Enhancement of anti-tumor efficacy of immune checkpoint blockade by alpha-TEA

Cancer immunotherapy such as anti-PD-1/anti-PD-L1 immune checkpoint blockade (ICB) can provide significant clinical benefit in patients with advanced malignancies. However, most patients eventually develop progressive disease, thus necessitating additional therapeutic options. We have developed a novel agent, a-TEA-LS, that selectively induces tumor cell death while sparing healthy tissues, leading to increased activation of tumor-reactive T cells and tumor regression. In the current study, we explored the impact of combined a-TEA-LS + ICB in orthotopic and spontaneously arising murine models of mammary carcinoma. We found that a-TEA-LS + ICB led to increased production of pro-inflammatory cytokines that were associated with a reduction in tumor growth and prolonged survival. Together, these data demonstrate the potential utility of a-TEA-LS + ICB for the treatment of breast cancer and provide the rationale for clinical translation of this novel approach.


Introduction
Immune checkpoint blockade (ICB) antibodies block the interactions between Programmed Death-1 (PD-1) expressed on activated and "exhausted" T cells and the PD-L1 ligand expressed on cancer cells and tumor-associated immune cells in the tumor microenvironment (TME) to unleash a T cell immune attack against cancer cells (1)(2)(3). ICB-targeted therapies have improved clinical outcomes in numerous clinical trials, and several are now approved for multiple cancer indications usually in combination with chemotherapy in the neoadjuvant and adjuvant settings (4)(5)(6)(7). While much success has been achieved using these agents, curative responses occur in only a fraction of patients.
Although ICB has made a difference in several cancer indications including melanoma, non-small cell lung cancer, Merkel cell carcinoma, head and neck cancer, and renal cell cancer amongst others, triple negative breast cancer (TNBC) has remained a clinical challenge until the recent FDA approval of pembrolizumab, an anti-PD-1 monoclonal antibody in combination with chemotherapy in early and late stage TNBC (7). Adjuvant chemotherapy, which frequently consists of DNA targeting agents, while necessary, can contribute additional acute toxicity and side effects to the immune-related toxicities inherent in the use of immune checkpoint inhibitors (8)(9)(10)(11)(12)(13)(14). These limitations can negatively impact the quality of life of patients during and after treatment and highlight the need for safer, less toxic anti-cancer agents that can be used in combination with ICB to improve patient outcomes.
In pre-clinical proof-of-concept in vivo studies, a-TEA increased the frequency of activated CD4 + and CD8 + T cells in the TME. Furthermore, in vivo depletion of CD4 + and CD8 + T cells in immune competent mice reduced overall survival of a-TEAtreated, tumor-bearing animals implicating T cells in the antitumor immune response (21). With a view to utilizing a-TEA as an effective adjuvant to improve the efficacy of ICB and restore antitumor activity, we evaluated the anti-tumor activity of a-TEA-LS, a scalable form of a-TEA (22) in combination with PD-1/PD-L1 blockade in three murine models of mammary carcinoma. We report here that a-TEA-LS significantly enhanced the anti-tumor efficacy of anti-PD-1 and anti-PD-L1 in multiple tumor models including inducing complete tumor regression in some instances. These anti-tumor effects were associated with increased effector T cell function at the tumor site.

Materials and methods
Preparation of a-tocopheryloxyacetic acid lysine salt (a-TEA-LS) Alpha-tocopheryloxyacetic acid lysine salt (a-TEA-LS) was synthesized by Olon Ricerca Biosciences LLC (Concord, OH) using a modification of a previously described procedure (22). Briefly, a-TEA-LS was prepared by reacting alpha-D-tocopherol with ethyl bromoacetate to form the ethyl ether intermediate. The ethyl ether intermediate was then reacted with potassium hydroxide to form a-TEA free acid. The lysine salt was formed by adding aqueous lysine solution to a solution of a-TEA in isopropyl alcohol. The lysine salt with its empirical formula of C 37 H 66 N 2 O 6 and its molecular weight of 634.93 g/mol is a stable crystalline off-white powder. a-TEA-LS was incorporated into the AIN93G diet by Envigo RMS LLC (Indianapolis, IN) at a concentration of 1g a-TEA per kg chow (0.1%).

Mice
Six to eight-week-old female wild-type BALB/c mice and MMTV-PyMT transgenic mice (FVB/N-Tg(MMTV-PyVT) 634Mul/J) were purchased from Jackson Labs (Bar Harbor, ME). All mice were maintained under specific pathogen-free conditions in the Earle A. Chiles Research Institute vivarium at Providence Portland Medical Center (Portland, OR) or at the Experimental Mouse Shared Resource facility at the University of Arizona Cancer Center (Tucson, AZ). Experimental procedures were performed according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals and in accordance with the Institutional Animal Care and Use Committees at both institutions.
In vivo tumor studies BALB/c mice were inoculated with 5 x 10 4 4T1 tumor cells subcutaneously (SC) into the right mammary fat pad. Eph4 1424 tumor cells (5 x 10 4 ) were injected with Matrigel (Becton Dickinson) into a cleared L4 mammary fat pad (MFP) of BALB/c mice in a total volume of 50 µl. The mice received normal diet until tumor establishment (day 10 post tumor-implantation;~25 mm 2 ) and were then switched to a-TEA-LS-containing mouse chow. Tumor growth was monitored twice weekly by measuring the tumor length (L) and width (W) using calipers and calculating the tumor area as: A = (L × W). Animals were euthanized when tumor area reached >175 mm 2 (4T1) or at the end of the experiment on Day 84 (Eph4 1424). For the spontaneous tumor model, MMTV-PyMT transgenic mice received normal diet until 6 weeks of age (day 42) and were then switched to a-TEA-LS chow. Control mice remained on matched control diet throughout the study. Mice were inspected for spontaneous tumor growth in the mammary fat pads beginning at week 6. As tumors presented, calipers were used to measure length (L) of the tumor's longest axis by perpendicular width (W) to calculate tumor area (L x W = A). Individual tumor area was collected for each tumor of each mouse twice weekly. The sum of tumor burden was calculated at each measurement. Mice were euthanized when the total tumor burden was >300 mm 2 .

Lymphocyte and TIL isolation
Tumor-infiltrating lymphocytes (TIL) were harvested by cutting tumors into small fragments followed by digestion in 1 mg/ml collagenase and 20 mg/ml DNase (Sigma Aldrich) in RPMI 1640 for 45 min at room temperature. TILs were filtered through 70 µm nylon mesh, then stained for analysis by flow cytometry as described above. Inguinal, axillary, and brachial lymph nodes were harvested and processed to obtain single-cell suspensions. ACK lysing buffer (Lonza) was used to lyse red blood cells. Cells were rinsed with complete RPMI and stained for flow cytometry analysis.

Statistical analysis
Statistical significance was determined by unpaired Student ttest (for comparison between two groups), one-way ANOVA for (comparison among three or more groups), or Kaplan-Meier survival (for tumor survival studies) using GraphPad Prism software (GraphPad, San Diego, CA); a p-value of <0.05 was considered significant.

Results
We investigated the extent to which a-TEA-LS treatment would augment the efficacy of ICB in mammary carcinoma tumor-bearing mice. 4T1 tumor cells, a model of triple-negative breast cancer, were injected orthotopically into wild-type female BALB/c mice and then treated with control diet or a-TEA-LS plus IgG (ctrl) or aPD-1 monoclonal antibody (mAb) ( Figure S1A). While monotherapy with a-TEA or aPD-1 had a minimal impact on tumor growth or survival ( Figures 1A, B, respectively), combined a-TEA-LS + aPD-1 therapy led to decreased tumor size ( Figure 1A) and significantly improved survival ( Figure 1B). When comparing individual tumor sizes (change in tumor area) between day 10 and 23 post-injection, we observed a transient reduction in tumor size in the majority (6/7; 86%) of a-TEA-LS + aPD-1 treated mice as compared to only 2/15 (13%), 4/15 (27%), and 1/6 (17%) of IgG, a-TEA-LS, and a-PD-1treated control cohorts, respectively ( Figure 1C).
Next, we investigated the extent to which a-TEA-LS + ICB affected the frequency and function of tumor infiltrating lymphocytes. 4T1 tumor-bearing mice were treated with a-TEA-Lys in the presence or absence of aPD-1 and then 7 days posttreatment, tumors were harvested and CD4 + and CD8 + T cell responses were evaluated by flow cytometry (Figures 1D-F). Combination a-TEA-LS + ICB therapy had minimal impact on the frequency of effector Foxp3 -CD4 + (Teff) or regulatory Foxp3 + CD4 + (Treg) T cells, although we detected an increase in Tregs following aPD-1 as compared to the a-TEA-LS-treated group. There was also a trend (though not statistically significant among the groups) towards decreased PD-1 expression on CD4 + Teff and Tregs receiving combination therapy ( Figure 1E, right panels).
Analysis of the CD8 + T cell compartment revealed no change in cell frequencies following mono-or combination therapy ( Figure 1F, top left), but a similar trend towards decreased PD-1 expression as was observed in the CD4 + T cell compartment ( Figure 1F, top right). Anti-PD-1 and a-TEA-Lys + anti-PD-1 increased granzyme A expression, although only aPD-1 showed a statistical difference compared to a-TEA-Lys alone ( Figure 1F, bottom left). However, a-TEA-LS + aPD-1 combination therapy elicited an increased proportion of effector IFN-g + CD8 + T cells ( Figure 1F, bottom right) as compared to ctrl/IgG and a-TEA-LS/IgG-treated controls. These data demonstrate that increased IFN-g + CD8 + T cells, rather than the extent of granzyme B expression, is associated with the improved survival observed following a-TEA-LS + aPD-1 therapy.
We also investigated the impact of a-TEA-LS + aPD-1 therapy on the frequency and functional status of myeloid cells including CD11b + Ly6C hi Ly6Gmonocytic myeloid-derived suppressor cells (Mo-MDSC), CD11b + Ly6C m e d Ly6G + polymorphonuclear MDSCs (PMN-MDSC), and CD11b + F4/ 80 + Ly6C -Ly6Gmacrophages ( Figure S2A). We did not observe any significant changes in the frequency of these myeloid cell subsets post-treatment ( Figures S2B-D). There was an increase in arginase within the Mo-MDSC population following a-TEA-LS monotherapy as compared to all other groups ( Figure S2B) and reduced iNOS following a-TEA-LS as compared to the control group in the PMN-MDSC group ( Figure  S2C), but no differences were observed in macrophages across treatments ( Figure S2D). Recent analyses have suggested differences in the therapeutic efficacy of aPD-1 vs. aPD-L1 for the treatment of TNBC (23)(24)(25)(26). It is also known that the preclinical aPD-1 and aPD-L1 mAbs have differential binding affinity to their targets, which may impact their efficacy (27). Therefore, we sought to evaluate the efficacy of a-TEA-LS + aPD-L1 therapy in 4T1 tumor-bearing mice. Combined a-TEA-LS + aPD-L1 therapy was associated with a slight decrease in tumor size (Figure 2A) and some improved survival ( Figure 2B), though these did not reach statistical significance. Our analysis of changes in individual tumor area revealed a reduction in tumor size in the majority (4/7; 57%) of a-TEA-LS + aPD-L1-treated mice as compared to only 0/7 (0%), 0/7 (0%), and 1/7 (14%) of IgG, a-TEA-LS, and a-PD-L1-treated control cohorts, respectively ( Figure 2C). In another cohort of mice, we observed a significant reduction in tumor size in the a-TEA-LS + a-PD-L1 combination group at an earlier time of tumor harvest as compared to monotherapy-treated controls ( Figure 2D) along with a significant increase in several effector proteins including MIP-1a/CCL3, MIP-1b/CCL4, and IFNg following a-TEA-LS + aPD-L1 ( Figure 2E). Global analysis of the cytokine milieu after a-TEA-LS + ICB therapy suggests that this combination also drives an anti-tumor proinflammatory Th1polarized profile to support tumor control as no significant changes were observed in Th2 cytokines (SF3).
Given the limited anti-tumor effect observed in 4T1 tumorbearing mice, we next asked whether combined a-TEA-LS + aPD-L1 therapy exhibited therapeutic efficacy in two additional models, Eph4 1424 tumor-bearing mice, which express constitutively active MEK1 (28), and the PyMT-MMTV Tg model of spontaneously arising mammary carcinoma. Combination a-TEA-LS + aPD-L1 therapy decreased tumor growth and improved survival in Eph4 1424 tumor-bearing mice (Figures 3A, B). Next, PyMT-MMTV Tg mice, which express the Polyoma Virus middle T antigen under the control of mammary tumor virus promoter/enhancer, were treated with control diet, IgG, a-TEA-LS, and/or aPD-L1 starting at 6 weeks of age ( Figure S1B) and then tumor growth and survival were determined. Combined a-TEA-LS + aPD-L1 therapy significantly improved survival ( Figure 3C), which was associated with a significant reduction in individual tumor size ( Figure 3D) and number of tumors per mouse ( Figure 3E) as compared to aPD-L1 alone.

Discussion
While ICB has provided a new treatment option for patients with advanced malignancies, most patients still develop progressive disease highlighting the critical need for additional therapies. In this study, we investigated the extent to which a novel therapy, a-TEA-LS, in conjunction with aPD-1 or aPD-L1 ICB would impact the growth of murine mammary carcinomas. Our data revealed that combined a-TEA-LS + aPD-1 ICB therapy significantly reduced tumor growth and was associated with the activation of effector CD8 + T cells within the TME as compared to controls (Figure 1). The increased therapeutic efficacy of a-TEA-LS + aPD-1 therapy was associated with a trend of lower PD-1 + CD4 + and CD8 + T cells ( Figures 1E, F), suggesting they may be less exhausted or suppressed than control-treated mice. Supporting this hypothesis, we observed a significant increase in effector IFN-g + CD8 + T cells following combination therapy as compared to controls ( Figure 1F, lower right panel).
In contrast, we observed less potent effects on tumor growth or survival following a-TEA-LS + aPD-L1 treatment in the 4T1 tumor model (Figures 2A, B, respectively). The reasons for this differential impact on tumor growth and survival as compared to a-TEA-LS + aPD-1 therapy remain unclear. For example, we observed a significant increase in the proinflammatory cytokine IFN-g following a-TEA-LS plus aPD-1 or aPD-L1 ( Figures 1F, 2E), but this was not sufficient for generating equivalent anti-tumor responses. The reduced efficacy of aPD-L1 may reflect differences in the biochemical properties of murine aPD-1 vs. a-PD-L1 mAbs such as affinity for their targets (27). We are conducting additional studies to explore further the differences in therapeutic efficacy between these agents in murine models of mammary carcinoma.
Given the slightly reduced responsiveness of a-TEA-LS + aPD-L1 ICB in the 4T1 model, we explored the therapeutic efficacy of this combination therapy in two additional tumor models, Eph4 1424 (MEK1-driven) and the PyMT-MMTV Tg model of spontaneously arising mammary carcinoma. Combination therapy led to improved survival in both models ( Figure 3) and we also observed a significant decrease in tumor foci and total tumor burden in the PyMT-MMTV Tg model following a-TEA-LS + aPD-L1 immunotherapy ( Figures 3D, E), highlighting the therapeutic impact of this approach.
Recent work has highlighted the important contribution of myeloid cells to immunosuppression with the TME (29-33). However, we did not observe any significant changes in the frequency or functional status of monocytic (Mo-MDSC) or polymorphonuclear myeloid-derived suppressor cells (PMN-MDSC) or tumor-associated macrophages post-treatment (SF2), suggesting that the efficacy of a-TEA-LS + ICB was likely primarily associated with promoting T cell function, rather than reducing MDSC-mediated immune suppression. Similarly, we detected few changes in the Treg compartment in terms of their frequency or phenotype ( Figure 1E and data not shown), although the impact on Treg (or MDSC)-specific suppression has not been assessed.
Previous studies have demonstrated the ability of a-TEA-LS to promote cross-presentation of tumor-associated antigens by driving increased tumor cell-specific autophagy (19). It is likely that the addition of ICB mitigates the induction of T cell exhaustion following treatment with a-TEA-LS, thus enabling a more potent anti-tumor response to occur. While the current study provides some insight into the mechanisms by which a-TEA-LS + ICB supports tumor regression through enhanced anti-tumor immunity, future work is planned to further understand how this approach alters the TME. For example, we plan to use single-cell RNA-seq + scTCR-seq analysis to fully characterize the entirety of the TME at single-cell resolution as well as using multiplex immunohistochemistry to investigate the impact of combination therapy on the spatial organization of the TME.
Most importantly, these data provide the rationale to explore combined a-TEA-LS + ICB therapy for patients with advanced malignancies, including breast cancer. A recent study evaluated the safety of a-TEA-LS in a phase I clinical trial as a monotherapy in patients with advanced cancer (NCT02192346). Currently, escalating doses of a-TEA-LS in combination with trastuzumab (anti-HER2 mAb) are being evaluated for safety in a Phase Ib trial in patients with treatment-refractory HER2 + metastatic breast cancer (NCT04120246). Based on the safety profile of a-TEA-LS and the enhanced anti-tumor efficacy demonstrated by a-TEA-LS + ICB in pre-clinical models of breast cancer, a-TEA-LS has the potential to be used as an adjuvant drug to improve the effectiveness of ICB in human breast cancer.

Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics statement
Experimental procedures were performed according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals and in accordance with the Earle A. Chiles Research Institute (EACRI) Institutional Animal Care and Use Committee (Animal Welfare Assurance No. A3913-01).

Author contributions
WR made substantial contributions to the conception, design, data analysis and interpretation, and drafted the manuscript. MK made substantial contributions to data acquisition and analysis. EA provided key reagents and contributed to the conception and design of the experiments and drafted the manuscript. All authors contributed to the article and approved the submitted version.

Funding
This work was funded by NIH grant #R43CA228721 to Veana Therapeutics and by the Providence Portland Medical Foundation.