Intrinsically Hyperactive and Hyperproliferative CD8⁺ T Cells In Cmah-/- Mice as a Model of Human Gene Transfer Responses.

Abstract 3773

Adeno-Associated Virus is a popular gene transfer vector that successfully treats many genetic diseases in animal models, yet has limited proof of efficacy in humans. While experimental animals have been useful in predicting vector dosing, they failed to provide a model for human T cell responses to the vector. Humans differ from other mammals with regard to a specific inactivating mutation in the CMAH gene, which encodes a hydroxylase that converts a precursor sialic acid donor (CMP-Neu5Ac) to one with an extra oxygen atom (CMP-Neu5Gc). Sialic acids play multiple roles in the immune system, especially as the ligands for Siglecs, which are ITIM-containing receptors broadly expressed on immune cells. The evolutionary loss of the original preferred ligand (Neu5Gc) for these inhibitory receptors may have rendered human immune cells active at a lower stimulation threshold. For multiple reasons the human loss of Neu5Gc also caused evolutionary pressure at Siglec gene loci. This led to the current loss of Siglec expression on human T cells, which contrasts with Siglec expression on other mammalian T cells, specifically non-human hominids. Recent studies have highlighted the relative over-reactivity of human T cells when compared to chimpanzee T cells at equivalent stimulus. Here we used a Cmah-/- mouse which mimics human Neu5Gc loss to characterize T cell responses in this mouse and investigate it as a model for human rAAV T cell response. Previously we reported the relative hyperproliferation and hyperactivation of Cmah-/- T cells in bulk splenocyte culture incubated with T cell activator beads [Molecular Therapy. 2010:vol18:Supplement 1 pg S211]. Now we show that hyperactivation and hyperproliferation is a T cell intrinsic mechanism, that the response can be blunted when the cells are fed the missing sialic acid (Neu5Gc) in culture, and that Cmah-/- T cells produce more IFNg when stimulated with PMA or anti-CD3. Isolated CD8+ T cells from Wild-type (WT) and Cmah mice were stimulated for 3 or 5 days with anti-CD3/anti-CD28 beads. At day 3, a greater percentage of Cmah-/- CD8+ T cells had low CD62L expression when compared to WT CD8s (39.8% vs. 16.1%). CD25 expression was also slightly higher at day 3 in Cmah-/- CD8s (81.2% high vs. 70% high). At day 5, Cmah-/- CD8s had proliferated much more than WTs as measured by CFSE dilution (75.4% CFSE low vs. 9.91% CFSE low). Additionally, the percentage of CD8+ T cells CD62L low (62% vs. 8.44% WT) and CD25 high (42.5% vs. 19.5% WT) in Cmah-/- mice highlighted the fact that Cmah-/- T cells were hyperactivated compared to wild type mice. This CD8+ hyperproliferation was slightly blunted at low anti-CD3 stimulation (50ng/mL) when cells were incubated with the missing Neu5Gc sialic acid (4.14% CFSE low Neu5Gc fed vs. 9.14% CFSE low controls fed with Neu5Ac). There was a striking difference in CD69 expression among the CFSE low (highly proliferated) cells, with only 4% expression on Neu5Gc fed Cmah-/- CD8s vs. 25% CD69 expression on control Cmah-/- CD8s, suggesting hyperactivation of the proliferated subset of CD8+ T cells. Intracellular IFNg staining on CD8+ T cells from naïve Cmah-/- and WT mice showed significant differences in %IFNg+ cells when stimulated with anti-CD3 (3.225% vs. 1.14%, p<0.05) or PMA (46.5% vs. 22.48%, p<0.01), indicating a lowered threshold for cytokine production. Finally, in vivo I.M. immunization of both strains with AdAAV2 led to a higher proportion of IFNg+ T cells at 10 days reactive to the immunodominant AAV2 peptide in Cmah-/- mice vs. WT controls (20.9 fold more IFNg+ cells over media vs. 8.51 fold over media, p<0.05). We conclude that the absence of Neu5Gc results in a T cell that is intrinsically hyperactive and hyperproliferative compared to WT cells, likely altering Siglec ligand expression and/or other as yet unknown effects. In support of the latter, feeding human T cells with Neu5Gc gave a similar blunting of responses. The Cmah-/- mouse may be useful as a model of relative over-reactive human immune CD8+ cytotoxic T cells responses to the AAV capsid.