Cross Priming of Transgene Product-Specific CD8+ T Cells in Hepatic AAV Gene Transfer Depends on IL-1 Receptor and XCR1+ Dendritic Cells but Not TLR9

Introduction: Adeno-associated virus (AAV) mediated gene transfer is currently evaluated in multiple Phase I/II and Phase III studies for the treatment of hemophilia. However, immune responses to both the AAV capsid and encoded transgene remain major impediments to clinical translation. Several studies have implicated innate immune sensors such as Toll-like receptors (TLR) and their downstream adaptor molecule MyD88 in sensing viral structures. TLR9-MyD88 signaling has been linked to cross-priming of CD8⁺ T cell responses to capsid and also to transgene product-specific CD8⁺ T cell responses. However, little is known about other signaling pathways that may lead to immune activation. Previously, our lab has shown that while liver gene transfer is capable of inducing immunological tolerance to AAV encoded transgene products, vector dose and design play a critical role. For instance, low hepatic gene expression levels may elicit a CD8⁺ T cell response to the AAV encoded transgene, resulting in loss of the model antigen ovalbumin (OVA) in C57BL/6 mice or of FIX expression in hemophilia B mice. We investigated innate immune sensing pathways that may play a role in initiating transgene specific CD8⁺ T cell response in the hepatic microenvironment. Further, we determined the contribution of hepatic antigen presenting cells (APC) by selectively depleting/neutralizing APCs and evaluating their effect on presentation of transgene product-derived antigen following AAV8-OVA liver gene delivery.

Methods: Wild-type (WT) C57BL6 and specific innate sensing knockout mice on the C57BL6 background were intravenously (IV) injected with a predetermined immunogenic dose (1x10⁹vg) of hepatotropic AAV8-OVA vector (Mol Ther 25:880, 2017). PBMCs were quantified at 4 weeks for OVA-specific CD8⁺ T cells using a class I MHC tetramer. Hepatic APC types [Kupffer cells, neutrophils, CD103⁺ dendritic cell (DC), CD11c⁺ DC, XCR1⁺ DC] involved in transgene specific CD8⁺ T cell activation were selectively depleted/inactivated by pre-treatment with gadolinium chloride (GdCl₃), Ly6G, CD103 antibody respectively, or by administering diphtheria toxin (DT) to CD11c-DTR and XCR1-DTR mice. This was followed by intravenous administration of AAV8-OVA and CellTrace violet labeled OT-1 cells.

Results: Similar to WT mice, TLR9^-/-, TLR2^-/-, TRIF^-/-, IFNaR^-/- and MDA5^-/- mice developed a CD8⁺ T cell response indicating that these sensors do not play a role in transgene specific CD8⁺ T cells response. Interestingly, adaptor protein MyD88^-/- mice did not elicit CD8⁺ T cell response to OVA, implying a MyD88 dependent but TLR9 independent response. Since MyD88 is an essential adaptor protein not only for TLR but also for interleukin-1 (IL-1) signaling pathways, we next analyzed IL-1R^-/- mice. Similar to MyD88^-/- mice, IL-1R^-/- mice did not show OVA specific CD8⁺ T cells (p=0.006, 0.007 respectively), indicating that transgene-specific adaptive responses are mediated by IL-1R/MyD88 signaling.

Kupffer cells and DCs are principal APCs in liver and infiltrating neutrophils could also act as APCs under inflammatory conditions in liver microenvironment. Using proliferation of OT-I cells as readout we tested if any of these cell types are required for presentation to transgene specific CD8⁺ T cells. In naïve control, GdCl₃ treated and a-Ly6G antibody treated mice, OT-I cell proliferation reached 60%, 65% and 48% on average, respectively. Depletion of CD11c DCs substantially reduced the proliferation of OT-I cells to ~6% (p<0.0001) indicating a critical role for DCs in mediating transgene specific CD8⁺ T cell responses. Since XCR1⁺ DCs are the major cross-presenting DCs and hepatic resident CD103⁺ DCs are shown to have intrinsically enhanced capacity to process and present antigen to naïve CD8⁺ T cells, we further sought to assess if any of these DCs plays a role in activation of transgene specific CD8⁺ T cells. Neutralization of CD103⁺ DCs reduced OT-I proliferation to 39% (p=0.01) whereas depletion of XCR1⁺ DCs reduced the proliferation to ~20% (p<0.0001) indicating a major role for XCR1⁺ DCs.

Conclusions: In summary, we uncovered a novel-signaling pathway that can activate CD8⁺ T cell responses during AAV gene transfer independent of TLR9 sensing. The IL-1R/MyD88 pathway drives activation of transgene specific CD8⁺ T cell, and XCR1⁺ DCs are critically involved in cross-presenting transgene product-derived antigen to CD8⁺ T cells.