Álvarez-Rodríguez et al. reported that age is positively correlated with the intracellular expression of TNFα, IL-6, and IL-1β in circulating monocytes (117), whereas O’Mahony et al. found no difference (118). Cao et al. and Puchta et al. also did not observe a difference in the TNFα, IL-6, and IL-1β expression of monocytes in young and old individuals (17, 119). Hearps et al. however, reported an increased TNFα expression in classical, intermediate, and non-classical monocytes of older individuals (67). Pence et al. showed a decreased IL-6 gene expression in older individuals (120), corroborating the findings of Mohanty et al. who observed a decrease in IL-6+ monocytes (71). In mice, unstimulated splenic macrophages, thioglycollate-elicited peritoneal macrophages, or bone marrow-derived macrophages from young and old mice did not differ in their spontaneous TNFα, IL-6, and IL-1β secretion (121–125).

The results on pro-inflammatory cytokine secretion or expression in resting monocytes and macrophages are inconclusive. Different analysis methods (intracellular cytokine staining, ELISA of cytokines in the supernatant) and detection limits of the expected low, spontaneous monocyte cytokine secretion might play a role.

Cao et al. reported a higher frequency of TNFα+ monocytes, a decreased frequency of IL-6+ monocytes, and an unchanged frequency of IL-1β+ monocytes in response to stimulation with bacterial lipopolysaccharide (LPS) in older individuals (17). Puchta et al. analyzed the TNFα and IL-6 response of classical, intermediate, and non-classical monocytes of old and young individuals. They found an increased TNFα and IL-6 expression in intermediate monocytes of older individuals in response to LPS and no difference in the other two subpopulations (119). In addition, the group analyzed the LPS-induced TNFα and IL-6 secretion of isolated monocytes and observed an increased response from older individuals. An increased TNFα expression was also observed by Hearps et al. in intermediate monocytes and non-classical monocytes (67). Wang et al. analyzed the intracellular expression of IL-6 and found an increased LPS-induced IL-6 expression in all three monocyte subpopulations from aged donors (60-70 years) vs. young donors (21–30) (70). However, Pence et al. showed a decreased IL-6 gene expression in response to LPS in older individuals (120). We analyzed CD56+ monocytes, a monocyte subset that expands with age, and found that they express more TNFα in response to LPS than CD56- monocytes (83).

The findings on an increased TLR4 cytokine response is not limited to monocytes. Bouchlaka et al. differentiated peripheral blood mononuclear cells (PBMCs) into macrophages and stimulated them with LPS. They found an increased TNFα and IL-6 secretion in response to LPS in macrophages differentiated from older individuals compared to young individuals (123). Gather et al. differentiated monocytes from young and old individuals towards M1 and found no difference in TNFα and IL-6 mRNA and TNFα protein expression (126).

Whereas there is a trend towards a stronger pro-inflammatory cytokine response in TLR4 activation in monocytes and also macrophages, the response to other TLR ligands seems to be attenuated. Van Duin et al. showed that the frequency of TNFα+ and IL-6+ monocytes is diminished in older individuals when the monocytes were activated with Pam₃CSK₄, a TLR1/2 heterodimer ligand (127). However, they observed no difference when the monocytes were activated with lipoteichoic acid (TLR2/6 heterodimer ligand), LPS (TLR4 ligand), or flagellin (TLR5 ligand). Nyugen et al. activated whole blood with Pam₃CSK₄ and analyzed the frequency of TNFα+ and IL-6+ monocytes (68). IL-6+ monocytes were decreased in all three monocyte subsets, TNFα+ monocytes were decreased in intermediate and non-classical monocytes in older individuals.

LPS-stimulated thioglycollate (TG)-elicited peritoneal macrophages from old mice produced less TNFα and IL-6 than those from young mice (121). The same group also showed that activation of splenic macrophages from old mice with LPS or zymosan led to a diminished TNFα and IL-6 production (122, 128). Fallah et al. made similar observations. They reported that splenic macrophages produced less TNFα in response to LPS or killed S. pneumoniae (129). Gomez et al. and Chelvarajan et al. also described a decreased LPS-induced TNFα, IL-6, and IL-1β secretion in splenic macrophages from old mice (130, 131). Beharka et al. observed no difference in the LPS-induced IL-6 secretion of peritoneal macrophages between young and old mice (125). Chen and Bradley found a decreased TNFα and IL-6 secretion in resident and complete Freund’s adjuvant-elicited peritoneal macrophages from old mice, but an increased TNFα and IL-6 secretion in TG-elicited peritoneal macrophages (132). Cecílio et al. showed that resident and TG-elicited peritoneal macrophages from old mice have a decreased TNFα response (133).

The overall findings on decreased cytokine secretion are not limited to TLR4 ligands but also found with other TLR ligands. Alveolar macrophages from mature mice (10-12 months) and old mice (19-21 months) had a decreased TNFα response to ethanol-killed pneumococci and purified pneumococcal cell wall (134), and purified CD11b+ peritoneal macrophages a decreased TNFα response to zymosan (135).

The data on murine bone marrow-derived macrophages (BMDM) from old mice are less clear. Several studies reported an increased TNFα and IL-6 secretion in response to LPS (123, 133, 136, 137) Kang et al. reported an increased production of IL-1β in BMDMs from old mice in response to activation with Staphylococcus aureus (89). Zhao et al., however, found a decreased LPS-induced IL-6 secretion (138), whereas Mahbub et al. observed no difference in the TNFα, IL-6, and IL-1β secretion between old and young mice (128). Ramirez et al. also did not find an age effect on the LPS-induced IL-1β secretion, however, they observed a decreased response to an inflammasome activation with LPS and ATP (139). BMDMs activated with Porphyromonas gingivalis from older mice also showed a decreased TNFα response (140).

Published data on the anti-inflammatory cytokine IL-10 and monocytes or macrophages are scarce. Cao et al. did not observe a difference in the IL-10 expression of monocytes in young and old individuals (17). They also reported an unchanged frequency of IL-10+ monocytes in response to stimulation with LPS in older individuals. Mohanty et al. found an increased frequency of IL-10+ monocytes in older individuals (71). Boehmer et al. showed that splenic macrophages of old mice produce more IL-10 than young mice in response to LPS as well as zymosan (122), whereas Cecílio et al. observed no effect on IL-10 in LPS-activated murine BMDM (133).

While an increased TLR4-induced pro-inflammatory cytokine response was most often observed in monocytes, the cytokine response was diminished in tissue macrophages. The age effect on bone marrow-derived macrophages is also less clear than on tissue macrophages from mice. Mahbub et al. already speculated that the surrounding aging microenvironment has an effect on macrophage development (128). Gomez et al. reported similar findings (141). Inflammaging and inflammatory cytokines and chemokines secreted by senescent cells (the senescence-associated secretory phenotype, SASP) have a profound influence on surrounding cells. Gather et al. showed that age has no effect on monocyte-derived macrophages, but that an aged microenvironment in co-culture experiments with aged dermal fibroblasts drives a more pro-inflammatory macrophage phenotype (126). The aged microenvironment also drives the accumulation and activation of M1-like CD38+ macrophages in adipose tissue and liver during aging (142, 143). McQuattie-Pimentel et al. also showed that age-related changes in alveolar macrophages are defined by their microenvironment in the lung (144). Chambers et al. described the recruitment of inflammatory monocytes by senescent fibroblasts in the human skin (145).

Pro-inflammatory cytokines play an important role in the pathogenesis of RA, as demonstrated by the successful therapies using TNFα and IL-6 neutralizing agents (146). We have shown that monocytes of RA patients show signs of premature aging. They express membrane TNFα and spontaneously secrete IL-1β, which leads to an increased survival of the monocytes (147). Paoletti et al. also found an increased expression of membrane TNFα on RA monocytes (148). It has also been demonstrated that the intermediate monocyte subpopulation is predominantly expanded in the synovial fluid of RA patients (29, 76) and this subpopulation is associated with aging as discussed above and characterized by a high IL-1β secretion (58, 149). Single-cell RNA sequencing of RA synovial tissue led to the identification of a IL1B ⁺ pro-inflammatory monocyte subset which was possibly shaped by the local microenvironment (150). Transcriptomic analysis of classical monocytes revealed that the expression of IL6 was higher in erosive disease than in non-erosive disease in premenopausal women (median age 39 vs. 37 years) (151).

We and others have shown that monocytes of treated RA patients and healthy controls have a comparable cytokine response to LPS (147, 152, 153). However, Leirisalo-Repo et al. observed an increased TNFα secretion in LPS-activated monocytes of untreated patients with early RA (152). This is in contrast to a study by Liou (154). They reported a normal cytokine response to LPS in monocytes from untreated RA patients. We also found no difference in the IL-1β secretion in treated and untreated RA patients following monocyte activation with LPS, however, we observed a lower IL-18 response in monocytes of untreated RA patients compared to healthy controls (153). The inflammasome activation with LPS and calcium led to an increased IL-1β and IL-18 secretion in monocytes from RA patients compared to healthy controls (153).

Published data on the effect of aging on human monocyte phagocytosis are conflicting. Hearps et al. report an impaired phagocytosis of Escherichia coli by human monocytes from old adults (median age 72) compared to monocytes from young adults (median age 28) (67) whereas several older publications describe no effect or a non-significant trend towards impaired phagocytosis in old individuals (155, 156). Mege et al. studied the phagocytosis of opsonized zymosan, immunoglobulin-coated sheep red cells and glutaraldehyde-treated sheep red cells (155). They observed an impaired phagocytosis of opsonized zymosan and immunoglobulin-coated sheep red cells by human monocytes from old adults (mean age 76) compared to monocytes from young adults (mean age 36) but this trend did not reach statistical significance (155). Gardner et al. found no difference in the phagocytosis and killing of Candida albicans in monocytes from healthy young controls (<35 years) and hospitalized patients >60 years and <35 years of age (156). Unrelated to an infectious setting, Bliederhaeuser et al. report an impaired phagocytosis of exosome-associated and free alpha-synuclein oligomers by human monocytes from old adults (median age 66) compared to monocytes from young adults (median age 24) (157). A recent study showed that monocytes and monocyte-derived macrophages from old adults (mean age 60.5) have an impaired phagocytosis of opsonized beads or Staphylococcus aureus in comparison to young individuals (mean age 23.7) (158).

Studies of aged monocytes and macrophages in mice and rats are as conflicting as in humans. The phagocytosis of latex beads by peritoneal macrophages was decreased in aged mice (85 weeks) compared to young mice (15 weeks) (159), whereas the phagocytosis of Klebsiella pneumoniae was increased in alveolar macrophages from old rats (18 month) compared to young rats (6 month) (160). The phagocytosis of apoptotic cells by thioglycolate-elicited peritoneal macrophages, resident peritoneal macrophages and bone marrow-derived macrophages was found to be decreased in aged mice (161–163).

Linehan et al. showed that the tissue microenvironment might play a role in age-related defects in phagocytosis (164). They reported a decreased phagocytosis of fluorescent particles by peritoneal macrophages from old mice (15-20 month) compared to young mice (8-12 weeks), whereas bone marrow-derived macrophages and bone marrow monocytes showed no age-related impairment of phagocytosis (164). Interestingly, peritoneal macrophages derived from young mice showed impaired phagocytosis when injected into the peritoneum of old mice (164). This suggests that microenvironmental factors in the aged peritoneum cause age-related defects and Linehan et al. identified IL-10 as one of the factors (164).

Several older studies reported no difference in phagocytosis of bacteria between RA patients and healthy controls (165–167). Steven et al. observed increased phagocytosis of opsonized Staphylococcus aureus and Proteus mirabilis by monocytes from RA patients compared to age-matched controls (168). The phagocytosis rate correlated positively with age when patients and controls were analyzed combined, and the phagocytosis rate of RA patients remained higher than in controls in the 30-60 age group (168). More recent studies also did not focus on age. Fragoulis et al. observed no difference in the phagocytosis of secondary necrotic cell remnants in RA patients and controls (169) and Tas et al. found no difference in the phagocytosis of apoptotic cells in monocyte-derived macrophages from RA patients and controls (170).

Interestingly, Lee et al. reported that activated platelet-induced CD16 (also known as FcγRIII) on classical monocytes participates in increased antibody-dependent cellular phagocytosis (171). The monocytes then resemble intermediate monocytes, a monocyte subpopulation expanded in the blood and synovial fluid of RA patients (58, 76) and reported to be expanded in older individuals (17, 23).

Aging leads to a decrease in ROS produced during the respiratory burst in monocytes and macrophages. Alvarez et al. have shown that the production of ROS decreases with age in rat peritoneal macrophages (174, 175). Phorbol Myristate Acetate (PMA)-activated 12-month-old rat macrophages showed a lower production of superoxide ions compared to 3-month-old rats. Macrophages from 24-month-old rats had a further reduction in ROS production compared to both 12-month-old and 3-month-old rats. Activation of macrophages with N-formylmethionyl-leucylphenylalanine (fMLP) or concanavalin A did not show a decrease in ROS production between 3 and 12-month-old rats, however, a significant reduction in 24-month-old rats compared to both 3 and 12-month-old rats was observed (175).

The PMA-induced superoxide generation in human monocytes is negatively correlated with (176) and the PMA-induced capacity to produce superoxide ions is reduced in aged subjects (65-75 years old) compared to young (25-35 years old) and mature (45-55 years old) (177). Lower ROS production was also reported in alveolar macrophages. Ganguly et al. measured the production of superoxide in alveolar macrophages from female Fisher-344 rats stimulated by opsonised zymosan and found that aged rats showed lower ROS production (178). One limitation of this study was the use of aged breeder rats which have been shown to exhibit a lower magnitude of respiratory burst compared to young and aged virgins (179). Esposito et al. compared alveolar macrophages obtained from virgin adult and aged female rats and observed that normal senescent mice released larger quantities of superoxide when activated by either PMA or zymosan (180), contrary to previous findings.

In contrast to the ROS produced in activated monocytes and macrophages, the general ROS content in resting monocytes and macrophages seems to increase with age. Sebastián et al. showed an increased ROS content in BMDM from old mice compared to young mice (88). Jacinto et al. reported an increased ROS content in circulating monocytes from aged mice (181). Their study focused primarily on the apoE-/- mouse model of spontaneous atherosclerosis, but they also observed a significant increase in ROS content in monocytes from 18-month-old mice compared to 2-month-old wild-type mice.

Saare et al. reported similar findings in human monocytes (105). Oxidative stress due to ROS production was determined by measuring the fluorescence of oxidized products of chloromethyl derivative of 2’,7’-dichlorodihydrofluorescein diacetate (CM-H₂DCFDA) in monocytes isolated from young (mean age 29.6 years) and old (mean age 79.4 years) donors (105). Unstimulated monocytes from older donors showed a twofold increase in ROS production, while LPS-stimulated monocytes showed a smaller but still significant increase. Higher cellular levels of ROS are associated with DNA damage (182), and Saare et al. observed that double-strand breaks were 1.5 times higher in the old cohort by analysing the percentage of monocytes showing a γH2AX signal (105).

Wang et al. also reported an increased ROS content (measured with DCFH-DA) in all three monocyte subpopulations (CM, IM, NCM) from elderly individuals (60-70 years) compared with those from young persons (21-30 years) (70). Hayashi et al. observed a correlation of monocyte ROS content (measured with dichlorofluorescein and hydroethidine) and age in combined exposed and unexposed Hiroshima atomic bomb survivors (65-95 years) (183). The monocyte ROS content also correlated with the radiation dose implying that atomic bomb radiation accelerates immunological aging. In our study on CD56+ monocytes, a monocyte subpopulation associated with aging, we found increased ROS content (measured with dihydrorhodamine 123) in CD56+ monocytes compared to CD56- monocytes (83).

Monocytes in RA are under high oxidative stress compared to healthy controls as shown by Ostrakhovitch and Afanas’ev (184). Oxygen radical production of RA monocytes and healthy controls were measured by chemiluminescence with luminol or lucigenin. PMA-stimulated RA monocytes showed 2.7 times more oxygen radical production compared to controls. Monocytes from RA patients also showed an increase in NADPH oxidase activity compared to healthy controls. In addition, oxygen radical production was decreased in monocytes from RA patients in the presence of mitochondrial inhibitors rotenone and antimycin A, while normal cells were unaffected, suggesting that mitochondrial superoxide production is another source of oxygen radicals in RA monocytes.

Other studies also show that superoxide anion production is increased in monocytes from RA patients compared to healthy controls. Repo et al. showed that monocytes from untreated, early RA patients showed significantly higher Luminol-enhanced chemiluminescence to stimulation by FMLP, PMA or opsonised zymosan particles compared to healthy controls (185). Hurst et al. observed an increased superoxide anion production in RA monocytes stimulated with IgG-treated zymosan or serum-treated zymosan (186, 187). No difference in basal rates of monocyte superoxide anion production was observed between controls and RA patients. Mateen et al. showed that RA patients have an increased ROS formation compared to age-matched controls. They stained ROS with 2’, 7’-dichlorofluorescein-diacetate (DCFH-DA) to observe the intracellular ROS production (188). Nitric oxide (NO), an important component of inflammatory oxidative burst, was also shown to be 1.56 times higher in monocytes isolated from the blood of RA patients compared to healthy individuals (189).

Maoldomhnaigh et al. compared macrophages differentiated from monocytes from adult blood (aged 18-69 years) and cord blood (term neonates) (193). Both MDMs responded with a rapid increase in glycolysis when activated with LPS, but only the MDMs from adults showed a decrease in oxidative phosphorylation.

Wang et al. reported an increased glycolysis, impaired oxidative phosphorylation, and mitochondrial dysfunction in all three monocyte subsets in aged adults (70). These changes were most prominent in non-classical monocytes, and the authors speculated that the increased glycolysis is a result of the mitochondrial dysfunction to meet the demand for energy metabolism.

Reduced cellular respiratory capacity with age has been attributed to dysfunctional mitochondria (194). Saare et al. also reported lower mitochondrial membrane potential despite higher mitochondrial mass in aged subjects (105). The oxygen consumption rate (OCR), basal and maximal of monocytes from young (mean age 29.6 years) and old (mean age 82.3 years) individuals were measured to calculate the spare respiratory capacity (SRC) which was significantly lower in the older group. Pence and Yarbro also reported impaired mitochondrial respiratory activity in classical monocytes from old individuals using flow cytometry (120). Their monocyte isolation strategy included CD16 depletion to only focus on classical monocytes. Both studies showed that basal respiration remains largely unaffected by age but the maximal respiratory capacity is reduced in monocytes of older individuals which leads to lower SRC.

Additionally, gene expression studies in monocytes of young (average age 35.7) and old individuals (average age 71.6 years) also highlighted the downregulation of oxidative phosphorylation genes PLA2G4B and ALOX15B in older individuals, while PDK4 which inhibits pyruvate dehydrogenase complex was highly upregulated. This correlates to a shift of glucose metabolism from oxidative phosphorylation to lactate production in monocytes of older individuals (195). Interestingly, the NADPH oxidase, the enzyme involved in respiratory burst, is not affected by age (177).

However, Pence and Yarbro did not find conclusive evidence of increased glycolysis in LPS-stimulated aged monocytes ex-vivo (196). The microenvironment plays a significant role in age-related changes which could be a reason for such an observation. As the authors point out, there could be several epigenetic modifications and alterations in sirtuin activity that could be linked to metabolism.

There is an increase in circulating fatty acids in older adults since aging leads to the redistribution of body fat from subcutaneous to visceral, the latter being less efficient in the storage of fatty acids (197). Additionally, Wang et al. reported decreased FAO and increased presence of lipid droplets (LD) in aged monocytes of mice and humans (198). They also observed that PPAR-α expression was downregulated in monocytes of aged humans (>60 years old) and subsequently were able to prove that this caused the FAO decrease and LD increase by downregulating PPAR-α in young monocytes to similar effects. This also affects monocyte polarisation as the accumulation of LD in aged monocytes leads to a pro-inflammatory phenotype.

As an important regulator of metabolism, NAD+ is thought to also undergo changes during aging. Endogenous NAD+ is depleted during aging which eventually disrupts cellular metabolism (199). NAD+ can be obtained by several pathways, the two most commonly studied ones being de novo biosynthesis pathway, also called kynurenine pathway (KP), or the salvage pathway through nicotinamide precursors. Clement et al. showed that the NAD+ precursors remain stable with age, while NAD+ levels are reduced (200). This suggests that the decline in NAD+ might be caused by an increase in the activity of NAD-consuming enzymes. It is currently not clear which pathway is primarily responsible for providing NAD+ in macrophages as previous reports show conflicting results based on the characteristics of the macrophage studied. Liu et al. showed that de novo synthesis majorly occurs in the liver (201). However, Minhas et al. showed that decreased cellular NAD+ can also be replenished by macrophages via de novo synthesis (202). Disruption of the salvage pathway using a NAMPT inhibitor in human monocyte-derived macrophages (MDMs) demonstrated a marked increase in KP enzymes and metabolites. They also confirmed that basal de novo NAD+ synthesis accounts for 40% of total NAD+, and almost all when the salvage pathway is blocked by the NAMPT inhibitor FK866. Furthermore, they also showed that loss of the KP disrupts mitochondrial respiration by inhibiting KP enzymes IDO1 and QPRT. Peritoneal macrophages from IDO1- or QPRT-deficient mice showed reduced cellular NAD+ amounts compared to wild type controls, along with suppressed oxygen consumption and increased glycolysis. This effect was reversed on supplementation with kynurenine. Similarly, human MDMs also showed a decreased NAD+ and mitochondrial respiration when IDO1 and QPRT were inhibited. Additionally, loss of QPRT activity also elevated lactate, pentose phosphate pathway intermediates and pro-inflammatory TCA intermediates that favour IL-1β and IL-6 production. Analysis of QPRT expression in human MDMs derived from young (<35 years old) and older (>65 years old) individuals showed a significant decline in older macrophages, the older macrophages having a metabolite profile similar to that observed in LPS-stimulated macrophages. Suppression of NAD+ synthesis in aged mice indicated that the KP derived NAD+ is critical in sustaining the cellular concentration. The age-associated loss of this pathway likely contributes to the proinflammatory shift of macrophages with age.

Macrophages after inflammation, however, seem to be dependent on the salvage pathway for NAD+. Cameron et al. observed that murine BMDM showed reduced NAD+ levels when stimulated by LPS compared to resting M0 macrophages or IL-4-stimulated macrophages (203). NAMPT inhibition by FK866 treatment induced dose-dependent cell death in LPS-stimulated macrophages while the viability of M0 and IL-4 macrophages was affected to a lesser extent. Inflammatory cytokines such as TNFα, IL-1β, IFNγ induce the rate limiting enzyme for the salvage pathway, NAMPT. The authors reported that this is necessary to counter the rapid NAD+ depletion on LPS activation. As mentioned earlier, these cytokines also increase with age which likely enables the macrophages to keep up with the demand due to an increase in NAD+ consuming enzyme activities resulting from glyceraldehyde-3-phosphate dehydrogenase activity and Warburg metabolism.

Several studies on RA monocytes and macrophages show a pro-inflammatory metabolic profile. Metabolomic profiling showed an increased presence of lactic acid, citrate and succinate in the synovial fluid of RA (204). Shime et al. showed that lactic acid promotes pro-inflammatory pathway (205). Succinate too is known to be induced on LPS activation, as a result of the metabolic shift from oxidative phosphorylation to glycolysis during inflammation leading to its accumulation as a TCA cycle intermediate. These suggest that the macrophages in the synovial fluid might show a pro-inflammatory phenotype. Michopoulos et al. showed that itaconic acid is a marker of RA using the Tg197 mouse model which correlates with findings from Lampropoulou et al. who reported that itaconate is involved in the regulation of succinate levels in LPS-stimulated mouse bone marrow-derived macrophages (206, 207).

Yoon et al. described that Solute carrier family 7 member 5 (SLC7A5) is increased in RA patients (208). LPS-stimulated human peripheral monocytes from RA patients showed a 5.85-fold increase in SLC7A5 expression compared to healthy controls. LPS-mediated expression was also observed in human macrophages. They further reported the role of SLC7A5 in being involved in the glycolytic shift and production of pro-inflammatory cytokines such as IL-1β. McGarry et al. (30) showed that RA monocytes have increased mitochondrial respiration and enhanced glycolysis after activation with LPS. Blockade of glucose consumption resulted in an inhibition of the release of pro-inflammatory mediators. They also reported that RA monocytes have an increased number of mitochondria and an increased mitochondrial respiration already ex vivo without any stimulation. These above studies show that monocytes in RA tend to have a pro-inflammatory metabolic profile, and thus RA patients exhibit signs of premature aging.