|Home | About | Journals | Submit | Contact Us | Français|
Studies comparing chronologically “young” versus “old” humans document age-related decline of classical immunological functions. However, older adults aged ≥65 years have very heterogeneous health phenotypes. A significant number of them are functionally independent and are surviving well into their 8th–11th decade life, observations indicating that aging or old age is not synonymous with immune incompetence. While there are dramatic age-related changes in the immune system, not all of these changes may be considered detrimental. Here, we review evidences for novel immunologic processes that become elaborated with advancing age that complement preserved classical immune functions and promote immune homeostasis later in life. We propose that elaboration such of late life immunologic properties is indicative of beneficial immune remodeling that is an integral component of successful aging, an emerging physiologic construct associated with similar age-related physiologic adaptations underlying maintenance of physical and cognitive function. We suggest that a systems approach integrating immune, physical, and cognitive functions, rather than a strict immunodeficiency-minded approach, will be key towards innovations in clinical interventions to better promote protective immunity and functional independence among the elderly.
In line with the damage paradigm of aging , studies comparing young and older persons, and their rodent counterparts, document age-related decline of classical immune functions. These include insufficiencies (a) in T cell receptor (TCR)-driven activation of T cells [2–4]; (b) in B cell activation, immunoglobulin (Ig) isotype switching, and humoral responses [5–7]; (c) in natural killer (NK) cell cytotoxicity [8, 9]; and (d) in innate cell functions [10–12]. Such observations are consistent with the idea of “immunosenescence”, a concept articulated by Walford in the late 1950s . By most accounts, the current terminology of immunosenescence refers to comparatively lower magnitudes, rather than absence, of classical T-/B-/NK-/innate cell-mediated responses among older adults (herein defined as persons aged ≥65 years old) relative to higher magnitudes of similar responses seen among younger adults. In recent years, an added aspect of immunosenescence is the observation about the unusual persistence of low levels of cytokines and acute phase reactants in circulation, e.g. interleukin (IL)-6, tumor necrosis factor (TNF)-α, and C-reactive protein (CRP) even in the absence of active inflammatory disease. This latter observation led to the coinage of the term “inflammaging” . In this regard, epidemiological studies indeed show associations of certain humoral factors with disability among older adults .
Nevertheless, old age or aging per se may not be synonymous with poor health or with immune incompetence. Older adults have very wide heterogeneity of health and immune phenotypes. They range from the chronically ill residents of nursing home/assisted-living institutions, to the community-dwellers who are highly functional despite history of chronic diseases and/or concurrent clinical conditions [16, 17]. Many older adults retain the capacity to mount vigorous immune responses as exemplified by their protection from the 2009 pandemic H1N1 influenza [18–20], and by the indicated efficacy of the varicella zoster vaccine [21, 22]. There is also mounting evidence that age may no longer be considered a barrier for organ transplantation from the perspective of either the recipient or the donor [23–25]. These observations suggest distinct mechanisms of immune homeostasis in old age that could either be determined in utero, and/or those elicited during childhood or in mid-life that are then maintained until the ultimate end of life. And with increasing evidences for the plasticity of immune functions [26–30], reappraisal of the notions of immune competence or incompetence in old age is warranted. Here, we draw lessons from clinical and epidemiological studies about maintenance and/or declines in the physical and cognitive function of older adults that are shaping the concept of successful or exceptional aging. In line with the general evolutionary principle that immunity is vital to individual fitness and survival , we suggest that definition(s) of successful aging needs to consider immune processes that are operational, if not optimal, in old age.
To better understand what might or might not constitute successful aging, it is fitting to emphasize a fundamental concept that aging is a biological process, rather than the late stages or later years of life. The process of aging encompasses the lifespan, rather than passive survival after reproductive maturity. In this context therefore, the physiology of old age is considered part of the continuum of life processes from postnatal development, to reproductive maturation (or adolescence), to adulthood, and to late life, hence the term “aging across the lifespan” .
This developmental view of aging is founded on observations that different organ-systems undergo varying rates of growth and retrogression [33, 34]. Organ/tissue retrogression, which has been intuitively thought to be a result of chronologic aging, may occur at any age or stage of life even when the individual is very highly functional. Examples are the rapid loss of insulin-like growth factor (IGF)-2 and cartilage growth immediately after birth, the cessation of bone and brain growth by the age of ~20 years, the cessation of oogenesis and mammary function by age ~40 years, and the near complete involution of the thymus by age ~50 years [33–38]. Moreover, some forms of tissue/organ damage/retrogression are programmed events of ontogeny; the best example is the massive tissue damage/loss that occurs during metamorphosis of frogs . For mammals, examples of programmed tissue damage are osteoclast destruction of bone as an essential part of osteogenesis , and apoptotic and autophagic cell death in tissue/organ homeostasis .
Thus, a view that the aging process is part of development invokes that tissue damage or retrogression may not be a determining event of aging, nor is tissue damage an essential phenotype of aging. Clearly, this is an antithesis of the damage theory of aging, which posits that aging and aging phenotypes are due to the accumulation of cell/tissue damage after reproductive maturity/maximal fecundity . In this regard, damage primarily due to reactive oxygen species (ROS), i.e. the oxidative damage theory of aging, has been argued and extensively studied for over half century since it was first proposed by Harman in the 1950’s  as a proximal cause or trigger of aging. However, this idea has recently come into question . Summative analysis of genetic studies of various strains of the laboratory mouse Mus musculus and the nematode Caenorhaditis elegans show that targeted deletion or over expression of genes encoding for various antioxidant factors have no significant impact on health outcomes of aging [43–45]. In fact, deletion of essential antioxidants such as SOD and clk-1 extend lifespan of mice and worms even though there is significant oxidative stress in such animals [46–48], indicating that ROS-mediated damage may not be critical in the initiation or progression of the aging process, or in the elaboration of aging phenotypes.
Emerging concepts on biology of aging therefore have lesser emphasis on wear and tear of tissues . There are now mounting evidences that tissue repair and regeneration pathways are naturally selected across the evolutionary scale [50, 51]. Thus, the aging process may be better viewed as the balance between consumption and replenishment of physiologic reserves. And so the rate of aging may not be synchronous with chronologic time. Phenotypes of aging, and the rates at which such aged phenotypes are elaborated, could depend on the size of the physiologic reserve at birth, and how robust the reserve is replenished/replaced thereafter .
In the last century, there has been an increasing trend in number of people living longer and have favorable health into their 8th–11th decade of life. This phenomenon has been argued to be a result of continued improvement in the standard of living propelled by better medical care. And so it is thought that better lifestyle choices and preventive medical care from childhood and onwards will be a way to alleviate future costs of care of the elderly in the 21st century and beyond .
However, medical care/lifestyle choice may not fully explain such increases in survival age and good health in late life. In 1928, Warthrix  noted that “what modern medicine has accomplished along the lines of hygiene and the prevention of disease has been to increase the number of individuals, both the fit and unfit – unfortunately too many of the latter kind – who come to maturity” and reach old age. Studies by Zheng et al  on the survival patterns of 42 countries over a 50-year period show that developing countries of Central America and Southeast Asia have in fact higher trajectory for longevity despite their poorer healthcare systems and overall more disadvantageous physical and socio-economic environments compared to industrialized countries such as the US, Japan, and European Union member countries. More significantly, the recent work by Barzilai and colleagues  show that lifestyle factors of centenarians and supercentenarians (i.e. those older than 100 years) are not significantly different from the general elderly population. This is a landmark finding in Biogerontology indicating that longer and healthier life of humans may be governed by yet unappreciated biological factors, which may or may not be modifiable by medical care/lifestyle.
The idea of successful aging, or aging well, or healthy aging, or exceptional aging comes from clinical studies describing the quality of life of elders. Although a consensus definition has yet to be established, the study by Christensen et al  on prospective health and performance assessment of Danish elders who were born in 1905 illustrates a framework for a working definition of successful aging as maintenance of functional independence. Function in this case is determined by objective measures of physical and cognitive performance. Physical function is measured by difficulty in performing basic activities of daily living (ADL), namely, transferring, toileting, bathing, eating, and dressing. Cognitive function is measured by the minimental state examination (MMSE) that scores global cognitive ability. This study shows that late-life independence is achievable in spite of a long history of disease and/or concurrent medical conditions.
Further improvements to define components of functional independence are being made by large cohort studies such as the Cardiovascular Health Study (CHS), a multicenter long-term observational study of aging for more than 20 years [57, 58]. CHS has developed and implemented standardized tests of physical and cognitive function that included various continuous measures such as gait speed, grip strength, MMSE, digit symbol substitution test, physical energy expenditure, ADL difficulty score, cardiovascular risk factors (carotid ultrasound, CRP, high/low density cholesterol, triglyceride), and physician diagnosis of incident disease. A comprehensive summary of all the findings of CHS from the various participating centers is beyond the scope of this paper. However, recent studies by Newman and colleagues [17, 59] on the survivors of the cohort CHS who are now in their 8th–9th decade of life show that measurement of physical and cognitive function, rather than a simple count of co-morbid conditions is a key component for a definition of successful aging or favorable health status in old age. Similar conclusion may be drawn from other studies such as the Long Life Family and the Framingham studies [60, 61], the Swedish Centenarian study , and the Honolulu Heart study .
A common theme of prospective cohort studies is that successful aging pertains to community-dwelling older adults. Residents of nursing homes and assisted-living communities generally have long-term, and usually severe, disabilities that preclude functional independence and so they have not been considered successful agers [64, 65]. A translational research challenge however, is the validation of particular physical and cognitive functional measures that might be easily adaptable in the routine medical care of the elderly. In this regard, a recent study by Studenski and colleagues  involving nine cohorts has shown that gait speed is a strong predictor of long-term survival among older adults.
Maintenance of functional independence into old age likely has a genetic basis vis a vis findings that long life runs in families [67, 68]. Polymorphisms in specific gene loci, such as the components of the signaling pathways for insulin/IGF system (IIS) and for pituitary hormones, have been identified as determinants of human longevity [69–72]. Remarkably, some genes are not only linked to lifespan extension, but also to better health characteristics in old age (hence the term “healthspan”). Among the best examples are allelic variants of cholesterol ester transfer protein that are associated with maintenance of cognitive function among centenarians [73–75]. Other examples are the various polymorphisms within the IGF gene domain that are associated with overall maintenance of physical activity, high muscle performance, and fat-free muscle mass among older adults [76–78].
Studies in mice support a genetic basis of successful aging and longevity. An example is a mouse strain expressing a transgene encoding for the cytoplasmic form of phosphoenolpyruvate carboxykinase (PEPCK-C). PEPCK-Cmus mice have high physical endurance; they are capable of running 5 km at a speed of 20m/min without stopping even among old mice . Another example is a mouse strain with targeted deletion of pregnancy-associated plasma protein A (PAPPA). PAPPA−/− mice not only have >30% extended lifespan compared to wildtype mice, but they also have significantly reduced tumor burden and have functional thymus even in their extended old age [80–82]. While PAPPA is a component of the IIS pathway found only in mammals, homozygous deletion or heteroallelic mutations of certain IIS homologs (e.g. daf-2, chico, etc) in invertebrates have been shown to extend lifespan and reproductive span of C. elegans worms and Drosophila fruit flies [83, 84]. There is now a large body of evidence from studies on yeast to mammals indicating conservation of certain genes linked to lifespan extension, referred to as longevity-assurance genes . Such genes are also thought to be important for the maintenance of population structure; the older individuals of the species facilitate reproductive fitness of the younger generation [86–88].
Due to their long lives, centenarians by default have been considered as models of successful aging . Longitudinal studies show that those who reach the 10th–11th decade of life have significantly shorter history of hospitalization, and have less severe clinical complications than their counterparts who died at younger ages . While centenarians may exhibit overt anatomical abnormalities in various organ systems, significant proportion of this very old population remains cognitively and physically functional at measurable levels comparable to their counterparts who are aged 60 to 90 years old [91–94]. Centenarians and supercentenarians have therefore become attractive subjects to examine immunologic features of extreme old age.
In line with the notion of inflammaging, there has been interest in evaluating relevance of genetic polymorphisms of IL-1, IL-6,and TNF-α. Results however are inconclusive; polymorphisms of these cytokine genes have not been reproduced in multiple populations. Analyses on the IL-1 gene cluster show no significant relationship of IL-1 allelic variants with actual levels of IL-1 production, age, and lifespan [95, 96]. Similar studies on TNF-α gene family show that while certain polymorphisms are associated with dementia among centenarians , many of these TNF-α variants have no apparent relationship with serum TNF-α levels in either young or old persons . Likewise, studies on IL-6 genotypes show conflicting results. Some studies show an association between IL-6 variants with increased production of IL-6 by mononuclear blood cells with age [99, 100], but other studies show no link with healthy aging or longevity [101–103].
Perhaps the most attractive notion for immune gene determinant(s) of longevity stems from the analyses of the human leukocyte histocompatibility antigen (HLA) multigene family because products of this family are central to antigen-specific immunity. Particular HFE alleles of class I HLA molecules have been associated with longevity of Sicilians and Sardinians [104, 105]. Several class II HLA molecules have also been identified, with HLA-DRB1*15 allele reportedly as the most significant longevity-associated HLA gene in these Italian populations [106–108]. Other alleles of the DRB1 family such as HLA-DRB1*11 and HLA-DRB1*1401 have also been found as longevity-associated genes in Caucasoid Mexican  and Okinawan  centenarians. Although more investigations are needed to ascertain whether there are distinct or shared HLA genotypes between different ethnic/racial groups, these studies underscore the central role of HLA gene products in immune responses. Since HLA molecules are known drivers of intrathymic maturation of T cells, and are required for maintenance of the peripheral T cell repertoire, it is more likely that longevity would be linked to an extended HLA haplotype(s), rather than to a single HLA allele. Interest on the role of HLA in longevity has been bolstered by studies showing relationships between HLA molecules, odor discrimination, mate selection or avoidance, and fertility [111–115]. Population genetic studies do show certain HLA genes and genes encoding olfactory molecules are in linkage disequilibrium [114–116], indicating tenability of a notion of an HLA-olfaction haplotype of longevity.
Complementing genetic studies, results of biological studies show novel aspects of the immune system of centenarians. In contrast to known age-related decline in the number of hematopoietic stem cells (HSC), centenarians have functionally active CD34+ HSC . Consistent with this observation, centenarians may also have robust naïve T cells characterized by high expression levels of CD127, CD45RA, and T cell receptor (TCR) excision circles (TRECs) [118, 119]. Such naïve T cells also appear to maintain their ability undergo mitosis given proper stimulation via the TCR and the CD28 costimulatory receptor . A naïve T cell pool among centenarians is clearly a physiologic deviation from that observed during normal chronologic aging. With advancing age, there is progressive involution of the thymus (indicated by reductions in thymic volume and TREC levels), and the naïve T cell pool becomes progressively depleted resulting in smaller size of the naïve T cell reserve, and in severe contraction of the TCR repertoire among 65–75 year olds .
Whether or not presence of TRECs and naïve T cells in centenarians translate into some degree of maintenance TCR repertoire diversity remains to be evaluated. And although mechanism(s) underlying the presence of a naïve T cell pool among centenarians has also yet to be elucidated, CD127 expression on the centenarian naïve T cells has been associated with systemic up regulation of IL-7 . IL-7 is known to drive normal thymus organogenesis and T cell lymphopoiesis, however it is not yet clear whether or not centenarians have functional residual thymus, or if they have a neothymopoietic tissue that is maintained by IL-7. A plausible alternative explanation is a genetic, or epigenetic, predisposition among centenarians to maintain long-lived naïve T cells. This latter suggestion is in line with a finding that centenarians generally have naïve T cells that are significantly less susceptible to apoptosis compared to naïve T cells of younger adults . For some centenarians, apoptosis-resistant naïve T cells is associated with polymorphisms of Fas and Fas ligand genes that attenuate expression levels of their otherwise pro-apoptotic protein products . A genetic/epigenetic argument about maintenance of a naïve T cell pool in old age is bolstered by studies in certain strains of old mice (aged ≥18 months) that have residual low level production of naïve T cells , and that such naïve T cells of aged mice tend have longer cellular lifespan than similar cells of younger mice . Irrespective of the underlying mechanisms, all these studies indicate a role of naïve T cells in immune function of centenarians.
There is limited information about features of other compartments of the immune system of centenarians. For B cells, centenarians carry much higher numbers of naïve IgD+CD27− cells compared to their younger counterparts [126, 127], contrary to well-documented decline in B cell lymphopoiesis with normal chronologic aging. Whether or not this is related to cumulative age-related deficits of Ig isotype switching , or to a possible protective role of naïve B cells in centenarians is not known. The notion of “protective” naïve B cells is consistent with lower levels of de novo autoantibody production among centenarians compared to adults aged 50–75 years old .
In the NK cell compartment, certain groups of centenarians have been found to carry higher numbers of NK cell subsets compared to either octogenarians or young people . These NK cells maintain their cytolytic activity, which has been linked to Zinc metabolism that is otherwise impaired with normal aging . The increased numbers of specific NK cell subsets is associated with the increased numbers of naïve IgD+CD27− B cells . Whether centenarians have a distinct NK cell - B cell regulatory circuit has yet to be examined.
Contrary to the notion of inflammaging, highly functioning centenarians have significantly higher levels of serum IL-6 than similarly functioning octogenarians and nonagenarians, or even younger adults [131, 132]. A proffered explanation to this latter paradox is a possible counter balancing effect of IL-10 as anti-inflammaging factor . This suggestion was inferred from findings that IL-10 gene polymorphisms linked to higher serum IL-10 levels are inversely associated with lower serum IL-6/TNF-α levels due to IL-6 or TNF-α gene polymorphisms [98, 134, 135]. However, this is an unsatisfactory explanation because IL-6 signaling generally leads to higher IL-10 production [136, 137], and that IL-10 has both pro- and anti-inflammatory activities . In view of the fact that IL-6 also has both pro- and anti-inflammatory activity, an interesting possibility that needs to be investigated is that systemic elevation of IL-6 might be a permissive cytokine environment upon which immune responses are generated in centenarians.
The phenomenal functional resilience of centenarians coupled with the identification of longevity-assurance genes and the immune characteristics of centenarians (contrary to the notion of age-related immune deficiency discussed above) suggest programming of biological pathways that promote good health in old age. Indeed, phylogenetic and population structure studies indicate evolutionary conservation for the association of certain biochemical processes such as sugar catabolism and cellular detoxification pathways with functional performance . In addition, long term follow up of populations of older adults show evidences for adaptive capacity of certain physiologic systems and promote long term survival beyond reproduction .
An example of such age-related adaptations occur in neuromuscular junctions (see review by Deschenes  and the references therein). While aging is associated with muscle fiber atrophy and reduced number of motor units, there is compensatory increased size of the average motor unit, as well as increased innervation and increased number of acethylcholine receptors per motor unit. Experimental studies show that age is the single most important factor associated with the expansion of slow-twitch and fast-twitch neuromuscular junctions . The physiologic end result is a greater strength of motor activity in the individual aged motor unit relative to the same size of the younger motor unit . Compensatory adaptations in the motor endplate help explain muscle strength among community-dwelling older adults [141, 144].
Similar age-related adaptations in the nervous system have been reported. Despite overall shrinkage of brain volume with age, Rosano and colleagues  have shown compensatory increased functional activity in the right posterior parietal cortex among elders with high cognitive function. Kantarci et al  also showed that memory, language, attention, and executive functions among elders without dementia are correlated with increased levels of activity of the cingulum of posterior cortex indicating this area of the brain as the main connectivity hub for cognitive brain networks. Similarly, Pelliciari et al  have reported that the aged somatosensory cortex has unusually plastic property. By paired associated stimulation testing, it was found that high cognitive function among elders was related to significant increase in N20-P25 activity compared to healthy young adults, indicating maintenance specialized neuronal circuits that otherwise diminish with normal chronologic aging. These functional compensations of the brain cortex is perhaps not surprising in light of the recent work by Kochunuv et al  that examined 1,031 subjects aged 11 to 90 years old, and who had history of psychiatric or neurologic conditions. This latter study showed that while the brain becomes fully developed by the age of ~20 years, parts of it such as the posterior cingulate cortex in fact undergoes regional thickening of the gray matter (GM) with further aging compared to other areas that retrogress with age; GM thickening is considered a measure of the preservation of integrity of brain structure.
Collectively, these studies show that vital organs such as brain, or at least parts of it, undergo physiologic remodeling that has beneficial outcome in old age. More recent studies show that cognitive function is closely linked to physical function in predicting a trajectory of continuing functional independence among older adults [148–150]. Such link underscores the importance of these two physiologic domains as components of successful aging. A research challenge is to determine factor(s) that triggers such compensatory pathways, and the biological context in which these compensatory pathways supersede, or perhaps override, the usual declines of physical and cognitive functions normally seen with advancing age.
In line with evidences for compensatory adaptations of physical and cognitive processes in old age, we have articulated the concept of immune remodeling as an ongoing dynamic change in the phenotype and function of immune cells in order to maintain/sustain immunological homeostasis into late life [151, 152]. Immune remodeling neither ignores nor argues against predictable age-related inefficiencies of classical immune function [2–12] as outlined above. Rather, immune remodeling pertains to elaboration of unique properties of the immune system (exemplified by centenarians described above), which may have begun at mid-life, or perhaps as early as adolescence, that become increasingly dominant with time. Immune remodeling complements classical immune functions that are preserved throughout life, such as the intrinsic ability of aged NK cells to produce various cytokines and chemokines [8, 153, 154], the retention of the ability of aged monocytes and dendritic cells to differentiate in response of granulocyte macrophage colony stimulating factor and IL-4 [155, 156], the phenomenal long lasting immune memory against H1N1 influenza virus [18–20] and herpes zoster virus , and the retention of aged CD8 T cells expressing the naïve markers CD127 and CD45RA along with TREC episomes . Beneficial immune remodeling with aging is consistent with the evolutionary concept that immunity is a determinant of individual health and fitness [31, 159, 160], and with the emerging evidences for the role of older individuals in the preservation of the species [32, 87, 161].
Early indications for beneficial remodeling of the immune cell repertoire with aging come from studies of older adults who maintain vigorous immune responsiveness especially in the setting of seasonal influenza vaccination [162, 163]. Those with measurable influenza vaccine responses have preponderance of a unique subset of CD8 T cells expressing high levels of CD25, CD62L, and IL-4 [164–166]. Similar remodeling in the NK cell compartment may be inferred from cross sectional studies. Current data show NK repertoire remodeling is more about the skewing of particular NK cell subsets rather than the emergence of unique NK cells with entirely new receptors or functional properties [8, 167]. Whether or not there are predominant NK subsets shared universally among populations of older adults, or what particular factors drive skewing of the NK cell subsets in old age are not yet known.
We first evaluated the idea of age-associated T cell repertoire remodeling based on several anecdotal reports about the increased level of expression of various NK cell receptors (NKR) such as CD56, CD16, and members of the CD158 KIR (killer cell immunoglobulin-like receptor) family among older adults [168–170]. Our experimental studies showed that persistent and/or cyclical stimulation and cellular senescence in vitro result in the emergence of T cell populations characterized by the induction of various NKRs, the irreversible loss of CD28, and the increased expression of mitotic inhibitors such as p16 and p53 [171, 172]. Our cross sectional studies showed an age-dependent increase in CD4 and CD8 T cells lacking CD28 but expressing CD56, p16, and p53 . Indeed, other investigators have shown that frequency of CD56+ T cells increases with further aging that such cell subset appears to be a feature of community-dwelling centenarians [129, 174]. And consistent with the broader concept that cellular senescence is not the same as quiescence nor does it precludes function, we have found that senescent CD56+ CD4 and CD8 T cells are functionally active. They capable of elaborating CD56-driven T cell effector function either in a TCR-dependent or TCR-independent manner [173, 175]. Other notable studies show that induction of NKRs on T cells is tightly regulated [176, 177], suggesting that accumulation of NKR+ T cells with advancing age is programmed.
Figure 1 summarizes findings from various laboratories including our own about the phenotypic characteristics of the “remodeled” CD28null NKR+ T cell.
The key observation is that such remodeled T cells have restricted usage of the clonotypic TCR, but they are nevertheless diverse considering the wide variations of co-dominantly expressed NKRs, such as CD16, CD56, NKG2D, NKG2A, CD94 and KIRs. It should be noted that these NK-like T cells express αβTCR in contradistinction with the NKT cells that express an invariant AV24 TCR and are usually CD161+ [178–180].
We suggest that remodeling or transformation of classical CD28+ T cells into NKR+ CD28null T cells with age is a compensatory mechanism to help maintain immune homeostasis into old age. With age-dependent involution of the thymus that severely impairs generation of new diverse naïve T cells , T cell homeostasis during middle to old age depends on the mobilization and maintenance of the pre-existing peripheral T cell pool. With age-related inefficiency of TCR-driven signaling [2–4], induction of NKRs on aged T cells may be a way to maintain T cell effector function.
We also suggest that NKR expression on the aged T cell could also compensate for the loss of classical NK cell function [8, 9]. As shown by data in Figure 2, there is an overall age-dependent decrease in the proportion of CD16+ and CD56+ NK cells with the corresponding increase in T cells expressing either CD16 and/or CD56. Furthermore, the data unequivocally show CD56 ligation is capable of eliciting T cell activation independent of the TCR.
Although biological situations in which T cells expressing CD56, CD16 and other NKRs confer protective immunity in old age remains to be investigated, we have recently reported that these cells comprise a distinctive immune fingerprint of octogenarians and nonagenarians who have high cognitive and physical function . Stimulatory NKRs such as CD56, CD16 and NKG2D are significant predictors of physical and cognitive resilience, whereas inhibitory NKRs such as NKG2A and CD158a predict mild disability. Moreover, these cellular signatures are coupled to a phenomenal global cytokine up regulation, with IL-4 and interferon-γ as corollary positive and negative predictors, respectively, of functional resilience. These observations indicate that immune remodeling occurs at the cellular and humoral levels as illustrated in Figure 1.
Beneficial humoral remodeling is not the same as the notion of inflammaging, which posits that immune deficiency in aging is due to the up regulation of cytokines that are presumed to be inflammatory . Notably, our data show co-existence of pro- and anti-inflammatory cytokines, and a similar co-existence of so-called Th1, Th2, and Th17 cytokines (see data in reference ). Along the same vein, a previous large cohort study of adults aged ≥85 years old has shown that pro-inflammatory factors, i.e. TNF-α, IL-1β, IL-6, and IL-1RA, are in fact beneficial to long term survivorship . Considering the myriad of influences of cytokines on T cell biology, we suggest the prevailing cytokine milieu dictates the differences in types and levels of NKRs expressed by T cells, and the elaboration of effector functions. It will therefore be of interest to examine what combinations of systemic cytokines and NKR+ T cells ultimately promote immune protection and better health in old age.
The large numbers of community-dwelling, functionally independent older adults who are surviving very well into their 8th to 11th decades of life underscore inaccuracy of the notion that the aged immune system is functionally incompetent. While it is true that old people are more vulnerable to disease and infection compared to younger adults, we may not ignore that many older adults are functionally resilient in spite of their histories of chronic disease and/or their concurrent medical conditions [17, 56, 59, 66, 182]. Geriatricians have made significant headway in defining physical and cognitive elements of successful aging. Therefore, burden is on immunologists to determine complementary immunologic elements of this physiologic construct.
The novel immune characteristics of older adults strongly indicate that the immune system in extreme old age is physiologically distinct, and may not be sweepingly categorized as a defective version of that seen at a young age. We suggest that the observed remodeling of the T cell repertoire  is consistent with the emerging findings about the plasticity of immune cell functions. Immune plasticity is illustrated by experimental data showing that T cell differentiation into Th1/Th2/Th17 effectors and regulatory Tregs is not as unidirectional as originally proposed by specific paradigms, but that T cells undergo dynamic shifts of cell phenotypes and function [28–30, 183]. The challenge then is to determine T cell differentiation mechanisms that promote continuous T cell function into old age. Studies showing distinct immunologic features of older adults who are either functionally impaired or unimpaired [152, 184] and those who retain the ability to mount vigorous immune responses [162–164] provide strong rationale towards systems approach, rather than an immune deficiency-minded approach, to better understand as to what constitutes immune competence, or true immune incompetence, in old age.
We thank Dr. Anne Newman for continued collaboration on the immunobiology of survivors of the CHS cohort. Original data shown in this paper were obtained from the secondary analysis of a large data set linked to our recently published study . We also thank Dr. Stephanie Studenski for ongoing research collaboration, and provocative discussions of ideas about exceptional aging.
We appreciate Amanda Way, Robert G. Mueller, David L. Hamel, and Jeffrey Dvergsten for their contributions in the processing of specimens, and to the secondary analyses of immunologic data that led to the summative data figures presented herein. We are indebted to the staff of the UPMC McGee Women’s Research Institute, and Jayne Rasmussen and the staff of the UPMC Children’s Hospital Laboratory Services for their assistance in procuring waste de-identified cord blood, and blood samples from healthy children, respectively.
Human subject’s research was performed in accordance with protocols approved by the Institutional Review Board of the University of Pittsburgh.
This work was supported by NIH research grant R01 AG030734.