To observe the effect of electroacupuncture (EA) on the cognitive ability, hypoxia-inducible factor-1 alpha (HIF-1α)/6-phosphofructo-2-kinase/fructose-2, 6-bisphosphatase 3 (PFKFB3) signaling pathway and glycolytic function in the hippocampus of Alzheimer’s disease (AD) model mice, so as to elucidate its potential mechanism underlying improvement of cognitive function.

Male APP/PS1 transgenic mice were used as the AD model and randomly divided into model, EA and EA plus dimethyloxaloylglycine (DMOG, antagonist of α-ketoglutarate cofactor and inhibitor of HIF prolylhydroxylase) groups (n=6 each). Age-matched C57BL/6J mice served as the control group (n=6). EA stimulation (2 Hz, 1 mA) was applied to bilateral “Shenshu”(BL23) and “Zusanli”(ST36), acupuncture was applied to “Baihui” (GV20) for 20 min, once every other day for 4 weeks. The mice in the EA+DMOG group was received intraperitoneal injection of DMOG (50 mg/kg) solution on the basis of EA starting from the 4^th week. The mouse’s cognitive function was evaluated by using Morris water maze and novel object recognition tests. Hematoxylin-eosin (H.E.) staining and immunohistochemistry were used to observe hippocampal neuronal morphology and β-amyloid (Aβ)1-42 deposition. The contents of pyruvate kinase M2 (PKM2) and lactate dehydrogenase A (LDHA) in the hippocampal tissue were assayed using ELISA, and the activity of hippocampal hexokinase (HK) and content of lactic acid were detected using biochemical methods. Western blot was employed to detect the expression levels of hippocampal HIF-1α and PFKFB3 protein, and real-time quantitative PCR was employed to detect the expression levels of hippocampal HIF-1α, PFKFB3, HK, PKM2, and LDHA mRNAs. The immunofluorescence histochemistry technique was used to measure the fluorescence intensity of hippocampal LDHA.

Compared with the control group, the model group showed significantly longer escape latency, fewer platform crossings and lower novel object recognition index (P<0.01). After EA intervention, the escape latency was evidently shorter, the platform crossings and the novel object recognition index were apparently increased in the EA group (P<0.01), suggesting an improvement of the recognition ability and learning-memory ability. Following administration of DMOG, the effects of EA in shortening the escape latency, and increasing the number of platform crossings and novel object recognition index were eliminated (P<0.01, P<0.05, compared with the EA group). In contrast to the control group, the model group had a considerable increase in the optical density of Aβ1-42, contents of PKM2, LDHA, lactate, and HK activity, expression levels of HIF-1α and PFKFB3 proteins and mRNAs, and HK, PKM2, LDHA mRNAs, and LDHA fluorescence density (P<0.01). Compared with the model group, the increased levels of Aβ1-42 optical density, contents of PKM2, LDHA, lactate and HK activity, expression levels of HIF-1α and PFKFB3 proteins and mRNAs, and HK, PKM2, LDHA mRNAs, and LDHA fluorescence density were all reversed in the EA group (P<0.01). In comparison with the EA group, the decreased levels of Aβ1-42 optical density, contents of PKM2, LDHA, lactate, and HK activity, expression levels of HIF-1α and PFKFB3 proteins and mRNAs, and HK, PKM2, LDHA mRNAs, and LDHA fluorescence intensity were reversed by EA+DMOG (P<0.01, P<0.05). HE staining showed swollen or disintegrated neuronal cell bodies in the hippocampal CA1 region, deepened nuclear staining and expanded intercellular space in the model group. In comparison with the model group, the EA group (not the EA+DMOG group) displayed relatively intact nuclei of neurons, mitigated swelling of cell body and reduction in the infiltration of inflammatory cells and uniform distribution of cells in the CA1 region.

EA can significantly improve the cognitive function in AD model mice, which may be related to its function in inhibiting the hippocampal HIF-1α/PFKFB3 signaling pathway to reduce glycolytic enzyme activity and abnormal lactate accumulation.

To observe the role of tyrosine hydroxylase (TH)-positive neurons in the nucleus tractus solitarius (NTS) and locus coeruleus (LC) in electroacupuncture (EA)-mediated improvement of blood-brain barrier (BBB) damage in APP/PS1 mice, so as to explore the mechanism of the NTS^TH-LC neural circuit underlying the effect of EA on prevention and treatment of Alzheimer’s disease (AD).

(1) Eight 4-month-old male C57BL/6 mice served as the control group, and 16 age-matched male APP/PS1 mice were randomly divided into the model and EA groups (n=8). The EA group received EA stimulation at “Neiguan” (PC6) and “Jianshi” (PC5) once every other day for 4 weeks. After intervention, Morris water maze and novel object recognition tests were used to evaluate learning and memory abilities. Western blot was performed to detect the expression levels of hippocampal tight junction proteins, including Occludin, Claudin-5, and zonula occludens-1 (ZO-1). Immunofluorescence was used to assess the co-localization of TH/c-Fos in NTS and LC, as well as TH/norepinephrine (NE) co-localization in LC. (2) Five TH-cre mice received retrograde tracing virus injection into LC to observe whether TH-positive neurons in NTS project to LC. (3) Twenty-one 7-month-old APP/PS1 mice were injected with chemogenetic activation virus (AAV2/9-hSyn-DIO-hM3D-mCherry-WPRE-hGH-pA) or empty virus (AAV2/9-Ef1α-DIO-mCherry-WPRE-hGH-pA) into the NTS, and AAVretro-TH-CRE-WPRE-hGH-pA virus into the LC. After 21 d of virus expression, 6 mice injected with empty virus were taken as mCherry+CNO+EA group; 3 mice from the activation virus group were randomly selected for brain slice patch-clamp to verify virus functionality. The remaining mice were randomly divided into 2 groups (n=6): hM3D+CNO+EA, and hM3D+saline+EA groups. EA was applied to PC6 and PC5 once daily for 15 d. After intervention, learning and memory abilities were evaluated by Morris water maze and novel object recognition tests. BBB permeability was detected by Evans blue (EB) staining. Immunofluorescence was used to measure TH/c-Fos and TH/NE co-localization in the LC.

(1) Compared with the control group, the model group showed significantly impaired learning and memory abilities (P<0.01), decreased expressions of hippocampal Occludin, Claudin-5, and ZO-1 (P<0.01), and increased TH/c-Fos co-localization (c-Fos expression in TH-positive neurons) in NTS and LC, as well as TH/NE co-localization in LC (P<0.01). Compared with the model group, the EA group exhibited improved learning and memory abilities (P<0.05, P<0.01), increased expressions of hippocampal tight junction proteins (P<0.01), and reduced TH/c-Fos and TH/NE co-localization (P<0.01, P<0.05). (2) Retrograde tracing confirmed that TH-positive neurons in NTS project to LC. (3) Compared with the mCherry+CNO+EA group and hM3D+saline+EA group, the hM3D+CNO+EA group showed significantly impaired learning and memory (P<0.01), increased EB content in brain tissue, and elevated TH/c-Fos and TH/NE co-localization in the LC (P<0.01).

EA at PC6 and PC5 can ameliorate BBB damage and learning/memory deficits in AD mice, and its mechanism may be related to inhibiting the activation of the NTS^TH-LC neural circuit.