Humanin, a Mitochondrial-Derived Peptide Released by Astrocytes, Prevents Synapse Loss in Hippocampal Neurons

Abstract

Astroglial cells are crucial for central nervous system (CNS) homeostasis. They undergo complex morpho-functional changes during aging and in response to hormonal milieu. Ovarian hormones positively affect different astroglia parameters, including regulation of cell morphology and release of neurotrophic and neuroprotective factors. Thus, ovarian hormone loss during menopause has profound impact in astroglial pathophysilogy and has been widely associated to the process of brain aging. Humanin (HN) is a secreted mitochondrial-encoded peptide with neuroprotective effects. It is localized in several tissues with high metabolic rate and its expression decreases with age. In the brain, humanin has been found in glial cells in physiological conditions. We previously reported that surgical menopause induces hippocampal mitochondrial dysfunction that mimics an aging phenotype. However, the effect of ovarian hormone deprivation on humanin expression in this area has not been studied. Also, whether astrocytes express and release humanin and the regulation of such processes by ovarian hormones remain elusive. Although humanin has also proven to be beneficial in ameliorating cognitive impairment induced by different insults, its putative actions on structural synaptic plasticity have not been fully addressed. In a model of surgical menopause in rats, we studied hippocampal humanin expression and localization by real-time quantitative polymerase chain reaction (RT-qPCR) and double immunohistochemistry, respectively. Humanin production and release and ovarian hormone regulation of such processes were studied in cultured astrocytes by flow cytometry and ELISA, respectively. Humanin effects on glutamate-induced structural synaptic alterations were determined in primary cultures of hippocampal neurons by immunocytochemistry. Humanin expression was lower in the hippocampus of ovariectomized rats and its immunoreactivity colocalized with astroglial markers. Chronic ovariectomy also promoted the presence of less complex astrocytes in this area. Ovarian hormones increased humanin intracellular content and release by cultured astrocytes. Humanin prevented glutamate-induced dendritic atrophy and reduction in puncta number and total puncta area for pre-synaptic marker synaptophysin in cultured hippocampal neurons. In conclusion, astroglial functional and morphological alterations induced by chronic ovariectomy resemble an aging phenotype and could affect astroglial support to neuronal function by altering synaptic connectivity and functionality. Reduced astroglial-derived humanin may represent an underlying mechanism for synaptic dysfunction and cognitive decline after menopause.

Keywords: astrocytes; hippocampus; humanin; mitochondria; ovarian hormones; synapse.

Figure 1

Behavioral characterization of an animal model of surgical menopause in adult rats. Adult Wistar female rats were ovariectomized (OVX) or sham-operated (SHAM). Twelve weeks after surgery, the animals were subjected to behavioral tests as described in “Materials and Methods” section. (A) Y-maze spontaneous alternation test: for each animal, the number of total arm entries and the number of triads in 6 min was recorded. (B) Elevated plus maze test: each animal was allowed to freely explore for 5 min and the number of times the animal entered an arm with all four paws was recorded. (C) Forced swimming test: every 5 s, a time-sampling technique was used to score the presence of immobility, swimming or climbing behavior. (D) SHAM or OVX rats were placed individually in the center of a field marked with a grid of 16 equal squares for 10 min. The number of times the animal crossed each line was registered. Each column represents the mean ± SEM of (A) the percentage of alternation, (B) the percentage of open arm entries, (C) the number of counts or (D) the number of crossings per session (n = 4–12 animals/group); *p < 0.05, ***p < 0.001, Student’s t-test.

Figure 2

Expression of HNr in the hippocampus. The expression of HNr RNA was evaluated in hippocampi from SHAM and OVX rats by real-time quantitative polymerase chain reaction (RT-qPCR). Each column represents the mean ± SEM of the concentration of HNr RNA relative to its internal control HPRT expressed as fold-changes relative to SHAM group (AU; n = 3 animals/group); *p < 0.05, Student’s t-test.

Figure 3

Double immunostaining for HNr and astroglial, neuronal or oligodendroglial markers. Coronal sections from the hippocampus of SHAM and OVX animals were processed for double immunohistochemistry for HNr and (A) astroglial (S100B), (B) neuronal (NF-200) or (C) oligodendroglial (OLIG-2) markers. Representative confocal microphotographs from CA1 stratum oriens (600×) show the expression of HNr (red), S100B, NF-200 or OLIG2 (green) and the merged image (yellow). PYR, pyramidal layer; CC, corpus callosum. Scale bar = 50 μm.

Figure 4

Effects of long-term ovarian hormone deprivation on HNr and glial fibrillary acidic protein (GFAP) immunoreactivity in the hippocampus. Coronal sections from the hippocampus of SHAM and OVX animals were processed for double immunohistochemistry for HNr and GFAP. (A) Representative confocal microphotographs from CA1 stratum oriens (600×) show the expression of HNr (red), GFAP (green) and the merged image (yellow). (B) Quantification of relative immunoreactive area (immunoreactive area/total area) for HNr and GFAP using ImageJ software. Each column represents the mean ± SEM of relative immunoreactive area (n = 3 animals/group). *p < 0.05, **p < 0.01, ***p < 0.001, Student’s t-test. (C) HNr and GFAP expression levels expressed as relative immunopositive area for each protein were normalized with respect to each corresponding hippocampal subregion (SHAM r = 0.74, p < 0.01; OVX r = 0.63, p < 0.05, Pearson correlation test). PYR, pyramidal layer; CC, corpus callosum; DG, dentate gyrus. Scale bar = 50 μm.

Figure 5

HNr production and release by astrocytes in vitro. Cultured astrocytes were incubated with estradiol (E) and progesterone (P). HNr secreted levels were determined in conditioned media by ELISA. Harvested cells from additional cultures were immunostained for HNr and analyzed by flow cytometry. Each column represents the mean ± SEM of (A) the concentration of HNr in conditioned media normalized to 50 μg of total protein in corresponding cell lysate, (B) the fluorescence intensity of HNr staining (Gmean) or (C) the percentage of HNr-positive cells. The upper panels in (B,C) show representative histograms and dot plots of HNr expression in astrocytes incubated with VEH or E + P (n = 3–4 wells per group from three independent experiments). (A) *p < 0.05 vs. respective control without E; ANOVA followed by Tukey’s test, (B,C) *p < 0.05; Student’s t-test.

Figure 6

Effects of glutamate and HN on hippocampal neuronal cell viability in vitro. Cultured embryonic neurons (DIV13) were subjected to a 3-min incubation with glutamate (5 μM) immediately followed by a 24-h incubation with HN (0.01–1 μM). Cell viability was assessed by MTT assay. Each column represents the mean ± SEM of five wells from one experiment representative of three independent experiments. ns, non-significant; ANOVA.

Figure 7

Effect of HN on neuronal dendritic tree area in hippocampal neurons in vitro. Hippocampal neurons in culture (DIV 13) were briefly exposed to glutamate (5 μM), immediately incubated with HN (0.01–1 μM) and evaluated 24 h later. (A) Representative microphotographs of hippocampal neurons in culture immunostained for MAP-2. Microphotographs corresponding to 0.01 μM HN are shown, (B) Quantification of MAP-2 immunostaining using ImageJ software. Each column represents the mean ± SEM of 20–40 neurons per experimental condition. ^∧∧∧p < 0.01 vs. respective control without glutamate, ***p < 0.001 vs. respective control without HN; two-way ANOVA followed by Tukey’s test. Scale bar = 50 μm.

Figure 8

Effect of HN on SYN synaptic profile in hippocampal neurons in vitro. Hippocampal neurons in culture (DIV 13) were briefly exposed to glutamate (5 μM), immediately incubated with HN (0.01–1 μM) and evaluated 24 h later. (A) Representative microphotographs of hippocampal neurons in culture immunostained for SYN. Microphotographs corresponding to 0.01 μM HN are shown. Insets (3× magnification) detail SYN immunostaining pattern. (B) Quantification of total SYN puncta area, SYN puncta number and individual SYN puncta area using ImageJ software. Each column represents the mean ± SEM of 20–40 neurons per experimental condition. ^∧∧∧p < 0.01 vs. respective control without glutamate, **p < 0.01, ***p < 0.001 vs. respective control without HN; two-way ANOVA followed by Tukey’s test. Scale bar = 50 μm.