Melatonin attenuates splanchnic hyperdynamic circulation in cirrhotic portal hypertension by disrupting POMK-VEGFR2 interaction

Abstract

BACKGROUND

Splanchnic hyperdynamic circulation is a major extra-hepatic hemodynamic alteration contributing to complications in cirrhotic portal hypertension (PHT). While melatonin (MLT) exhibits potent anti-fibrotic properties, the specific impact on splanchnic hemodynamics remains unexplored.

AIM

To investigate the protective effects and underlying molecular mechanisms of MLT on splanchnic hyperdynamic circulation in a cirrhotic PHT model.

METHODS

Cirrhotic PHT was induced in Sprague-Dawley rats via common bile duct ligation. The rats were randomly assigned to MLT or vehicle treatment for four weeks. Hemodynamic parameters, mesenteric angiogenesis, and vascular remodeling were subsequently evaluated. Transcriptomic alterations in mesenteric arteries were assessed using bulk RNA sequencing. Mechanistic pathways were further validated in vivo using the VEGFR-2 inhibitor AZD2171, and in vitro using human umbilical vein endothelial cells and immunoprecipitation (IP)-mass spectrometry.

RESULTS

MLT treatment significantly reduced portal pressure and ameliorated splanchnic hyperdynamic circulation, alongside marked improvements in mesenteric angiogenesis and vascular remodeling. RNA-sequencing of mesenteric arteries identified significant inhibition of the VEGF-VEGFR signaling pathway following MLT administration. In vivo VEGFR-2 inhibition with AZD2171 recapitulated these protective effects, effectively mitigating PHT and mesenteric angiogenesis. In vitro assays confirmed the anti-angiogenic properties of both MLT and AZD2171 in human umbilical vein endothelial cells. Mechanistically, IP-mass spectrometry identified POMK as a potential VEGFR-2-associated protein, which was subsequently validated by IP assays. MLT disrupted the POMK-VEGFR-2 interaction, which may contribute to its anti-angiogenic effects.

CONCLUSION

MLT attenuates splanchnic hyperdynamic circulation and mesenteric angiogenesis in cirrhotic PHT by inhibiting VEGFR-2 signaling. This protective effect is driven by the targeted disruption of the POMK-VEGFR-2 interaction, highlighting a novel protective axis for the management of PHT.

Key Words: Portal hypertension; Splanchnic hyperdynamic circulation; Melatonin; VEGFR-2; POMK; Liver cirrhosis

Core Tip: This study uncovers a novel protective role for melatonin (MLT) in alleviating splanchnic hyperdynamic circulation in cirrhotic portal hypertension. Using in vivo and in vitro models, we demonstrate that MLT effectively reduces portal pressure by inhibiting pathological mesenteric angiogenesis. Mechanistically, our multi-omics and functional analyses reveal for the first time that MLT exerts its anti-angiogenic effects by specifically disrupting the physical interaction between POMK and VEGFR-2, thereby dampening the VEGF signaling cascade. These findings highlight the POMK-VEGFR-2 axis as a critical pathophysiological driver and propose MLT as a promising, translatable candidate for managing cirrhotic hemodynamic complications.

INTRODUCTION

Portal hypertension (PHT) is a syndrome featured by elevated portal pressure (PP) especially occurring at the end stage of chronic liver disease[1]. During the development of liver cirrhosis in chronic liver disease, the elevation of PP and increased intrahepatic vascular resistance (IHVR) are accompanied by splanchnic hemodynamic and structural abnormalities including increased blood flow, decreased vascular resistance, positive vascular remodeling, and enhanced angiogenesis[2]. Generally, the changes in visceral vessels are identified as splanchnic hyperdynamic circulation, leading to a vicious cycle increasing portal blood flow against high IHVR, which further promotes the progression of PHT[3]. Moreover, the progressive formation of portal-systemic collateral vessels driven by angiogenesis will finally result in gastroesophageal variceal bleeding as a major contributor to mortality of PHT[1].

When cirrhotic PHT was initiated by increased IHVR due to advanced liver fibrosis and structural distortion, the congestion of the portal vein system disrupted the homeostasis of vasoconstrictors and vasodilators in visceral vessels[4]. The response and sensitivity of visceral vessels to vasoactive agents were also altered. As one of the main vasodilatory substances produced by endothelial cells, nitric oxide (NO) was regarded as the major factor associated with extrahepatic vasodilation in PHT. NO synthesized and released from visceral vascular endothelium was increased by upregulated activating phosphorylation of endothelial NO synthase[5]. Besides functional changes, visceral vessels were also undergoing structural remodeling under hyperdynamic circulation. Enhanced mesenteric angiogenesis and positive remodeling also synergistically drive the splanchnic hyperdynamic circulation, which is a potential protective target of PHT[4,6]. Previous studies have proved the protective effects of anti-angiogenic therapy and anti-hypertensive medications against extrahepatic hemodynamic alterations and elevated PP[7,8]. However, there are still limitations and contraindications administrating the drugs mentioned above, especially for those targeted anti-angiogenic substances. Therefore, discovering and developing novel protective strategies warranted greater attention.

Melatonin (MLT) is a derivative of indoleamine secreted by the pineal gland in a rhythmic manner, exerting various beneficial effects including antioxidant, antiaging, anti-inflammation, and immunomodulation[9]. So far, mounting evidence has suggested a pleiotropic role of MLT in both liver diseases and cardiovascular diseases. The vasoprotective effects of MLT against atherosclerosis, abdominal aortic aneurysm, and hypertension have been demonstrated[10-12]. MLT was also proved to alleviate liver fibrosis via its antioxidative and anti-inflammatory effects on both gut and liver[13-15]. Notably, MLT was considered as an angiogenesis inhibitor to combat cancer, while the effects of MLT on extrahepatic angiogenesis during cirrhotic PHT have never been reported[16].

Therefore, in the current study, we aimed to investigate how MLT affects cirrhotic PHT, especially on extrahepatic angiogenesis and splanchnic hyperdynamic circulation based on bile duct ligation (BDL)-induced rat model of liver cirrhosis and PHT.

MATERIALS AND METHODS

Animals

Male Sprague-Dawley rats (200-250 g, 6-8 weeks old) were procured from the Experimental Animal Center of the School of Medicine, Shanghai Jiao Tong University (Shanghai, China). All Sprague-Dawley rats were maintained at our facility under specific pathogen-free conditions, including a temperature of 22 °C, humidity range of 40%-60%, and a 12-hour light/dark cycle with ad libitum access to food and water. Common BDL or sham operation (Sham) were performed on rats. The rats were randomly divided into four groups with six rats per group: (1) Sham group; (2) Sham + MLT (25 mg/kg/day, ab libitum in drinking water) group; (3) BDL group; and (4) BDL + MLT (25 mg/kg/day, ab libitum in drinking water) group. The administration of MLT lasted until the rats were sacrificed at the end of the 4^th week. Ethical approval for all animal-related protocols was obtained from the Ethical Committee of Shanghai Ninth People’s Hospital, School of Medicine, Shanghai Jiao Tong University.

Hemodynamic studies

The rats were anesthetized by 3% isoflurane. Briefly, PE50 tube was placed in the left femoral artery or portal vein of rats. After connecting the tube to the pressure sensor, the mean arterial pressure (MAP), heart rate and PP were measured by using an ALC-MPA multi-channel biological information analysis system (Shanghai Alcott Biotech Co., Ltd., Shanghai, China). Cardiac output (CO) was measured by echocardiography. Briefly, Sprague-Dawley rats were placed in supine position under 3% isoflurane anesthesia. The chest area of the rats was depilated using a depilatory cream and then rinsed with water for image acquisition. Subsequently, the superior mesenteric artery (SMA) was carefully separated from its origin in the aorta. The pulsed-Doppler flow transducer was used to detect the flow in the SMA.

The stroke volume of the rat heart was determined by dividing the CO by the heart rate, measured in milliliters per beat (mL/beat). The cardiac index (CI), expressed in mL/minute/100 g, was calculated by dividing the CO by the body weight. Systemic vascular resistance, given in mmHg/mL/minute/100 g, was obtained by dividing the MAP by the CI. The SMA resistance, expressed in mmHg/mL/minute/100 g, was calculated by dividing the difference between MAP and PP by the SMA flow, all per 100 g[17].

Biochemical analysis

Blood samples were collected using anticoagulant-free tubes to obtain serum by centrifuging at 3000 rpm for 15 minutes at 4 °C. The serum samples were then subjected to analysis for the levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), and total bilirubin. This analysis was performed using an automatic clinical analyzer (HITACHI Ltd, Japan).

Enzyme-linked immunosorbent assay

The levels of pro-inflammatory cytokines including interleukin (IL)-1β, IL-6, and tumor necrosis factor-alpha were measured by corresponding enzyme-linked immunosorbent assay kits according to the manufacturer’s instructions.

Histological and immunohistochemical examinations

The hepatic lobes and mesenteries were obtained from sacrificed rats and fixed in 10% formalin buffer solution with pH of 7.4, followed by embedding in paraffin blocks for histological analysis. Subsequently, the liver and mesentery sections were subjected to staining with hematoxylin and eosin. In addition, Sirius red staining and Masson’s trichrome staining methods were employed specifically for the sections of liver. To ensure unbiased evaluation, all tissue sections underwent random examination by an experienced pathologist utilizing an optical microscope.

Immunohistochemical examination was conducted on sections obtained from paraffin-embedded liver and mesenteric tissues. To block endogenous peroxidase activity, a methanol solution containing 0.3% H₂O₂ was applied. The citrate buffer was used for antigen retrieval. To prevent nonspecific antibody binding, incubation with 10% normal horse serum was performed. The sections were then incubated overnight at 4 °C with the following primary antibodies: (1) Anti-collagen I antibody (1:200, Abcam, United Kingdom); (2) Anti-α-smooth muscle actin antibody (1:200, Abcam, United Kingdom); (3) Anti-CD31 antibody (1:100, Servicebio, China); and (4) Anti-CD68 antibody (1:200, Abcam, United Kingdom). Phosphate-buffered saline was used as the negative control. Subsequently, the bound antibodies were visualized using diaminobenzidine as the chromogen, followed by re-staining with hematoxylin.

RNA sequencing

Total RNA was isolated from the SMA tissues using RNAiso Plus (TaKaRa, Japan). RNA quality was examined by gel electrophoresis and a Nanodrop spectrophotometer (Thermo, United States).

Total RNA was isolated from the primary astrocytes using RNAiso Plus (TaKaRa, Japan). RNA quality was examined by gel electrophoresis and a Nanodrop spectrophotometer (Thermo, United States). For RNA-sequencing, samples from BDL and BDL + MLT groups with 6 biological replicates per group were used. Briefly, TruSeq RNA sample preparation kit (Illumina, United States) was used to construct strand-specific libraries, followed by sequencing using the Illumina HiSeq X. The raw data was converted into Fastq format. The number of transcripts was calculated by fragments per kilobase of transcript per million fragments mapped (FPKM). Further bioinformatic analysis was performed by R 3.4.0 and corresponding R packages.

Cell culture and treatment

Human umbilical vein endothelial cells (HUVECs) purchased from Cellverse (Shanghai, China) were cultured in 1640 medium (Gibco, United States) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin (100 IU/mL)/streptomycin (100 μg/mL) solution (P/S). The culture was maintained in a humidified incubator at 37 °C with 5% CO₂. HUVECs were incubated in VEGF-A (50 ng/mL), MLT (1 mM), or AZD2171 (100 nM) for further mechanistic investigations. To interfere the gene expression of VEGFR-2 and POMK, siRNA (VEGFR2 Human siRNA: 5’-GCGGCACGAAATATCCTCTTA-3’; POMK Human siRNA: 5’-GTCTTGGATACACTTAGA-3’) were transfected by Lipofectamine 3000 (Invitrogen, United States) according to manufacturer’s instructions.

Cell proliferation assay

HUVECs were cultured in 96-well plates, treated with VEGF-A, MLT, and/or AZD2171 at the corresponding concentrations for days. Then the cells were incubated with CCK8 reagent (Beyotime, China) for 1 hour. The absorbance at 450 nm was measured with a scanning multiwell spectrophotometer.

Tube formation assay

Firstly, the 1640 medium with 10% FBS on HUVECs was changed to 1640 medium without FBS for overnight serum starvation. Then HUVECs incubated with vehicle, VEGF-A, MLT, and/or AZD2171 were seeded onto a 96-well culture plate (2000 cells/well) coated with growth factor-reduced Matrigel (Corning, United States) and incubated at 37 °C for 4 hours. The morphological changes of HUVECs were microscopically captured after fixation with 4% paraformaldehyde. The tube networks were assessed in 3 different fields per well, when tube formation was evaluated by tube lengths, number of branching points and loops using ImageJ software.

Wound healing assay

Wound healing assay was used to examine the migration ability of HUVECs. Cell monolayers were wounded by dragging a sterile 1000 μL pipette tip across the culture, and the disassociated cells and debris were washed away by phosphate-buffered saline for 3 times. During wound healing assay, HUVECs incubated with vehicle, VEGF-A, MLT, and/or AZD2171. Images of the wounds were acquired at 0 hour and 24 hours after wound dragging. Images were captured and analyzed with ImageJ software. The cell migration distance was calculated by dividing the reduced wound area with the length of the wound.

Transwell assay

Vertical migrating ability of HUVECs was investigated using Transwell migration assay (Boyden Chamber, 8.0 μm pore polycarbonate membranes, Corning, NY, United States). Cells (40000 cells/insert) were seeded on Transwell inserts in 1640 medium without FBS. The lower chamber was filled with 1640 medium with 10% FBS. After 48 hours, non-migrated cells from the topside of membranes were wiped, and membranes were fixed with 4% paraformaldehyde and stained with 0.5% crystal violet. Images were captured and analyzed with ImageJ software by counting migrated cells.

Immunoprecipitation

After incubation, the HUVECs were lysed by NP40 buffer (Boster Biological Technology, AR0107) at 4 °C for 15 minutes and centrifuged at 14000 g and 4 °C for 15 minutes. The protein concentrations in the lysate supernatants were determined using BCA protein analysis kit (Beyotime, China) and the concentrations of supernatants were adjusted to approximately 1 μg/μL. Total 500 μL of the lysates were incubated with rabbit monoclonal anti-VEGFR-2 antibody (1:100, Abcam, United States) or Normal Rabbit IgG (1:100, Abcam, United States) at 4 °C overnight. Total 50 μL Protein G Magnetic Beads (Thermo Scientific™, 88847) were added to immunoprecipitate the protein-antibody complexes. After 2 hours incubation at 4 °C, the samples were washed three times with the wash buffer. The beads containing the protein-antibody complexes were mixed with loading buffer, boiled at 100 °C for 10 minutes, and subjected to mass spectrometry and western blotting analysis.

Western blotting

Liver and mesenteric tissues of rats were collected and stored at -80 °C. The samples were homogenized with RIPA buffer (Beyotime, China) to extract protein from these tissues, following the manufacturer’s instructions. The tissue extracts were centrifuged for 15 minutes at 4 °C at a speed of 12000 g. Subsequently, the resulting supernatant was collected and the total protein concentration was determined using a BCA protein analysis kit (Beyotime, China). Equal amounts of proteins were loaded onto sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and subsequently transferred onto PVDF membranes through electrotransfer. The membranes were incubated with primary antibodies against VEGFR-2 (1:1000, ABclonal Technology, China), phospho-VEGFR2 (1:1000, ABclonal Technology, China), POMK (1:1000, Proteintech, United States), DSG1 (1:1000, Proteintech, United States), DOCK5 (1:1000, Proteintech, United States), and GAPDH (1:10000, Proteintech, United States). Following that, the membranes were incubated with secondary antibodies for 1 hour at 24 °C and visualized by an enhanced chemiluminescence detection kit (Millipore Corporation, United States).

Quantitative real-time polymerase chain reaction

The total RNA was extracted from mesenteric tissues by RNAiso Plus (TaKaRa, Japan). The cDNA was synthesized by RT SuperMix for qPCR (Vazyme, China). Quantitative real-time polymerase chain reaction was performed with ChamQ Universal SYBR qPCR Master Mix (Vazyme, China) in LightCycler^® 480 Real-Time PCR System (Roche, Germany). The results from corresponding analysis software were presented as Ct values. The mRNA levels of targeted genes were normalized to GAPDH. The primer sequences used in this study were listed as below: (1) Vegfa, forward 5’-AGAAAGCCCATGAAGTGGTG-3’, reverse 5’-ACTCCAGGGCTTCATCATTG-3’; (2) Vegfb, forward 5’-TGGTACCTCTGAGCATGGAA-3’, reverse 5’-GAGGATCTGCATTCGGACTT-3’; (3) Vegfc, forward 5’-TGCCGGTGCATGTCTAAA-3’, reverse 5’- CTGCCTGACACTGTGGTAAT-3’; (4) Vegfd, forward 5’-TGGGACAGAAGACCACTCTTA-3’, reverse 5’-TCCAGGACATGGTGCTTTAC-3’; and (5) Gapdh, forward 5’-TTGTGCAGTGCCAGCCTC-3’, reverse 5’-GGTAACCAGGCGTCCGATAC-3’.

Statistical analysis

Statistical analysis was conducted using IBM SPSS Statistics version 23.0 (United States). The data were presented as means ± SD. Statistical significance was determined using Student’s t-test. For multiple comparisons, one-way analysis of variance was employed. A P value of less than 0.05 was considered statistically significant for all analyses.

RESULTS

MLT alleviated BDL-induced liver cirrhosis and PHT

To investigate the effects of MLT on cirrhotic PHT, BDL rats were administrated with MLT (25 mg/kg/day) or distilled water as vehicle (Figure 1A). MLT significantly lowered the PP in portal hypertensive rats without significantly changing MAP (Table 1).

Figure 1 Melatonin alleviated bile duct ligation-induced liver cirrhosis and portal hypertension. A: Construction of animal model of liver cirrhosis and melatonin administration; B: The levels of plasma liver injury markers alanine aminotransferase, aspartate aminotransferase and total bilirubin in rats; C: The histological staining of liver tissue sections using hematoxylin and eosin, Masson’s trichrome, and Sirius red. Immunohistochemical staining of collagen I and α-smooth muscle actin in rat livers; D: Quantitative analysis of histopathological and immunohistochemical staining in C; E: Western blotting of hepatic VEGFR-2 and its phosphorylated form; F: Quantitative analysis of VEGFR-2 phosphorylation. ^aP < 0.05. ^bP < 0.01. BDL: Bile duct ligation; MLT: Melatonin; α-SMA: Α-smooth muscle actin; HE: Hematoxylin and eosin; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase.

Table 1 Body weight and hemodynamic parameters, mean ± SD.

Characteristics	Sham	Sham + MLT	BDL	BDL + MLT
Body weight (g)	408.5 ± 11.3	404.7 ± 12.0	349.2 ± 22.0b	364.3 ± 20.2
MAP (mmHg)	119.5 ± 4.3	119.3 ± 8.9	99.2 ± 5.9b	104.0 ± 6.3
PP (mmHg)	8.0 ± 0.9	7.7 ± 1.4	20.2 ± 2.9b	14.5 ± 2.3b
HR (beats/minute)	343 ± 23.4	336.8 ± 16.2	350.5 ± 14.6	349.3 ± 19.0
SV (mL/beats)	0.41 ± 0.04	0.42 ± 0.04	0.58 ± 0.08b	0.43 ± 0.05b
CO (mL/minute)	140.7 ± 9.6	137.1 ± 16.1	203.4 ± 31.3b	149.3 ± 22.6b
CI (mL/minute/100 g)	34.4 ± 2.7	34.0 ± 4.9	58.5 ± 10.1b	41.1 ± 6.9b
SVR (mmHg/mL/minute/100 g)	3.5 ± 0.3	3.6 ± 0.7	1.7 ± 0.4b	2.6 ± 0.3a
SMA flow (mL/minute/100 g)	3.1 ± 0.5	3.3 ± 0.5	6.4 ± 0.4b	4.5 ± 0.5b
SMA resistance (mmHg/mL/minute/100 g)	37.1 ± 6.1	35.0 ± 6.1	12.4 ± 1.5b	20.0 ± 1.9a

^aP < 0.05.

^bP < 0.01.

MLT: Melatonin; BDL: Bile duct ligation; MAP: Mean arterial pressure; HR: Heart rate; SMA: Superior mesenteric artery; SV: Stroke volume; SVR: Systemic vascular resistance; PP: Portal pressure; CO: Cardiac output; CI: Cardiac index.

Moreover, we investigated the changes occurred in liver upon MLT treatment. BDL rats exhibited significant liver injury and fibrosis. It was shown that the liver injury-related serum ALT and AST was significantly reduced by MLT, while the serum level of bilirubin was not decreased (Figure 1B). The pathohistological staining suggested a significant amelioration of liver fibrosis as the collagen staining and fibrotic genes were significantly downregulated by MLT (Figure 1C and D).

The activation of vasoreactivity factors in liver were also investigated. In BDL rats, phosphorylation of VEGFR-2 was significantly decreased compared to sham-operated rats, which was reversed by administration with MLT (Figure 1E and F).

MLT alleviated BDL-induced mesenteric vascular remodeling and angiogenesis

As shown in Table 1, the hyperdynamic splanchnic circulation exhibited in cirrhotic PHT was improved by MLT, featured by significantly increased systemic vascular resistance and SMA resistance, accompanied with decreased stroke volume, CO, CI, and SMA blood flow. Administration of MLT didn’t alter the hemodynamics in sham-operated rats.

Therefore, the vascular structure of mesenteric arteries was evaluated in the current study. An eccentric remodeling featured by thinner wall and dilated lumen was observed in mesenteric arteries from BDL rats, which was mitigated after administration with MLT (Figure 2A and B). Meanwhile, the elevated levels of pro-inflammatory cytokines including IL-1β, IL-6, and tumor necrosis factor-alpha in portal hypertensive mesenteric arteries were significantly lowered by MLT (Figure 2C). The infiltration of macrophage detected by immunostaining of CD68 was also downregulated by MLT, suggesting an anti-inflammatory effect of MLT in cirrhotic rats (Figure 2D and E).

Figure 2 Melatonin alleviated bile duct ligation-induced mesenteric vascular remodeling and angiogenesis. A: Representative hematoxylin and eosin staining images of mesenteric artery. The yellow arrows represent the thickness of vascular wall; B: Quantitative analysis of the wall thickness; C: Enzyme-linked immunosorbent assay of inflammatory cytokines in mesenteric artery tissues; D: Representative immunohistochemical images of CD68 in the mesentery; E: Quantitative analysis of CD68-positive cells; F: Representative immunohistochemical images of CD31 in the mesentery; G: Quantitative analysis of CD31 staining intensity; H: Western blotting of mesenteric VEGFR-2 and its phosphorylated form; I: Quantitative analysis of VEGFR-2 phosphorylation. ^aP < 0.05. ^bP < 0.01. BDL: Bile duct ligation; MLT: Melatonin; HE: Hematoxylin and eosin; IL: Interleukin; TNF-α: Tumor necrosis factor alpha.

In addition, we also investigated the angiogenesis in mesentery. CD31⁺ vascular density was significantly downregulated by MLT, suggesting a suppression against angiogenesis by MLT, which might also account for the reduced SMA blood flow (Figure 2F and G). Furthermore, as we observed a significant phosphorylation of VEGFR-2 in mesenteric arteries from BDL rats, a remarkable downregulation of VEGFR-2 phosphorylation by MLT might also contribute to the increased SMA resistance (Figure 2H and I).

RNA-sequencing of mesenteric arteries from BDL rats treated with MLT

To further illustrate the underlying mechanisms of MLT alleviating both intra- and extra-hepatic alterations in cirrhotic PHT, we performed bulk RNA-sequencing with separated SMA tissues. The volcano plot and heatmap of differential expressed genes between BDL rats treated with vehicle or MLT was shown by Figure 3A and B, respectively. Protein-protein interaction network based on differential expressed genes was used to identify hub genes including Atf3, Nr4a1, Itgad, Myc, Fosb, Fgf2, Fox, Jun, Egr1, and Ptgs2 (Figure 3C). The GO analysis suggested a significant enrichment of PDGF, VEGF, and NO-related vascular cell proliferation and angiogenesis (Figure 3D). Kyoto Encyclopedia of Genes and Genomes enrichment analysis showed that MLT treatment is related to cytokines signaling and cell death-related pathways (Figure 3E). Interestingly, we have observed significant downregulations of VEGF family in mesenteric arteries after MLT administration, which might account for the suppression of mesenteric vascular remodeling and angiogenesis by MLT (Figure 3F). Additionally, the downregulation of VEGF family was validated by quantitative real-time polymerase chain reaction of mesenteric tissues (Figure 3G).

Figure 3 RNA-sequencing of mesenteric arteries from bile duct ligation rats treated with melatonin. A: The volcano plot of differential expressed genes (DEGs) between bile duct ligation (BDL) and BDL + melatonin (MLT) groups; B: The heatmap of DEGs between BDL and BDL + MLT groups; C: The protein-protein interaction network of DEGs with 10 hub genes; D: The top 10 terms enriched by Gene Ontology enrichment analysis based on DEGs; E: The top 10 pathways enriched by Kyoto Encyclopedia of Genes and Genomes analysis based on DEGs; F: The expression profile of VEGFs between BDL and BDL + MLT groups; G: The validation of the expression levels of VEGFs by quantitative real-time polymerase chain reaction. ^aP < 0.05. ^bP < 0.01. BDL: Bile duct ligation; MLT: Melatonin.

VEGFR-2 inhibition alleviated BDL-induced cirrhotic PHT

Regarding the significant downregulations of VEGFs, we speculated that VEGFs were core mediators of the protective effects against cirrhotic PHT by MLT. Therefore, the VEGFR-2 inhibitor AZD2171 was administrated to BDL rats, which showed similar protective effects with MLT (Figure 4A). AZD2171 significantly reduced PP and SMA blood flow, accompanied with increased SMA resistance, suggesting alleviation of both portal and splanchnic hemodynamics (Figure 4B). Meanwhile, reduced collagen deposition and fibrotic gene expression in liver showed anti-fibrotic effects of AZD2171 on BDL rats (Figure 4C and D). The extra-hepatic alterations were investigated by immunostaining of CD68 and CD31 (Figure 4E-H). Reduced macrophage infiltration and vascular density suggested both anti-inflammatory and anti-angiogenic effects of AZD2171.

Figure 4 VEGFR-2 inhibition alleviated bile duct ligation-induced cirrhotic portal hypertension. A: Construction of animal model of liver cirrhosis and AZD2171 administration; B: The portal hypertension hemodynamic parameters in rats from 4 groups; C: The histological staining of liver tissue sections using hematoxylin and eosin, Masson’s trichrome, and Sirius red. Immunohistochemical (IHC) staining of Collagen I and α-smooth muscle actin; in rat livers; D: Quantitative analysis of histopathological and IHC staining in C; E: Representative IHC images of CD68 in the mesentery; F: Quantitative analysis of CD68-positive cells; G: Representative IHC images of CD31 in the mesentery; H: Quantitative analysis of CD31 staining intensity. ^aP < 0.05. ^bP < 0.01. α-SMA: Α-smooth muscle actin; HE: Hematoxylin and eosin; BDL: Bile duct ligation.

MLT and AZD2171 inhibited angiogenesis in vitro

HUVECs were used to investigate the effects of VEGFR-2 inhibition in vitro. As shown by Figure 5A and B, VEGF-A significantly activated the phosphorylation of VEGFR-2, which was reversed by incubation with MLT and AZD2171. The enhanced proliferation by VEGF-A was also suppressed by MLT and AZD2171 (Figure 5C). Furthermore, the anti-angiogenic effects of MLT and AZD2171 were validated by tube formation assay, when the tube length and branching points were significantly decreased compared to the positive control treated with VEGF-A (Figure 5D and E). The wound healing assay also suggested that both MLT and AZD-2171 impeded the horizontal migration of endothelial cells (Figure 5F and G). The vertical migration ability of endothelial cells was also suppressed by MLT and AZD2171, as demonstrated by Transwell assay (Figure 5H and I).

Figure 5 Melatonin and AZD2171 inhibited angiogenesis in vitro. A: Western blotting of VEGFR-2 and its phosphorylated form in human umbilical vein endothelial cells (HUVECs) treated with VEGFA, melatonin (MLT), or AZD2171; B: Quantitative analysis of VEGFR-2 phosphorylation; C: The proliferation assay of HUVECs treated with VEGFA, MLT, or AZD2171; D: The tube formation assay of HUVECs treated with VEGFA, MLT, or AZD2171; E: Quantitative analysis of tube formation assay; F: The wound healing assay of HUVECs treated with VEGFA, MLT, or AZD2171; G: Quantitative analysis of wound healing assay; H: Transwell migration assay of HUVECs treated with VEGFA, MLT, or AZD2171; I: Quantitative analysis of Transwell migration assay. ^aP < 0.05. ^bP < 0.01. MLT: Melatonin.

MLT suppressed VEGF signaling and interactions between VEGFR-2 and POMK

After elucidating the involvement of VEGF-VEGFR signaling in the protective effects of MLT, the downstream signaling pathways were also investigated in HUVECs. The downstream phosphorylation of PLCγ and ERK1/2 were suppressed by both MLT and AZD2171 (Figure 6A and B). We performed immunoprecipitation (IP)-mass spectrometry by primary antibody against VEGFR-2 to figure out the underlying mechanisms how MLT suppressed VEGFR-2 signaling (Figure 6C). The top-enriched proteins in IP-mass spectrometry were shown in Figure 6D. Further IP verified the binding of target proteins with VEGFR-2 upon VEGF-A and MLT incubation (Figure 6E-H). The bindings of DOCK5, DSG1, and POMK with VEGFR-2 were significantly upregulated by VEGF-A. MLT slightly suppressed the interaction between DOCK5 and VEGFR-2, while the binding of DSG1 was not significantly changed. The binding of POMK was remarkably downregulated by MLT, which suggested the potential role of POMK in MLT-induced VEGFR-2 signaling inhibition.

Figure 6 Melatonin suppressed VEGF signaling and interactions between VEGFR-2 and POMK. A: Western blotting of VEGFR-2 signaling in human umbilical vein endothelial cells (HUVECs) treated with VEGFA, melatonin, or AZD2171; B: Quantitative analysis of the activation of VEGFR-2 signaling; C: Immunoprecipitation (IP) blotting of VEGFR-2 in HUVECs and subjected of IP-mass spectrometry; D: PPI network of enriched proteins by IP-mass spectrometry; E: IP blotting of DOCK5 by VEGFR-2 in HUVECs; F: IP blotting of DSG1 by VEGFR-2 in HUVECs; G: IP blotting of POMK by VEGFR-2 in HUVECs; H: Quantitative analysis of IP blotting. ^aP < 0.05. ^bP < 0.01. MLT: Melatonin.

MLT suppressed VEGFR-2 signaling via a POMK-dependent manner

To further validate the role of POMK in mediating MLT's inhibition of VEGFR-2 signaling, siRNAs targeting VEGFR-2 and POMK were transfected into HUVECs. The efficiency of VEGFR-2 knockdown was shown in Figure 7A. VEGFR-2 knockdown reduced the phosphorylation of downstream signaling, including PLCγ and ERK1/2, even without activation of VEGF-A (Figure 7B-D). The efficiency of POMK knockdown was shown in Figure 7E. Crucially, the lack of POMK not only diminished the overall activation of VEGFR-2 but also abolished the inhibitory effect of MLT. In POMK-depleted HUVECs, the addition of MLT failed to further suppress VEGFR-2 phosphorylation compared to the vehicle-treated knockdown group (Figure 7F-I). This genetic evidence definitively demonstrated that the inhibitory effect of MLT on VEGFR-2 signaling was highly dependent on the presence of POMK.

Figure 7 Melatonin suppressed VEGFR-2 signaling via a POMK-dependent manner. A: Validation of VEGFR-2 knockdown by siRNA in human umbilical vein endothelial cells (HUVECs); B: Western blotting of VEGFR-2 signaling in HUVECs with VEGFR-2 knockdown; C and D: Quantitative analysis of PLCγ and ERK phosphorylation in B; E: Validation of POMK knockdown by siRNA in HUVECs; F: Western blotting of VEGFR-2 signaling in HUVECs with POMK knockdown; G-I: Quantitative analysis of VEGFR-2, PLCγ, and ERK phosphorylation in F. ^bP < 0.01. MLT: Melatonin.

DISCUSSION

The pathogenesis of cirrhotic PHT involves a progressive cycle that persistently elevated IHVR initiates the disease, further exacerbated by subsequent development of splanchnic hyperdynamic circulation[18]. Although previous studies have extensively documented the anti-fibrotic properties of MLT in reducing intrahepatic resistance, the precise mechanisms governing its effects on the extrahepatic splanchnic vasculature have remained largely unexplored. In the present study, we expanded upon these findings by demonstrating that MLT interrupted this pathological progression by suppressing mesenteric angiogenesis and vascular remodeling. By identifying the POMK-VEGFR2 axis as a novel regulatory pathway, our results provided a mechanistic explanation for how MLT effectively mitigated splanchnic hyperdynamic circulation (Figure 8).

Figure 8 Proposed mechanism underlying the protective effects of melatonin against pathological angiogenesis in cirrhotic portal hypertension. Melatonin disrupts the interaction between POMK and VEGFR-2, resulting in reduced VEGFR-2 phosphorylation and attenuation of downstream PLC-γ/ERK1/2 signaling activation. This inhibition suppresses VEGF-induced endothelial cell proliferation and angiogenesis, thereby contributing to the improvement of splanchnic hyperdynamic circulation in cirrhotic portal hypertension. HUVEC: Human umbilical vein endothelial cell.

Pathological angiogenesis in the mesenteric bed promotes the formation of portosystemic collaterals and results in increased splanchnic blood flow, thereby limiting the efficacy of traditional therapeutic strategies that primarily target hepatic fibrosis. While VEGFR2-targeted tyrosine kinase inhibitors exhibit robust anti-angiogenic efficacy, their clinical application in patients with cirrhosis is limited by severe systemic adverse events. In this context, MLT presented a favorable safety profile while exerting potent pharmacological effects. Our findings revealed that MLT restricted the proliferation and migration of vascular endothelial cells by specifically disrupting the interaction between POMK and VEGFR2. This structural disruption effectively prevented VEGFR2 phosphorylation. These findings suggest that modulation of VEGFR2-associated angiogenic signaling represents an important mechanism through which MLT alleviates extrahepatic vascular abnormalities in cirrhotic PHT.

Notably, while evaluating the intrahepatic effects of MLT, we observed that MLT significantly reduced serum AST and ALT levels but failed to decrease bilirubin levels. This differential biochemical response was associated with the characteristics of the BDL model. Because BDL induces persistent mechanical biliary obstruction, MLT cannot reverse the primary cholestatic process, thereby limiting its effects on hyperbilirubinemia. However, its hepatoprotective properties effectively attenuated secondary hepatocellular injury and necrosis caused by toxic bile acid accumulation, resulting in reduced transaminase levels. These findings highlighted the distinct yet complementary roles of MLT in alleviating hepatocellular injury while ameliorating extrahepatic vascular dysfunction.

The VEGFA/VEGFR2 axis is well recognized as a major driver of pathological angiogenesis[19,20]. Extensive evidence has demonstrated that VEGF signaling is markedly upregulated in liver cirrhosis, where its persistent activation drives splanchnic angiogenesis and the formation of portosystemic collateral circulation[17]. In the present study, transcriptomic profiling of mesenteric arteries evidenced that MLT suppressed angiogenesis-associated signaling pathways in BDL rats, concurrently downregulating VEGF protein expression. While the precise molecular cascades have historically been elusive, growing evidence identifies VEGF as a crucial target through which MLT exerts its anti-angiogenic effects[18,21]. Previous research highlighted that MLT modulated the expression of VEGF and its receptors across various pathophysiological conditions, including embryonic development and tumorigenesis[22,23]. Collectively, these findings strongly align with our transcriptomic results, providing a rationale for further investigation of how MLT modulates the VEGF/VEGFR axis in cirrhotic PHT.

Previous studies have established that pharmacological antagonism of VEGFR2 (e.g., using sorafenib, imatinib, or rapamycin) yields robust anti-angiogenic and anti-fibrotic efficacy[24,25]. To investigate the role of VEGFR2 in our observed outcomes, we employed AZD2171, a selective, orally available VEGFR2 tyrosine kinase inhibitor. Strikingly, AZD2171 treatment phenocopied the protective benefits of MLT. In BDL rats, targeted blockade of the VEGFR2 by AZD2171 significantly reduced PP and portal blood flow, driven by a marked decrease in SMA inflow and the marked suppression of mesenteric angiogenesis. These in vivo findings were corroborated by in vitro assays demonstrating that AZD2171 significantly inhibited HUVEC proliferation. Because the targeted inhibition by AZD2171 completely recapitulates the outcomes achieved with MLT, our data compellingly confirmed that MLT ameliorated splanchnic hyperdynamic circulation in PHT by blunting VEGFR2 phosphorylation and halting pathological mesenteric angiogenesis. More importantly, while conventional VEGFR2 tyrosine kinase inhibitors like AZD2171 are frequently plagued by severe systemic toxicities that limit their chronic use in cirrhotic patients, MLT possesses an exceptionally favorable safety profile. These findings support further investigation of MLT as a potential strategy for targeting pathological angiogenesis in cirrhotic PHT.

VEGFA, a major regulator of angiogenesis, orchestrates endothelial cell proliferation primarily through VEGFR2 activation and the subsequent engagement of downstream effectors, notably PLC-γ and ERK[26]. In the present study, we demonstrated that MLT suppressed VEGFA-driven endothelial proliferation, migration, and angiogenesis by disrupting the interaction between VEGFR2 and POMK. This disruption reduced VEGFR2 phosphorylation and attenuated downstream signaling without significantly altering total VEGFR2 abundance, suggesting that MLT primarily regulates receptor activation rather than VEGFR2 expression or stability. Consistently, genetic depletion of POMK in HUVECs recapitulated the inhibitory effects of MLT on VEGFA/VEGFR2 signaling, supporting the functional importance of the POMK-VEGFR2 axis in angiogenic regulation. Although these findings establish the functional relevance of the POMK-VEGFR2 interaction, the precise molecular basis of this regulation requires further investigation. POMK may regulate VEGFR2 signaling through O-mannosylation-related enzymatic activity or through protein-protein interaction-mediated scaffolding. In addition, whether MLT modulates this complex directly or through upstream MT1/MT2-dependent signaling remains to be clarified.

Several considerations should be addressed regarding the clinical translation of our findings. Although the BDL model is widely used to investigate cholestatic cirrhosis-associated PHT, it does not fully represent the heterogeneous etiologies of human cirrhosis. Nevertheless, pathological angiogenesis and VEGFR2 signaling represent common mechanisms contributing to PHT progression across different disease contexts, suggesting potential broader relevance of the POMK-VEGFR2 axis. Validation in additional etiologically distinct models will further establish its general applicability. In addition, the dose of MLT used in this preclinical study was selected based on previous experimental studies investigating the pharmacological effects of MLT in liver disease and vascular dysfunction models. Although higher than doses commonly used for sleep-related indications, such pharmacological dosing is frequently applied in experimental disease models. Given the favorable safety profile of MLT, further pharmacokinetic and dose-optimization studies are warranted to determine an appropriate clinical dosing strategy. Collectively, our findings identify a novel regulatory pathway linking MLT to VEGFR2 activation and provide new insights into the potential protective role of MLT in controlling pathological angiogenesis in cirrhotic PHT.

CONCLUSION

In conclusion, our study demonstrated that MLT attenuated splanchnic hyperdynamic circulation in cirrhotic PHT by inhibiting mesenteric angiogenesis. The identification of the POMK-VEGFR2 axis provided new insights into the regulation of pathological angiogenesis and highlighted a potential protective strategy for PHT.