Regulation of CD45 isoforms during human effector and memory CD8 T cell differentiation: Implications for T cell nomenclature

This study shows a detailed analysis of the expression of CD45 isoforms on virus specific CD8 T cells after YFV vaccination (YFV-17D) longitudinally starting at two weeks after infection going out to over 10 years. As expected, effector CD8 T cells at day 14 express CD45RO but interestingly within 4-6 weeks all of the virus specific CD8 T cells become CD45RA and then remain CD45 RA for > 10yrs. The journey for these YFV specific CD8 T cells goes from naïve (CD45RA+ CCR7+) to effector/EM (CD45RO+ CCR7-) to TEMRA (CD45RA+ CCR7-) to TSCM (CD45RA+ CCR7+). These YFV specific CD8 T cells rarely acquire the canonical Tcm phenotype (CD45RO+ CCR7+). The few CM cell found in the YFV CD8 T cell response are effector cells in the first 11 days of the infection and their numbers become almost undetectable in subsequent samples. We found that CD8 T cells returned to expressing a mixture of CD45 isoforms similar but not identical to those found in naive T cells. CD45RA expression correlated closely with reduced expression of the splicing mediator hnRNP LL. This CD45RO to RA switch happens around one month post-infection and coincides with clearance of YFV, so we hypothesized that antigen may be playing a role in which CD45 isoform is expressed by the virus specific CD8 T cells. We addressed this issue by ex vivo analysis and provide evidence that this switch is indeed regulated by antigen. Sorted CD45RO positive YFV specific effector CD8 T cells re-express CD45RA when cultured ex vivo in the absence of antigen. We then show that SARS-CoV-2 spike specific CD8 T cells undergo the same CD45RA to RO to RA transition in vivo in a longitudinal analysis of a cohort of COVID-19 patients receiving the SARS-CoV-2 mRNA vaccine. These studies have important implications for T cell nomenclature and in particular for defining human memory CD8 T cell subsets. The re-expression of CD45RA in CD8 T cells likely explains the large number of TEMRA and absence of CM CD8 T cells. However, comparison to CD4 populations suggests that these findings are only applicable to CD8 T cells.