Age-Associated Decline in Dendritic Cell Function and the Impact of Mediterranean Diet Intervention in Elderly Subjects

Abstract

Introduction: Aging is accompanied by increased susceptibility to infection and age-associated chronic diseases. It is also associated with reduced vaccine responses, which is often attributed to immunosenescence and the functional decline of the immune system. Immunosenescence is characterized by a chronic, low-grade, inflammatory state termed inflammaging. Habitants of Mediterranean (MED) regions maintain good health into old age; often attributed to MED diets.

Hypothesis: Adoption of a MED-diet by elderly subjects, in Norfolk (UK), may improve immune responses of these individuals and in particular, dendritic cell (DC) function.

Experimental approach: A total of 120 elderly subjects (65-79 years old) recruited onto the Nu-AGE study, a multicenter European dietary study specifically addressing the needs of the elderly, across five countries, and were randomized to the control or MED-diet groups, for one year. Blood samples were taken pre- and post-intervention for DC analysis and were compared with each other, and to samples obtained from 45 young (18-40 years old) subjects. MED-diet compliance was assessed using high performance liquid chromatography-with tandem mass spectrometry analysis of urine samples. Immune cell and DC subset numbers and concentrations of secreted proteins were determined by flow cytometric analysis.

Results: As expected, reduced myeloid DC numbers were observed in blood samples from elderly subjects compared with young. The elevated secretion of the adipokine, resistin, after ex vivo stimulation of peripheral blood mononuclear cells from elderly subjects, was significantly reduced after MED-diet intervention.

Conclusion: This study provides further evidence of numerical and functional effects of aging on DCs. The MED-diet showed potential to impact on the aging immune cells investigated and could provide an economical approach to address problems associated with our aging population.

Keywords: Mediterranean diet; aging; cytokines; dendritic cells; resistin.

Figure 1

Myeloid DC (mDC) counts for the Im-AGE (young) and pre-intervention Nu-AGE (elderly) cohorts. Whole blood was stained with antibodies reactive with CD1c to identify mDCs, CD303 to identify pDCs, and to CD14 and CD19 to exclude CD14⁺ monocytes and CD19⁺ B cells, of which a high proportion express CD1c. n = 45 in the young cohort and n = 120 in the elderly cohort. Box and whisker plot of mDC (A) and pDC (B) counts for young and elderly subjects extending from the 25th to the 75th percentiles with the line through the box representing the median and plus (+) representing the mean value. Whiskers were determined using Tukey’s method using the 25th and 75th percentile plus 1.5 times the interquartile range (IQR) as the end of the whiskers. Dots represent individual participants where the values fell above the 25th or 75th quartile plus 1.5 times the IQR. Welch–Satterthwaite t-test on rank transformed data was used to determine the presence of differences between the young and elderly cohorts; significance assumed at P < 0.05, **P < 0.01, P = 0.0043 (mDCs) and P = 0.3108 (pDCs).

Figure 2

Scatter plots showing proportion of cytokine producing dendritic cells from young and elderly subjects. Peripheral blood mononuclear cells (PBMCs) were cultured in vitro for 3 h with tissue culture media alone (A,C) or in media containing lipopolysaccharide (LPS) and R848 (B,D) in the presence of 2 µM monensin. PBMC samples were subsequently surface stained with monoclonal antibodies against HLA-DR, CD1c, CD303, and CD304, CD14, CD16, CD19, and CD3, permeabilized and stained with anti-IL-1β, IL-6, and CXCL8 antibodies. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 indicate significance between age groups as measured by Mann–Whitney U tests and post hoc analyses, performed by one-way analysis of variance using the Kruskal–Wallis test with Dunn’s multiple comparisons post hoc test to identify any differences in proportions of the three cell types. Error bars show SEM and a horizontal line at the mean. CXCL8, C-X-C chemokine ligand 8; IL, interleukin; LPS, lipopolysaccharide, R848, resiquimod.

Figure 3

Myeloid DC (mDC) and plasmacytoid DC (pDC) counts after dietary intervention for control and MED diet study groups. Whole blood from control and dietary intervention groups was stained with anti-CD1c to identify mDCs, CD303 to identify pDCs, and anti-CD14 and CD19 to exclude CD14⁺ monocytes and CD19⁺ B cells, of which a high proportion express CD1c. The data show box and whisker plots of change in mDC (A) and pDC (B) counts from pre- to post-intervention in elderly subjects extending from the 25th to the 75th percentiles, with the horizontal line representing the median. Whiskers were determined using Tukey’s method using the 25th and 75th percentile plus 1.5 times the interquartile range (IQR) as the end of the whiskers. Squares represent individual participants where the values fell above the 25th or 75th quartile plus 1.5 times the IQR. A Mann–Whitney U test was used to determine the presence of differences between the control and dietary intervention cohorts. n = 58, Control group, n = 62, Mediterranean (MED)-diet group, significance assumed at *P < 0.05, **P < 0.01, P = 0.9104 (mDCs) and P = 0.3553 (pDCs).

Figure 4

Cytokine producing dendritic cells from elderly subjects at pre- and post-dietary intervention. Peripheral blood mononuclear cells from control subjects and Mediterranean (MED)-diet intervention subjects were cultured for 3 h in complete tissue culture media alone (A,C), or in media containing lipopolysaccharide (LPS) and R848 (B,D) in the presence of 2 µM monensin. Samples were subsequently surface stained with anti-HLA-DR, CD1c, CD303, and CD304, CD14, CD16, CD19, and CD3 antibodies, permeabilized and stained with anti-IL-1β IL-6 and CXCL8 antibodies. Squares represent values for each individual subject with filled squares (■) representing pre-intervention values and open squares (□) representing post-intervention values. Significance assumed at P < 0.05; *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 indicates significance between pre- and post-dietary intervention as measured by Wilcoxon matched pairs signed rank test, and post hoc analyses, performed by one-way analysis of variance using the Kruskal–Wallis test with Dunn’s multiple comparisons post hoc test to identify any differences in proportions of the three cell types. Error bars show SEM and a horizontal line at the mean. CXCL8, C-X-C chemokine ligand 8; IL, interleukin; LPS, lipopolysaccharide, R848, resiquimod.

Figure 5

Change in cytokine production by peripheral blood mononuclear cells (PBMCs) in response to lipopolysaccharide (LPS) and R848 stimulation in pre- versus post-dietary intervention. PBMCs from control subjects and Mediterranean (MED)-diet intervention subjects were cultured for 3 h in complete tissue culture media alone, or in media containing LPS and R848 in the presence of 2 µM monensin. PBMC culture supernatants were analyzed by multiplex bead based immunoassay to determine absolute concentrations of MCP-1 (A), CXCL8 (B), resistin (C) and TNF-α (D) in the PBMC samples. Concentrations (pg/ml) in non-stimulated samples were subtracted from stimulated samples to give a change in concentration as a result of the stimulus. Squares represent individual values for change in secretion of each cytokine by PBMCs, error bars represent the SEM. Determination of significant differences between pre- and post-intervention by paired t-tests, significance assumed at *P < 0.05, **P < 0.01, ***P < 0.001. CXCL8, C-X-C chemokine ligand 8, MCP-1, monocyte chemoattractant protein-1 ns; not statistically significant; TNF-α, tumor necrosis factor alpha.

Figure 6

Change from baseline (CFB) in cytokine production by PBMCs in response to lipopolysaccharide (LPS) and R848 stimulation after 1 year of control and MED-diet intervention. PBMCs from control subjects and MED-diet intervention subjects were cultured for 3 h in complete tissue culture media containing LPS and R848 in the presence of 2 µM monensin. PBMC culture supernatants were analyzed by multiplex bead based immunoassay to determine absolute concentrations of MCP-1 (A), CXCL8 (B), resistin (C) and TNF-α (D) in the PBMC samples. Concentrations (pg/mL) of each cytokine at pre-intervention were subtracted from the post-intervention values to give a CFB value. CFB in the control and MED-diet groups was compared using a Mann–Whitney U test, with significance assumed at *P < 0.05. CFB, change from baseline; CXCL8, C–X–C chemokine ligand 8; MCP-1, monocyte chemoattractant protein-1; MED: Mediterranean; PBMC, peripheral blood mononuclear cell; TNF-α, tumor necrosis factor alpha.