Hydrogen-rich saline promotes microglia M2 polarization and complement-mediated synapse loss to restore behavioral deficits following hypoxia-ischemic in neonatal mice via AMPK activation

Abstract

Background: Hypoxia-ischemia (HI) during the perinatal period is one of the most common causes of acute mortality and chronic neurologic morbidity. Hydrogen-rich saline (HS) treatment in neonatal mice has been reported to alleviate brain injury following HI, but the mechanisms involved are not known.

Methods: A modified version of the Rice-Vannucci method for the induction of neonatal HI brain injury was performed on postnatal day 7 mouse pups. Animals or BV2-cells received HS and an AMPK inhibitor at indicative time post-injury.

Results: In the current study, we show that HS treatment attenuated the accumulation of CD11b⁺/CD45^high cells, suppressed HI-induced neuro-inflammation, induced microglial anti-inflammatory M2 polarization, was associated with promoting AMPK activation, and inhibited nuclear factor-κB activation as demonstrated both in vivo and in vitro. In addition, HS treatment reversed HI-induced neurological disabilities, was associated with improving damaged synapses, and restored the expression levels of synaptophysin and postsynaptic density protein 95 following HI insult. Furthermore, HI insult which increased levels of complement component C1q, C3, and C3aR1 was observed. Importantly, C1q deposited in the infarct core and lesion boundary zone following HI injury, was found to co-localize within regions of synapse loss, whereas HS treatment reversed these effects of HI on synapse loss and complement component levels. Notably, the AMPK inhibitor reversed the beneficial effects of HS as described above.

Conclusions: These results demonstrate that HS restored behavioral deficits after HI in neonatal mice. These beneficial effects, in part, involve promoting microglia M2 polarization and complement-mediated synapse loss via AMPK activation.

Keywords: Complement; Hydrogen-rich saline; Hypoxia-ischemia; M2 polarization; Synapse loss.

Fig. 1

Effects of HS on AMPK activation. a BV-2 cells treated with different concentration of HS for 2 h. b BV-2 cells treated with 1 μM HS for different times. c BV-2 cells treated with 1 μM HS with/without 500 ng/mL LPS for 2 or 4 h. d BV-2 cells treated with 1 μM HS with/without 500 ng/mL LPS or 1 μM Compound C (CC) for 2 h. Levels of p-AMPK, AMPK and β-actin were determined with use of Western blot. e Representative immunoblots of protein levels for phosphorylated AMPK (p-AMPK), AMPK and β-actin in ipsilateral cortex at 3 days following HI insult. N = 4/group. Values represent the mean ± SD, *p < 0.05, **p < 0.01 according to ANOVA

Fig. 2

Effects of HS on HI-induced microglial activation. a Representative Iba-1 staining within the core of infarcted cortex at 3 days following HI insult. N = 4/group. b Representative flow cytometric lots of CD11b⁺/CD45^high cells within ipsilateral cortex at 3 days following HI insult. N = 4/group. Right histograms of average percentage of CNS-associated phagocytes (CD11b⁺/CD45^high cells) in live single immune cells from brains. c Representative brain images as obtained at 72 h following HI. Arrows indicate sites of significant edema. Brain water content was determined at 72 h following HI insult, N = 6/group. d Representative samples stained with TTC. Quantified infarct volume is indicated by the white area. N = 6/group. Values represent the mean ± SD, *p < 0.05, **p < 0.01, ***p < 0.001 according to ANOVA

Fig. 3

HS significantly alters expression profiles of inflammation mediators. a Representative RT-PCR photographs showing mRNA levels of IL-1β, CD86, TGFβ and YM-1 within ipsilateral cortex at 3 days following HI insult. Quantitative results of relative mRNA levels within each group. Values were normalized to β-actin. N = 4/group. b Representative photographs of double immunofluorescent staining of CD16 (green), CD206 (green), and Iba-1 (red) within the ipsilateral cortex at 3 days following HI insult. Scale bar = 20 μm. Six randomly captured images (× 20) were obtained for each section per animal. N = 4/group. c BV-2 cells treated with 1 μM HS with/without 500 ng/mL LPS or 1 μM Compound C for 2 h. The mRNA levels of IL-1β, CD86, TGFβ and YM-1 were determined with use of RT-PCR. N = 4/group. Values represent the mean ± SD, * p < 0.05, **p < 0.01, ***p < 0.001 according to ANOVA

Fig. 4

Effect of HS on NF-κB activation in lesioned cortex and BV-2 cells. a Representative immunoblots are shown for phosphorylated NF-κB (p-NF-κB), total NF-κB, phosphorylated-IκBα (p-IκBα), and total IκBα within ipsilateral cortex at 3 days following HI insult as determined with western blotting analysis. Quantification of the relative levels of p-NF-κB/NF-κB, p-IκBα/ IκBα within each group. N = 4/group. b BV-2 cells treated with 1 μM HS and with/without 500 ng/mL LPS for 2 h. The level of p-NF-κB/NF-κB, p-IκBα/ IκBα were measured by Western blot. N = 4/group. Values were normalized to β-actin. Values represent the mean ± SD, *p < 0.05, **p < 0.01, ***p < 0.001 according to ANOVA

Fig. 5

Effects of HS on structural remodeling of synapses following HI injury. a Representative electron micrographs at 28 days following HI indicating: * - presynaptic vesicles, Yellow arrow - postsynaptic partners, Green arrowhead - widened synaptic clefts, and Yellow arrowhead - destroyed synapses. Scale bar = 2 μm. b-d Levels of PSD95 and Syn within the ipsilateral cortex as examined at 3, 14 or 28 days post-HI with use of Western blot. Bar graphs show quantifications of protein levels at 3, 14 or 28 days. N = 4/group. Values represent the mean ± SD, * p < 0.05, **p < 0.01, ***p < 0.001 according to ANOVA

Fig. 6

HS suppresses HI-induced complement activation in the lesioned cortex. a Representative RT-PCR photographs showing mRNA levels of C1q, C3 and C3aR1 within ipsilateral cortex at 3 days post-HI. Quantitative results of relative mRNA levels within each group. Values were normalized to β-actin. N = 4 mice/group. b Representative photographs of C1q staining within ipsilateral cortex taken at 72 h following HI. c Magnified views of boxed regions in B showing C1q staining. Quantification of C1q⁺ cells. N = 6/group. d BV-2 cells treated with 1 μM HS with/without 500 ng/mL LPS or 1 μM Compound C for 2 h. The mRNA levels of C1q, C3 and C3aR1 were determined with use of RT-PCR. Values represent the mean ± SD, *p < 0.05, **p < 0.01, ***p < 0.001 according to ANOVA

Fig. 7

HS suppresses HI-induced synapse deficits linked to regulating complement expression in the lesioned cortex. a Double staining PSD95 (red) and C1q (Green) within the ipsilateral cortex were examined at 3 days post-HI. C1q labeled cells (green) were detected in the infract core and lesion boundary zone (IBZ). Scale bar = 500 μm. b Magnified views of boxed regions in A showing PSD95 and C1q staining. Scale bar = 50 μm. c Magnified views of boxed regions in b showing PSD95 and C1q staining. Yellow arrows showing PSD95 and C1q co-location. Red arrows showing C1q-positive cells and PSD95-negative cells. Scale bar = 20 μm. N = 4/group. Values represent the mean ± SD, **p < 0.01, ***p < 0.001 according to ANOVA

Fig. 8

Effects of HS on neurological reflexes following HI insult. Daily performance of a front-limb suspension reflexes and b negative geotaxis of each group. c Comparisons of NORT test discrimination ratios which represent the ability to discriminate between a novel versus familiar object: novel/(novel + familiar time). d During each acquisition trial in the MWM test, escape latencies (in seconds) were measured and analyzed over days 1 to 5. e On the sixth day, the escape latency and percent time spent in the target quadrant were recorded (in seconds) and analyzed. f Swimming speeds of all groups from days 1 to 6 in the MWM test. g Display of tracks of all groups on day 6. Values represent the mean ± SD, N = 10/group. *p < 0.05, **p < 0.01, ***p < 0.001 according to ANOVA