Graduation Year


Document Type




Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Biology (Cell Biology, Microbiology, Molecular Biology)

Major Professor

Brian Ruffell, Ph.D.

Committee Member

Amer Beg, Ph.D.

Committee Member

Jose R. Conejo-Garcia, M.D., Ph.D.

Committee Member

Conor Lynch, Ph.D.

Committee Member

Hatem Soliman, M.D.


immunotherapy, CD8 T cells, tumor immunology


Intratumoral CD103+ dendritic cells (cDC1) are required for anti-tumor immune responses. In tumors that are poorly responsive to immunotherapeutic approaches targeting T cells, targeting cDC1 represents an alternative approach that may be useful in improving patient response rates. As such, it is critical to understand cDC1 function within tumors, and what may be preventing optimal function of cDC1. TIM-3 is a receptor that is highly expressed by cDC1 in murine and human mammary tumors, and TIM-3 blocking antibodies are currently being evaluated in clinical trials for a number of solid and hematological malignancies. In order to best design combinatorial therapeutic approaches, it is important to understand the mechanism by which therapies act. This project aims to understand the mechanism by which combination therapy with paclitaxel (PTX) and TIM-3 blockade reduces the rate of mammary tumor growth.

In order to do so, we identify Cxcl9, Cxcl10, Cxcl11, and Il12b as being upregulated by cDC1 following PTX/TIM-3 blockade. By blocking signaling through CXCR3 (the receptor for CXCL9, CXCL10, and CXCL11) we determine that signaling through the CXCR3 axis was responsible for the observed reduction in the rate of tumor growth. We next evaluate whether the chemokines themselves are necessary for therapeutic response, using mixed bone marrow chimeras and diphtheria toxin depletion. In doing so, we identify CXCL9 production by cDC1 as a requirement for the observed reduction in tumor growth. As one role for CXCL9 has been shown to be mediating lymph node migration by CD8+ T cells, we then inhibit lymph node egress and use a fluorescently labelled tumor model to assess involvement of the lymph node following initiation of anti-tumor immunity. We did not observe changes in antigen presentation by cDC1 in the lymph node, or alterations in responsiveness to PTX/TIM-3 blockade when lymph node egress was inhibited. Together, these data point suggest that cDC1 may be interacting with CD8+ T cells within the tumor, and that these interactions may be supporting CD8+ T cell functionality.

To evaluate this, we first use flow cytometry to assess T cell production of IFNγ, as a measure of the effector function. We find that treatment with PTX/TIM-3 blockade increases IFNγ production by CD8+ T cells, and that this is increase is prevented when signaling through CXCR3 is blocked. We next assess the distance between CD8+ T cells and their nearest cDC1 in tumors treated with PTX/TIM-3 blockade as compared to a PTX/IgG2a control. We find that CD8+ T cells are located closer to cDC1 in tumors treated with PTX/TIM-3 blockade, without increases in either CD8+ T cell or cDC1 infiltration. We then use mixed bone marrow chimeras and diphtheria toxin depletion to evaluate the role of two main functions of cDC1 within tumors: antigen cross-presentation on MHCI and production of IL-12. We find that IL-12 production by cDC1 is required for response to PTX/TIM-3 blockade, while antigen presentation by MHCI is dispensable. Taken together, the data described herein suggests that treatment with TIM-3 blockade drives increased CXCL9 production by cDC1, bringing CD8+ T cells into closer proximity and increasing their exposure to IL-12, thereby supporting CD8+ T cell functionality.

Overall, these studies propose a mechanism by which TIM-3 blockade functions in mammary tumors and provide direct evidence for the ability of cDC1 to support CD8+ T cells within the tumor. Understanding these aspects has implications for the design of therapeutic strategies using antibodies against TIM-3.