Vascular endothelial cells in the brain are closely linked to each other through various junction proteins, and interact with pericytes and astrocytes to form a special blood-brain barrier (BBB) barrier system. BBB strictly limits the entry of neurotoxic substances, inflammatory factors and immune cells in the blood into central nervous system (CNS), and excludes the metabolites and neurotoxic substances in CNS from the brain. However, BBB also prevents certain drugs and large-molecule therapeutics from entering the brain (Banks, 2016). Clearly, higher levels of therapeutic drugs in the brain will be beneficial, reducing dose requirements and providing greater therapeutic indices. Reflecting this demand, many brain delivery technologies have been developed and tested in laboratory and clinical studies, comprising invasive technologies such as direct brain injection, intrathecal brain deliver, intracerebral grafts and deep brain stimulation and non-invasive technologies including nanoparticulate systems, focused ultrasound, biological mechanisms and intranasal brain delivery (Terstappen et al., 2021). A number of factors contribute to the physical barrier of BBB and the major are endothelial transport systems and cellular junction molecules.

Our previous study found that 2/100 Hz Electroacupuncture (EA) stimulation increased BBB permeability, reduced ZO-1 and occludin levels, and induced ultrastructural changes in TJ morphology (Zhang et al., 2020), but the molecular and signaling mechanisms of BBB opening mediated by EA are still lacking. Therefore, in this manuscript, transcriptome sequencing combined with bioinformatics analysis was used to explore the effect of specific stimulation mode electroacupuncture on the transcriptome of rat cerebral cortex, focusing on the molecules and signaling pathways related to the blood-brain barrier, in order to reveal the potential mechanism of electroacupuncture opening the blood-brain barrier, and provide potential targets and new research ideas for mediating central nervous system therapeutic drugs into the brain.

In the development of drugs for the treatment of CNS diseases, achieving adequate BBB transmission is a key challenge. The BBB is formed by brain microvascular endothelial cells that are sealed by tight junctions, which makes it an important obstacle to most brain therapies. Therefore, poor BBB permeability is the main reason for limiting the application of newly developed CNS therapeutic drugs in clinical practice.

Multiple ABC proteins are expressed in the BBB lumen and blood-oriented inner cortical membrane, which limits the permeability of a large number of toxins, including therapeutic agents (Miller, 2015). ABC transporters are ATP-driven efflux pumps of exogenous and endogenous metabolites, and their high expression in BBB leads to drug resistance in CNS. The decreased expression and/or functional activity of ABC-BBB transporters in Alzheimer’s disease (AD) and Parkinson’s disease (PD) patients (Zlokovic, 2011), and in the brain of AD animal models proved that the decreased expression and/or functional activity of ABC-BBB transporters led to the accumulation of amyloid beta peptide (Aβ) (Cirrito et al., 2005). However, the clinical potential of ABC transporters for disease management and drug delivery improvement remains unclear. EA can reduce the expression of Abcc6 (Table 3), suggesting that further studies can explore whether electroacupuncture can mediate the clearance of harmful sediments such as Aβ and the delivery of CNS therapeutic drugs by affecting the expression of Abcc6. Furthermore, EA intervention can change the expression of multiple SLC, including Slc7a1, Slc9a3r2, Slc38a3, Slco2b1, and Slc25a25 (Table 3). The non-invasive delivery of therapeutic drugs through BBB may use endogenous processes, such as adsorption-mediated cell transport, CMT and RMT (Terstappen et al., 2021). Among them, the CMT system is expressed by genes in the SLC transporter gene family, including more than 300 transporter genes encoding membrane-binding proteins, which are helpful for the trans-membrane transport of multiple substrates (Lin et al., 2015). In BBB, SLC proteins promote cross-cellular transport of multiple molecules, including carbohydrates, amino acids, monocarboxylic acids, hormones, fatty acids, nucleotides, organic anions, amines, choline and vitamins (Zlokovic, 2008; Daneman and Prat, 2015; Pardridge, 2015). It is worth noting that Cavin2 is also a gene related to transcytosis, which can be significantly down regulated by electroacupuncture; Other genes related to endocytosis include Cttn, Cdc45, Hspa12b, Cdcc117, Cdc42ep3, Cblb and Hspa1b (Table 3). The role of these genes in maintaining the integrity of BBB should be further studied.

Basement membrane (BM) is one of the important components of BBB because endothelial cells are connected to the BM through integrins. In addition, integrins also interact with extracellular matrix proteins such as laminin, collagen and perlecan, and mediate signal transduction by activating ECM ligands, growth factors and growth factor receptors (Baeten and Akassoglou, 2011). Our GO analysis showed that the mechanism of EA intervention involved integrin binding (GO:0005178, P = 0.0426) and collagen-containing extracellular matrix (GO:0062023, P = 0.0071) (Table 4). Mice lacking β1- integrin in endothelial cells produce abnormal cadherin signaling, claudin-5 deficiency and immature BBB (Yamamoto et al., 2015). Similarly, mice lacking laminin secreted by astrocytes showed BBB breakdown (Yao et al., 2014).

It is well known that the TJ in the BBB is a dynamic structure. TJ proteins undergo changes in expression, subcellular localization, post-translational modifications and protein-protein interactions under physiological and pathophysiological conditions (Huber et al., 2001). Interactions between TJ proteins and intracellular signaling pathways have begun to be elucidated (Bazzoni and Dejana, 2004). The cultured epithelial cells in the medium lacking Ca^{2 +} showed the loss of cell membrane ZO-1, ZO-2 and thromboxane, and the increase of related paracellular permeability (Klingler et al., 2000). Interestingly, although artificially increasing intracellular [Ca^{2 +}] does not change BBB TJ (Brown et al., 2004), the treatment with Ca^{2 +} channel blocker SKF 96365 can improve the permeability increase effect of hypoxia/hyperglycemia on BBB through a mechanism that seems to involve the recruitment of occludin into the cell membrane (Brown and Davis, 2005). The results of KEGG analysis showed that the most significant enrichment pathway was the calcium signaling pathway, and the differential genes associated with this pathway are Plcd3, Ryr2, Itpr3, Htr2a, Grin2c, Sphk2 (Figure 6 and Table 5). Itpr3 is an inositol 1,4,5-triphosphate receptor, type 3 (InsP3R3), encoding a receptor for inositol 1,4,5-trisphosphate, a second messenger that mediates the release of intracellular calcium. Although EA intervention did not significantly change the mRNA level of TJ proteins in this experiment, the previous study of our research group found that EA intervention could affect the expression level of ZO-1 and occluding (Zhang et al., 2020). Therefore, it is not excluded that EA regulates the subcellular localization and post-translational modification of TJ proteins through the calcium signaling pathway. The next step can study the effect of EA intervention on the protein expression level of TJ proteins and the possible role of calcium signaling in it.

Autophagy persists at the basic level of cells and is the mechanism of organelle or protein degradation and recycling (Lakhan et al., 2013). It is stimulated by a variety of pathological processes and participates in the elimination of organelle and protein decomposition, thereby avoiding excessive damage and cell dysfunction in various organs and cells (De Meyer et al., 2015). In vivo experiments showed that rapamycin (an autophagy inducer) led to the decrease of ZO-1 and the increase of BBB permeability, and confirmed the role of autophagy in regulating paracellular permeability (Zhang et al., 2018). However, studies have reported that CLDN5 (claudin 5) is abnormally accumulated in the cytoplasm of BMECs in stroke patients, accompanied by autophagy activation. Studies on zebrafish in vivo and in vitro cells showed that the BBB decomposition was caused by the redistribution of CLDN5 mediated by CAV1 (caveolin 1) into the cytoplasm under hypoxia. At the same time, autophagy is activated, which is mainly helpful degradation of CAV1 and CLDN5 in BMECs cytoplasm. These data suggest that autophagy plays a role in BBB integrity by regulating the redistribution of CLDN5 (Yang et al., 2021). Also, autophagy involves the dynamic disposition of intracellular and extracellular components, theoretically affecting BBB penetration. For example, autophagy-mediated lysosomal degradation pathway is involved in the degradation of PPR/p TRAIL in brain capillary endothelial cells and prevents them from penetrating BBB. The pre-inhibition of BBB autophagy by wortmannin loaded liposomes (Wtmn-Lip) can increase the accumulation of non-viral gene vector PPR in the brain without damaging the BBB tight junction (Wang et al., 2020). In the study of APP/PS1 mice, LSP (Lychee seed polyphenol) induced autophagy in bEnd.3 cells through AMPK/mTOR/ULK1 pathway and increased TJs expression (Xiong et al., 2021). The above data show that autophagy can regulate BBB permeability and material transport, but there is still no direct evidence to explain the exact relationship between the two, which may be related to diseases and the types of models used in the study. Our GSEA results showed that EA at specific stimulation mode could activate autophagy (Figure 7) and inhibit TORC2 signal (GO: 0038203, P = 0.0311), and the genes involved were significantly down-regulated Prr5l (Table 4). Therefore, we believe that autophagy is closely related to the translocation and degradation of TJ proteins in BBB, which is also an idea to study the mechanism of EA opening BBB. However, the protein expression and mRNA expression are not completely consistent, because the sample of this experiment is the total RNA of rat cortical tissue, not a single cell, which is also one of the limitations of this experiment. Future studies should first confirm whether the autophagy pathway activated by EA is involved in the translocation and degradation of BBB TJ protein. Secondly, whether the degree of autophagy activation is enough to cause cell damage. Finally, is the autophagy activation after EA a direct effect or is it because EA changes the cellular distribution of TJ proteins (such as translocation from cell membrane to cytoplasm), thus promoting autophagy activation.

In addition, EA may have an effect on the clathrin-coated pit (GO:0005905, P = 0.0424), involving genes including Cttn and Sh3bp4 (Table 4). Clathrin-coated pit is a part of the endomembrane system in the form of an invagination of a membrane upon which a clathrin coat forms, and that can be converted by vesicle budding into a clathrin-coated vesicle. Coated pits form on the plasma membrane, where they are involved in receptor-mediated selective transport of many proteins and other macromolecules across the cell membrane. Experimental studies on various animal models have shown that Aβ is mainly cleared through the BBB across blood vessels (70–85%), while a small part is cleared through the ISF flow (Bading et al., 2002; Deane et al., 2004; Tarasoff-Conway et al., 2015) and the molecular mechanism of Aβ clearance through BBB has been clarified in more detail recently. In short, Aβ produced in the brain binds to LRP1 on the albumin side of BBB, resulting in its rapid internalization into endothelial cells and clearance through blood. Phosphatidylinositol-binding clathrin assembly protein (PICALM) is crucial for the internalization of LRP1 – Aβ complex mediated by endothelial clathrin/PICALM, and guides the transport of endocytosis vesicles containing Aβ in endothelial cells by sequential fusion with Rab5 positive early endosomes and Rab11 positive sorting endosomes used for the endocytosis of BBB luminal (Zhao et al., 2015).

However, as we mentioned above, our sample is a mixture of all cells in the cortex, which can not directly prove that these genes and proteins are expressed in which kind of cells of the blood-brain barrier. Our research mainly focuses on the role of endothelial cells, and the blood-brain barrier is not a single entity. It is composed of astrocytes, pericytes and endothelial cells, which allow the endothelial cells of the central nervous system to strictly regulate the movement of ions, molecules and cells between blood and neural tissues. Astrocytes are the most abundant cells in the brain, which are very important in regulating brain and BBB function. A recent breakthrough study (Kitchen et al., 2020) proved that targeting astrocytes at the blood-brain barrier (BBB) and spinal cord barrier (BSCB) is a feasible option for the treatment of TBI and spinal cord injury. This study showed that pharmacological inhibition of these signal events could prevent the development of central nervous system edema and promote the functional recovery of injured rats, which was confirmed by Sylvain et al. (2021). Recently, the molecular mechanism of clearing Aβ through BBB has been elucidated in more detail. Lymphatic pathway is a waste removal system, which plays an important role in dementia. It uses the unique system of perivascular channels to promote the effective removal of soluble proteins and metabolites from the central nervous system. AQP4 is highly expressed in astrocytes. The enrichment of perivascular AQP4 at BSCB/BBB shows that it plays a role in lymphatic function. The development of new drugs for this system will have great therapeutic potential (Salman et al., 2021). In addition, pericytes also play an important role in the maintenance of BBB function. In addition to communicating with astrocytes, pericytes also support the maintenance of BBB in postnatal brain. Pericyte deficient mutant mice showed increased permeability of BBB to low molecular weight and high molecular weight tracers, and pericytes can directly regulate the expression of transporters (Arvanitis et al., 2020).

Achieving adequate BBB transmission is a key challenge in the development of drugs for the treatment of central nervous system (CNS) diseases. However, most traditional BBB opening strategies are difficult to be applied in clinic because of their extensive and non-specific regulation of BBB, resulting in damage to normal brain tissue. Therefore, non-invasive methods that selectively and effectively cross the BBB, such as receptor-mediated cell transport and the use of neurophilic viruses, nanoparticles and exosomes, have become popular in recent years (Terstappen et al., 2021). For example, in the process of glioma treatment, the poor permeability of BBB is the main reason that restricts the application of newly developed treatment methods for glioma in clinic. Researchers have proposed a variety of strategies to target and effectively regulate the BBB of tumor, so as to obtain more effective glioma treatment (Luo and Shusta, 2020).

Our research will provide valuable resources for discovering and studying new targets and signaling pathways that mediate crosstalk between different cell types in NVU. This may lead to the development of new transgenic animal models – BBB, NVU, and pluripotent stem cell models of different neurological diseases, as a valuable way to discover and test new drug delivery methods.

The original contributions presented in the study are publicly available. This data can be found here: https://www.ncbi.nlm.nih.gov/geo/, GSE192885.

This study was approved by the Animal Protection and Use Committee of Zhejiang University of Traditional Chinese Medicine.

CM was responsible for drafting and writing the manuscript. LG and HW analyzed and sorted out the data. LR and YL provided help for the experiment. SZ and PG contributed to make important modifications to the manuscript. XL approved the final manuscript to be published. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.