a decrease in the CD56brightCD16− NK cell subset and an expansion of CD56dimCD16+ NK cells [75]. Considering that CD56brightCD16− NK cells have a high capacity to produce different cytokines and chemokines in response to cytokines released by other activated immune cells, their decreased proportion may be responsible for the defective overall production of cytokines and chemokines by NK cells stimulated with IL-2 or IL-12 observed in elderly individuals including nonagenarians [76, 77]. On the contrary, old individuals show an increased production of granzyme A and IFN by CD56bright cells, potentially representing a compensatory mechanism to maintain the protective and immunoregulatory role of these cells in older individuals [70]. The decrease in old individuals of the CD56bright subset is due to a decreased output of these immature NK cells from the bone marrow to the peripheral blood reflecting possible alterations in hematopoietic stem cells [78]. CD56dim CD16+ NK cell subsets are associated with decreased granzyme A but no change in GrzB and perforin levels sustaining their cytotoxicity either directly or through CD16 which does not change with aging [71]. Therefore, the observed decrease in cytotoxicity at the single cell level is due most probably to the defect of the recognition of the target cells as a consequence of the decreased expression of the NK cell-activating receptors observed in aged individuals [79–81].
Fig. 1. Phenotype and functional changes of NK cells in relation to their maturation, aging, and influence of exposure to viruses (CMV) and vaccination (VZV). Major receptors defining NK subpopulations are shown. More intense color of some receptor symbols indicates higher expression in some subpopulations; aging-associated changes in the levels of expression of surface molecules are not shown. Thick arrows inside the cell symbols and in the boxes describing major functionalities indicate decreased or increased proportions (black) and functionalities (light grey) of specific populations associated with aging and/or CMV infection. Refer to the main text for detailed description.
Furthermore, the effect of aging on the expression (and, likely, function) of NK cell receptors has been recently described [82]. Although there was no alteration in the CD16 expression or function in the total NK cells, expression of the activating natural cytotoxicity receptors NKp30 and NKp46 and of costimulatory molecule DNAX accessory molecule (DNAM)-1 was decreased [17, 18, 71, 79, 83, 84]. Analysis of NK cell inhibitory receptors showed an age-related increase in KIR expression and a decrease in CD94/NKG2A expression, although discrepancies could be found in different studies [17, 73].
Concomitantly, CD57 expression is increased with aging, which highly suggests a shift towards NK cell maturity with aging; however, it is of note that these changes cannot be always distinguished from chronic viral infections especially from that caused by the presence of CMV [85–87]. These CD57+ terminally differentiated NK cells have low proliferative capacity and have low responses to cytokines but are highly cytotoxic and produce high levels of IFNγ in response to CD16 ligation but only in CMV-infected old individuals [18]. As a functional consequence, NK cytotoxicity against classic NK cell targets was impaired especially on a single cell level as correlated with the decrease in NCR receptors such as NKp46, but antibody-dependent NKCC was not affected by aging. The expression of these NCRs, especially NKp46, was inversely proportional to the total number of NK cells [18]. The NKp46 receptors may recognize not only the foreign epitopes but also autologous ones, mediating some autoimmune effect such as seen in type 1 diabetes [88]. This regulation seems to be maintained in aging as we witness an increase in the total NK cell number with a decrease in NKp46 expression reducing some autoimmune conditions in aging. In the meantime, these changes may explain the higher incidence of infectious diseases (especially viral infections), malignancies, and atherosclerosis-based cardiovascular diseases [17, 89–92]. The gender differences in old individuals are controversial, with some studies reporting an increased functionality in aged women and a higher maturation state and lower NK cell activity in elderly men, but others have demonstrated no changes with gender during aging [18, 76].
Functional changes of NK cells may be also important for the physiological aging process [25, 89, 93]. Numbers of senescent cells in different tissues increase with aging and are considered among the major causes of aging (either directly or indirectly by the process of inflammaging) [94, 95]. NK cells may be major players in the elimination of senescent cells [25, 96]. It was shown that NK cells directly eliminate senescent cells, and as such they may contribute to maintenance of what is called the health span. NK cells eliminate senescent cells through the granule exocytosis (perforin/granzyme) proapoptotic pathway [97]. As these cytotoxic functions decrease with age, the elimination of senescent cells is also decreased. Moreover, senescent cells express increased levels of the nonclassical MHC-class Ib molecule HLA-E [25]. HLA-E has the capacity to inhibit the NK cytotoxicity by interacting with the inhibitory receptor NKG2A. This receptor is increased on senescent cells by IL-6 [25].
A few studies have addressed the mechanistic changes underlying the decreased NKCC with aging. Granule exocytosis is the predominant mechanism utilized by NK cells to eliminate their targets. It was proposed that a defective perforin secretion into the immunological synapse mediates the reduction in NKCC that accompanies physiological aging. This age-associated impairment in perforin secretion was associated with defective polarization of lytic granules towards the immunological synapse [98].