The results of PLS-DA in this study indicated metabolite differences among the sham, ST36, and PC6 groups in the cerebral cortices, hippocampi, and hypothalamus. This finding suggested that the signals from the sham, ST36, and PC6 points were transmitted through different pathways according to their specific location in the cerebral cortex, hippocampus, and hypothalamus (Fig. 3). In addition, our results also indicated that EA at ST36 increased adrenaline levels in the cerebral cortex and hippocampus, whereas EA at PC6 could not produce similar results. EA at PC6 could reduce the GABA and glycine levels of the cerebral cortex and hippocampus, whereas EA at ST36 could not produce similar results, suggesting that the EA at ST36 and PC6 could cause different changes in the neurotransmitters. Collectively, the results were consistent with the idea that different points modulate the activity of a corresponding brain area and have specificity1. GABA is a neurotransmitter that can rapidly inhibit synaptic transmission. The GABA system has been widely used as a target in the CNS for various drugs, such as anxiolytic, sedative–hypnotic, and anticonvulsant medications15,16. Glycine is a secondary rapid inhibitory neurotransmitter, and its receptor involves motor reflexes and nociceptive pathways, especially in the spinal cord15,16. GABA-A controls inhibitory synaptic neurotransmission, whereas glutamate controls excitatory synaptic neurotransmission. A reduction in the number of GABA-A receptors reduces synaptic inhibition, and an increase in the number of glutamatergic receptors causes an imbalance between excitation and inhibition, resulting in epileptic seizure17. GABA dysfunction and glycine disruption involve the mechanism of neuropathic pain18. The results of the present study indicated that EA at PC6 reduces GABA and glycine levels in the cerebral cortex and hippocampus (Fig. 2A, B). Further studies are required to determine whether the reduction of GABA and glycine by EA at PC6 results in epileptic seizure or pain thresholds to cause seizures or neuropathic pain to occur.

Figure 3


The summary of differential effects on the putative meridian paths between ST36 and PC6. The linkage of dash line implies the potential role of hypothalamus in coordinating modulation of metabolic change in cerebral cortex or hippocampus by referring the previous studies31,32.

Additionally, our present results also indicated that the glutamate levels in the hippocampus were also lower in the PC6 group than in the sham and ST36 groups (Fig. 2B). Glutamate is mainly an excitatory neurotransmitter in the mammalian CNS. Glutamatergic excitotoxicity plays a critical role in the pathogenesis following an ischemic stroke; additionally, it contributes to neuropsychiatric disorders such as schizophrenia and major depressive disorder19. The results imply that applying EA at PC6 may be feasible to treat ischemic stroke or neuropsychiatric diseases.

Our results also indicated that adenosine levels of the cerebral cortices in the ST36 and PC6 groups were lower than those in the sham group (Fig. 2A). Adenosine is a neurotransmitter that serves as an endogenous distress signal and plays a modulatory role in tissue damage and repair in the nervous system20,21. The formation of adenosine mainly results from the breakdown of intracellular or extracellular adenine nucleotides when cell-membrane damage occurs, such as when transient global ischemia causes a massive increase in the extracellular adenosine triphosphate (ATP), which results in the rapid formation of extracellular adenosine. By contrast, extracellular adenosine levels decrease with the increase in intracellular adenosine formation21,22. Acupuncture may induce increases in extracellular ATP, adenosine diphosphate, adenosine monophosphate, and adenosine levels; however, the increase in adenosine level only lasts for a short duration because of the facilitated uptake of nucleoside transporters21,23. Collectively, these findings indicate that the reduction of adenosine levels in the ST36 and PC6 groups may be attributable to the movement of adenosine from the extracellular to intracellular space.

Our results also indicted that the adrenaline levels of the cerebral cortices and hippocampi in the ST36 group were greater than those in the sham and PC6 groups (Fig. 2A, B). Adrenaline is one of the principal neurotransmitters in the CNS and is involved in many CNS functions. It is suggested that noradrenaline can control the neuronal activity of the cerebral cortex or amygdala in the consolidation process of long-term memory formation24. Adrenaline can enhance hippocampal glucose metabolism after stress to support stress-related memory processing25. Therefore, EA at ST36 is perhaps more suitable than EA at PC6 for improving memory.

From the results, arginine levels in the hippocampus were greater in the ST36 group than in the PC6 group, whereas the arginine levels in the cerebral cortex were not significantly different between the two groups (Fig. 2A, B). l-arginine involves nitric oxide (NO) synthase, which is related to the generation of endogenous NO. l-arginine induces vasodilation, and its long-term oral administration can alleviate the clinical symptoms of cardiovascular disorders26. NO levels are elevated in acupoints, inducing vasodilation and leading to an increase in local blood flow and also contributes to the specificity of acupoints27. The change of arginine level in these results partly support to explain of specificity of acupoints.

One of our results requires further explanation, namely why no significant difference was observed in the adenosine, arginine, GABA, and glycine levels of the hypothalamus among the three groups, except for higher adrenaline levels of the hypothalamus in the PC6 group than in the ST36 group (Fig. 3). The specificity of acupoints and the meridian in the hypothalamic paraventricular nucleus was reported in a study using acupuncture stimulation at different 33 points28. EA at PC6 or ST36 can elevate gonadotropin-releasing hormone (GnRH) mRNA levels in the hypothalamus of female rats29. GnRH acts as a primary brain signal in the hypothalamic–pituitary–ovarian axis response to the release of follicle-stimulating hormone, luteinizing hormone, and neurotransmitters and neuropeptides such as GABA and glutamate29,30. The functional connectivity between the amygdala and hypothalamus is observed in patients with carpal tunnel syndrome subjected to acupuncture stimuli31. The hypothalamus centrally communicates acupuncture stimulation because the 2 Hz EA at LI4 can induce hypothalamus activation, which in turn extends to the insula, anterior cingulate cortex, and other brain regions32. Although acupuncture stimulation reaches hypothalamus, only moderate change has been observed on these neurotransmitters in this study. Therefore, hypothalamus may mainly coordinate signals and transmit to the cerebral cortex and hippocampus. Considering all of these findings together, we suggest that an acupuncture signal transfers from the point to the hypothalamus, after which it transfers to the cerebral cortex and hippocampus, producing physiological action possibly playing a partially critical role in acupuncture stimulation, but further study is still needed..

In conclusion, our results indicated differences among the sham, PC6, and ST36 groups in the cerebral cortex, hippocampus, and hypothalamus. EA produces different changes at ST36 and PC6 in the aforementioned neurotransmitters in the cerebral cortex and hippocampus, suggesting point specificity. The hypothalamus is a characterized central expression of acupuncture stimulation; some signals from acupuncture stimulation extend to other brain areas, possibly through the hypothalamus, but further study is needed in the future.