In this study, the effect and possible mechanism of NBP on the improvement of cognitive function in mBCCAO rats were explored. The experimental results proved that NBP improved the performance of CCH rats in the Morris water maze, and its amelioration of cognitive function may be achieved by protecting the integrity of BBB.
The most classic method to build a CCH rat model is BCCAO, also known as 2-vessel occlusion (2-VO)[17], in which both sides of bilateral common carotid arteries of rats are permanently ligated at the same time, causing ischemic-hypoxic damage in vulnerable areas of brain tissue. However, the conventional 2-VO blocks blood flow too quickly and causes a high postoperative mortality rate, which could not reflect the true pathological characteristics of CCH state in clinical conditions[18]. And the occlusion of ophthalmic arteries impairs visual function in rats[19]. Therefore, in this study, a modified version of BCCAO was used, in which the unilateral common carotid artery was ligated first and the contralateral common carotid artery was ligated one week later. This staggered ligation of the common carotid artery reduces one-time injury and allows for a rapid recovery, mimicking a chronic state of cerebral hypoperfusion in rats.
We innovatively used PET technology to verify the successful construction of the model and found that the CBF changes in different brain regions were not the same in mBCCAO rats. There were obvious changes in CBF in the hippocampus, thalamus, and motor and sensory cortex regions, then the CBF partially recovered over time, which was consistent with the conclusions of other similar studies[20, 21]. The reason for this phenomenon is related to angiogenesis and vascular remodeling[22]. Under CCH state, the expression levels of vascular endothelial growth factor (VEGF) are different in the cortex and hippocampus, which may explain the differences in CBF in different brain regions[23].
The BBB, composed of pericytes together with vascular endothelial cells, specially differentiated basement membrane and astrocyte terminal foot, is an important component of the neurovascular unit (NVU) and an indispensable physical barrier that protects the central nervous system[24, 25]. In the state of CCH, reduced CBF disrupts the integrity of the BBB through the induction of excitotoxicity, inflammation, oxidative stress, and expression of matrix metalloproteinases[26]. And then, BBB injury induces a series of downstream events, such as astrocyte proliferation, microglia activation, oligodendrocyte apoptosis and white matter lesions, contributing to secondary brain injury[19]. In a word, endothelial dysfunction and BBB injury play a key role in the pathology of progression from CCH to VCI. Clinical studies have also identified BBB leakage as an important pathological change in VCI patients using magnetic resonance imaging[27–30]. Therefore, the early control of BBB injury is important for subsequent white matter and neuronal protection.
DL-3-n-butylphthalide has been validated for cognitive function improvement in many animal models of CCH[31–38], but most of the existing studies focus on its direct protective effects on neurons[34, 36–38], as well as its mechanisms in promoting angiogenesis[31, 38], inhibiting neuroinflammation[32, 33], and reducing reactive astrocyte proliferation[33, 38], while the number of studies about BBB is very limited. In this experiment, we paid more attention to the changes in BBB permeability after CCH, and used EB leakage to reflect the degree of BBB destruction.
TJs are macromolecular complexes composed of transmembrane proteins, cytoplasmic adhesion proteins and cytoskeletal proteins, which play an essential part in maintaining the integrity of BBB. TJ proteins mainly include two broad categories: cytoplasmic zonula occludens (ZO), and three transmembrane proteins - claudins, occludins, and junctional adhesion molecules (JAMs)[39]. Claudin-5 is the most abundant member of the claudin family in BBB and governs selective permeation of the TJ by controlling paracellular permeation of small molecules[40]. ZO-1 belongs to the membrane-associated guanylate kinase-like proteins family, interacts with occludin and claudins, anchoring TJ to the cytoskeletal scaffold and actin of endothelial cells[41], and is sensitive to the BBB damage in pathological conditions, which can be used as a good indicator of the structure and function of BBB. In other CCH models, researchers have also observed a decrease in the expression levels of ZO-1 and Claudin-5[42, 43]. In clinical practice, ZO-1 and Claudin-5 are also often used as markers for BBB alteration, the levels of which are positively associated with disease prognosis[44, 45].
In this study, we found that NBP improved EB leakage and reduced the loss of ZO-1 and Claudin-5 in the mBCCAO rats with some dose-dependent effects. The high-dose NBP was effective, while the low-dose group did not reflect significant improvements in cognitive function and blood-brain barrier. The clinical dose for oral administration of butylphthalide is 0.2g, tid, which is higher than 40mg/kg·d and lower than 80mg/kg·d based on the conversion relationship between oral administration in mice and humans, which may explain the reason for the ineffectiveness of low-dose NBP. It is worth to note that the NBP has side effects, long-term administration of which can result in mild elevation of transaminases and mild gastrointestinal symptoms. However, previous randomized double-blind controlled trials have confirmed that NBP-related adverse events are uncommon[12].
During embryonic development, pericytes are recruited by endothelial cells to surround neovascularization and are involved in the regulation of angiogenesis, vascular maturation and vascular permeability[46, 47]. The pericyte coverage ratio determines the relative permeability of vasculature, and it has been shown that there is a significant decrease in pericyte coverage in corpus callosum in the BCCAO rat model[16]. However, in our study, we found no difference in cortex and corpus callosum(data unshown) pericyte coverage between control and model group. We guess, the underlying reason could be that the ischemic impairment in mBCCAO was more slight than in BCCAO, and the manual counting error could also play a part.
There are some limitations of this study. First, our experiment only explored the protective effect of NBP on BBB under CCH state at several individual time points. It is meaningful to extend the experimental time and set more time points to observe the efficacy of NBP in the chronic process. Second, further investigations are needed to find the possible molecular mechanisms by which CCH affects TJ protein expression.