Caspase-8 Mutation Promotes HNSCC Development and Resistance to Anti

Abstract

Heterozygous mutations of the gene encoding caspase-8 protease occur in 10% of human head and neck squamous cell carcinomas (HNSCCs). Cell line studies indicate that these mutations block apoptosis induced by death ligands. However, the in vivo role of caspase-8 mutations in the development of HNSCC and their impact on response to immune checkpoint blockade have not been determined. We generated mice with heterozygous, epithelium-specific knock-in of a representative, HNSCC-associated caspase-8 mutation (D305G). The impact of the caspase-8 mutation was assessed following treatment with the carcinogen 4-nitroquinoline-1-oxide (4NQO) in drinking water. Treatment of the D305G caspase-8 mutant mice with 4NQO resulted in a greater number of tongue tumors per mouse and a higher percentage of advanced-stage invasive carcinomas than was observed in 4NQO-treated mice with wild-type caspase-8, and tumors from the mutant mice were more resistant to anti–PD-1. We also engineered the murine oral cancer cell line MOC1 for heterozygous expression of caspase-8 mutations. Tumors generated from these engineered cells in syngeneic, immunocompetent mice demonstrated reduced responsiveness to anti–PD-1, relative to tumors with wild-type caspase-8. Further, the caspase-8 mutant tumors displayed reduced intratumoral and splenic CD8⁺ T cells and impaired recruitment of monocytes and dendritic cells during PD-1 blockade. Collectively, these findings demonstrate that caspase-8 mutation promotes carcinogen-induced HNSCC development and resistance to anti–PD-1. Investigation of caspase-8 mutations as potential biomarkers of poor response to immunotherapy in patients with HNSCC is warranted.

Keywords

mice transgenic disease models animal immunotherapy programmed cell death 1 receptor dendritic cells T-Lymphoctes

Introduction

Head and neck squamous cell carcinomas (HNSCCs) are responsible for almost 5% of all deaths from cancer worldwide (Barsouk et al 2023; Bray et al 2024; Siegel et al 2024). Treatment of these cancers has historically relied on surgical resection, which can cause disfigurement and difficulties in speech and swallowing, and/or treatment with radiation and conventional chemotherapy drugs, which frequently leads to adverse toxicities (Chow 2020; Johnson et al 2020). The advent of immune checkpoint inhibitors (ICIs) has yielded improvement in clinical outcomes and a reduction in unwanted side effects. In particular, treatment with nivolumab or pembrolizumab, monoclonal antibodies targeting the immune checkpoint protein programmed cell death protein 1 (PD-1), can lead to potent and durable responses against HNSCC (Chow et al 2016; Ferris et al 2016; Seiwert et al 2016; Mehra et al 2018; Burtness et al 2019; Johnson et al 2020). However, only a minority of patients with HNSCC (~15%–20%) demonstrate clinical benefit from nivolumab or pembrolizumab. Efforts to rationally select patients who are likely to benefit from ICI therapies have been severely hampered by a lack of knowledge regarding potential biomarkers of response or resistance.

Characterization of the genomic landscape of HNSCC tumors has provided insight into possible mechanisms of tumor development and drug resistance (Cerami et al 2012; Gao et al 2013; Cancer Genome Atlas 2015). Notably, the gene encoding the intracellular cysteine protease caspase-8 is among the more frequently mutated genes in HNSCC tumors, with heterozygous mutations occurring in approximately 10% of human tumors (Agrawal et al 2011; Stransky et al 2011; Cancer Genome Atlas 2015). Mutations in the caspase-8 gene are scattered throughout the coding region, which encodes a proenzyme protease consisting of a small and large subunit.

Caspase-8 protease mediates apoptosis induced by ligand-activated cell surface death receptors, including the receptors for tumor necrosis factor (TNF), TNF-related apoptosis-inducing ligand (TRAIL), and Fas ligand (Ashkenazi and Dixit 1998; Leonard and Johnson 2018). The activation of caspase-8 following death receptor activation propagates the extrinsic apoptosis pathway, leading to cell death. We previously reported that exogenous overexpression of HNSCC-associated caspase-8 mutations in cell line models blocks apoptosis induction by TRAIL or Fas ligand, indicating that the mutant proteins have the capacity to act as dominant-negative proteins (Li et al 2014; Cui et al 2021). Since effector T cells induce apoptosis in target cells, in part, via death ligand activation of caspase-8 and the extrinsic apoptosis pathway, mutation of caspase-8 may provide tumors a means of immune evasion and resistance to ICIs.

In this report, we determined the impact of a heterozygous caspase-8 mutation on the development of HNSCC tumors in the context of an immunocompetent murine model. We further examined whether heterozygous expression of caspase-8 mutants would alter the sensitivity of HNSCC tumors to ICI treatment. We engineered wild-type, immunocompetent mice for epithelium-specific, heterozygous knock-in of an HNSCC-associated caspase-8 mutation. The mutant mice exhibited increased tumorigenicity relative to mice with wild-type (WT) caspase-8 following treatment with 4-nitroquinoline-1-oxide (4NQO), demonstrating a clear role for caspase-8 mutation in promoting carcinogen-induced HNSCC. We also found that HNSCC tumors expressing heterozygous caspase-8 mutations were resistant to the antitumor effects of anti–PD-1. These findings provide a first-time demonstration that HNSCC-associated caspase-8 mutations promote tumor development in vivo and confer resistance to ICI therapy.

Materials and Methods

Extensive experimental details are presented in the Appendix.

Results

Generation of Mice with Conditional, Tissue-Specific Expression of the D305G Caspase-8 Mutation

Caspase-8 gene mutations in human HNSCC occur as heterozygous mutations. To investigate the role of caspase-8 mutation in the development of HNSCC, we engineered immunocompetent mice for heterozygous, conditional, epithelium-specific expression of the D305G caspase-8 mutation (D303 in humans) under the control of the endogenous caspase-8 promoter (Fig. 1A, Appendix Fig. 1). In The Cancer Genome Atlas (TCGA) PanCancer Atlas Studies (Cerami et al 2012; Gao et al 2013), mutation at the D303 site has been observed in 2 different human HNSCC tumors (D303G and D303V) and in a lung squamous cell carcinoma tumor (D303V). We have previously shown that the D303G and D303V caspase-8 mutant proteins retain the ability to homodimerize following stimulation of cells with the death ligand TRAIL but fail to mediate TRAIL-induced apoptosis (Cui et al 2021).

Figure 1.

Generation of the caspase-8 D305G knock-in mice and tamoxifen treatment strategies. (A) (i) genomic organization of the mouse caspase-8 gene, including exon 8 encoding the D305 site and exon 9 encoding the stop codon. (ii) The caspase-8 D305 targeting construct. The 5′ homologous arm is a 2,526-bp DNA sequence representing a portion of intron 7, while the 3′ homologous arm contains exon 8 (wherein the D305G mutation has been introduced), intron 8, and exon 9 sequences up to and including the stop codon. Located 5′ to the 5′ arm are sequences encoding diphtheria toxin A (DTA), used for negative selection. Between the 5′ and 3′ homologous arms, loxP sites flank the internal cassettes. The first cassette begins with a splice acceptor (SA) sequence, followed by complementary DNA sequences encoding wild-type exons 8 and 9, including the stop codon, and then by rabbit beta-globin polyadenylation sequences (rBG pA). The second cassette includes a neomycin resistance gene flanked by self-deletion anchoring sequences. (iii) The genomic targeted allele following recombination to incorporate the targeting construct. Embryonic stem cells incorporating the construct were negatively selected with DTA and positively selected with G418. (iv) The targeted allele following self-excision of the neomycin resistance gene. (B) Depicted are the breeding strategy and tamoxifen treatment strategy used to generate epithelium-specific, heterozygous expression of the caspase-8 D305 mutant.

The targeting vector containing the D305G mutation is shown in Figure 1A. The D305 site is located in exon 8 of the murine caspase-8 gene and the stop codon in exon 9. The targeting vector contained a 5′ arm representing a portion of intron 7 (between exons 7 and 8). The 3′ arm of the targeting vector contained exon 8, into which the D305G mutation was introduced; intron 8 (between exons 8 and 9); and a portion of exon 9 containing the stop codon. Between the 5′ and 3′ arms, the construct contained a neomycin resistance gene flanked by self-deletion anchor (SDA) sequences, as well as the WT complementary DNA (cDNA) sequence encoding exons 8 and 9 (including the stop codon) flanked by loxP sites. The construct was designed so that in the absence of Cre, the construct would generate WT caspase-8, but following Cre-mediated excision of the WT cDNA sequence, the D305G mutant would be generated.

The targeting construct was introduced into C57BL/6 embryonic stem cells, and neomycin-resistant cells were injected into C57BL/6 blastocysts, followed by reimplantation into pseudo-pregnant female mice. The neomycin resistance gene was then removed by self-excision. Founder mice were genotyped for the presence of the desired targeting sequences, as described in the Materials and Methods, and Appendix Figure 1A, B. Founder mice that contained the targeting sequences were designated “conditional CASP8^D305G” founders and were bred to homozygosity (Appendix Fig. 1C, D).

Homozygous, conditional CASP8^D305G mice were then bred with heterozygous Krt5-CreERT2 mice (Fig. 1B, Appendix Figs. 2 and 3), in which CreERT2 expression is under the control of the keratin 5 promoter, enabling selective expression of tamoxifen-activated CreERT2 in epithelial tissues, including the oral epithelium. This breeding generated 2 lines of progeny mice (Fig. 1B). The first line of progeny mice lacked Krt5-CreERT2 sequences, were heterozygous for the conditional CASP8^D305G sequence, and were designated “no-Cre/conditional CASP8^D305G” (Fig. 1B, Appendix Figs. 2 and 3). Due to the lack of CreERT2 expression in these mice, treatment with tamoxifen did not result in excision of the loxP-flanked, WT caspase-8 exon 8/9 sequences and did not enable expression of mutant caspase-8 (Appendix Fig. 4A, C). Following tamoxifen treatment, these mice were referred to simply as “CASP8^WT” mice (Fig. 1B). The second line of progeny mice, which were heterozygous for Krt5-ERT2 and heterozygous for the conditional CASP8^D305G sequence, was designated “Cre/conditional CASP8^D305G.” Tamoxifen treatment of “Cre/conditional CASP8^D305G” led to excision of the WT caspase-8 exon 8/9 cDNA sequence and expression of the D305G caspase-8 mutant (Fig. 1B, Appendix Fig. 4B, C). Following tamoxifen treatment, these mice were referred to as “CASP8^D305G” mice (Fig. 1B).

The D305G Caspase-8 Mutation Promotes Carcinogen-Induced Oral Tumor Formation

To determine the impact of caspase-8 mutation on HNSCC tumor development, CASP8^WT and CASP8^D305G mice were treated with the chemical carcinogen 4NQO (100 μg/mL in drinking water), and the formation, size, and pathological classification of resulting tumors were assessed. The study was initiated following a 2-wk resting period after tamoxifen treatment, with 31 CASP8^WT mice and 24 CASP8^D305G mice. Mice were given 4NQO for 16 wk, then returned to normal drinking water and examined for the development of tongue tumors 14, 28, and 41 d after the cessation of 4NQO treatment (Fig. 2A, B). No differences in body weights were detected following 4NQO treatment in the CASP8^WT versus CASP8^D305G groups (Appendix Fig. 5). For CASP8^WT mice, the mean numbers of tongue tumors observed per mouse were 0.68, 0.81, and 1.3 on posttreatment days 14, 28, and 41, respectively (Fig. 2C, Appendix Table 1). In contrast, CASP8^D305G mice exhibited means of 1.2, 1.3, and 2.9 tongue tumors per mouse on posttreatment days 14, 28, and 41, respectively (Fig. 2C, Appendix Table 1). Similarly, the percentage of mice with tongue tumors was markedly higher at each time point in the CASP8^D305G group compared with the CASP8^WT group (Fig. 2D, Appendix Table 1). However, although the mean numbers of tumors/mouse and the percentage of mice with tumors were higher in CASP8^D305G mice, we did not observe a strong trend toward larger tumors in the CASP8^D305G mice. Comparison of the CASP8^D305G and CASP8^WT groups 41 d after the end of 4NQO treatment revealed statistically significant differences between the groups only for smaller tumor sizes (<1 mm, 1–2 mm) and not for larger tumors (2–3 mm, >3 mm) (Fig. 2E, Appendix Tables 1 and 2).

Figure 2.

The caspase-8 D305G mutation promotes 4-nitroquinoline-1-oxide (4NQO)–induced development and progression of oral tumors. (A) Experimental strategy for 4NQO treatment of CASP8^WT and CASP8^D305G mice. (B) Tongue tumor formation in 4NQO-treated mice. Shown are 2 large tumors arising on the tongue of a 4NQO-treated CASP8^D305G mouse. (C) Mean number of tumors per mouse 14, 28, or 41 d after cessation of 4NQO treatment. The total number of gross tumors observed in CASP8^WT or CASP8^D305G mice was divided by the number of mice in each group (31 and 24, respectively) to obtain the mean number of tumors per mouse. The graph depicts the mean ± standard error of the mean. Unpaired Student’s t test was used to compare the number of tumors per mouse between mutant and wild-type groups at days 14, 28, and 41, respectively. (D) Percentage of mice harboring gross tumors 14, 28, or 41 d after the end of 4NQO treatment is shown. The CASP8^WT and CASP8^D305G groups were compared by Fisher’s exact test. (E) Average number of tumors of differing sizes per mouse in the CASP8^WT group (31 mice) and the CASP8^D305G group (24 mice) 41 d after cessation of 4NQO treatment is shown. The plotted data represent the mean ± 95% confidence intervals. Data were analyzed using a generalized linear mixed model, and comparisons within different tumor sizes used ad hoc contrast. (F) Percentage of mice exhibiting tongue dysplasia or carcinoma in situ (CIS) 41 d after the end of 4NQO. (G) Percentage of mice with tongue invasive squamous cell carcinoma 41 d after the end of 4NQO is shown. Groups were compared using Fisher’s exact test. (H) Hematoxylin and eosin staining of a tongue invasive squamous cell carcinoma (SCC) from a 4NQO-treated CASP8^D305G mouse is shown. (I) Cleaved caspase-3 immunofluorescence was quantified in 4NQO-treated tongue tissues from CASP8^WT and CASP8^D305G mice. Lesions were grouped into dysplasia (including normal epithelium and mild-to-severe dysplasia) and SCC (CIS and invasive). Quantification was performed using a composite score generated by multiplying field count by signal score, and individual values are shown. Group sizes were as follows: wild-type (WT) dysplasia, n = 18; WT SCC, n = 8; mutant (MT) dysplasia, n = 8; and MT SCC, n = 13. WT and MT groups were compared within each histologic category using a 2-tailed Mann–Whitney test.

Caspase-8 Mutation Promotes Progression of Oral Tumors

We next performed hematoxylin and eosin (H&E) staining and pathology assessment of tumors harvested on day 41 after cessation of 4NQO treatment. At this time point, 80.7% of CASP8^WT mice and 100% of CASP8^D305G mice had observable tumors. H&E staining and pathology assessment were performed on the tongues from all mice, irrespective of the presence or absence of observable tumors. The percentage of mice with tongues exhibiting dysplasia (100% for both CASP8^WT and CASP8^D305G) or carcinoma in situ (58.1% for CASP8^WT and 58.3% for CASP8^D305G) did not differ between the CASP8^WT and CASP8^D305G groups (Fig. 2F, Appendix Table 3). By contrast, the percentage of mice with tongues exhibiting invasive squamous cell carcinoma was elevated in CASP8^D305G mice (29.2%) versus CASP8^WT mice (16.1%) (Fig. 2G, H, Appendix Table 3). Immunofluorescence staining of tumor tissues revealed reduced expression of cleaved caspase-3 in both dysplasia and squamous cell carcinoma from CASP8^D305G mice compared with CASP8^WT mice (Fig. 2I, Appendix Fig. 6). Taken together, our analyses of tongue tumor development and progression indicate that heterozygous mutation of caspase-8, as observed in human HNSCC, promotes carcinogen-induced oral tumor formation and progression in mice.

Caspase-8 Mutation Abrogates the Antitumor Activity of Anti–PD-1

The antitumor effects of T cells are mediated, in part, by activating caspase-8 in tumor cells. Therefore, we hypothesized that heterozygous expression of HNSCC-associated caspase-8 mutations in tumor cells would inhibit the antitumor activity of anti–PD-1. To test this, we used the MOC1 syngeneic tumor model. MOC1 cells were derived from a tongue tumor in carcinogen-treated C57BL/6 mice and grow well as tumors in these immunocompetent mice (Ku et al 2007; Judd et al 2012a, 2012b; Onken et al 2014). Blockade of PD-L1/PD-1 interactions has been reported to inhibit the growth of MOC1 tumors (Moore et al 2016a; Moore et al 2016b; Tran et al 2017). We first examined the impact of loss of caspase-8 on sensitivity to anti–PD-1 by comparing the response of tumors derived from parental (WT equivalent) MOC1 cells versus MOC1 caspase-8 knockout (MOC1 CASP8 KO) cells. Wild-type C57BL/6 mice were inoculated subcutaneously in the flank with parental MOC1 cells or MOC1 CASP8 KO cells. When tumors reached 100 mm³, mice in both groups were randomized to receive treatment with an isotype-matched control antibody or anti–PD-1 (0.2 mg/mouse). As shown in Figure 3A, treatment of tumor-bearing mice with anti–PD-1 resulted in significant growth inhibition of MOC1 parental tumors (P < 0.001) but not MOC1 CASP8 KO tumors, demonstrating the requirement for tumor expression of caspase-8 for the antitumor activity of anti–PD-1.

Figure 3.

Heterozygous expression of caspase-8 mutants abrogates MOC1 tumor sensitivity to anti–PD-1. (A) C57BL/6 mice were inoculated into the right flank with MOC1 parental cells (2 million cells/mouse, n = 16) or MOC1 CASP8 knockout (KO) cells (2 million cells/mouse, n = 16). When tumors reached 100 mm³, mice in each group were randomized: half received treatment (Tr) with anti–PD-1 (triangles; 0.2 mg/mouse) and half had isotype-matched control IgG (circles). Mice were treated every 3 d (arrows) for a total of 3 doses, and tumor dimensions were measured roughly every 3 d. Error bars represent standard error of the mean, and 2-way analysis of variance was used to compare tumor volumes on the final day in mice treated with anti–PD-1 versus isotype-matched IgG (P < 0.001 for MOC1 parental tumors, but nonsignificant [NS] for MOC1 CASP8 KO tumors). (B) MOC1 cells engineered for DOX-inducible expression of FLAG-tagged, murine versions of the D305G or S388* caspase-8 mutants (MOC1 CASP8^D305G or MOC1 CASP8^S388* cells, respectively) were treated in the absence or presence of DOX (1 μg/mL) for 20 h. Whole-cell lysates were prepared and subjected to immunoblot analysis with anti-FLAG or anti–caspase-8. Detection of GAPDH was used as a loading control. (C) MOC1 parental, MOC1 CASP8^D305G, or MOC1 CASP8^S388* cells (2 million) were injected into the flank of C57BL/6 mice (10 mice/group for parental and CASP8^D305G group; 5 mice/group for the CASP8^S388* group). Immediately following inoculation, mice were given drinking water containing DOX (0.5 mg/mL), which they continued to receive until the conclusion of the experiment. When tumors reached 100 mm³, mice in each group were randomized: half received anti–PD-1 (triangles; 0.2 mg/mouse), and half received isotype-matched control IgG (circles; 0.2 mg/mouse). Error bars represent standard error of the mean, and 2-way analysis of variance was used to compare tumor volumes in mice treated with anti–PD-1 versus isotype-matched IgG (P < 0.001 for MOC1 parental tumors, but nonsignificant for MOC1 CASP8^D305G and MOC1 CASP8^S388* tumors). (D) CASP8^WT and CASP8^D305G mice with 4-nitroquinoline-1-oxide (4NQO)–induced tongue tumors were treated with isotype IgG or anti–PD-1 once a tumor lesion reached 0.5 mm in length. At each time point, the volumes of all detectable tongue tumors within each mouse were summed to generate a single total tumor burden value per mouse. Longitudinal total tumor burden for each group is shown as mean ± standard error of the mean. Group sizes were as follows: CASP8^WT IgG, n = 11 mice (18 tumors); CASP8^WT anti–PD-1, n = 7 mice (16 tumors); CASP8^D305G IgG, n = 11 mice (19 tumors); CASP8^D305G anti–PD-1, n = 10 mice (18 tumors). P values were determined by linear mixed-effects modeling of longitudinal tumor burden, with group, time, and the group-by-time interaction as fixed effects and mouse as a random effect.

Because HNSCC-associated caspase-8 mutations are heterozygous in human tumors, we next examined the impact of heterozygous expression of caspase-8 mutants in MOC1 tumors on response to anti–PD-1. In addition to the D305G mutant, we also studied the effects of a second caspase-8 mutation, S388* (S386* in humans), observed in 2 different human HNSCC tumors in TCGA. MOC1 cells, which encode WT caspase-8, were engineered for doxycycline (DOX)–inducible expression of FLAG epitope-tagged D305G and S388* mutants, and DOX-induced expression of the mutant proteins was confirmed by immunoblotting with anti-FLAG (Fig. 3B). The cells were then inoculated into the flanks of C57BL/6 mice, and the mice were simultaneously given drinking water containing DOX (0.5 mg/mL) to induce expression of the mutant proteins. Treatment with isotype-matched IgG or anti–PD-1 was initiated when tumors reached 100 mm³. While MOC1 parental tumors again demonstrated significant growth inhibition in response to anti–PD-1, the MOC1 CASP8^D305G and MOC1 CASP8^S388* tumors failed to respond (Fig. 3C).

To exclude artifacts resulting from DOX treatment and the inducible system, we next generated MOC1 cells with constitutive, heterozygous expression of the caspase-8 D305G mutant using a T2A-mCherry cassette that drives separate expression of caspase-8 D305G and mCherry (Appendix Fig. 7A). Using the same treatment regimen, we observed that tumors constitutively expressing the caspase-8 D305G mutant were resistant to anti–PD-1, recapitulating our findings with the DOX-inducible system (Appendix Fig. 7B).

We next examined responsiveness to anti–PD-1 in 4NQO-induced tongue tumors in the CASP8^WT and CASP8^D305G mice. Tumor growth was quantified at each time point as total tumor burden per mouse. Consistent with our prior results, in this cohort, tumor burden developed earlier and was greater in CASP8^D305G mice than in CASP8^WT mice (Appendix Fig. 7C). Treatment with anti–PD-1 reduced tumor burden in CASP8^WT mice but not in CASP8^D305G mice (Fig. 3D). Collectively, these results demonstrate that tumors expressing HNSCC-associated caspase-8 mutations exhibit resistance to anti–PD-1 in immunocompetent mouse models.

Caspase-8 Mutation Modulates CD8⁺ T-Cell and Dendritic Cell Levels upon Anti–PD-1 Treatment

To further investigate how caspase-8 mutation may modulate response to anti–PD-1, we performed immunoprofiling. C57BL/6 mice were inoculated with MOC1 parental cells or MOC1 cells with constitutive, heterozygous expression of the caspase-8 D305G mutant. Fourteen days after inoculation, mice received a single dose of IgG or anti–PD-1 (0.2 mg/mouse), and tumors and spleens were harvested 36 h later for flow cytometric analysis (Fig. 4A).

Figure 4.

Caspase-8 mutation modulates CD8⁺ T-cell and dendritic cell (DC) levels after anti–PD-1. (A) Experimental scheme for immune profiling. Mice bearing parental MOC1 tumors or MOC1 tumors with heterozygous caspase-8 D305G were treated with isotype IgG (n = 5) or anti–PD-1 (n = 5); tissues were analyzed 36 h after a single dose. (B, C) Flow cytometric quantification of lymphocytic populations within CD45⁺ immune cells. Total T cells were gated as CD3⁺ NK1.1^– cells. Natural killer (NK) cells were gated as CD3^– NK1.1⁺ cells and NKT cells as CD3⁺ NK1.1⁺ cells. CD8⁺ T cells were then gated as CD8⁺ CD4^– cells within the CD3⁺ T-cell population. From the CD3⁺ CD8⁺ CD4^– gate, naive (CD62L⁺ CD44^–), central memory (cm, CD62L⁺ CD44⁺), and effector (eff, CD62L^– CD44⁺) CD8⁺ T-cell subsets were further defined. Data are shown for tumors (B) and spleens (C) from mice bearing D305G-mutant tumors. (D, E) Intratumoral monocyte/DC-lineage populations (monocytes, total DCs, and DC subsets) under the same conditions in MOC1 tumors expressing wild-type (WT) (parental, D) or mutant (D305G, E) caspase-8. Within CD45⁺ immune cells, monocytes were gated as CD11b⁺ Ly6C^high. Total DCs were gated as F4/80^– MHCII⁺ CD11c⁺. From the total DC gate, cDC1 (CD103⁺ CD11b^–) and cDC2 (CD103^– CD11b⁺) were further gated. Y-axes show absolute cell numbers per 1 million total tumor or spleen cells. Error bars indicate the standard error of the mean with individual mice overlaid. Two-tailed Student’s t tests were used to compare anti–PD-1 versus IgG within each genotype; exact P values are shown.

In mice bearing tumors derived from MOC1 parental cells, 1 dose of anti–PD-1 treatment did not significantly alter the intratumoral or splenic levels of total T cells (Appendix Fig. 8). By contrast, tumors with caspase-8 D305G exhibited a significant reduction in tumor-infiltrating and splenic total T cells following anti–PD-1 treatment, with a pronounced decrease in CD8⁺ T cells and their subsets (naive, central memory and effector CD8⁺ T cells), alongside reductions in natural killer (NK) and NKT cells (Fig. 4B, C). The levels of tumor-infiltrating CD4⁺ T cells were not impacted by anti–PD-1 treatment in both parental and mutant tumors (Appendix Fig. 9A), whereas splenic CD4⁺ T cells were significantly reduced in the caspase-8 mutant cohort, but not in the parental cohort, following anti–PD-1 treatment (Appendix Fig. 9B). B cells and Tregs were not significantly altered in tumors or spleens of both genotypes by anti–PD-1 versus IgG treatment (Appendix Fig. 10).

We further assessed the levels of myeloid lineage cells following treatment with anti–PD-1. In MOC1 parental tumors, anti–PD-1 increased intratumoral monocyte/dendritic cells but not granulocytes (Fig. 4D, Appendix Fig. 11A, B). Specifically, monocytes, total dendritic cells (DCs), and the conventional DC2 (cDC2, CD103^–/CD11b⁺) subset, but not the monocytic myeloid-derived suppressor cells (mMDSCs) or tumor-associated macrophages (TAMs), were significantly elevated (Fig. 4D, Appendix Fig. 11C), with no corresponding changes in spleen (Appendix Fig. 11D), consistent with prior reports of monocyte and DC recruitment to tumors following PD-1 blockade (Schetters et al 2020; Lee et al 2021). In contrast, in caspase-8 mutant tumors, anti–PD-1 significantly reduced intratumoral monocytes, total DCs, and both cDC1 (CD103⁺/CD11b^–) and cDC2 subsets relative to IgG, without changes in mMDSCs, TAMs, neutrophil, or granulocytic MDSCs (gMDSCs) (Fig. 4E, Appendix Fig. 11E). Most splenic myeloid populations were unchanged in mutant tumor-bearing mice, with only a slight increase in neutrophils observed following anti–PD-1 treatment (Appendix Fig. 11F). No significant differences in intratumoral M1/M2 ratio were observed across genotypes or treatments (Appendix Fig. 11G). Although the splenic M1/M2 ratio was increased after anti–PD-1 in both parental and caspase-8–mutant groups, these increases did not reach statistical significance (Appendix Fig. 11G). Collectively, these data suggest that tumor expression of the caspase-8 D305G mutant impairs the early intratumoral monocyte and DC response to PD-1 blockade and coincides with diminished CD8⁺ T-cell abundance locally and systemically.

Caspase-8 Mutation Disrupts Anti–PD-1–Induced CD8⁺ T-cell Redistribution and Chemokine Responses in Tumors

To further investigate how caspase-8 mutation affects the tissue distribution of immune cells after PD-1 blockade, we analyzed spleen, lymph node, and tumor samples from mice treated with anti–PD-1 or isotype IgG. PD-1 staining on CD8⁺ T cells was markedly reduced after anti–PD-1 treatment in all 3 tissues, consistent with receptor binding by the therapeutic antibody (Fig. 5A). In mutant tumor-bearing mice, anti–PD-1 induced a distinct tissue distribution of CD8⁺ T cells, characterized by reduced CD8⁺ T-cell abundance in tumors and spleens but increased CD8⁺ T-cell numbers in lymph nodes, a pattern not observed in parental tumors (Fig. 5B). Although anti–PD-1 increased the proportion of Ki-67⁺ CD8⁺ T cells within tumors in both parental and mutant groups, no increase was observed in spleens or lymph nodes (Fig. 5C), indicating that residual intratumoral CD8⁺ T cells in mutant tumors retain some proliferative responsiveness despite an overall reduction in tumor CD8⁺ T-cell abundance. IFN-γ increased only in splenic CD8⁺ T cells from mutant tumor-bearing mice after anti–PD-1 treatment, whereas TIM-3 decreased only in splenic CD8⁺ T cells from the mutant group (Appendix Fig. 12A, B). In contrast, LAG-3 and granzyme B were not significantly altered by genotype or treatment in any tissue examined (Appendix Fig. 12C). By comparison, monocyte and dendritic cell numbers were not significantly changed in spleens or lymph nodes after anti–PD-1 treatment (Appendix Fig. 12D, E). Anti–PD-1 increased cleaved caspase-3 in 3 of 5 tumors from parental mice but not in tumors from CASP8^D305G mice, suggesting reduced tumor-cell apoptosis after PD-1 blockade (Fig. 5D). By contrast, phospho-MLKL was not detected in either group under these conditions, indicating that robust necroptosis was not induced by PD-1 blockade.

Figure 5.

Caspase-8 mutation disrupts anti–PD-1–induced CD8+ T-cell redistribution, apoptotic engagement, and chemokine responses. (A–C) C57BL/6 mice bearing parental MOC1 tumors or MOC1 tumors with heterozygous caspase-8 D305G were treated with isotype IgG (n = 5) or anti–PD-1 (n = 5), and spleen, draining lymph node (LN), and tumor tissues were analyzed 36 h after a single dose. (A) Percentages of PD-1⁺ cells within CD8⁺ T cells in spleen, LN, and tumor. (B) Absolute numbers of CD8⁺ T cells per 1 million total cells in spleen, LN, and tumor. (C) Percentages of Ki-67⁺ cells within CD8⁺ T cells in spleen, LN, and tumor. CD8+ T cells were gated as CD3⁺ CD8⁺ cells within CD45⁺ immune cells. (D) Immunoblot analysis of cleaved caspase-3 and GAPDH in parental and CASP8^D305G tumors harvested 36 h after treatment with IgG or anti–PD-1. (E, F) Relative messenger RNA expression of CXCL9 (E) and CXCL10 (F) in parental and CASP8^D305G tumors after IgG or anti–PD-1 treatment (n = 5 per group), normalized to the corresponding IgG control. Error bars indicate the standard error of the mean, with individual mice overlaid. Two-tailed Student’s t tests were used to compare anti–PD-1 versus IgG within each genotype and tissue; exact P values are shown.

To determine whether chemokine responses might contribute to the anti–PD-1 resistance of mutant tumors, we measured tumor chemokine expression after treatment with IgG or anti–PD-1. In parental tumors, anti–PD-1 increased CXCL9 expression, whereas this induction was absent in caspase-8–mutant tumors (Fig. 5E). In addition, CXCL10 expression decreased in anti–PD-1–treated caspase-8–mutant tumors but not in parental tumors (Fig. 5F), while CCL2 and CCL5 remained unchanged (Appendix Fig. 13). Together, these findings suggest that caspase-8 mutation impairs anti–PD-1–induced local immune remodeling, characterized by blunted CXCL9/CXCL10 chemokine response, thereby limiting the transition of CD8⁺ T cells from lymph nodes to tumors.

Discussion

Genomic studies have revealed that the gene encoding the apoptotic protease caspase-8 is among the more frequently mutated genes in HNSCC. Mutations in the caspase-8 gene are dispersed throughout the coding region, with the lack of hotspot mutations suggesting that the caspase-8 protein may have tumor suppressor activity. Further, heterozygous expression of HNSCC-associated caspase-8 mutations has been shown to block, in a dominant-negative fashion, death ligand–induced apoptosis (Li et al 2014; Cui et al 2021). Our studies now reveal that heterozygous expression of a caspase-8 mutation under the control of the endogenous caspase-8 gene promoter stimulates HNSCC tumor development in vivo in immunocompetent mice, clearly implicating the importance of the mutations.

Our studies also provide important insight into a possible mechanism of resistance to ICI therapy in HNSCC. Because effector T cells exert their antitumor effects, in part, by activating cell death induced by death ligands, we examined whether heterozygous caspase-8 mutations would impact the in vivo response of HNSCC tumors to anti–PD-1. Our findings show that the antitumor activity of PD-1–targeting antibody is blocked when tumor cells express heterozygous caspase-8 mutations. This suggests the need to evaluate whether caspase-8 mutation is associated with failure to respond to immunotherapy in the subset of patients with HNSCC whose tumors harbor such mutations.

Unexpectedly, in mice bearing caspase-8 mutant tumors, anti–PD-1 treatment reduced CD8⁺ T-cell levels in both tumor and spleen, as well as diminished DC and monocyte levels in tumors. PD-1 blockade typically results in rapid TNF-α production from functionally restored T cells or other innate immune cells (Zhou et al 2010; Simon et al 2016; Howard et al 2021). Our immune profiling analysis in the draining lymph node suggests this phenotype aligns more with impaired local immune-cell recruitment within the mutant tumors than with a broad defect in immune activation. The absence of CXCL9 induction and reduction of CXCL10 in caspase-8 mutant tumors after anti–PD-1 treatment further supports this interpretation. In addition, the reduced expression of cleaved caspase-3 in caspase-8–mutant tumors likely reflects both reduced local immune-cell accumulation and reduced cell susceptibility to apoptosis. However, the precise mechanisms by which caspase-8 mutations modulate these chemokine responses after anti–PD-1 treatment and how this affects the redistribution of DCs and T cells require further investigation.

At present, there is no publicly available HNSCC dataset that integrates caspase-8 mutation status with clinical response to ICIs. Examination of TCGA HNSCC data reveals that CASP8-mutant tumors do not uniformly exhibit an immune-excluded phenotype, and in some HPV-negative HNSCC cases, mutation is associated with higher immune infiltration and stronger interferon-related signatures despite poorer clinical outcomes. This apparent difference from our mouse models likely reflects the biological heterogeneity of human HNSCC tumors, including heterogeneity of co-occurring genetic alterations, and underscores the need for future studies integrating tumor genetics with immunotherapy response in HNSCC clinical cohorts.

In summary, our preclinical results strongly support clinical investigation to determine whether the caspase-8 mutation is associated with failure to respond to ICI therapy in patients with HNSCC. Moreover, as pembrolizumab and/or nivolumab have been approved by the US Food and Drug Administration in several other cancers where caspase-8 mutations are prevalent (endometrial, stomach, cervical, colorectal, and bladder cancers and melanoma), the impact of this potential mechanism of resistance may be broad.

Author Contributions

Z. Cui, H. Wu, contributed to conception and design, data analysis, drafted and critically revised the manuscript; L.C. Woerner, S.R. Long, C.N. Peterson, N.K. VanLandingham, A. Nazarenko, M.-O. Kim, J.R. Grandis, D.E. Johnson, contributed to data analysis, critically revised the manuscript. All authors gave their final approval and agree to be accountable for all aspects of the work.

Supplemental Material

sj-docx-1-jdr-10.1177_00220345261457066 – Supplemental material for Caspase-8 Mutation Promotes HNSCC Development and Resistance to Anti–PD-1

Supplemental material, sj-docx-1-jdr-10.1177_00220345261457066 for Caspase-8 Mutation Promotes HNSCC Development and Resistance to Anti–PD-1 by Z. Cui, H. Wu, L.C. Woerner, S.R. Long, C.N. Peterson, N.K. VanLandingham, A. Nazarenko, M.-O. Kim, J.R. Grandis and D.E. Johnson in Journal of Dental Research

Supplemental Material

sj-pdf-2-jdr-10.1177_00220345261457066 – Supplemental material for Caspase-8 Mutation Promotes HNSCC Development and Resistance to Anti–PD-1

Supplemental material, sj-pdf-2-jdr-10.1177_00220345261457066 for Caspase-8 Mutation Promotes HNSCC Development and Resistance to Anti–PD-1 by Z. Cui, H. Wu, L.C. Woerner, S.R. Long, C.N. Peterson, N.K. VanLandingham, A. Nazarenko, M.-O. Kim, J.R. Grandis and D.E. Johnson in Journal of Dental Research

Footnotes

Acknowledgements

Cyagen US generated the tamoxifen-inducible CASP8^D305G knock-in mice. The authors thank the Core Animal Facility of the Helen Diller Family Comprehensive Cancer Center for support and advice with mouse breeding and analysis. Flow cytometric data in this study were acquired at the Optical Imaging and Analysis Facility, School of Dental Medicine, State University of New York at Buffalo.

Declaration of Conflicting Interests

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: J.R.G. and D.E.J. are coinventors of a cyclic STAT3 decoy oligonucleotide and report financial interests in Bluedot Bio LLC, the exclusive licensee of the cyclic STAT3 decoy patents.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by National Institutes of Health grants R01DE024728 (D.E.J.) and R35CA231998 (J.R.G.) and by a California Tobacco-Related Disease Research Program award (T29FT0328) and New York State Research Foundation to Z.C.

Ethics Approval and Consent to Participate

This study did not require ethics approval or consent to participate because no human subjects, human patient data, or human-derived materials were used. All experiments were performed with established cell lines or animal models. Ethics approval for studies using mice was obtained from the University of California–San Francisco Institutional Animal Care and Use Committee (IACUC protocol #AN197429-01C), and all experiments were performed in accordance with established guidelines.

ORCID iDs

Z. Cui

H. Wu

N.K. VanLandingham

Data Availability

All data were generated by the authors and are included in the article, and no other public data were used in this study.

A supplemental appendix to this article is available online.

References

Agrawal

, et al. 2011. Exome sequencing of head and neck squamous cell carcinoma reveals inactivating mutations in NOTCH1. Science. 333(6046):1154–1157. https://doi.org/10.1126/science.1206923

Ashkenazi

Dixit

VM.

1998. Death receptors: signaling and modulation. Science. 281(5381):1305–1308. https://doi.org/10.1126/science.281.5381.1305

Barsouk

Aluru

Rawla

Saginala

Barsouk

2023. Epidemiology, risk factors, and prevention of head and neck squamous cell carcinoma. Med Sci (Basel). 11(2):42. https://doi.org/10.3390/medsci11020042

Bray

, et al. 2024. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 74(3):229–263. https://doi.org/10.3322/caac.21834

Burtness

, et al. 2019. Pembrolizumab alone or with chemotherapy versus cetuximab with chemotherapy for recurrent or metastatic squamous cell carcinoma of the head and neck (KEYNOTE-048): a randomised, open-label, phase 3 study. Lancet. 394(10212):1915–1928. https://doi.org/10.1016/s0140-6736(19)32591-7

Cancer Genome Atlas Network. 2015. Comprehensive genomic characterization of head and neck squamous cell carcinomas. Nature. 517(7536):576–582. https://doi.org/10.1038/nature14129

Cerami

, et al. 2012. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer Discov. 2(5):401–404. https://doi.org/10.1158/2159-8290.cd-12-0095

Chow

LQM

. 2020. Head and neck cancer. N Engl J Med. 382(1):60–72. https://doi.org/10.1056/nejmra1715715

Chow

LQM

, et al. 2016. Antitumor activity of pembrolizumab in biomarker-unselected patients with recurrent and/or metastatic head and neck squamous cell carcinoma: results from the phase Ib KEYNOTE-012 expansion cohort. J Clin Oncol. 34(32):3838–3845. https://doi.org/10.1200/jco.2016.68.1478

10.

Cui

, et al. 2021. Caspase-8 mutations associated with head and neck cancer differentially retain functional properties related to TRAIL-induced apoptosis and cytokine induction. Cell Death Dis. 12(8):775. https://doi.org/10.1038/s41419-021-04066-z

11.

Ferris

, et al. 2016. Nivolumab for recurrent squamous-cell carcinoma of the head and neck. N Engl J Med. 375(19):1856–1867. https://doi.org/10.1056/nejmoa1602252

12.

Gao

, et al. 2013. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal. 6(269):pl1. https://doi.org/10.1126/scisignal.2004088

13.

Howard

, et al. 2021. PD-1 blockade on tumor microenvironment-resident ILC2s promotes TNF-α production and restricts progression of metastatic melanoma. Front Immunol. 12:733136. https://doi.org/10.3389/fimmu.2021.733136

14.

Johnson

, et al. 2020. Head and neck squamous cell carcinoma. Nat Rev Dis Primers. 6(1):92. https://doi.org/10.1038/s41572-020-00224-3

15.

Judd

Allen

Winkler

Uppaluri

2012a. Comparative analysis of tumor-infiltrating lymphocytes in a syngeneic mouse model of oral cancer. Otolaryngol Head Neck Surg. 147(3):493–500. https://doi.org/10.1177/0194599812442037

16.

Judd

, et al. 2012b. ERK1/2 regulation of CD44 modulates oral cancer aggressiveness. Cancer Res. 72(1):365–374. https://doi.org/10.1158/0008-5472.can-11-1831

17.

, et al. 2007. Loss of p53 expression correlates with metastatic phenotype and transcriptional profile in a new mouse model of head and neck cancer. Mol Cancer Res. 5(4):351–362. https://doi.org/10.1158/1541-7786.mcr-06-0238

18.

Lee

, et al. 2021. Neoadjuvant PD-1 blockade induces T cell and CDC1 activation but fails to overcome the immunosuppressive tumor associated macrophages in recurrent glioblastoma. Nat Commun. 12(1):6938. https://doi.org/10.1038/s41467-021-26940-2

19.

Leonard

Johnson

DE.

2018. Signaling by cell surface death receptors: alterations in head and neck cancer. Adv Biol Regul. 67:170–178. https://doi.org/10.1016/j.jbior.2017.10.006

20.

Egloff

Sen

Grandis

Johnson

DE.

2014. Caspase-8 mutations in head and neck cancer confer resistance to death receptor-mediated apoptosis and enhance migration, invasion, and tumor growth. Mol Oncol. 8(7):1220–1230. https://doi.org/10.1016/j.molonc.2014.03.018

21.

Mehra

, et al. 2018. Efficacy and safety of pembrolizumab in recurrent/metastatic head and neck squamous cell carcinoma: pooled analyses after long-term follow-up in KEYNOTE-012. Br J Cancer. 119(2):153–159. https://doi.org/10.1038/s41416-018-0131-9

22.

Moore

, et al. 2016a. Established T cell-inflamed tumors rejected after adaptive resistance was reversed by combination STING activation and PD-1 pathway blockade. Cancer Immunol Res. 4(12):1061–1071. https://doi.org/10.1158/2326-6066.cir-16-0104

23.

Moore

, et al. 2016b. Enhanced tumor control with combination mTOR and PD-L1 inhibition in syngeneic oral cavity cancers. Cancer Immunol Res. 4(7):611–620. https://doi.org/10.1158/2326-6066.cir-15-0252

24.

Onken

, et al. 2014. 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(11):2873–2884. https://doi.org/10.1158/1078-0432.ccr-14-0205

25.

Schetters

STT

, et al. 2020. Monocyte-derived APCs are central to the response of PD1 checkpoint blockade and provide a therapeutic target for combination therapy. J Immunother Cancer. 8(2):e000588. https://doi.org/10.1136/jitc-2020-000588

26.

Seiwert

, et al. 2016. Safety and clinical activity of pembrolizumab for treatment of recurrent or metastatic squamous cell carcinoma of the head and neck (KEYNOTE-012): an open-label, multicentre, phase 1b trial. Lancet Oncol. 17(7):956–965. https://doi.org/10.1016/s1470-2045(16)30066-3

27.

Siegel

Giaquinto

Jemal

2024. Cancer statistics, 2024. CA Cancer J Clin. 74(1):12–49. https://doi.org/10.3322/caac.21820

28.

Simon

, et al. 2016. PD-1 expression conditions T cell avidity within an antigen-specific repertoire. Oncoimmunology. 5(1):e1104448. https://doi.org/10.1080/2162402x.2015.1104448

29.

Stransky

, et al. 2011. The mutational landscape of head and neck squamous cell carcinoma. Science. 333(6046):1157–1160. https://doi.org/10.1126/science.1208130

30.

Tran

, et al. 2017. Cisplatin alters antitumor immunity and synergizes with PD-1/PD-L1 inhibition in head and neck squamous cell carcinoma. Cancer Immunol Res. 5(12):1141–1151. https://doi.org/10.1158/2326-6066.cir-17-0235

31.

Zhou

, et al. 2010. Blockade of programmed death-1 pathway rescues the effector function of tumor-infiltrating T cells and enhances the antitumor efficacy of lentivector immunization. J Immunol. 185(9):5082–5092. https://doi.org/10.4049/jimmunol.1001821

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.03 MB

1.53 MB

Caspase-8 Mutation Promotes HNSCC Development and Resistance to Anti–PD-1

Abstract

Keywords

Introduction

Materials and Methods

Results

Generation of Mice with Conditional, Tissue-Specific Expression of the D305G Caspase-8 Mutation

The D305G Caspase-8 Mutation Promotes Carcinogen-Induced Oral Tumor Formation

Caspase-8 Mutation Promotes Progression of Oral Tumors

Caspase-8 Mutation Abrogates the Antitumor Activity of Anti–PD-1

Caspase-8 Mutation Modulates CD8+ T-Cell and Dendritic Cell Levels upon Anti–PD-1 Treatment

Caspase-8 Mutation Disrupts Anti–PD-1–Induced CD8+ T-cell Redistribution and Chemokine Responses in Tumors

Discussion

Author Contributions

Supplemental Material

sj-docx-1-jdr-10.1177_00220345261457066 – Supplemental material for Caspase-8 Mutation Promotes HNSCC Development and Resistance to Anti–PD-1

Supplemental Material

sj-pdf-2-jdr-10.1177_00220345261457066 – Supplemental material for Caspase-8 Mutation Promotes HNSCC Development and Resistance to Anti–PD-1

Footnotes

Acknowledgements

Declaration of Conflicting Interests

Funding

Ethics Approval and Consent to Participate

ORCID iDs

Data Availability

References

Supplementary Material

Caspase-8 Mutation Modulates CD8⁺ T-Cell and Dendritic Cell Levels upon Anti–PD-1 Treatment

Caspase-8 Mutation Disrupts Anti–PD-1–Induced CD8⁺ T-cell Redistribution and Chemokine Responses in Tumors