Cerebral small-vessel disease (CSVD) is the clinical, radiological, and pathological syndrome caused by lesions involving cerebral microvessels (Pantoni, 2010), causing various lesions such as white matter hyperintensities (WMH) usually identified by neuroimaging with MRI (Wardlaw et al., 2013; van Veluw et al., 2017). As one of the most important causes of vascular cognitive impairment (VCI) (Inzitari et al., 2009; Gorelick et al., 2011; Hachinski, 2015), CSVD affects almost everyone older than 90 years. Because of its high age correlation, CSVD is related to pathological changes similar to those in neurodegenerative diseases, such as amyloid β-protein (Aβ) deposition (Sweeney et al., 2018). The pathology of vascular Aβ in cerebral amyloid angiopathy (CAA, a subtype of CSVD) coexists with the pathology of Alzheimer disease (AD).

The imbalance between Aβ production and clearance is an important pathological mechanism of Aβ deposition in AD (Hardy and Selkoe, 2002; Selkoe, 2003). In the process of clearance, there has been evidence that soluble metabolites in the brain can be eliminated along the perivascular space (PVS) (Iliff et al., 2013; Abbott et al., 2018). Patients with CSVD often have enlarged perivascular space (EPVS), which was considered likely reflected perivascular clearance disorders. This perivascular pathway that clears interstitial solutes was termed the “glymphatic pathway” for its functional similarity to the peripheral lymphatic system (Iliff et al., 2012; Mestre et al., 2017). It has been found that EPVS is significantly associated with vascular Aβ deposition (Perosa et al., 2022). Moreover, the glymphatic system has been linked to plaque formation in AD (Iliff et al., 2012; Arbel-Ornath et al., 2013; Xie et al., 2013). The arterial pulsation is the main driver of the glymphatic flow, and arteriosclerosis may impede this process. Mortensen et al. (2019) reported that glymphatic clearance was inhibited in stroke-prone spontaneously hypertensive rats (SHRSP), suggesting a new link between vascular pathology and AD. Previous studies also showed that reduced cerebral blood flow increased the deposition of Aβ in patients with AD, which might be caused by glymphatic system disorder that due to the abnormal blood flow (Hama et al., 2015; Saito and Ihara, 2016). The glymphatic system adjacent to the blood vessels is surrounded by the end feet of astrocytes, which are rich in aquaporin 4 (AQP4) (Nedergaard and Goldman, 2016). AQP4 plays an important role in maintaining the function of the glymphatic system. Loss of AQP4 in mice impaired glymphatic clearance of soluble macromolecules (Iliff et al., 2012; Xu et al., 2015). In addition, the improvement in AQP4 polarity promoted the clearance of Aβ by the glymphatic system (He et al., 2017). These studies suggested that astrocyte AQP4 was involved in the process of Aβ clearance, but the existence of related pathological changes in CSVD has not been supported by evidences.

On the other hand, as an important cause of Aβ formation, neuroinflammation has been widely thought to be closely related to chronic degenerative diseases, such as AD and Parkinson disease (Kinney et al., 2018). Also, microglia, in particular, has been proved to trigger the inflammatory cascades in the brain (Gogoleva et al., 2019). Activated microglia has two phenotypes: proinflammatory MI type and anti-inflammatory M2 type. Aβ can induce neuronal oxidative stress, activate microglia to convert into MI type, and trigger neuroinflammation and neuronal apoptosis. At the same time, inflammation may lead to more Aβ deposition, which leads to the transformation of acute injury into chronic injury (Garcia-Alloza et al., 2013; Saito and Saido, 2018). In the chronic cerebral hypoperfusion model, the activation and proliferation of glial cells has been observed to be an important pathological process leading to cognitive impairment (Barrientos et al., 2015; Wang et al., 2019), as well in RHRSP rats (Cai et al., 2020). Therefore, regulating the transformation of microglia into M2 type, an anti-inflammatory phenotype, is a new target for disease treatment. Studies on macrophages revealed that signal transduction and activator of transcription (STAT) family played an important role in microglia or macrophage phenotypic transformation (Sica and Bronte, 2007). STAT1 promoted the transformation of IFN-γ-induced M1 phenotype (Qin et al., 2012), and STAT6 was found to be involved in transforming microglia into beneficial phenotype and functional recovery in MCAO mice (Liu et al., 2016; Cai et al., 2019). Therefore, STAT6 might play an essential role in regulating neuroinflammation induced by cerebral ischemia. However, whether STAT6 is involved in the phenotypic regulation of microglia in CSVD brain lesions remains unclear.

Mesenchymal stem cells are multipotent cells mainly reflected in neuroprotection (Sasaki et al., 2009), neovascularization (Onda et al., 2008), anti-neuro inflammation, and neuroplasticity (Suzuki et al., 2013), supporting the use as a potential therapy to promote functional recovery after stroke and chronic degenerative diseases. Wu et al. found that MSCs could alleviate AQP4-dependent glymphatic clearance in mice with Huntington disease (Wu et al., 2020). MSCs were also found to promote the M2-type transformation of microglia in vitro and the expression of anti-inflammatory factors, but the mechanism remains unclear (Gao et al., 2014). Currently, studies on the use of MSCs in CSVD-induced VCI are few. Whether MSCs can improve the pathological changes related to chronic degenerative diseases in CSVD, such as Aβ deposition, and whether STAT6 is involved in the regulation of MSCs on the microglia phenotype in this disease are still unclear.

The present study investigated the cognitive function, Aβ deposition, distribution of AQP4 polarity, and neuroinflammation in RHRSP rats, and the effects of MSCs on cognitive impairment and aforementioned pathological changes, as well as the potential mechanism.

The present study used RHRSP rats to determine the effects of MSCs on the cognitive impairment and Aβ deposition in the brain of CSVD. In the RHRSP rats, significant spatial cognitive impairment, white matter lesions (WMLs) and Aβ deposition were detected. MSCs effectively improved the cognition and reduced Aβ deposition despite no improvement in WMLs. MSCs also significantly restored AQP4 polarity and alleviated neuroinflammation by regulating the phenotypic distribution of glial cells. Therefore, this study provided evidence of the effectiveness of MSCs against plaque deposition and cognitive impairment CSVD rats, suggesting the potential therapeutic effect of MSCs on CSVD.

Aβ deposition is one of the important causes of cognitive decline and is associated with neurodegenerative diseases, especially AD. It is increasingly found to be associated with age-related cerebral vascular diseases. Previous studies proposed that cerebrovascular injury was a major risk factor for AD dementia (Arvanitakis et al., 2016). It has been suggested that the dysfunction of efflux transport at the BBB causes the accumulation of Aβ in AD, leading to progressive cognitive dysfunction. On the other hand, influx of neurotoxic cell fragments results in inflammatory responses (van Assema et al., 2012). CSVD is a kind of cerebrovascular disease highly correlated with aging. Besides vascular Aβ lesions found in CAA, an association between Aβ and CSVD has been reported in many studies. Liu et al. found that serum Aβ42 combined with the CSVD score could predict cognitive impairment in patients with CSVD (Liu et al., 2021). An 8-year longitudinal study suggested that Aβ positivity in amyloid PET was an independent predictor of functional disability regardless of the occurrence of stroke in patients with subcortical VCI (Kang et al., 2022). The present study showed the deposition of Aβ in the brain of rats with CSVD, which was consistent with clinical findings. Many studies suggested that the intravenously injected stem cells could cross the blood-brain barrier and migrate to the lesion, thus inducing nerve regeneration, re-myelination and immune regulation. In particular, they reduced neuroinflammation and Aβ deposition (Urdzíková et al., 2014; Lykhmus et al., 2019). This study found that MSCs reduced Aβ deposition, thus providing new ideas that MSC intervention could alleviate Aβ deposition and cognitive impairment in CSVD.

Although Aβ decreased after treatment, no significant improvement was found in WMLs. The pathogenesis of WMLs is still unclear. The most possible pathogenic factor is BBB destruction; in addition, there are oxidative stress, immune inflammation, and impaired cerebral blood flow automatic regulation (Lin et al., 2017). The permeation of toxic substances including Aβ may also be involved in the formation of WMLs (Zhang et al., 2019; McAleese et al., 2021). At present, the temporal and spatial changes in Aβ and WML expression in the brain and the mechanism of promoting cognitive decline are rarely reported. The present study found that MSC treatment reversed Aβ deposition, but not WMLs, which indirectly explained the complexity of the formation of WMLs. Thus, it was hypothesized that the aggregation of Aβ was an important segment in the development of WMLs, but the formation of WMLs might be the result of a combination of many other factors and predated Aβ aggregation. This finding suggested that the dynamic change in WMLs and Aβ deposition should be observed in future studies.

Aβ accumulation in the brain is due to the imbalance between its production and clearance. Aβ clearance has been thought to be associated with the dysfunction of glymphatic system (Rasmussen et al., 2018). The glymphatic system in the brain plays an indispensable role in clearing waste products, and is responsible for a substantial part of the total Aβ clearance during sleep (Xie et al., 2013). The paravascular material exchange, mainly through cerebrospinal fluid (CSF) and interstitial fluid (ISF), is supported by the astrocytes involving AQP4 (Kress et al., 2014), which normally localizes to the perivascular astrocytic end feet. The knockout of AQP4 is known to significantly affect the clearance process, which gets aggravated when accompanied by neuroinflammation and the failure of perivascular AQP4 polarization (Iliff et al., 2012). The polarity of AQP4 refers to AQP4 located in the astrocyte foot, but astrocytes dysfunction, such as activation, leads to AQP4 redistribution to the astrocyte body. As a result, the clearance of brain metabolic waste gets blocked (Heppner et al., 2015). The present study showed thatAQP4 polarization was lost, and more AQP4 was redistributed from the foot processes to the cell body in CSVD rats. However, AQP4 expression significantly increased in the brains of RHRSP rats, which was consistent with the findings of several studies related to chronic cerebral hypoperfusion. It was probably related to astrocyte edema after ischemia (Manley et al., 2000). This study demonstrated that MSCs effectively downregulated AQP4 expression and restored AQP4 polarization, thus relieving cerebral edema and promoting Aβ clearance.

Neuroinflammation, known as glial cell activation, releases inflammatory cytokines that are thought to be partly responsible for cognitive impairment (Singh and Abraham, 2017). It usually begins with the activation of microglia and is considered to be closely associated with neurodegenerative disorders such as AD, PD, and multiple sclerosis (Lucin and Wyss-Coray, 2009; Huang et al., 2015; Subedi et al., 2017; Li et al., 2019; Thawkar and Kaur, 2019). It is usually accompanied by Aβ deposition because Aβ can activate microglia. Studies have shown that Aβ42 can increase the expression of M1-type markers and decrease the expression of M2-type markers in microglia cultured in vitro(Cheyuo et al., 2012). Meanwhile, the activated glial cells secrete a variety of inflammatory factors, which cause tissue damage and further aggravate the accumulation of Aβ. Thus, the inhibition of inflammatory response can relieve the severity of such neurodegenerative diseases (Vlad et al., 2008). Later, Rosenberg (2017) put forward that chronic hypertension could lead to neuroinflammation. The present study revealed the phenomenon in the CSVD rats as well. Activated microglia consists of two types: M1 and M2 (Allendorf et al., 2020). M1 microglia produces various pro-inflammatory factors, such as IL-1β, TNF-α, and IFN-γ, and promotes the inflammatory process (Gupta et al., 2018). However, M2 microglia release anti-inflammatory cytokines, such as IL-10 and TGF-β, engulf damaged cell fragments, and promote tissue restoration and neuron regeneration (Liu et al., 2019). In analogy with the “M1”/ “M2” microglia(Martinez and Gordon, 2014; Heppner et al., 2015), two types of activated astrocytes exist, which are termed “A1” (neurotoxic phenotype) and “A2” (neuroprotective phenotype), respectively in the case of neuroinflammation. Importantly, a recent study uncovered the decisive effect of reactive microglia on astrocyte polarization (Liddelow et al., 2017), confirming that the neurotoxic phenotype, A1 astrocytes were motivated by activated neuroinflammatory microglia. Similarly, A1 astrocytes significantly upregulated the intensity of complement cascade shown to be destructive to synapses, while A2 astrocytes upregulated many neurotrophic factors. In this study, the infusion of MSCs significantly attenuated the activation of microglia and astrocytes. Furthermore, it downregulated the M1/A1 phenotype and upregulated the M2/A2 phenotype in RHRSP rats, reduced inflammatory cytokine production that was M1-related and promoted the expression of M2-related anti-inflammatory cytokines. The possible reasons were that the reduction of MSC-induced Aβ reduced the conversion of microglia into M1, and MSCs might also be directly involved in regulating the glial cell phenotype.

The STAT family is involved in regulating the functional status of microglia/macrophages; the STAT6 signaling pathway appears to induce the transformation of macrophages into an anti-inflammatory phenotype (Brunn et al., 2014; Hu et al., 2015). The knockout of STAT6 has been found to result in the dysfunction of dead neuron clearance and increased expression of pro-inflammatory factors in mice with cerebral ischemia (Cai et al., 2019). Thus, the present study investigated the relationship between inflammatory cytokines and pSTAT6 to explore the mechanism underlying the neural immunomodulatory functions of MSCs. We discovered that MSCs could regulate the expression of inflammatory cytokines by activating the STAT6 signaling pathway while the pharmacological inhibitor of pSTAT6 (AS1517499) suppressed the effect, suggesting that pSTAT6 played a crucial role in the immunomodulatory process of MSCs.

In conclusion, this study showed the cognitive impairment and Aβ deposition in CSVD rats. MSC treatment improved the cognition and reduced Aβ accumulation despite no WML improvement, possibly in part through restoration of AQP4 polarity and alleviation of neuroinflammation. Furthermore, MSCs might regulate the microglia phenotype by activating the STAT6 signaling pathway to further relieve neuroinflammation.

A limitation of this study was that it used young rather than aged animals. The cognitive impairment and neuroimaging and pathological changes, except the aging of human CSVD, were better simulated in the chronic hypertensive rats (Fan Y. L. et al., 2015; Lan et al., 2015). Aging is an independent risk factor for stroke, and studies suggested that the self-regulation of cerebral blood flow might be altered with age (Wollner et al., 1979; Lartaud et al., 1993). Moreover, the incidence and imaging markers of CSVD increased with aging (Mu et al., 2022). Despite its importance, it is difficult to replicate the factor in rats because of their harder feeding and longer life span compared with mice. Therefore, it has always been a difficult problem in the animal research of CSVD. We will explore a more appropriate animal model of CSVD combined with the factor of aging in future studies.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding authors.

The animal study was reviewed and approved by the Animal Research Ethics Committee at Sun Yat-sen University.

XL, ZP, and YF designed the research. XL, FO, LH, PS, JY, YS, ML, and LL performed the experiments. XL drafted the manuscript. All authors approved the final version of the manuscript.

This study was supported by the grants from the Guangzhou Science and Technology Program Key Project (No. 202007030010), National Natural Science Foundation of China (No. 82071294), Guangzhou Municipal Science and Technology Project (No. 202002030073), Guangdong Provincial Key Laboratory of Diagnosis and Treatment of Major Neurological Diseases (No. 2020B1212060017), Guangdong Provincial Clinical Research Center for Neurological Diseases (No. 2020B1111170002), the Southern China International Cooperation Base for Early Intervention and Functional Rehabilitation of Neurological Diseases (Nos. 2015B050501003 and 2020A0505020004), and Guangdong Provincial Engineering Center for Major Neurological Disease Treatment, and Guangdong Provincial Translational Medicine Innovation Platform for Diagnosis and Treatment of Major Neurological Disease.

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.