Untargeted metabolomics reveals the effect of MYH9 on the metabolism of vascular endothelial cells

Summary

MYH9 plays a crucial role in vascular endothelial cells (VECs) and is a significant cause of abnormal bleeding in MYH9-related diseases (MYH9-RDs). However, research on how MYH9 regulates VECs has been limited so far, and there is a lack of biomarkers to assist in diagnosing MYH9-RD. In this study, we discovered that inhibiting MYH9 (via Blebbistatin, Bleb) or knocking down its gene significantly impairs angiogenesis, vasculogenesis, and VEC functions, including migration and tube formation. Further untargeted metabolomics identified 152 differential metabolites, such as sulfate and glycerophosphocholine, in MYH9-knockdown VECs, enriched in sulfur metabolism and glycerophospholipid metabolism. Importantly, MYH9 overexpression in VECs validated these key metabolites and pathways. These results reveal a close association between metabolites, such as sulfate and MYH9-RD vascular abnormalities, and provide potential diagnostic biomarkers and therapeutic targets for MYH9-RD.

Graphical abstract

Highlights

• •MYH9 regulates angiogenesis, vasculogenesis, and the functions of VECs
• •Untargeted metabolomics identifies differential metabolites in MYH9-knockout VECs
• •MYH9 overexpression validates key metabolites and their associated metabolic pathways
• •The findings offer potential diagnostic biomarkers and therapeutic targets for MYH9-RD

Molecular network; Metabolic flux analysis; Metabolomics

Introduction

MYH9 gene encodes the non-muscle myosin heavy chain IIA (NMMHC IIA), a crucial protein that is essential for various cellular processes, including cell division, adhesion, and migration.^1,2,3 Mutations in the MYH9 gene lead to a group of autosomal dominant genetic disorders collectively referred to as MYH9-RD, which include May-Hegglin anomaly (MHA),⁴ Epstein syndrome (EPS),⁵ Fechtner syndrome (FTS),⁶ and Sebastian platelet syndrome (SPS).⁷ MYH9-RD is typically characterized by spontaneous bleeding as a primary clinical manifestation. This bleeding tendency is generally attributed to platelet defects, including reduced platelet count, increased platelet volume, and impaired clot retraction.^8,9,10 However, the role of vascular endothelial cells (VECs) in this process has often been overlooked.

Recent studies have increasingly shown that MYH9 is strongly expressed in VECs and plays a crucial role in in various biological functions essential for endothelial integrity, such as the regulation of cell adhesion and migration.¹¹ Studies have confirmed that altered MYH9 expression in VECs can lead to abnormal bleeding.^7,12 MYH9-deficient mice exhibit impaired release of von Willebrand factor (VWF) from adrenaline-stimulated VECs, resulting in prolonged bleeding times and defective thrombus formation.¹³ Luo et al. demonstrated that the MYH9 E1841K mutation in VECs disrupts cAMP-induced VWF secretion, worsening the bleeding phenotype in MYH9-RD mouse models.¹² Moreover, MYH9 has been considered a regulator of angiogenesis, and abnormal angiogenesis has been identified as a key contributor to abnormal bleeding. In mice, the abnormal replacement of the MYH9 gene causes embryonic death before placental formation due to vascular development defects, with further studies showing that MYH9 regulates angiogenesis by modulating nucleolin localization.^11,14 These findings suggest that VECs in patients with MYH9-RD undergo pathological changes, potentially contributing to the spontaneous bleeding observed. This highlights the need for further investigation into the role of VECs in the disease mechanism of MYH9-RD.

Although there are few studies on MYH9 regulating cellular metabolism, research has shown that NMMHC IIA plays a key role in the metabolic regulation of VECs, particularly in lipid metabolic reprogramming. Under oxygen-glucose deprivation (OGD) conditions, NMMHC IIA modulates the transformation of major lipid components such as phospholipids, sphingolipids, and neutral lipids in VECs, while knockdown of NMMHC IIA reduces the sensitivity of cells to OGD-induced lipid reprogramming.¹⁵ These findings indicate that MYH9 exerts a specific regulatory role in metabolism of VECs, which is crucial for maintaining cellular function. Currently, metabolic abnormalities in VECs have been observed in various diseases, including cancers, atherosclerosis, and diabetic vascular complications, correlating with corresponding treatment recommendations.^16,17 Despite the extensive attention that has been paid to VEC metabolism in various diseases, the specific regulatory mechanisms by which MYH9 affects metabolic pathways in VECs remain unclear. By analyzing the metabolic changes of VECs, we can not only elucidate key biological processes that may be involved in the pathogenesis of MYH9-RD but also identify key diagnostic biomarkers.

Identifying specific diagnostic biomarkers is crucial for patients with MYH9-RD, as it enables accurate diagnosis and better management. In clinical practice, due to the lack of specific diagnostic markers and clinical features, patients with MYH9-RD are often misdiagnosed and undergo prolonged, unnecessary treatments. For instance, misdiagnosis as autoimmune thrombocytopenia may lead to treatments such as steroid use, intravenous immunoglobulin administration, or even splenectomy for up to 10 years.^18,19,20,21 Fortunately, metabolomics-based approaches are emerging as a new tool for this purpose. Metabolomics allows for the analysis of cellular or tissue metabolites, uncovering disease-specific biomarkers and elucidating pathogenic mechanisms.

Therefore, employing adenovirus-mediated MYH9 gene knockdown or overexpression models in VECs, combined with untargeted metabolomics using ultra performance liquid chromatography-mass spectrometry (UPLC-MS) to detect and analyze differential metabolites, holds significant promise. This approach aims to elucidate the specific mechanisms by which MYH9 regulates metabolism in VECs, offering new insights and theoretical foundations for understanding the pathophysiological processes of MYH9-RD and related vascular diseases. This study not only improves early diagnostic capabilities for MYH9-RD but also contributes to the development of novel therapeutic strategies targeting MYH9 regulatory pathways and their pharmacological mechanisms, ultimately enhancing patient prognosis and quality of life.

Results

Bleb inhibits angiogenesis and vasculogenesis

Based on previous literature studies,^22,23 the concentration of Bleb was set at 5 μM, and CCK-8 assay results also indicated that 5 μM Bleb had no effect on cell viability (Figure 1A). In the mouse thoracic aorta ring assay, the administration of the MYH9 inhibitor Bleb significantly suppressed angiogenesis, as evidenced by a marked reduction in the number of branches (Figure 1B). Similarly, in a hESC-derived vascular organoid model, Bleb markedly inhibited vasculogenesis in vascular organoids (Figure 1C). qPCR analysis in vascular organoids revealed significant downregulation of endothelial-associated genes (PECAM1, VE-Cadherin), pericyte-associated gene (PDGFR-β), and smooth muscle-associated gene (α-SMA), while the expression of the cell proliferation gene Ki67 remained unaffected (Figure 1D). Consistent results were observed via immunofluorescence staining for CD31 and α-SMA in vascular organoids (Figure 1E). Additionally, flow cytometry results showed a significant decrease in the proportion of CD31-positive cells in the vascular organoids following Bleb treatment (Figure 1F). These results indicate that the inhibition of MYH9 activity leads to the significant suppression of angiogenesis and vasculogenesis, with this effect possibly mediated through VECs.

Figure 1.

Bleb inhibits angiogenesis and vasculogenesis

(A) Cell viability of HUVECs treated with different concentrations of Bleb for 24 and 48 h, assessed by CCK-8 assay (n = 4, number of independent experiments). Statistical analysis was performed using two-way ANOVA, ∗∗∗p < 0.001.

(B) Angiogenesis in mouse ex vivo thoracic aorta rings from the control and Bleb-treated (5 μM) groups, with the quantification of branch numbers (n = 4, number of independent experiments). Scale bars, 200 μm. Statistical analysis was performed using Student’s t test, ∗∗∗p < 0.001.

(C) Vasculogenic morphology of vascular organoids in the control and Bleb-treated (5 μM) groups. Scale bars, 200 μm.

(D) Relative mRNA expression of PECAM1, VE-Cadherin, PDGFRB, α-SMA, and Ki67 in vascular organoids from the control and Bleb-treated (5 μM) groups, measured by qRT-PCR (n = 3–4, number of independent experiments). Statistical analysis was performed using Student’s t test, ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001, ns: not significant.

(E) Immunofluorescence staining of DAPI (blue), α-SMA (green), CD31 (red), and their merge in vascular organoids from the Control and Bleb-treated (5 μM) groups, with the quantification of relative fluorescence intensity (n = 3, number of independent experiments). Scale bars, 100 μm. Statistical analysis was performed using Student’s t test, ∗p < 0.05 and ∗∗∗p < 0.001.

(F) Flow cytometry analysis of the proportion of CD31-positive cells in vascular organoids from the Control and Bleb-treated (5 μM) groups, including histograms and statistical results (n = 4, number of independent experiments). Statistical analysis was performed using Student’s t test, ∗∗∗p < 0.001.

MYH9 affects the functions of VECs: migration and tube formation

To further investigate the impact of MYH9 on VECs, we treated human umbilical vein endothelial cells (HUVECs) with a MYH9 inhibitor (Bleb) and a MYH9 knockdown adenovirus, then examined the related functions of HUVECs. Wound healing assay showed that compared with the control group, the Bleb-treated group exhibited a decreasing trend in HUVEC migration rate at 24 h but no significant difference, while the migration rate was significantly reduced at 48 h (Figures 2A and 2B). This result was consistent with that obtained after MYH9 gene knockdown using adenovirus: The MYH9 gene knockdown (MYH9_KD) group had a significantly lower migration rate at 48 h compared with the MYH9 gene knockdown control (MYH9_KDC) group (Figures 2C and 2D). Furthermore, transwell migration assay results indicated that the number of migrated HUVECs in the Bleb-treated group was significantly reduced compared with the control group (Figure 2E), and the number of migrated cells in the MYH9_KD group was also significantly less than that in the MYH9_KDC group (Figure 2F).

Figure 2.

Bleb and MYH9 knockdown adenovirus inhibit HUVECs migration and tube formation

(A) Wound healing assay of HUVECs in control and Bleb-treated (5 μM) groups at 0, 24, and 48 h. Scale bars, 200 μm.

(B) Statistical analysis of migration percentage in the wound healing assay depicted in A (n = 3, number of independent experiments). Statistical analysis was performed using two-way ANOVA, ∗p < 0.05, ns: not significant.

(C) Wound healing assay of HUVECs in MYH9_KDC and MYH9_KD groups at 0, 24, and 48 h. Scale bars, 200 μm.

(D) Statistical analysis of migration percentage in the wound healing assay depicted in C (n = 4, number of independent experiments). Statistical analysis was performed using two-way ANOVA, ∗∗∗p < 0.001, ns: not significant.

(E) Transwell migration assay of HUVECs in control and Bleb-treated (5 μM) groups, with the quantification of migrating cells. Scale bars, 50 μm. Statistical analysis was performed using Student’s t test (n = 3, number of independent experiments), ∗∗∗p < 0.001.

(F) Transwell migration assay of HUVECs in MYH9_KDC and MYH9_KD groups, with the quantification of migrating cells. Scale bars, 50 μm. Statistical analysis was performed using Student’s t test (n = 3, number of independent experiments), ∗∗∗p < 0.001.

(G) Tube formation assay of HUVECs in control and Bleb-treated (5 μM) groups. Scale bars, 200 μm.

(H) Quantitative analysis of isolated segments, master junctions, master segments, and meshes in the tube formation assay depicted in G (n = 3, number of independent experiments). Statistical analysis was performed using Student’s t test, ∗p < 0.05.

(I) Tube formation assay of HUVECs in MYH9_KDC and MYH9_KD groups. Scale bars, 200 μm.

(J) Quantitative analysis of isolated segments, master junctions, master segments, and meshes in the tube formation assay depicted in I (n = 3, number of independent experiments). Statistical analysis was performed using Student’s t test, ∗∗∗p < 0.001, ns: not significant.

Furthermore, tube formation assays showed that the ability of HUVECs to form tubular structures was significantly impaired: This impairment was observed when MYH9 activity was inhibited (5 μM Bleb-treated group vs. control group, Figure 2G) and when MYH9 gene expression was knocked down (MYH9_KD group vs. MYH9_KDC group, Figure 2I). Specifically, analysis using AngioTool software demonstrated a marked reduction in the number of master junctions, master segments, and meshes in HUVECs following two distinct MYH9-targeting treatments. The only difference was that Bleb could increase the number of isolated segments, while MYH9 knockdown adenovirus had no significant effect (Figures 2H and 2J). In addition, we found that neither Bleb nor MYH9 knockdown adenovirus had a significant effect on the apoptosis or proliferation of HUVECs compared with their corresponding control groups (Figures S1B–S1D). These findings suggest that MYH9 significantly influences the functionality of VECs.

Metabolomics analysis

To investigate the influence of MYH9 on metabolic products and pathways in VECs, we initially used adenoviral vectors to knock down MYH9 in HUVECs and analyzed its metabolic products. Subsequently, we validated our findings by overexpressing MYH9 in HUVECs using adenoviral vectors. Here are the specific experimental results.

Analysis and selection of differential metabolites after MYH9 gene knockdown in HUVECs

The scree plot illustrates the contribution rates of eigenvalues for each principal component (PC). As shown in the plot, PC1 contributes 59.7%, and PC2 contributes 17.2% (Figure 3A). We used PCA to evaluate the reproducibility of sample data and differences in the metabolome. The permutation test plot depicts the frequency distribution of model accuracies from 100 permutation tests, with an arrow indicating the position of the observed statistic within the model accuracy distribution. From the results, we observe that p = 0.96 in the observed statistic (Figure 3B), indicating that the PCA model possesses excellent explanatory capabilities. Cross-validation results indicate that the coefficient of determination (R²) is close to 1, and Q² is greater than 0.8, suggesting that the established model can effectively represent true data. The proximity of R² and Q² values further underscores the model’s high predictive power (Figure 3C).

Figure 3.

Analysis and screening of differential metabolites in HUVECs after MYH9 knockdown

(A) Scree plot showing the contribution rates of each principal component’s eigenvalues.

(B) Model validation plot. The x axis represents model accuracy in permutation tests, and the y axis represents the frequency of accuracy across 100 models in permutation tests. An arrow indicates the position of the accuracy of this model.

(C) Cross-validation results using R² and Q² to assess the predictive ability of the model on training and test data.

(D) PCA score plot. PC1 on the x axis represents the first principal component, PC2 on the y axis represents the second principal component, percentages denote the variance explained by each component, and circles indicate 95% confidence intervals.

(E) Volcano plot of metabolites between MYH9_KD and MYH9_KDC groups, with 50 downregulated, 102 upregulated, and 382 non-significant metabolites.

(F) Venn diagram illustrates the overlap of differential metabolites meeting the criteria of VIP >1 and |Log₂FC| ≥ 1 (p < 0.05), identifying 152 significant differential metabolites.

From the PCA scatterplot, it is evident that QC samples cluster tightly, indicating the stability and reproducibility of the detection system. Additionally, in the PCA plot, there is noticeable intra-group clustering and inter-group separation between MYH9_KD group and MYH9_KDC group, suggesting significant metabolic differences in HUVECs after MYH9 knockdown (Figure 3D). Further analysis through volcano plots and variable importance in projection (VIP) scores identified 152 (|Log₂FC|≥1, p < 0.05) and 289 potential differential metabolites (VIP>1) (Figure 3E) (Tables 1, 2, and S3). The VIP values of all 152 potential differential metabolites were observed to be greater than 1 (Figure 3F). To enhance the accuracy of subsequent analyses, we have selected these 152 potential differential metabolites for further study.

Table 1.

Metabolites of MYH9_KD vs. MYH9_KDC, with P-value<0.05 and |Log₂FC| ≥ 1 (sort by P-value, top 30 metabolites)

Compound	FC	log2(FC)	raw.pval	-log10(p)	q value
Glu-Ile	0.30752	−1.7012	8.01E-07	6.0965	6.84E-05
L-Acetylcarnitine (AcCa(2:0))	0.2356	−2.0856	9.00E-07	6.0458	6.84E-05
Succinic acid	2.7627	1.4661	2.51E-06	5.5995	0.000123
L-Proline	0.31523	−1.6655	3.82E-06	5.4177	0.000123
M1X-RT511MZ431	0.41489	−1.2692	4.06E-06	5.3913	0.000123
N-Acetylaspartylglutamic acid	0.22509	−2.1514	5.97E-06	5.2243	0.00013
3beta-Hydroxy-5-cholestenoic acid	0.16355	−2.6122	6.61E-06	5.1795	0.00013
Hydroxybutyrylcarnitine (AcCa(4:0-OH))	0.32593	−1.6174	7.59E-06	5.12	0.00013
Alanine	0.48249	−1.0514	8.21E-06	5.0858	0.00013
Palmitic acid-[13C]16	3.5842	1.8416	8.58E-06	5.0664	0.00013
N-Hexosyl Isoleucine	7.5245	2.9116	1.11E-05	4.9546	0.000141
N-Hexosyl Leucine	7.5245	2.9116	1.11E-05	4.9546	0.000141
Pyroglutamic acid	2.4706	1.3049	1.21E-05	4.9176	0.000141
ADMA	0.28525	−1.8097	1.31E-05	4.8815	0.000143
N-Palmitoyl Serine	2.1197	1.0838	1.73E-05	4.7616	0.000175
N,N,N-trimethyl-5-aminovaleric acid	0.361	−1.4699	3.66E-05	4.4361	0.000322
N-Hexosyl Glutamate	4.1073	2.0382	3.68E-05	4.4344	0.000322
N-Hexosyl Glutamine	64.119	6.0027	3.85E-05	4.4143	0.000322
3-Hydroxy-N6,N6,N6-trimethyl-L-lysine	0.097598	−3.357	4.16E-05	4.3813	0.000322
Palmitic acid (FFA(16:0))	2.0437	1.0312	4.23E-05	4.3732	0.000322
N-Hexosyl Tyrosine	6.1473	2.6199	4.59E-05	4.3383	0.000332
LysoPA(0:0/20:4)	6.9434	2.7956	5.20E-05	4.2841	0.000359
2-Hydroxyglutaric acid	2.5378	1.3436	6.29E-05	4.2012	0.000403
Choline	2.4931	1.3179	6.37E-05	4.196	0.000403
Eicosatrienoic acid (FFA(20:3n6))	5.4807	2.4544	6.76E-05	4.1702	0.00041
Oleoylethanolamide	2.3612	1.2395	7.02E-05	4.1537	0.00041
Stearic acid-D35	3.035	1.6017	7.70E-05	4.1134	0.000434
LysoPE(0:0/22:5n6)	2.7395	1.4539	8.56E-05	4.0676	0.000465
Indole-3-pyruvic acid	2.461	1.2992	9.07E-05	4.0423	0.00047
LysoPI(0:0/20:4)	3.1164	1.6399	9.50E-05	4.0221	0.00047

Table 2.

Metabolites of MYH9_KD vs. MYH9_KDC, with VIP>1(sort by P-value, top 30 metabolites)

Compound	VIP[t]	VIP[ortho-t]
Decaethylene glycol (PEG-10)	1.402895243	0.11744852
Glu-Ile	1.401991101	0.04587863
L-Acetylcarnitine (AcCa (2:0))	1.401871995	0.06859109
2-Chlorophenylalanine	1.400890926	0.00135006
L-Proline	1.398008266	0.03848398
Dodecaethylene glycol (PEG-12)	1.396035882	0.06096702
Alanine	1.395995669	0.0350452
3beta-Hydroxy-5-cholestenoic acid	1.395342349	0.34066186
Hydroxybutyrylcarnitine (AcCa (4:0-OH))	1.394201239	0.09845583
Chloramphenicol-D5	1.393957625	0.20551478
Methyl hexadecanoic acid (FFA (17:0-CH3))	1.393535057	0.14842366
Succinic acid	1.39307799	0.20402434
N-Acetylaspartylglutamic acid	1.392421715	0.31214488
Palmitic acid-[13C]16	1.390649529	0.34651822
Octadecanamide	1.390565071	0.26063241
Dodecenoic acid (FFA (12:1))	1.390229497	0.04908951
Palmitic amide	1.389437595	0.02083037
Pyroglutamic acid	1.386854088	0.09330142
N-Palmitoyl Serine	1.386801809	0.31785156
ADMA	1.384784352	0.20160349
L-Leucine	1.384759729	0.012541
L-Isoleucine	1.384759729	0.012541
Asp-Ile	1.384211364	0.16459965
CDCA-D4	1.383098096	0.04027316
N-Hexosyl Leucine	1.381718169	0.02962467
N-Hexosyl Isoleucine	1.381718169	0.02962467
N, N, N-trimethyl-5-aminovaleric acid	1.380530233	0.17850634
LysoPA (0:0/20:4)	1.377623317	0.45213924
N-Hexosyl Glutamate	1.377205643	0.22054869
3-Hydroxy-N6, N6, N6-trimethyl-L-lysine	1.376508386	0.18981398

Enrichment analysis and pathway analysis of differential metabolites after MYH9 knockdown in HUVECs

To visually present the expression patterns of the 152 significantly differential metabolites between the MYH9_KD and MYH9_KDC groups, we performed heatmap and clustering analysis, and the results demonstrated that these metabolites exhibited distinct expression differences across the two groups (Figure 4A). Subsequently, to further investigate the biological functions associated with these differential metabolites and their enrichment in metabolic pathways, we employed MetaboAnalyst tools to perform enrichment analysis and pathway analysis.

Figure 4.

Pathway analysis and enrichment analysis of differential metabolites in MYH9 knockdown HUVECs

(A) Heatmap of 152 differential metabolites between MYH9_KD groups (KD1–KD4) and MYH9_KDC groups (KDC1–KDC4). The horizontal axis represents the names of differential metabolites, and the vertical axis represents different groups.

(B) Bar chart of enrichment overview (top 25) for differential metabolites. It displays the top 25 significantly enriched biological functions perturbed by MYH9 knockdown. The vertical axis enumerates the biological function terms; the horizontal axis represents the enrichment ratio.

(C) Network view of enriched biological functions. This diagram illustrates the interconnected relationships between the enriched biological processes. Nodes represent individual functional terms, and edges denote their functional associations, revealing the coordinated perturbation of related biological processes.

(D) Bubble plot of metabolic pathway enrichment analysis, showing metabolic pathway enrichment results. The horizontal axis is pathway impact, the vertical axis is -log₁₀(p), and bubbles represent pathways perturbed by MYH9 knockdown.

(E–J) Schematics of key metabolic pathways perturbed by MYH9 knockdown, corresponding to the prominent pathways in D (outside the dashed line, labeled in red). Specifically, E displays the lysine degradation pathway; F displays the biosynthesis of unsaturated fatty acids pathway; G displays the alanine, aspartate, and glutamate metabolism pathway; H displays the glycerophospholipid metabolism pathway; I displays the sulfur metabolism pathway; J displays the taurine and hypotaurine metabolism pathway.

Enrichment analysis revealed that the differential metabolites were primarily enriched in 686 biological functions, including but not limited to: “SLC-mediated transmembrane transport,” “synthesis, secretion, and inactivation of glucagon-like peptide-1 (GLP-1),” “incretin synthesis, secretion, and inactivation,” “free fatty acid receptors,” “RORA activates gene expression,” and “BMAL1: CLOCK, NPAS2 activates circadian gene expression” (Figure 4B). These functions exhibit complex interrelationships (Figure 4C), reflecting the broad impact of MYH9 knockdown on metabolic functions in HUVECs.

To further elucidate the metabolic alterations induced by MYH9 knockdown, we conducted pathway enrichment analysis and found that the differential metabolites were significantly enriched in 28 metabolic pathways (Table 3). Notably, these included “lysine degradation,” “biosynthesis of unsaturated fatty acids,” “alanine, aspartate and glutamate metabolism,” “glycerophospholipid metabolism,” “sulfur metabolism,” and “taurine and hypotaurine metabolism,” among others (Figure 4D). Moreover, the specific metabolites involved in these pathways (highlighted in red with corresponding KEGG Compound IDs) were visualized within their respective metabolic networks (Figures 4E–4J), explicitly demonstrating the roles of these differential metabolites in driving the enrichment of the aforementioned metabolic pathways.

Table 3.

Differential metabolite pathway analysis results of MYH9_KD vs. MYH9_KDC

Pathway Name	Match Status	p	-log(p)	Holm p	FDR	Impact
Lysine degradation	5/30	3.35E-04	3.4748	0.0268	0.0268	0.11451
Biosynthesis of unsaturated fatty acids	5/36	8.08E-04	3.0928	0.0638	0.0323	0
Alanine, aspartate and glutamate metabolism	4/28	0.0025708	2.5899	0.2005	0.0686	0.13462
Glycerophospholipid metabolism	3/36	0.040381	1.3938	1	0.8076	0.15489
Sphingolipid metabolism	2/32	0.15021	0.82329	1	1	0.01688
Sulfur metabolism	1/8	0.16053	0.79445	1	1	0.21277
Taurine and hypotaurine metabolism	1/8	0.16053	0.79445	1	1	0.42857
Fatty acid degradation	2/39	0.20485	0.68857	1	1	0
Pyrimidine metabolism	2/39	0.20485	0.68857	1	1	0.09634
Primary bile acid biosynthesis	2/46	0.26144	0.58262	1	1	0.00758
Fatty acid biosynthesis	2/47	0.26959	0.56929	1	1	0.01473
Butanoate metabolism	1/15	0.28024	0.55247	1	1	0
Nicotinate and nicotinamide metabolism	1/15	0.28024	0.55247	1	1	0.13816
Glycerolipid metabolism	1/16	0.29593	0.52882	1	1	0.04361
Selenocompound metabolism	1/20	0.35542	0.44926	1	1	0
Ether lipid metabolism	1/20	0.35542	0.44926	1	1	0
Pantothenate and CoA biosynthesis	1/20	0.35542	0.44926	1	1	0.0068
Citrate cycle (TCA cycle)	1/20	0.35542	0.44926	1	1	0.03273
Propanoate metabolism	1/22	0.38331	0.41645	1	1	0
Purine metabolism	2/70	0.4515	0.34535	1	1	0
Glutathione metabolism	1/28	0.46013	0.33712	1	1	0.00709
Porphyrin metabolism	1/31	0.49497	0.30542	1	1	0.07453
Glycine, serine and threonine metabolism	1/33	0.51697	0.28653	1	1	0
Cysteine and methionine metabolism	1/33	0.51697	0.28653	1	1	0.02089
Arginine and proline metabolism	1/36	0.54825	0.26103	1	1	0.01744
Fatty acid elongation	1/39	0.57755	0.23841	1	1	0
Tryptophan metabolism	1/41	0.59605	0.22472	1	1	0
Arachidonic acid metabolism	1/44	0.62233	0.20598	1	1	0.27659

HUVEC overexpression of MYH9 validates differential metabolites and metabolic pathways

To further validate the previous analysis, we utilized MYH9 overexpression (MYH9_OE) adenoviruses and MYH9 overexpression control (MYH9_OEC) adenovirus to infect HUVECs, followed by metabolomics profiling. Volcano plot analysis revealed that compared to MYH9 knockdown, MYH9 overexpression did not induce a significant number of differential metabolites (Figure 5A). Therefore, we applied a threshold of p < 0.05, resulting in 52 potential differential metabolites, all of which had VIP values greater than 1 upon further analysis (Figure 5B; Tables 4 and 5).

Figure 5.

Validation of differential metabolites and metabolic pathways by MYH9 overexpression in HUVECs

(A) Volcano plot of metabolites between MYH9_OE and MYH9_OEC groups, with 41 downregulated, 11 upregulated, and 471 non-significant metabolites.

(B) Venn Diagram of differential metabolites between MYH9_OE and MYH9_OEC groups, identifying metabolites that meet the criteria of VIP>1 and p < 0.05, identifying 52 significant differential metabolites.

(C) Petal Venn diagram shows the distribution of differential metabolites among MYH9_KD, MYH9_KDC, MYH9_OE, and MYH9_OEC groups, illustrating the overlap and specificity of metabolic changes induced by MYH9 knockdown and overexpression.

(D) Heatmap of expression levels of 7 key differential metabolites in MYH9_KD and MYH9_KDC groups.

(E) Heatmap of expression levels of 7 key differential metabolites in MYH9_OE and MYH9_OEC groups.

(F) Bar chart of enrichment analysis for 7 key differential metabolites, highlighting the top enriched biological processes and pathways.

(G) Network view of enrichment analysis for 7 key differential metabolites, illustrating the functional connections between the enriched pathways and metabolites.

(H) Bubble plot of pathway analysis for 7 key differential metabolites, with the vertical axis is −log₁₀(p), horizontal axis is pathway impact, and bubble size/color indicating enrichment significance.

Table 4.

Metabolites of MYH9_OE vs. MYH9_OEC, with P-value<0.05 (sort by P-value, top 30 metabolites)

Compound	FC	log2(FC)	raw.pval	-log10(p)
N6, N6, N6-Trimethyl-L-lysine	0.84661	−0.24022	0.0018759	2.7268
Oxoglutaric acid	0.57492	−0.79857	0.0043771	2.3588
Stearidonic acid (FFA (18:4n3))	0.58018	−0.78544	0.0061946	2.208
Methionine sulfoxide	0.71233	−0.48939	0.0064072	2.1933
L-Proline	0.84795	−0.23796	0.0065731	2.1822
Docosadienoate (FFA (22:2))	2.3897	1.2569	0.0066825	2.1751
Taurodeoxycholic acid	0.6774	−0.56192	0.0079919	2.0974
4-Guanidinobutanoic acid	0.74583	−0.42307	0.012627	1.8987
Linoleic acid (FFA (18:2n6))	0.68822	−0.53907	0.012965	1.8872
3beta-Hydroxy-5-cholestenoic acid	0.6127	−0.70676	0.01638	1.7857
Sphingosine 1-phosphate	0.28003	−1.8363	0.016503	1.7824
Sedoheptulose 7-phosphate	1.9302	0.94878	0.016872	1.7728
Homostachydrine	0.34578	−1.5321	0.01809	1.7426
N, N, N-trimethyl-5-aminovaleric acid	0.75292	−0.40942	0.018118	1.7419
Hypoxanthine	3.2968	1.7211	0.01939	1.7124
Deoxycholic acid glycine conjugate	0.75216	−0.41089	0.020656	1.685
15-HETE	1.9642	0.97396	0.024909	1.6036
4-Methyl-2-oxovaleric acid	0.64271	−0.63775	0.024934	1.6032
Isovalerylglycine	0.53656	−0.89818	0.025134	1.5997
Glucuronic acid	0.22317	−2.1638	0.025316	1.5966
Goshuyic acid (FFA (14:2))	0.60099	−0.73459	0.025599	1.5918
L-Acetylcarnitine (AcCa (2:0))	0.79374	−0.33326	0.025794	1.5885
Alpha-Linolenic acid (FFA (18:3n3))	0.66286	−0.59323	0.026449	1.5776
Gamma-Linolenic acid (FFA (18:3n6))	0.66286	−0.59323	0.026449	1.5776
Tetradecenoylcarnitine (AcCa (14:1))	0.6229	−0.68293	0.026532	1.5762
Tryptamine	1.411	0.49674	0.027178	1.5658
Tetracosahexaenoic acid (FFA (24:6))	2.4788	1.3096	0.028988	1.5378
9-HODE	0.5964	−0.74565	0.029134	1.5356
Lithocholic acid glycine conjugate	0.70626	−0.50172	0.029251	1.5339
Leu-Pro	0.21458	−2.2204	0.030275	1.5189

Table 5.

Metabolites of MYH9_OE vs. MYH9_OEC, with VIP>1 (sort by P-value, top 30 metabolites)

Compound	VIP[t]	VIP[ortho-t]
N6, N6, N6-Trimethyl-L-lysine	2.00816425	0.51962914
Oxoglutaric acid	1.95000869	0.26648908
Docosadienoate (FFA (22:2))	1.9029944	0.62846192
Stearidonic acid (FFA (18:4n3))	1.90052644	0.6633154
Methionine sulfoxide	1.89842723	0.52036898
L-Proline	1.88533042	0.38106837
Taurodeoxycholic acid	1.88286943	0.66841011
4-Guanidinobutanoic acid	1.81601864	0.75543491
Linoleic acid (FFA (18:2n6))	1.8023179	0.64998067
Sphingosine 1-phosphate	1.79043055	0.6913896
Sedoheptulose 7-phosphate	1.78740567	0.61422256
Homostachydrine	1.77610679	0.65223994
3beta-Hydroxy-5-cholestenoic acid	1.77034963	0.65856819
Hypoxanthine	1.76044241	0.73873002
N, N, N-trimethyl-5-aminovaleric acid	1.75661806	0.48692726
Deoxycholic acid glycine conjugate	1.75214449	0.76019064
4-Methyl-2-oxovaleric acid	1.71864766	0.72414018
Glucuronic acid	1.71010525	0.66383233
15-HETE	1.70911686	0.87870757
Isovalerylglycine	1.70896739	0.86742839
L-Acetylcarnitine (AcCa (2:0))	1.70089256	0.77030913
Goshuyic acid (FFA (14:2))	1.69991473	0.79627918
Tetracosahexaenoic acid (FFA (24:6))	1.68731916	0.82970056
Tryptamine	1.68649784	0.80235672
Tetradecenoylcarnitine (AcCa (14:1))	1.68605475	0.60661387
Gamma-Linolenic acid (FFA (18:3n6))	1.68416963	0.69467665
Alpha-Linolenic acid (FFA (18:3n3))	1.68416963	0.69467665
Lithocholic acid glycine conjugate	1.67996234	0.85812307
Hexadecadienoic acid (FFA (16:2))	1.67963202	0.81044222
9-HODE	1.67706848	0.88786159

Comparison with MYH9 knockdown data showed that among these 52 differential metabolites, 2 metabolites were downregulated in the MYH9 knockdown group and upregulated in the MYH9 overexpression group, while 5 metabolites were upregulated in the MYH9 knockdown group and downregulated in the MYH9 overexpression group (Figure 5C). We conducted heatmap and clustering analysis for these 7 metabolites, clearly illustrating their expression patterns across the groups (Figures 5D and 5E).

Next, we conducted enrichment analysis and pathway analysis on these 7 metabolites to further elucidate their roles. Enrichment analysis revealed that these metabolites primarily enriched in 109 biological processes, including but not limited to: “defective SLC26A2 causes chondrodysplasias,” “clock-controlled autophagy in bone metabolism,” and “classical antibody-mediated complement activation” (Figures 5F and 5G). Pathway analysis showed that these metabolites were significantly enriched in 6 metabolic pathways, namely: “sulfur metabolism,” “ether lipid metabolism,” “sphingolipid metabolism,” “glycerophospholipid metabolism,” “tryptophan metabolism,” and “purine metabolism” (Figure 5H). Particularly noteworthy were the pathways sulfur metabolism and glycerophospholipid (GPL) metabolism, which aligned with previous findings from the MYH9 knockdown group (Figure 4D). Specific metabolites involved in these pathways included sulfate and glycerophosphocholine.

MYH9 overexpression enhances VECs functions

To verify the effect of MYH9 overexpression on VEC functions, we used MYH9 overexpression and control adenovirus in HUVECs, and detected cell functional changes through wound healing assay, transwell migration assay, and tube formation assay. The results of the wound healing assay showed that at 24 and 48 h, the migration rate of HUVECs in the MYH9_OE group was higher than that in the MYH9_OEC group (Figure 6A), with significant statistical differences (Figure 6B). In the transwell migration assay (Figures 6C and 6D), crystal violet staining revealed that the number of migrating cells in the MYH9_OE group was significantly increased compared with the MYH9_OEC group. The results of the tube formation assay (Figures 6E and 6F) indicated that the capacity of HUVECs in the MYH9_OE group to form tubular structures was significantly enhanced, as evidenced by a marked increase in the number of master segments and meshes compared with the MYH9_OEC group. In summary, MYH9 overexpression can promote the migration and tube formation abilities of HUVECs, further confirming the crucial role of MYH9 in regulating the biological functions of VECs.

Figure 6.

Functional validation of MYH9 overexpression on HUVEC migration and tube formation

(A) Wound healing assay of HUVECs in MYH9_OEC and MYH9_OE groups at 0, 24, and 48 h. Red dashed lines mark the scratch boundaries. Scale bars, 200 μm.

(B) Statistical analysis of migration percentage in the wound healing assay shown in A (n = 4, number of independent experiments). Statistical analysis was performed using two-way ANOVA, ∗p < 0.05, ns: not significant.

(C) Representative images of the transwell migration assay for MYH9_OEC and MYH9_OE groups, stained with crystal violet. Scale bars, 50 μm.

(D) Quantitative analysis of migrating cells in C (n = 3, number of independent experiments). Statistical analysis was performed using Student’s t test, ∗p < 0.05.

(E) Tube formation assay of HUVECs in MYH9_OEC and MYH9_OE groups. Scale bars, 200 μm.

(F) Quantitative analysis of master segments and meshes in the tube formation assay shown in E (n = 3, number of independent experiments). Statistical analysis was performed using Student’s t test, ∗p < 0.05 and ∗∗∗p < 0.001.

Discussion

Previous studies have confirmed the significant role of MYH9 in vascular remodeling and angiogenesis.^24,25 However, research on how MYH9 influences these physiological processes through the regulation of metabolic pathways remains relatively limited. Consistent with previous studies that verified the role of MYH9 in VECs function and angiogenesis,^11,12 our study found that inhibition or knockdown of MYH9 significantly impairs endothelial function, and subsequent metabolomic analysis further identified the changes in key metabolites and related metabolic pathways associated with these potential metabolic disorders. Using untargeted UPLC-MS metabolomic technology, we identified 152 significantly different metabolites in MYH9 knockdown HUVECs, and these metabolites were mapped to 28 distinct metabolic pathways. After validation with the MYH9 overexpression model, we further screened these differential metabolites into 7 key molecules, including indole-3-pyruvate, sulfate, 3-methylglutaric acid, sphingosine-1-phosphate, glycerophosphocholine, tetradecenoic acid (FFA(14:1n9)), and palmitoleic acid (FFA(16:1)). Among these, changes in sulfur metabolism and GPL metabolism were particularly prominent. GSEA reanalysis of the GEO dataset GSE169225²⁶ further validated these findings (Figure S4A), providing transcriptomic evidence to support our metabolomic observation of disrupted metabolic homeostasis.

Sulfur metabolism plays a crucial role in the development of cardiovascular diseases.^27,28 The intracellular sulfur-containing amino acid metabolic network centered on methionine can generate key metabolic intermediates such as homocysteine (Hcy) and cysteine.²⁸ Methionine-derived s-adenosylmethionine (SAM) mediates epigenetic regulation via DNA methylation,^29,30 while cysteine serves as the rate-limiting substrate for glutathione (GSH) synthesis; notably, GSH is a key substance for maintaining endothelial antioxidant capacity.²⁸ Dysregulation of this metabolic network can trigger the production of reactive oxygen species (ROS) via spontaneous oxidation or activation of NADPH oxidase, inhibit the activity of endothelial nitric oxide synthase (eNOS), and disrupt the balance between nitric oxide (NO) and peroxynitrite anion (ONOO⁻). Meanwhile, cysteine metabolic disorders reduce GSH levels, exacerbate endothelial oxidative stress, and thereby promote low-density lipoprotein (LDL) oxidation and atherosclerotic plaque formation.^27,31 Furthermore, hydrogen sulfide (H₂S), as a major component of sulfur metabolism, exerts significant influence on angiogenesis and has been extensively studied in cardiovascular research, oncology, nephrology, and other fields.^30,32,33,34

Our study found a significant upregulation of sulfate in MYH9 knockdown HUVECs, providing new insights into the interaction between MYH9 and sulfur metabolism. We hypothesize that sulfate accumulation may be associated with abnormalities in enzymes such as 3′-phosphoadenosine 5′-phosphosulfate synthase 2 (PAPSS2)—PAPSS2 is a key enzyme that catalyzes the conversion of inorganic sulfate (SO₄²⁻) to 3′-phosphoadenosine 5′-phosphosulfate (PAPS), and PAPS is the only sulfonation substrate in mammals.³⁵ Consistent with this hypothesis, the analysis of the GSE169225 dataset²⁶ confirmed significant upregulation of PAPSS2 and other sulfur metabolism genes (SUOX, SULF2, GLC) (Figure S4B), supporting dysregulated sulfur metabolism in MYH9-deficient cells. Enhanced PAPSS2 expression likely reflects a compensatory response to impaired ATP-dependent PAPS synthesis, driving sulfate accumulation. Since PAPS synthesis depends on ATP, and non-muscle myosin IIA (NMIIA) encoded by MYH9 has ATPase activity,³⁶ functional defects of NMIIA may disrupt intracellular ATP levels, thereby exacerbating sulfate accumulation.³⁷ This may represent another potential mechanism by which MYH9 contributes to the significant upregulation of sulfate and dysregulation of sulfur metabolism in VECs, though further experimental validation is required.

In addition, glycerophosphocholine, as a key metabolic product of GPLs, undergoes significant changes, and these changes can reflect the metabolic state of GPLs. As the core lipids of this pathway, GPLs play a crucial role in maintaining the physicochemical properties and osmotic barrier function of cell membranes.³⁸ In mammalian cell membranes, phosphatidylcholine (PC) and phosphatidylethanolamine (PE) are the most abundant GPLs,³⁹ and their de novo synthesis mainly occurs in the endoplasmic reticulum, relying on high-energy intermediates CDP-choline and CDP-ethanolamine.⁴⁰ PC synthesis depends on the synergistic enzymatic reactions of choline kinase (ChoK), choline-phosphate cytidylyltransferase (CCT), and 1,2-diacylglycerol cholinephosphotransferase (CPT),⁴¹ while PE synthesis requires the involvement of ethanolamine kinase (EK), ethanolamine-phosphate cytidylyltransferase (CET), and 1,2-diacylglycerol ethanolaminephosphotransferase (EPT).⁴² Abnormal expression of MYH9 may affect the functions of the aforementioned enzymes, thereby causing abnormalities in GPL metabolism.

Dysregulation of GPL metabolism contributes to the pathogenesis of cardiovascular diseases: Abnormal PC metabolism affects the levels of very low-density lipoprotein (VLDL) and high-density lipoprotein (HDL), increasing atherosclerotic risk⁴³; oxidized PC induces endothelial dysfunction and plaque instability^44,45; and L-α-glycerophosphocholine is associated with an increased 10-year stroke risk.⁴⁶ Metabolites such as lysophosphatidic acid (LPA) can activate platelets and promote foam cell formation via the LPA5 pathway,^47,48 while lysophosphatidylcholine (LPC) triggers the activation of NF-κB or NLRP3 inflammasome, exacerbating endothelial injury.^49,50 Furthermore, the transcriptomic analysis of the GSE169225 dataset²⁶ revealed the dysregulation of GPL metabolism genes in MYH9-knockout cells, with LPCAT4 showing the most significant upregulation (Figure S4C). As a key enzyme mediating phosphatidylcholine remodeling and preserving membrane fluidity and integrity, LPCAT4 upregulation indicates MYH9 deficiency disrupts GPL synthesis and remodeling, thereby impairing endothelial membrane function and exacerbating vascular pathogenesis. Nevertheless, the specific pathways and mechanisms underlying MYH9 function in this process require further experimental validation.

By linking MYH9 to sulfur and GPL metabolism, this study establishes a novel framework for understanding how MYH9 dysfunction elicits endothelial injury and underlies vascular pathologies in MYH9-RD, and our findings confirm via established models that MYH9 modulates these two key metabolic pathways in VECs, alters critical metabolites including sulfate and glycerophosphocholine, and regulates core VEC functions such as migration and tube formation, thereby directly contributing to the vascular pathological processes of MYH9-RD. This work fills the critical mechanistic gap in MYH9-mediated vascular endothelial metabolic regulation, provides promising potential diagnostic biomarkers for MYH9-RD, and offers essential guidance for subsequent in-depth mechanistic investigations and the development of targeted therapeutic strategies for this disease.

Limitations of the study

This study has several limitations that need to be acknowledged. First, the VECs used were HUVECs, which exhibit anatomical heterogeneity compared with VECs from adult individuals^51,52—this may limit the applicability of the experimental results. Second, ex vivo blood vessels used in vascular ring experiments were derived from male C57BL/6 mice, and this was intended to reduce interference from sex hormones (e.g., estrogen and progesterone) in female mice on vascular endothelial function, ensuring the stability of experimental results.^53,54 Nevertheless, blood vessels from different sexes may inherently exhibit differences. Third, despite reanalysis using public datasets,²⁶ in vivo verification in MYH9-RD pathological models is still lacking. Additionally, our study has not fully elucidated the regulatory mechanisms and relationships among MYH9, VECs, and differential metabolites. In the future, we will construct MYH9-RD mouse models to supplement in vivo metabolomic analysis and pathological validation of vascular lesions, addressing the aforementioned issues.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Xiuli Wang (wangxl01@dmu.edu.cn).

Materials availability

This study did not generate new unique reagents.

Data and code availability

The metabolomic datasets generated during and/or analyzed during the current study are available in the MetaboLights repository, accession number MTBLS12819, and the access link is: https://www.ebi.ac.uk/metabolights/MTBLS12819.

This article does not report original code.

Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Acknowledgments

We thank Prof. Peiyuan Yin from The First Affiliated Hospital of Dalian Medical University for providing the UPLC-MS instrument and supporting the relevant experiments. This study was supported by the National Key R&D Program (2019YFE0117700) and the Scientific Research Fund of the Liaoning Provincial Department of Education (LJ212410161002).

Author contributions

R.Y.: methodology, formal analysis, data curation, validation, writing-original draft, and writing – review and editing. L.G.: methodology, formal analysis, investigation, writing – review and editing. H.S.: methodology and formal analysis. X.Z.: methodology, formal analysis, and validation. M.W.: conceptualization and supervision. X.W.: funding acquisition, conceptualization, supervision, resources, writing-review and editing, and project administration.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Alexa Fluor 647 Mouse anti-Human CD31	BD Pharmingen	Cat#558094; AB_397020
α-SMA Primary Antibody	Sigma	Cat#ABT1487; AB_3094617
Alexa Fluor® 488-conjugated Goat Anti-Rabbit IgG	ZSGB-BIO	Cat#ZF-0511; AB_2864279
Anti-MYH9 antibody	Proteintech	11128-1-AP; AB_2147294
Anti-beta Actin antibody	Abcam	ab8226; AB_306371
HRP-Goat Anti-Mouse Secondary Antibody	Proteintech	SA00001-1; AB_2722565
HRP-Goat Anti-Rabbit Secondary Antibody	Proteintech	SA00001-2; AB_2722564
Bacterial and virus strains
MYH9-targeting shRNA Adenovirus (MYH9_KD)	Vigene Biosciences	N/A
Non-targeting shRNA Adenovirus (MYH9_KDC)	Vigene Biosciences	N/A
MYH9 Overexpression Adenovirus (MYH9_OE)	Vigene Biosciences	N/A
Empty Vector Adenovirus (MYH9_OEC)	Vigene Biosciences	N/A
Chemicals, peptides, and recombinant proteins
Blebbistatin (Bleb)	MCE	Cat#HY-13441
Endothelial Cell Medium	ScienCell (via Yuhengfeng Co)	Cat##1001
DMEM/F12	Gibco	Cat#12634010
Fetal Bovine Serum	ScienCell (via Yuhengfeng Co)	Cat#0025
Endothelial Cell Growth Supplement	ScienCell (via Yuhengfeng Co)	Cat#1052
Penicillin/Streptomycin	ScienCell (via Yuhengfeng Co)	Cat#0503
mTeSR™ Plus Basal Medium	STEMCELL	Cat#100-0274
mTeSR™ Plus 5 × Supplement	STEMCELL	Cat#100-0275
Matrigel	Corning	Cat#354230
BMP4	Chamot Bio	Cat#CM065-20HP
CHIR99021	APExBIO	Cat#A3001
VEGFA	Proteintech	Cat#Ag13500
Forskolin	Sigma	Cat#66575-29-9
StemPro®-34 SFM	Gibco	Cat#10639011
bFGF	Peprotech	Cat#100-18B-100UG
GlutamMAX	Gibco	Cat#35050061
B27	Gibco	Cat#17504044
N2 supplement	Gibco	Cat#17502048
β-mercaptoethanol	Sigma	Cat#60-24-2
CCK-8	APExBIO	Cat#K1018
Trizol Reagent	ACCURATE	Cat#AG21101
Reverse Transcription Kit	ACCURATE	Cat#AG11603
SYBR Green Mix	ACCURATE	Cat#AG11701
Annexin V-FITC/PI Apoptosis Detection Kit	Beyotime	Cat#C1062S
RNase A/PI Solution	Beyotime	Cat#C1052
DAPI	Beyotime	Cat#C1002
Crystal violet	Beyotime	Cat#Y268090
LC/MS-grade Methanol	Thermo Fisher Scientific	Cat#67-56-1
LC/MS-grade Acetonitrile	Thermo Fisher Scientific	Cat#75-05-8
LC/MS-grade Formic Acid	Thermo Fisher Scientific	Cat#28905
LC/MS-grade Ammonium Bicarbonate	LiChropur	Cat#1066-33-7
Ultra-pure water (18.2 MΩ·cm)	Milli-Q water purification system from Merck KGaA	N/A
Deposited data
Metabolomic data	MetaboLights	https://www.ebi.ac.uk/metabolights/reviewer846c81b9-3f01-4d3f-a168-f9ab3d8cf0ec
Experimental models: Cell lines
HUVECs	ScienCell (via Yuhengfeng Co., BJ, China)	Cat#8000
Human Embryonic Stem Cells (hESCs, H9)	WiCell	Cat#WA09
C57BL/6	Dalian Medical University	N/A
Oligonucleotides
Primer sequences	Sangon Biotech	See Table S1 Primer Sequences for q-PCR
Software and algorithms
ImageJ	National Institutes of Health (NIH)	N/A
AngioTool	Plugin in ImageJ software	N/A
ModFit LT	Verity Software House	N/A
FlowJo_V10	BD Life Sciences	N/A
MetaboAnalyst 6.0	McGill	https://www.metaboanalyst.ca
GraphPad Prism 10.0	GraphPad Software	N/A
TECAN Spark Microplate Reader System	TECAN	N/A
Leica SP8 Laser Confocal Microscope System	Leica	N/A
Ultimate™ 3000 UPLC System and Q Exactive™ Quadrupole-Orbitrap MS System	Thermo Scientific	N/A

Experimental model and study participant details

Cells

HUVECs were obtained from ScienCell Research Laboratories (Cat#8000) We performed immunofluorescence staining to identify HUVECs, and they highly expressed CD31 (Figure S1A). HUVECs were maintained in ScienCell-specific Endothelial Cell Medium (ECM, Cat#1001). This medium consists of 500 ml basal medium, 25 ml fetal bovine serum (FBS; ScienCell, Cat#0025), 5 ml endothelial cell growth supplement (ECGS; ScienCell, Cat#1052), and 5 ml penicillin/streptomycin solution (P/S; ScienCell, Cat#0503). All the aforementioned cells and reagents were purchased from Yuhengfeng Co, BJ, China.

Human embryonic stem cells (hESCs) (H9) (WiCell, Cat#WA09) were maintained in mTeSR™ Plus Basal Medium (STEMCELL, Cat#1100-0274) supplemented with 5 × Supplement (STEMCELL, Cat#100-0275) on matrigel (Corning Cat#354234)-coated 6-well plates. Medium was replaced every two days, cells subcultured at 70–80% confluence.

All cells were routinely tested for mycoplasma contamination and confirmed negative.

Animals

Male wild-type C57BL/6 mice (8–10 weeks of age) were purchased from the Laboratory Animal Center of Dalian Medical University (Dalian, China) and housed under standard specific-pathogen-free (SPF) conditions at 23°C ± 1°C with 55% relative humidity and a 12 hours light/dark cycle, with ad libitum access to standard rodent chow and filtered water. All animal experiments were approved by the Animal Ethics Committee of Dalian Medical University (Approval No. AEE24027) and performed in accordance with institutional animal care guidelines; all efforts were made to minimize animal pain and distress, and mice were humanely euthanized via CO₂ inhalation at the end of the study. As only male mice were used to avoid confounding effects of sex hormones, the present findings are specific to male mice, and sex-related differences will be investigated in future work.

Method details

CCK-8 assay

To assess the effect of Bleb (MCE, Cat#HY-13441,) on cell viability, HUVECs were seeded in 96-well plates at a density of 1 × 10⁴ cells per well and cultured until reaching ∼70% confluence. After cell attachment, the medium was refreshed, and Bleb was administered at concentrations of 0, 2.5, 5, 10, 20, 40, and 80 μM. The cells were then incubated for 24 and 48 hours, respectively. At each time point, CCK-8 reagent was added following the manufacturer’s instructions, and the plates were incubated for an additional 2 hours. Absorbance at 450 nm was measured using a microplate reader (TECAN Spark). Cell viability was calculated relative to the control group (0 μM Bleb), and the experiment was independently repeated three times.

Aortic ring sprouting assay in mice

Based on previous experimental results and animal ethics requirements, four male mice were selected, and their thoracic aortas were isolated for the experiment. First, mice were euthanized using CO₂ and immersed in 75% ethanol for disinfection. The thoracic aorta was then dissected out and placed in sterile PBS. Subsequently, the vessels were cut into 1-2 mm vascular ring and placed in a 24-well plate pre-coated with matrigel. 200 μL of matrigel was added and allowed to solidify at 37°C for 10 minutes, forming a matrigel-vascular ring-matrigel structure. Next, 600 μL of complete ECM was added and the plates were incubated at 37°C with 5% CO₂ for 4 days, with medium changed every two days. Finally, vascular sprouting was observed and analyzed. The experimental group received 5 μM Bleb throughout the entire process, while the control group received an equivalent volume of vehicle.

Vascular organoid experiment

Referring to the reported method,^55,56 we constructed human vascular organoids by first culturing hESCs in suspension to generate embryoid bodies (EBs). For mesodermal differentiation, EBs were transferred to an induction medium consisting of equal volumes of neurobasal and DMEM/F12 (Gibco, Cat#12634010), supplemented with 0.5% GlutaMAX (Cat#35050061), 2% B27 (Cat#17504044), 1% N2 supplement(Cat#17502048), 1% P/S (ScienCell, Cat#0503) (all Gibco); 0.143 mM β-mercaptoethanol (Sigma, Cat#60-24-2); 30 ng/mL BMP4 (Chamot Bio, Cat#CM065-20HP); and 12 μM CHIR99021 (APExBIO, Cat#A3001). After 3 days, vascular lineage specification was initiated using the same base medium (equal parts Neurobasal/DMEM/F12) with the above shared supplements, plus 100 ng/mL VEGFA (Proteintech, Cat#Ag13500) and 2 μM Forskolin (Sigma, Cat#66575-29-9). On day 7, the medium was switched to StemPro®-34 SFM (vascular induction medium, Gibco, Cat#10639011) supplemented with 1% GlutaMAX, 1% P/S (both Gibco); 15% FBS; 100 ng/mL bFGF (Peprotech, Cat#100-18B-100UG); and 100 ng/mL VEGFA. Organoids were embedded in a 3:1 (v/v) collagen/matrigel hydrogel and cultured in this vascular induction medium, with medium changes every 3 days. Mature hVOs were obtained after 5 days for subsequent use.

Starting from the mesodermal differentiation stage and continuing until the completion of the vascular organoid induction experiment, the culture medium of the experimental group was supplemented with 5 μM Bleb, while the culture medium of the control group was supplemented with an equivalent volume of vehicle. Fresh medium was replaced daily during this period.

Throughout the induction process, the morphology of vascular organoid was observed and documented. After collecting vascular organoid samples, the proportion of CD31-positive cells was analyzed using flow cytometry, and q-PCR technology and immunofluorescence methods were employed to assess the expression levels of relevant genes.

q-PCR

q-PCR was used to detect mRNA expression of target genes in vascular organoids. For sample preparation, cells from each group were collected; total RNA was extracted with trizol (ACCURATE, Cat#AG21101), with purity checked by nanodrop and integrity verified via agarose gel electrophoresis. cDNA was synthesized from total RNA using a reverse transcription kit (ACCURATE, Cat#11603). Amplification used specific primers (Sangon Biotech) with GAPDH as the reference gene. Primer sequences see Table S1. Reaction mix (per well): 10 μL 2 × SYBR Green Mix, 0.4 μL each primer, 2 μL cDNA, 7.2 μL nuclease-free water. Conditions: 95°C pre-denaturation for 5 minutes; 40 cycles of 95°C denaturation (10 seconds) and 60°C annealing/extension (30 seconds); melting curve analysis for specificity. Relative expression was calculated via the 2^-ΔΔCt method.

Knockdown or overexpression of MYH9 gene in HUVECs

Based on experimental requirements, HUVECs were divided into 4 groups:

• a)MYH9_KD group treated with recombinant adenoviruses containing MYH9-targeting shRNA sequence (5'-GCAAGCTGCCGATAAGTATCTTTCAAGAGAAGATACTTATCGGCAGCTTGCTTTTTT-3').
• b)MYH9_KDC group treated with recombinant adenoviruses containing non-targeting shRNA sequence (5'-TTCTCCGAACGTGTCACGTTTCAAGAGAACGTGACACGTTCGGAGAATTTTTT-3'), using the same vector as MYH9_KD group but with non-targeting shRNA.
• c)MYH9_OE group treated with adenoviruses harboring full-length human MYH9 cDNA (corresponding gene ID: NM_002473.6) via AdMax packaging system for vector recombination into adenoviruses.
• d)MYH9_OEC group treated with empty vector adenoviruses, sharing the same vector backbone as MYH9_OE group but without MYH9 cDNA. The MYH9 adenovirus was obtained from Vigene Biosciences.

Specific procedures: 3.5 × 10⁵ HUVECs were seeded into 100 mm culture dishes, and 8 mL of ECM (containing 10% fetal bovine serum, 5 ng/mL endothelial cell growth supplement, 100 U/mL penicillin, and 100 μg/mL streptomycin) was added. Cells were cultured in a 37°C incubator with 5% CO₂ and 95% humidity for 12 hours to allow adherence. The original medium was discarded, and 8 mL of medium containing the corresponding adenovirus (multiplicity of infection, MOI=150) was added. After 12 hours of infection, the virus-containing medium was removed and replaced with fresh medium, followed by continuous culture for 72 hours (with fresh medium changed every 24 hours during this period).

Western blot verification: After 72 hours of culture, total proteins were extracted from each group using RIPA lysis buffer (supplemented with protease inhibitor). Equal amounts of protein were separated by SDS-PAGE, transferred to PVDF membranes, and probed with anti-MYH9 antibody (Proteintech, 11128-1-AP, diluted 1:10000) and anti-β-actin antibody (Abcam, ab8226, diluted 1:5000) overnight at 4°C. After incubation with HRP-conjugated secondary antibodies, protein bands were visualized by ECL. Relative MYH9 protein density was quantified using ImageJ, normalized to β-actin.

Flow cytometry

Flow cytometry was used to analyze the proportion of CD31-positive cells in vascular organoids, as well as the cell cycle distribution and apoptosis of HUVECs.

• a)Proportion of CD31-positive cells: Single-cell suspensions of vascular organoids were prepared, washed twice with pre-cooled PBS, and incubated with Alexa Fluor 647 Mouse anti-Human CD31 antibody (BD Pharmingen, Cat# 558094, 1:100) in the dark for 30 minutes. After three washes with PBS, samples were analyzed by flow cytometry.
• b)HUVEC cell cycle analysis: HUVECs (control, Bleb-treated, MYH9_KD and MYH9_KDC) were trypsinized, centrifuged at 1000 × g for 5 minutes, and washed twice with pre-cooled PBS. Cells were then fixed with 70% ice-cold ethanol at 4°C overnight. Fixed cells were washed, stained with RNase A/PI solution at 37°C in the dark for 30 minutes, and analyzed by flow cytometry. The proportions of cells in G0/G1, S, and G2/M phases were determined using ModFit LT software.
• c)HUVEC apoptosis analysis: Annexin V-FITC/PI double-staining was performed. HUVECs (Control, Bleb-treated) were washed with PBS, resuspended in Binding Buffer, and incubated with Annexin V-FITC (1:50) and PI (1:100) in the dark for 15 minutes. Samples were analyzed by flow cytometry within 1 hour. Cells were categorized as live (Annexin V^-/PI^-), early apoptotic (Annexin V⁺/PI^-), late apoptotic (Annexin V⁺/PI⁺), or dead (Annexin V^-/PI⁺); the total apoptosis rate was calculated using FlowJo_V10 software.

Immunofluorescence staining

Immunofluorescence staining was performed to observe the expression and localization of CD31 and α-SMA in vascular organoids (whole-mount immunostaining) and to identify HUVECs.

• a)Whole-mount immunostaining of vascular organoids: Mature vascular organoids were collected, fixed with 4% paraformaldehyde at room temperature for 30 minutes, and washed 3 times with PBS. Following permeabilization with 0.3% Triton X-100 at room temperature for 4 hours, the organoids were washed 3 times with PBS and then blocked with 5% BSA at room temperature for 1 hour. They were incubated with Alexa Fluor® 647-conjugated CD31 antibody (BD Biosciences, Cat#558094, 1:200) and α-SMA primary antibody (Sigma, Cat#ABT1487, 1:200) at 4°C in the dark for 16–20 hours. After washing twice with PBS, the organoids were incubated with Alexa Fluor® 488-conjugated Goat Anti-Rabbit IgG antibody (ZSGB-BIO, Cat#ZF-0511, 1:500) at room temperature in the dark for 1 hour. Following another PBS wash, DAPI solution was added, and samples were incubated at room temperature in the dark for 2 hours. Finally, organoids were mounted with anti-fluorescence quenching medium, observed, and imaged using a Leica SP8 laser confocal microscope. Fluorescence intensity was quantified with ImageJ software, with three replicates per group.
• b)Identification of HUVECs: HUVECs were seeded on coverslips. After adherence, they were fixed with 4% paraformaldehyde at room temperature for 20 minutes and washed 3 times with PBS. Cells were permeabilized with 0.3% Triton X-100 for 15 minutes, washed with PBS, and then blocked with 5% BSA at room temperature for 1 hour. HUVECs were incubated with Alexa Fluor® 647-conjugated CD31 antibody (1:200) at 4°C overnight. The next day, after washing 3 times with PBS, DAPI solution was added, and cells were incubated at room temperature in the dark for 10 minutes. Coverslips were mounted with anti-fluorescence quenching medium, then observed and imaged using a laser confocal microscope.

HUVEC wound healing assay

HUVECs (MYH9_KD, MYH9_KDC, MYH9_OE, MYH9_OEC, Bleb-treated, and normal control groups) were seeded in 6-well plates at 5 × 10⁴ cells/well, cultured in ECM with 10% FBS at 37°C, 5% CO₂ until 90% confluence. Three parallel scratches were made vertically with a sterile 200 μL pipette tip; cells were gently washed twice with PBS to remove detached cells. Serum-free ECM was added with 5 μM Bleb supplemented in the Bleb-treated group, and cultures were continued. Images of the same fields were captured at 0, 24, and 48 hours under an inverted phase-contrast microscope. Scratch width was measured by ImageJ, and healing rate was calculated as: (initial scratch area - residual area at each time point)/initial area × 100%.

HUVEC transwell migration assay

HUVECs treated with MYH9 adenoviruses for 72 hours and normal HUVECs were resuspended in serum-free ECM (5 μM Bleb for the Bleb-treated group) to 5 × 10⁴ cells/mL. 200 μL cell suspension was added to the upper chamber of a transwell insert (pore size: 8 μm), and 600 μL ECM with 10% FBS was added to the lower chamber (5 μM Bleb for the Bleb-treated group). After 24 hours incubation at 37°C, 5% CO₂, cells were washed with PBS, fixed with 4% paraformaldehyde for 15 minutes, and stained with 0.1% crystal violet for 10 minutes. Non-migrated cells in the upper chamber were wiped off, and migrated cells were imaged and counted by ImageJ.

HUVEC tube formation assay

HUVECs within passage 3–6 (normal or pre-treated with MYH9 adenoviruses for 72 hours) were used. Matrigel was thawed on ice, 100 μL added to each 24-well plate well and solidified at 37°C with 5% CO₂ for 10 minutes. Then, 3 × 10⁴ HUVECs were evenly seeded into matrigel-coated wells. Each well received 600 μL complete ECM (with 10% FBS, ECGS, and P/S). For the Bleb-treated group, ECM containing 5 μM Bleb was used throughout the entire process. Cells were incubated at 37°C with 5% CO₂ for 7 hours. Tube morphology was observed and imaged via an inverted phase-contrast microscope. Vascular networks were analyzed using AngioTool software; parameters quantified included isolated segments, master junctions, master segments, and meshes.

Metabolomics sample preparation

After cell confluency, discard the culture medium and wash cells three times with PBS to remove residual medium. Next, add 100 μL of deionized water, scrape cells using a cell scraper, and take out 10 μL of the sample for protein concentration measurement. Subsequently, add 1 ml of 80% methanol to the sample three times for extraction. Combine the extracts, centrifuge at 12000 rpm for 15 minutes at 4°C and collect the supernatant as the sample for subsequent UPLC-MS analysis.

Separation conditions of ultra performance liquid chromatography

The hydrophilic fraction of the metabolite extracts was submitted for untargeted metabolomics analysis using two different analytical methods on an Ultimate™ 3000 UPLC system coupled with a Q Exactive™ quadrupole-Orbitrap mass spectrometer (Thermo Scientific, San Jose, CA, USA). For Methods 1 and 2, the metabolite extracts were analyzed using reverse phase chromatographic separation with positive ionization for Method 1 and negative ionization for Method 2. Metabolites were separated using an ACE C18-PFP column (Advanced Chromatography Technologies Ltd, Aberdeen, Scotland) in Method 1. Elution was performed with a linear gradient of 0.1% formic acid in water and acetonitrile, increasing from 2% to 98% organic phase over 10 minutes. For Method 2, metabolites were separated on an Acquity™ HSS C18 column (Waters Corporation, Milford, USA; 1.8μm, 2.1 × 100 mm) with mobile phases of water and acetonitrile/methanol, both containing ammonium bicarbonate buffer salt. The gradient for Method 2 was 2% organic phase ramped to 100% in 10 minutes, followed by 5 minutes for column washing and equilibration. Conditions were consistent across methods with a flow rate of 0.4 mL/min, an injection volume of 5 μL, and a column temperature of 50°C. For reagents and materials: Optima™ LC/MS-grade methanol, acetonitrile, and formic acid were purchased from Fisher Scientific (Fair Lawn, NJ, USA). LC-MS-grade ammonium bicarbonate was obtained from LiChropur. Ultra-pure water (18.2 MΩ·cm) was purified using a Milli-Q water purification system from Merck KGaA (Darmstadt, Germany).

Separation conditions of mass spectrometry

For both Method 1 and Method 2, the quadrupole-Orbitrap mass spectrometer was operated with identical ionization parameters using a heated electrospray ionization source, except for the ionization voltage. Parameters included sheath gas at 45 arb, auxiliary gas at 10 arb, heater temperature at 355°C, capillary temperature at 320°C, and S-Lens RF level at 55%. Metabolome extracts were analyzed in full scan mode with a resolution of 70,000 FWHM, AGC target of 1E6, and a maximum injection time of 200 ms, covering a scan range of 70-1000 m/z. QC samples were injected repeatedly to acquire Top 10 data-dependent MS2 spectra (full scan-ddMS2) for detailed metabolite and lipid structural annotation. Full MS/MS data were acquired with a resolution of 17,500 FWHM, using apex trigger, dynamic exclusion, and isotope exclusion. The precursor isolation window was set to 1.0 Da. Stepped normalized collision energy was applied for collision-induced dissociation of metabolites using ultra-pure nitrogen as the fragmentation gas. All data were acquired in profile mode.

Multivariate statistical analysis and identification of potential differential metabolites

UPLC-MS data were standardized and subjected to multivariate statistical analysis using MetaboAnalyst 6.0, including principal component analysis (PCA) and Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA). PCA was employed to discern differences among groups, and the OPLS-DA model was validated through 100 permutation tests. Based on VIP scores, fold change (FC), and p-values obtained from the OPLS-DA model, the Benjamini-Hochberg (BH) method was strictly applied for multiple test correction to calculate the false discovery rate (FDR, q-value). The final criteria for screening significant differential metabolites were VIP > 1, p < 0.05, q-value (BH-FDR) < 0.05, and |log₂FC| ≥ 1 (or meeting VIP > 1, p < 0.05, and q-value < 0.05). This approach aimed to uncover metabolic differences between different experimental groups. We calibrated the proteomics data using the protein concentrations of each sample (Table S2).

Enrichment analysis and pathway analysis

Using MetaboAnalyst 6.0, differential metabolites were subjected to Enrichment and Pathway Analysis to identify their significant enrichment in biological functions and metabolic pathways, and to explore their roles and positions within these pathways. Analysis results will be presented in the form of bar charts, network views, and scatter plots in the metabolome view.

Quantification and statistical analysis

Experimental results are expressed as mean ± standard error of the mean (SEM), with the number of independent experimental replicates (n) and sample size per group specified in the corresponding figure legends. Prior to statistical analysis, the normality of data distribution was verified using the Shapiro-Wilk test. For comparisons between two independent groups, unpaired Student's t-test was applied; for single-factor multiple group comparisons, one-way analysis of variance (one-way ANOVA) was performed, followed by Tukey’s post hoc test for pairwise comparisons; for multiple group comparisons, two-way ANOVA was performed first, followed by Tukey’s post hoc test for pairwise comparisons to correct for multiple testing. Statistical analyses were conducted using GraphPad Prism 10.0. A p-value of less than 0.05 was considered statistically significant, and significance levels were marked as ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001 in figures for clarity.

Published: March 18, 2026