Moyamoya disease (MMD) is a rare, chronic, and progressive cerebrovascular disorder that is characterized by stenosis and occlusion of the distal carotid, proximal, middle, and anterior cerebral arteries and is accompanied by the development of a fine collateral vascular network at the base of the brain (moyamoya vessels) (1, 2). Besides, patients with MMD can develop spontaneous collaterals (innate collateral vessels) in areas of ischemia. These alternative conduits vary from individual to individual, resulting in different clinical presentations, the severity of disease, and outcomes. Successful compensatory collateralization is considered a defensive measure against ischemic and hemorrhagic stroke in MMD (3, 4). These spontaneous collaterals are of such importance that, to some extent, they determine, or rather indicate, the fate of patients with MMD. Although many articles have reported relevant factors that affect the formation of these collateral vessels in MMD, the mechanisms involved in spontaneous vessels formation are not fully understood.

Indirect revascularization is an essential component of the surgical treatment of MMD. This procedure improves cerebral perfusion by attaching pedicled, vascularized grafts to the cortical surface, and facilitating the in-growth of angiogenesis. Compared to direct revascularization, indirect revascularization is a much simpler, less strenuous technique with lower risks of post-operative complications, and as demonstrated in many recent studies, may achieve long-term outcomes as satisfying as those for a direct bypass surgery (5–9). However, the mechanisms involved in synangiosis induced by indirect revascularization are also not well-understood. The uncertainty of post-surgical synangiosis is the main consideration that limits the use of this method. According to earlier reports, collateral vessel formation in encephaloduroarteriosynangiosis (EDAS), a commonly used indirect revascularization surgery, may be primarily driven by angiogenesis (10, 11). Angiogenesis is induced by hypoxia and involves the formation of new collateral vessels by sprouting or splitting from the pre-existing vascular structures (12, 13). Studies have shown that a subtype of immature circulating cells is involved in endothelial repair, neo-angiogenesis, and tissue homeostasis (14, 15). The surface markers used to identify such endothelial progenitor cells (EPCs) have not been consistent across studies, creating confusion (16). However, phenotypical and functional tests show that CD34^briCD133⁺KDR⁺ cells are a population of endothelial precursor cells (17). Based on previous studies (18, 19), here we used the markers CD34^briCD133⁺CD45^dimKDR⁺ to define putative EPCs, although this definition is not consensual.

Genetics has contributed greatly to understanding the pathophysiology of MMD. A polymorphism, R4810K (p.Arg4810Lys), in the Ring Finger Protein 213 (RNF213) gene was identified as the strongest genetic susceptibility factor for MMD in East Asian populations, which could increase the risk of MMD by ≈300 times (20, 21). Moreover, patients with a homozygous RNF213 mutation are predisposed to significantly early-onset age, rapid disease progression, cerebral infarction at diagnosis, and poor prognosis (22–24). Thus, RNF213 is a major susceptibility gene for MMD, and it has been recognized as a key molecule to understand the pathophysiology of MMD. In vitro and in vivo experiments revealed that RNF213 is related to angiogenesis and vascular inflammation; however, the exact physiologic functions of RNF213 remain unknown (25).

Therefore, we designed an angiography-based study using patients from a prospective trial of EDAS surgery to study the locus (c.14429G>A) p.R4810K of the RNF213 and EPCs in patients with MMD, in order to understand the genetic basis and molecular mechanism of these two types of collateral vessels in MMD (native arteriogenesis and therapeutic synangiosis).

In our study, we found that the heterozygous mutations of RNF213 p.R4810K were significantly correlated with spontaneous collateral circulation. In addition, we found that the CD34^briCD133⁺CD45^dimKDR⁺ cell count in patient with post-EDAS was significantly higher as compared to the count before EDAS in the good Matsushima grade group. This implied that the mutations of RNF213 p.R4810K affect the establishment of spontaneous collateral circulation, while CD34^briCD133⁺CD45^dimKDR⁺ cells are involved in the process of newly formed EDAS collaterals. To the best of our knowledge, this is the first report to explore the mechanism of these two types of collateral vessels in MMD.

Recently, several studies have indicated that in patients the homozygous c.14429G>A (p.R4810K) variant, rather than the heterozygous variant, was significantly associated with a young age at disease onset, cerebral infarction at presentation, and poor cognitive outcomes (22, 23, 34–36). However, we found that patients carrying heterozygous mutations of RNF213 p.R4810K had a more severe pattern, such as infarction on presentation, more severe hemodynamic impairment, worse neurological status, and good post-operative Matsushima grade in our study (Table 4). One explanation could be that our cases were collected from a group with homogeneous age (30–35 years), while the homozygous p.R4810K variant of RNF213 always predicts early onset of MMD in patients (23, 36). Patients with a young age at disease onset themselves have a more severe clinical course of MMD (37, 38). Another explanation could be that the patients with heterozygous mutations had a higher rate of PCA involvement (50%) in our study, which was similar to that of a previous Chinese study (48.7%) (39).

In fact, severe MMD with an unfavorable prognosis has been attributed to insufficient compensatory collateralization, which plays a pivotal role in the pathophysiology of cerebral ischemia. The collateral circulation, especially the leptomeningeal system, plays the most important role in the collateral supply of the ischemic cortex in the ACA and MCA territory in patients with MMD (40). Recently, we proposed a new collateral grading system to capture the entire scope of spontaneous collateral circulation, aiming to assess the intracranial blood supply (31). We found that in patients with good leptomeningeal collateral that there was a prolonged TTP as compared to normal but that in those with poor LM collateral the TTP was further prolonged. Further, the heterozygous mutations were significantly correlated with poor collateral circulation, which may be the cause of the severe symptoms and worse neurological status. These findings suggest that not only homozygous but heterozygous c.14429G>A (p.R4810K) may be useful as a specific biomarker for severe MMD with an unfavorable prognosis that requires timely intervention for revascularization surgery.

Indirect cerebral revascularization through EDAS was demonstrated to establish new collateral vessels through the formation of vessels from the external carotid artery (ECA) to the internal carotid artery (ICA) in patients with MMD (5–9). Although this phenomenon has been extensively described in the literature, the mechanism involved in the formation of collateral vessels is poorly understood. In our study, good Matsushima grade was occurred significantly more frequently in those with more severe ischemia, such as lower spontaneous collateral grade and longer TTP delay, suggesting that ischemia and hypoxia are the main driving forces of the newly formed EDAS collaterals. In our study, we also found that the CD34^briCD133⁺CD45^dimKDR⁺ cell count in patient post-EDAS was significantly higher as compared to the count prior to EDAS in the good Matsushima grade group. However, this change was not observed in the poor Matsushima grade group. The results supported the theory that ischemia and hypoxia can also increase the EPC count in the peripheral blood of patients, leading to angiogenesis (41). Hence, we infer that neovascularization induced by EDAS might involve the recruitment, proliferation, and assembly of CD34^briCD133⁺CD45^dimKDR⁺ cells and their ability to promote angiogenesis.

Spontaneous collateral circulation, or native collaterals, occurs through a process known as arteriogenesis, which is flow-dependent and hypoxia-independent. In arteriogenesis, pre-existing vessels with stenotic segments are exposed to increased shear stress due to a high-pressure gradient between two vascular territories (42). Arteriogenesis involves the recruitment and enlargement of pre-existing vessels, and sprouting and splitting are not parts of this process. A previous study showed that when a collateral develops into a vessel distal to a stenosis, higher flow and increased shear stress are associated with the enlargement of the collateral vessel (11). We found no significant difference in the CD34^briCD133⁺CD45^dimKDR⁺ cell count among the three spontaneous collateral circulation groups, although the heterozygous c.14429G>A (p.R4810K) of RNF213 was significantly correlated with spontaneous collateral circulation. Knockdown of RNF213 in zebrafish leads to abnormal sprouting and irregular diameter of intracranial vessels, suggesting its possible contribution to vascular formation (21). These findings suggest that mutations of RNF213 p.R4810K may inhibit spontaneous collateral circulation. However, the pathophysiological role of RNF213 and the mechanisms by which RNF213 polymorphisms lead to MMD have not been elucidated. Further research on its contribution to the pathogenesis of MMD is necessary.

The seminal findings of Asahara et al. describe the isolation and characterization of putative EPCs (14). EPCs play an essential role in the integrity and functionality of neo-vasculature under specific programmatic pathways that include angiogenesis and are especially relevant for tissue engineering strategies (43, 44). Since its initial discovery, the phenotypic definition of EPCs has undergone a gradual transformation to include a wide number of phenotypic cell surface markers (45, 46). The ongoing stratification and sub-stratification of cell surface markers have made inter-laboratory findings regarding tissue neo-vascularization difficult to compare at both the qualitative and quantitative levels. More recently, experts in the field of lymphatics and vascular biology offer a perspective article attempting to initiate consensus terminology for EPCs along with their phenotypic expression (16). However, numerous peer-reviewed studies that span decades have utilized different cell surface markers to demonstrate these EPCs to incorporate into neo-vasculature (18, 47–50). Within this context, we have utilized a population of cells defined as CD34^briCD133⁺CD45^dimKDR⁺ (18, 19). Our study suggested that the CD34^briCD133⁺CD45^dimKDR⁺ cells might play a significant role in the formation of collaterals post-surgery, which is a potential mechanism behind therapeutic collaterals in MMD.

This study had certain limitations that should be noted. First, it remains unknown whether CD34^briCD133⁺CD45^dimKDR⁺ cells are myeloid angiogenic cells (MACs) as defined by Medina et al. (16), since we characterized them based on a different set of markers. In addition, in vitro or in vivo testing was not performed to assess whether the CD34^briCD133⁺CD45^dimKDR⁺ cells exhibit endothelial properties and could indeed contribute to angiogenesis. The identification of these cells and their role in the formation of collateral circulation in MMD are worthy of further study.

Secondly, we only analyzed the relationship between RNF213 p.R4810K and the two types of collateral vessel formation. However, MMD is a complex disease with genetic heterogeneity and a classic pattern of inheritance, and therefore the single gene-focused study cannot sufficiently elucidate its genetic susceptibility. Therefore, future studies should include other mutations to comprehensively determine the genetic basis of this pathology. Thirdly, this study was based on a single-center hospital, so selection bias was unavoidable. Fourthly, we only enrolled patients who were aged 30–35 years, which may also have contributed to an unavoidable selection bias. Finally, the value of cerebral hemodynamics determined by perfusion MRI was relative and not absolute, and further technical refinements are required to enhance its diagnostic value.

Our study implied that the mutations of RNF213 p.R4810K affect the establishment of spontaneous collateral circulation, while CD34^briCD133⁺CD45^dimKDR⁺ cells are involved in the process of newly formed EDAS collaterals. Heterozygous c.14429G>A (p.R4810K) may be useful as a specific biomarker for severe MMD with an unfavorable prognosis that requires timely intervention for revascularization surgery.

The datasets presented in this article are not readily available due to ethical and privacy restrictions. Requests to access the datasets should be directed to the corresponding author.

The studies involving human participants were reviewed and approved by the Research Ethics Board at the Fifth Medical Center of Chinese PLA General Hospital. The patients/participants provided their written informed consent to participate in this study.

X-YB: drafting the article. Y-NF: acquisition of data and analysis and interpretation of data. Q-NW: acquisition of data and statistical analysis. X-PW, R-MY, and Z-XZ: acquisition of data. QZ and D-SL: technical support. X-GY and LD: conception and design and critically revising the article. All authors contributed to the article and approved the submitted version.

This study was supported by a grant from the National Natural Science Foundation of China (Grant No. 82171280) and the Hygiene and Health Development Scientific Research Fostering Plan of Haidian District Beijing (HP2021-04-80202).

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.