Liu, Xin

Abstract Description CD8 T cell programming creates a pool of naïve, effector, memory, or exhausted subsets, however the regulation of these trajectories is not expansively elucidated and has serious implications for cancer immunity and immune checkpoint inhibitor (ICI) therapy. We found a significant decrease of Dual Specificity Phosphatase 2 (DUSP2) in CX3CR1+ CD8s, a highly resilient subset, from ICI responders. Thus, we hypothesized that DUSP2 negatively regulates CD8 T cells in responses to ICI therapy. DUSP2 knockdown (KD) statistically improved CD8 cytotoxicity and expression of NKG7, a vital effector molecule. DUSP2 KD also increased SMARCA4 expression, a BRG/BRM-associated factor (BAF) complex ATPase, and increased open chromatin regions of BAF components. BAF complexes are chromatin remodelers implicated in CD8 differentiation pathways. In line with this, we found CD8s from DUSP2 conditional knockout (cKO) mice demonstrated a trend in defective memory generation. DUSP2 cKO CD8s also trended towards decreased PD-1 and increased CX3CR1 expression, demonstrating a more effector like phenotype. DUSP2 cKO mice did not show a significant difference in tumor protection in-vivo, presumably due to skewing of CD8 fates via BAF. Promisingly, DUSP2 KD increased ex-vivo cytotoxic functionality in about 50 percent of tested cancer patient PBMCs. Adjuvant BAF inhibition with DUSP2 KO versus KD in a specific CD8 subset may enhance responses further, providing a novel avenue for improving ICI efficacy. Funding Sources Supported by NCI 1F31CA287495-01 Topic Categories Immune Response Regulation: Molecular Mechanisms (IRM)

Abstract Background: Although targeting myeloid cells has become one of the next generation targets for cancer immunotherapy, there is a lack of therapeutic tools to specifically target these cells. Our group has developed a novel humanized PD-L1-targeting antibody, H1A, that induces PD-L1 degradation via blockade of PD-L1 and CMTM6, therefore limiting PD-L1 intrinsic signaling while also preventing interactions with PD-1. In the immune cell compartment, CMTM6 expression is restricted to myeloid cells, making these cells the primary target of H1A. Recently, we unexpectedly found that targeting intrinsic PD-L1 in myeloid cells via H1A resulted in increased release of CCL2 from myeloid cells. In this study, we examined the role of CCL2/CCR2 signaling in CD8 T cell function using human peripheral blood-derived immune cells and evaluated the therapeutic potential of H1A antibody using our humanized PD-L1/PD-1 mouse model. Methods: Human PBMCs derived from healthy donors were treated with H1A or Atezolizumab, a blocking PD-L1 antibody, followed by mass cytometry and flow cytometry to evaluate immune cell phenotype and functional states. PD-L1 expressing myeloid cells isolated from human PBMCs were treated with H1A or Atezolizumab and assessed for CCL2 secretion using multiplex immunoassays. CCL2 production and transcription were confirmed by flow cytometry and single-cell RNA-sequencing. CCR2 blockade was used on human PBMCs to evaluate the effects of CCL2/CCR2 signaling on T cell function alongside H1A treatment. Finally, in vivo antitumor activity of H1A and Atezolizumab were compared in two humanized PD-L1 tumor models. Results: H1A reduced the frequency of regulatory T cells and increased effector T cells, in whole PBMCs upon global T cell activation with anti-CD3/CD28. H1A enhanced the secretion and production of CCL2 from myeloid cells, which was confirmed at a transcript level. CCR2 blockade diminished CD8 T cells’ ability to differentiate into effector cells in H1A treated PBMCs. H1A demonstrated comparable therapeutic effects with Atezolizumab in treatment of immunogenic and ICI-responsive mouse tumor MC38, which is CD8 T cell-dependent, but showed superior therapeutic effects in treatment of ICI-insensitive mouse tumor E0771 compared to Atezolizumab. Both treatments elicited durable memory specific immune responses upon E0771 tumor rechallenge. Conclusion: H1A demonstrated a superior antitumor activity compared to Atezolizumab in our preclinical models via a new mechanism by which H1A caused degradation of intrinsic PD-L1 in myeloid cells. This causes more release of CCL2 to improve T cell function via CCR2 upon T cell activation. Thus, H1A is emerging as a new immunotherapy tool to target PD-L1/CMTM6 in myeloid cells for treatment of cancers that are refractory to current immune checkpoint inhibitor therapy. Citation Format: Michelle Alexandria Hsu, Xin Liu, Ying Li, Whitney Barham, Kevin Pavelko, Jake Hirdler, Fabrice Lucien, Haidong Dong. Targeting PD-L1/CMTM6 in myeloid cells results in elevated release of CCL2 to improve CD8 T cell-mediated antitumor response [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2664.

Abstract Although immune checkpoint blockade that targets the PD-L1 has revolutionized the treatment of human cancers, only a small proportion of patients positively respond to this form of treatment. A limitation of these therapies is that they only target and inhibit PD-L1’s extracellular function, which is to bind its receptor on PD-1 on T cells and inhibit T cell cytotoxicity. Recent discoveries have shown that PD-L1 expressed on tumor cells can signal intrinsically and independently of PD-1 ligation to promote tumor cell survival. Besides tumor cells, host myeloid cells have been shown to have high expression of PD-L1, therefore contributing to the immunosuppressive nature of the tumor. However, PD-L1 intrinsic signaling in human myeloid cells has not been studied. Therefore, the purpose of our work is to determine the intrinsic functions of PD-L1 in myeloid cells and how this impacts anti-tumor immunity. To study this, we used a novel PD-L1 antibody (H1A) that does not block PD-L1’s binding with PD-1, but directs PD-L1 for degradation, thereby decreasing total PD-L1 expression. Additionally, we developed a chimeric mouse model in which the murine PD-1/PD-L1 has been replaced with the human version. This allows us to evaluate PD-L1’s intrinsic function in vivo. In our preliminary studies, the degradation of PD-L1 in human monocytes led to increased chemokine release as well as increased conversion of effector T cells. Additionally, H1A treatment of tumor-bearing mice led to lower tumor burden in a CD8 T cell dependent manner. Based on these results, PD-L1 intrinsic signaling has the potential to regulate the immune-stimulatory response of monocytes, thereby limiting effector T cell responses against cancer. Supported by grants from NIH (R0192389) and Mayo Clinic graduate school of biomedical sciences

Abstract Though anti-PD-1/PD-L1 immune checkpoint blockade therapy has demonstrated robust anti-tumor immunity in human patients with advanced cancers such as melanoma, only a fraction of patients achieved such durable anti-tumor responses. Among all the non-responders, some patients responded initially but developed resistance to the therapy in the later treatments. This reflects an urgent need to understand the role of PD-1 signals in CD8+ T cells at priming and at effector stages during tumor development for the benefit of immunotherapy. The controversy of rapid tumor progression in PD-1 conditional knockout in regulatory T cells and tumor regression in PD-1 targeted deletion in regulatory T cells indicated germline PD-1 knockout mice might not be sophisticated enough to study the role of PD-1 in CD8+ T cells during tumor development. Thus, we produced a CD8 T cell-specific PD-1 knockout mouse model (CD8 E8ICre-PD-1fl/fl, also called CD8-PD-1 cKO mice) in which PD-1 is deleted in CD8+ T cells in the peripheral tissues. When CD8-PD-1 cKO mice were challenged with B16-OVA mouse melanoma cells, we found a delayed tumor growth and prolonged survival of CD8-PD-1 cKO compared to control mice. However, after 20 days of tumor challenge, the delayed tumor growth rebound in CD8-PD-1 cKO mice. In an OVA antigen immunization model, we found increased effector cells and tissue resident cells in the spleen of CD8-PD-1 cKO mice compared with control mice on day 7 after immunization. Taken together, our results suggest that PD-1 may restrain the expansion of effector T cells that contribute to the early phase of tumor control, but the augmented effector T cell response may cause an attrition of memory T cell pool that will be needed for long term protection. R01CA256927

Abstract Seven different anti–PD-1 and PD-L1 mAbs are now widely used in the United States to treat a variety of cancer types, but no clinical trials have compared them directly. Furthermore, because many of these Abs do not cross-react between mouse and human proteins, no preclinical models exist in which to consider these types of questions. Thus, we produced humanized PD-1 and PD-L1 mice in which the extracellular domains of both mouse PD-1 and PD-L1 were replaced with the corresponding human sequences. Using this new model, we sought to compare the strength of the immune response generated by Food and Drug Administration–approved Abs. To do this, we performed an in vivo T cell priming assay in which anti–PD-1/L1 therapies were given at the time of T cell priming against surrogate tumor Ag (OVA), followed by subsequent B16-OVA tumor challenge. Surprisingly, both control and Ab-treated mice formed an equally robust OVA-specific T cell response at the time of priming. Despite this, anti–PD-1/L1-treated mice exhibited significantly better tumor rejection versus controls, with avelumab generating the best protection. To determine what could be mediating this, we identified the increased production of CX3CR1+PD-1+CD8+ cytotoxic T cells in the avelumab-treated mice, the same phenotype of effector T cells known to increase in clinical responders to PD-1/L1 therapy. Thus, our model permits the direct comparison of Food and Drug Administration–approved anti–PD-1/L1 mAbs and further correlates successful tumor rejection with the level of CX3CR1+PD-1+CD8 + T cells, making this model a critical tool for optimizing and better utilizing anti–PD-1/L1 therapeutics.

Abstract Though anti-PD-1/L1 antibody therapies have generated unprecedented responses in a limited number of patients, the biological and immunological parameters that define their effective use are still largely unknown. The absence of appropriate pre-clinical models has limited our ability to compare the mechanistic differences between the approved PD-1/L1 antibodies in vivo and to determine the context in which each might be most effective. To address this critical need, we have produced humanized PD-1 and PD-L1 mice in which the extracellular domains (antibody binding sites) of mouse PD-1 and PD-L1 are replaced with the sequences of human PD-1 and PD-L1. The PD-1 and PD-L1 in the resulting “HuPD” mice can be recognized by FDA-approved antibodies, while the intracellular domains are kept intact for normal downstream signals. Following characterization of the new mouse model, in our first in vivo treatment studies, HuPD mice were immunized with OVA/Poly I:C in the context of FDA-approved PD-1/L1 antibodies and antigen-specific T cell production was quantified. Nivolumab, Pembrolizumab, and Atezolizumab all increased effector CD8+T cell production against the OVA antigen as expected given their immune stimulating function. Surprisingly, OVA-tetramer staining was unable to detect antigen-specific CD8+T cells in the spleens of immunized mice treated with Avelumab. This may reflect the unique ability of Avelumab to induce ADCC, and indicates that our model is capable of distinguishing key differences among the antibodies in vivo. Ongoing studies will identify how these differences might be exploited to render the most effective results for patients in a variety of tumor types and immune environments.

Abstract We utilized CD11a, an integrin up-regulated on effector and memory CD8+ T cells, as a biomarker to identify, measure, and monitor endogenous tumor-reactive cytotoxic lymphocytes (CTLs) in two mouse tumor models and in melanoma patients. CD11ahigh CD8+ T cells were detected in primary and metastatic tumor sites after tumor cell injection, as well as in spontaneous murine breast cancer tissues. In the peripheral blood of melanoma patients, tumor antigen-specific T cells were detected within the population of CD11ahigh CD8+ T cells that express high levels of PD-1. Infusion of lethally irradiated tumor cells failed to stimulate the expansion of CD11ahigh CD8+ T cells, suggesting that the access of tumor cells to target tissues is required to induce CD11ahigh CD8+ T cell responses. Phenotypic analysis revealed that the CD11ahigh CD8+ T cells detected in our studies were proliferating (Ki67+) activated T cells (CD62L-, CD69+). Although CD11ahigh CD8+ T cells were differentiated effector cells with CTL effector functions, they were typically unable to control tumor growth in our mouse models, possibly due to high expression of PD-1. Our results demonstrate that CD11a can be used as a biomarker to identify and track endogenous tumor-reactive CTLs.

Abstract Spontaneous antitumor immunity remains challenging to detect and monitor largely due to the absence of established biomarkers. We selected the integrin CD11a as a biomarker to define antitumor T cell responses. CD11a is up-regulated on effector and memory T cells and is involved in the killing of tumor cells by cytotoxic T lymphocytes. In primary (subcutaneously inoculated) and metastatic (intravenously injected) tumors, CD11a high CD8 T cells were detected in the primary (skin) or metastatic (lung) tumor sites, but not in secondary lymphoid organs. Once tumors were established, CD11a high CD8 T cells were identified in tumor draining lymph nodes and peripheral blood. Phenotype analysis revealed that CD11a high CD8 T cells were proliferating (Ki67+) activated T cells (CD62L low, CD69 high, PD-1 high). CD11a high CD8 T cells expressed the transcription factor T-bet but not Foxp3. Live, but not dead, tumor cells, induced expansion of CD11a high CD8 T cells in the lung, suggesting that the access of tumor cells to target tissues is required to induce CD11a high CD8 T cell responses. Our results suggest that CD11a may provide a new T cell biomarker to measure ongoing natural tumor-reactive T cell responses and to predicate the efficacy of tumor immunotherapy.

Abstract Seven different anti-PD-1 and PD-L1 monoclonal antibodies are now widely used in the US to treat a variety of cancer types, but there have been no clinical trials comparing them directly. Furthermore, because many of these antibodies do not cross react between mouse and human proteins, no preclinical models exist in which to consider these types of questions. Thus, we produced humanized PD-1 and PD-L1 mice in which the extracellular domains of both mouse PD-1 and PD-L1 were replaced with the corresponding human sequences. Using this new model, we sought to compare the strength of the immune response generated by FDA-approved antibodies. To do this, we performed an in vivo T cell priming assay in which anti-PD-1/L1 therapies were given at the time of T cell priming against surrogate tumor antigen (OVA), followed by subsequent B16-OVA tumor challenge. We found that anti-PD-1/L1-treated mice exhibited significantly better tumor rejection than controls, although both the control and antibody-treated mice generated comparable numbers of OVA-specific T cells at the time of priming. To determine what could mediate this strong antitumor immune response, we identified the increased production of CX3CR1+PD-1+CD8+ cytotoxic T cells in the anti-PD-1/L1-treated mice, the same subset of effector T cells known to increase in clinical responders to PD-1/L1 therapy. Thus, our model permits the direct comparison of FDA-approved anti-PD-1/L1 monoclonal antibodies and further correlates successful tumor rejection with the level of CX3CR1+PD-1+CD8 + T cells, making this model a critical tool for future studies to optimize and better utilize anti-PD-1/L1 therapeutics.