So Grateful for My X: Sex Chromosomes Drive Differences in Glioblastoma Immunity

Higher cancer incidence in males compared with females can be observed across many nonreproductive tissues, including bladder, skin, and lung (1). The mechanisms mediating this difference are poorly defined. Recently, sex-based disparities in antitumor T-cell responses was found to contribute to better tumor control in females compared with males in mouse models of bladder and colorectal cancers (2–4). In this issue of Cancer Discovery, Lee and colleagues add to this emerging narrative of sex-based differences in antitumor immunity by investigating the function of male versus female T-cell populations infiltrating glioblastoma in both mouse models and human patients (Fig. 1; ref. 5).

Glioblastoma is the most common cancer of the brain and is relatively intractable to many standard-of-care therapeutic strategies and immunotherapy. Thus, new insight into the mechanisms that drive control versus resistance in glioblastoma is critically needed. Glioblastoma also demonstrates a sex bias in the human population, with females experiencing 1.6-fold lower incidence compared with males (5). Lee and colleagues observed that mice recapitulate this disparity, and report that female mice survived significantly longer than males following intracranial injection of multiple diverse glioblastoma cell lines. By comparing tumor growth in immunocompetent versus immunodeficient animals, the authors establish that sex-based differences in glioblastoma are mediated in part by CD8⁺ T cells.

Lee and colleagues then defined the phenotypic characteristics of T cells infiltrating mouse glioblastoma tumors that contribute to sex disparities in disease progression. Tumors generated in female mice contained greater frequencies of CD44⁺TCF1⁻TIM3⁻ polyfunctional effector CD8⁺ T cells. In contrast, tumors generated in male mice contained greater frequencies of progenitor exhausted CD8⁺ T (T_PE) cells defined by high expression of programmed cell death protein 1 (PD-1) and TCF1 and low expression of TIM3. These results are similar to observations recently reported in the MB49 mouse bladder cancer model (2). However, other studies have shown that MC38 colorectal tumors in female mice and in female patients with melanoma and renal cancers contain greater frequencies of T_PE cells compared with tumors in males (3). Lee and colleagues may provide some continuity between these seemingly contradictory findings when they compared the function of male versus female effector, T_PE, and terminally exhausted CD8⁺ T cells. They observed that female CD8⁺ T cells from mice and humans maintained greater production of cytokines and cytotoxicity compared with male CD8⁺ T cells across the exhaustion spectrum. This suggests that fundamental aspects of the biology of T-cell exhaustion display sex-based variability. Furthermore, these findings may indicate that regardless of the phenotype assigned to female cells, the T-cell response against tumors in females is simply more functional than the male antitumor response.

Because higher frequencies of T_PE cells are associated with better immunotherapy response (6), Lee and colleagues further investigated whether superior tumor control in female mice could also be observed in the setting of αPD-1 treatment. However, αPD-1 resulted in enhanced survival of male tumor-bearing animals, while having only a limited impact on the survival of female tumor-bearing animals. This finding echoes recent observations that male sex is significantly associated with better response following immune checkpoint therapy (ICT) in many tumor types (7). Much of the narrative surrounding the idea of sex-based differences in immunotherapy response has stemmed from retrospective analyses of clinical trials. However, clinical trial cohorts have a male majority and therefore may lack the ability to accurately identify sex-based differences in outcomes. The findings by Lee and colleagues are among the first to translate this trend of better ICT response in male compared with female humans to an experimental setting that is well controlled and specifically designed to interrogate sex as a variable in immunotherapy. However, a limitation of this work is that only one ICT strategy was used. It is possible that while targeting PD-1 showed efficacy predominantly in the male cohort, a different ICT strategy would show preferential efficacy in females or would improve survival in both sexes.

Previous reports investigating the mechanisms through which sex influences tumor-infiltrating T-cell function have focused on the role of sex hormones, particularly the immunosuppressive impact of androgen signaling (2, 3, 8). While a role for cell-extrinsic mechanisms was noted for the phenotypic differences observed between male and female T cells in glioblastoma, Lee and colleagues uncover a major T cell–intrinsic mechanism using elegant bone marrow transplant experiments. The authors first transferred male and female T cell–depleted bone marrow to irradiated recipients of either the same or opposite sex. They observed that survival following intracranial injection of glioblastoma was dependent on the sex of the bone marrow donor rather than the sex of the recipient. Next, congenically labeled male and female bone marrow was mixed at equivalent ratios prior to transplant into male or female hosts that were later inoculated with tumor cells. These experiments allowed for the interrogation of intrinsic differences in T-cell phenotype between cells of differing sex within the same tumor microenvironment. Within male hosts, male-derived tumor-infiltrating CD8⁺ T cells contained greater frequencies of T_PE cells compared with the female-derived CD8⁺ T-cell population. Similar trends were observed in female hosts, although the difference between the frequency of male and female T_PE cells did not reach significance (P = 0.0540). In combination with previous reports identifying the role of androgens in establishing CD8⁺ T-cell suppression, these results highlight that sex chromosomes and hormones both influence tumor immunity in complex ways.

On the basis of the observation that a cell-intrinsic mechanism was playing a role in determining phenotypic differences between male and female T cells in glioblastoma, Lee and colleagues then investigated whether expression differences in X chromosome–linked genes could be involved. The dosage of X-linked genes between male (XY) and female (XX) cells is typically normalized through the process of X inactivation, in which one X is randomly silenced in every cell of a female individual. However, escape from X inactivation has been reported for approximately 10% to 20% of the genes encoded by the X chromosome, and immune cells have displayed particularly dynamic levels of X inactivation across the entire chromosome (9). This can result in overexpression of X-linked genes in females compared with males and has been attributed to greater incidence of autoimmunity in the female population (9). The authors observed that CD8⁺ T cells of female mice and humans expressed greater levels of the histone demethylase UTX than their male counterparts. When UTX was inhibited by the small-molecule GSK-J4 in in vitro T-cell exhaustion assays, IFNγ production by female T cells decreased and the expression of exhaustion markers TOX, PD-1, and TIM3 increased in T cells of both sexes. While these findings indicate that differential expression of X-linked genes is a regulator of enhanced CD8⁺ T-cell function in females, the relevance of UTX overexpression in female cells to better control of glioblastoma was not established. In addition, the epigenetic targets of UTX that are relevant for T-cell function versus exhaustion will be an interesting area of future investigation.

More robust immune reactions in females compared with males following vaccination, in response to infections, and against self are well characterized (10). Our appreciation of sex-based differences in the immune response against cancer is only beginning to develop. The report by Lee and colleagues expands our understanding of the types of cancer that are influenced by and the potential mechanisms driving sex-based differences in T cell–mediated tumor immunity. The authors also add important evidence to the growing body of literature showing that males and females do not respond to ICT equivalently. These findings further indicate that to improve immunotherapy response and patient survival, we must understand sex as a variable. Future investigations should be targeted at extending our understanding beyond CD8⁺ T cells and into the multitude of other cell types that influence tumor fate.