The liver is characterized by a microenvironment that promotes immune tolerance, which can be exploited by gene therapy. In the case of hepatitis viral infections such as HBV or HCV however, ineffective or non-functional responses can result in chronic disease. To understand the parameters and mechanisms that govern the CD8+ T cell response to a virally encoded antigen expressed in the liver, we administered an adeno-associated virus expressing ovalbumin (AAV8-EF1α-ovalbumin) to immune competent C57BL/6J mice at various doses (low: 1×108 vg, medium: 1×109 vg, and high: 1×1010 vg). Interestingly, all doses resulted in sustained vector-dose dependent systemic ova expression, albeit with distinct immune profiles. At the low dose and the high doses, no response to ova was observed. At the medium dose, circulating ova-specific CD8+ T cells were detected in 40-50% of mice at high frequency (5-35%) by tetramer stain. These emerged within 1 month and, while declining in frequency over time, can persist for up to one year. Despite their inefficiency in eliminating vector-derived ova expression, functional assays demonstrated that induced CD8+ T cells retained cytolytic activity to SIINFEKL (the dominant ovalbumin MHC class I epitope in C57BL/6J mice) peptide-loaded target cells in vivo or in vitro, and that they produced Th1 type cytokines such as interferon-γ (IFN-γ) and tumor necrosis factor-α (TNF-α) in response to ex vivo stimulation with SIINFEKL epitope. Further phenotypic characterization revealed that most of these cells expressed PD-1 and 2B4. Other markers of exhaustion such as CD160, LAG3 and TIM3 were expressed to a lesser extent. Cells acquired memory phenotype over time as judged by CD44, CD62L, and CCR7 markers. Since higher vector doses failed to induce the CD8+ T cell response entirely, a potent antigen-dependent “off switch” has to exist in the liver. Previously, we and others identified FoxP3+ Treg, activation induced apoptotic cell death, and expression of the immune suppressive cytokine IL-10 as contributors to immune tolerance induction by hepatic gene transfer. Therefore, we used Foxp3-DTR mice (for transient depletion of Treg), FasL−/−, and IL-10−/− mice to determine if these pathways were required in preventing ova-specific CD8+ T cell activation in the high-dose group. Transient depletion of Foxp3+ Treg in Foxp3-DTR mice led to a low but detectable CD8+ T cell response in all animals. In contrast, only 1/5 (20%) of FasL−/− or IL-10−/− mice developed a response. Therefore, it is likely that multiple (and perhaps redundant) mechanisms work in concert to achieve high-dose tolerance to proteins expressed in the liver. These results have implications for viral infections of the liver and for gene and cell therapies, illustrating the potential for chronic albeit only partially functional T cell responses at low doses, which can be entirely prevented at sufficient antigen doses.