Age-dependent alterations of monocyte subsets and monocyte-related chemokine pathways in healthy adults

Monocytes represent about 5-10% of peripheral blood leukocytes in humans and mice. They originate from a myeloid precursor in the bone marrow, circulate in the blood, bone marrow and spleen, and then enter tissues [1, 2]. Monocytes are established circulating precursors for tissue macrophages and dendritic cells (DCs). Migration of monocytes into tissues and differentiation into macrophages or DCs is believed to be largely determined by the inflammatory milieu, i.e. adhesion molecules, chemokines and pathogen-associated pattern-recognition receptors [1]. The identification of different monocyte subsets in mice [3] has prompted intensive research over the last years to understand the true contribution of monocyte subpopulations to the macrophage and tissue DC pool in inflammatory disorders, but also in steady state [1].

Heterogeneity among human monocytes was recognized about thirty years ago [4, 5], and several markers like CD64 [6] or CD16 [7] have been suggested for differentiating subpopulations of monocytes. The differential expression of CD14 (part of the receptor for lipopolysaccharide) and CD16 (also known as FcγRIII) are commonly used to define two major subsets in peripheral blood: 'classical' CD14⁺⁺CD16^- monocytes, typically representing up to 95% of the monocytes in a healthy individual, and the 'non-classical' CD14⁺CD16⁺ cells comprising the remaining fraction of monocytes [7]. These subsets differ in many respects, including adhesion molecule and chemokine receptor (CCR) expression. CD14⁺⁺CD16^- monocytes express CCR2, CD62L (L-Selectin) and FCγRI (CD64), whereas CD14⁺CD16⁺ monocytes lack CCR2, and have higher levels of MHC-II and FCγRII (CD32). Both subsets express the receptor for fractalkine, CX₃CR1, but CD14⁺CD16⁺ monocytes characteristically express higher levels [2, 8]. Based on similar adhesion molecule and chemokine receptor as well as similar gene expression profiles, murine Gr1^hi (Ly6C^hi) monocytes are considered counterparts of human CD14⁺⁺CD16^- monocytes, and murine Gr1^lo (Ly6C^lo) cells may represent the subpopulation comparable to human CD14⁺CD16⁺ monocytes [8].

Intensive research efforts have unraveled important functional characteristics of both monocyte subsets in mice. While the Gr1^hi monocytes are rapidly recruited to sites of inflammation, such as in atherosclerosis, peritonitis or after organ damage into the injured liver [9‐12], Gr1^lo monocytes appear to have a more patrolling behavior at the endothelium [13]. Consequently, Gr1^hi monocytes were found to give rise to pro-inflammatory macrophages and TNF-producing DCs in inflamed tissue [1, 9, 10, 14], while Gr1^lo monocytes have been proposed as precursors for alternatively activated macrophages, possibly fulfilling functions in tissue repair and resident macrophage/DC turnover [1, 13, 15]. All these studies have raised the question how this knowledge can be translated into understanding physiology and pathophysiology of monocyte subsets in humans. We therefore aimed at characterizing peripheral blood monocyte subsets in healthy volunteers based on frequency, phenotype, chemokine and chemokine receptor expression and functionality assays ex vivo. Data from mice [2] and humans [16] indicated that monocyte subsets considerably change with age. Here, we show that the 'non-classical' CD14⁺CD16⁺ monocytes specifically increase with age in healthy volunteers, that monocyte-related chemokines and chemokine receptors are regulated within individuals of different ages and that the peripheral monocyte population preserves its cytokine-secreting function during ageing in adults.

Alterations in the numbers, phenotype and functionality of monocyte subsets have recently gained remarkable attention, because the identification of murine monocyte subpopulations has revealed new insights into their critical functions for macrophage and dendritic cell homeostasis as well as during inflammatory responses. For instance, the appearance of monocyte subsets in peripheral murine blood is the result of several well regulated processes at the level of bone marrow exit, half-life in blood and release from the spleen that can serve as an important reservoir [1, 14, 23]. Their differential recruitment in steady state or to sites of inflammation involves a cascade of contacts with endothelial proteins, adhesion molecule expression and chemokine-chemokine receptor interactions [10, 13, 24], and some of these mechanisms are highly specific for each monocyte subset, resulting in their selective or sequential accumulation and determining their function in health and disease [1, 9, 10, 15, 24]. It is an ongoing challenge to establish which of these mechanisms can be translated into human physiology and pathophysiology.

We here conducted a study aiming at thoroughly characterizing alterations of monocyte subpopulations in healthy human volunteers at different ages. The effects of ageing on adaptive and innate immune responses in humans remain incompletely understood. The increased susceptibility to infections and the reduced response to vaccinations in old people have been linked to T- and B-cell functions [25], but also changes in the innate immune systems, including neutrophils, monocytes, macrophages, natural killer and natural killer T (NKT) cells and dendritic cells, have been reported to contribute to the "immunosenescence" observed at old age [26]. Our study demonstrates a significant shift from 'classical' CD14⁺⁺CD16^- to 'non-classical' CD14⁺CD16⁺ monocytes with increasing age in healthy adults. These results are well in agreement with a prior smaller study that reported a significant expansion of CD14⁺CD16⁺ monocytes in the elderly, although both studies are not directly comparably, because the prior study compared very old (mean age 88 years) with young (mean age 30 years) subjects [16].

Moreover, the CD14⁺CD16⁺ monocytes undergo phenotypic changes during ageing, as they display reduced levels of the activation molecule HLA-DR and of the key chemokine receptor CX₃CR1. Especially the latter finding might be functionally relevant, because CX₃CR1 has been implicated to promote monocyte/macrophage survival in some conditions like atherosclerosis [27]. Given experimental data from mice, reduced CX₃CR1 expression on non-classical monocytes in humans is likely to affect their migration to inflammatory sites and to the spleen [10, 28], but could also reduce the half-life of monocyte-derived tissue macrophages [29]. Therefore, increased numbers of circulating CD14⁺CD16⁺ monocytes may not necessarily mean enhanced availability and functionality in aged persons.

At the level of circulating chemokines, we observed a marked increase of serum MCP-1 (CCL2) concentrations with age. Independent studies have reported increasing MCP-1 concentrations in older volunteers previously and speculated that this could be related to development of (subclinical) atherosclerosis or an age-dependent shift between T-helper cell dependent cytokine patterns (Th1/Th2) [30, 31]. In our cohort, we observed a correlation between serum MCP-1 and the body mass index, possibly indicating an association with obesity and adipose tissue-derived macrophages or obesity-induced hepatic MCP-1 expression as important sources of circulating MCP-1 [32, 33]. Interestingly, neither was CCR2 expression downregulated on CD14⁺⁺CD16^- monocytes in response to MCP-1 levels nor did serum concentrations of fractalkine (CX₃CL1) or the CCR1/CCR5 ligands CCL3 (MIP1α) and CCL4 (MIP1β) change with age.

Surprisingly, our experiments further indicate that the observed changes in the number and phenotype of circulating monocyte subsets do not largely impact their overall functionality, at least not in the studied age-range of healthy adults (mean age 41 years, 95%-interval 20-68 years). This is unexpected, because CD14⁺CD16⁺ monocytes, which are significantly expanded in older volunteers, are generally believed to have a much higher capacity to secrete proinflammatory cytokines [2]. We recently confirmed these assumptions, as CD14⁺CD16⁺ monocytes derived from healthy volunteers produce significantly more TNF, IL6, MIP1α, MIP1β, IFNγ and MIG upon culture without specific stimulation, while CD14⁺⁺CD16^- monocytes readily release more MCP-1 and IL-10 in vitro [19].

In our study, the constitutive and inducible synthesis of cytokines and chemokines was similar between total monocytes cultured from young and old(er) adults, although the fraction of the CD14⁺CD16⁺ subset was expanded in the cells derived from older persons. Possibly, the expansion of the 'non-classical' monocytes with age might represent a physiological counter-regulatory reaction, in order to preserve the overall functionality of the monocyte pool throughout life-time. However, it is important to note that our experimental setting did not allow to differentiate the distinct contribution of each subset to the overall cytokine release and that our in-vitro results do not necessarily have to reflect the true behaviour of monocyte-derived cells in their respective (inflamed or non-inflamed) in-vivo microenvironment.

Our results of a preserved cytokine secretion function of monocyte-derived macrophages in older volunteers confirm a report that compared macrophages only from women [34], but are in contrast to prior studies [16, 35]. However, these divergent studies analyzed cells from remarkably older people than in our study, suggesting that monocyte functionality may decrease during further ageing. In elderly persons with a mean age 88 years, monocyte-derived macrophages were reported to display an imbalanced production of cytokines, with higher IL1β and IL6 at steady state as well as lower IL1β and higher IL6 and IL10 secretion upon stimulation [16]. Similarly, higher TNF, IL1β and IL6 levels were reported in macrophages from old (mean age 80 years) compared to young (mean age 27 years) subjects [35].

Our study indicates that the circulating monocyte pool dynamically changes during ageing in humans. The expansion of the non-classical CD14⁺CD16⁺ subtype, alterations of surface protein and chemokine receptor expression as well as circulating monocyte-related chemokines likely allow the preservation of functionality for the monocyte pool throughout the majority of adulthood.