TNF Drives Monocyte Dysfunction with Age and Results in Impaired Anti-pneumococcal Immunity

Abstract

Monocyte phenotype and output changes with age, but why this occurs and how it impacts anti-bacterial immunity are not clear. We found that, in both humans and mice, circulating monocyte phenotype and function was altered with age due to increasing levels of TNF in the circulation that occur as part of the aging process. Ly6C+ monocytes from old (18-22 mo) mice and CD14+CD16+ intermediate/inflammatory monocytes from older adults also contributed to this "age-associated inflammation" as they produced more of the inflammatory cytokines IL6 and TNF in the steady state and when stimulated with bacterial products. Using an aged mouse model of pneumococcal colonization we found that chronic exposure to TNF with age altered the maturity of circulating monocytes, as measured by F4/80 expression, and this decrease in monocyte maturation was directly linked to susceptibility to infection. Ly6C+ monocytes from old mice had higher levels of CCR2 expression, which promoted premature egress from the bone marrow when challenged with Streptococcus pneumoniae. Although Ly6C+ monocyte recruitment and TNF levels in the blood and nasopharnyx were higher in old mice during S. pneumoniae colonization, bacterial clearance was impaired. Counterintuitively, elevated TNF and excessive monocyte recruitment in old mice contributed to impaired anti-pneumococcal immunity since bacterial clearance was improved upon pharmacological reduction of TNF or Ly6C+ monocytes, which were the major producers of TNF. Thus, with age TNF impairs inflammatory monocyte development, function and promotes premature egress, which contribute to systemic inflammation and is ultimately detrimental to anti-pneumococcal immunity.

Fig 1. Ly6C^high monocytes are increased with age, express more CCR2 and less F4/80.

(A) Total numbers of Ly6C^high and Ly6C^low monocytes were quantitated in the blood of old (18–22 mo) WT C57Bl6/J mice and compared to that from young (10–14 wk) mice. The data represent the mean (± SEM) of 6 mice. (B) Analysis of the Ly6C^high monocytes as a percentage of CD45⁺ cells in the blood and bone marrow of young and old mice (± SEM; n = 6). (C) CCR2 expression on Ly6C^high monocytes in the bone marrow and blood of old mice is higher than young controls as determined by flow cytometry (n = 6–8). (D) The mean expression of the macrophage maturity marker, F4/80, on Ly6C^high monocytes in the bone marrow and blood of young and old mice (n = 6–8). (E) Cells recruited to the peritoneum were quantitated 4 hours after administration of 100 nM CCL2. The recruitment of Ly6C^high and Ly6C^low monocytes was greater in old mice (± SEM; n = 5). Statistical significance was determined by two-tailed Mann-Whitney-Wilcoxon test or two-way ANOVA with Fisher's LSD post-test where appropriate. * indicates p < .05, ** indicates p < 0.005, *** indicates p < 0.0005 and **** indicates p < 0.00005. (A-D) is representative of 4 independent experiments; (E) is representative of 2 independent experiments.

Fig 2. Ly6C^high monocytes contribute to elevated levels of serum IL6 and TNF in aged mice.

Young and old mice were injected with 500 nm negatively-charged polystyrene microparticles (PS-MPs) previously shown to reduce numbers of circulating Ly6C^high monocytes. Circulating monocyte populations (A) and IL6 levels in whole blood (B) were quantitated after 24 hours. Statistical significance was determined by two-tailed Mann-Whitney-Wilcoxon test. * indicates p < .05, ** indicates p < 0.005, *** indicates p < 0.0005 and **** indicates p < 0.00005. (A-B) is representative of ± SEM of 5 mice from 2 independent experiments.

Fig 3. Human CD14⁺⁺CD16⁺HLA-DR⁺ (intermediate) monocytes produce more inflammatory cytokines with age.

Intracellular production of TNF (A) and IL-6 (B) in classical (CD14⁺⁺), intermediate (CD14⁺⁺CD16⁺) and non-classical (CD14⁺CD16⁺) monocytes from young and elderly donors in response to LPS (50 ng/ml) and S. pneumoniae (5 x10 ⁶ CFU). C) The secretion of TNF and D) IL-6 for isolated CD14+ monocytes in response to LPS for young and older donors. E) The frequency of intermediate monocytes were found to have a significant, positive correlation with the levels of serum TNF (β = 2.78, p<0.006). (A-D) is representative of ± SEM of n = 7 young donors (26–52 yrs) and n = 6 older donors (63–70 yrs) *indicates p<0.05, and ** indicates p< 0.05. Intermediate monocyte (CD14++CD16+HLA-DR+) count (cells per microlitre of whole blood) increases relative to serum levels of TNF in older donors (n = 94, 61-100yrs).

Fig 4. TNF drives increases in circulating Ly6C^high monocytes.

(A) Total numbers of Ly6C^high and Ly6C^low monocytes in the blood of heterochronic bone marrow chimeric mice. Old recipient mice which receive young donor marrow have increased numbers of circulating Ly6C^high and Ly6C^low monocytes which are comparable to old recipient mice that receive old donor marrow. Young recipient mice that receive old donor marrow do not have an increase in Ly6C^high and Ly6C^low monocytes. The data represent the mean (± SEM) of 5 mice. (B) CCR2 expression on circulating monocytes is elevated when the recipient mouse is old, indicating that the bone marrow microenvironment drives changes in CCR2 expression (CCR2 MFI± SEM; n = 5). (C-D) The percent Ly6C^high monocytes as a proportion of CD45⁺ cells in the (C) blood or (D) bone marrow of young and old WT and TNF KO mice was quantitated (± SEM; n = 4–6). (E-F) Expression of CCR2 on Ly6C^high monocytes in the (E) blood or (F) bone marrow of young and old WT and TNF KO mice (n = 4–8) demonstrate that the presence of TNF drives CCR2 expression with age. (G) IL6 production in whole blood from young and old TNF KO mice stimulated with 100 ng/ml of LPS or a vehicle control for 24 hours was quantitated by ELISA (± SEM; n = 5). Statistical significance was determined by two-tailed Mann-Whitney-Wilcoxon test, one-way or two-way ANOVA with Fisher's LSD post-test where appropriate. * indicates p < .05, ** indicates p < 0.005, *** indicates p < 0.0005 and **** indicates p < 0.00005. (A-B) is representative of 2 independent experiments; (C-G) is representative of 3 independent experiments.

Fig 5. Anti-TNF therapy can reverse the age-associated increase in circulating Ly6C^high monocytes.

(A-B) Young mice were give 200 ng/ml of TNF intraperitoneally every other day for 3 weeks. Numbers of circulating Ly6C^high and Ly6C^low monocytes (A) and serum IL6 (B) were quantitated. The data represent the mean (± SEM) of 5 mice. (C) Young and old WT mice were treated for 3 weeks with a neutralizing TNF antibody or IgG control and total numbers of circulating Ly6C^high monocytes were quantitated by flow cytometry. The data represent the mean (± SEM) of 4 mice. (D) The mean CCR2 expression on circulating Ly6C^high monocytes in young and old mice treated with either anti-TNF or IgG was quantitated and found to be reduced with anti-TNF treatment (n = 4). (E) Intracellular staining of IL6 and TNF on blood monocytes after a 4 hour stimulation with LPS from young and old WT mice treated with either anti-TNF or IgG demonstrates that the number of monocytes that stain positive for IL6 or TNF are decreased with anti-TNF therapy(± SEM; n = of 4). (F) Serum IL6 is reduced in old mice treated with anti-TNF but not the IgG control. (G) IL6 production in whole blood following stimulation with LPS or a vehicle control after 24 hours from young and old WT mice given either anti-TNF or IgG (± SEM; n = 4). Statistical significance was determined by two-tailed Mann-Whitney-Wilcoxon test, one-way or two-way ANOVA with Fisher's LSD post-test where appropriate. * indicates p < .05, ** indicates p < 0.005, *** indicates p < 0.0005 and **** indicates p < 0.00005. (A-G) are representative of 1 experiment with n = 4 mice.

Fig 6. Old mice have increased numbers of circulating and recruited Ly6C^high monocytes during the course of S. pneumoniae colonization.

(A) Colony forming units (CFUs) in nasal lavages from young and old WT mice were quantified on days 3, 7, 14 and 21 following intranasal colonization with S. pneumoniae (± SEM; n = 5–22). (B) CFUs of S. pneumoniae in the lungs at day 3 following intranasal colonization (± SEM; n = 9–22). (C) Survival of young and old mice after intranasal S. pneumoniae colonization (± SEM; n = 12). (D) Total serum CCL2 in young and old mice following intranasal S. pneumoniae colonization was measured by a high sensitivity ELISA. The data represent the mean (± SEM) of 3 mice per time point. (E) Ly6C^high monocytes as a percent of CD45⁺ cells in the blood of young and old WT mice during nasopharyngeal S. pneumoniae colonization (± SEM; n = 5–8) was measured by flow cytometry. (F) Mean expression of F4/80 on Ly6C^high monocytes in the blood of old mice during S. pneumoniae colonization is decreased as compared to young mice. (G) Levels of CCL2 transcript in the nasopharynx during the course of S. pneumoniae colonization were measured by quantitative PCR. (± SEM; n = 3). (H-I) Total numbers of (H) Ly6C^high monocytes and (I) macrophages detected by flow cytometry in the nasopharnyx of young and old mice during S. pneumoniae colonization (± SEM; n = 3–8). (J) Mean F4/80 expression on nasopharyngeal macrophages is lower in old mice during S. pneumoniae colonization (± SEM; n = 3–8). Statistical significance was determined by two-tailed Mann-Whitney-Wilcoxon test, one-way or two-way ANOVA with Fisher's LSD post-test. (K) Circulating blood monocytes from old mice bind fewer TRITC-labelled S. pneumoniae (4°C) but there is no difference in internalization of the bacteria (37°C). Survival in (C) was determined by the Mantel-Cox Log-rank test. * indicates p < .05, ** indicates p < 0.005, *** indicates p < 0.0005 and **** indicates p < 0.00005. (A-J) is representative of 3 independent experiments.

Fig 7. Reducing TNF-regulated recruitment of Ly6C^high monocytes during S. pneumoniae colonization in old mice reduced nasopharyngeal bacterial loads.

(A-B) TNF in the (A) nasopharnyx and (B) serum of young and old mice during S. pneumoniae colonization as measured by qPCR and ELISA, respectively (± SEM; n = 3–5). (C) CFUs in nasal lavages of old WT and old TNF mice on day 4 after colonization with S. pneumoniae (± SEM; n = 6–8, one independent experiment of two shown). (D) Ly6C^high monocytes as a percent of circulating CD45+ cells in old WT and TNF KO mice on day 4 of S. pneumoniae colonization (± SEM; n = 3–4, one independent experiment of two shown). Statistical significance was determined by two-tailed Mann-Whitney-Wilcoxon test, one-way ANOVA or two-way ANOVA with Fisher's LSD post-test where appropriate. * indicates p < .05, ** indicates p < 0.005, *** indicates p < 0.0005 and **** indicates p < 0.00005.

Fig 8. Depletion of inflammatory monocytes improves outcome to S. pneumoniae infection in old mice.

Mice (n = 7-10/group) were injected with PS-MP day -1, 0, +1, +3 and +5 during colonization with S. pneumoniae. A) The percentage of Ly6C^high monocytes was significantly reduced in old mice treated with PS-MP (see S3 Fig). B) Survival was significantly improved in old mice treated with PS-MP (p = 0.005, Mantel-Cox log-rank test). C) Both young and old mice treated with PS-MP lost less weight than their control counterparts (*,p<0.05, one-way ANOVA with uncorrected Fisher's LSD). Levels of S. pneumoniae in the D) nasal wash, E) lungs and F) spleen were lower in old mice treated with PS-MP. Fewer young mice had bacteria in their lungs and spleens when they were treated with PS-MP. (*,p<0.05, **,p<0.005 one-way ANOVA with uncorrected Fisher's LSD). CFU count for mice that reached endpoint before day 7 are not included.