In vivo CRISPR screening in head and neck cancer reveals Uchl5 as an immunotherapy target

Fu, Cong; Saddawi-Konefka, Robert; Chinai, Jordan M.; Kim, Sarah Y.; Kammula, Ashwin V.; Perera, Jonathan J.; Jiang, Aiping; Tiwari, Payal; Kistler, Erin N.; Tang, Shiqi; Luna, Sarah M.; Colvin, Kayla J.; Dubrot, Juan; Anderson, Seth; Fetterman, Rachel A.; Chuong, Cun Lan; Lane-Reticker, Sarah K.; Cheruiyot, Collins K.; Muscato, Audrey J.; Alipour, Zahra; Adkins, Douglas R.; Griffin, Gabriel K.; Bernstein, Bradley E.; Egloff, Ann Marie; Yates, Kathleen B.; Chernock, Rebecca D.; Gutkind, J. Silvio; Uppaluri, Ravindra; Manguso, Robert T.

doi:10.1038/s41467-025-63592-y

Download PDF

Article
Open access
Published: 29 September 2025

In vivo CRISPR screening in head and neck cancer reveals Uchl5 as an immunotherapy target

Nature Communications volume 16, Article number: 8572 (2025) Cite this article

1746 Accesses
19 Altmetric
Metrics details

Subjects

Abstract

Recurrent/metastatic head and neck squamous cell carcinoma (HNSCC) is an aggressive malignancy with a significant unmet need for enhancing immunotherapy response given current modest efficacy. Here, we perform an in vivo CRISPR screen in an HNSCC mouse model to identify immune evasion genes. We identify several regulators of immune checkpoint blockade (ICB) response, including the ubiquitin C-terminal hydrolase 5 (UCHL5). Loss of Uchl5 in tumors increases CD8⁺ T cell infiltration and improved ICB responses. Uchl5 deficiency attenuates extracellular matrix (ECM) production and epithelial-mesenchymal-transition (EMT) transcriptional programs, which contribute to stromal desmoplasia, a histologic finding we describe as associated with reduced anti-PD1 response in human HNSCCs. COL17A1, a collagen highly and specifically expressed in HNSCC, mediates in part Uchl5-mediated immune evasion. Our findings suggest an unappreciated role for UCHL5 in promoting EMT in HNSCC and highlight ECM modulation as a strategy to improve immunotherapy responses.

Acquired resistance to immunotherapy and chemoradiation in MYC amplified head and neck cancer

Article Open access 23 May 2024

In vivo CRISPR screens reveal the landscape of immune evasion pathways across cancer

Article 23 September 2022

Mucosa-like differentiation of head and neck cancer cells is inducible and drives the epigenetic loss of cell malignancy

Article Open access 02 October 2024

Introduction

Head and neck squamous cell carcinoma (HNSCC) is the seventh most prevalent cancer worldwide, primarily affecting the mucosal surfaces of four key anatomical sites: the oral cavity, sinonasal cavity, pharynx, and larynx^1,2. Despite therapeutic advancements, HNSCC not associated with human papillomavirus (HPV) infection remains a leading cause of global cancer mortality, with approximately 202,000 cases/year and a 5-year survival rate of only 50%³. This emphasizes the critical need for novel therapeutic approaches to improve outcomes for HNSCC patients. In this regard, immunotherapy has given hope for recurrent/metastatic HNSCC (R/M HNSCC) patients. Immune checkpoint blockade (ICB) targeting PD1 and CTLA-4 has shown remarkable success in various cancers, and FDA has approved two PD1 inhibitors to treat HNSCC⁴. However, the anti-PD1 response rate in R/M HNSCC patients (15-20%) remains modest, underscoring the pressing need for further investigation and design of more efficacious therapies. Consequently, elucidating the mechanisms underlying resistance to immune-checkpoint blockade will reveal potential biomarkers for predicting patient response and validate promising therapeutic combinations to improve overall survival.

Many preclinical studies have highlighted the key role of epigenetic regulation in the immunomodulation of cancer^5,6,7. We previously identified EZH2 as a potential therapeutic target to augment antigen presentation and antitumor immunity in HNSCC⁸. Both genetically attenuated EZH2 expression and pharmacologic EZH2 inhibition resulted in significantly higher MHC-I expression and antigen presentation capacity and antigen-specific T cell proliferation. The combination of an EZH2 inhibitor and anti-PD1 significantly suppressed tumor progression of the anti-PD1-resistant HNSCC model MOC1-esc1⁸. Furthermore, our previous work identifying epigenetic modifiers that modulate anti-tumor immunity identified SETDB1 and the HUSH complex as factors that mask tumor intrinsic endogenous retrovirus expression and immunogenicity⁹. In addition to preclinical investigations, many clinical trials are exploring combinations of epigenetic agents with immunotherapy across various cancers¹⁰. In HNSCC, a notable clinical trial revealed that combining anti-PD1 with an HDAC inhibitor achieved ~32% overall response rate in clinical trials, surpassing the typical 15-20% seen with anti-PD1 alone in HNSCC^10,11. Two ongoing clinical trials are evaluating whether decitabine or azacitidine, hypomethylating agents, enhances responses to anti-PD1 therapy in recurrent/metastatic HNSCC (NCT03019003, NCT04624113). Given the emergence of combination epigenetic therapy and immunotherapy as a treatment strategy in HNSCC, we set out to identify unexplored epigenetic therapeutic targets to enhance anti-tumor immunity and advance the field of HNSCC immunotherapy.

We used an in vivo CRISPR screening approach to identify epigenetic regulators of ICB response in HNSCC. Among the most promising targets, we identified that the ubiquitin c-terminal hydrolase enzyme, UCHL5, modulates sensitivity to ICB. Using transcriptional profiling and functional studies, we show that Uchl5 regulates ICB efficacy by controlling the expression of extracellular matrix (ECM) components, especially collagens. These results not only underscore the potential impact of Uchl5 on the tumor microenvironment and treatment outcomes but also emphasize the significance of our epigenetic regulator screening, offering valuable insights for devising unexplored therapeutic approaches in HNSCC.

Results

A head and neck cancer in vivo CRISPR screen identifies epigenetic regulators of tumor immunity

We performed an In vivo CRISPR-Cas9 screen in a preclinical murine oral carcinoma MOC1-esc1 model using our previously published epigenetic regulator-focused sgRNA library^12,13. The MOC1-esc1 cell line was generated by isolating a tumor that escaped anti-PD1 immunotherapy and found to be intrinsically resistant to anti-PD1 upon subsequent rechallenge in naive C57BL/6 mice¹⁴. To perform an in vivo CRISPR screen in MOC1-esc1, we cloned the epigenetic regulators (936 genes) focused sgRNA library⁹, positive control genes (n = 3), and intergenic targeting control guides (n = 605) (Supplementary Data 1) into the non-immunogenic SCAR lentiviral vector system¹⁵. First, we engineered the MOC1-esc1 cell line to express pSCAR-Cas9 and confirmed high genome editing efficiency (Supplementary Fig. 1A and 1B). Next, we transduced the sgRNA library into the Cas9⁺ MOC1-esc1 cell line. After selection, CRISPR editing, and IDLV-Cre transduction to remove immunogenic vector components, the library-edited MOC1-esc1 tumor cells were implanted into three groups of mice with increasing levels of immune selective pressure: immunodeficient NSG (NOD-SCID-Il2rg null) mice, immunocompetent WT C57BL/6 mice, and WT mice treated with ICB (anti-PD1 and anti-CTLA-4) (Fig. 1A). Treatment with ICB resulted in tumor regression compared to untreated WT and NSG mice (Fig. 1B). The engrafted tumors were harvested for analysis 18 days after transplantation. Meanwhile, the library-edited tumor cells were maintained in vitro for the duration of the In vivo experiments to enable the identification of genes essential for growth. We performed PCR amplification and sequenced the sgRNA region from tumor genomic DNA, quantified sgRNA counts, and then conducted quality control analyses to confirm an adequate representation of the sgRNA library. These analyses indicated excellent screen performance, with the vast majority of sgRNAs well represented in all conditions and individual replicates (Supplementary Fig. 1C–1F).

Our screen design enabled the identification of genes essential for the growth of the MOC1-esc1 cell line both in vitro and In vivo (in immunodeficient NSG mice). To assess whether there is consistency between the intrinsic dependencies observed in vitro and in vivo, we performed a comparative analysis between the NSG mouse model (in vivo) and the in vitro samples. This approach allowed us to identify genes critical for the growth of the MOC1-esc1 cell line either in vitro, In vivo, or in both conditions. While some genes required for in vitro growth were also important In vivo, we observed differences, indicating that certain genes play a more prominent role in the In vivo context (Supplementary Fig. 1G). To identify genes that regulate the tumor response, we compared sgRNA frequencies between the ICB-treated, untreated, and NSG groups (Supplementary Data 2). As expected based on previous data, sgRNAs targeting H2-T23, Ptpn2 and Adar were significantly depleted in the untreated and immunotherapy-treated groups compared with the NSG group, as well as in the immunotherapy-treated versus untreated group (Fig. 1C–E and Supplementary Fig. 2A)^16,17,18. We compared the results from the immunotherapy-treated group versus NSG group or untreated group and found genes encoding factors with established roles in anti-tumor immunity (such as Asf1a) among the top depleted genes (Fig. 1F)¹⁹, observing high concordance between the two comparisons. We also identified previously uncharacterized factors, including genes for INO80 (Uchl5, Nfrkb, and Tfpt), TFIID (Taf4, Taf6, and Taf8), SIN3 (Sap18, Arid4a, and Suds3), EP300 (Ep300, Ing5, and Crebbp), and E3 ligase or deubiquitinase (DUB) (Bap1, Jade2, Cul5, and Rnf2) complex proteins (Fig. 1F).

To further validate candidates from the In vivo CRISPR screen, top targets from these complexes were used to construct a validation pool library (Supplementary Data 3), focusing on 13 candidate genes from the primary screen (Fig. 1G). The validation pool library transduced MOC1-esc1 cells were transplanted into NSG mice, immunocompetent WT C57BL/6 mice, and WT mice treated with either anti-PD1 alone or a combination of anti-PD1 and anti-CTLA-4. As anticipated, we observed depletion of Ptpn2 which plays a pivotal role in anti-tumor immunity, alongside enrichment of genes like Jmjd1c, aligning with trends in the primary screen (Fig. 1H and Supplementary Fig. 2B). Additionally, we noted a striking depletion of the top genes identified in the primary screen, such as Nfrkb, Bap1, and Uchl5, in mice treated with immunotherapy compared to NSG mice (Fig. 1H and Supplementary Fig. 2B).

NFRKB mediates recruitment of UCHL5 to the INO80 complex, a nucleosome remodeling complex, and inhibits UCHL5 by blocking the ubiquitin-binding site and by disrupting the enzyme active site. Conversely, UCHL5 protects NFRKB from proteasomal degradation via its deubiquitinating function^20,21. BRCA1-associated protein-1 (BAP1) and UCHL5 both belong to the ubiquitin C-terminal hydrolases (UCHs) family of deubiquitinating enzymes²². BAP1 has emerged as a critical tumor suppressor across multiple cancer types, predisposing to tumor development when mutated in the germline as well as somatically²³. We utilized the SCAR system to generate Nfrkb and Bap1 KO MOC1-esc1 cells. In validation experiments, tumors lacking Nfrkb exhibited increased sensitivity to anti-PD1 treatment, although no discernible difference was observed in NSG mice (Supplementary Fig. 2C). The knockout of Bap1 rendered MOC1-esc1 cells more susceptible to anti-PD1/CTLA-4 treatment yet demonstrated no response to monotherapy anti-PD1 or in groups without ICB treatment (Supplementary Fig. 2D-S2E). Notably, Bap1 KO MOC-esc1 tumors exhibited a growth advantage in NSG mice compared with control tumors, consistent with its role as a tumor suppressor (Supplementary Fig. 2E).

UCHL5, also known as UCH37, is a deubiquitinating enzyme known for its specificity for the distal subunit of Lys48-linked poly Ub chains^24,25. UCH37/UCHL5 functions in two large and very different complexes, the 26S proteasome and the INO80 chromatin remodeler, influencing diverse signaling pathways implicated in tumor progression, including TGF-β, Wnt, DNA repair, cell cycle, and NF-κB activation^{21,26,27,28,29,30}. Both the primary and validation library pooled screens identified Uchl5 as one of the most depleted genes regulating tumor response to immunotherapy. Uchl5 is consistently overexpressed in multiple cancers, including esophageal squamous cell carcinoma, and is often associated with poor prognosis^31,32. Several compounds, including the antifouling paint biocide copper pyrithione (CuPT) and b-AP15, target the proteasomal deubiquitinases UCHL5 and USP14^33,34,35. In vivo studies further demonstrate b-AP15’s efficacy in halting tumor progression, including in HNSCC models^36,37, highlighting the potential of therapeutically targeting UCHL5 deubiquitinating activity in HNSCC. Moreover, when comparing the MOC1-esc1 screen with our previously published In vivo screens across various cancer models¹⁷, we found that Uchl5 sgRNAs were uniquely depleted in the MOC1-esc1 screen but not in others (Fig. 1I), suggesting a specific role for UCHL5 in HNSCC. Consistent with finding from the In vivo screens across various cancer models, no significant differences were observed between control and Uchl5 KO Lewis lung carcinomas (LLCs) in either NSG or WT mice with or without treatment using a combination of anti-PD1 and anti-CTLA-4 (Supplementary Fig. 2F), further demonstrating the specific role for UCHL5 in HNSCC.

We also validated the enhanced sensitivity of Uchl5-deficient tumor cells compared to controls through an In vivo competition assay. Mixtures of control and Uchl5-deficient MOC1-esc1 cells, generated with the pSCAR system (Supplementary Fig. 3A), were implanted into NSG mice, untreated WT mice, and cohorts treated with anti-PD1 (Fig. 1J). Consistent with both primary and validation pool screens, Uchl5-deficient cells showed selective depletion in untreated, and anti-PD1-treated mice compared to NSG mice (Fig. 1K). Based on these results, we hypothesized that Uchl5 promotes the suppression of tumor immunity in HNSCC and prioritized Uchl5 for further investigation.

Uchl5 deficiency in tumors enhances sensitivity to anti-PD1 treatment

To determine whether Uchl5-deficient MOC1-esc1 tumor cells display enhanced response to ICB, we implanted them into NSG mice, untreated WT mice, or WT mice treated with either monotherapy anti-PD1 or combination anti-PD1 + anti-CTLA-4 (Fig. 2A). MOC1-esc1 cells lacking Uchl5 implanted in NSG or untreated WT mice grew similarly to control tumors In vivo (Fig. 2A). In contrast, Uchl5-deficient tumors exhibited dramatically enhanced response to immunotherapy, both anti-PD1 monotherapy and combination anti-PD1/CTLA-4, compared with control tumors (Fig. 2A). Survival analysis showed that ICB-treated mice with Uchl5-deficient tumors had prolonged survival and a greater number of complete responses (75% cure rate for monotherapy anti-PD1 treatment and 80% for combination of anti-PD-1 + anti-CTLA-4) compared with control tumors (Fig. 2B). Reexpression of UCHL5 into Uchl5-deficient MOC1-esc1 tumor cells (Supplementary Fig. 3B) rescued their tumor progression in response to ICB (Supplementary Fig. 3C), supporting the specificity of this effect for UCHL5. To determine whether responses against Uchl5-deficient tumors would generate durable immune memory, we performed a rechallenge experiment in animals that had previously cured Uchl5-deficient MOC1-esc1 tumors (Fig. 2C). We implanted unmodified MOC1-esc1 tumor cells in both naive and survivor mice and showed that while tumors grew progressively in naive mice, all survivor mice rejected subsequent tumor challenge (Fig. 2D). Thus, loss of Uchl5 in immunotherapy-treated tumors elicits potent and durable anti-tumor immunity.

To determine whether the role for UCHL5 is specific to the MOC1-esc1 model or more generalizable to HNSCC, we evaluated Uchl5 deletion in the MOC2 and 4MOSC1 murine models of HNSCC^12,38. MOC2, which is a highly aggressive and treatment refractory cell line that spontaneously metastasizes to draining lymph nodes following flank transplantation¹², displays complete resistance to combination of anti-PD1 and anti-CTLA-4 treatment³⁹. While Uchl5 deletion in MOC2 cells had no effect on tumor growth in untreated NSG or WT mice, it moderately but significantly decreased tumor growth in response to ICB treatment (Supplementary Fig. 3D and 3E). To further confirm the role of Uchl5 in anti-tumor immunity in HNSCC, we investigated whether Uchl5 deficiency could lead to sensitivity to anti-PD1 treatment in the 4MOSC1 model, another syngeneic animal model of tobacco-associated oral cancer³⁸. Control and Uchl5-deficient 4MOSC1 cells (Supplementary Fig. 3F) were implanted orthotopically in the buccal mucosa of NSG mice, untreated WT mice, and WT mice treated with anti-PD1. Uchl5 deficiency in 4MOSC1 demonstrated minimal impact on tumor growth in untreated WT mice and an increase in tumor growth in NSG mice, but caused a striking sensitivity to anti-PD1 treatment (70% cure rate for Uchl5 KO and 0% for control) (Fig. 2E, F). The rechallenge experiment demonstrated that immune responses against Uchl5-deficient 4MOSC1 tumors generate durable immune memory (Supplementary Fig. 3G). Thus, Uchl5 deficiency significantly enhanced the sensitivity of several mouse HNSCC tumor models to immune checkpoint blockade. These findings confirmed the primary screen results and established UCHL5 as a promising target for immunotherapy in HNSCC.

Radiotherapy is a standard treatment for head and neck cancer, often used alone or in combination with surgery, chemotherapy and immunotherapy, depending on the cancer stage and location². To assess the antitumor effects of radiation, radiation dosage optimization showed that 5 Gy modestly impacted tumor growth while 15 Gy led to reduced growth. Next, we investigated whether Uchl5 deficiency influenced the response to combined radiotherapy and immunotherapy (Supplementary Fig. 3H). Control and Uchl5 KO MOC1-esc1 cells were transplanted into WT mice, followed by treatment with a combination of anti-PD1 therapy and 5 Gy irradiation. Uchl5 KO MOC1-esc1 cells showed enhanced sensitivity to the combination treatment compared to control cells (Supplementary Fig. 3I).

Tumor cell intrinsic Uchl5 deficiency promotes intratumoral CD8⁺ T cell infiltration

To determine how Uchl5 deletion impacts the tumor microenvironment (TME), we compared the composition of immune cell subsets in control and Uchl5-deficient MOC1-esc1 tumors using flow cytometry (Supplementary Fig. 4). We observed no significant difference in the total number of infiltrating CD45⁺ cells, CD4⁺ T cells, Foxp3⁺ regulatory T cells, NK cells, or myeloid cells in Uchl5-deficient tumors relative to control tumors, or in the percentage of CD45⁺ cells (Fig. 3A–C and Supplementary Fig. 5A–H). However, Uchl5-deficient tumors exhibited a significantly greater number and percentage of CD8⁺ T cells (Fig. 3A–C). Moreover, both the number and percentage of Perforin⁺ cytotoxic CD8⁺ T lymphocytes (CTLs) showed a substantial increase in Uchl5-deficient tumors (Fig. 3B, C). This was further supported by a notable elevation in the ratio of CTLs to Treg cells within Uchl5-deficient tumors (Fig. 3D). Reduced percentage of Myeloid-Derived Suppressor Cells (MDSCs) within total CD45⁺ cells was also observed in Uchl5-deficient tumors compared to control tumors (Fig. 3C). We next investigated the immune cell subset composition within the tumor microenvironment of control and Uchl5-deficient orthotopic 4MOSC1 tumors using flow cytometry. Interestingly, we found that Uchl5-deficient 4MOSC1 tumors exhibited a significantly higher percentage of CD8⁺ and CD4⁺ T cells within total CD45⁺ cells, consistent with observations in MOC1-esc1 tumors (Supplementary Fig. 5I and J). Thus, Uchl5 deletion significantly enhances the infiltration of CD8⁺ T cells with a cytotoxic phenotype in HNSCC tumors.

To determine the importance of CD8⁺ T cells in Uchl5 deficiency-induced immunotherapy sensitivity, we used a well-established antibody-based approach to deplete CD8⁺ T cells⁴⁰. Depleting CD8⁺ T cells abrogated the growth delay of Uchl5-deficient tumors treated with anti-PD1 (Fig. 3E), confirming that CD8⁺ T cells are necessary for sensitivity to Uchl5 deletion. Next, we investigated the impact of Uchl5 deficiency on CTL-mediated tumor cell killing using OT-1 transgenic T cells. We expressed the model antigen OVA in both control and Uchl5-deficient MOC1-esc1 cells. Flow cytometry using a monoclonal antibody specific to the H2-Kb conjugated SIINFEKL epitope from OVA showed surface antigen expression but revealed no discernible difference between Uchl5-deficient and control MOC1-esc1 cells after IFNγ stimulation (Supplementary Fig. 6A), confirming that MHC-I presentation is not impacted by Uchl5 deletion. We generated a 1:1 mixture of control and Uchl5-deficient OVA⁺ MOC1-esc1 tumor cells, which were then implanted into NSG mice followed by PBS or OT-1 CD8⁺ T cell transfer on day 7 post-implantation (Fig. 3F). Transfer of OT-1 CD8⁺ T cells led to tumor regression compared with the PBS group, demonstrating the technical success of the experiment (Supplementary Fig. 6B). Uchl5-deficient OVA⁺ MOC1-esc1 tumor cells exhibited markedly higher sensitivity to OT-1-mediated killing compared to control cells In vivo (Fig. 3G). These findings suggest that CD8⁺ T cells are both necessary and sufficient for Uchl5 deficiency-induced immunotherapy sensitivity.

Tumor cell intrinsic Uchl5 deficiency remodels extracellular matrix

We next focused on the molecular mechanism underlying the enhanced CTL-mediated killing observed in Uchl5-deficient tumor cells. We found that the surface expression pattern of classical MHC-I (H2-Kb/H2Db) remained unaltered on the surface of IFNγ-stimulated Uchl5-deficient MOC1-esc1 cells compared to control cells (Supplementary Fig. 6C). Likewise, Uchl5 deficiency did not elicit any discernible effect on the IFNγ-mediated expression induction of PD-L1 or non-classical MHC-I Qa-1^b in MOC1-esc1 cells (Supplementary Fig. 6C).

Given that UCHL5 is a deubiquitinating enzyme known for its specificity for the distal subunit of Lys48-linked poly Ub chains^24,25, we next tested whether the total level of ubiquitinated protein in cells was changed. As a positive control, we observed a dramatic accumulation of ubiquitinated proteins in cells treated with b-AP15 (Supplementary Fig. 7A), which effectively inhibits UCHL5 and USP14 activity while sparing the 20S core particle³⁵. We also observed an accumulation of ubiquitinated proteins in Uchl5-deficient MOC1-esc1 cells compared to controls, and reintroduction of UCHL5 into these cells attenuated the accumulation (Supplementary Fig. 7B), suggesting that UCHL5 target specific protein for deubiquitination rather than exerting a global effect. UCHL5 can also function as part of the chromatin remodeling INO80 complex, which regulates gene expression, DNA repair, and replication by sliding nucleosomes²⁰. Although several members of this complex did not score as hits, one potential binding partner, Nfrkb, was significantly depleted in the PD1-treated group compared with NSG mice (Fig. 1H and Supplementary Fig. 2C). These data indicate that Uchl5 likely regulates ICB efficacy at both transcriptional and post-translational levels.

To further elucidate the molecular mechanisms underlying the augmented CTL-mediated killing of Uchl5-deficient tumor cells, we inoculated control or Uchl5-deficient MOC1-esc1 tumor cells into wild-type C57BL/6 mice treated with or without anti-PD-1. Subsequently, RNA-seq analysis was performed on both control and Uchl5-deficient tumor cells isolated from bulk tumor tissue (Fig. 4A), with over 95% of the purified cells identified as tumor cells (Supplementary Fig. 7C). The Principal Component Analysis (PCA) revealed a clear and distinct separation between all experimental groups (Fig. 4A). We identified roughly one thousand significantly up and downregulated genes (log₂FoldChange > 0.5, q < 0.01) in Uchl5-deficient tumors compared to control, in the presence or absence of anti-PD1 treatment (Fig. 4B, Supplementary Data 4). Gene set enrichment analysis (GSEA) revealed that Uchl5-deficient MOC1-esc1 cells displayed downregulation of genes in extracellular matrix (ECM) related pathways, including core matrisome^41,42, Hallmark Epithelial-Mesenchymal Transition (EMT) and partial-EMT (p-EMT)43 compared to control MOC1-esc1 cells under anti-PD1 treatment (Fig. 4C, D, Supplementary Data 5). The matrisome encompasses ECM and ECM-associated proteins shaping the cellular microenvironment, with the core matrisome comprising ECM glycoproteins, collagens, and proteoglycans⁴². One study in primary and metastatic oral squamous cell carcinoma (OSCC) revealed a distinct subset of cell populations at the tumor-stroma interface by single-cell transcriptomics analysis⁴³. These cells expressed genes associated with an Extracellular Matrix (ECM) program and exhibited features of Epithelial-Mesenchymal Transition while retaining epithelial markers, a phenomenon termed p-EMT⁴³. Alterations in ECM composition, including changes in collagen types, glycoproteins, and proteoglycans, or modifications in ECM stiffness and architecture mediated by proteins like fibronectin and laminins, have been implicated in inducing or supporting EMT^44,45,46. In Uchl5-deficient tumors isolated from mice treated with anti-PD-1 antibody, numerous genes encoding glycoproteins, collagens, and proteoglycans were downregulated compared with control tumors (Fig. 4D). We also observed similar results in Uchl5-deficient MOC1-esc1 tumors without anti-PD1 treatment (Supplementary Fig. 7D and E). Similar observations were made using in vitro RNA-seq analysis of Uchl5-deficient and control tumors, with or without stimulation with IFNγ, ruling out that signal from ECM-related genes results from differential abundance of non-tumor cells, such as fibroblasts or macrophages, in the TME (Supplementary Fig. 8A–F, Supplementary Data 6). Interestingly, we found that the top differentially downregulated or enriched genes in Uchl5-deficient MOC1-esc1 cells were similarly affected in Nfrkb-deficient MOC1-esc1 cells (Supplementary Fig. 8G and H, Supplementary Data 7). Additionally, we also observed a downregulation of p-EMT in Nfrkb-deficient MOC1-esc1 cells compared with control cells (Supplementary Fig. 8G and H, Supplementary Data 7). These findings suggest a potential role of UCHL5 in orchestrating the extracellular matrix within MOC1-esc1 tumors through its interaction with NFRKB.

**Fig. 4: Tumor cell intrinsic *Uchl5* deficiency remodels extracellular matrix.**

Subsequently, we explored whether UCHL5 expression levels correlated with ECM-related gene sets in tumor cells utilizing single-cell RNA-seq data from human HNSCC patients (Fig. 4E)⁴³. We derived gene sets of differentially upregulated and downregulated genes in mouse MOC1-esc1 Uchl5-deficient tumors relative to controls and checked for an association between UCHL5 expression in the patient scRNA-seq data. Encouragingly, we found that the top 50 differentially downregulated genes in Uchl5-deficient MOC1-esc1 tumors strongly correlated with the expression of UCHL5 in human HNSCC, and the top 50 upregulated genes showed a modest, though not significant, negative correlation with UCHL5 expression (Fig. 4F). Remarkably, we also observed an enrichment of p-EMT in human HNSCC patients with relatively higher UCHL5 expression levels (Fig. 4F). This result suggests the involvement of UCHL5 in remodeling the extracellular matrix in human HNSCC patients.

ECM accumulation is frequently observed within the tumor microenvironment, and often results in a pronounced fibrotic reaction known as stromal desmoplasia^43,47. Stromal desmoplasia is associated with the development of a dense ECM, characterized by elevated levels of total fibrillar collagen, fibronectin, proteoglycans, and tenascin C⁴⁸. To assess the possible contribution of stromal desmoplasia to resistance to anti-PD1, we focused on tumor samples from our previous clinical studies, where neoadjuvant anti-PD-1 therapy prior to surgery in HPV-negative HNSCCs showed pathologic anti-tumor responses in a subset of patients. Pathologic analysis of stromal desmoplasia was conducted on pretreatment biopsy hematoxylin and eosin (HE) stained sections from these patients (Fig. 4G)⁴⁹. Among the 51 enrolled patients, stromal desmoplasia was observed in 27 cases (Fig. 4H)^49,50. Intriguingly, among 24 patients without stromal desmoplasia, significant post-treatment regression (pTR-2) was observed in 11 patients (45.87%), while moderate PRT (pTR-1) was noted in 3 patients (12.5%). Conversely, only 4 patients (14.8%) with stromal desmoplasia exhibited pTR-2, with no instances of pTR-1 observed in this subgroup (Fig. 4H, Fisher’s exact test: P = 0.003; Supplementary Data 8). These findings suggest a negative association between stromal desmoplasia and anti-PD1 treatment response, possibly suggesting that UCHL5 can serve as a potential prognostic marker and therapeutic target to improve HNSCC cancer immunotherapy.

Uchl5-mediated control of Col17a1 expression contributes to immune evasion and ICB resistance in HNSCC

Collagens represent a predominant component within the extracellular matrix of tumor tissue, playing pivotal roles in tumor development and progression^51,52. Transcriptomic analysis of ex vivo sorted tumor cells revealed a notable downregulation in mRNA expression levels of many collagens in Uchl5-deficient tumors, including Col12a1, Col16a1, Col17a1, and Col6a1 (Supplementary Fig. 9A). Of particular interest, collagen XVII, encoded by the COL17A1 gene, is a transmembrane-type collagen molecule crucial for anchoring hemidesmosomes at the dermal–epidermal junction⁵³. In the context of epidermal squamous cell carcinomas, COL17A1 emerges as an early marker of malignant transformation. Immunohistochemical analysis has revealed a heightened immunoreaction of collagen XVII ectodomain in primary oral SCC tumors and associated metastases⁵⁴. It has also been identified as an emerging biomarker for predicting the prognosis of HNSCC cancer. Collagen XVII immunohistochemistry (IHC) staining was a positive predictor of mortality in HNSCC patients, underscoring its clinical significance⁵⁵. However, the role of COL17A1 in immune evasion in HNSCC is poorly understood. COL17A1 is highly and specifically expressed in HNSCC compared to other cancer types, as evidenced by TCGA analysis (Supplementary Fig. 9B) and corroborated by the Cancer Cell Line Encyclopedia (CCLE) dataset (Fig. 5A), which highlights specific expression of COL17A1 in HNSCC cell lines. These findings underscore the potential significance of COL17A1 in the context of HNSCC tumorigenesis.

We thus next validated whether COL17A1 protein levels were reduced in Uchl5-deficient MOC1-esc1 tumor cells, as we observed in our RNA-seq analysis (Fig. 4D and S9A). Western blotting confirmed a decrease in the expression of COL17A1 in Uchl5-deficient MOC1-esc1 cells compared with control cells. Intriguingly, reintroduction of Uchl5 successfully restored the protein level of COL17A1 (Fig. 5B). Our in vitro and In vivo RNA-seq data indicate that Col17A1 expression is transcriptionally reduced in Uchl5 KO MOC1-esc1 tumor cells. To determine whether Col17A1 is regulated at the transcriptional level, post translationally, or both, we performed qPCR to assess whether reintroducing UCHL5 could restore Col17a1 mRNA levels. We observed a significant increase in Col17a1 mRNA expression compared to control MOC1-esc1 cells (Supplementary Fig. 9C), suggesting that UCHL5 primarily regulates Col17a1 through transcriptional mechanisms. This observation was further corroborated by Western blot results obtained from In vivo tumor samples, where a similar pattern of decreased COL17A1 expression in Uchl5-deficient tumors was observed (Fig. 5C, D). Moreover, immunohistochemistry staining demonstrated a reduced proportion of COL17A1⁺ tumor cells and staining intensity for COL17A1 in Uchl5-deficient tumors compared to controls (Fig. 5E, F; Supplementary Fig. 9D and E). Thus, Col17a1, along with other collagen and ECM factors, is downregulated in Uchl5-deficient MOC1-esc1 cell lines and tumors. To determine whether this downregulation of Col17a1 has a functional impact on anti-tumor immunity, we generated Col17a1 knockout MOC1-esc1 cells (Supplementary Fig. 9F). Subsequently, we implanted control or Col17a1-deficient tumor cells into NSG mice, untreated WT mice, or WT mice treated with anti-PD1. MOC1-esc1 cells lacking Col17a1 displayed comparable growth rates to control tumors when implanted in NSG mice. However, Col17a1-deficient tumors exhibited significantly reduced growth in untreated WT mice and showed an enhanced response to anti-PD-1 treatment compared to control tumors (Fig. 5G and Supplementary Fig. 9G). As deletion of Uchl5 caused the downregulation of many collagens and ECM components in addition to Col17a1, the less pronounced ICB sensitizing effect of Col17a1 deletion compared to Uchl5 suggested that functional redundancy between ECM factors could mask the effect of a single gene. However, overexpression studies would not be limited by functional redundancy and could thus further support a link between Col17a1 expression and loss of Uchl5. We proceeded to overexpress COL17A1 in Uchl5-deficient MOC1-esc1 tumor cells (Fig. 5H). Subsequently, these cells, along with Uchl5-deficient MOC1-esc1 cells overexpressing human truncated CD19 as a control, were implanted into WT mice treated with anti-PD1. Remarkably, we observed that the tumor growth of Uchl5-deficient MOC1-esc1 cells was rescued by COL17A1 overexpression (Fig. 5I), suggesting that the decreased expression of COL17A1 resulting from Uchl5 deficiency at least partially contributes to the suppression of MOC1-esc1 tumor growth. Thus, Uchl5 mediates immune evasion in HNSCC by modulating the expression of collagen and ECM factors.

Discussion

Immune-checkpoint blockade has improved outcomes for patients with R/M HNSCC, but the majority remain unresponsive to immunotherapy³. The modest response rate highlights an urgent need for continued investigation of tumor cell intrinsic resistance mechanisms. Here, we conducted an unbiased In vivo CRISPR KO screen using a syngeneic head and neck cancer model to map the epigenetic immune evasion landscape of HNSCC tumors. We discovered that deletion of the deubiquitinating enzyme Uchl5, not previously linked to immunotherapy responses, led to increased CD8⁺ T cell infiltration, enhanced sensitivity to ICB, and improved survival in mouse models of HNSCC. Mechanistic investigations revealed that Uchl5 loss reshaped the ECM, reducing matrix deposition and suppressing the EMT process associated with stromal desmoplasia in human HNSCC patients. Additionally, our study demonstrated that loss of Col17a1, a highly expressed HNSCC-specific collagen, sensitized cancer cells to anti-PD1 treatment, while overexpression of Col17a1 in Uchl5-deficient cells partially reversed the sensitivity to anti-PD1 response. Our findings, therefore, identify Uchl5 as a previously unrecognized regulator of anti-tumor immunity and underscore the role of Uchl5 in regulating ECM composition and modulating the effects of immunotherapies.

UCHL5 belongs to the Ubiquitin C-terminal hydrolases subfamily of deubiquitinating enzyme and catalyzes the removal of ubiquitin from target proteins, generating free monomeric Ub^24,56. Although Uchl5 has consistently been observed to be overexpressed in numerous cancers, including esophageal squamous cell carcinoma, and is often correlated with poor prognosis^31,32, its specific role in HNSCC remains largely unexplored. While some studies have suggested that its dysregulation can be linked to cancer progression, notably in HNSCC^31,32, UCHL5 has not previously been linked to immune evasion or cancer immunotherapy response. In our study, we provided direct evidence implicating the role of Uchl5 in anti-tumor immunity in head and neck cancer. According to RNA-seq data and functional studies, Uchl5 appears to regulate ICB efficacy by controlling the expression of collagen and extracellular matrix (ECM) components at the transcriptional level, consistent with its role as a member of the INO80 nucleosome remodeling complex. As previously mentioned, another INO80 complex member, the transcriptional regulator NFRKB, inhibits UCHL5 by disrupting its active site, while UCHL5 protects NFRKB from proteasomal degradation through its deubiquitinating function^20,21. We observed similar change between Uchl5-deficient and Nfrkb-deficient MOC-esc1 cells compared with control. Therefore, it is possible that UCHL5 promotes the expression of ECM genes through the stabilization of NFRKB. Since Col17a1 overexpression compensates for loss of Uchl5, it is indeed possible that a non-degradable/stable version of NFRKB could compensate for the loss of UCHL5, given their functional relationship. However, experimentally generating and validating a stable version of NFRKB presents technical challenges. To date, the overall structure of NFRKB remains unclear, with approximately 65% of NFRKB predicted to be in a disordered state⁵⁷, and only two structural domains associated with protein-protein interactions have been identified), called as Winged-Helix Like Domain and DEUBAD^58,59, Further, its degradation and turnover mechanisms remain uncharacterized. More studies are needed to elucidate the interplay between UCHL5 and NFRKB. Additionally, while we observed an accumulation of K48-ubiquitinated protein in Uchl5-deficient MOC1-esc1 cells, further investigation is required to identify the specific substrate driving the sensitivity of Uchl5-deficient cells to ICB treatment.

Increasing evidence has highlighted the pivotal role of ECM in driving tumor advancement and resistance to diverse treatments, including immunotherapy^60,61. Throughout tumor growth, the ECM undergoes substantial remodeling, transitioning from a normal state to a tumor-specific ECM characterized by heightened collagen density and increased stiffness^47,62. As a predominant component within the extracellular matrix of tumor tissue, elevated collagen density and aligned collagen fibers have been associated with poor prognosis in various cancers, including oral squamous cell carcinomas^47,62,63,64. Recent studies have also underscored its crucial immune modulatory functions within the tumor microenvironment, influencing both cancer progression and the effectiveness of cancer immunotherapy^65,66,67,68. In the context of HNSCCs, which are known to be collagen-rich environments, individual collagen subtypes are expressed by both cancer-associated fibroblasts (CAFs) and malignant epithelium^51,69. However, the precise role of collagens in mediating resistance to immune checkpoint blockade in HNSCC is poorly defined. In our study, we uncover previously uncharacterized insights by demonstrating that Col17a1 expression in malignant cells contributes to anti-PD1 sensitivity driven by Uchl5 deficiency. Intriguingly, COL17A1 is highly and specifically expressed in HNSCC, likely due to its predominant expression in epithelial cells at the basal layer^53,70,71. Collagen XVII also emerges as a pivotal regulator governing the growth and progression of pancreatic cancer, orchestrating intricate interactions within the tumor microenvironment⁷². Future studies exploring the effect of collagen XVII downregulation on various signaling pathways in HNSCCs will help us to understand better how collagen XVII affects PD1 sensitivity.

We also observed downregulation of several other genes encoding collagen subunits, including Col16a1, Col6a1, Col6a2, and Col18a1, in Uchl5-deficient MOC1-esc1 tumors. These findings suggest that these collagens may also contribute to the PD1 sensitivity induced by Uchl5 deficiency. Further investigations are warranted to elucidate whether overexpression of these collagens in tumor cells can similarly reverse the anti-PD1 sensitivity phenotype. Moreover, it is essential to consider factors beyond collagen abundance, such as collagen alignment and distribution, which have been shown to impact clinical outcomes in cancer^72,73,74. We have yet to determine whether the alignment and distribution of collagens are altered in Uchl5-deficient tumors, and cannot dismiss the possibility of their involvement in the sensitivity to anti-PD1 treatment. Using CollaTIL, a unique computational pathology tool that quantifies the relationship between immune cells and collagen structure in the tumor microenvironment (TME) of gynecologic cancers, a study from Madabhushi’s lab found that increased immune cell infiltration is associated with chaotic collagen architecture, while immune-sparse TMEs exhibit more ordered collagen structures and lower entropy⁷⁵. Future studies using similar tools will be essential to explore the relationship between collagen architecture and immune cell infiltration in HNSCC tumor models.

Loss of Uchl5 not only induced downregulation of collagens, but also many other proteins of the extracellular matrix^42,76,77, especially the components of the basement membrane, including laminins, nidogens, periostin, SPARC, thrombospondins, tenascins, perlecan, agrin, versican, and PRELP (Fig. 4D)⁷⁸. These proteins can either bridge the laminin and Collagen IV networks to increase their stability and influence the structural integrity of the basement membrane or interact with surface receptors contributing to tissue-specific functions^79,80. Some of these components, such as versican and hyaluronan, have already been suggested to have direct immune modulatory function^81,82. This suggests that, in addition to collagen, other unidentified factors may regulate the anti-PD-1 response in Uchl5-deficient tumors. Although we observed enhanced infiltration of CD8⁺ T cells in Uchl5 KO tumors, we were surprised to find lower levels of cytokines and chemokines, such as Ccl2, Cxcl9, and Cxcl10, which are known to regulate CD8⁺ T cell recruitment, in Uchl5 KO tumor cells based on RNA-seq data (Supplementary Data 4, Supplementary Fig. 7F). This suggests that Uchl5 is unlikely to directly regulate the recruitment of CD8⁺ T cells. RNA-seq analysis was performed on tumor cells isolated from bulk tumor tissue, making it likely that other factors within the TME, such as immune cells and ECM components, contribute to modulating CD8⁺ T cell recruitment. Future studies are needed to determine whether Uchl5 has a direct role in immune cell communication beyond ECM regulation.

Stromal desmoplasia is a common feature observed across various cancer types, including pancreatic, breast, colorectal cancer and head and neck cancer. Its presence significantly influences cancer progression and treatment response by impacting tumor growth, invasion, metastasis, and the delivery of therapeutic agents to the tumor site^83,84,85,86. Marked by excessive ECM deposition and stromal cell activation, stromal desmoplasia shapes T cell distribution within tumors⁴⁸. The ECM modulates immune cell differentiation, migration, infiltration and polarization, thereby influencing antitumor immunity cell migration^{61,62,63,64,65,66,67,87}, Distinct collagen deposition patterns in tumors can form structural barriers that hinder T cells from leaving stromal regions or create “collagen highways” that misdirect T cells away from target cells⁸⁸. High-density collagen matrices reduce T cell migration speed, and collagen fibers can guide T cell movement, collectively limiting T cell infiltration into tumors^89,90,91. Several studies have demonstrated that modulating desmoplasia could potentially improve the efficacy of immunotherapy in melanoma and pancreatic ductal adenocarcinoma (PDAC)^92,93,94. For instance, inhibiting hematopoietic cell kinase (HCK) in myeloid cells has been shown to reduce the desmoplastic microenvironment, thereby improving immunotherapy outcomes in PDAC⁹⁵. However, the role of stromal desmoplasia in anti-PD1 responses in HNSCC remains uncertain. Our findings revealed that Uchl5 KO tumors exhibit reduced ECM components, particularly collagens, indicating an alleviation of stromal desmoplasia. This reduction is accompanied by a significant increase in CD8⁺ T cell infiltration, further supporting the notion that stromal desmoplasia and dense collagen create barriers that hinder CD8⁺ T cells trafficking. Our investigation identified a negative correlation between stromal desmoplasia and anti-PD1 pathologic response through the analysis of tumor samples from our prior clinical studies. Given its demonstrated role in ECM regulation, as revealed in our mouse data, UCHL5 emerges as a promising prognostic marker and therapeutic target for enhancing the efficacy of immunotherapy in HNSCC.

Emerging studies have revealed that different cancer interventions, including radiotherapy, chemotherapy, and immunotherapy, significantly remodel the ECM and contribute to an immunosuppressive microenvironment⁹⁶. Radiotherapy and chemotherapy activate cancer-associated fibroblasts (CAFs) and macrophages, increasing TGF-β, MMPs, and collagen secretion, leading to fibrosis and tumor invasiveness^97,98,99. Radiation-induced hypoxia upregulates HIF-1α and HIF-2α, promoting ECM remodeling and radiotherapy resistance, with both pathways strongly linked to treatment failure in head and neck cancer^98,100. CAFs also contribute to chemotherapy resistance by supporting cancer cell recovery and metabolic adaptations¹⁰¹, while chemotherapy-resistant HNSCC spheroids have been shown to upregulate EMT-associated stem markers, further emphasizing the role of ECM in tumor aggressiveness¹⁰². Although less studied in HNSCC, immunotherapy has been reported to induce fibrosis, as seen in non-small cell lung cancer, where 12% of patients treated with immune checkpoint inhibitors developed interstitial lung disease¹⁰³. While our study highlights UCHL5’s role in ECM regulation, it is conceivable that immunotherapy may also contribute to ECM changes in HNSCC, potentially affecting tumor progression and treatment response.

Overall, using In vivo CRISPR screens, we identified Uchl5 as a key regulator of immune evasion. Our study underscores the intricate interplay between extracellular matrix (ECM) and anti-tumor immunity in head and neck squamous cell carcinoma (HNSCC) and discovered an immune inhibitory role for the transmembrane collagen gene Col17a1, uniquely expressed in HNSCC. Our findings reveal a previously undescribed dependency on UCHL5 for anti-PD-1 resistance and propose a combinatorial therapeutic approach to improve treatment efficacy in HNSCC patients.

Methods

Mice

Female wildtype C57BL/6 wildtype mice aged 6 to 10 weeks were obtained from Jackson Laboratories. Immunodeficient NOD.Cg-PrkdcscidIl2rgtm1Wjl/SzJ (NSG) mice were bred on-site at the Broad Institute. All mice were housed in the Broad Institute’s animal facility, adhering to the institute’s animal care and use guidelines. Tumor challenges were carried out in age-matched mice under protocol number 0110-08-16-1, approved by the Broad Institute Institutional Animal Care and Use Committee. Experiments were terminated when tumors reached a maximum allowable size of 2 cm in the largest dimension, showed ulceration, or if mice exhibited signs of distress as defined by the institutional animal care guidelines.

Animal studies using the 4MOSC1 tumor model were conducted at Moores Cancer Center, San Diego, under protocol ASP #S15195, approved by the Institutional Animal Care and Use Committee (IACUC) of the University of California. The maximal tumor size permitted by this protocol was 8 mm in any dimension. At no point did the tumor size or burden in any animal exceed this approved limit. All mice were housed in specific pathogen-free (SPF) conditions at the Broad Institute or the Moores Cancer Center animal facilities. Mice were maintained under a 12-hour light/12-hour dark cycle at a controlled ambient temperature of 20–26 °C and a relative humidity of 30–70%, with ad libitum access to food and water.

Mice were monitored at least once daily for general health, tumor growth, and behavior throughout the study. Tumor dimensions (length and width) were measured with digital calipers, and volumes were calculated using the formula: volume = (length × width²) / 2. Disease progression was primarily assessed by tumor size, rate of growth, and the presence of ulceration. Humane endpoint criteria included ≥15% body weight loss, persistent hunched posture, ruffled fur, reduced mobility or activity, labored breathing, or failure to access food/water. Mice meeting any humane endpoint criteria were euthanized immediately via CO₂ inhalation followed by cervical dislocation, in accordance with protocols approved by the Institutional Animal Care and Use Committee (IACUC) of the Broad Institute and the Moores Cancer Center.

Murine cell lines

Mouse oral squamous cell carcinoma models, including MOC1-esc1 and MOC2, were maintained in IMDM/Hams-F12 (2:1) supplemented with 5% heat-inactivated FBS, 100 U/mL penicillin–streptomycin, 5 ng/mL EGF (Millipore), 40 ng/mL hydrocortisone (Sigma Aldrich), and 5 mg/mL insulin (Sigma Aldrich). 4MOSC1 cells were cultured in Defined Ketatinocyte-SFM medium supplemented with EGF Recombinant Mouse Protein (5 ng/ml), Cholera Toxin (50 pM) and 1% antibiotic/antimycotic solution. All Cell lines were routinely tested for Mycoplasma and underwent short tandem repeat cell line authentication within 6 months of use.

Murine oral squamous cell carcinoma lines (MOC1-esc1 and MOC2) were propagated in a 2:1 mixture of IMDM and Ham’s F-12 containing 5% heat-inactivated FBS, penicillin–streptomycin (100 U/mL), EGF (5 ng/mL; Millipore), hydrocortisone (40 ng/mL; Sigma-Aldrich), and insulin (5 μg/mL; Sigma-Aldrich). The 4MOSC1 line was maintained in Defined Keratinocyte-SFM supplemented with recombinant mouse EGF (5 ng/mL), cholera toxin (50 pM), and 1% antibiotic/antimycotic. Mycoplasma testing was performed on a routine schedule, and STR profiling confirmed cell identity within six months prior to use.

In vivo chromatin regulator CRISPR screen in the MOC1-esc1 model

A lentiviral CRISPR library containing 7806 sgRNAs targeting 936 genes (6 sgRNAs per gene) plus non-targeting controls⁹, was cloned into the pSCAR_sgRNA-060 vector. MOC1-esc1 cells stably expressing pSCAR_Cas9-hygromycin were transduced at ~30% infection efficiency, ensuring approximately 1000-fold representation for each sgRNA. After puromycin selection (3 µg/mL; Millipore) in IMDM/Ham’s-F12 medium, cells were maintained at >1000× coverage throughout the experiment. Following one week of selection and editing, the pool was infected with IDLV-Cre at a similar coverage level. SCAR vector expression was monitored via fluorescent reporters, and cells were cultured until fewer than 10% expressed GFP or mKate2.

For in vivo screening, 4 × 10⁶ MOC1-esc1 cells per tumor were prepared in a 1:1 mixture of growth factor–reduced Matrigel (Corning) and Hanks’ Balanced Salt Solution (HBSS) and implanted subcutaneously into the right flank of 40 NSG and 80 WT mice, providing ~250× coverage per sgRNA. Of the WT group, 40 mice were treated intraperitoneally with 100 µg anti–PD-1 and 100 µg anti–CTLA-4 antibodies on days 9, 12, and 15 after tumor inoculation. Parallel in vitro cultures were maintained with >1000× sgRNA coverage for the duration of the screen.

18 days after implantation, tumors and in vitro pools were collected. Tumors were finely minced, pooled by equal tissue mass (five tumors per pool), and digested with proteinase K. Genomic DNA was extracted using the QIAGEN Blood Maxi Kit. The sgRNA cassettes were PCR-amplified and sequenced on an Illumina HiSeq platform. Guide count data were normalized by z-score prior to comparison between groups. sgRNA enrichment or depletion was assessed, and statistically significant hits were identified using the STARS algorithm¹⁰⁴.

Generation and validation of gene deficient cell lines

MOC1-esc1, MOC2, and 4MOSC1 CRISPR knockout cell lines were generated by sequential lentiviral transductions. First, cells were infected with a Cas9-expressing vector (pSCAR_Cas9-hygromycin_GFP, Addgene 162075) and selected with hygromycin for 1–2 weeks. Cas9-expressing populations were then transduced with pSCAR_sgRNA_1 (pXPR_060, Addgene 162076) carrying a single guide RNA, followed by puromycin selection for 1–2 weeks. To remove immunogenic sequences present in the Cas9 and sgRNA constructs, the resulting lines were transduced with IDLV-Cre lentivirus¹⁵. Validation of knockout efficiency was performed by flow cytometry or Western blotting.

CRISPR sgRNA sequences

Gene name and sequence are as follows: Control sgRNA 1 GCGAGGTATTCGGCTCCGCG; control sgRNA 2 GCTTTCACGGAGGTTCGACG; control sgRNA 3 ATGTTGCAGTTCGGCTCGAT; control sgRNA 4 ACGTGTAAGGCGAACGCCTT; control sgRNA 5 ATTGTTCGACCGTCTACGGG; Cd47 sgRNA CCACATTACGGACGATGCAA; Uchl5 sgRNA 1 TCTTGTTTCAGGTTGCCGAG; Uchl5 sgRNA 2 TACACTTACGATAGCCTGAG; Bap1 sgRNA ACCTGTCTGAGTGCACTCAG; Nfrkb sgRNA CCTTGTGTGAAATATGACAT; Col17a1 sgRNA 1 ACGTCATGATATAGCACCT; Col17a1 sgRNA 2 ACACTCCCCATCCCCAAGAA.

In vivo tumor challenge and immune checkpoint blockade (ICB) treatments

In the mouse MOC model tumor challenge experiments, either 2 × 10⁶ MOC1-esc1 cells or 1 × 10⁵ MOC2 tumor cells were harvested and washed twice with ice-cold Hanks balanced salt solution (HBSS, Gibco). They were then suspended in HBSS and injected subcutaneously (s.c.) into the left and right flanks of mice. At specified time points, mice received treatment with 200 μg of rat monoclonal anti-PD-1 (BioXCell, clone 29 F.1A12) via intraperitoneal (i.p.) injection or a combination of 100 μg anti-PD-1 and 100 μg anti-CTLA-4 (Bio X Cell, clone 9D9). For the mouse 4MOSC1 model, 1 × 10⁶ cells were transplanted into the buccal mucosa of female C57Bl/6 mice (4–6 weeks of age, weighing 16–18 g). Upon tumor formation (around day 5–6), mice were randomized into groups. Treatment involved IP injection of anti-PD-1 antibody at specified intervals. In experiments involving CD8⁺ T cell depletion, mice received 200 μg of anti-CD8β (BioXCell, clone 53-5.8) via i.p. injection every 4 days, beginning 1 day prior to tumor inoculation and continuing throughout the experiment.

In vivo tumor growth competition assays

For in vivo competition assays, equal numbers of the indicated cell lines (1:1 ratio) were mixed and cultured for one passage before subcutaneous implantation of 2 × 10⁶ cells into NSG and WT mice. The same mixed cell populations were also maintained in parallel in vitro cultures for the duration of the assay. Immunotherapy treatments were administered to mice as described elsewhere in the text. Fifteen to eighteen days after implantation, tumors and corresponding in vitro samples were collected. Tumor tissues were finely minced and digested using Proteinase K with Buffer ATL (QIAGEN), and genomic DNA was purified using the QIAGEN Blood Maxi Kit. For each sample, 1–10 μg of genomic DNA was used as input for PCR amplification of the sgRNA cassettes, employing P5 and P7 primers according to protocols provided at https://portals.broadinstitute.org/gpp/public/resources/protocols. Amplicons were sequenced on an Illumina MiSeq platform. Raw base call files were converted to FASTQ format and demultiplexed using Illumina’s bcl2fastq v2.20.0 software. sgRNA counts and log-transformed abundance matrices (log[RPM + 1]) were generated using PoolQ v2.2.0.

Flow cytometry

Adherent cell lines were detached using Trypsin-EDTA (0.25%, Gibco) and subsequently rinsed with MACS buffer (PBS + 2% FBS + 5 mM EDTA). For cell surface staining, cells were exposed to fluorochrome-conjugated monoclonal antibodies for 15 to 30 minutes at 4 °C, followed by two washes with MACS buffer. Samples were then analyzed using a Beckman Coulter CytoFLEX flow cytometer and FlowJo software (FlowJo).

Analysis of tumor-infiltrating lymphocytes by flow cytometry

For the MOC1-esc1 model, mice received subcutaneous injections of 2 × 10⁶ CRISPR–Cas9 engineered MOC1-esc1 cells. Tumors were excised on day 15, weighed, and cut into small fragments before enzymatic dissociation using the Miltenyi tumor dissociation kit and the gentleMACS Dissociator (program m-TDK-1). Cell suspensions were filtered through 70-μm strainers, blocked with anti-mouse CD16/32 (BioLegend), and incubated on ice for 15–30 minutes with the specified surface antibodies. To exclude dead cells, LIVE/DEAD™ Fixable Near-IR dye (1:5000; Invitrogen) was included during surface staining. After washing, cells were fixed using the Foxp3/Transcription Factor Staining Buffer Set (eBiosciences), blocked with mouse and rat serum, and subsequently stained with intracellular antibodies. For quantification, Spherotech AccuCount Rainbow particles were added to samples before flow cytometric acquisition. Data were collected on an LSR Fortessa using single-color compensation and fluorescence-minus-one controls to define gates, and analyzed in FlowJo. Statistical comparisons between groups were performed using Student’s t-tests. Antibodies used at 1:100 dilution for flow cytometry: CD45 (Biolegend, 30-F11), TCRβ (Biolegend, H57-597), CD8α (Biolegend, QA17A07), CD4(Biolegend, GK1.5), NK1.1 (Biolegend, S17016D), CD19 (Biolegend, 1D3/CD19), Perforin (Biolegend, S16009A), Foxp3 (Biolegend, FJK-16s), CD11b (Biolegend, M1/70), MHCII (Biolegend, AF6-120.1), F4_80 (Biolegend, BM8), Gr.1 (Biolegend, RB6-8C5), CD163 (Biolegend, S15049I), CD64 (Biolegend, X54-5/7.1), Ter119 (Biolegend, TER-119), CD90.2 (Biolegend, 53-2.1)

In the 4MOSC1 tumor experiments, either control or Uchl5-deficient 4MOSC1 cells (1 × 10⁶ per tumor) were orthotopically implanted into the buccal mucosa of female C57BL/6 mice (4–6 weeks old). Mice received intraperitoneal injections of anti–PD-1 antibody (clone J43, BE0033-2) or PBS on days 6, 8, and 10 post-implantation. On day 12, tumors were harvested, minced, and processed into single-cell suspensions using the Miltenyi tumor dissociation kit and gentleMACS Dissociator, following the manufacturer’s tumor protocol. Cell suspensions were passed through 70-μm strainers, washed with PBS, and stained with Zombie viability dye (BioLegend) for live/dead discrimination. After additional washing with cell staining buffer (BioLegend 420201), surface staining was performed for 30 minutes at 4 °C in the dark using the following BioLegend antibodies at 1:100: CD45 (30-F11), CD3 (17A2), CD8α (53-6.7), CD4 (RM4-4), and CD19 (6D5). Samples were washed with MACS buffer and analyzed on a BD LSRII Fortessa, with data processed in FlowJo.

Primary CD8⁺ T cell isolation and culture

Primary murine CD8⁺ T cells were obtained from spleens of C57BL/6 mice. Spleens were mechanically dissociated, passed through a cell strainer, and treated with ACK lysing buffer to remove red blood cells. After washing with MACS buffer, CD8⁺ T cells were purified using the mouse CD8⁺ T Cell Isolation Kit (Miltenyi Biotec) following the manufacturer’s protocol.

Purified cells were seeded onto plates coated with purified NA/LE hamster anti-mouse CD3ε (1 µg/mL; BD Pharmingen, clone 145-2C11) and cultured in RPMI-1640 medium (GlutaMAX supplemented) containing 10% FBS, antibiotics (50 U/mL penicillin and 50 µg/mL streptomycin), 1× MEM non-essential amino acids, 10 mM HEPES, 1 mM sodium pyruvate, and 55 µM 2-mercaptoethanol. The medium was supplemented with recombinant human IL-2 (PeproTech) and purified NA/LE hamster anti-mouse CD28 (2 µg/mL; BD Pharmingen). After 24 h of activation, cells were maintained in recombinant human IL-2 (100 U/mL) and expanded in plates or flasks, with medium refreshed every 24–48 h for 5–7 days before downstream assays.

Lymphocyte depletion

To investigate the involvement of specific subsets of immune effector cells in mice, we utilized depletion methods targeting CD8⁺ T cells. This involved intraperitoneal injections of 200 μg of anti-CD8β antibody (BioXcell, clone 53-5.8) on days −1, 2, 6, and 10. As controls, equal amounts of IgG isotype antibodies (BioXcell) were administered.

OT-1 T cell culture and in vivo CD8⁺ T cell killing competition assay

OT-1 CD8⁺ T cells, expressing a transgene encoding a TCR that specifically recognizes the SIINFEKL peptide bound to mouse H-2Kb, were isolated from the spleens of OT-1 C57BL/6 mice. These OT-1 CD8⁺ T cells were cultured and activated according to the previous description. For the in vivo CD8+ T cell killing experiment, 1:1 mixes of the SIINFEKL peptide-expressing control and Uchl5-deficient MOC1-esc1 cells were cultured for one passage before implanting 2 × 106 cells into NSG mice subcutaneously. OT1 CD8⁺ T cells were injected via I.V. on day 7 after tumor inoculation. Throughout the duration of the in vivo competition assay, the tumor cell mixtures were maintained in in vitro culture. Tumors and in vitro cultures were harvested 6 days after OT1 CD8⁺ T cells injection. Genomic DNA extraction, amplification of the sgRNA region, sequencing, and sgRNA abundance analysis were done according to the previous description.

RNA sequencing

The control and Uchl5-deficient MOC1-esc1 cell line was treated with either DMSO or 10 ng/ml IFNγ for 24 h. Single-cell suspensions from cells cultured in a monolayer were homogenized in the RNA lysis buffer. RNA was extracted from homogenized samples using the RNA Miniprep Plus Kit in accordance with the manufacturer’s suggested protocol. For collected MOC1-esc1 tumors grown in vivo, tumors were collected on day 15, weighed, mechanically diced, digested with a tumor dissociation kit (Miltenyi) and a gentleMACS dissociator (Miltenyi) using the m-TDK-1 program. After filtering through a 70-μm strainer, tumor cells were isolated using the mouse tumor cell isolation kit from Miltenyi Biotec according to the manufacturer’s instructions. RNA was extracted from cell pellets using the Qiagen RNA Miniprep Plus Kit according to the manufacturer’s instructions. Illumina-barcoded libraries for RNA sequencing were generated utilizing the Takara SMARTer Stranded Total RNA-Seq Kit v2 following the manufacturer’s protocol. Subsequent sequencing was executed on an Illumina NovaSeq 6000 platform. Data underwent quality trimming using the Trimmomatic pipeline, with specified parameters: LEADING:15, TRAILING:15, SLIDINGWINDOW:4:15, MINLEN:16. Pre-and post-trimming quality control was done using FastQC (v0.11.7). Subsequently, data were aligned to the mouse reference genome mm10 using Kallisto (v0.46.0). HTSeq was utilized for mapping aligned reads to genes and generating a gene count matrix. PCA was used to assess sample quality and replicate concordance. One in vivo anti-PD-1 sgUchl5 sample was removed from the in vivo RNA-seq experiment as an outlier based on PCA and subsequent further analysis. Gene counts were aggregated from transcript counts using the tximport (v1.24.0) R package. Gene set enrichment analysis was also conducted, as outlined in prior studies^105,106.

Single cell analysis of data from human HNSCC patients

We investigated the gene sets correlating with UCHL5 expression levels in a human single-cell RNA-sequencing dataset of HNSCC patients⁴³. TPM counts were obtained from the Curated Cancer Cell Atlas (3CA) (https://www.weizmann.ac.il/sites/3CA/)¹⁰⁷. We excluded cells with fewer than 1,000 detected genes and samples with fewer than 10 malignant cells. Data were log2(TPM/10 + 1) normalized. Only cells annotated as malignant and from primary tumor samples were included (n = 13 patients, 1711 cells). No new human participant data were collected in this study.

We calculated a UCHL5 score with a control size of 50 using the Scanpy (1.9.8) ‘sc.tl.score_genes’ function. Cells were grouped into UCHL5 high (score > 0) and low (score ≤ 0) expression categories. The log fold change (LFC) in mean expression was calculated between the groups for each gene. P-values were determined using Welch’s t-test and adjusted for multiple hypothesis testing with the Benjamini-Hochberg method. -log10(adjusted p-value) values were calculated, with infinite values replaced by the maximum finite value plus one. Genes were sorted by -log10(adjusted p-value), directionalized by the sign of the LFC, and Gene Set Enrichment Analysis¹⁰⁶ was performed for hallmark gene sets, the ‘partial’ EMT gene set⁴³, and the top 50 upregulated and downregulated differentially expressed genes derived from comparison between Uchl5-deficient MOC1-esc1 mouse tumors compared with control tumors.

Clinical sample collection and pathologic assessment of stromal desmoplasia

Tumor samples used for stromal desmoplasia analysis were obtained from a previously published clinical study of neoadjuvant anti-PD-1 therapy in HPV-negative head and neck squamous cell carcinoma (HNSCC) patients (NCT02296684). All clinical samples were collected with informed consent under protocols approved by the appropriate Institutional Review Boards. Clinical response (pTR-0, pTR-1, pTR-2) was assessed following anti‑PD‑1 therapy. For this study, pretreatment biopsy samples were retrieved and hematoxylin and eosin (H&E)–stained sections were evaluated for the presence or absence of stromal desmoplasia by Dr. Rebecca D. Chernock. Comparative analyses of pathologic response and desmoplasia status were performed using Fisher’s exact test. No new patient data were collected for this study, and all analyses were conducted in accordance with approved ethical guidelines.

Western blot and antibodies

After washing with ice-cold PBS, cells were detached from the culture plate utilizing Cell lifter (VWR International LLC) in ice-cold PBS. Cell pellets were then acquired through centrifugation at 300 x g at 4 °C for 5 minutes, followed by lysis in the RIPA lysis buffer (Thermo Scientific) on ice for 20-30 min. Post-lysis, the cell lysates underwent centrifugation at > 13,000 x g at 4 °C for 15 min to collect the supernatants. Quantification of protein concentrations was performed utilizing the BCA protein assay kit (Thermo Fisher Scientific) and a SpectraMax M5 microplate reader (Molecular Devices). Lysates from samples (40-60 μg) were mixed with LDS sample buffer (Thermo Fisher Scientific) and subjected to protein denaturation at 70 °C according to the manufacturer’s instructions. The prepared samples were then electrophoresed on 4-12% NuPAGE Bis-Tris Gels (Thermo Fisher Scientific) and subsequently transferred onto 0.45 µm PVDF membranes (Thermo Fisher Scientific). The following primary antibodies were used to detect designated proteins: UCHL5 (Santa Cruz, sc-271003, 1:500), COL17A1 (Abcam, ab184996, 1:1000), α-Tubulin (Abcam, Ab7291, 1:10000), β-actin (Abcam, Ab6276, 1:10000), GAPDH (Abcam, Ab8245, 1:10000), Vinculin (Santa Cruz, sc-73614, 1:2000) and Flag (Sigma, F1804, 1:2000). IRDye 680RD or 800CW secondary antibodies (Li-COR Bioscience) were used at a 1:5000 and the images were acquired with the Odyssey Imager (Li-COR Bioscience).

Immunohistochemistry/immunofluorescence

MOC1-esc1 tumors were fully fixed in 10% neutral-buffered formalin for a duration of 24 h, followed by an overnight permeabilization step in 70% ethanol. After fixation, the tissues were embedded in paraffin, sectioned, and affixed onto slides for subsequent staining using mouse COL17A1 antibody (Abcam, Ab186415, 1:200). Imaging of the slides was conducted using Leica Aperio VERSA Scanning System, and analysis was performed utilizing Qupath software.

Ectopic gene expression

Full-length mouse UCHL5 construct was generated via polymerase chain reaction (PCR) using plasmids pcDNA3.1-mUchl5 obtained from Genscript as templates. These fragments were subsequently cloned into the pDONR211 vector (Thermo-Fisher) using BP Clonase II, and then transferred to the pLX-311 expression vector (Thermo-Fisher) utilizing LR Clonase II gateway technology for gene expression. A flag tag was appended to the 3′-end of UCHL5. The same procedure was applied to mouse Col17a1. Lentiviral infection was employed to deliver the expression vectors to MOC1-esc1 cells, which were subsequently harvested for western analyses.

The PCR primers used to as follow:

mUchl5-F, 5′-ggggacaagtttgtacaaaaaagcaggcttcgccaccatgtcgagcaatgccggggagtg-3′

mUchl5-R, 5′-ggggaccactttgtacaagaaagctgggtctcacttatcgtcgtcatccttgta-3′

mCol17a1-F, 5’-ggggacaagtttgtacaaaaaagcaggcttcgccaccatggatgtgaccaagaaaagcaa-3’

mCol17a1-R,5’-ggggaccactttgtacaagaaagctgggtcttacggcttgatggcaatacttc-3’.

Statistics & reproducibility

Statistical analyses were performed with GraphPad Prism 10.2.3 software, with statistical significance set at p < 0.05. Two-way ANOVA was used for multiple comparisons in tumor growth delay experiments, while mouse survival analyses were conducted using log-rank tests. For experiments involving comparisons between two groups, unpaired two-sided Student’s t-tests were employed. Furthermore, fisher’s exact test was utilized to assess the association between stromal desmoplasia and the response to anti-PD1 treatment. All samples that have received the proper procedures with confidence were included for analysis. Animals and cells were randomized before treatments. n values in the figure legends represent the number of biologically independent samples. Experiments were repeated independently two to four times with similar results, and representative images are displayed. Immunohistochemistry staining was performed on 4-7 independent samples, and representative images are shown.

Ethics

All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of Broad Institute, under protocol number 0110-08-16-1, and carried out in accordance with institutional and national guidelines. Animal studies using 4MOSC1 tumor model were conducted at Moores Cancer Center, San Diego under protocol ASP #S15195, approved by Institutional Animal Care and Use Committee (IACUC) of University of California. No human subjects or patient-derived samples were involved in this study. Both male and female mice were used in experiments unless otherwise noted.

Reporting summary

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

Data availability

All genomic sequencing data that support the findings of this study have been deposited in the Gene Expression Omnibus database with the accession code GSE269446: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE269446 and GSE287094: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE287094. The previously published human single-cell RNA-sequencing dataset of HNSCC patients is available through the Gene Expression Omnibus with accession number GSE103322: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE103322. Source data are provided with this paper.

References

Ferlay, J. et al. Estimating the global cancer incidence and mortality in 2018: GLOBOCAN sources and methods. Int. J. Cancer 144, 1941–1953 (2019).
Article CAS PubMed Google Scholar
Johnson, D. E. et al. Head and neck squamous cell carcinoma. Nat. Rev. Dis. Prim. 6, 92 (2020).
Article PubMed Google Scholar
Cramer, J. D., Burtness, B. & Ferris, R. L. Immunotherapy for head and neck cancer: Recent advances and future directions. Oral. Oncol. 99, 104460 (2019).
Article CAS PubMed PubMed Central Google Scholar
Zhang, X., Zeng, X., Li, Y. & Ju, D. Combinational Immunotherapy of Cancer: Novel Targets, Mechanisms, and Strategies. (Frontiers Media SA, 2024).
Villanueva, L., Álvarez-Errico, D. & Esteller, M. The contribution of epigenetics to cancer immunotherapy. Trends Immunol. 41, 676–691 (2020).
Article CAS PubMed Google Scholar
Yang, J. et al. Epigenetic regulation in the tumor microenvironment: molecular mechanisms and therapeutic targets. Signal Transduct. Target Ther. 8, 210 (2023).
Article ADS PubMed PubMed Central Google Scholar
Zhou, L., Xu, N., Shibata, H., Saloura, V. & Uppaluri, R. Epigenetic modulation of immunotherapy and implications in head and neck cancer. Cancer Metastasis Rev. 40, 141–152 (2021).
Article PubMed PubMed Central Google Scholar
Zhou, L., Mudianto, T., Ma, X., Riley, R. & Uppaluri, R. Targeting EZH2 enhances antigen presentation, antitumor immunity, and circumvents Anti-PD-1 resistance in head and neck cancer. Clin. Cancer Res. 26, 290–300 (2020).
Article CAS PubMed Google Scholar
Griffin, G. K. et al. Epigenetic silencing by SETDB1 suppresses tumour intrinsic immunogenicity. Nature 595, 309–314 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Olino, K., Park, T. & Ahuja, N. Exposing Hidden Targets: Combining epigenetic and immunotherapy to overcome cancer resistance. Semin. Cancer Biol. 65, 114–122 (2020).
Article CAS PubMed Google Scholar
Rodriguez, C. P. et al. A phase II trial of pembrolizumab and vorinostat in recurrent metastatic head and neck squamous cell carcinomas and salivary gland cancer. Clin. Cancer Res. 26, 837–845 (2020).
Article CAS PubMed Google Scholar
Judd, N. P. et al. ERK1/2 regulation of CD44 modulates oral cancer aggressiveness. Cancer Res. 72, 365–374 (2012).
Article CAS PubMed Google Scholar
Onken, M. D. et al. A surprising cross-species conservation in the genomic landscape of mouse and human oral cancer identifies a transcriptional signature predicting metastatic disease. Clin. Cancer Res. 20, 2873–2884 (2014).
Article CAS PubMed PubMed Central Google Scholar
Zhou, L. et al. Checkpoint blockade-induced CD8+ T cell differentiation in head and neck cancer responders. J. Immunother. Cancer 10, e004034 (2022).
Dubrot, J. et al. In vivo screens using a selective CRISPR antigen removal lentiviral vector system reveal immune dependencies in renal cell carcinoma. Immunity 54, 571–585.e6 (2021).
Article CAS PubMed Google Scholar
Manguso, R. T. et al. In vivo CRISPR screening identifies Ptpn2 as a cancer immunotherapy target. Nature 547, 413–418 (2017).
Article CAS PubMed PubMed Central Google Scholar
Dubrot, J. et al. In vivo CRISPR screens reveal the landscape of immune evasion pathways across cancer. Nat. Immunol. 23, 1495–1506 (2022).
Article CAS PubMed Google Scholar
Ishizuka, J. J. et al. Loss of ADAR1 in tumours overcomes resistance to immune checkpoint blockade. Nature 565, 43–48 (2019).
Article ADS CAS PubMed Google Scholar
Li, F. et al. Epigenetic CRISPR Screen Identifies as an Immunotherapeutic Target in -Mutant Lung Adenocarcinoma. Cancer Discov. 10, 270–287 (2020).
Article PubMed Google Scholar
Vander Linden, R. T. et al. Structural basis for the activation and inhibition of the UCH37 deubiquitylase. Mol. Cell 57, 901–911 (2015).
Article Google Scholar
Nishi, R. et al. Systematic characterization of deubiquitylating enzymes for roles in maintaining genome integrity. Nat. Cell Biol. 16, 1016–1026 (2014).
Article CAS PubMed PubMed Central Google Scholar
Fang, Y. & Shen, X. Ubiquitin carboxyl-terminal hydrolases: involvement in cancer progression and clinical implications. Cancer Metastasis Rev. 36, 669–682 (2017).
Article CAS PubMed Google Scholar
International Agency for Research on Cancer. Who Classification of Tumours of the Lung, Pleura, Thymus and Heart. (World Health Organization, 2015).
Dewson, G., Eichhorn, P. J. A. & Komander, D. Deubiquitinases in cancer. Nat. Rev. Cancer 23, 842–862 (2023).
Article CAS PubMed Google Scholar
Du, J. et al. A cryptic K48 ubiquitin chain binding site on UCH37 is required for its role in proteasomal degradation. Elife 11, e76100 (2022).
Article CAS PubMed PubMed Central Google Scholar
Liu, D., Song, Z., Wang, X. & Ouyang, L. Ubiquitin C-terminal hydrolase L5 (UCHL5) accelerates the growth of endometrial cancer via activating the Wnt/β-catenin signaling pathway. Front. Oncol. 10, 865 (2020).
Article PubMed PubMed Central Google Scholar
Wicks, S. J. et al. The deubiquitinating enzyme UCH37 interacts with Smads and regulates TGF-beta signalling. Oncogene 24, 8080–8084 (2005).
Article CAS PubMed Google Scholar
Han, W., Lee, H. & Han, J.-K. Ubiquitin C-terminal hydrolase37 regulates Tcf7 DNA binding for the activation of Wnt signalling. Sci. Rep. 7, 42590 (2017).
Article ADS CAS PubMed PubMed Central Google Scholar
Randles, L., Anchoori, R. K., Roden, R. B. S. & Walters, K. J. The proteasome ubiquitin receptor hrpn13 and its interacting deubiquitinating enzyme uch37 are required for proper cell cycle progression. J. Biol. Chem. 291, 8773–8783 (2016).
Article CAS PubMed PubMed Central Google Scholar
Mazumdar, T. et al. Regulation of NF-kappaB activity and inducible nitric oxide synthase by regulatory particle non-ATPase subunit 13 (Rpn13). Proc. Natl Acad. Sci. USA 107, 13854–13859 (2010).
Article ADS CAS PubMed PubMed Central Google Scholar
Zhang, J. et al. Deubiquitinase UCHL5 is elevated and associated with a poor clinical outcome in lung adenocarcinoma (LUAD). J. Cancer 11, 6675–6685 (2020).
Article CAS PubMed PubMed Central Google Scholar
Chen, Y. et al. Expression and clinical significance of UCH37 in human esophageal squamous cell carcinoma. Dig. Dis. Sci. 57, 2310–2317 (2012).
Article CAS PubMed Google Scholar
Chen, X. et al. Metal-based proteasomal deubiquitinase inhibitors as potential anticancer agents. Cancer Metastasis Rev. 36, 655–668 (2017).
Article CAS PubMed PubMed Central Google Scholar
Liu, N. et al. A novel proteasome inhibitor suppresses tumor growth via targeting both 19S proteasome deubiquitinases and 20S proteolytic peptidases. Sci. Rep. 4, 5240 (2014).
Article ADS CAS PubMed PubMed Central Google Scholar
D’Arcy, P. et al. Inhibition of proteasome deubiquitinating activity as a new cancer therapy. Nat. Med. 17, 1636–1640 (2011).
Article PubMed Google Scholar
Chow, P.-M. et al. The UCHL5 inhibitor b-AP15 overcomes cisplatin resistance via suppression of cancer stemness in urothelial carcinoma. Mol. Ther. Oncolytics 26, 387–398 (2022).
Article CAS PubMed PubMed Central Google Scholar
Sha, B. et al. Deubiquitylatinase inhibitor b-AP15 induces c-Myc-Noxa-mediated apoptosis in esophageal squamous cell carcinoma. Apoptosis 24, 826–836 (2019).
Article PubMed Google Scholar
Wang, Z. et al. Syngeneic animal models of tobacco-associated oral cancer reveal the activity of in situ anti-CTLA-4. Nat. Commun. 10, 5546 (2019).
Article ADS CAS PubMed PubMed Central Google Scholar
Zolkind, P. et al. Cancer immunogenomic approach to neoantigen discovery in a checkpoint blockade responsive murine model of oral cavity squamous cell carcinoma. Oncotarget 9, 4109–4119 (2018).
Article PubMed Google Scholar
Chandramohan, V. et al. Improved efficacy against malignant brain tumors with EGFRwt/EGFRvIII targeting immunotoxin and checkpoint inhibitor combinations. J. Immunother. Cancer 7, 142 (2019).
Article PubMed PubMed Central Google Scholar
Naba, A. Ten years of extracellular matrix proteomics: accomplishments, challenges, and future perspectives. Mol. Cell. Proteom. 22, 100528 (2023).
Article CAS Google Scholar
Naba, A. et al. The extracellular matrix: Tools and insights for the ‘omics’ era. Matrix Biol. 49, 10–24 (2016).
Article CAS PubMed Google Scholar
Puram, S. V. et al. Single-cell transcriptomic analysis of primary and metastatic tumor ecosystems in head and neck cancer. Cell 171, 1611–1624.e24 (2017).
Article CAS PubMed PubMed Central Google Scholar
Lamouille, S., Xu, J. & Derynck, R. Molecular mechanisms of epithelial-mesenchymal transition. Nat. Rev. Mol. Cell Biol. 15, 178–196 (2014).
Article CAS PubMed PubMed Central Google Scholar
Pickup, M. W., Mouw, J. K. & Weaver, V. M. The extracellular matrix modulates the hallmarks of cancer. EMBO Rep. 15, 1243–1253 (2014).
Article CAS PubMed PubMed Central Google Scholar
Muncie, J. M. & Weaver, V. M. The Physical and Biochemical Properties of the Extracellular Matrix Regulate Cell Fate. Curr. Top. Dev. Biol. 130, 1–37 (2018).
Article CAS PubMed PubMed Central Google Scholar
Martins Cavaco, A. C., Dâmaso, S., Casimiro, S. & Costa, L. Collagen biology making inroads into prognosis and treatment of cancer progression and metastasis. Cancer Metastasis Rev. 39, 603–623 (2020).
Article CAS PubMed Google Scholar
Gkretsi, V. & Stylianopoulos, T. Cell Adhesion and matrix stiffness: coordinating cancer cell invasion and metastasis. Front. Oncol. 8, 145 (2018).
Article PubMed PubMed Central Google Scholar
Oliveira, G. et al. Preexisting tumor-resident T cells with cytotoxic potential associate with response to neoadjuvant anti-PD-1 in head and neck cancer. Sci. Immunol. 8, eadf4968 (2023).
Article CAS PubMed PubMed Central Google Scholar
Uppaluri, R. et al. Neoadjuvant and adjuvant pembrolizumab in resectable locally advanced, human papillomavirus-unrelated head and neck cancer: a multicenter, phase II Trial. Clin. Cancer Res. 26, 5140–5152 (2020).
Article CAS PubMed PubMed Central Google Scholar
Saint, A. & Van Obberghen-Schilling, E. The role of the tumor matrix environment in progression of head and neck cancer. Curr. Opin. Oncol. 33, 168–174 (2021).
Article CAS PubMed Google Scholar
Yuan, Z. et al. Extracellular matrix remodeling in tumor progression and immune escape: from mechanisms to treatments. Mol. Cancer 22, 48 (2023).
Article CAS PubMed PubMed Central Google Scholar
Tuusa, J., Kokkonen, N. & Tasanen, K. BP180/Collagen XVII: A Molecular View. Int. J. Mol. Sci. 22, 12233 (2021).
Article CAS PubMed PubMed Central Google Scholar
Stelkovics, E. et al. Collagen XVII/BP180 protein expression in squamous cell carcinoma of the skin detected with novel monoclonal antibodies in archived tissues using tissue microarrays and digital microscopy. Appl. Immunohistochem. Mol. Morphol. 16, 433–441 (2008).
Article CAS PubMed Google Scholar
Meng, X. et al. BP180 Is a Prognostic Factor in Head and Neck Squamous Cell Carcinoma. Anticancer Res. 41, 1089–1099 (2021).
Article CAS PubMed Google Scholar
Deol, K. K. et al. Proteasome-Bound UCH37/UCHL5 Debranches Ubiquitin Chains to Promote Degradation. Mol. Cell 80, 796–809.e9 (2020).
Article CAS PubMed PubMed Central Google Scholar
Peng, Q. et al. The biological function of metazoan-specific subunit nuclear factor related to kappaB binding protein of INO80 complex. Int J. Biol. Macromol. 203, 176–183 (2022).
Article CAS PubMed Google Scholar
Kumar, A. et al. Structure of a novel winged-helix like domain from human NFRKB protein. PLoS One 7, e43761 (2012).
Article ADS CAS PubMed PubMed Central Google Scholar
Sanchez-Pulido, L., Kong, L. & Ponting, C. P. A common ancestry for BAP1 and Uch37 regulators. Bioinformatics 28, 1953–1956 (2012).
Article CAS PubMed Google Scholar
Lorusso, G., Rüegg, C. & Kuonen, F. Targeting the extra-cellular matrix-tumor cell crosstalk for anti-cancer therapy: emerging alternatives to integrin inhibitors. Front. Oncol. 10, 1231 (2020).
Article PubMed PubMed Central Google Scholar
Deb, G., Cicala, A., Papadas, A. & Asimakopoulos, F. Matrix proteoglycans in tumor inflammation and immunity. Am. J. Physiol. Cell Physiol. 323, C678–C693 (2022).
Article CAS PubMed PubMed Central Google Scholar
Li, H.-X. et al. Expression of αvβ6 integrin and collagen fibre in oral squamous cell carcinoma: association with clinical outcomes and prognostic implications. J. Oral. Pathol. Med. 42, 547–556 (2013).
Article ADS CAS PubMed Google Scholar
Wang, J. et al. Collagen-related gene expression level predicts the prognosis and immune therapy response. Gastric Cancer 26, 891–903 (2023).
Article CAS PubMed Google Scholar
Esbona, K. et al. The presence of cyclooxygenase 2, tumor-associated macrophages, and collagen alignment as prognostic markers for invasive breast carcinoma patients. Am. J. Pathol. 188, 559–573 (2018).
Article CAS PubMed PubMed Central Google Scholar
Rømer, A. M. A., Thorseth, M.-L. & Madsen, D. H. Immune Modulatory Properties of Collagen in Cancer. Front. Immunol. 12, 791453 (2021).
Article PubMed PubMed Central Google Scholar
Flies, D. B., Langermann, S., Jensen, C., Karsdal, M. A. & Willumsen, N. Regulation of tumor immunity and immunotherapy by the tumor collagen extracellular matrix. Front. Immunol. 14, 1199513 (2023).
Article CAS PubMed PubMed Central Google Scholar
Tripathi, S. et al. Cancer-associated fibroblast-secreted collagen is associated with immune inhibitor receptor LAIR1 in gliomas. J. Clin. Invest. 134, e176613 (2024).
Peng, D. H. et al. Collagen promotes anti-PD-1/PD-L1 resistance in cancer through LAIR1-dependent CD8 T cell exhaustion. Nat. Commun. 11, 4520 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Lai, S. L. et al. Collagen induces a more proliferative, migratory and chemoresistant phenotype in head and neck cancer via DDR1. Cancers 11, 1766 (2019).
Article CAS PubMed PubMed Central Google Scholar
Kozawa, K. et al. The CD44/COL17A1 pathway promotes the formation of multilayered, transformed epithelia. Curr. Biol. 31, 3086–3097.e7 (2021).
Article CAS PubMed Google Scholar
Natsuga, K., Watanabe, M., Nishie, W. & Shimizu, H. Life before and beyond blistering: The role of collagen XVII in epidermal physiology. Exp. Dermatol. 28, 1135–1141 (2019).
Article CAS PubMed Google Scholar
Kashiwagi, R. et al. Collagen XVII regulates tumor growth in pancreatic cancer through interaction with the tumor microenvironment. Cancer Sci. 114, 4286–4298 (2023).
Article CAS PubMed PubMed Central Google Scholar
Matte, B. F. et al. Matrix stiffness mechanically conditions EMT and migratory behavior of oral squamous cell carcinoma. J. Cell Sci. 132, jcs224360 (2019).
PubMed PubMed Central Google Scholar
Xu, S. et al. The role of collagen in cancer: from bench to bedside. J. Transl. Med. 17, 309 (2019).
Article PubMed PubMed Central Google Scholar
Aggarwal, A. et al. Computational pathology identifies immune-mediated collagen disruption to predict clinical outcomes in gynecologic malignancies. Commun. Med (Lond.) 4, 2 (2024).
Article CAS PubMed Google Scholar
Naba, A. et al. The matrisome: in silico definition and in vivo characterization by proteomics of normal and tumor extracellular matrices. Mol. Cell. Proteom. 11, M111.014647 (2012).
Article Google Scholar
Socovich, A. M. & Naba, A. The cancer matrisome: From comprehensive characterization to biomarker discovery. Semin. Cell Dev. Biol. 89, 157–166 (2019).
Article CAS PubMed Google Scholar
Popova, N. V. & Jücker, M. The functional role of extracellular matrix proteins in cancer. Cancers 14, 238 (2022).
LeBleu, V. S., Macdonald, B. & Kalluri, R. Structure and function of basement membranes. Exp. Biol. Med. 232, 1121–1129 (2007).
Article CAS Google Scholar
Gerarduzzi, C., Hartmann, U., Leask, A. & Drobetsky, E. The matrix revolution: matricellular proteins and restructuring of the cancer microenvironment. Cancer Res. 80, 2705–2717 (2020).
Article CAS PubMed Google Scholar
Wight, T. N. et al. Versican-a critical extracellular matrix regulator of immunity and inflammation. Front. Immunol. 11, 512 (2020).
Article CAS PubMed PubMed Central Google Scholar
Bollyky, P. L. et al. Cutting edge: high molecular weight hyaluronan promotes the suppressive effects of CD4+CD25+ regulatory T cells. J. Immunol. 179, 744–747 (2007).
Article CAS PubMed Google Scholar
Ren, B. et al. Tumor microenvironment participates in metastasis of pancreatic cancer. Mol. Cancer 17, 108 (2018).
Article PubMed PubMed Central Google Scholar
Nandi, T., Pradyuth, S., Singh, A. K., Chitkara, D. & Mittal, A. Therapeutic agents for targeting desmoplasia: current status and emerging trends. Drug Discov. Today. 25, 2046–2055 (2020).
Cho, C. et al. Cancer-associated fibroblasts downregulate type I interferon receptor to stimulate intratumoral stromagenesis. Oncogene 39, 6129–6137 (2020).
Article CAS PubMed PubMed Central Google Scholar
Almangush, A. et al. Staging and grading of oral squamous cell carcinoma: An update. Oral. Oncol. 107, 104799 (2020).
Article PubMed Google Scholar
Hallmann, R. et al. The regulation of immune cell trafficking by the extracellular matrix. Curr. Opin. Cell Biol. 36, 54–61 (2015).
Article CAS PubMed Google Scholar
Melssen, M. M., Sheybani, N. D., Leick, K. M. & Slingluff, C. L. Jr. Barriers to immune cell infiltration in tumors. J. Immunother. Cancer 11, e006401 (2023).
Article PubMed PubMed Central Google Scholar
Nicolas-Boluda, A. et al. Tumor stiffening reversion through collagen crosslinking inhibition improves T cell migration and anti-PD-1 treatment. Elife 10, e58688 (2021).
Tabdanov, E. D. et al. Engineering T cells to enhance 3D migration through structurally and mechanically complex tumor microenvironments. Nat. Commun. 12, 2815 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Pruitt, H. C. et al. Collagen fiber structure guides 3D motility of cytotoxic T lymphocytes. Matrix Biol. 85-86, 147–159 (2020).
Article CAS PubMed Google Scholar
Kasperska, A. et al. Current challenges in targeting tumor desmoplasia to improve the efficacy of immunotherapy. Curr. Cancer Drug Targets 21, 919–931 (2021).
Article CAS PubMed Google Scholar
Killock, D. Immunotherapy: Desmoplasia is no barrier to PD-1 blockade in melanoma. Nat. Rev. Clin. Oncol. 15, 200–201 (2018).
PubMed Google Scholar
Mehra, S. et al. Remodeling of stromal immune microenvironment by urolithin a improves survival with immune checkpoint blockade in pancreatic cancer. Cancer Res Commun. 3, 1224–1236 (2023).
Article CAS PubMed PubMed Central Google Scholar
Poh, A. R. et al. Inhibition of HCK in myeloid cells restricts pancreatic tumor growth and metastasis. Cell Rep. 41, 111479 (2022).
Article CAS PubMed PubMed Central Google Scholar
Prakash, J. & Shaked, Y. The interplay between extracellular matrix remodeling and cancer therapeutics. Cancer Discov. 14, 1375–1388 (2024).
Article CAS PubMed PubMed Central Google Scholar
Gonzalez-Molina, J. et al. Chemotherapy as a regulator of extracellular matrix-cell communication: Implications in therapy resistance. Semin Cancer Biol. 86, 224–236 (2022).
Article CAS PubMed Google Scholar
Barker, H. E., Paget, J. T. E., Khan, A. A. & Harrington, K. J. The tumour microenvironment after radiotherapy: mechanisms of resistance and recurrence. Nat. Rev. Cancer 15, 409–425 (2015).
Article CAS PubMed PubMed Central Google Scholar
Nguyen, D. H. et al. Radiation acts on the microenvironment to affect breast carcinogenesis by distinct mechanisms that decrease cancer latency and affect tumor type. Cancer Cell 19, 640–651 (2011).
Article CAS PubMed PubMed Central Google Scholar
Koukourakis, M. I. et al. Endogenous markers of two separate hypoxia response pathways (hypoxia inducible factor 2 alpha and carbonic anhydrase 9) are associated with radiotherapy failure in head and neck cancer patients recruited in the CHART randomized trial. J. Clin. Oncol. 24, 727–735 (2006).
Article CAS PubMed Google Scholar
Wang, Y. et al. Cancer-associated Fibroblasts Promote Irradiated Cancer Cell Recovery Through Autophagy. EBioMedicine 17, 45–56 (2017).
Article CAS PubMed PubMed Central Google Scholar
Melissaridou, S. et al. The effect of 2D and 3D cell cultures on treatment response, EMT profile and stem cell features in head and neck cancer. Cancer Cell Int 19, 16 (2019).
Article PubMed PubMed Central Google Scholar
Yamagata, A., Yokoyama, T., Fukuda, Y. & Ishida, T. Impact of interstitial lung disease associated with immune checkpoint inhibitors on prognosis in patients with non-small-cell lung cancer. Cancer Chemother. Pharm. 87, 251–258 (2021).
Article CAS Google Scholar
Doench, J. G. et al. Rational design of highly active sgRNAs for CRISPR-Cas9-mediated gene inactivation. Nat. Biotechnol. 32, 1262–1267 (2014).
Article CAS PubMed PubMed Central Google Scholar
Liberzon, A. et al. The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst. 1, 417–425 (2015).
Article CAS PubMed PubMed Central Google Scholar
Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl Acad. Sci. USA 102, 15545–15550 (2005).
Article ADS CAS PubMed PubMed Central Google Scholar
Gavish, A. et al. Hallmarks of transcriptional intratumour heterogeneity across a thousand tumours. Nature 618, 598–606 (2023).
Article ADS CAS PubMed Google Scholar
Yates, K. (2025) https://BioRender.com/h4afkwv.
Yates, K. (2025) https://BioRender.com/x518cbo.
Yates, K. (2025) https://BioRender.com/qlx09zl.
Yates, K. (2025) https://BioRender.com/071x1qm.
Yates, K. (2025) https://BioRender.com/cc3i4xm.
Yates, K. (2025). https://BioRender.com/u6hk4f0.

Download references

Acknowledgements

The authors would like to thank all members of Manguso lab at MGH and the Broad Institute, Uppaluri lab at the Dana-Farber Cancer Institute. We thank the Genetic Perturbation Platform (GPP) at the Broad Institute for CRISPR KO sgRNA library design and preparation. We thank Patricia Rogers and the staff of the Broad Flow Cytometry core for assistance with cell analysis and sorting. We would also like to thank Dr. Meng-Ju Wu of MGH for the discussion on the project. Finally, we thank Jacquelyn Stathopoulos, Dr. Tyler Caron, and the staff of the Broad Institute Comparative Medicine program for assistance with animal studies. Schematics were generated in Adobe Illustrator with the support of Biorender (https://biorender.com). This work was supported by Cancer Moonshot grant from the National Cancer Institute (NCI) and the National Institute of Dental and Craniofacial Research (NIDCR) U01DE029188.

Author information

Authors and Affiliations

Department of Medical Oncology, Dana-Farber Cancer Institute, Harvard Medical School, Massachusetts, USA
Cong Fu, Jordan M. Chinai, Payal Tiwari, Ann Marie Egloff & Ravindra Uppaluri
Broad Institute of MIT and Harvard, Cambridge, MA, USA
Cong Fu, Jordan M. Chinai, Sarah Y. Kim, Ashwin V. Kammula, Jonathan J. Perera, Aiping Jiang, Payal Tiwari, Erin N. Kistler, Kayla J. Colvin, Juan Dubrot, Seth Anderson, Rachel A. Fetterman, Cun Lan Chuong, Sarah K. Lane-Reticker, Collins K. Cheruiyot, Audrey J. Muscato, Gabriel K. Griffin, Bradley E. Bernstein, Kathleen B. Yates & Robert T. Manguso
Center for Cancer Research, Massachusetts General Hospital, Boston, MA, USA
Cong Fu, Jordan M. Chinai, Sarah Y. Kim, Ashwin V. Kammula, Jonathan J. Perera, Aiping Jiang, Payal Tiwari, Erin N. Kistler, Kayla J. Colvin, Juan Dubrot, Seth Anderson, Rachel A. Fetterman, Cun Lan Chuong, Sarah K. Lane-Reticker, Collins K. Cheruiyot, Audrey J. Muscato, Kathleen B. Yates & Robert T. Manguso
Department of Otolaryngology-Head and Neck Surgery, University of California San Diego, La Jolla, CA, USA
Robert Saddawi-Konefka & Shiqi Tang
Department of Head and Neck Surgery, The University of Texas MD Anderson Cancer Center, Houston, TX, USA
Robert Saddawi-Konefka
Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA
Ashwin V. Kammula, Seth Anderson & Zahra Alipour
Stanford University School of Medicine, Stanford, CA, USA
Jonathan J. Perera
Department of Bioengineering, Stanford University, Stanford, CA, USA
Shiqi Tang
Moores Cancer Center, University of California, San Diego, La Jolla, CA, USA
Sarah M. Luna & J. Silvio Gutkind
Division of Medical Sciences, Harvard Medical School, Boston, MA, 02115, USA
Kayla J. Colvin, Sarah K. Lane-Reticker, Collins K. Cheruiyot & Audrey J. Muscato
Solid Tumors Program, Cima Universidad de Navarra, Cancer Center Clínica Universidad de Navarra (CCUN), Pamplona, Spain
Juan Dubrot
UMass Chan Medical School, Worcester, MA, USA
Rachel A. Fetterman
Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri, USA
Zahra Alipour & Rebecca D. Chernock
Division of Medical Oncology, Washington University School of Medicine, Robert Ebert and Greg Stubblefield Head and Neck Tumor Center, Alvin J. Siteman Cancer Center, and Barnes Jewish Hospital, St. Louis, MO, USA
Douglas R. Adkins
Department of Pathology, Dana-Farber Cancer Institute, Harvard Medical School, Massachusetts, USA
Gabriel K. Griffin
Department of Cancer Biology, Dana-Farber Cancer Institute, Harvard Medical School, Massachusetts, USA
Bradley E. Bernstein
Department of Surgery, Department of Medical Oncology, Brigham and Women’s Hospital, Boston, MA, USA
Ann Marie Egloff & Ravindra Uppaluri

Authors

Cong Fu
View author publications
Search author on:PubMed Google Scholar
Robert Saddawi-Konefka
View author publications
Search author on:PubMed Google Scholar
Jordan M. Chinai
View author publications
Search author on:PubMed Google Scholar
Sarah Y. Kim
View author publications
Search author on:PubMed Google Scholar
Ashwin V. Kammula
View author publications
Search author on:PubMed Google Scholar
Jonathan J. Perera
View author publications
Search author on:PubMed Google Scholar
Aiping Jiang
View author publications
Search author on:PubMed Google Scholar
Payal Tiwari
View author publications
Search author on:PubMed Google Scholar
Erin N. Kistler
View author publications
Search author on:PubMed Google Scholar
Shiqi Tang
View author publications
Search author on:PubMed Google Scholar
Sarah M. Luna
View author publications
Search author on:PubMed Google Scholar
Kayla J. Colvin
View author publications
Search author on:PubMed Google Scholar
Juan Dubrot
View author publications
Search author on:PubMed Google Scholar
Seth Anderson
View author publications
Search author on:PubMed Google Scholar
Rachel A. Fetterman
View author publications
Search author on:PubMed Google Scholar
Cun Lan Chuong
View author publications
Search author on:PubMed Google Scholar
Sarah K. Lane-Reticker
View author publications
Search author on:PubMed Google Scholar
Collins K. Cheruiyot
View author publications
Search author on:PubMed Google Scholar
Audrey J. Muscato
View author publications
Search author on:PubMed Google Scholar
Zahra Alipour
View author publications
Search author on:PubMed Google Scholar
Douglas R. Adkins
View author publications
Search author on:PubMed Google Scholar
Gabriel K. Griffin
View author publications
Search author on:PubMed Google Scholar
Bradley E. Bernstein
View author publications
Search author on:PubMed Google Scholar
Ann Marie Egloff
View author publications
Search author on:PubMed Google Scholar
Kathleen B. Yates
View author publications
Search author on:PubMed Google Scholar
Rebecca D. Chernock
View author publications
Search author on:PubMed Google Scholar
J. Silvio Gutkind
View author publications
Search author on:PubMed Google Scholar
Ravindra Uppaluri
View author publications
Search author on:PubMed Google Scholar
Robert T. Manguso
View author publications
Search author on:PubMed Google Scholar

Contributions

Conceptualization: C.F., R.U. and R.T.M.; Format analysis: C.F., S.Y.K., A.V.K. and J.J.P.; Investigation: C.F., R.S.K., J.M.C., C.L.C., A.J., P.T., E.N.K., S.T., S.M.L., K.J.C., J.D., S.A., R.A.F., S.K.L.R., C.K.C., A.J.M.; Resources: G.K.G., B.E.B., Z.A., D.R.A., A.M.E., R.D.C., J.S.G., and R.U.; Supervision: R.U. and R.T.M.; Visualization: C.F., S.Y.K., K.B.Y., R.U. and R.T.M.; Writing - original draft: C.F.; Writing - review & editing: R.U. and R.T.M.

Corresponding authors

Correspondence to Ravindra Uppaluri or Robert T. Manguso.

Ethics declarations

Competing interests

R.T.M. receives research funding from Calico Life Sciences, LLC. R.T.M. has received speaking or consulting fees from Bristol Myers Squibb, Gilead Sciences, Kumquat Biosciences, Immunai Therapeutics, and BioNTech. R.U. reports grants and personal fees from Merck, personal fees from Regeneron, and Daichi-Sankyo. The MOC models developed by R.U. have been filed with the Washington University Office of Technology Management and are licensed for distribution by Kerafast. The remaining authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.

Additional information

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

Supplementary information

Supplementary Infomation

Description of Addtional Supplementary File

Supplementary Data 1

Supplementary Data 2

Supplementary Data 3

Supplementary Data 4

Supplementary Data 5

Supplementary Data 6

Supplementary Data 7

Supplementary Data 8

Reporting Summary

Transparent Peer Review file

Source data

Source Data

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

Fu, C., Saddawi-Konefka, R., Chinai, J.M. et al. In vivo CRISPR screening in head and neck cancer reveals Uchl5 as an immunotherapy target. Nat Commun 16, 8572 (2025). https://doi.org/10.1038/s41467-025-63592-y

Download citation

Received: 07 December 2024
Accepted: 25 August 2025
Published: 29 September 2025
DOI: https://doi.org/10.1038/s41467-025-63592-y