GPR30 Functions as a Shared Inhibitory Signal for Both Pericyte‐Mediated Angiogenesis and Oligodendrocyte Myelination

ABSTRACT

Myelinogenesis is insufficient in numerous myelin‐related diseases in the CNS, leading to functional impairments. Myelinogenesis couples with angiogenesis to ensure adequate need of oxygen and nutrients for oligodendrocyte (OL) differentiation. However, approaches to synchronize myelino‐vascular coupling remain unavailable. We hypothesize the identification of shared signaling pathways in vascular cells and oligodendroglia may yield novel strategies to promote myelin repair through strengthening the blood vessel‐myelination coupling. Here, single‐cell sequencing and in situ hybridization revealed high expression of G‐protein‐coupled receptor 30 (Gpr30) in both vascular cells and oligodendroglia, with selective enrichment in pericytes and oligodendrocyte precursor cells (OPCs). Cell‐specific deletion of GPR30 in pericytes driven by PDGFRβ^CreERT2 resulted in enhanced angiogenesis and myelination in developing brains. GPR30 deletion in OPCs or antagonizing GPR30 by G15 resulted in increased MBP‐positive cell density and enhanced nanofiber wrapping capacity in vitro, thereby demonstrating an inhibiting role of GPR30 on OPC differentiation. To elucidate the coordinative role of GPR30 in both cell types, we employed NG2^CreERT to induce a conditional knockout of GPR30 in both NG2‐positive pericytes and OPCs. The conditional deletion of GPR30 enhanced myelination and increased vascular density in developing brains. Further, GPR30 cKO or G15 treatment enhanced myelin repair and functional recovery in the chronic neonatal hypoxia and lysolecithin‐induced demyelination model, suggesting that antagonizing GPR30 is a promising strategy to synchronize angiogenesis with myelination to promote myelinogenesis. These findings establish GPR30 antagonism as a promising approach to enhance myelin repair through synchronizing pericyte‐mediated angiogenesis and OPC differentiation.

• GPR30 is selectively expressed by pericytes and oligodendroglia in developing brains.

• GPR30 deletion in pericytes enhances angiogenesis and subsequently OL myelination.

• GPR30 cKO or antagonism directly promotes OPCs differentiation in cultures.

• GPR30 deletion in OPCs and pericytes promotes myelin repair and functional recovery against hypoxia and demyelination.

1. Introduction

Myelin sheaths in the central nervous system (CNS) accelerate action potential conduction along axons and play a crucial role in maintaining various neurofunctions (Baumann and Pham‐Dinh ²⁰⁰¹; McKenzie et al. ²⁰¹⁴; Stadelmann et al. ²⁰¹⁹; Xin and Chan ²⁰²⁰). Oligodendrocytes (OLs) generate myelin sheaths and also provide metabolites to maintain axonal integrity (Chamberlain et al. 2021; Lee et al. ²⁰¹²; Nave ²⁰¹⁰). Myelin deficits significantly contribute to neuronal dysfunction in the developing brain and in neurodegenerative diseases, such as neonatal white matter injury (WMI), multiple sclerosis (MS), and Alzheimer's disease (AD) (Back 2017; Chen, Liu, et al. ²⁰²¹; Reich et al. ²⁰¹⁸; Soria Lopez et al. ²⁰¹⁹). Oligodendrocyte precursor cells (OPCs) are widely distributed throughout the CNS, including in lesions and technically can differentiate into myelinating OLs (Hughes et al. 2013; Plemel et al. ²⁰¹⁷). Therefore, pro‐myelination strategies are considered as promising approaches to improve functional recovery in developing brains and in demyelinating diseases (Bottes and Jessberger 2021; Call and Bergles ²⁰²¹; Cordano et al. ²⁰²²; Hooijmans et al. ²⁰¹⁹; Mei et al. ²⁰¹⁶; Najm et al. ²⁰¹⁵). Notably, once the differentiation program is activated in OPCs, the cells must produce 3000 to 6000 times more membranes to form myelin sheaths within hours (Baron and Hoekstra 2010; Chong et al. ²⁰¹²). Thus, myelination is an energy‐intensive process that needs sufficient nutrients and oxygen support. It has been repeatedly demonstrated that myelination is closely coupled with angiogenesis in developing brains (Su et al. 2023; Yuen et al. ²⁰¹⁴; Zhao et al. ²⁰²²). Supporting this view, a recent study showed that pericyte‐mediated angiogenesis is required for myelination in the brain under both normoxic and hypoxic conditions (Ren et al. 2024). However, effective strategies to synergistically regulate myelinogenesis and angiogenesis remain unavailable.

The G‐protein‐coupled receptor (GPCR) superfamily is the target of most clinically used selective drugs so far (Katritch et al. 2013; Lorente et al. ²⁰²⁵; Zhang et al. ²⁰²⁴). The GPCRs regulate many biological processes, and the antagonists or agonists targeting these receptors accordingly exert high potency on myelination and angiogenesis (Folts et al. 2019; Hirahara et al. ²⁰¹³; McDonald et al. ²⁰²⁵). A few GPCRs, including GPR56, S1PR, and the Wnt receptor, have been reported to regulate myelination and angiogenesis, respectively (Danuta and Betty 2015; Elisabetta ²⁰¹⁰; Giera et al. ²⁰¹⁵; Guo et al. ²⁰¹⁵; Henry et al. ²⁰¹⁹). Some of them exerted opposing effects on myelination and vascularization. Activation of GPR56 inhibited VEGF production from melanoma cell lines and impeded melanoma angiogenesis and growth; on the contrary, GPR56 played a positive role in myelination in vivo (Giera et al. 2015; Yang et al. ²⁰¹¹). Given the close coupling between myelinogenesis and angiogenesis, identification of GPCRs concurrently regulate blood vessel growth and oligodendrocyte differentiation may yield a novel strategy for treating myelin deficits.

Here, we identified that GPR30 is selectively expressed in pericytes and oligodendroglia in the developing mouse brain, as revealed by single‐cell sequencing and in situ hybridization. Conditional deletion of GPR30 in PDGFRβ‐positive pericytes increased vascular density and consequently myelination in developing brains. To confirm the direct effect of GPR30 on regulating OPC differentiation and myelination, cortical OPCs were purified and cultured in vitro on coverslips with engineered nanofibers (pseudo‐axons). The OPCs with GPR30 ablation or exposed to G15 (the GPR30 antagonist) exhibited enhanced differentiation and nanofiber wrapping, demonstrating a direct effect on OL differentiation. We further used the NG2^CreERT to induce GPR30 deletion in both pericytes and OPCs. We found enhanced myelination and increased vascular density in the developing brain, suggesting that GPR30 antagonism synchronizes angiogenesis and myelination. More importantly, the GPR30 cKO in pericytes and OPCs or G15 treatment promotes angiogenesis and myelinogenesis upon hypoxic insults or lysolecithin‐induced demyelination. Together, these results demonstrate that antagonizing GPR30 is a promising strategy for myelin repair by coupling angiogenesis and myelinogenesis.

2. Materials and Methods

2.1. Mice

All experimental mice in this study were derived from the C57BL/6J genetic background. No sex differences were observed in the morphology and behavior of male and female mice used in the study during the developmental stage. For the demyelination model in adulthood, male mice were used for experiments. The GPR30 floxed mouse line was generated by using the CRISPR/Cas9 strategy, in which exon 3 was flanked by loxP sites (GemPharmatech Co. Ltd., Cat #T052092). To obtain NG2^CreERT; Tau‐mGFP mice, the NG2^CreERT line (The Jackson Laboratory, Cat #008538) was crossed with the Tau‐mGFP (The Jackson Laboratory, Cat #021162) line. The GPR30 floxed mouse line was crossed with the NG2^CreERT mice or the PDGFRβ^CreERT2 (The Jackson Laboratory, Cat #029684) mice (Gerl et al. 2015). The mT/mG line was crossed with the NG2^CreERT mice or the PDGFRβ^CreERT2 mice (Cheng et al. 2024; Ren et al. ²⁰²⁴). The animal breeding center of Third Military Medical University provided wildtype mice. The offspring were genotyped by PCR using specific DNA extraction, PCR procedures, and electrophoresis detection. The mice for all studies were housed in individually ventilated cages (IVC) systems under standard 12 h light/dark cycle conditions with free choice feeding. The Laboratory Animal Welfare and Ethics Committee of the Third Military Medical University approved all the animal experiments.

2.2. Human Brain Samples

The human brain sample was obtained from the Department of Neurosurgery at Xinqiao Hospital (Third Military Medical University, Chongqing, China). The patient was diagnosed with spontaneous cerebral hemorrhage and received emergency surgery. The brain tissue was sampled from the peri‐lesion area and fixed immediately in 4% paraformaldehyde (PFA) for 24 h before further processing. The study was conducted in accordance with the guidelines of the Medical Ethics Committee of Xinqiao Hospital and with the ethical standards outlined in the Declaration of Helsinki (Approval Number: 2022‐361‐01). The patient and their guardians were fully informed, and written informed consent was obtained from both the patient and their guardians.

2.3. Administration of Tamoxifen

Tamoxifen (Sigma‐Aldrich, Cat: T5648) was dissolved in corn oil at concentrations of 10 mg/mL (during the developmental stage) and 30 mg/mL (after adulthood). As previously described (Ren et al. 2024), mice were treated with tamoxifen at a dose of 50 mg/kg by oral gavage to induce Cre recombination.

2.4. Drug Treatment

The GPR30 antagonist, G15 (Selleck.cn, Cat: S6651), was used in this study. Animals were dosed at 10 mg/kg (dissolved in 2% dimethyl sulfoxide (DMSO, v/v), 20% polyethylene glycol 300 (PEG300, v/v), 5% Tween‐80 (v/v), and 73% double‐distilled water (v/v) at 2 mg/mL). The G15 solution must be prepared as needed. The vehicle group was treated with a mixture of DMSO, PEG300, Tween‐80, and double‐distilled water at the same concentration, dose, and formulation. The hypoxic pups were administered G15 or vehicle from P3 to P10. The adult mice were given G15 continuously every day for 2 weeks after stereotaxic injection.

2.5. Stereotaxic Injection

Adult C57BL/6 mice were anesthetized and fixed on a stereotaxic instrument after scalp disinfection and skull exposure. According to the Mouse Brain Atlas for target coordinates, a burr hole was drilled, and lysolecithin was microinjected at 0.2 μL/min (total 1.5 μL). The syringe was retained for 5 min post‐injection before slow withdrawal, followed by incision suturing. Mice recovered on a heating pad with 24 h postoperative monitoring.

2.6. Brain Tissue Preparation

Mice were subjected to deep anesthesia via administration of 20% Urethane. Following an initial perfusion with 0.01 M phosphate‐buffered saline (PBS) to flush out residual blood, transcardial perfusion was subsequently performed using 4% paraformaldehyde dissolved in 0.1 M phosphate buffer (PB). Dissected brains were harvested immediately after perfusion and then immersed in the same 4% paraformaldehyde‐PB solution for overnight post‐fixation. After post‐fixation, the isolated tissues were dehydrated by incubation in 30% sucrose prepared with 0.01 M PBS. For cryosectioning, brain samples were embedded in optimal cutting temperature (OCT) compound (SAKURA, Cat. No. 4583, USA), after which coronal sections (20 μm in thickness) were obtained using a cryostat microtome (Leica MS 1850, Wetzlar, Germany).

2.7. Immunofluorescence Staining and Image Acquisition

Brain slices were rinsed with phosphate‐buffered saline (PBS) and subsequently blocked in PBS supplemented with 5% bovine serum albumin (BSA) and 0.5% Triton X‐100 for 2–3 h at room temperature in the dark. Following this, the slices were incubated overnight at 4°C with the primary antibody solution, then for 2 h at room temperature in the dark with the secondary antibody solution. Nuclear staining was performed using DAPI for 8–10 min. Thereafter, the slices were mounted, coverslipped, and subjected to microscopic observation and imaging (Ren et al. 2024). Primary antibodies were to the following proteins: GPR30 (1:500, rabbit, Cat #PA5‐28647, Thermo Fisher), MBP (1:500, rat, Cat #MAB386, Millipore), CC1 (1:500, mouse, Cat #OP80, Calbiochem), PDGFRα (1:500, goat, Cat #AF1062, R&D systems), GFP (1:500, goat, Cat #ab5450, Abcam), NG2 (1:500, rabbit, Cat #ab5320, Millipore), NF200 (1:500, Cat #N4142, Sigma‐Aldrich), SOX10 (1:500, rabbit, Cat #ab180862, Abcam), Olig2 (1:500, rabbit, Cat #AB9610, Millipore), Caspr (1:500, mouse, Cat #75‐001, Davis/NIH NeuroMab Facility), CD31 (1:500, rat, Cat #553370, BD Pharmingen), PDGFRβ (1:250, goat, Cat #AF1042, R&D systems). Secondary antibodies included: AlexaFluor‐488‐, AlexaFluor‐568‐, or AlexaFluor‐647‐conjugated secondary antibodies against goat, rabbit, mouse, or rat (1:1000; Invitrogen). DAPI was used to counterstain nuclei. Fluorescent images were captured using a confocal laser‐scanning microscope (Olympus, FV3000, Shinjuku, Tokyo), a spinning disk confocal super‐resolution microscope (Olympus, SpinSR10, Shinjuku, Tokyo), or a fluorescence microscope (Olympus, VS200, Shinjuku, Tokyo). Images were analyzed using CellSens Dimension software and ImageJ (version Java 1.8.0_77).

2.8. 3D Imaging

OLs were scanned using a confocal microscope (Olympus, FV3000) with a 60× objective (oil, NA = 1.42) and imaged on z‐stacks at an interval of 0.50 μm. Individual pericytes and vessels were scanned with a confocal microscope (Olympus, SpinSR10) using a 60× objective (oil, NA = 1.42). Isosurface reconstruction of a single OL or pericyte and vessel images was processed with Imaris ×64 software (Bitplane, USA).

2.9. Analysis of scRNA‐Seq

The scRNA‐seq dataset from mouse brain samples at P10 was obtained from our recent report (Ren et al. 2024) (GEO: GSE262996). For quality control, cells expressing fewer than 200 genes and more than 5000 genes were removed, as were those with more than 10% mitochondrial content. Only genes expressed in three or more cells were included for analysis. Next, the R workflow based on the Seurat (v4) package and ggplot2 was applied to visualize the scRNA‐seq results.

2.10. Purification of OPCs and Drug Treatment

The NG2^CreERT; GPR30 fl/fl mice and littermates were used for purification of OPCs at P7 after tamoxifen treatment for 4 times from P3. OPC purification via immunopanning and subsequent OPC‐nanofiber co‐culture were conducted according to a previously established protocol (Lee et al. 2013). Before OPC isolation, panning dishes were coated overnight at room temperature with goat anti‐mouse IgG/M secondary antibodies (10 μg/mL) reconstituted in 50 mM Tris–HCl buffer (pH 9.5). Cerebral cortices dissected from postnatal day 7 (P7) mice were enzymatically digested with papain solution (Sigma‐Aldrich) at 37°C to generate a single‐cell suspension. This suspension was then incubated with mouse anti‐PDGFRα primary antibody (1:250 dilution; Abcam, Catalog No. ab96569) for 30 min at room temperature to label the target OPCs. OPCs captured by a secondary antibody‐coated plate were seeded onto glass coverslips or aligned nanofiber inserts (Sigma‐Aldrich) and maintained in basal medium (BM). The BM was formulated with Dulbecco's modified Eagle medium (DMEM; Gibco) supplemented with B27 supplement (Gibco), N2 supplement (Gibco), N‐acetylcysteine (Sigma‐Aldrich), forskolin (Sigma‐Aldrich), and 12.5 ng/mL platelet‐derived growth factor‐AA (PDGF‐AA; Peprotech). After the initial 48 h culture period, isolated OPCs were cultured in differentiation medium (identical to BM but without PDGF‐AA supplementation) for an additional 48 h. During this differentiation phase, cells were either treated with G15 (Selleck.cn, Cat: S6651) at a final concentration of 0.5 μM or maintained in G15‐free medium as controls.

2.11. RNAscope In Situ Hybridization

RNAscope in situ hybridization was carried out following the manufacturer's instructions for the RNAscope Multiplex Fluorescent Reagent Kit v2 (Cat: 323100, ACD). Brain tissues from a human sample or a 3‐month‐old regular mouse were sectioned at 10 μm and stored at −80°C. Sections were first incubated with RNAscope Hydrogen Peroxide for 10 min at room temperature, followed by treatment with RNAscope Target Retrieval Solution at 99°C for 5 min. Thereafter, sections were incubated in RNAscope Protease Plus at 40°C for 30 min. RNAscope probes for Hs‐GPER (Cat: 553361, ACD) or Mm‐Gper1‐C3 (Cat: 475251‐C3, ACD) were hybridized to sections at 40°C for 2 h, and Opal 570 fluorescent dye (Akoya Biosciences) was used for imaging.

2.12. Behavioral Tests

2.12.1. Beam Walking Test

A 0.55 cm‐wide beam was set up in a dark room, with one end illuminated by a lamp and the other placed in a 20 × 10 × 20 cm³ black box. During the acquisition phase, mice were placed on the beam at 30, 50, and 70 cm from the illuminated end for 3 days of training. For the test phase, videos were recorded bilaterally along the beam. The frequency of hindlimb slips evaluated mouse performance during their traversal of the 80 cm‐long beam (Brooks and Dunnett 2009).

2.12.2. Morris Water Maze Task

This experiment was conducted according to a previously reported protocol (Vorhees and Williams 2006). Briefly, the water maze was partitioned into four quadrants. During the acquisition phase, all mice underwent four daily training sessions for 3–4 consecutive days to locate the hidden platform in the third quadrant. Following training, the platform was removed, and the mice were placed into the water maze from the first quadrant. Long‐term spatial memory was assessed by quantifying three parameters: time spent in the platform‐containing quadrant, distance traveled within this quadrant, and the number of crossings over the former platform area. All investigators were blinded to mouse genotypes.

2.13. Statistical Analysis

Data analysis was performed with GraphPad Prism 10.1.2. All graphical data are presented as mean ± SD. Group differences were assessed using an unpaired t‐test or one‐way ANOVA. Repeated‐measures two‐way analysis of variance (two‐way RM‐ANOVA) was applied for data analysis of the Morris water maze (MWM) acquisition phase. The p values were indicated on the figures, and significance was reported as *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns: not significant.

3. Results

3.1. GPR30 Is Highly Expressed in Pericytes and Oligodendroglia

To identify potential shared pathways for myelination and angiogenesis, we analyzed our previously published single‐cell RNA sequencing (scRNA‐seq) dataset using neonatal mouse brains (GSE262996) (Ren et al. 2024). We chose to focus on orphan G‐protein‐coupled receptors (GPCRs), those without known endogenous ligands (Alavi et al. 2019; Keifi Bajestani et al. ²⁰²⁴). The expression levels of orphan GPCR families in the oligodendrocyte lineage cells, pericytes, endothelial cells, microglia, astrocytes, and neurons were analyzed (Dennis et al. 2024; Hou et al. ²⁰²³; Jäkel et al. ²⁰¹⁹; Liu et al. ²⁰²⁵) (Figure 1A). We then screened for receptors co‐expressed in oligodendroglia and vascular cells, respectively. There were 12 orphan GPCRs co‐expressed in OPCs and vascular cells, including Gpr146, Gpr161, Gpr19, Gpr48, Gpr56, Gpr85, Gpr30, Gpr153, Gpr162, Gpr173, Gpr27, and Gpr45 (Figure 1B). Then, those candidates that were also highly expressed in neurons, astrocytes, and microglia were excluded, including Gpr161, Gpr19, Gpr48, Gpr56, Gpr85, Gpr162, Gpr173, Gpr27, and Gpr45 (Figure S1). We found that Gpr30 exhibits comparably selective expression in the oligodendrocyte lineage and vascular cells, with particular enrichment in pericytes (Figure 1C). Then, single‐molecule fluorescence in situ hybridization RNAscope was performed to validate Gpr30 expression in oligodendroglia and pericytes. Consistent with the scRNA‐seq data, Gpr30 mRNA was markedly expressed and specifically localized in platelet‐derived growth factor receptor β (PDGFRβ)‐positive pericytes in both human and mouse brains (Figure 1D,E). In addition, Gpr30 mRNA was also found in both PDGFRα‐positive OPCs in mouse brains and SOX10‐positive oligodendrocyte lineage cells in human brains (Figure 1D,E). Further, immunostaining results showed that GPR30 was expressed in both PDGFRα‐positive OPCs (green, left panels) and PDGFRβ‐positive pericytes in mouse brains (Figure 1F). These results align with previous studies indicating GPR30 expression in vascular and oligodendroglial cells in the CNS via cell‐lineage tracing or immunostaining (Alexander et al. 2017; Hirahara et al. ²⁰¹³; Wu et al. ²⁰²⁵). These findings suggest GPR30 signaling as a shared signaling pathway in oligodendroglial and vascular cells.

FIGURE 1.

GPR30 expression in pericytes and oligodendroglia. (A) UMA plots identifying oligodendrocytes (Oligos), OPC, endothelial, pericytes, neurons, microglial and astroglia in P10 mouse brains. (B) Venn diagram showing the co‐expressed GPCR genes in OPCs, pericytes, and endothelial cells, respectively. (C) Expression level of Gpr56, Gpr30, Gpr19, and Gpr45 in various cell types in brains. (D) Representative images showing Gpr30 mRNA expression (red) in PDGFRα‐positive OPCs (green, upper panels) and PDGFRβ‐positive pericytes (green, lower panels) in mouse brains. Right panels show enlarged images of dotted box in left panel, respectively. White arrows indicate expressing cells. Scale bar: 20 μm (left panels); 2 μm (right panels). (E) Representative images show Gpr30 mRNA expression (red) in SOX10‐positive oligodendrocyte lineage cells (green, upper panels) and PDGFRβ‐positive pericytes (green, lower panels) in human brain tissues. Right panels show enlarged images of dotted box in left panel, respectively. White arrows indicate expressing cells. Scale bar: 20 μm (left panels); 2 μm (right panels). (F) Representative images show GPR30 expression (red) in PDGFRα‐positive OPCs (green, left panels) and PDGFRβ‐positive pericytes (green, right panels) in wildtype mouse brains. White arrows indicate expressing cells. Scale bar: 20 μm (left panels); 10 μm (right panels).

3.2. Conditional Deletion of GPR30 in Pericytes Enhances Angiogenesis

To elucidate the function of GPR30 in pericytes, we generated a GPR30 floxed line by integrating loxP sites on both sides of exon 3 and crossed it with PDGFRβ^CreERT2 mice to obtain the PDGFRβ^CreERT2; GPR30 fl/fl line (Gerl et al. 2015) (Figure 2A). The efficiency of PDGFRβ^CreERT2 was examined by crossing with the mTomato/mGFP (mT/mG) line (Figure 2A). After induction for 4 days, the mGFP‐positive cells were selectively colocalized with ~90.1% PDGFRβ‐positive cells revealed by immunostaining in the postnatal day 11 (P11) brains (Figure 2B,C). To validate the knockout efficiency, RNAscope and co‐immunostaining for PDGFRβ showed that GPR30 mRNA was sharply decreased in PDGFRβ‐positive pericytes in PDGFRβ^CreERT2; GPR30 fl/fl mice, compared with the littermates (Figure 2D). Furthermore, GPR30 expression was sharply decreased in PDGFRβ‐positive pericytes in PDGFRβ^CreERT2; GPR30 fl/fl mice using immunostaining for PDGFRβ and GPR30 (Figure 2E). Then we explored the effects of special GPR30 deletion on angiogenesis and subsequent myelination. The PDGFRβ^CreERT2; GPR30 fl/fl or GPR30 fl/fl controls (GPR30 fl/fl or PDGFRβ^CreERT2) mice were induced from P3, and the brains were collected at P11 for examining blood vessel density and myelination (Figure 2B). Given the known limitations of the PDGFRβ^CreERT2 mouse line (https://www.jax.org/strain/029684), the brains of littermate PDGFRβ^CreERT2; GPR30 fl/+ mice were also collected. Intriguingly, the densities of CD31‐positive vessels and PDGFRβ‐positive pericytes were increased in the GPR30 conditional knockout (cKO) brains, compared with either GPR30 fl/fl or PDGFRβ^CreERT2; GPR30 fl/+ controls. Interestingly, the CD31‐positive blood vessels in the GPR30 cKO brains exhibited a larger lumen and more branches per vessel than those of the GPR30 fl/fl or PDGFRβ^CreERT2; GPR30 fl/+ controls (Figure 2F,G). To assess whether the increased angiogenesis could alter myelination, immunostaining for myelin basic protein (MBP) and CC1 was performed to identify myelin (MBP‐positive) and mature OLs (CC1‐positive), respectively. MBP‐positive myelin areas and CC1‐positive cell density in the GPR30 cKO brains were significantly increased, compared with those of GPR30 fl/fl or PDGFRβ^CreERT2; GPR30 fl/+ controls, without changing the density of PDGFRα‐positive OPCs (Figure 2H,I). These results indicated that GPR30 in pericytes plays an inhibitory role in regulating angiogenesis, and subsequently governs myelinogenesis in developing brains.

FIGURE 2.

GPR30 deletion in pericytes promotes angiogenesis and myelination. (A) Schematic diagram showing genetic strategy for generating PDGFRβ^CreERT2; GPR30 fl/fl and PDGFRβ^CreERT2; mT/mG reporter lines, and schematic diagram showing selective GPR30 deletion in PDGFRβ‐positive pericytes. (B) Schematic displaying the time course for tamoxifen induction and histology in PDGFRβ^CreERT2; GPR30 fl/fl mice and littermate controls or PDGFRβ^CreERT2; mT/mG mice. (C) Representative images show co‐expression of mGFP (green)‐positive signal and PDGFRβ (red), and percentage of mGFP/PDGFRβ double‐positive cells in PDGFRβ‐positive cells. Scale bar: 20 μm. n = 3 biologically independent mice. (D) Gpr30 RNAscope (red) counterstained with PDGFRβ (green) in GPR30 cKO and GPR30 CTL mouse brains (white arrows indicating expressing cells and yellow arrows indicating non‐expressing cells). Scale bar: 20 μm. (E) Representative images showing the co‐localization of GPR30 (red) and PDGFRβ (green) in GPR30 cKO and GPR30 CTL mouse brains, and quantification of percentage of GPR30/PDGFRβ double‐positive cells to PDGFRβ‐positive cells in GPR30 cKO and GPR30 CTL brains (white arrows indicating colocalized cells and yellow arrows indicating non‐colocalized cells). Scale bar: 20 μm. n = 3 biologically independent mice for each group. (F) Representative images of PDGFRβ (green, yellow dotted lines indicating corpus callosum)‐positive pericytes, and CD31 (red)‐positive endothelium in PDGFRβ^CreERT2; GPR30 fl/fl brains (lower panels) and controls (PDGFRβ^CreERT2; GPR30 fl/+ or GPR30 fl/fl, middle and upper panels). Middle panel in each row shows isosurface reconstruction of PDGFRβ‐positive pericytes and CD31‐positive endothelium of white dotted boxes in the corresponding left panel, and right panel showing enlarged images of white dotted box in the corresponding left panel. Scale bar: 50 μm (left panels); 20 μm (middle and right panels). (G) Quantification of CD31‐positive area, PDGFRβ‐positive area, branches of vessel, length and width of vessel in PDGFRβ^CreERT2; GPR30 fl/fl brains and littermate controls (PDGFRβ^CreERT2; GPR30 fl/+ or GPR30 fl/fl). n = 3 biologically independent mice for each group. (H) Representative images of MBP expression (red, left panels, yellow dotted lines indicating corpus callosum), CC1‐positive OLs (green, middle panels, yellow dotted lines indicating corpus callosum) and PDGFRα‐positive OPCs (red, right panels) in PDGFRβ^CreERT2; GPR30 fl/fl brains and littermate controls (PDGFRβ^CreERT2; GPR30 fl/+ and GPR30 fl/fl). Scale bar: 200 μm (left panels); 50 μm (middle and right panels). (I) Quantification of MBP‐positive area, CC1‐positive OLs and PDGFRα‐positive OPCs in PDGFRβ^CreERT2; GPR30 fl/fl brains and littermate controls (PDGFRβ^CreERT2; GPR30 fl/+ or GPR30 fl/fl). n = 3 biologically independent mice for each group. Error bars represent mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns: not significant. The unpaired t test was used to assess the significance between control and GPR30 cKO groups, one‐way ANOVA was used to analyze the significance among GPR30 fl/fl, PDGFRβ^CreERT2; GPR30 fl/+ and PDGFRβ^CreERT2; GPR30 fl/fl mice. Dots represent number of animals.

3.3. GPR30 Antagonism Promotes OPC Differentiation In Vitro

Since GPR30 is expressed in oligodendroglial cells, we next sought to determine its role in oligodendrocyte differentiation. We purified OPCs from P8 mouse brains and examined their differentiation in culture by immunostaining for MBP (OLs) and NG2 (OPCs) (Figure 3A). G15 is a selective GPR30 antagonist with very high efficacy and specificity (Dennis et al. 2009). After being exposed to 0.5 μM G15 for 48 h, MBP‐positive OL density was significantly increased in the purified OPC cultures (Figure 3B). To examine the wrapping capacity of OLs, OPCs were cultured on coverslips engineered with plastic nanofibers as pseudo‐axons with the presence of geometrical cues. Interestingly, the MBP‐positive OLs wrapped around nanofibers and the number of wrapped segments per MBP‐positive OL was increased after treatment with G15 in contrast to the vehicle‐treated controls (Figure 3C). To determine if the G15 effect on OPC differentiation is dependent on GPR30, we isolated OPCs with GPR30 conditional ablation from the NG2^CreERT; GPR30 floxed mouse line (Figure 3D). Next, OPCs were purified from P8 cortex of NG2^CreERT; GPR30 fl/fl mice or littermate GPR30 fl/fl controls after induction from P3 (Figure 3E). After 2 days of culture, the density of MBP‐positive OLs was increased with GPR30 deletion, in contrast to wildtype controls (Figure 3F,G). Significantly, 0.5 μM G15 did not change the density of MBP‐positive cells and wrapping segments on nanofibers per OL after GPR30 deletion, in contrast to the control (Figure 3F,G), demonstrating that the effect of G15 on OPC differentiation is mediated by GPR30. These in vitro results demonstrate that GPR30 antagonism can directly promote OPC differentiation and myelination.

FIGURE 3.

G15 or GPR30 deletion on OPCs differentiation in vitro. (A) Schematic diagram displaying workflow of OPC purification and cultures. (B) Representative images and quantification of MBP‐positive cells (red), NG2‐positive cells (green) upon 0.5 μM G15 treatment for 48 h. Scale bar: 50 μm. n = 4 independent coverslips. (C) Representative images (left panels) and isosurface reconstruction (right panels) of MBP (red)‐positive signal wrapping along nanofibers (gray), and quantification of segment per MBP‐positive OL. Scale bar: 10 μm. n = 4 independent coverslips. (D, E) Genetic strategy and time course for tamoxifen induction and purification of OPCs for the NG2^CreERT; GPR30 fl/fl mice and littermate controls. (F) Representative images and quantification of MBP (red)‐positive cells, NG2 (green)‐positive cells in GPR30 CTL OPCs, GPR30 cKO OPCs, and GPR30 cKO OPCs with 0.5 μM G15 treatment. Scale bar: 50 μm. n = 4 independent coverslips. (G) Representative images (left panels) and isosurface reconstruction (right panels) of MBP (red)‐positive signal wrapping along nanofibers (gray), and quantification of segment per MBP‐positive OL in GPR30 CTL OPCs, GPR30 cKO OPCs, and GPR30 cKO OPCs with 0.5 μM G15 treatment. Scale bar: 10 μm. n = 4 independent coverslips. Error bars represent mean ± SD. **p < 0.01, ****p < 0.0001, ns: not significant. The unpaired t test was used to assess the significance between vehicle and G15 groups; one‐way ANOVA was used to assess the significance among CTL, cKO‐vehicle, and cKO‐G15 groups. Dots represent coverslip numbers.

3.4. Deletion of GPR30 in Pericytes and OPCs Promotes Angiogenesis and Myelination In Vivo

Next, we wanted to know if conditional knockout of GPR30 in both pericytes and OPCs can accelerate myelination in developing brains. The NG2^CreERT line drives Cre recombinase expression in both pericytes and OPCs in vivo. In the NG2^CreERT line, tamoxifen treatment can induce recombination in NG2‐positive pericytes and OPCs (Figure 4A). The NG2^CreERT; GPR30 fl/fl mice and littermate controls (GPR30 fl/fl) were induced with tamoxifen from P3 for 4 days, and brain tissues were harvested on P11 (Figure 4B). To evaluate Cre recombinase expression, we generated the NG2^CreERT; mT/mG line and immunostaining revealed selective mGFP expression in NG2‐positive pericytes and OPCs in the P11 mouse brains after tamoxifen induction from P3 to P6 (Figure 4A–C). Furthermore, immunostaining results showed that GPR30 expression was sharply decreased both in PDGFRβ‐positive pericytes and PDGFRα‐positive OPCs in NG2^CreERT; GPR30 fl/fl mice, compared with the littermate controls, suggesting specific deletion of GPR30 (Figure 4D). Then we explored the role of GPR30 in NG2‐positive cells in myelination and angiogenesis. Notably, immunostaining for MBP and CC1 demonstrated that MBP expression and density of CC1‐positive mature OLs were markedly increased in the GPR30 cKO mouse brains, compared with the littermate controls (GPR30 fl/fl) (Figure 4E). In contrast, the number of PDGFRα‐positive or PDGFRα/Olig2 double‐positive OPCs was not significantly altered upon GPR30 deletion (Figure 4F). Furthermore, the length of CD31‐positive blood vessels in GPR30 cKO mice was significantly increased compared with that in control mice, with larger lumens, but vessel branches showed no significant change (Figure 4G). We further examined the effects of G15 on angiogenesis and myelination in developing brains. G15 was administered to the P3 wildtype (C57BL/6) mice for 7 days without severe adverse effects (Figure S2A). The brains were collected at P11, and immunostaining results showed significantly increased MBP area and density of CC1‐positive OLs, but PDGFRα‐positive OPCs showed no significant change (Figure S2A–C). Similarly, the CD31‐positive vessel density was also increased in the G15‐treated brains, with longer vessel length and more branches, but without altering the lumen size. These findings phenotyped the effects of GPR30 deletion in both pericytes and OPCs (Figure S2D), demonstrating GPR30 signaling pathway can concurrently regulate angiogenesis and myelinogenesis, contributing to myelination in developing brains.

FIGURE 4.

GPR30 deletion in pericytes and OPCs promotes angiogenesis and myelination in developing brains. (A) Schematic diagram showing genetic strategy for generating the NG2^CreERT; GPR30 fl/fl mouse line and NG2^CreERT; mT/mG reporter line and displaying specific deletion of GPR30 in NG2‐positive pericytes and OPCs in the NG2^CreERT; GPR30 fl/fl mouse. (B) Schematic diagram displaying the time course for induction and histology in the NG2^CreERT; GPR30 fl/fl mice and littermate controls or NG2^CreERT; mT/mG mice. (C) Representative images show co‐expression of mGFP (green)‐positive signal and PDGFRα (red, upper panels)/PDGFRβ (red, lower panels), and percentage of mGFP/PDGFRα double‐positive cells in PDGFRα‐positive cells and mGFP/PDGFRβ double‐positive cells in PDGFRβ‐positive cells in NG2^CreERT; mT/mG brains. White arrows indicating colocalized cells and yellow arrows indicating non‐colocalized cells. Scale bar: 20 μm. n = 4 biologically independent mice. (D) Representative images of GPR30 expression (red) in PDGFRα (green, left panels) or PDGFRβ (green, middle panels) in NG2^CreERT; GPR30 fl/fl mice and controls, and quantification of GPR30/PDGFRα double‐positive cells in PDGFRα‐positive cells and GPR30/PDGFRβ double‐positive cells in PDGFRβ‐positive cells. White arrows indicate colocalized cells and yellow arrows indicating non‐colocalized cells. Scale bar: 10 μm (left panels); 20 μm (middle panels). n = 3 biologically independent mice. (E) Representative images and quantification of MBP (red, left panels), and CC1 (green, middle panels)‐positive OLs in NG2^CreERT; GPR30 fl/fl mice and littermate controls. Dotted lines indicating corpus callosum (middle panels). Scale bar: 200 μm (left panels); 50 μm (middle panels). n = 3–4 biologically independent mice for each group. (F) Representative images and quantification of Olig2 (green)/PDGFRα (red) double‐positive OPCs in NG2^CreERT; GPR30 fl/fl mice and littermate controls. Scale bar: 20 μm. n = 3–4 biologically independent mice for each group. (G) Representative images of PDGFRβ (green)‐positive pericytes and CD31 (red)‐positive endothelium in NG2^CreERT; GPR30 fl/fl mice and littermate controls, and quantification of CD31‐positive area, PDGFRβ‐positive area, branches of vessel, length and width of vessel in NG2^CreERT; GPR30 fl/fl mice and littermate controls. Middle panels show isosurface reconstruction of PDGFRβ‐positive pericytes and CD31‐positive endothelium of white dotted boxes in left panels, and right panels show enlarged images of white dotted boxes in left panels. Yellow dotted lines indicating corpus callosum (left panels). Scale bar: 100 μm (left panels); 20 μm (middle and right panels). n = 3–4 biologically independent mice for each group. Error bars represent mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns: not significant. The unpaired t test was used to assess the significance between NG2^CreERT; GPR30 fl/fl mice and littermate controls. Dots represent animal numbers.

3.5. GPR30 cKO Promotes Myelin Repair and Functional Recovery After Hypoxia

Chronic hypoxia constitutes the primary etiology of myelin deficits and associated long‐term neurofunctional impairments in neonatal white matter injury (WMI) (Back 2017; Cree et al. ²⁰¹⁸; Wang, Yang, et al. ²⁰¹⁸). One mechanism underlying the myelin deficit is decoupling of angiogenesis and myelination. Then we asked whether GPR30 antagonism can strengthen the angio‐myelinogenesis coupling as a therapeutic strategy for treating hypomyelination and associated functional defects caused by chronic hypoxia. To that end, the neonatal NG2^CreERT; GPR30 fl/fl mice and littermate controls were subjected to a hypoxic environment (10% oxygen v/v) from P3 to P10, corresponding to the late gestational period, with tamoxifen from P3 to P6 (Figure 5A). In the P11 hypoxic brains, the density of CD31‐positive blood vessels was obviously increased after GPR30 cKO, displaying increased vascular lumens and branches, as compared with corresponding littermate controls (Figure 5B). Again, we examined the changes in myelination by immunostaining for MBP and CC1. As expected, the MBP‐positive area was significantly increased in hypoxic GPR30 cKO brains, and CC1‐positive OLs were also increased at P11 without changing the densities of NF200‐positive axons and PDGFRα‐positive OPCs (Figure 5C,D). In the P21 GPR30 cKO brains, the MBP‐positive area and Olig2/CC1 double‐positive OLs remained evidently increased, as compared with the controls (Figure 5E,F). Next, we calculated the number of Ranvier nodes by immunostaining for Contactin‐associated protein 1 (Caspr1), a paranodal marker expressed by axons in the cortical regions. The density of paired Caspr‐positive nodes along MBP‐positive sheath was significantly increased in the hypoxic GPR30 cKO brains at P21, compared with age‐matched littermates (Figure 5E). Nodes constitute functional units that mediate conduction velocity. This result indicates potential functional improvement after GPR30 deletion in both pericytes and OPCs. Together, these results suggest that GPR30 deletion in pericytes and OPCs promotes angiogenesis and myelination in response to hypoxia‐induced hypomyelination.

FIGURE 5.

GPR30 deletion improves functional recovery via enhancing myelinogenesis and angiogenesis in neonatal mice. (A) Schematic diagram displaying the time course for hypoxia exposure, tamoxifen induction, histology and behavior tests. (B) Representative images of CD31 (gray, yellow dotted lines indicating corpus callosum)‐positive vessels in P11 (left panels) and P21 (right panels), and quantification of CD31‐positive area, branches of vessel, length and width of vessel in P11 (upper panels) and P21 (lower panels) hypoxic NG2^CreERT; GPR30 fl/fl and littermate controls brains. Scale bar: 50 μm. n = 4 biologically independent mice for each group. (C) Representative images and quantification of MBP (red, yellow dotted lines indicating corpus callosum) and NF200 (green) in hypoxic NG2^CreERT; GPR30 fl/fl and littermate controls brains at P11 (middle and right panels showing images of white dotted boxes in left panels). And quantification of MBP‐positive area and NF200‐positive area in P11 hypoxic NG2^CreERT; GPR30 fl/fl and littermate controls brains. Scale bar: 200 μm (left panels); 50 μm (middle and right panels). n = 4 biologically independent mice for each group. (D) Representative images and quantification of CC1 (green, left panels, dotted lines indicating corpus callosum)‐positive OLs, and PDGFRα (red, middle panels)‐positive OPCs in hypoxic NG2^CreERT; GPR30 fl/fl mice and littermate controls at P11. Scale bar: 50 μm. n = 4 biologically independent mice for each group. (E) Representative images of Caspr (red) and MBP (green) double‐positive nodes in hypoxic NG2^CreERT; GPR30 fl/fl and littermate controls brains at P21, and quantification of MBP‐positive area and density of Nodes (white arrows indicating each node). Scale bar: 10 μm. n = 4 biologically independent mice for each group. (F) Representative images and quantification of CC1 (green, dotted lines indicating corpus callosum) and Olig2 (red) double‐positive OLs (left panels) and PDGFRα (red, middle panels)‐positive OPCs in hypoxic NG2^CreERT; GPR30 fl/fl and littermate controls brains at P21. Scale bar: 50 μm. n = 4 biologically independent mice for each group. (G) Crossing latency and mean number of foot slips of NG2^CreERT; GPR30 fl/fl and littermate controls mice in the beam walking test at P40. n = 8 biologically independent mice for each group. (H) Escape latencies to platform in the acquisition phase, number of crossing platform, proportion of time and distance spent in target sector of NG2^CreERT; GPR30 fl/fl and littermate controls mice in Morris water maze at P40. n = 8 biologically independent mice for each group. Error bars represent mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ns: not significant. The unpaired t test was used to assess the significance between NG2^CreERT; GPR30 fl/fl and littermate controls. Two‐way repeated‐measures analysis of variance (two‐way RM‐ANOVA) was applied for data analysis of the Morris water maze acquisition phase. Dots represent animal numbers.

Chronic hypoxia during the neonatal stage may lead to neurological impairments in adolescent mice (Back 2017; Zeng et al. ²⁰¹⁹). Next, we wanted to further examine whether GPR30 cKO in pericytes and OPCs can improve functional deficits caused by hypoxia at P40. Remarkably, the GPR30 cKO mice showed decreased latency walking along the beam, compared with their littermates, reflecting improved motor coordination (Figure 5G). In MWM, the GPR30 cKO mice exhibited decreased latency during the acquisition stage, with more crossings into the target area, and more time and higher ratio of distance traveled in the target quadrant to total distance in the test phase, compared with age‐matched hypoxic controls (Figure 5H), suggesting improved spatial learning and memory capacity. These findings indicate that conditional knockout of GPR30 can relieve chronic hypoxia‐induced neurological impairments in motor coordination and spatial memory.

3.6. GPR30 Antagonism Promotes Myelin and Functional Recovery After Hypoxia

We then explored whether the G15 treatment could recapitulate the effects of GPR30 deletion in NG2‐positive cells. G15 was administered to the mice during hypoxia exposure from P3 to P10, and brain tissues were collected at P11 (Figure 6A). Similarly, the CD31‐positive vessel density was significantly increased in the G15‐treated corpus callosum and cortex, compared with the vehicle‐treated controls (Figure 6B). Furthermore, the width and length of CD31‐positive vessels, and even branches of the vessels, were increased upon G15 treatment (Figure 6B). Similarly, immunostaining results showed that the MBP‐positive area and the density of CC1‐positive mature OLs were significantly increased in the G15‐treated hypoxic brains (Figure 6C), but the density of PDGFRα‐positive OPCs was not significantly changed (Figure 6D). These findings suggested G15 treatment during chronic hypoxia exposure promoted angiogenesis and myelin development. To examine the functional changes, the G15‐treated mice were subjected to behavioral assessments at P40. In the beam walking test, the G15‐treated mice exhibited fewer foot slips than vehicle‐treated controls (Figure 6E). In the MWM test, the G15‐treated mice spent more time in the targeted quadrant and showed a higher ratio of distance traveled in the target quadrant to total distance, without a significant change in latency during the acquisition stage (Figure 6F). These results compellingly suggest that G15 treatment during hypoxia substantially ameliorates impairments in long‐term motor coordination and spatial memory.

FIGURE 6.

G15 treatment increases angiogenesis and myelinogenesis and promotes functional recovery after hypoxia. (A) Schematic diagram displaying the time course for hypoxia exposure, G15 treatment, histology, and behavior tests. (B) Representative images of CD31‐positive blood vessel (gray, dotted lines indicating corpus callosum) and quantification of CD31‐positive area, branches of vessel, length and width of vessel in vehicle‐ and G15‐treated hypoxic mice. Scale bar: 50 μm. n = 3–4 biologically independent mice for each group. (C) Representative images and quantification of MBP (red, left panels, dotted lines indicating corpus callosum) and CC1 (green, middle panels, dotted lines indicating corpus callosum) in the vehicle‐ and G15‐treated hypoxic mice. Scale bar: 50 μm (left panels); 20 μm (middle panels). n = 3–4 biologically independent mice for each group. (D) Representative images and quantification of PDGFRα (red)‐positive OPCs in the vehicle‐ and G15‐treated hypoxic mice. n = 3–4 biologically independent mice for each group. (E) Crossing latency and mean number of foot slips of vehicle‐ and G15‐treated hypoxic mice in the beam walking test. n = 10–12 biologically independent mice for each group. (F) Escape latencies to platform in the acquisition phase, proportion of time and distance spent in target sector of vehicle‐ and G15‐treated hypoxic mice in Morris water maze. n = 10–12 biologically independent mice for each group. Error bars represent mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns: not significant. The unpaired t test was used to assess the significance between vehicle and G15 groups. Two‐way repeated‐measures analysis of variance (two‐way RM‐ANOVA) was applied for data analysis of the Morris water maze acquisition phase. Dots represent animal numbers.

3.7. GPR30 Antagonism to Promote Remyelination After Demyelination

Myelin repair after demyelination depends on OPC differentiation in the lesion. To understand if antagonizing GPR30 is also effective for remyelination after demyelination, we next used a lysolecithin‐induced demyelination model. Upon injection of lysolecithin into the corpus callosum, myelin breakdown occurs immediately, and OPCs repopulate the lesion and undergo initial differentiation about 10 days post injury (d.p.i.) (Franklin and Ffrench‐Constant ²⁰¹⁷). The lesion can be visualized as a DAPI‐aggregated region. To exclude potential disruptions by sex hormones, male GPR30 cKO and control mice were administered with tamoxifen, then they were stereotactically injected with lysolecithin (Figure 7A). The brains were collected at 14 d.p.i. and subjected to immunostaining for CD31, CC1, and MBP (Figure 7A). The immunostaining results revealed that the density of CD31‐positive vessels was increased in the lesion in the GPR30 cKO mice, compared with the age‐matched controls (Figure 7B). Consistent with that, MBP‐positive area and CC1‐positive mature OLs were increased in the lesion of GPR30 cKO mice (Figure 7B). These results confirmed that GPR30 deletion in pericytes and OPCs promotes angiogenesis and remyelination after lysolecithin‐induced demyelination. To determine whether G15 can replicate the effect of GPR30 deletion in promoting remyelination in adult male mice, NG2^CreERT; Tau‐mGFP mice were induced and stereotactically injected with lysolecithin, followed by continuous G15 treatment (Figure 7C). Only newly formed OLs and associated myelin sheaths can be seen for expressing mGFP in the NG2^CreERT; Tau‐mGFP line, because the mGFP is under the control of the Tau gene, which is solely expressed in mature OLs and associated myelin, with negligible expression in OPCs nor pericytes (Chen, Liu, et al. ²⁰²¹; Chen, Ren, et al. ²⁰²¹; Wang et al. ²⁰²⁰; Zhi et al. ²⁰²³) (Figure 7D,E; Figure S3). The brains treated with G15 for 14 days displayed increased CD31‐positive vessels in demyelinated lesions and, more importantly, the density of mGFP/CC1 double‐positive newly formed OLs and associated myelin sheaths was significantly increased in lesions of G15‐treated brains, compared with vehicle‐treated brains (Figure 7F,G). These findings indicated that antagonizing GPR30 is effective in promoting remyelination after focal demyelination.

FIGURE 7.

Antagonizing GPR30 promotes vascular regrowth and remyelination. (A) Schematic diagram displaying the time course for tamoxifen induction, lysolecithin stereotaxic injection and histology in NG2^CreERT; GPR30 fl/fl mice. (B) Representative images and quantification for CD31 (gray, left panels, yellow dotted lines indicating demyelinating lesion), CC1 (red)‐positive OLs, and MBP (green) expression (middle panels) in NG2^CreERT; GPR30 fl/fl and littermate controls mice. Middle and right panels show enlarged images of white dotted boxes in left panels. Scale bar: 200 μm (left panels); 50 μm (middle and right panels). n = 4–5 biologically independent mice for each group. (C) Schematic diagram display the time course for tamoxifen induction, lysolecithin stereotaxic injection, G15 treatment, and histology in the NG2^CreERT; Tau‐mGFP mice. (D) Genetic strategy for the NG2^CreERT; Tau‐mGFP reporter line showing mGFP expression pattern in the NG2^CreERT; Tau‐mGFP mice. (E) Representative image show mGFP (green) and CC1 (red) double‐positive OLs (arrowhead). Scale bar: 10 μm. (F) Representative images and quantification of CD31 (gray, yellow dotted lines indicating demyelinating lesion) in the vehicle‐ and G15‐treated mice upon demyelination. Middle and right panels show enlarged images of white dotted boxes in left panels. Scale bar: 200 μm (left panels); 50 μm (middle panels). n = 4 biologically independent mice for each group. (G) Representative images and quantification of anti‐mGFP (green, yellow dotted lines indicating demyelinating lesion) and CC1 (red) in the vehicle‐ and G15‐treated mice upon demyelination. Scale bar: 50 μm. n = 4 biologically independent mice for each group. Error bars represent mean ± SD. *p < 0.05, **p < 0.01. The unpaired t test was used to assess the significance between vehicle‐ and G15‐treated mice, as well as between GPR30 cKO and control mice. Dots represent animal numbers.

4. Discussion

Myelinogenesis, depending on OPC differentiation, needs to couple with angiogenesis to obtain sufficient nutrients and oxygen for myelin sheath synthesis. Since GPCRs are the most common drug targets and play important roles in regulating angiogenesis and myelination, identifying novel GPCRs that can concurrently regulate myelination and angiogenesis may yield strategies for myelin‐deficit‐related diseases. Here, we identified that GPR30 is selectively expressed in pericytes and oligodendroglia in the developing mouse and human brains. Conditional deletion of GPR30 in pericytes is sufficient to promote blood vessel growth and myelination. GPR30 antagonism can directly enhance OPC differentiation and wrapping in vitro. As a result, deletion of GPR30 in both pericytes and OPCs significantly increases myelination in vivo, and more importantly, antagonizing GPR30 by genetic approaches or G15 treatment improves myelin repair and functional recovery after hypoxic insults in neonatal mice or lysolecithin‐induced demyelination. These results together demonstrate that antagonizing GPR30 is a potent strategy to synchronize angiogenesis and myelination against myelin‐defect disorders.

4.1. Myelination and Angiogenesis Coupling

Coupling between myelination and angiogenesis has been repeatedly demonstrated in previous studies (Ren et al. 2024; Tsai et al. ²⁰¹⁶; Wang et al. ²⁰²⁴; Xiao and Czopka ²⁰²³; Yuen et al. ²⁰¹⁴). Oligodendroglial cells regulate angiogenesis by expressing pro‐angiogenic factors, including those involved in the Wnt signaling pathway (Yuen et al. 2014). Therefore, the high demand for energy and oxygen during OPC differentiation and myelination can be supported by the blood vessels. The hypoxia‐sensitive transcription factor HIF‐1 regulates these pro‐angiogenic factors (Yuen et al. 2014). Conversely, secreted factors from vascular cells can directly modulate OPC differentiation and myelination. Lama2, secreted by pericytes, and endothelin, a vasoactive peptide expressed by endothelial cells, are required for myelinogenesis in developing or demyelinated brains (De La Fuente et al. 2017; Shibahara et al. ²⁰²⁰; Wang et al. ²⁰²⁴; Zhao et al. ²⁰²²). These previous findings have demonstrated either vessel‐to‐oligodendroglia or oligodendroglia‐to‐vascular crosstalk, which governs myelination or angiogenesis, respectively. Here, we identified GPR30 as a novel signal that can simultaneously regulate myelination and angiogenesis, thereby synchronizing their coupling. Previous studies using cell‐lineage tracing or immunostaining have demonstrated GPR30 expression in blood vessels and oligodendroglia (Hirahara et al. 2013; Wu et al. ²⁰²⁵). Our results show that antagonizing GPR30 can induce OPC differentiation in vitro. This finding is in line with a previous study showing that GPR30 agonism enhances OPC proliferation in vitro. The high OPC cell density resulting from GPR30 agonism may contribute to increased remyelination (Hirahara et al. 2013).

4.2. GPR30 Signaling on Pericytes and Oligodendrocytes

Deletion of GPR30 in PDGFRβ‐positive pericytes contributed to increased angiogenesis in developing brains, indicating that pericytes play a crucial role in mediating blood vessel growth. This finding aligns with previous studies, which have shown that pericytes can promote vascular sprouting and bridging to form new blood vessels (Ren et al. 2024). The enhanced myelination resulting from promoting pericyte‐mediated angiogenesis indicates that blood perfusion is essential for myelinogenesis (Li and Fan 2023; Ren et al. ²⁰²⁴). On the other hand, deletion of GPR30 or antagonism of GPR30 directly promotes OPC differentiation and wrapping in purified OPC cultures. The downstream signals remain unclear from the current dataset. GPCR functions depend on the subunits Gαs, Gαi/o, Gq/11, or Gβγ, and the biased downstream signaling determines versatile physiological processes across different cell types (Lorente et al. 2025). Generally, GPR30 may alter ERK/MAPK, PI3K/Akt, and cAMP/PKA signaling pathways, which regulate myelination and angiogenesis (McDonald et al. ²⁰²⁵; Mogha et al. ²⁰¹⁶; Prossnitz and Barton ²⁰²³). In addition, GPR30 can interact with transcription factor HIF1α in hypoxic conditions, including the tumor microenvironment and ischemic injury (Lappano et al. 2016; Wang, Li, et al. ²⁰¹⁸), suggesting that GPR30 function in vascular cells is involved in oxygen‐related changes.

4.3. GPR30 Antagonism as a Promising Strategy for Myelin Repair

The G‐protein‐coupled receptor (GPCR) superfamily is the target of most clinically used selective drugs so far (Lorente et al. 2025). GPR30, also known as membrane‐associated estrogen receptor 1 (GPER1), mediates rapid non‐genomic signaling transduction by binding to estrogen (Prossnitz and Barton 2023). However, a recent study showed that 17‐α estrogen may indirectly activate GPER1 through structural biology analysis (Liu et al. 2024). Nevertheless, G15 and G1 are highly selective and potent antagonists and agonists, respectively (Bologa et al. 2006; Dennis et al. ²⁰⁰⁹). Our results showed that G15 can promote angiogenesis and myelinogenesis in a hypoxia‐induced white matter injury model and a demyelination mouse model. Though G15 is not an FDA‐approved drug yet, G15 treatment did not display significant adverse effects on the experimental animals. As a proof of concept, GPR30 antagonism is implicated as a potential strategy for myelin repair. Developing novel drugs targeting GPR30 or modification of G15 may yield safe and potent therapeutic drugs for treating myelin‐defect diseases.

Taken together, we identified GPR30, which is concurrently expressed in pericytes and OPCs, and plays crucial roles in angiogenesis and myelination, respectively. Deletion of GPR30 in both pericytes and OPCs or GPR30 antagonism promoted angiogenesis, improved myelin deficits and functional impairments upon hypoxic insults or demyelination, indicating that antagonizing GPR30 is a promising approach to synchronize pericyte‐mediated angiogenesis and OPCs differentiation, contributing to recovery of myelin deficits and associated functional impairments.

Author Contributions

F.M., H.R., and F.W. conceived and supervised this project. S.‐W.H., X.G., B.Y., Q.‐J.L., Z.‐Y.L., S.‐L.W., Y.‐J.C., L.X., and T.L. performed experiments. S.‐W.H., X.G., F.M., H.R., and F.W. analyzed the data and wrote the paper. H.R. and T.C. optimized the text content and figures. All authors read and approved the manuscript.

Funding

This work was supported by Chinese STI‐2030‐Major Projects (2022ZD0207200), the Foundation for Innovative Research Groups of the National Natural Science Foundation of China (32471026, 82271226, W2511025), and the New Chongqing Innovative Talent Program (CSTB2024NSCQ‐QCXMX0045).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: (Related to Figure 1). The expression level of orphan receptors in the GPCR family within the mouse brain. The expression levels of all screened‐out orphan receptors belonging to the GPCR family in oligodendrocyte precursor cells (OPC), oligodendrocytes (OL), pericytes, endothelial cells, astrocytes, microglia and neurons in mouse brain, including Gpr3, Gpr4, Gpr6, Gpr12, Gpr17, Gpr19, Gpr20, Gpr21, Gpr22, Gpr25, Gpr26, Gpr27, Gpr34, Gpr35, Gpr37, Gpr39, Gpr45, Gpr50, Gpr52, Gpr61, Gpr62, Gpr63, Gpr65, Gpr68, Gpr82, Gpr83, Gpr84, Gpr85, Gpr87, Gpr88, Gpr132, Gpr135, Gpr139, Gpr141, Gpr146, Gpr149, Gpr150, Gpr153, Gpr160, Gpr161, Gpr162, Gpr171, Gpr173, Gpr174, Gpr176, Gpr182, Gpr183, Lgr4, Lgr5, Lgr6, Mrgpre, Mrgprf, P2ry10, Ffar1, Kiss1r, Adgrg1, Adgrd1, Hcar2, Adgra2, Adgrg6, Gper1.

Figure S2: (Related to Figure 6). (A) Schematic illustration showing the time course for G15 treatment and histology in wildtype mice. (B) Representative images and quantification of MBP‐positive area (green) and CC1‐positive OLs (red) in the G15‐ or vehicle‐treated mice at P11. Scale bar: 200 μm (left panels); 50 μm (middle and right panels). n = 4 biologically independent mice for each group. (C) Representative images and quantification of PDGFRα‐positive OPCs (red) in the G15‐ or vehicle‐treated mice at P11. Scale bar: 20 μm. n = 4 biologically independent mice for each group. (D) Representative images and quantification of CD31‐positive area (gray, dotted lines indicating corpus callosum) in the G15‐ or vehicle‐treated mice at P11. Scale bar: 100 μm. n = 4 biologically independent mice for each group. Error bars represent mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001. The unpaired t test was used to assess the significance between vehicle‐ and G15‐treated mice. Dots represent animal numbers.

Figure S3: (Related to Figure 7). (A) Schematic illustration showing the time course for tamoxifen induction and histology in NG2^CreERT; Tau‐mGFP mice. (B) Representative images showing the co‐labeling of mGFP (green) with MBP (red, upper panels), CC1 (red, middle panels) and PDGFRα (red, lower panels). Scale bar: 20 μm.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.