Sex-Dependent Pathology in the HPA Axis at a Sub-acute Period After Experimental Traumatic Brain Injury

Abstract

Over 2.8 million traumatic brain injuries (TBIs) are reported in the United States annually, of which, over 75% are mild TBIs with diffuse axonal injury (DAI) as the primary pathology. TBI instigates a stress response that stimulates the hypothalamic-pituitary-adrenal (HPA) axis concurrently with DAI in brain regions responsible for feedback regulation. While the incidence of affective symptoms is high in both men and women, presentation is more prevalent and severe in women. Few studies have longitudinally evaluated the etiology underlying late-onset affective symptoms after mild TBI and even fewer have included females in the experimental design. In the experimental TBI model employed in this study, evidence of chronic HPA dysregulation has been reported at 2 months post-injury in male rats, with peak neuropathology in other regions of the brain at 7 days post-injury (DPI). We predicted that mechanisms leading to dysregulation of the HPA axis in male and female rats would be most evident at 7 DPI, the sub-acute time point. Young adult age-matched male and naturally cycling female Sprague Dawley rats were subjected to midline fluid percussion injury (mFPI) or sham surgery. Corticotropin releasing hormone, gliosis, and glucocorticoid receptor (GR) levels were evaluated in the hypothalamus and hippocampus, along with baseline plasma adrenocorticotropic hormone (ACTH) and adrenal gland weights. Microglial response in the paraventricular nucleus of the hypothalamus indicated mild neuroinflammation in males compared to sex-matched shams, but not females. Evidence of microglia activation in the dentate gyrus of the hippocampus was robust in both sexes compared with uninjured shams and there was evidence of a significant interaction between sex and injury regarding microglial cell count. GFAP intensity and astrocyte numbers increased as a function of injury, indicative of astrocytosis. GR protein levels were elevated 30% in the hippocampus of females in comparison to sex-matched shams. These data indicate sex-differences in sub-acute pathophysiology following DAI that precede late-onset HPA axis dysregulation. Further understanding of the etiology leading up to late-onset HPA axis dysregulation following DAI could identify targets to stabilize feedback, attenuate symptoms, and improve efficacy of rehabilitation and overall recovery.

Keywords: astrocytosis; diffuse axonal injury; diffuse traumatic brain injury; glucocorticoid receptors; hypothalamic-pituitary-adrenal axis; microglia; neuroinflammation; sex-differences.

Figure 1

Males lost more weight after diffuse TBI. (A) Righting reflex times for male and female rats did not significantly differ (t₁₈ = 1.685; p = 0.1093). (B) mFPI had a significant effect on weight [F_{(1, 37)} = 20.73; p = 0.002], with 7 DPI males losing more weight than their sex-matched sham controls (p = 0.0003; male 7 DPI n = 10; female 7 DPI n = 11). There was also an interaction between injury and sex [F_{(1, 37)} = 5.340; p = 0.0265], where weight loss at 7 DPI depends on the rat's sex. Data are represented by the mean + SEM; *Difference from same-sex sham; ^†overall injury effect; male sham n = 11; 7 DPI male n = 9; female sham n = 10; 7 DPI female n = 11.

Figure 2

Gene expression of CRH in the hypothalamus did not change at 7 DPI. (A) CRH gene expression in the hypothalamus was similar in both males (t₉ = 0.7959; p = 0.4466), and females (t₁₀ = 0.1258; p = 0.9024) after injury. (B) Results in the hippocampus were similar to hypothalamus, males (t₈ = 0.3830; p = 0.7117) and females (t₈ = 0.3348; p = 0.7463). Data are represented by the mean + SEM; hippocampus: male sham n = 4; 7 DPI male n = 6; female sham n = 4; 7 DPI female n = 6; hypothalamus: male sham n = 4; 7 DPI male n = 7; female sham n = 4; 7 DPI female n = 8.

Figure 3

ACTH and adrenal gland weights did not change at 7DPI. (A) There was no injury effect on ACTH [F_{(1, 31)} = 1.591; p = 0.217]. Female rats also had significantly higher ACTH levels compared with males [F_{(1, 31)} = 8.742; p = 0.006]. (B) Adrenal weights were normalized to the body weights of each animal to calculate the organ index. There was no effect of injury [F_{(1, 37)} = 2.944; p = 0.0946], but the female organ index was significantly higher compared with the organ index of males [F_{(1, 37)} = 83.84; p < 0.0001]. Data are represented by the mean + SEM. ^#difference from opposite sex; ACTH: male sham n = 8; 7 DPI male n = 9; female sham n = 8; 7 DPI female n = 10. Adrenals: male sham n = 11; 7 DPI male n = 9; female sham n = 10; 7 DPI female n = 11.

Figure 4

Diffuse TBI activated microglia in the PVN of male rats at 7 DPI. (A) 40× representative images of Iba-1 staining in the male (top), female (bottom), sham (left), and 7 DPI (right). (B–D) Results from Skeleton Analysis of microglia in the PVN. (B) There were significantly more Iba-1 positive cells in males at 7 DPI compared to same-sex shams, but not in females [F_{(1, 16)} = 12.32; p = 0.0029]. (C) The endpoints/microglia approached statistical significance at 7 DPI as a function of injury [F_{(1, 16)} = 3.337; p = 0.0865]. When stratified by sex and analyzed for an injury effect, there was a statistically significant difference among males (p < 0.05) but not among females (p = 0.673). There was no overall effect of sex [F_{(1, 16)} = 0.3917; p = 0.5402]. (D) There was an injury effect on the branch length/microglia [F_{(1, 16)} = 4.879; p = 0.0421] but there were no distinct sex effects [F_{(1, 16)} = 2.078; p = 0.1688]. Scale bar = 100 μm; data are represented by the mean + SEM; ^†overall injury effect; *difference from same-sex sham; n = 5/group.

Figure 5

There was no evidence of reactive astrogliosis in the PVN at 7 DPI. (A) 40× representative images of GFAP in the male (top) and female (bottom), sham (left), and 7 DPI (right). (B) The image within the box in (A) is magnified to demonstrate how cell counts were made, where inclusion criteria required the presence of the soma (yellow arrows). (C) Pixel density of GFAP staining did not change as a function of injury [F_{(1, 16)} = 1.276; p = 0.2753] or sex [F_{(1, 16)} = 0.4245; p = 0.5240]. (D) Cell counts of GFAP did not change as a function of injury [F_{(1, 16)} = 0.5068; p = 0.4868] or sex [F_{(1, 16)} = 0.4628; p = 0.5061]. Scale bar = 100 μm; data are represented by the mean + SEM; n = 5/group.

Figure 6

Diffuse TBI activated microglia in the hippocampus across both sexes at 7 DPI. (A) 40× representative images of Iba-1 staining in the male (top), female (bottom), sham (left), and 7 DPI (right). (B) There was an injury effect on average microglia cell counts at 7 DPI [F_{(1, 16)} = 5.58; p < 0.0001] in the hippocampus. Both injured males and females had more microglia compared to uninjured shams. Additionally, there was an overall sex effect on number of microglia in the hippocampus [F_{(1, 16)} = 8.603; p = 0.0097]. There were more microglia in males compared with females. There was an interaction between sex and injury [F_{(1, 16)} = 16.60; p = 0.0009]. (C) There was an overall injury effect on microglia process endpoints per cell [F_{(1, 16)} = 22.51; p = 0.0002]. endpoints per cell [F_{(1, 16)} = 22.51; p = 0.0002]. post-hoc analysis indicated there were fewer microglia process endpoints per cell at 7 DPI in males (p = 0.0034), whereas females approached significance (p = 0.095) compared to respective shams in the hippocampus. (D) There was an overall injury effect on process branch lengths [F_{(1, 16)} = 20.03; p = 0.0004] in the hippocampus. post-hoc analysis indicated that male injured rats had shorter process branch lengths compared with male shams. Scale bar = 100 μm; data are represented by the mean + SEM; ^†overall injury effect; ^#difference from opposite sex; *difference from same-sex sham; n = 5/group.

Figure 7

Diffuse TBI led to astrocytosis in the DG across both sexes at 7 DPI. (A) 40× representative images of GFAP in the male (top), female (bottom), sham (left), and 7 DPI (right). (B) There was an injury effect on pixel density of GFAP [F_{(1, 16)} = 124.2; p < 0.0001], where both injured males (p < 0.0001) and injured females (p < 0.0001) had significantly greater in comparison to their sex-matched shams. There was also an effect of sex [F_{(1, 16)} = 6.771, p = 0.0193], with females having a significantly lower density of GFAP in comparison to males. (C) There was an injury effect on cell counts of GFAP stained astrocytes [F_{(1, 16)} = 37.25; p < 0.0001] but no sex effect [F_{(1, 16)} = 2.481; p = 0.1348]. Scale bar = 100 μm; data are represented by the mean + SEM; ^#difference from opposite sex; *difference from same-sex sham; n = 5/group.

Figure 8

Diffuse TBI led to higher GR protein levels in the hippocampus of females but not males at 7 DPI. (A) Gene and protein expression of GR in the hypothalamus did not change as a function of injury in males (t₁₀ = 1.689; p = 0.1220) or females (t₁₀ = 0.3562; p = 7291). (B) There were no TBI-induced differences in gene expression of GR in the hippocampus of males (t₈ = 2.042; p = 0.0754) or females (t₁₂ = 1.651; p = 0.1246). (C) There were no TBI-induced differences in GR protein levels in the hypothalamus in males (t₁₄ = 0.022134; p = 0.9833) or females (t₉ = 1.263; p = 0.2365). (D) There was a significant injury-induced difference in protein levels of GR in the hippocampus in females (t₁₀ = 2.797; p = 0.0189), but not males (t₁₇ = 0.1527; p = 0.8805). Data are represented by the mean + SEM; *difference from same-sex sham; male sham n = 4–5; 7 DPI male n = 6–7; female sham n = 4–6; 7 DPI female n = 8.