The Innate Immune System and the TRAIL-Bcl-XL Axis Mediate a Sex Bias in Lung Cancer and Confer a Therapeutic Vulnerability in Females

Abstract

There is a significant sex bias in lung cancer, with males showing increased mortality compared with females. A better mechanistic understanding of these differences could help identify therapeutic targets to personalize cancer therapies to each sex. After observing a clear sex bias in humanized mice, with male patient-derived xenograft lung tumors being more progressive and deadlier than female patient-derived xenograft lung tumors, we identified mouse tumor models of lung cancer with the same sex bias. This sex bias was not observed in models of breast, colon, melanoma, and renal cancers. In vivo, the sex bias in growth and lethality required intact ovaries, functional innate NK cells and monocytes/macrophages, and the activating receptor NKG2D. Ex vivo cell culture models were sensitized to the anticancer effects of NKG2D-mediated NK cell and macrophage killing through the TRAIL-Bcl-XL axis when cultured with serum from female mice with intact ovaries. In both flank and orthotopic models, the Bcl-XL inhibitor navitoclax (ABT-263) improved tumor growth control in female mice and required NK cells, macrophages, and the TRAIL signaling pathway. This research suggests that navitoclax and TRAIL pathway agonists could be used as a personalized therapy to improve outcomes in women with lung cancer. Significance: Lung cancers in females are more susceptible to killing through a TRAIL-Bcl-XL axis, indicating that targeting this axis therapeutically could represent a personalized approach to treat female patients with lung cancer.

Figure 1.. Sex Differences in the Control of Lung Cancer PDX Growth in Humanized Mouse Models.

(a) Flow cytometry for CD45⁺ populations taken from the bone marrow of NSG and humanized NSG mice. (b) Average tumor volume of two independent female and male lung PDX flank tumors in NSG and humanized NSG mice. Aggregate of two independent CD34⁺ donors for each humanized mouse group. N ≥ 3 animals per donor with a total N ≥ 6 animals per group. (c) Kaplan-Meier plots of panel (b). Significance shown as inset. (d) Flow cytometry assessment of hCD45⁺ population from humanized mice. Specific immune cell populations are presented as a percentage of hCD45⁺ cells. N ≥ 6 mice per group. ND = Not determined. * = p-value <0.05, two-way ANOVA.

Figure 2.. An Observed Sex Difference in Tumor Growth in Mouse Models of Lung Cancer.

(a) Tumor volume of flank CMT-167 and Lewis Lung Carcinoma (LLC) lung cancer models in female or male C57BL/6 mice. N ≥ 8 animals per group. (b) Kaplan-Meier survival plots of mice shown in panel (a). Significance shown as inset. (c) Tumor volume of flank CT26 colon cancer, Renca kidney cancer, E0771 and 4T1 breast cancer, and B16-F10 melanoma models grown in either male or female syngeneic C57BL/6 or BALB/C mice. N >8 animals per group. (d) Tumor volume of flank CMT-167 and LLC tumors grown in castrated (Ctx) or ovariectomized (Ovx) C57BL/6 mice with wild-type (WT) controls. N ≥ 6 mice per group. All panels * = p-value <0.05, non-parametric Friedman’s test or log-rank test.

Figure 3.. The Innate Immune System and NKG2D are Required for Sex Differences in CMT-167 and LLC Tumor Growth.

(a) Log2-transformed complete blood counts (CBC) of C57BL/6 mice without tumors, and those bearing CMT-167 or LLC tumors. * = p-value <0.05, t-test. N ≥ 4 animals per group. (b-g) Tumor volume of flank CMT-167 and LLC lung cancer models in (b) NSG, (c) RAG1 KO, (d) treated with PBS, neutral liposome control or clodronate liposome, (e) treated with control antibody (MPC11) or NK1.1 depleting (PK136), or (f) control antibody (MPC11) or NKG2D blocking (CX5). N ≥ 6 animals per group. (g) Tumor volume of flank CMT-167 and LLC lung cancer models in control C57BL/6 (B6) or NCR1 GFP/GFP (NCR1) mice. N ≥ 6 animals per group. All panels * = p-value <0.05, non-parametric Friedman’s test or mixed-model test.

Figure 4.. Sex Differences in Tumor Cell Apoptosis and Immune Cell Populations in CMT-167 and LLC Tumors.

(a,b) Histological analysis of cleaved caspase 3 (CC3) and Ki67 staining in tumors taken from (a) C57BL/6 or (b) NSG mice. N >5 tumors per group. (c) Histological analysis of cleaved caspase 3 (CC3) in tumors taken from C67BL/6 mice treated with control (MPC11) or anti-NKG2D (CX5) mAb. (d) Multi-color flow cytometry analysis of tumor resident immune cells from male, female, and Ovx female tumor-bearing mice. N = 3–4 biological replicates. (e) Bar graphs for NK cell populations from (d) that show significant differences between males, females, and Ovx females. (f) Bar graph for macrophage population from LLC tumors from (d) that show significant differences between males, females, and Ovx females. (g) Bar graphs for M-MDSC and G-MDSC populations from CMT-167 tumors from (d). All panels * = p-value <0.05, t-test or ANOVA.

Figure 5.. Sex Differences in Innate Immune Cell Killing of CMT-167 and LLC Cells Ex Vivo Requires NKG2D and TRAIL.

(a,b) Analysis of CFSE-labeled CMT-167 or LLC killing (7AAD uptake) after 12 hours co-culture with (a) monocyte/macrophages or (b) NK cells harvested from female, ovariectomized (Ovx) female, or male mice. Co-culture was done with cells preincubated with and in the presence of serum from female, Ovx female, or male mice matching the origin of the immune cells. (c,d) Same assays as in panel (b) after 48 hours co-culture with (c) monocytes/macrophages or (d) NK cells harvested from female, Ovx female, or male mice. (e,f) Same assays as in panel (a) but with matched (male + male and female + female) and unmatched (male + female and female + male) co-culturing conditions with monocytes/macrophages (e) or NK cells (f). (g) Crystal violet staining of cells after 4 days co-cultured with 50 ng/ml FASL, TRAIL, IFNγ or TNFα. (h-k) Same assays in panel (b) after 48 hours co-culture with (h,j) monocyte/macrophages (i,k) NK cells harvested from female, Ovx female, or male mice. (h,i) Co-cultures were treated with NKG2D blocking (CX5) or control (MPC11) mAb. (j,k) Co-cultures were treated with TRAIL blocking or control mAb. (l) Cell counting over a time course of CMT-167 and LLC cell lines expressing transgene human Bcl2 or Bcl-XL, or vector-only controls, treated with either vehicle or 50 ng/ml TRAIL. (m) Tumor volume of flank LLC and CMT-167 lung cancer models expressing human Bcl2 or Bcl-XL in female and male C57BL/6 mice. All panels N ≥ 3 – 6 replicates per group. * = p-value <0.05, t-test, one way ANOVA, or non-parametric Freidman’s test.

Figure 6.. Sex Differences in the Innate Immune Control of CMT-167 and LLC Tumor Growth Requires the TRAIL and Bcl-XL Pathways.

(a) Tumor volume of flank CMT-167 and LLC lung cancer models in C57BL/6 mice treated with vehicle or 100 mg/kg ABT-263 on starting 14 days post-transplant for five doses over 10 days. N ≥ 8 animals per group. (b) Kaplan-Meier survival plots for C57BL/6 mice from panel (a). (c) Tumor volume of flank CMT-167 and LLC lung cancer models in C57BL/6 mice treated with combinations of NK1.1 (PK136) or control mAb, and 100 mg/kg of ABT-263 or vehicle, two weeks post-transplant. N ≥ 6 animals per group. (d) Tumor volume of flank CMT-167 and LLC lung cancer models in male or female C57BL/6 mice treated with combinations of liposomal clodronate or control liposomes, and 100 mg/kg of ABT-263 or vehicle, two weeks post-transplant. N ≥ 5 animals per group. (e) Annexin V/7AAD staining of CMT-167 and LLC cell lines after 7 days co-cultured with 50 ng/ml TRAIL or/and in combination with ABT-263. N = 3 per group. * = p-value <0.05, one way ANOVA. (f) Volume of flank TRAIL receptor (DR5) KD, or vector-only control, CMT-167 and LLC tumors in C57BL/6 mice treated with vehicle or 100 mg/kg ABT-263 for two weeks post-transplant. N ≥ 5 per group. (g, h) Volume measurements of CMT-167 (g) and LLC (h) flank tumors in C57BL/6 mice receiving 15 mg/kg etoposide (CMT-167) or 8 mg/kg oxaliplatin (LLC) in sequence with 100 mg/kg ABT-263. N ≥ 5 per group. Panels a,c,d,f,g,h * = p-value <0.05, non-parametric Friedman’s test or mixed models test.

Figure 7.. Orthotopic CMT-167 and LLC Tumor Models Recapitulate Sex Differences in Tumor Growth, Sensitivity to ABT-263 and Dependence on the TRAIL Pathway.

(a) Bioluminescent imaging (BLI) average integrated signal intensity of orthotopic CMT-167-tdTomato-Luc and LLC-tdTomato-Luc and tumors in C57BL/6 mice. N ≥ 5 biological replicates. (b) Kaplan-Meier survival plots for mice shown in panel (a). (c) Weights of whole lungs taken from male and female CMT-167-tdTomato-Luc and LLC-tdTomato-Luc tumor-bearing mice normalized to lung weights from non-tumor-bearing mice. N= 4 biological replicates, * = p-value <0.05, t-test. (d) BLI average signal intensity of CMT-167-tdTomato-Luc and LLC-tdTomato-Luc tumors in an orthotopic model of lung cancer in C57BL/6 mice receiving either vehicle or 100 mg/kg ABT-263. N ≥5 biological replicates. (e) BLI average signal intensity of orthotopic CMT-167-tdTomato-Luc or LLC-tdTomato-Luc cells transduced with shRNAs to TRAIL receptor (DR5), or vector-only control, in C57BL/6 mice treated with treated with vehicle or 100 mg/kg ABT-263 two weeks post-transplant. N ≥5 animals per group. (f,g) BLI average signal intensity of orthotopic (f) CMT-167-Luc tumors receiving 15 mg/kg etoposide or (g) LLC-Luc tumors 8 mg/kg oxaliplatin, 100 mg/kg ABT-263 alone or in sequence in C57BL/6 mice. N = 5 biological replicates. Panels a,d,e,f,g * = p-value <0.05, non-parametric Friedman’s test or mixed models test.