ T cells showed better overall survival (Extended Data Fig. 11g). Moreover, a higher T_ex-PSR signature in patients with cancer also correlated with poor responses to immunotherapies, including both CAR T cells and ICBs (anti-PD1, and anti-PD1 with anti-CTLA4) (Fig. 5o and Extended Data Fig. 11h).

Discussion

The question of how protein quality control might differ between T cell activation and the exhaustion program has not been clearly answered. A key finding of our study was the activation of a distinct PSR in T_ex cells, which we term T_ex-PSR. This PSR is characterized by a high rate of protein translation, accumulation of SGs and global protein aggregation despite increased protein catabolism (Extended Data Fig. 12). The high translational rate in T_ex cells was not associated with the production of functional molecules. Our study provides an explanation to this paradox in that many proteins, such as granzymes and perforin, in T_ex cells aggregate instead of being properly folded, which may be a consequence of an overwhelmed protein quality control system. We also demonstrated that the introduction of misfolded proteins to T cells under optimal conditions for T_eff cell differentiation caused exhaustion. It has previously been reported⁵⁰ that tumour cells can evade T cell immunity by competing for methionine to alter T cell histone modifications. This result suggests that amino acid metabolism might have important roles in regulating T cell function by concurrently affecting translation and the epigenetic landscape⁵⁰. Nonetheless, the causal relationship between PSR and T_ex suggests that dysregulated proteostasis can be a viable target for immunotherapeutic purposes. In this regard, it is worth noting that the HSP90 inhibitor ganetespib, which also inhibits the T_ex-PSR chaperone gp96, has been shown to promote ICB efficacy⁵¹.

Our work identified AKT signalling as a central regulator of T_ex-PSR and T cell exhaustion. It is well established that PI3K–AKT–mTOR signalling is essential for T cell activation and differentiation by upregulating metabolic programs and supporting their bioenergetic needs^46,47. However, the implications of AKT in T cell exhaustion are controversial and under-explored^52,53. We demonstrated here that T_ex cells maintain chronic AKT signalling. Enforced expression of constitutively active AKT drives T_ex-PSR and a bona fide T cell exhaustion program. We therefore posit that AKT signalling is required for T cell survival, but its persistent activation disrupts the proteostatic equilibrium, which triggers T_ex-PSR and promotes exhaustion. Meanwhile, although often thought to operate on the same axis, mTOR inhibition by rapamycin did not show a substantial effect on preventing T cell exhaustion in our study. It may be because mTOR signalling is already suppressed in T_ex cells^47,52,54. The plasticity of these signalling pathways suggests that the crosstalk of these signalling hubs in T cell exhaustion warrants further investigation.

Another intriguing aspect of T_ex-PSR activation in T_ex cells is the selective upregulation of T_ex-PSR chaperones. The role of T_ex-PSR chaperones presents a conundrum here in T_ex cells. Previous studies have reported that chaperones extend their impact beyond protein folding^55,56,57,58. Moreover, AKT is a known client of HSP90 (ref. ⁵⁹). Furthermore, ER chaperones such as BiP can translocate into the nucleus and function as TFs⁶⁰. It is possible that the actions of T_ex-PSR chaperones in T_ex cells go beyond the fundamental role of protein folding and instead mediate signal transduction. An alternative and more simplistic explanation is that the chaperone machinery in T_ex cells is qualitatively suboptimal owing to the upregulation of some but not all chaperones. In T_ex cells, the T_ex-PSR chaperone stoichiometry is in disarray because of chaperone imbalance as well as substrate accumulation, which all result in pathological proteotoxic stress.

Under chronic stimulation, T cells must navigate a delicate balance between effector function and self-survival. We demonstrated that modulation of T_ex-PSR can enhance effector cytokine production at the expense of survival. Meanwhile, maintaining a high rate of protein synthesis might be advantageous for T_ex cells. This strategy ensures the production of essential proteins for their survival, albeit in a manner that is not cost-effective. Our study molecularly characterized the aggregation proteome. We demonstrated that protein aggregation in T_ex cells is a global event without selectivity, a result that highlights that it is the protein quality control machinery itself that is defective in T_ex cells. Our findings indicate that increased protein expression of T cell effector molecules per se without correcting the pathological T_ex-PSR in T_ex cells will not lead to functional improvement or reversal of the exhaustion phenotype. Moreover, we showed that the introduction of misfolded proteins alone, even in the absence of persistent TCR stimulation, effectively induced a T cell exhaustion phenotype. Thus, the fate of T_ex cells is intricately linked to protein quality control. How T cells sense aggregates and subsequently reprogram T_eff cells to T_ex cells remains an open question that warrants further exploration.

Methods

Cell lines

The MC38 cell line was purchased from Kerafast (ENH204-FP). The MB49 cell line was purchased from Sigma-Aldrich (SCC148). The HEK293T cell line was purchased from the American Type Culture Collection (CRL-3216). The MB49-gp33 cell line was shared by W. Cui (Northwestern University). The B16-OVA cell line was generated as previously described⁶⁴ and shared by L. Deng (Memorial Sloan Kettering Cancer Center). HEK293T, MC38 and MB49 cells were cultured in Dulbecco’s modified Eagle medium (DMEM; Gibco, 11965-092) with 10% FBS (Gibco, 10082-147) and 1% penicillin–streptomycin (Gibco, 15140-122) at 37 °C and 5% CO₂. B16-OVA cells were cultured in RPMI-1640 (Gibco, 11875-093) with 10% FBS and 1% penicillin–streptomycin. Cell lines were regularly tested for mycoplasma contamination.

Mice

WT C57BL/6J mice (strain 000664) were purchased from The Jackson Laboratory. CD8-specific gp96-deficient mice were generated by crossing E8i-Cre mice (The Jackson Laboratory, strain 008766) and Hsp90b1^flox/flox mice, previously generated and described by our group⁶⁵. The P14 mouse strain was a gift from W. Cui (Northwestern University). OT-1 (strain 003831) and Rag2^–/– (strain 033526) mice were purchased from The Jackson Laboratory. These mice were maintained in the animal facility at the Ohio State University under standard conditions (ambient temperature of 20–24 °C, relative humidity of 30–70% and a 12-h dark–light cycle (lights on from 6:00 to 18:00)). Mice aged 6–8 weeks were used for experiments. All procedures were performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health (NIH). The protocol was approved by the Committee on the Ethics of Animal Experiments of the Ohio State University.

T cell isolation, stimulation and drug treatment

Spleens were isolated from C57BL/6J mice and minced into single-cell suspensions. CD8⁺ T cells were isolated using an immunomagnetic negative selection kit (Stemcell, 19853). Isolated CD8⁺ T cells were first stimulated with 3 μg ml^–1 plate-bound anti-CD3 (BioLegend, 100359) and 1 μg ml^–1 anti-CD28 (BioLegend, 102121) antibodies in T cell medium made with RPMI-1640 with 10% FBS, 1% penicillin–streptomycin, 1 mM sodium pyruvate (Gibco, 11360-070), 1× MEM NEAA (Gibco, 11140-050), 10 mM HEPES (Gibco, 15630-080) and 50 μM 2-mercaptoethanol (Gibco, 21985-023) supplemented with 100 U ml^–1 recombinant human IL-2 (acquired from the Biological Resources Branch at the NIH) in 12-well plates at a density of 10⁶ cells per well for 48 h at 37 °C and 5% CO₂. For chronic stimulation, CD8⁺ T cells were re-stimulated every 2 days by passaging to new plates with plate-bound anti-CD3 in T cell medium with IL-2. For acute stimulation, CD8⁺ T cells were passaged every 2 days and maintained in T cell medium with IL-2. In some experiments, cells were treated with MK2206 (Cayman, 11593), LY294002 (Sigma-Aldrich, 440202) or rapamycin (Sigma-Aldrich, 553210) 2 days after initial activation and replenished concurrently with cell passage.

To measure cytokine production, activated cells were collected, plated and re-stimulated with 0.5× cell stimulation cocktail (Thermo Fisher, 00-4970-93) in T cell medium for 3 h at 37 °C and 5% CO₂.

Tumour challenge and TIL isolation

For the MC38 tumour model, 1 × 10⁶ cells were subcutaneously injected into the right flank of shaved C57BL/6J mice. Mice were euthanized for tumour collection 16 days after tumour implantation for cell sorting. For the MB49 tumour model, 5 × 10⁵ cells were subcutaneously injected into the right flank of shaved C57BL/6J mice. Tumours were collected 13 days after tumour implantation. To prepare single-cell suspensions, isolated tumours were chopped and washed with PBS before incubation with collagenase I (200 U ml^–1, Worthington, LS004196) in serum-free RPMI-1640 for 30 min at 37 °C with gentle agitation. After digestion, 2% BSA in PBS was added to cell suspensions to neutralize collagenase. Cell suspensions were washed with PBS and filtered through a 70 μm nylon filter. Single-cell suspensions were centrifuged and resuspended in PBS for downstream assays. For cell sorting, immune cells were enriched using a mouse TIL CD45 positive selection kit (Stemcell, 100-0350).

Flow cytometry

Cells were washed with PBS twice. Dead cells were stained using Live/Dead fixable blue (Invitrogen, L23105) or Zombie UV (BioLegend, 423108) at 4 °C for 15 min. Cells were washed with FACS buffer twice and a surface molecule staining antibody cocktail was applied for 30 min at 4 °C. After incubation, cells were washed twice with FACS buffer and then fixed and permeabilized using a FOXP3 fixation and permeabilization kit (eBioscience, 00-5523-00) overnight. After overnight fixation, cells were washed twice in permeabilization buffer and an intracellular staining antibody cocktail was added to the cells. After 2 h of incubation at room temperature, cells were washed twice with FACS buffer and analysed using Cytek Aurora. Acquired data were analysed with FlowJo software (v.10.10, BD Life Sciences) or OMIQ (Dotmatics) for high dimensional analysis. The gating strategy for TIL analysis is provided in Supplementary Fig. 2. A list of antibodies used for the multispectral flow cytometry study is provided in Supplementary Table 1.

For protein aggregation staining, cells were washed with HBSS (Sigma-Aldrich, H6648) twice and stained with 100 nM NIAD-4 (Cayman, 18520) or 50 μM CRANAD-2 (Cayman, 19814) in HBSS for 30 min at 37 °C and 5% CO₂. Cells were stained using Live/Dead fixable Near IR (Invitrogen, L34975) at 4 °C for 15 min, followed by fixation (BD Biosciences, 554655) for 15 min and DAPI staining for 5 min at room temperature. Cells were then analysed by ImageStream for acquiring fluorescent images or Cytek Aurora for quantification.

For SG analysis, cells were collected and stained using Live/Dead fixable NIR, followed by fixation in BD Cytofix fixation buffer (BD Biosciences, 554655) for 15 min and permeabilization using a FOXP3 fixation and permeabilization kit for 30 min at room temperature. Cells were then stained with anti-G3BP1 antibody (Proteintech, 13057-2-AP) in permeabilization buffer for 1 h at room temperature and then FITC-conjugated anti-rabbit antibody for 30 min. DAPI was added to the cell suspension and incubated for 5 min. Data were collected by ImageStream and analysed using IDEAS (v.6.2). Live cells were gated for SG analysis. Cells with SG loci were determined by gating on the Bight Detail Intensity feature high population on the FITC–G3BP1 channel.

Protein synthesis rate measurement

Nascent proteins were labelled using a Click-iT HPG Alexa Fluor 488 Protein Synthesis Assay kit (Thermo Fisher, C10428). Cells were incubated with 50 μM HPG (Thermo Fisher, C10186) in T cell medium made with methionine-free RPMI (Gibco, A14517-01) for 30 min at 37 °C and 5% CO₂. Cycloheximide (Sigma-Aldrich, 239763) was added to the negative control group at 50 μg ml^–1 to inhibit translation. In some experiments, 2.5 μM MG132 (Sigma-Aldrich, M7449-1ML) or 10 nM bafilomycin A1 (Sigma-Aldrich, SML1661) was added to cells after HPG incubation. Cells were then labelled following the manufacturer’s protocol and analysed using Cytek Aurora.

For measuring translation in TIL subsets in vivo, 50 mg kg^–1 OPP (Vector Laboratories, CCT-1407-25) was administered into tumour-bearing mice by intraperitoneal injection. Mice were killed exactly 1 h after injection. Tumours were isolated and processed into single-cell suspensions. Cells were stained with surface markers and OPP was labelled using a Click-iT reaction kit following the manufacturer’s protocol (Thermo Fisher, C10457).

Cell sorting

Single-cell suspensions were stained using Live/Dead fixable blue (Invitrogen, L23105) at 4 °C for 15 min. Cells were then washed twice with FACS after viability dye staining. Tumour cells were enriched for CD45⁺ lymphocytes using a mouse TIL positive selection kit (Stemcell, 100-0350) and spleen samples from mice infected with LCMV were enriched for CD8⁺ T cells with a negative selection kit (Stemcell, 19853) before viability staining. Cells were then incubated with a surface staining antibody cocktail for 30 min at 4 °C. Cells were washed twice with FACS buffer and filtered through a 70 μm nylon filter immediately before loading into a Cytek Aurora CS for sorting. For sorting, a 100 μm nozzle was used for tumour-derived samples and a 70 μm nozzle for spleen-derived samples.

LCMV infection model

For acute LCMV infection, 8–10-week-old male mice were intraperitoneally inoculated with 2 × 10⁵ p.f.u. LCMV Armstrong. For chronic LCMV infection, 8–10-week-old male mice were intravenously inoculated with 2 × 10⁶ p.f.u. LCMV clone 13 in 400 µl RPMI-1640. Mice were euthanized on day 8 and day 30 after infection.

Gene editing in T cells by CRISPR–Cas9

The sgRNAs targeting each candidate were designed and purchased from IDT. The sequences of sgRNAs are provided in Supplementary Table 2. Two days before electroporation, splenic CD8⁺ T cells were isolated and activated with 3 μg ml^–1 plate-bound anti-CD3 and 1 μg ml^–1 anti-CD28 antibodies in T cell medium supplemented with 100 U ml^–1 IL-2. On the day of electroporation, RNPs were assembled by mixing 1.5 μl sgRNA and 1 μg Cas9 nuclease V3 (IDT, 1081059) and incubated at room temperature for 20 min. Electroporation was prepared using a P4 Primary Cell 4D-Nucleofector kit (Lonza, V4XP-4032). The activated T cells were washed with PBS twice and resuspended with P4 nucleofector solution with supplement provided by the kit. RNPs and 1 μl HDR Enhancer (IDT, 10007921) were added to the cell suspensions. The reaction mix was loaded into a Nucleocuvette after incubation at room temperature for 2 min. 4D-Nucleofector and program CMT137 were used for electroporation. Cells were rested in T cell medium with 50 U ml^–1 IL-2 for 2 days and received re-stimulation every 2 days afterwards. At 8 days after electroporation, cells were collected for downstream analyses.

Protein electrophoresis and western blotting

Cells were pelleted and lysed in NP-40 buffer (50 mM Tris 7.4, 150 mM NaCl, 1% NP-40 and 0.1% sodium deoxycholate) supplemented with protease and phosphatase inhibitor cocktail (Thermo Fisher, 78440) and incubated on a roller for 30 min at 4 °C. Samples were centrifuged at 18,000g, 4 °C for 15 min and supernatant was transferred to fresh tubes as the detergent-soluble fraction. The detergent-insoluble fraction was resuspended in NP-40 buffer supplemented with 4% SDS. The protein concentration was quantified using a BCA assay (Pierce, 23227).

Native samples were diluted with native sample buffer (Thermo Fisher, NP) and run on 3–8% Tris-acetate gels (Thermo Fisher, EA0378) with Tris-glycine native running buffer (Thermo Fisher, LC2672). Samples were electrophoresed at 150 V for 3 h at 4 °C. SDS–PAGE samples were boiled in NuPAGE LDS sample buffer (Thermo Fisher, NP0007) and resolved on 4–12% Bis-Tris gels (Thermo Fisher, NP0335) with MOPS SDS running buffer (Thermo Fisher, NP0001). Samples were electrophoresed at 150 V for 1 h at room temperature. A list of antibodies used for western blot analyses is provided in Supplementary Table 1.

Retrovirus packaging and T cell transduction

The retroviral EV plasmid pMIG and pMIG-myrAKT were purchased from Addgene (52107, 65063). The open-reading frame for CFTR^ΔF508 was synthesized and cloned into the pMIG plasmid for this study. To generate retrovirus for mouse T cell transduction, HEK293T cells were transfected with pMIG and pCL-Eco in Opti-MEM. The cell culture supernatant was collected 48 h after transfection and concentrated overnight with Retro-X Concentrator (Takara, 631456). Concentrated retrovirus was added onto plates coated with RetroNectin (Takara, T100B) and spun at 1,800g at 32 °C for 2 h. Virus supernatant was removed after centrifugation and washed with PBS twice. Polyclonal, P14 cells and OT-1 CD8⁺ T cells that have been activated for 16–48 h were added to the virus-coated plate and cultured for 24 h. Cells were washed twice and plated into new plates for another 3–6 days for downstream analyses. For the generation of retrovirus for human T cell transduction, a similar approach to that used for mouse cells was used, with the key modification of using the Plat-A cell line for virus packaging. To transduce human CD8⁺ T cells, CD8⁺ T cells were magnetically isolated from peripheral blood mononuclear cells (Stemcell, 17953) and activated with Dynabeads (Gibco, 11131D) for 1 day. After activation, the cells were transduced with the indicated virus. In brief, the cells were spinoculated at 1,000g in a RetroNectin-virus-coated plate. After 24 h, the virus was removed, and subsequent analyses were performed after an additional 6–8 days of activation and maintenance.

ACT experiment

P14 cells were isolated from the spleens of P14 mice and activated with 1 μg ml^–1 gp33 peptide. Two days after activation, cells were edited by CRISPR–Cas9 as described above and expanded for another 2 days with 100 U ml^–1 IL-2. Next, 1 × 10⁶ P14 cells were intravenously transferred per mouse. Then 5 × 10⁵ MB49-gp33 cells were subcutaneously injected into the right flank of shaved WT C57BL/6J mice or Rag2^–/– mice. WT mice were lymphodepleted using 5 Gray of total body irradiation on the day before cell transfer and randomized for treatment groups. OT-1 cells were activated and transduced with retroviral vector as described above. Transduced OT-1 cells were purified by cell sorting on the basis of positive GFP expression. In total, 2.5 × 10⁵ OT-1 cells were intravenously transferred to B16-OVA tumour-bearing Rag2^–/– mice. For OT-1 ACT experiments, 5 × 10⁵ cells B16-OVA cells were subcutaneously injected into the right flank of Rag2^–/– mice 8 days before adoptive transfer and randomized into treatment groups.

Immunofluorescence analysis by confocal microscopy

T cells were collected and spun onto glass coverslips in a 12-well plate. For protein aggregation staining, cells were stained with NIAD-4 and fixed as described above. For CFTR staining, cells were fixed with fixation buffer (BD, 554655) for 15 min, permeabilized with 0.5% Triton X-100 in PBS for 20 min and blocked with 2% BSA for 1 h. Cells were stained with primary anti-CFTR antibody (Proteintech, 20738-1-AP) and then Alexa Fluor 647-conjugated goat anti-rabbit IgG antibody (Thermo Fisher, A-21244). After staining, coverslips were mounted onto glass slides with mountant and DAPI (Thermo Fisher, P36962). Images were taken using an Olympus FV3000 microscope with ×60 magnification and processed with Olympus OlyVIA (v.4.2). For analysis, images were imported into ImageJ as .tiff files and adjusted to RGB stack format for downstream processing. Thresholds for positive detection of aggregates were determined through normalized autodetection and maintained across all images with a lower threshold of 100 and an upper threshold of 255 to generate binary image masks. The area, average size per particle, percentage of area and mean fluorescence intensity were analysed using the Analyze Particles function selected for area, area fraction, fluorescence intensity, particle count and average particle size.

MS sample processing

Cell samples were collected and washed with PBS once. Cell pellets were frozen at −80 °C if not immediately processed. Cells were lysed in lysis buffer made with 5% SDS (Thermo Fisher, AM9820), 50 mM TEAB (Thermo Fisher, 90114) and 2 mM MgCl₂ (Thermo Fisher, AM9530G) with HALT protease inhibitor cocktail (Thermo Fisher, 78441). Lysates were homogenized using either a probe sonicator or a Biorupter. DNA was removed by centrifugation at 13,000g for 10 min and the pellet discarded. For in vitro cell samples, the protein concentration was quantified using a BCA assay (Pierce, 23227) and 50 μg protein of each sample was used for subsequent steps. For in vivo samples, total lysates were used assuming accurate FACS cell counts. Cell lysates were then treated with 20 mM DTT (Sigma-Aldrich, 10197777001) at 95 °C for 10 min, followed by the addition of 40 mM iodoacetamide (Pierce, A39271) at room temperature for 30 min in the dark and then quenched with 20 mM DTT for 15 min at room temperature. Phosphoric acid (1.2%; Sigma-Aldrich, 345245) was used to acidify proteins. Binding buffer with 100 mM TEAB in methanol (Thermo Fisher, A4581) was added to samples that were then loaded onto S-traps (ProtiFi, C01-micro-80) and washed with binding buffer 3 times. Proteins were digested with trypsin (Pierce, 90058) at 47 °C for 2 h. Digested peptides were eluted from S-traps with 0.2% formic acid (Thermo Fisher TS-28905) followed by a second elution with 50% acetonitrile (Sigma-Aldrich, T7408) in 0.2% formic acid. Eluates were pooled and lyophilized for storage at −80 °C.

MS acquisition

Peptides were reconstituted with 2% acetonitrile in 0.1% formic acid and separated using either an Easy-nLC 1200 coupled to an Thermo Exploris 480 tandem mass spectrometer (Thermo Fisher) or an UltiMate 3000 UHPLC coupled to a Thermo Fusion tandem mass spectrometer (Thermo Fisher). In both set ups, peptides were first desalted online using an Acclaim PepMap 100 Trap column (75 μm inner diameter, 150 mm length, 3 μm C18 packing) and then separated and ionized using either a 50 cm (Easy-nLC) or 25 cm (Ultimate 3000) Easy-Spray HPLC column (75 μm inner diameter, 2 μm C18 packing) with a 90-min linear gradient.

All data-independent acquisition (DIA) measurements were configured in a staggered window pattern using boundaries optimized to place window boundaries in forbidden zones. The Thermo Fusion was configured to use two DIA injections (covering peptide precursors from 400 to 700 m/z and from 700 to 1,000 m/z) of 38 ×8 m/z-wide windows in a staggered window pattern. These windows were configured to have 17,500 resolution and an automatic gain control (AGC) target of 4 × 10⁵. Precursor spectra were placed every 38 scans (1 per cycle) using 35,000 resolution and an AGC target of 4 × 10⁵. Similarly, the Thermo Exploris 480 was configured to use single-injection DIA measurements (covering peptide precursors from 400 to 1,000 m/z) of 38 × 16 m/z-width windows. These windows were configured to have 30,000 resolution and an AGC target of 1 × 10⁶. Precursor spectra were placed every 38 scans (1 per cycle) using 60,000 resolution and an AGC target of 1 × 10⁶.

For each dataset, a sample pool was made from subaliquots and used for library generation. We used gas-phase fractionation (GPF) DIA following the chromatogram library approach^66,67. For this, we injected each peptide pool 6 times using different 100 m/z regions (400–500 m/z, 500–600 m/z, 600–700 m/z, 700–800 m/z, 800–900 m/z and 900–1,000 m/z). Each injection was configured to use 4 m/z staggered DIA windows and appropriate precursor windows. Otherwise, all measurements were performed as for normal DIA above on their respective instrument.

Proteomic data analysis

Raw files were demultiplexed using MSConvert in the Proteowizard package (v.3.0.20169)⁶⁸ and then searched using EncyclopeDIA (v.2.12.31). EncyclopeDIA was configured with the default settings for Orbitraps: 10 ppm precursor, fragment and library tolerances. EncyclopeDIA was allowed to consider both B and Y ions, and trypsin digestion was assumed. Searches were performed using a two-step procedure. First, the GPF-DIA injections were searched using a Prosit^69,70 predicted spectrum library to generate a chromatogram library based on the Mus musculus UniProt FASTA database (downloaded on 22 October 2019, containing 17,025 entries). All z = +2 or z = +3 peptides from 396.4 to 1002.7 m/z (with a maximum of one missed cleavage) were predicted assuming a normalized collision energy of 33. Peptides detected in the six GPF-DIA injections at a 1% peptide-level false discovery rate (FDR) were compiled into the chromatogram library. Quantitative DIA injections were searched against this chromatogram library, again filtered to a 1% peptide-level FDR. A normalized protein expression matrix for all proteomics generated in this study is provided in Supplementary Table 3. Bubble plots of protein expression were generated using the R package tidyverse (v.1.3.1)⁷¹ based on z score-normalized protein expression values. Gene set enrichment analysis for protein clusters was performed using Enrichr^72,73,74.

Bulk RNA-seq sample preparation and data analysis

Acutely and chronically stimulated T cells were collected on day 8 after initial activation. Cells were washed with PBS twice and pelleted. RNA was first extracted using TRIzol and chloroform and then cleaned up using a RNeasy Micro kit (Qiagen, 74004). Sample library preparation and sequencing were performed by Azenta Life Sciences. Poly(A) selection was used for library preparation. Sequencing was performed using an Illumina NovaSeq platform with a depth of 50 million reads per sample. The raw bulk sequences were checked, trimmed and filtered using Fastp (v.0.23.4)⁷⁵. The filtered reads were mapped to the mouse reference genome mm10 using HISAT2 (v.2.2.1)⁷⁶, and samtools (v.1.17)⁷⁷ was used to convert and sort BAM files. Last, the subread tool (v.2.0.6)⁷⁸ was used for gene quantification and generating the raw expression matrix. Raw expression data were first log-normalized, and the R package Limma (v.3.56.2)⁷⁹ was used to fit the model and perform differential expression analysis. To avoid NA values, a pseudo count of 1 was added to the raw count matrix. Genes with an absolute log[fold change] value greater than 1.5 and FDR-adjusted P value smaller than 0.05 are considered as differentially expressed genes.

Statistical comparison of protein expression and gene expression

To accurately compare protein and gene expression levels, we created a hash table (Supplementary Table 4) that included the protein accession number, protein name, gene name and Mouse Genome Informatics (MGI) number. Each protein and RNA matrix needed to match the hash table, and only the overlapped proteins and genes were kept.

We compared the normalized and log-transformed protein expression and gene expression levels in samples of the sample condition (for example, day 8 T_ex samples). Only proteins and genes that overlapped in both protein and RNA data were retained for comparison. A Pearson’s correlation test was applied to calculate the correlation coefficient between protein expression and gene expression levels. We also compared the log[fold change] of proteins and genes between different conditions. The log[fold change] of proteins and genes were calculated in the analysis of differentially expressed genes described above.

We generated a functional gene list to further evaluate the expression level of proteins and genes undergoing specific cell functions, including 13 gene ontology terms, one EIF2A-dependent and one EIF2A-independent gene list. Specifically, the EIF2A-dependent and EIF2A-independent genes were determined according to the EIF2A-regulated upstream open reading frames³⁵. As previously described³⁵, EIF2A-regulated upstream open reading frames were defined as the ratio of 5′ untranslated region (UTR) translation in control/5′ UTR translation in Eif2a KO > 4. The remaining mRNAs with a ratio <4 were defined as non-EIF2A regulated (EIF2A-independent). The 5′ UTR translation rate was quantified for mRNAs with an average of more than 16 reads over all replicates. Genes in each of the 26 lists are highlighted on the scatter plot to compare the protein and gene expression/log[fold change].

Gene signature score analysis

For each of the gene lists mentioned above, we also calculated a gene signature score based on the single-sample gene set enrichment analysis (ssGSEA) method. An in-house script was used to perform the ssGSEA analysis. The R package heatmaply (v.1.4.2)⁸⁰ or Morpheus (https://software.broadinstitute.org/morpheus) was used to draw the heatmap. For gene signature score analysis for scRNA-seq data, the raw expression matrix of LCMV scRNA-seq data was downloaded from GSM3701181 (ref. ³¹). Cells were divided into three categories on the basis of gene expression levels: progenitor state (Slamf6 > 0 and Cx3cr1 = 0); intermediate state (Cx3cr1 > 0); and terminal state (Slamf6 = 0 and Cx3cr1 = 0). Cells in each category were randomly divided into three equal subgroups. Pseudo bulk gene expression was defined by the average expression of genes in each cell subgroup. Then, the same ssGSEA method was performed on the pseudo bulk expression data to calculate the gene signature scores and to generate the heatmap.

Pan-cancer scRNA-seq data collection

To construct a comprehensive pan-cancer scRNA-seq dataset, we compiled transcriptomic profiles from 346 tumour samples derived from 251 individuals across 20 publicly available scRNA-seq datasets^{81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100} (Supplementary Table 5). To ensure data consistency and to minimize platform-related biases, only datasets generated using the 10x Genomics droplet-based platform were included for our analyses.

Quality control and preprocessing of the pan-cancer scRNA-seq data

We applied rigorous quality control measures using the package Scanpy (v.1.9.5)¹⁰¹ to filter and preprocess single-cell transcriptomic data. The following inclusion criteria were applied: (1) each cell expressed at least 200 genes; and (2) mitochondrial gene content remained below 20% of total counts. Further filtering steps removed the following data: (1) low-quality barcodes indicative of debris (<400 detected genes, <500 unique molecular identifiers); and (2) potential duplicate cells (>5,500 detected genes or >30,000 unique molecular identifiers). After quality control, raw count matrices and AnnData objects were concatenated, and counts were normalized to transcripts per million using sc.pp.normalize_total, followed by log-transformation with sc.pp.log1p. Non-tumour cells were excluded before normalization, which produced 1,030,968 high-quality single cells and 14,090 genes for downstream analyses.

Batch correction and data integration

To harmonize datasets across studies while preserving biological signals, we used the Python package scVI (scvi-tools v.1.0.4)¹⁰² for batch-effect correction and data integration. The scVI model was trained with sample identity as a covariate, mitigating inter-sample technical variability while ensuring robust integration of multiple datasets. The efficiency of batch correction was assessed by quantifying the reduction in batch-specific effects while maintaining key biological variance. After correction, downstream analyses—including clustering, differential gene expression and trajectory inference—were performed on the integrated dataset. UMAP was used for visualization, depicting cellular heterogeneity across batches, datasets, sex, organ origins and cancer types.

Cell-type annotation of pan-cancer scRNA-seq data

To annotate cell populations, we leveraged the scANVI algorithm (scVI-tools v.1.0.4), which provided pre-labelled reference annotations for epithelial, endothelial, fibroblast, lymphoid, myeloid and plasma cells. Initial clustering was performed in the scANVI latent space, followed by Leiden clustering to assign cell identities. The scANVI model was trained with max_epochs=20, and cluster annotations were transferred with n_samples_per_label=100. For detailed characterization of T cell subpopulations, we further integrated corresponding AnnData objects and applied scVI-based batch correction.

Functional signature calculation for scRNA-seq data

We used the scanpy.tl.score_genes function from the Python package Scanpy (v.1.9.5) to compute gene set scores across individual cells, which enabled the quantification of functional signatures in the scRNA-seq dataset.

RNA velocity and trajectory inference

RNA velocity analysis was performed to infer the directionality of cellular state transitions using spliced and unspliced transcript counts. Velocities were computed using the scVelo toolkit (v.0.3.3)^103,104, which estimates transcriptional dynamics across single cells. The resulting velocity vectors were projected onto the UMAP embedding to visualize the flow of differentiation. To infer developmental trajectories, the Slingshot algorithm was applied to the UMAP coordinates, incorporating RNA velocity information to identify lineage structures. Slingshot fit smooth curves (principal curves) through the data and assigned pseudotime values along each inferred lineage. Two dominant lineages were identified: one progressing towards a T_ex cell phenotype (lineage 1) and the other towards an effector-like phenotype (lineage 2). Signature scores for naive, exhaustion and T_ex-PSR gene modules were calculated across pseudotime for each lineage using averaged normalized expression of predefined marker genes.

Validation of the T_ex-PSR signature in CD8⁺ T cells and its prognostic impact

To assess the clinical significance of the T_ex-PSR signature in CD8⁺ T cells, we analysed public processed scRNA-seq data from 116 liver cancer samples obtained from 94 male patients¹⁰⁵. Survival analyses were restricted to primary tumours and metastatic samples. After quality filtering, batch correction and cell-type annotation using the established preprocessing pipeline, CD8⁺ T cells were isolated and T_ex-PSR signature scores were computed using the scanpy.tl.score_genes function from the Scanpy package (v.1.9.5).

T_ex-PSR signature expression in CD8⁺ T cells and its impact on patient survival

To evaluate the prognostic significance of T_ex-PSR expression levels in CD8⁺ T cells, we performed survival analyses using Kaplan–Meier curves, with statistical comparisons conducted using the log-rank test and univariate Cox proportional hazards (Cox PH) models, as specified in each figure. Two additional multivariable Cox PH models were fitted to account for potential confounders. The hazard ratio and 95% confidence intervals were reported on the basis of these models. Kaplan–Meier survival curves were generated to compare high versus low T_ex-PSR expression in liver cancer scRNA-seq datasets, with P values computed using univariate Cox PH models. To determine the optimal cut-off value for T_ex-PSR signature expression in relation to survival outcomes, we used the surv_cutpoint function from the R package survminer. This approach uses maximally selected rank statistics from the R package maxstat¹⁰⁶ to stratify patients into low-risk and high-risk groups. Moreover, continuous variables included in the Cox PH¹⁰⁷ models were assessed for linearity to ensure model validity.

T_ex-PSR expression in immunotherapy-treated patients

We further investigated T_ex-PSR expression in responders and non-responders across independent scRNA-seq datasets from patients receiving diverse immunotherapy treatments, including CAR T cell therapy for refractory B cell lymphoma⁶¹, anti-PD1 therapy for lung cancer and advanced renal cell RCC^62,63, and anti-CTLA-4 with anti-PD1 combination therapy for RCC^64,108. For each dataset, we applied the same preprocessing pipeline, including quality filtering, batch correction and cell-type annotation, as described for the pan-cancer scRNA-seq dataset.

Statistical analysis

Statistical analyses were performed using GraphPad Prism (v.10). Two-tailed unpaired Student’s t-test was used for comparison between two groups. One-way ANOVA was used for comparisons among three or more groups. Two-way ANOVA was used to compare curves of time-course studies, including cell and tumour growth curves. P < 0.05 was considered significant.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

Proteomic data are available from the MassIVE repository (MSV000098609) and through ProteomeXchange (PXD066433). Normalized protein expression data are provided in Supplementary Table 3. Bulk RNA-seq data from this paper have been deposited into the NCBI Gene Expression Omnibus database with the identifier GSE303401. Source data are provided with this paper.

Code availability

Scripts generated for analysis are available from Zenodo (https://doi.org/10.5281/zenodo.16323779)¹⁰⁹.

References

Zajac, A. J. et al. Viral immune evasion due to persistence of activated T cells without effector function. J. Exp. Med. 188, 2205–2213 (1998).
Article PubMed PubMed Central CAS Google Scholar
Chow, A., Perica, K., Klebanoff, C. A. & Wolchok, J. D. Clinical implications of T cell exhaustion for cancer immunotherapy. Nat. Rev. Clin. Oncol. 19, 775–790 (2022).
Article PubMed PubMed Central Google Scholar
Wherry, E. J. T cell exhaustion. Nat. Immunol. 12, 492–499 (2011).
Article PubMed CAS Google Scholar
Blank, C. U. et al. Defining ‘T cell exhaustion’. Nat. Rev. Immunol. 19, 665–674 (2019).
Article PubMed PubMed Central CAS Google Scholar
Hetz, C., Zhang, K. & Kaufman, R. J. Mechanisms, regulation and functions of the unfolded protein response. Nat. Rev. Mol. Cell Biol. 21, 421–438 (2020).
Article PubMed PubMed Central CAS Google Scholar
Costa-Mattioli, M. & Walter, P. The integrated stress response: from mechanism to disease. Science https://doi.org/10.1126/science.aat5314 (2020).
McLane, L. M., Abdel-Hakeem, M. S. & Wherry, E. J. CD8 T cell exhaustion during chronic viral infection and cancer. Annu. Rev. Immunol. 37, 457–495 (2019).
Article PubMed CAS Google Scholar
Hudson, W. H. et al. Proliferating transitory T cells with an effector-like transcriptional signature emerge from PD-1⁺ stem-like CD8⁺ T cells during chronic infection. Immunity 51, 1043–1058 (2019).
Article PubMed PubMed Central CAS Google Scholar
Siddiqui, I. et al. Intratumoral Tcf1⁺PD-1⁺CD8⁺ T cells with stem-like properties promote tumor control in response to vaccination and checkpoint blockade immunotherapy. Immunity 50, 195–211 (2019).
Article PubMed CAS Google Scholar
Im, S. J. et al. Defining CD8⁺ T cells that provide the proliferative burst after PD-1 therapy. Nature 537, 417–421 (2016).
Article ADS PubMed PubMed Central CAS Google Scholar
Miller, B. C. et al. Subsets of exhausted CD8⁺ T cells differentially mediate tumor control and respond to checkpoint blockade. Nat. Immunol. 20, 326–336 (2019).
Article PubMed PubMed Central CAS Google Scholar
Kurtulus, S. et al. Checkpoint blockade immunotherapy induces dynamic changes in PD-1^–CD8⁺ tumor-infiltrating T cells. Immunity 50, 181–194 (2019).
Article PubMed PubMed Central CAS Google Scholar
Schietinger, A. et al. Tumor-specific T cell dysfunction is a dynamic antigen-driven differentiation program initiated early during tumorigenesis. Immunity 45, 389–401 (2016).
Article PubMed PubMed Central CAS Google Scholar
Lynn, R. C. et al. c-Jun overexpression in CAR T cells induces exhaustion resistance. Nature 576, 293–300 (2019).
Article ADS PubMed PubMed Central CAS Google Scholar
Chen, J. et al. NR4A transcription factors limit CAR T cell function in solid tumours. Nature 567, 530–534 (2019).
Article ADS PubMed PubMed Central CAS Google Scholar
Seo, H. et al. TOX and TOX2 transcription factors cooperate with NR4A transcription factors to impose CD8⁺ T cell exhaustion. Proc. Natl Acad. Sci. USA 116, 12410–12415 (2019).
Article ADS PubMed PubMed Central CAS Google Scholar
Long, A. H. et al. 4-1BB costimulation ameliorates T cell exhaustion induced by tonic signaling of chimeric antigen receptors. Nat. Med. 21, 581–590 (2015).
Article PubMed PubMed Central CAS Google Scholar
Vogel, C. & Marcotte, E. M. Insights into the regulation of protein abundance from proteomic and transcriptomic analyses. Nat. Rev. Genet. 13, 227–232 (2012).
Article PubMed PubMed Central CAS Google Scholar
Schwanhausser, B. et al. Global quantification of mammalian gene expression control. Nature 473, 337–342 (2011).
Article ADS PubMed Google Scholar
Maier, T., Guell, M. & Serrano, L. Correlation of mRNA and protein in complex biological samples. FEBS Lett. 583, 3966–3973 (2009).
Article PubMed CAS Google Scholar
Jiang, L. et al. A quantitative proteome map of the human body. Cell 183, 269–283 (2020).
Article PubMed PubMed Central CAS Google Scholar
Taniguchi, Y. et al. Quantifying E. coli proteome and transcriptome with single-molecule sensitivity in single cells. Science 329, 533–538 (2010).
Article ADS PubMed PubMed Central CAS Google Scholar
Hukelmann, J. L. et al. The cytotoxic T cell proteome and its shaping by the kinase mTOR. Nat. Immunol. 17, 104–112 (2016).
Article PubMed CAS Google Scholar
Cuadrado, E. et al. Proteomic analyses of human regulatory T cells reveal adaptations in signaling pathways that protect cellular identity. Immunity 48, 1046–1059 (2018).
Article PubMed CAS Google Scholar
Nicolet, B. P. & Wolkers, M. C. The relationship of mRNA with protein expression in CD8⁺ T cells associates with gene class and gene characteristics. PLoS ONE 17, e0276294 (2022).
Article PubMed PubMed Central CAS Google Scholar
Chang, C. H. et al. Posttranscriptional control of T cell effector function by aerobic glycolysis. Cell 153, 1239–1251 (2013).
Article PubMed PubMed Central CAS Google Scholar
Vardhana, S. A. et al. Impaired mitochondrial oxidative phosphorylation limits the self-renewal of T cells exposed to persistent antigen. Nat. Immunol. 21, 1022–1033 (2020).
Article PubMed PubMed Central CAS Google Scholar
Belk, J. A. et al. Genome-wide CRISPR screens of T cell exhaustion identify chromatin remodeling factors that limit T cell persistence. Cancer Cell 40, 768–786 (2022).
Article PubMed PubMed Central CAS Google Scholar
Nusinow, D. P. et al. Quantitative proteomics of the Cancer Cell Line Encyclopedia. Cell 180, 387–402 (2020).
Article PubMed PubMed Central CAS Google Scholar
Chu, Y. et al. Pan-cancer T cell atlas links a cellular stress response state to immunotherapy resistance. Nat. Med. 29, 1550–1562 (2023).
Article PubMed PubMed Central CAS Google Scholar
Zander, R. et al. CD4⁺ T cell help is required for the formation of a cytolytic CD8⁺ T cell subset that protects against chronic infection and cancer. Immunity 51, 1028–1042 (2019).
Article PubMed PubMed Central CAS Google Scholar
Beusch, C. M. et al. Longitudinal proteomic profiling of T cell differentiation in vivo unveils dynamic proteome remodeling. Preprint at bioRxiv https://doi.org/10.1101/2024.05.14.593504 (2024).
Nesterov, E. E. et al. In vivo optical imaging of amyloid aggregates in brain: design of fluorescent markers. Angew. Chem. Int. Ed. Engl. 44, 5452–5456 (2005).
Article PubMed CAS Google Scholar
Gebauer, F. & Hentze, M. W. Molecular mechanisms of translational control. Nat. Rev. Mol. Cell Biol. 5, 827–835 (2004).
Article PubMed PubMed Central CAS Google Scholar
Sendoel, A. et al. Translation from unconventional 5′ start sites drives tumour initiation. Nature 541, 494–499 (2017).
Article ADS PubMed PubMed Central CAS Google Scholar
Suzuki, C. et al. PDCD4 inhibits translation initiation by binding to eIF4A using both its MA3 domains. Proc. Natl Acad. Sci. USA 105, 3274–3279 (2008).
Article ADS PubMed PubMed Central CAS Google Scholar
Marchingo, J. M. & Cantrell, D. A. Protein synthesis, degradation, and energy metabolism in T cell immunity. Cell. Mol. Immunol. 19, 303–315 (2022).
Article PubMed PubMed Central CAS Google Scholar
Signer, R. A., Magee, J. A., Salic, A. & Morrison, S. J. Haematopoietic stem cells require a highly regulated protein synthesis rate. Nature 509, 49–54 (2014).
Article ADS PubMed PubMed Central CAS Google Scholar
Franchini, D. M. et al. Microtubule-driven stress granule dynamics regulate inhibitory immune checkpoint expression in T cells. Cell Rep. 26, 94–107 (2019).
Article PubMed CAS Google Scholar
Curdy, N. et al. The proteome and transcriptome of stress granules and P bodies during human T lymphocyte activation. Cell Rep. 42, 112211 (2023).
Article PubMed CAS Google Scholar
Trotter, E. W. et al. Misfolded proteins are competent to mediate a subset of the responses to heat shock in Saccharomyces cerevisiae. J. Biol. Chem. 277, 44817–44825 (2002).
Article PubMed CAS Google Scholar
Qian, S. B. et al. mTORC1 links protein quality and quantity control by sensing chaperone availability. J. Biol. Chem. 285, 27385–27395 (2010).
Article PubMed PubMed Central CAS Google Scholar
Yoshimura, K. et al. Expression of the cystic fibrosis transmembrane conductance regulator gene in cells of non-epithelial origin. Nucleic Acids Res. 19, 5417–5423 (1991).
Article PubMed PubMed Central CAS Google Scholar
Rommens, J. M. et al. Identification of the cystic fibrosis gene: chromosome walking and jumping. Science 245, 1059–1065 (1989).
Article ADS PubMed CAS Google Scholar
Riordan, J. R. CFTR function and prospects for therapy. Annu. Rev. Biochem. 77, 701–726 (2008).
Article PubMed CAS Google Scholar
Chi, H. Regulation and function of mTOR signalling in T cell fate decisions. Nat. Rev. Immunol. 12, 325–338 (2012).
Article PubMed PubMed Central CAS Google Scholar
Huang, H., Long, L., Zhou, P., Chapman, N. M. & Chi, H. mTOR signaling at the crossroads of environmental signals and T-cell fate decisions. Immunol. Rev. 295, 15–38 (2020).
Article PubMed PubMed Central CAS Google Scholar
Pellman, D., Garber, E. A., Cross, F. R. & Hanafusa, H. An N-terminal peptide from p60src can direct myristylation and plasma membrane localization when fused to heterologous proteins. Nature 314, 374–377 (1985).
Article ADS PubMed CAS Google Scholar
Kharas, M. G. et al. Constitutively active AKT depletes hematopoietic stem cells and induces leukemia in mice. Blood 115, 1406–1415 (2010).
Article PubMed PubMed Central CAS Google Scholar
Bian, Y. et al. Cancer SLC43A2 alters T cell methionine metabolism and histone methylation. Nature 585, 277–282 (2020).
Article PubMed PubMed Central CAS Google Scholar
Mbofung, R. M. et al. HSP90 inhibition enhances cancer immunotherapy by upregulating interferon response genes. Nat. Commun. 8, 451 (2017).
Article ADS PubMed PubMed Central Google Scholar
Staron, M. M. et al. The transcription factor FoxO1 sustains expression of the inhibitory receptor PD-1 and survival of antiviral CD8⁺ T cells during chronic infection. Immunity 41, 802–814 (2014).
Article PubMed PubMed Central CAS Google Scholar
Utzschneider, D. T. et al. Active maintenance of T cell memory in acute and chronic viral infection depends on continuous expression of FOXO1. Cell Rep. 22, 3454–3467 (2018).
Article PubMed PubMed Central CAS Google Scholar
Ando, S. et al. mTOR regulates T cell exhaustion and PD-1-targeted immunotherapy response during chronic viral infection. J. Clin. Invest. https://doi.org/10.1172/JCI160025 (2023).
Liu, B. & Li, Z. Endoplasmic reticulum HSP90b1 (gp96, grp94) optimizes B-cell function via chaperoning integrin and TLR but not immunoglobulin. Blood 112, 1223–1230 (2008).
Article PubMed PubMed Central CAS Google Scholar
Xu, Y. et al. Heat shock protein gp96 drives natural killer cell maturation and anti-tumor immunity by counteracting Trim28 to stabilize Eomes. Nat. Commun. 15, 1106 (2024).
Article ADS PubMed PubMed Central CAS Google Scholar
Zhang, Y. et al. GP96 is a GARP chaperone and controls regulatory T cell functions. J. Clin. Invest. 125, 859–869 (2015).
Article ADS PubMed PubMed Central Google Scholar
Amankwah, Y. S. et al. Structural transitions modulate the chaperone activities of Grp94. Proc. Natl Acad. Sci. USA 121, e2309326121 (2024).
Article PubMed PubMed Central CAS Google Scholar
Basso, A. D. et al. Akt forms an intracellular complex with heat shock protein 90 (Hsp90) and Cdc37 and is destabilized by inhibitors of Hsp90 function. J. Biol. Chem. 277, 39858–39866 (2002).
Article PubMed CAS Google Scholar
Liu, Z. et al. ER chaperone GRP78/BiP translocates to the nucleus under stress and acts as a transcriptional regulator. Proc. Natl Acad. Sci. USA 120, e2303448120 (2023).
Article PubMed PubMed Central CAS Google Scholar
Haradhvala, N. J. et al. Distinct cellular dynamics associated with response to CAR-T therapy for refractory B cell lymphoma. Nat. Med. 28, 1848–1859 (2022).
Article PubMed PubMed Central CAS Google Scholar
Liu, B. et al. Temporal single-cell tracing reveals clonal revival and expansion of precursor exhausted T cells during anti-PD-1 therapy in lung cancer. Nat. Cancer 3, 108–121 (2022).
Article PubMed CAS Google Scholar
Bi, K. et al. Tumor and immune reprogramming during immunotherapy in advanced renal cell carcinoma. Cancer Cell 39, 649–661 (2021).
Article PubMed PubMed Central CAS Google Scholar
Budhu, S. et al. Blockade of surface-bound TGF-β on regulatory T cells abrogates suppression of effector T cell function in the tumor microenvironment. Sci. Signal. 10, eaak9702 (2017).
Article PubMed PubMed Central Google Scholar
Yang, Y. et al. Heat shock protein gp96 is a master chaperone for Toll-like receptors and is important in the innate function of macrophages. Immunity 26, 215–226 (2007).
Article PubMed PubMed Central Google Scholar
Pino, L. K., Just, S. C., MacCoss, M. J. & Searle, B. C. Acquiring and analyzing data independent acquisition proteomics experiments without spectrum libraries. Mol. Cell. Proteomics 19, 1088–1103 (2020).
Article PubMed PubMed Central Google Scholar
Searle, B. C. et al. Chromatogram libraries improve peptide detection and quantification by data independent acquisition mass spectrometry. Nat. Commun. 9, 5128 (2018).
Article ADS PubMed PubMed Central Google Scholar
Chambers, M. C. et al. A cross-platform toolkit for mass spectrometry and proteomics. Nat. Biotechnol. 30, 918–920 (2012).
Article PubMed PubMed Central CAS Google Scholar
Gessulat, S. et al. Prosit: proteome-wide prediction of peptide tandem mass spectra by deep learning. Nat. Methods 16, 509–518 (2019).
Article PubMed CAS Google Scholar
Searle, B. C. et al. Generating high quality libraries for DIA MS with empirically corrected peptide predictions. Nat. Commun. 11, 1548 (2020).
Article ADS PubMed PubMed Central CAS Google Scholar
Wickham, H. et al. Welcome to the Tidyverse. J. Open Source Softw. 4, 1686 (2019).
Article ADS Google Scholar
Chen, E. Y. et al. Enrichr: interactive and collaborative HTML5 gene list enrichment analysis tool. BMC Bioinformatics 14, 128 (2013).
Article PubMed PubMed Central Google Scholar
Kuleshov, M. V. et al. Enrichr: a comprehensive gene set enrichment analysis web server 2016 update. Nucleic Acids Res. 44, W90–W97 (2016).
Article PubMed PubMed Central CAS Google Scholar
Xie, Z. et al. Gene set knowledge discovery with Enrichr. Curr. Protoc. 1, e90 (2021).
Article PubMed PubMed Central CAS Google Scholar
Chen, S., Zhou, Y., Chen, Y. & Gu, J. fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics 34, i884–i890 (2018).
Article PubMed PubMed Central Google Scholar
Kim, D., Paggi, J. M., Park, C., Bennett, C. & Salzberg, S. L. Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype. Nat. Biotechnol. 37, 907–915 (2019).
Article PubMed PubMed Central CAS Google Scholar
Li, H. et al. The Sequence Alignment/Map format and SAMtools. Bioinformatics 25, 2078–2079 (2009).
Article PubMed PubMed Central Google Scholar
Liao, Y., Smyth, G. K. & Shi, W. The Subread aligner: fast, accurate and scalable read mapping by seed-and-vote. Nucleic Acids Res. 41, e108 (2013).
Article PubMed PubMed Central Google Scholar
Ritchie, M. E. et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43, e47 (2015).
Article PubMed PubMed Central Google Scholar
Galili, T., O’Callaghan, A., Sidi, J. & Sievert, C. heatmaply: an R package for creating interactive cluster heatmaps for online publishing. Bioinformatics 34, 1600–1602 (2018).
Article PubMed CAS Google Scholar
Li, R. Y. et al. Mapping single-cell transcriptomes in the intra-tumoral and associated territories of kidney cancer. Cancer Cell 40, 1583–1599 (2022).
Article PubMed PubMed Central CAS Google Scholar
Zhang, Y. P. et al. Single-cell analyses of renal cell cancers reveal insights into tumor microenvironment, cell of origin, and therapy response. Proc. Natl Acad. Sci. USA 118, e2103240118 (2021).
Article PubMed PubMed Central CAS Google Scholar
Werba, G. et al. Single-cell RNA sequencing reveals the effects of chemotherapy on human pancreatic adenocarcinoma and its tumor microenvironment. Nat. Commun. 14, 797 (2023).
Article ADS PubMed PubMed Central CAS Google Scholar
Ma, L. C. et al. Tumor cell biodiversity drives microenvironmental reprogramming in liver cancer. Cancer Cell 36, 418–430 (2019).
Article PubMed PubMed Central CAS Google Scholar
Regner, M. J. et al. A multi-omic single-cell landscape of human gynecologic malignancies. Mol. Cell 81, 4924–4941 (2021).
Article PubMed CAS Google Scholar
Chen, Y. P. et al. Single-cell transcriptomics reveals regulators underlying immune cell diversity and immune subtypes associated with prognosis in nasopharyngeal carcinoma. Cell Res. 30, 1024–1042 (2020).
Article PubMed PubMed Central CAS Google Scholar
Xu, J. F. et al. Single-cell RNA sequencing reveals the tissue architecture in human high-grade serous ovarian cancer. Clin. Cancer Res. 28, 3590–3602 (2022).
Article PubMed PubMed Central CAS Google Scholar
Li, C. B. et al. Single-cell transcriptomics reveals cellular heterogeneity and molecular stratification of cervical cancer. Commun. Biol. 5, 1208 (2022).
Article PubMed PubMed Central CAS Google Scholar
Luo, H. et al. Pan-cancer single-cell analysis reveals the heterogeneity and plasticity of cancer-associated fibroblasts in the tumor microenvironment. Nat. Commun. 13, 6619 (2022).
Article ADS PubMed PubMed Central CAS Google Scholar
Sathe, A. et al. Single-cell genomic characterization reveals the cellular reprogramming of the gastric tumor microenvironment. Clin. Cancer Res. 26, 2640–2653 (2020).
Article PubMed PubMed Central CAS Google Scholar
Kim, N. et al. Single-cell RNA sequencing demonstrates the molecular and cellular reprogramming of metastatic lung adenocarcinoma. Nat. Commun. 11, 2285 (2020).
Article ADS PubMed PubMed Central CAS Google Scholar
Lambrechts, D. et al. Phenotype molding of stromal cells in the lung tumor microenvironment. Nat. Med. 24, 1277–1289 (2018).
Article PubMed CAS Google Scholar
Chen, S. J. et al. Single-cell analysis reveals transcriptomic remodellings in distinct cell types that contribute to human prostate cancer progression. Nat. Cell Biol. 23, 87–98 (2021).
Article PubMed CAS Google Scholar
Ma, X. S. et al. Identification of a distinct luminal subgroup diagnosing and stratifying early stage prostate cancer by tissue-based single-cell RNA sequencing. Mol. Cancer 19, 147 (2020).
Article PubMed PubMed Central CAS Google Scholar
Luo, H. et al. Characterizing dedifferentiation of thyroid cancer by integrated analysis. Sci. Adv. 7, eabf3657 (2021).
Article ADS PubMed PubMed Central CAS Google Scholar
Chen, Z. H. et al. Single-cell RNA sequencing highlights the role of inflammatory cancer-associated fibroblasts in bladder urothelial carcinoma. Nat. Commun. 11, 5077 (2020).
Article ADS PubMed PubMed Central CAS Google Scholar
Qian, J. B. et al. A pan-cancer blueprint of the heterogeneous tumor microenvironment revealed by single-cell profiling. Cell Res. 30, 745–762 (2020).
Article PubMed PubMed Central CAS Google Scholar
Zhang, M. et al. Single-cell transcriptomic architecture and intercellular crosstalk of human intrahepatic cholangiocarcinoma. J. Hepatol. 73, 1118–1130 (2020).
Article PubMed CAS Google Scholar
Peng, J. Y. et al. Single-cell RNA-seq highlights intra-tumoral heterogeneity and malignant progression in pancreatic ductal adenocarcinoma. Cell Res. 29, 725–738 (2019).
Article PubMed PubMed Central CAS Google Scholar
Lin, W. et al. Single-cell transcriptome analysis of tumor and stromal compartments of pancreatic ductal adenocarcinoma primary tumors and metastatic lesions. Genome Med. 12, 80 (2020).
Article PubMed PubMed Central CAS Google Scholar
Wolf, F. A., Angerer, P. & Theis, F. J. SCANPY: large-scale single-cell gene expression data analysis. Genome Biol. 19, 15 (2018).
Article PubMed PubMed Central Google Scholar
Gayoso, A. et al. A Python library for probabilistic analysis of single-cell omics data. Nat. Biotechnol. 40, 163–166 (2022).
Article PubMed CAS Google Scholar
Bergen, V., Lange, M., Peidli, S., Wolf, F. A. & Theis, F. J. Generalizing RNA velocity to transient cell states through dynamical modeling. Nat. Biotechnol. 38, 1408–1414 (2020).
Article PubMed CAS Google Scholar
Weiler, P., Lange, M., Klein, M., Pe’er, D. & Theis, F. CellRank 2: unified fate mapping in multiview single-cell data. Nat. Methods 21, 1196–1205 (2024).
Article PubMed PubMed Central CAS Google Scholar
Xue, R. et al. Liver tumour immune microenvironment subtypes and neutrophil heterogeneity. Nature 612, 141–147 (2022).
Article ADS PubMed CAS Google Scholar
Lausen, B. & Schumacher, M. Maximally selected rank statistics. Biometrics 48, 73–85 (1992).
Article Google Scholar
Grambsch, P. M. & Therneau, T. M. Proportional hazards tests and diagnostics based on weighted residuals. Biometrika 81, 515–526 (1994).
Article MathSciNet MATH Google Scholar
Krishna, C. et al. Single-cell sequencing links multiregional immune landscapes and tissue-resident T cells in ccRCC to tumor topology and therapy efficacy. Cancer Cell 39, 662–677 (2021).
Article PubMed PubMed Central CAS Google Scholar
Wu, W. Proteostatic stress response drives T cell exhaustion and immune evasion. Zenodo https://doi.org/10.5281/zenodo.16323779 (2025).

Download references

Acknowledgements

We thank K. Weller and B. McElwain for assisting with cell sorting and ImageStream analysis; Y. Koguchi, K. Reynolds and R. Davenport, J. Li and J. Jin for technical support; and all members of the Z.L. Laboratory and the Pelotonia Institute for Immuno-Oncology (PIIO) for discussions throughout the course of the study. This work was supported by NIH grants R01CA262069 (Z.L.), R01CA282501 (Z.L.), R35GM150723 (B.C.S.), R21CA267394 (B.C.S.) and R01AI170926 (H.E.G.). M.S.A., D.E.G. and C.M.B. were supported by an Emory MP3 Seed Grant (121861), a Winship Cancer Center Invest$ Pilot Award (147150) and a Swedish Research Council grant (2023-00510), respectively.

Author information

Authors and Affiliations

Pelotonia Institute for Immuno-Oncology, The Ohio State University Comprehensive Cancer Center–The James, Columbus, OH, USA
Yi Wang, Anjun Ma, No-Joon Song, Ariana E. Shannon, Yaa S. Amankwah, Ziyu Wang, Jay K. Mandula, Maria Velegraki, Tong Xiao, Haitao Wen, Hazem E. Ghoneim, Gang Xin, Brian C. Searle & Zihai Li
Department of Biomedical Informatics, College of Medicine, The Ohio State University, Columbus, OH, USA
Anjun Ma, Ariana E. Shannon, Weidong Wu, Gautam Ghosh & Brian C. Searle
Johns Hopkins University, Baltimore, MD, USA
Xingyu Chen
Department of Microbial Infection and Immunity, College of Medicine, The Ohio State University, Columbus, OH, USA
Abbey A. Saadey, Amir Yousif, Haitao Wen, Stanley Ching-Cheng Huang, Hazem E. Ghoneim & Gang Xin
Center for Childhood Cancer, Hematology/Oncology and BMT, Abigail Wexner Research Institute at Nationwide Children’s Hospital, Department of Pediatrics, The Ohio State University, Columbus, OH, USA
Ruoning Wang
Department of Pathology and Laboratory Medicine, Emory University, Atlanta, GA, USA
Christian M. Beusch, Abdelhameed S. Dawood, David E. Gordon & Mohamed S. Abdel-Hakeem
Department of Surgical Sciences, Uppsala University, Uppsala, Sweden
Christian M. Beusch
Division of Medical Oncology, Department of Medicine, College of Medicine, The Ohio State University, Columbus, OH, USA
Zihai Li

Authors

Yi Wang
View author publications
Search author on:PubMed Google Scholar
Anjun Ma
View author publications
Search author on:PubMed Google Scholar
No-Joon Song
View author publications
Search author on:PubMed Google Scholar
Ariana E. Shannon
View author publications
Search author on:PubMed Google Scholar
Yaa S. Amankwah
View author publications
Search author on:PubMed Google Scholar
Xingyu Chen
View author publications
Search author on:PubMed Google Scholar
Weidong Wu
View author publications
Search author on:PubMed Google Scholar
Ziyu Wang
View author publications
Search author on:PubMed Google Scholar
Abbey A. Saadey
View author publications
Search author on:PubMed Google Scholar
Amir Yousif
View author publications
Search author on:PubMed Google Scholar
Gautam Ghosh
View author publications
Search author on:PubMed Google Scholar
Jay K. Mandula
View author publications
Search author on:PubMed Google Scholar
Maria Velegraki
View author publications
Search author on:PubMed Google Scholar
Tong Xiao
View author publications
Search author on:PubMed Google Scholar
Haitao Wen
View author publications
Search author on:PubMed Google Scholar
Stanley Ching-Cheng Huang
View author publications
Search author on:PubMed Google Scholar
Ruoning Wang
View author publications
Search author on:PubMed Google Scholar
Christian M. Beusch
View author publications
Search author on:PubMed Google Scholar
Abdelhameed S. Dawood
View author publications
Search author on:PubMed Google Scholar
David E. Gordon
View author publications
Search author on:PubMed Google Scholar
Mohamed S. Abdel-Hakeem
View author publications
Search author on:PubMed Google Scholar
Hazem E. Ghoneim
View author publications
Search author on:PubMed Google Scholar
Gang Xin
View author publications
Search author on:PubMed Google Scholar
Brian C. Searle
View author publications
Search author on:PubMed Google Scholar
Zihai Li
View author publications
Search author on:PubMed Google Scholar

Contributions

Z.L., B.C.S. and Y.W. conceived the project, designed the experiments and wrote the paper with input from the rest of the authors. Y.W. was involved in all aspects of this study, including performing experiments, data analyses and interpretation. B.C.S. led the MS-based proteomic study. B.C.S., A.E.S. and G.G. performed liquid chromatography and tandem MS data acquisition and analyses. A.M., W.W., A.E.S. and B.C.S. performed bioinformatic analysis on the proteomic data and bulk RNA-seq data. X.C. performed pan-cancer scRNA-seq dataset collection and analyses. A.A.S., A.Y., T.X., G.X. and H.E.G. performed LCMV infection experiments. C.M.B., A.S.D., D.E.G. and M.S.A.-H. performed additional proteomic analyses on LCMV model data. N.-J.S., Y.S.A., M.V., J.K.M. and Z.W. assisted in performing experiments and manuscript preparation. H.W., S.C.-C.H. and R.W. assisted in data interpretation and manuscript preparation. Z.L. provided overall supervision of the study.

Corresponding author

Correspondence to Zihai Li.

Ethics declarations

Competing interests

Z.L. serves as a member of the scientific advisory board for HanchorBio. B.C.S. is a founder and shareholder in Proteome Software, which operates in the field of proteomics. No conflicts of interest were disclosed by the other authors.

Peer review

Peer review information

Nature thanks the anonymous reviewer(s) for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data figures and tables

Extended Data Fig. 1 Study of the proteome changes during T cell exhaustion using an in vitro exhaustion model and quantitative mass spectrometry.

a, Schematic diagram of generation of mouse CD8⁺ Tex cells in vitro by repetitive TCR stimulations followed by parallel bulk RNA-seq and mass spectrometry. b, Proliferation and viability of Teff and Tex cells during the 8-day course post initial activation (n = 4, two-way ANOVA). c, Representative flow cytometry plots and statistics of the percentage of PD-1⁺TIM-3⁺ (upper) and IFNγ⁺TNFα⁺ (bottom) cells over the total CD8⁺ T cells in Teff versus Tex cells (n = 3, two-tailed t test). d, Schematic diagram of the workflow of the chromatogram library-based data-independent acquisition (DIA) mass spectrometry. e, Expression levels of indicated proteins measured by flow cytometry (line) and mass spectrometry (scatter-bar). Teff cells were day 3 acutely activated CD8⁺ cells in vitro. Early and late Tex cells were day 3 and day 7 chronically activated CD8⁺ T cells. Data are presented as mean ± SD. Diagrams in a and d were created in BioRender. Wang, Y. (2025) https://BioRender.com/n0ibmgq.

Source Data

Extended Data Fig. 2 Association as well as discordance between RNA and protein expression level in T cells.

a, Correlation of RNA and protein expression levels of all detected molecules in Teff and Tex cells 8 days post initial activation. Each dot represents one molecule. A regression line with a shaded 95% confidence interval is shown. (r: Pearson correlation coefficient). b-c, Correlation between RNA levels and protein expression fold changes of molecules across one customer gene set and seven different gene ontologies between Tex and Teff cells (r: Pearson correlation coefficient; n: numbers of proteins/genes mapped to each gene ontology, A regression line with a shaded 95% confidence interval is shown). d, Bar plots of protein and mRNA expression changes of selected transcription factors in Tex vs Teff cells. e, Bar plots of protein and RNA expression changes of molecules belonging to the indicated gene ontologies in Tex vs. Teff cells.

Extended Data Fig. 3 Study of T cell exhausion using an antigen-specific system against LCMV.

a, Gating strategies for sorting antigen-specific CD8⁺ T cells into different subpopulations from the spleens of LCMV infected mice. Tet, tetramer. b, Heatmap of expression levels of 17 gene signatures in gp33⁺ T cell subpopulations on day 30 post LCMV Clone 13 infection (the scRNA-seq data were from Zander et al.). T_prog, T_int, T_tex subpopulations were determined based on their expression level of Slamf6 and Cx3cr1. c, Schematic diagram of longitudinal proteome analysis of P14 T cells transferred to mice and harvested 8, 15 and 30 days after infection with LCMV Armstrong or Clone 13. Created in BioRender. Wang, Y. (2025) https://BioRender.com/n0ibmgq. d, Heatmap of expression levels of proteins belonging to different T cell signatures in P14 T cells from LCMV Armstrong or Clone 13 infected mice. e, Expression dynamics of protein chaperones in P14 cells analyzed as in c (n = 4, two-tailed t test). Data are presented as box plots displaying the median (center line), 25th and 75th percentiles (bounds of the box), and minimum and maximum values (whiskers).

Extended Data Fig. 4 Mass spectrometry-based quantitative proteomics uncovers proteome changes during T cell exhaustion.

a, Bubble plot of the expressions of proteins associated with T cell stemness, activation and exhaustion. Bubble color intensity is proportional to protein expression level. Bubble size is proportional to the absolute z-score value. b, Enrichment analysis of proteins in cluster 1–5 and 7, 8 by Enrichr (one-sided Fisher’s exact test with Benjamini–Hochberg correction). Cluster 6 data was show in main Fig. 1. Top 10 significant gene ontologies are shown. c, PCA analysis of the proteome of in vitro-generated Teff and Tex cells at indicated time points. d, Schematic diagram of generating Teff, early and late Tex in vitro. Created in BioRender. Wang, Y. (2025) https://BioRender.com/n0ibmgq. e, Gene set enrichment analysis of differentially expressed proteins in early Tex versus Teff (top) and late Tex versus Teff cells (bottom). f, Volcano plots of differentially expressed proteins in late Tex versus Teff cells (two-sided t-tests with Benjamini–Hochberg correction). Proteins belonging to GO term “Responses to ER stress” are colored.

Extended Data Fig. 5 Exhausted T cells do not have expected features of translational attenuation.

a, Bar plot of protein and RNA expression changes of IRES motif-containing genes in Tex vs Teff cells 8 days post initial activation from the in vitro exhaustion model. b, Correlations of RNA and protein expression fold changes of EIF2A-dependent and EIF2A-independent genes in Tex vs Teff cells. A regression line with a shaded 95% confidence interval is shown. c, Enrichment analysis by Enrichr on proteins showing upregulation in both RNA and protein level (circled, left) in Tex versus Teff cells (one-sided Fisher’s exact test with Benjamini–Hochberg correction). A regression line with a shaded 95% confidence interval is shown in the left panel. d, PDCD4 expression across different CD8⁺ T cells subpopulations from in vitro exhaustion and in vivo tumor models by mass spectrometry (n = 5 for in vitro model, n = 3 for MC38 and MB49 tumor models, one-way ANOVA). Data are presented as mean (center line) and min to max (box bounds). e, The left panel is the heatmap of protein expression levels associated with selected Gene Ontology (GO) terms in in vitro-generated Teff and Tex cells 8 days post initial activation. Gene Set Enrichment Analysis (GSEA, right) compares Tex versus Teff cells, with normalized enrichment score (NES), gene rank distribution, and associated unadjusted P values are shown. P values were estimated using an adaptive multi-level split Monte Carlo scheme. f, Volcano plots of differentially expressed proteins in Tex versus Teff cells. Red: translation-related proteins; Blue: transcription-related proteins; Grey: other proteins (two-sided t-tests with Benjamini–Hochberg correction). g, Mass spectrometry analysis of protein expression levels of RPL13, eIF4E and eEF2 (n = 3 for Teff and n = 4 for Tex, two-tailed t test).

Source Data

Extended Data Fig. 6 Protein synthesis is upregulated in exhausted T cells in both tumor and chronic LCMV infection models.

a, Volcano plots of differentially expressed proteins in P14 T cells from LCMV acute and chronic infections as described in Extended Data Fig. 3c. Proteins belonging to the protein translation (GO:0006412) are colored in red. b, eIF2D and eIF4G3 expression in P14 T cells from LCMV acute and chronic infections (n = 4, multiple t test). Data are presented as box plots displaying the median (center line), 25th and 75th percentiles (bounds of the box), and minimum and maximum values (whiskers). c-d, Flow cytometry histogram and bar plot of 30-min HPG incorporation ex vivo in tumor-infiltrating CD8⁺ T cell subpopulations from mouse MC38 (c) and MB49 (d) tumors (spleen: n = 4, TIL: n = 7, one-way ANOVA). Data in c and d are presented as mean ± SD.

Source Data

Extended Data Fig. 7 Stress granules, proteasomal and lysosomal activities are increased in Tex cells.

All experiments were performed using mouse Teff and Tex cells generated in vitro 8 days post initial activation. a, ImageStream analysis of G3BP1 stress granules (SGs) in Teff and Tex cells. b, Quantification of the frequency of G3BP1 SGs in Teff and Tex cells determined with Bright Detail Intensity by ImageStream analysis. At least 2,000 live cells were examined for each replicate (n = 4, two-tailed t test). c, Flow cytometry quantification of G3BP1 expression in Teff and Tex cells (n = 4, two-tailed t test). d, Representative flow cytometry plots and statistics of G3BP1 expression in G3bp1 knockout and control wild-type Tex cells (n = 4, two-tailed t test). e, Representative flow cytometry plots and statistics of the percentage of IFNγ⁺TNFα⁺ cells out of the total CD8⁺ T cells in G3bp1 knockout and wild-type Tex cells (n = 4, two-tailed t test). f, Viability of G3bp1 knockout and wild-type Tex cells (n = 4, two-tailed t test). g, Proteasomal activity in Teff and Tex cells (n = 4, two-tailed t test). h, Flow cytometry quantification of lysosome mass in Teff and Tex cells (n = 6, two-tailed t test). i, Degradation kinetics of newly synthesized proteins in Teff and Tex cells (n = 3). j, Fold changes of HPG MFI 8 h post “pulsing” with MG132, Bafilomycin A1(BafA1) or DMSO in Tex cells (n = 5, one-way ANOVA). Data in b-h and j are presented as mean ± SD. Data in i are presented as mean ± SEM.

Source Data

Extended Data Fig. 8 Effector molecule granzyme B but not inhibitory receptor PD-1 is enriched in insoluble aggregates of Tex cells.

a, Schematic diagram of fractionate soluble and insoluble proteins in the cell lysates of Teff and Tex cells 8 days post initial activation in vitro for mass spectrometry analysis. Created in BioRender. Wang, Y. (2025) https://BioRender.com/n0ibmgq. b, Representative chromatograms of the retention time and intensity for peptides indicative of granzyme B in insoluble and soluble protein. Raw intensity is on the y-axis and retention time in minutes is on the x-axis. Each colored line represents a fragment ion associated with the peptide for granzyme B. Solid lines indicate fragment ions used for quantitation. Dotted lines represent interfering ions that were excluded for quantitation. The plot on the far right showed the fold change of granzyme B in the insoluble over soluble fractions in Tex and Teff cells. ΔΔ: Tex[Log2FC(insoluble/soluble)] - Teff[Log2FC(insoluble/soluble)]. c, Representative chromatograms of the retention time and intensity for peptides indicative of PD-1 in the insoluble and soluble fractions. The fold changes of PD-1 in the insoluble over soluble fractions in Tex and Teff cells are also shown.

Extended Data Fig. 9 Moderate attenuation of Akt signaling but not PI3K and mTOR pathways prevents T cell exhaustion without compromising cell survival and proliferation.

Mouse CD8⁺ T cells were first activated in vitro for 48 h, followed by chronic TCR stimulation in the presence of various pharmacological inhibitors and the final downstream analysis. a-f, Representative flow cytometry plots and statistics of the percentages of live cells (a), Ki67⁺ population (b), SLAMF6⁺TIM-3^- population (c), IFNγ⁺TNFα⁺ population (d), CD44⁺CD62L⁺ population (e) and CD107a production (f). For measuring cytokine production (IFNγ, TNFα and CD107a), cells were rested for two additional days after 6-day treatment and restimulated with PMA/Ionomycin (n = 4, one-way ANOVA). Data in a -f are presented as mean ± SD.

Source Data

Extended Data Fig. 10 Impact of deletion of TexPSR chaperones on T cell exhaustion.

Experiments were performed using in vitro exhausted T cells (a-g) and LCMV model in vivo (h). a, Expressions of the 10 candidate proteins in T cell exhaustion across 4 differentiation states as described in Extended Data Fig. 4d by mass spectrometry (n = 5, one-way ANOVA). Data are presented as mean (center line) and min to max (box bounds). b, Confirmation of CRISPR knockout of respective genes by RT-PCR. c, Confirmation of CRISPR knockout of indicated genes at the protein level by flow cytometry. d, Bar plot of the percentage of IFNγ⁺TNFα⁺ cells out of the total CD8⁺ T cells 8 days post CRISPR-Cas9 knockout of indicated genes and chronic stimulation (n = 3, experiment was repeated three times). e, Representative flow cytometry plot and percentages of SLAMF6^-TIM-3⁺ population over total live CD8⁺ T cells after Hspa5, Ero1a or Hsp90b1 knockout or control (n = 3, one-way ANOVA). f, UMAP visualization of Tex cells with Hspa5, Ero1a or Hsp90b1 knockout or control T cells analyzed with 26 markers associated with T cell activation and exhaustion indicated in g by multi-spectral flow cytometry. g, Heatmap of expression pattern of 26 markers across 12 cell clusters. Data was shown using all concatenated cells. h, gp96 expression in different Tex subsets gated on CD44⁺PD-1⁺ CD8⁺ T cells from day 30 LCMV Clone 13 infected wild type mouse spleens. Naïve CD8⁺ T cells were CD44^-CD8⁺ T cells from uninfected mice (n = 5 for naïve CD8⁺, n = 10 for T_prog, T_int and T_tex, one-way ANOVA). Data in b-e and h are presented as mean ± SD.

Source Data

Extended Data Fig. 11 Patients with high TexPSR signature in tumor-infiltrating T cells do poorly in response to immunotherapy.

a, Overview of CD8⁺ T cell counts and subtype distributions across different cancer types from integrated analysis of publicly available pan-cancer scRNA-seq datasets. Left panel: total number of CD8⁺ T cells detected in each cancer type, plotted on a logarithmic scale. Middle left panel: sample counts in each cancer type, plotted on a logarithmic scale. Middle right panel: number of independent studies for each cancer type. Right panel: CD8⁺ T cell subtype proportion. BLCA: bladder urothelial carcinoma; BRCA: breast invasive carcinoma; CESC: cervical squamous cell carcinoma and endocervical adenocarcinoma; CHOL: cholangiocarcinoma; CRC: colorectal cancer; GIM: gastric and intestinal malignancies; HNSC: head and neck squamous cell carcinoma; LC: lung cancer; LIHC: liver hepatocellular carcinoma; OC: ovarian cancer; PAAD: pancreatic adenocarcinoma; PDAC: pancreatic ductal adenocarcinoma; PRAD: prostate adenocarcinoma; RCC/CRCC: renal cell carcinoma/clear cell renal cell carcinoma; STAD: stomach adenocarcinoma; THCA: thyroid carcinoma; UCEC: uterine corpus endometrial carcinoma. b, Heatmap showing the expression level of TexPSR signature genes across various T cell populations. c, Correlation of RNA and protein expression fold changes of TexPSR signature genes between Tex versus Teff cells from the in vitro exhaustion model. d, TexPSR signature scores across different T cell subpopulations (two-sided Wilcoxon rank sum test). e, UMAP embedding of scRNA-seq data from T cells colored by annotated cell types. Overlaid are RNA velocity-based trajectories, inferred with the Slingshot algorithm, delineating two major differentiation pathways: Trajectory 1 (teal) toward exhaustion and Trajectory 2 (orange) toward effector differentiation. f. Signature scores along pseudotime for each trajectory. Trajectory 1 (left) shows decreasing Naive and increasing Exhaustion and TexPSR scores, consistent with an exhaustion trajectory. Trajectory 2 (right) shows dynamic changes in TexPSR and Naive scores, indicative of an alternative effector trajectory. g, Kaplan-Meier survival curves for patients with liver cancer stratified by high vs. low TexPSR signature, calculated from the CD8⁺ T cell scRNAseq data. P values were calculated using a univariate Cox proportional hazards model. h, TexPSR scores in scRNA-seq datasets of CD8⁺ T cells from non-responders (NR) and responders (R) in response to αPD-1 combined with αCTLA-4 therapy for RCC (NR: n = 5354 cells from 6 patients, R: n = 15207 cells from 9 patients). Data are presented as mean ± SEM, and statistical significance was determined using two-sided Wilcoxon rank sum tests.

Extended Data Fig. 12 Working model of TexPSR as a mechanistic driver of CD8⁺ T cell exhaustion.

The model presented illustrates a cancer cell at the top interacting with an exhausted T cell at the bottom through the MHC class I–T cell receptor complex. The intracellular events inside of the Tex cells underscore key novel findings from the study, which collectively elucidate a proteotoxic stress response (PSR) driving T cell exhaustion, a phenomenon we term TexPSR. TexPSR is characterized by an elevated rate of protein synthesis, the accumulation of intracellular protein aggregates, the upregulation of specific endoplasmic reticulum (ER) chaperones, and an accelerated protein catabolism pathway. Central to these processes is the sustained activation of AKT signaling, which orchestrates this complex proteotoxic environment. These findings not only provide deeper insight into the molecular mechanisms of T cell dysfunction within the tumor microenvironment but also open avenues for possible therapeutic strategies aimed at mitigating TexPSR to restore immune cell function in cancer immunotherapy. The figure was reproduced with the permission of The Ohio State University.

Supplementary information

Supplementary Information

Supplementary Figs. 1 and 2, which display the uncropped blots and gating strategies.

Reporting Summary

Supplementary Table 1

List of antibodies used.

Supplementary Table 2

sgRNA sequences.

Supplementary Table 3

Normalized protein expression tables.

Supplementary Table 4

Protein to gene name conversion table.

Supplementary Table 5

Summary of pan-cancer single-cell RNAseq datasets.

Source data

Source Data Fig. 2

Source Data Fig. 3

Source Data Fig. 4

Source Data Fig. 5

Source Data Extended Data Fig. 1

Source Data Extended Data Fig. 5

Source Data Extended Data Fig. 6

Source Data Extended Data Fig. 7

Source Data Extended Data Fig. 9

Source Data Extended Data Fig. 10

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Wang, Y., Ma, A., Song, NJ. et al. Proteotoxic stress response drives T cell exhaustion and immune evasion. Nature (2025). https://doi.org/10.1038/s41586-025-09539-1

Download citation

Received: 12 June 2024
Accepted: 15 August 2025
Published: 01 October 2025
DOI: https://doi.org/10.1038/s41586-025-09539-1

Subjects

Abstract

Main

Discordance between RNA and protein levels

PSR and specialized chaperone enrichment

Dysregulated proteostasis in Tex cells

Molecular definition of protein aggregates

Misfolded proteins drive exhaustion

Sustained AKT activity causes exhaustion

Tex-PSR chaperones underlie exhaustion

Clinical relevance of Tex-PSR in cancer

Discussion

Methods

Cell lines

Mice

T cell isolation, stimulation and drug treatment

Tumour challenge and TIL isolation

Flow cytometry

Protein synthesis rate measurement

Cell sorting

LCMV infection model

Gene editing in T cells by CRISPR–Cas9

Protein electrophoresis and western blotting

Retrovirus packaging and T cell transduction

ACT experiment

Immunofluorescence analysis by confocal microscopy

MS sample processing

MS acquisition

Proteomic data analysis

Bulk RNA-seq sample preparation and data analysis

Statistical comparison of protein expression and gene expression

Gene signature score analysis

Pan-cancer scRNA-seq data collection

Quality control and preprocessing of the pan-cancer scRNA-seq data

Batch correction and data integration

Cell-type annotation of pan-cancer scRNA-seq data

Functional signature calculation for scRNA-seq data

RNA velocity and trajectory inference

Validation of the Tex-PSR signature in CD8+ T cells and its prognostic impact

Tex-PSR signature expression in CD8+ T cells and its impact on patient survival

Tex-PSR expression in immunotherapy-treated patients

Statistical analysis

Reporting summary

Data availability

Code availability

References

Acknowledgements

Author information

Authors and Affiliations

Contributions

Corresponding author

Ethics declarations

Competing interests

Peer review

Peer review information

Additional information

Extended data figures and tables

Supplementary information

Source data

Rights and permissions

About this article

Cite this article

Share this article

Search

Quick links

Dysregulated proteostasis in T_ex cells

T_ex-PSR chaperones underlie exhaustion

Clinical relevance of T_ex-PSR in cancer

Validation of the T_ex-PSR signature in CD8⁺ T cells and its prognostic impact

T_ex-PSR signature expression in CD8⁺ T cells and its impact on patient survival

T_ex-PSR expression in immunotherapy-treated patients