Iron overload induces cerebral endothelial senescence in aged mice and in primary culture in a sex-dependent manner

Abstract

Iron imbalance in the brain negatively affects brain function. With aging, iron levels increase in the brain and contribute to brain damage and neurological disorders. Changes in the cerebral vasculature with aging may enhance iron entry into the brain parenchyma, leading to iron overload and its deleterious consequences. Endothelial senescence has emerged as an important contributor to age-related changes in the cerebral vasculature. Evidence indicates that iron overload may induce senescence in cultured cell lines. Importantly, cells derived from female human and mice generally show enhanced senescence-associated phenotype, compared with males. Thus, we hypothesize that cerebral endothelial cells (CEC) derived from aged female mice are more susceptible to iron-induced senescence, compared with CEC from aged males. We found that aged female mice, but not males, showed cognitive deficits when chronically treated with ferric citrate (FC), and their brains and the brain vasculature showed senescence-associated phenotype. We also found that primary culture of CEC derived from aged female mice, but not male-derived CEC, exhibited senescence-associated phenotype when treated with FC. We identified that the transmembrane receptor Robo4 was downregulated in the brain vasculature and in cultured primary CEC derived from aged female mice, compared with those from male mice. We discovered that Robo4 downregulation contributed to enhanced vulnerability to FC-induced senescence. Thus, our study identifies Robo4 downregulation as a driver of senescence induced by iron overload in primary culture of CEC and a potential risk factor of brain vasculature impairment and brain dysfunction.

Keywords: cellular senescence; cerebral endothelial cells; iron overload; molecular biology of aging; sex characteristics.

FIGURE 1

Iron overload impaired recognition and conditioned memory in aged female mice. (a) Relative body weight in mice from FC‐treated and control groups. (b) Absolute body weight 6 weeks after the initiation of the treatments. (c–i) Open field: velocity (c), distance moved (d), percentage of time not moving (e), percentage of time spent in the center (f), number of visits to the center (g), percentage of time spent in the borders (h), and number of visits to the borders (i). (j) Percentage of alternation. (k) Novel object recognition index. (l) Percentage of freezing time. All data are mean ± SEM with n = 6–8. Two‐way ANOVA test, Tukey's multiple comparisons test, *p < 0.05, **p < 0.01.

FIGURE 2

FC treatment enhanced iron deposition in brains of aged female mice, but not in male mouse brains. (a) Representative images of brain sections (cortex region) from aged male and female mice treated with FC or a vehicle stained with Prussian blue (blue) and Nuclear Fast Red (pink). Black arrows depict discrete iron deposits. Scale bar, 100 μm. (b) Number of Prussian blue‐positive deposits per area from (a). Six brain sections per mouse were stained and quantified. Data are mean ± SEM with n = 4. Two‐way ANOVA test, Tukey's multiple comparisons test, *p < 0.05. (c) Iron content in blood cells of mice of the experimental groups. (d) Iron content in the brain lysates of the experimental groups. Data are mean ± SEM with n = 7–8. ***p < 0.001, ****p < 0.0001. Data are mean ± SEM with n = 4–6.

FIGURE 3

The brain vasculature of aged female mice was more vulnerable to FC than that of aged male mice. (a) Representative images of the hippocampi of mice from control or FC groups stained with anti‐γH2AX (brown) and Hematoxylin (purple). Scale bar, 250 μm. (b) Relative area positive to γH2AX from (a). Data are mean ± SEM with n = 5. (c) Representative image of a cerebral vessel positive to γH2AX (black arrows). Scale bar, 25 μM. (d) Representative images of cerebral microvessels from mice from control or FC groups stained with anti‐γH2AX (green) anti‐ICAM‐1(red) and Hoechst (blue). Scale bar, 25 μM. (e) Puncta index of γH2AX from (d). Data are mean ± SEM pooled from >100 microvessels from four mice/group. (f–h) Gene expression of p16 (f), p21 (g), and IL6 (h) relative to Gapdh. Data are mean ± SEM pooled from four mice/group. All experiments were analyzed by two‐way ANOVA test, Tukey's multiple comparisons test, *p < 0.05.

FIGURE 4

The transcriptome of cultured CEC isolated from aged mice showed sex differences. (a) Heatmap of the 2500 most expressed genes in cultured CEC isolated from 18 to 20 m/o male and female mice. (b,c) Representation of the most enriched KEGG pathways containing the most significant upregulated (b) and downregulated (c) genes in aged female‐derived CEC, compared with aged male‐derived CEC. Each KEGG pathway shows a p‐value and, in brackets, the number of genes in each group. (d) Heatmap of a SA genes in cultured CEC isolated from 18 to 20 m/o male and female mice. (e) Dot plot and bar graph of the percentage of SA‐β‐Galactosidase activity in CD31‐positive cells isolated from aged male and female mice. Data are mean ± SEM pooled from three mice/group. Student's t test, *p < 0.05.

FIGURE 5

Cultured primary CEC isolated from aged female mice were more susceptible to FC than CEC from male mice. (a) Representative images of the scratch assay in cultured CEC isolated from aged male and female mice and treated with FC (50 or 150 μM), or a vehicle. Images show wound right after (0 h) and 48 h after (48 h) a scratch and treatment initiation. Scale bar, 200 μm. (b,c) Percentage of the wound closure in male CEC (b) and female CEC (c) treated with FC or a vehicle. (d) Percentage of wound closure 48 h after the initiation of the treatments. Data are mean ± SEM pooled from three mice/group. (e) Representative images of cultured CEC from male and female mice treated with FC (50 or 150 μM), or a vehicle. 7 days after the treatments, cells were fixed and stained with anti‐γH2AX (green) and the nuclear Hoechst dye (blue). Scale bar, 25 μm. (f) Puncta index of γH2AX in CEC from (e). Data are mean ± SEM pooled from >500 cells from three mice/group. (g–i) Gene expression of p16 (g), p21 (h), and IL6 (i) relative to Gapdh. Data are mean ± SEM pooled from three mice/group. All experiments were analyzed by two‐way ANOVA test, Tukey's multiple comparisons test, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

FIGURE 6

Robo4 downregulation sensitized aged male mice‐derived CEC to FC. (a) Volcano plot showing fold changes for genes differentially expressed between aged female mouse‐derived CEC versus aged male mouse‐derived CEC in culture. (b) Gene expression of Robo4 relative to Gapdh in cultured CEC from aged mice of both sexes. Data are mean ± SEM pooled from three mice/group. Student's t test, *p < 0.05. (c) Gene expression of Robo4 relative to Gapdh in brain microvessel fractions from aged male and female mice. Data are mean ± SEM pooled from three mice/group. Student's t test, *p < 0.05. (d) Cultured CEC from male mice were transfected with siRNA targeting Robo4, or non‐targeting siRNA, and treated with FC, or a vehicle. Line graph shows the percentage of the wound closure every 8 h for 48 h. (e) Percentage of wound closure 48 h after the initiation of the treatments. (f) Puncta index of γH2AX in CEC 7 days after the treatments. Data are mean ± SEM pooled from >250 cells from three mice/group. Two‐way ANOVA test, Tukey's multiple comparisons test, *p < 0.05, **p < 0.01. (g–i) Gene expression of p16 (g), p21 (h), and IL6 (i) relative to Gapdh. Data are mean ± SEM pooled from two independent experiments from three mice/group. Two‐way ANOVA test, Tukey's multiple comparisons test, *p < 0.05, **p < 0.01.