Declining neutrophil production despite increasing G-CSF levels is associated with chronic inflammation in elderly rhesus macaques

Abstract

Aging is characterized by a loss of bone marrow hematopoietic tissue, systemic chronic inflammation, and higher susceptibility to infectious and noninfectious diseases. We previously reported the tightly regulated kinetics and massive daily production of neutrophils during homeostasis in adult rhesus macaques aged 3 to 19 yr (equivalent to approximately 10 to 70 yr of age in humans). In the current study, we observed an earlier release of recently dividing neutrophils from bone marrow and greater in-group variability of neutrophil kinetics based on in vivo BrdU labeling in a group of older rhesus macaques of 20-26 yr of age. Comparing neutrophil numbers and circulating cytokine levels in rhesus macaques spanning 2 to 26 yr of age, we found a negative correlation between age and blood neutrophil counts and a positive correlation between age and plasma G-CSF levels. Hierarchic clustering analysis also identified strong associations between G-CSF with the proinflammatory cytokines, IL-1β and MIP-1α. Furthermore, neutrophils from older macaques expressed less myeloperoxidase and comprised higher frequencies of polymorphonuclear myeloid-derived suppressor cells (PMN-MDSCs) compared to the young adult macaques. In summary, we observed an earlier release from bone marrow and a reduced production of neutrophils despite the increased levels of plasma G-CSF, especially in the elderly rhesus macaques. This lower neutrophil production capacity associated with increased production of proinflammatory cytokines as well as an earlier release of less mature neutrophils and PMN-MDSCs may contribute to the chronic inflammation and greater susceptibility to infectious and noninfectious diseases during aging.

Keywords: aging; hematopoiesis; myeloid-derived suppressor cells; polymorphonuclear cells.

FIGURE 1.. Neutrophils exhibited shorter bone marrow post-mitotic development time, earlier transit into blood, and higher in-group variability among older rhesus macaques aged 20 – 26 years of age compared to younger adults aged 3 – 19 years of age.

Animals each received a single bolus of BrdU (60 mg/kg) intravenously and EDTA-treated blood samples were collected various days later for antibody staining and flow cytometry analysis. Neutrophils were gated from single cells, FSC/SSC^high/dim, HLA-DR^-, CD3^-, CD20^-, CD123^-, and the percentages of BrdU⁺ cells were gated from the neutrophil population. (A&B) Percentages of BrdU-positive neutrophils four days (A) and seven days (B) after administration of BrdU were compared between the two age groups using the nonparametric Mann-Whitney U test. (C) The kinetics of BrdU-labeled blood neutrophil kinetics were compared between rhesus macaques aged 20–26 years old (red line) and previously reported data from rhesus macaques between 3 to 19 years old (blue line) [5]. Medians and interquartile ranges were shown at different time points after BrdU injection. (D) BrdU-labeled blood neutrophil kinetics of individual rhesus macaques aged 20–26 years old demonstrates in-group variability in this age group. ****, P <0.0001. Medians and interquartile ranges were shown.

FIGURE 2.. Declining blood neutrophil numbers despite elevated plasma G-CSF levels were observed in elderly rhesus macaques.

(A) Blood neutrophil counts in 142 animals aged 3 – 26 years old were examined by nonparametric Spearman correlation. (B) Blood neutrophil counts and medians with interquartile range were plotted and compared between age groups of rhesus macaques by Kruskal-Wallis test followed by Dunn’s post-test for pairwise comparisons. (C) Plasma G-CSF levels (pg/ml) were measured in 132 samples from 126 rhesus macaques as well as 12 specimens collected from six animals at different ages. Results were analyzed by nonparametric Spearman correlation. (D) G-CSF levels in age groups of rhesus macaques were plotted along with medians with interquartile ranges and compared by Kruskal-Wallis test with Dunn’s post-test for pairwise comparisons. *, P < 0.05; **, P < 0.01; ***, P < 0.001. P < 0.05 was considered statistically significant.

FIGURE 3.. Hierarchical clustering analysis and pairwise correlation matrix of cytokine profiles and age revealed that G-CSF levels are closely associated with IL-1β and MIP-1α levels as well as with aging.

Multiplex cytokines profiles were measured in 123 rhesus macaques aged 2 – 24 years old. (A) Hierarchical clustering analysis was performed on 18 plasma cytokines levels. Sex; red = female and blue = male. (B) A pairwise Spearman correlation matrix was performed for cytokine levels and age. Red indicates positive correlations (r > 0) and blue indicates negative correlations (r < 0). P < 0.05 was considered statistically significant. Scatter plots were shown between G-CSF and IL-1β (C), MIP-1α (D) with the ages of the animals indicated by the color intensity and size of the dots. Spearman nonparametric correlation coefficient (r) and P value were calculated.

FIGURE 4.. Cytokines that directly correlated with G-CSF levels in rhesus macaques aged 2 – 24 years old.

(A) Spearman nonparametric correlations and pairwise scatter plots between G-CSF and representative examples of cytokines are shown. Note the levels for IL-8 were plotted on a log scale. P < 0.05 was considered significant. NS, not significant. (B) Hierarchical clustering analysis of 18 plasma cytokines levels in 123 rhesus macaques that were sorted by age. (i.e. young to old: dark to light, Sex; red = female and blue = male)

FIGURE 5.

Neutrophils in elderly rhesus macaques express lower levels of MPO. Whole blood samples from rhesus macaques ranging in age from 5.6 – 27.8 years old were stained for surface and intracellular markers. MPO expression was measured as median florescent intensity (MFI) in neutrophils gated from single cells/live/CD45⁺/CD3^-/CD20^-/HLA-DR^-/CD66abce⁺. (A) Spearman nonparametric correlation test was applied. (B) Unpaired Mann-Whitney U test was used to compare MPO MFI expression in the younger animals aged 5.6 – 8.8 years old with that in older animals aged 17.9 – 27.83 years old. ** P < 0.01.

FIGURE 6.. Increased frequency and absolute numbers of CD33+ PMN-MDSC-like cells were observed in elderly rhesus macaques over 20 years of age.

(A) A representative gating strategy is shown for PMN-MDSCs in PBMCs of younger (5–10 year-old) and elderly (20–26 year-old) rhesus macaques. PMN-MDSCs were gated as single/live/CD45⁺/CD3^-/CD20^-/HLA-DR^-/CD66abce⁺/CD33⁺. (B) The absolute counts of CD33⁺ PMN-MDSCs were plotted against age of each animal and analyzed by Spearman correlation test. P < 0.05 was considered statistically significant. (C) The percentages of CD33⁺ PMN-MDSCs in CD66abce⁺ low-density neutrophils were plotted along with median and interquartile range for comparison between the younger and older rhesus macaques by Mann-Whitney U Test.