OBJECTIVETo observe the effects of different frequencies of electroacupuncture (EA) at "Baihui (GV 20)" and "Shenshu(BL 23)" for the learning and memory ability as well as glycogen synthase kinase-3β (GSK-3β) and growth associated protein-43 (GAP-43) in hippocampal tissue of rats with Alzheimer's disease(AD), so as to explore the mechanism of different frequencies of EA for the prevention and treatment of AD.

METHODSOne hundred and twelve healthy Wistar male rats were divided into seven groups by random number table, namely a normal group, a sham operation group, a model group, an acupuncture group, a 2 Hz EA group, a 30 Hz EA group, and a 50 Hz EA group, 16 rats in each one. The rats in the normal group were conventionally raised in the laboratory without any treatment. 0.9% NaCl solution was injected into bilateral dentate convolution of hippocampus in rats of the sham operation group. AD model was established by β-amyloid protein1-42 (Aβ1-42) injected into bilateral dentate convolution of hippocampus in the other groups. 15 days after establishment, no treatment was applied in the model and sham operation groups, and EA with corresponding frequencies at "Baihui (GV 20)" and "Shenshu (BL 23)" was used in the three EA groups for 2 sessions, once a day and 7 times as one session. There was 1 day between the two sessions. The same acupoints were adopted in the acupuncture group, without electrical connection. The escape latency, the first spanning platform time, and the number of crossing platform were tested in the Morris water maze immediately after treatment. The expressions of GSK-3β and GAP-43 were examined by immunohistochemistry and Western blot.

RESULTS①Morris water maze tests showed that the escape latency and the first spanning platform time significantly increased in the model group compared with those in the normal group (both<0.01), and the number of crossing platform decreased (<0.01). Compared with the model group, the escape latency and the first spanning platform times decreased in the acupuncture and three EA groups (all<0.01), and the numbers of crossing platform increased (<0.01). Compared with the acupuncture and 2 Hz, 30 Hz EA groups, the escape latency decreased in the 50 Hz EA group (<0.01,<0.05); the first spanning platform time reduced (all<0.01); the number of crossing platform increased (<0.01,<0.05). ②The expressions of GSK-3β and GAP-43 of the model group increased compared with those of the normal group(both<0.01). The expressions of GSK-3β in the acupuncture and three EA groups decreased compared with that in the model group (all<0.01), and the expressions of GAP-43 increased (all<0.01). The expressions of GSK-3β decreased and GAP-43 increases in the 50 Hz EA group compared with those in the acupuncture group and 2 Hz, 30 Hz groups (all<0.01).

CONCLUSIONSEA may promote synaptic damage rehabilitation by down regulating GSK-3β and up regulating GAP-43 to improve learning and memory ability of AD rats. The effect of 50 Hz EA is better than those of 30 Hz and 2 Hz EA and acupuncture.