Accumulations of KIR⁺ T cells in the elderly

Mechanisms controlling the age-associated gradual acquisition by CD8 T cells of natural killer cell receptors negatively regulating lymphocyte stimulation have remained obscure. Now, data from Li and colleagues¹  imply that a state of transcriptional activation and histone stoichiometry in CD8 but not CD4 T cells from elderly people results in epigenetic control of KIR genes by a mechanism involving displacement of DNA methyltransferase 1.

Late-stage differentiated CD8 T cells accumulate in the elderly with a corresponding reduction of naive T cells. Pathogen exposure throughout life drives the differentiation of naive cells toward differentiated cells, which are required for maintaining specific immunologic memory at the expense of responses to new infectious agents.²  T-cell homeostasis ensures that the number of peripheral T cells remains constant; hence, accumulations of late-differentiated cells leave less “space” for other cells.³  Late-stage CD8 cells no longer express the costimulatory molecule CD28, but, as shown in earlier work by Goronzy et al, this may not represent a permanent differentiation marker.⁴  Later in the differentiation pathway, CD8 T cells begin to express molecules that mediate signals blocking T-cell proliferation; many of these belong to families of receptors expressed on natural killer cells but not on naive T cells. The regulation of the expression of these molecules, the signals responsible for their acquisition, and the functional consequences of their expression have remained obscure until now.

In this issue of Blood, work from Goronzy's group¹  goes some way toward dissecting the responsible mechanisms. They study the representative KIR2DL3 gene encoding surface receptor CD158b. They elegantly demonstrate that in CD8, but not CD4 cells, the KIR2DL3 promoter exhibits what they term age-associated patchy and stochastic demethylation, and hence activation. Moreover, they show that this is likely caused by decreased recruitment of DNA-methyltransferase 1 (DNMT1) to the promoter, and suggest that this may, in turn, be caused by displacement of DNMT1 due to the state of transcriptional activation and histone stoichiometry of CD8 cells. The big question here concerns the functional consequences of these events. The authors suggest that age-associated increased expression of negative receptors implies a contribution to immune defects in the elderly. However, there is an alternative explanation that does not view these events as pathological.

First, we need to ask whether these results are generally applicable. KIR2DL3 may not be typical of the range of negative receptors expressed by CD8 cells. However, Li et al specifically selected this gene because its promoter shares much of its sequence with other KIR promoters, so this seems unlikely. Second, their study design was cross-sectional, so by definition the young and elderly population may have been different from the beginning, and the observed effects not necessarily ascribed to aging but to variables such as medication, nutrition, and so forth. However, data from several other studies are consistent with greater expression of KIR and other negative receptors on CD8 cells from older persons, so this is also unlikely. Third, the proportions of later-stage differentiated cells are higher in older than in younger people, presumably due to greater pathogen exposure in the former. Therefore, Li et al sorted CD8 subpopulations and found that late-stage differentiated cells from young persons showed less demethylation than the same subsets from the elderly. This is a strong argument in favor of some age-associated component in addition to the large effect seen as a result of altered cell subset distribution due to the immunological history of the person. Although an accumulation of such cells may indeed contribute to immune defects in the elderly, as suggested by the authors, it may also be the case that they are actually crucial to survival. This argument is based on the fact that the one pathogen that most effectively drives the accumulation of these late-stage differentiated cells in humans is the persistent herpesvirus, cytomegalovirus.⁵  Because most elderly people are infected with this virus, it is difficult to separate its effects from those of aging. Where this has been done, it was concluded that it is primarily CMV that is responsible for driving these “age-associated” changes.⁶  Therefore, it may be proposed that the increased levels of expression of negative receptors by CD8 cells is in response to the necessity of maintaining these cells in an activated state as effectors responsible for CMV immunosurveillance. An antiproliferative, apoptosis-resistant, CD28⁻, KIR⁺ phenotype (commonly but perhaps erroneously assumed to reflect “senescence”) might, in fact, be protective and help to retain essential effector cells because CMV-specific cells constantly exposed to viral antigens might otherwise be stimulated to proliferate until they reached exhaustion and were clonally deleted. There is some published evidence from longitudinal studies that these CD8 clonal expansions occurring late in life are indeed predominantly CMV-specific, and that at the end stage of life, the loss of these clones is the immunological factor directly associated with mortality.⁷  Because younger persons may not yet have acquired CMV infection, or may not have cohabited with the virus for such an extended period, the majority of their memory cells will not be CMV-specific and might behave differently vis-à-vis KIR promoter demethylation. It would be fascinating to know whether and what fraction of the subjects studied by Li et al were CMV-positive and whether this alternative explanation is the more likely one.