Age is not just a number: Naive T cells increase their ability to persist in the circulation over time

Abstract

The processes regulating peripheral naive T-cell numbers and clonal diversity remain poorly understood. Conceptually, homeostatic mechanisms must fall into the broad categories of neutral (simple random birth–death models), competition (regulation of cell numbers through quorum-sensing, perhaps via limiting shared resources), adaptation (involving cell-intrinsic changes in homeostatic fitness, defined as net growth rate over time), or selection (involving the loss or outgrowth of cell populations deriving from intercellular variation in fitness). There may also be stably maintained heterogeneity within the naive T-cell pool. To distinguish between these mechanisms, we confront very general models of these processes with an array of experimental data, both new and published. While reduced competition for homeostatic stimuli may impact cell survival or proliferation in neonates or under moderate to severe lymphopenia, we show that the only mechanism capable of explaining multiple, independent experimental studies of naive CD4+ and CD8+ T-cell homeostasis in mice from young adulthood into old age is one of adaptation, in which cells act independently and accrue a survival or proliferative advantage continuously with their post-thymic age. However, aged naive T cells may also be functionally impaired, and so the accumulation of older cells via ‘conditioning through experience’ may contribute to reduced immune responsiveness in the elderly.

The natural history of naive T cells from birth to maturity

“…We confronted these models with 3 datasets relating to naive T‐cell homeostasis under healthy conditions or in partial lymphopenia; the kinetics of T‐cell numbers in both euthymic and thymectomized mice reported by den Braber et al, the kinetics of naive T‐cell replacement in busulfan chimeras of different ages, and the results of adoptive transfers of naive CD4 T cells from hosts of different ages, reported in Tsukamoto et al Only the adaptation model was able to simultaneously explain all 3 datasets (Figure ), with fitness increasing slowly on a timescale of roughly 100 days. This pace of accrual of fitness is somewhat at odds with the shorter timescales of RTE maturation, and indeed a model of a conveyor‐belt mechanism of RTE dynamics, a special case of adaptation in which all cells progress to maturity (and higher fitness) after a fixed time in the periphery, explains these diverse datasets poorly . The study also supports the conclusion of Tsukamoto et al that selection alone is unable to explain the trend in naive T‐cell survival with host age.…”

“…The expected residence times of naive CD4 and CD8 T cells (the average time taken after thymic export to leave the naive pool due to loss or differentiation) are two‐ or threefold shorter than this . Because total naive T‐cell numbers in mice fall only by a factor of 2 between 100 and 500 days of age, this simple analysis confirms that naive cells in adult mice are sustained largely by thymic export …”

“…The long‐term dynamics of the naive T cell pools in healthy adult mice can be described remarkably well by a model pairing a declining thymic source with constant rates of division and death,where λ is the net effect of loss of naive cells through death or differentiation and cell division . Ki67 is a nuclear protein that is detectable for 3‐4 days following cell division, and is detectable in roughly 4% of naive CD4 and CD8 T cells in adult mice .…”

“…Heterogeneity could also emerge progressively through selection or adaptation. In a pure selection scenario, natural variation in the fitness of cells exported from the thymus generates heterogeneity in the mature naive T‐cell pool which develops over an individual's lifetime through the accumulation of longer lived or more proliferative cells. If this fitness distribution derives from cell‐cell variation in the average affinity of the TCR for self‐peptide–MHC, the naive pool may become progressively enriched for strongly self‐reactive cells, potentially increasing the risk of autoimmune disease with age .…”

“…One can assess the support for different models using statistical criteria, but another test of a model's strength is its ability to explain multiple independent sets of observations. Taking this approach, a recent study of ours compared a suite of candidate models of naive T‐cell homeostasis (Figure ), describing constant rates of division and loss (equation 1), density‐dependent rates of division or death, adaptation, selection, and population heterogeneity with incumbent cells. We confronted these models with 3 datasets relating to naive T‐cell homeostasis under healthy conditions or in partial lymphopenia; the kinetics of T‐cell numbers in both euthymic and thymectomized mice reported by den Braber et al, the kinetics of naive T‐cell replacement in busulfan chimeras of different ages, and the results of adoptive transfers of naive CD4 T cells from hosts of different ages, reported in Tsukamoto et al Only the adaptation model was able to simultaneously explain all 3 datasets (Figure ), with fitness increasing slowly on a timescale of roughly 100 days.…”