According to a recent report in Science, CD8αα could be the long-sought marker of effector T cells that are destined to enter the memory pool. Only a small fraction of the effector T-cell population that expands during a primary response survives as long-lived memory T cells after antigen has been cleared, but until now, the mechanism of this selection process has been unclear.

Madakamutil et al. used tetramers of the non-classical MHC class I molecule thymic leukaemia antigen (TL) — which is a unique ligand for CD8αα — to show that mature CD8αCD8β+ T cells upregulate cell-surface expression of CD8αα after stimulation through CD3. CD8αα upregulation was also seen after stimulation of OT-1 T-cell receptor (TCR)-transgenic splenocytes with their cognate antigen (ovalbumin, OVA). This effect was greatest when OT-1 splenocytes were stimulated with OVA presented by TL-expressing antigen-presenting cells, indicating that interaction between CD8αα and TL is necessary to maintain CD8αα expression. TL is expressed by activated dendritic cells and monocytes, indicating that such an interaction could occur physiologically.

Expression of CD8αα correlated with high levels of expression of the anti-apoptotic proteins Bcl-2 and Bcl-XL and with increased T-cell survival in vitro. These results were confirmed in vivo after infection of mice with lymphocytic choriomeningitis virus (LCMV); CD8αα was upregulated on a subset of LCMV-specific CD8αβ+ T cells after infection, and these T cells expressed high levels of Bcl-XL and survived long term (for up to 60 days after infection).

As CD8αα is only upregulated on a subset of effector T cells and promotes long-term survival, it might be a marker of memory T-cell precursors. To test this, they transferred TCR-transgenic P14 T cells (specific for an LCMV peptide) to naive recipients, then re-isolated the P14 T cells 8 days after LCMV infection. CD8αα+ and CD8αα− P14 fractions were then transferred separately to naive recipients, and the mice were challenged with LCMV 40 days later. Only those mice that had received the CD8αα+ fraction had a significant memory P14 T-cell response to the virus, and their T cells produced interferon-γ in vitro when stimulated with LCMV peptide. An enhancer deletion in the mouse Cd8 locus that prevents high-level expression of CD8α to form CD8αα homodimers, but still allows CD8αβ heterodimer formation, did not affect the primary response to LCMV but led to low numbers and a poor in vitro response of LCMV-specific T cells 50 days after infection, which confirms the essential role of CD8αα in T-cell survival and memory-cell differentiation.