The (gradual) rise of memory inflation

Abstract

Memory inflation, as a term, has been used for 15 years now to describe the longitudinal development of stable, expanded CD8⁺ T memory pools with a distinct phenotype and functional profile which emerge in specific infection and vaccine settings. These settings have in common the persistence of antigen, especially cytomegalovirus infection but also more recently adenoviral vector vaccination. However, in contrast to chronic infections which lead to "exhaustion" the repeated antigen encounters experienced by CD8⁺ T cells lead to development of a robust T-cell population structure which maintains functionality and size. In this review, I will discuss how the ideas around this form of memory have evolved over time and some new models which can help explain how these populations are induced and sustained. These models are relevant to immunity against persistent viruses, to novel vaccine strategies and to concepts about aging.

Keywords: CD8+ T cell; adenovirus; cytomegalovirus; exhaustion; inflation; memory; vaccination.

Figure 1

A pictorial “definition” of memory inflation. The definition includes the framing of the phenomenon within the context of antigen persistence and also includes the fact that not all T cells included in the response to a given epitope will have an “inflationary” phenotype. The 2 negatives in the definition (lack of full contraction and lack of exhaustion) could still be improved if a transcriptional programme can be defined more accurately which underpins the distinct phenotypic and functional profile seen

Figure 2

Memory inflation and viral control. The infections discussed represent different examples on a continuum of viral replication dynamics. The inflationary response can vary between an essential role in virus suppression to effectively a bystander role

Figure 3

Antigen presentation and memory inflation. The infection/transduction of a stable cell type which is not a dendritic cell is suggested by different experimental data in MCMV and adenoviral models. Priming requires cross‐presentation and leads to generation of multiple responses to antigens which are not expanded during the memory phase—only those antigens presented on the unconventional APC drive memory inflation. Whether priming can additionally occur on the original APC is not fully defined, but is likely to be much less efficient

Figure 4

Models for memory inflation. The original model is based on 2 populations of cells, 1 in lymph nodes and the rest expanded in the periphery. Model 2 includes an intermediate cell type which does not possess lymph node homing markers but retains proliferative capacity. Exposure to antigen on a non‐classical APC in the periphery (at sites to be defined fully in the MCMV and adenovirus models) could drive proliferation and differentiation of these cells. They may also self‐renew or potentially de‐differentiate toward the long‐lived central memory pool