Serum proteomics identify biomarkers and pathogenesis of portopulmonary hypertension diagnosed based on echocardiography

Abstract

Portal pulmonary hypertension (PoPH), a severe complication of portal hypertension (PHTN), is marked by elevated pulmonary arterial pressure, but its pathophysiological mechanisms are unclear. This study used proteomics to identify differentially expressed proteins (DEPs) and genes in Patients with PoPH compared to those with PHTN and healthy controls (HC), aiming to uncover potential biomarkers for diagnosis and treatment. Patients with liver cirrhosis and PHTN, admitted between January 2023 and May 2024, were classified into PoPH and non-PoPH (PHTN) groups based on echocardiography. Serum from 12 PoPH, 12 PHTN, and 6 HC was analyzed using data-independent acquisition (DIA) proteomics to identify DEPs. Protein-protein interaction (PPI) networks identified key DEPs, and ELISA was performed for biomarker validation. Compared to HC, 374 proteins were upregulated and 115 downregulated in PoPH, while 18 were upregulated and 38 downregulated in PHTN. KEGG and GO analyses linked DEPs to immune response, metabolism, and cell signaling. Thirty-five proteins distinguish PoPH from HC and PHTN. Vitronectin (VTN, P04004) was correlated with RDW (R = -0.56, P < 0.01) and PLT (R = 0.52, P < 0.01). ELISA confirmed lower VTN levels in PoPH (P < 0.05). This study identified 35 serum proteins involved in PoPH, with VTN as a potential biomarker for distinguishing PoPH from PHTN and HC. Further research is needed to explore these findings.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-08376-6.

Introduction

Portopulmonary hypertension (PoPH) is pulmonary arterial hypertension (PH) in association with portal hypertension (PHTN). The pathogenesis of PoPH is complex and closely related to a variety of diseases causing PHTN, such as liver cirrhosis, bile duct obstruction, and cholestatic disease. Any disease that can cause PHTN, including congenital extrahepatic portosystemic shunts, may result in PoPH, regardless of the presence or absence of liver disease. According to the 2022 European Society of Cardiology/European Respiratory Society guidelines (ESC/ERS), PoPH is classified as Group 1 PH¹. The prevalence of PoPH has been reported variably in different studies, and it accounts for approximately 5-15% of all PAH causes^2,3, with prevalence rates of 2-6% in patients with PH and 1-2% in those with cirrhosis^4,5. Data reported from the United States indicate that the prevalence of PoPH increases to 5-8% among liver transplant patients⁶, while the latest epidemiological data from China suggest that Patients with PoPH account for 6.3% of all liver transplant patients⁷.

Despite the absence of a definitive elucidation of the precise pathogenesis of PoPH, many potential factors have been posited as contributors to its development. These include hyperdynamic circulatory states, imbalances in vasoactive mediators, local thrombosis, genetic variants, systemic inflammatory responses, immune injury, and oxidative stress. A distinctive feature of PoPH is its association with complex interactions between the hepatic and pulmonary circulations, which may involve pathophysiological processes that have not yet been fully elucidated. Current research suggests that the occurrence of PoPH involves a multifactorial and multistep process. Nevertheless, the precise molecular basis has yet to be fully elucidated. While existing diagnostic methods (e.g., echocardiography) can provide an initial assessment of PoPH, they are limited in revealing its molecular features. This underscores the pressing need to develop innovative diagnostic methods and research tools to elucidate the molecular mechanisms of PoPH further and enable effective differentiation from other forms of pulmonary hypertension.

The clinical presentation of PoPH is often non-specific, which significantly increases the difficulty of early diagnosis and leads to delayed recognition and intervention in the early stages of the disease. Such delays exacerbate the diagnostic challenges and may result in the loss of critical opportunities for effective treatment. Of particular concern is the dismal overall prognosis of patients with PoPH, with a 5-year survival rate ranging from only 30–40%^8,9. Notably, this rate is further reduced to 14% in cases of untreated moderate to severe PoPH¹⁰. Therefore, a comprehensive investigation into the pathogenesis of PoPH and refinement of early diagnostic strategies are crucial for improving patient prognosis and survival. By identifying the molecular and pathological features of PoPH and developing more precise and reliable diagnostic tools, earlier detection and intervention can be achieved, ultimately leading to substantial improvements in clinical outcomes.

In recent years, proteomics has emerged as a significant instrument in studying molecular mapping of diseases, facilitating elucidation of the molecular mechanisms underlying complex diseases such as PoPH. By comparing the proteomic data of Patients with PoPH with that of PHTN patients and HC, it is possible to identify differentially expressed proteins (DEPs) and genes, which may serve as biomarkers for early diagnosis, prognosis, and therapeutic targets. Consequently, proteomics has the potential to not only elucidate the molecular pathological features of PoPH but also to establish the foundation for future targeted therapeutic interventions.

In this study, we analysed serum samples from patients with cirrhotic portal hypertension who were admitted between January 2023 and May 2024. We used data-independent acquisition (DIA) quantitative proteomics technology to analyze the samples. DEPs were identified in patients with PoPH compared to patients with PHTN and HC. These DEPs were subjected to comprehensive functional enrichment analyses to understand their biological roles. To distinct molecular characteristics of PoPH, we integrated multiple bioinformatics approaches, including Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG)¹¹, InterPro database, and Protein-Protein Interaction (PPI) network analyses.

Identifying key DEPs and their functional roles in hepatogenic PoPH can enhance our comprehension of the pathogenesis of the disease. Furthermore, this process may facilitate the development of potential biomarkers, which could not only improve early diagnosis of PoPH but also provide a theoretical foundation for the development of personalised therapeutic strategies. The results are expected to have a significant impact on diagnosis, treatment, and management of patients with PoPH.

Materials and methods

Patients

This study retrospectively collected patients with cirrhosis and portal hypertension who attended the First Hospital of Jilin University between January 2023 and June 2024. All patients had undergone cardiac ultrasonography. The echocardiographic results were then used to divide the patients into two groups: diagnosed with PoPH and non-PoPH (PHTN). Concurrently, relevant clinical data and serum samples were collected for further analysis.

A total of 30 subjects were enrolled in this study for proteomic analysis, including 12 Patients with PoPH, 12 PHTN patients, and 6 HC. In addition, to validate the biomarkers, another 10 patients with PoPH, 10 patients with PHTN, and 10 HC admitted in the same period were selected for the validation experiments. It is important to note that all subjects were rigorously screened according to predefined inclusion and exclusion criteria. This process was done to ensure the representativeness and reliability of the samples. Furthermore, to minimize potential confounding factors, the control groups (PHTN and HC) were age- and sex-matched to the PoPH group.

Inclusion criteria

The study’s inclusion criteria were: First, all cirrhotic patients met the diagnostic criteria in the Guidelines for the Diagnosis and Treatment of Liver Cirrhosis (2019)¹². Second, all cirrhotic patients met the diagnostic criteria for PHTN in the Cirrhosis Portal Hypertension Baveno VII—consensus (2021 Edition)¹³. Third, enrolled patients with pulmonary hypertension met the criteria for diagnosis by echocardiography in the 2022 ESC/ERS Guidelines for the Diagnosis and Treatment of Pulmonary Arterial Hypertension¹.

Exclusion criteria

First, referring to the classification of PH in the Chinese Guidelines for the Diagnosis and Treatment of Pulmonary Arterial Hypertension (2021 Edition)¹, patients with pulmonary, cardiac, haematological, connective tissue diseases, chronic kidney disease, infections (HIV, etc.) and metabolic diseases (hyperthyroidism, hypothyroidism) that may lead to PH were excluded; Second, patients with incomplete clinical data were excluded; and third patients with incomplete blood samples were excluded.

Sample processing

Serum samples were obtained through venipuncture and allowed to clot at room temperature. Following collection, the samples should be processed as soon as possible (preferably within 2 h at 2–8 °C) and centrifuged at 3000 × g for 15 min to isolate the upper serum layer. The separated serum was carefully transferred into sterile tubes and stored at -80 °C for further analysis. All collection dates, times, processing procedures, and storage conditions were meticulously recorded to ensure full data traceability.

Enzyme-Linked immunosorbent assay (ELISA)

ELISA (enzyme-linked immunosorbent assay) is a widely used technique for detecting and quantifying antigens or antibodies. It is based on the specific interaction between an antigen and its corresponding antibody. An enzyme-conjugated antibody (or streptavidin) reacts with a substrate to produce a colorimetric change, the intensity of which is measured using a photometer. This signal is proportional to the concentration of the target protein in the sample.

Detection methods for VTN, APOA1, and APOA2 levels

We measured the levels of VTN, APOA1, and APOA2 using kits called ELISA kits (VTN: CSB-E08983h; APOA1: CSB-E08103h; APOA2: CSB-E13504h; CUSABIO). All reagents, standards, and serum samples were prepared according to the manufacturer’s instructions. Before use, reagents were equilibrated at room temperature (18–25 °C) for 20–30 min. Standards and diluted serum samples (50–100 µL) were added to specific wells. Dilution ratios were 1:5000 for VTN and APOA2, and 1:8000 for APOA1. Plates were gently shaken, covered, and incubated at 37 °C for 60–120 min, depending on the assay.

After that, wells were washed 3–5 times with 200–350 µL of washing buffer, with each cycle including a 1–2-minute soak. Next, 100 µL of horseradish peroxidase (HRP)-labeled or biotin-labeled antibody was added to each well (except the blanks), followed by a second 60-minute incubation at 37 °C and another round of washing. Substrate solution (50–90 µL) was then added, and plates were incubated in the dark at 37 °C for 15–30 min. The reactions were terminated by adding 50 µL of stop solution to each well.

Optical density (OD) was measured at 450 nm within 5–15 min using a microplate reader. The average OD of the blank was calculated from all readings. A standard curve was generated by plotting standard concentration (x-axis) against corrected absorbance value (y-axis). Sample concentrations were calculated based on this curve. All assays were performed in triplicate to ensure reproducibility, and statistical analyses were conducted to confirm reliability.

ROC curve analysis

To evaluate the diagnostic performance of the ELISA-validated candidate biomarkers, receiver operating characteristic (ROC) curve analysis was performed using GraphPad Prism 9.0. Comparisons were made between PoPH vs. HC and PoPH vs. PHTN groups. The area under the curve (AUC), 95% confidence interval (CI), sensitivity, and specificity were calculated. AUC values greater than 0.5 were considered to indicate diagnostic relevance, while CIs including 0.5 suggested statistical uncertainty. Confidence intervals were computed using the nonparametric bootstrap method.

Protein detection, identification, and quantification

This study utilised Data-Independent Acquisition (DIA) quantitative proteomics technology to analyse serum samples. DIA is a novel mass spectrometry data acquisition approach developed in recent years¹⁴. Phase A (100% water, 0.1% formic acid) and Phase B (80% acetonitrile, 0.1% formic acid). A 10 µL volume of Phase A was used to dissolve the lyophilised powder, followed by centrifugation at 14,000 g for 20 min at 4 °C. The supernatant was used for a 200 ng sample injection. LC-MS analysis was performed using a Vanquish Neo UHPLC system with a C18 pre-column and an ES906 C18 analytical column (Thermo). The Thermo Orbitrap Astral mass spectrometer with Easy-Spray ESI was used for ionization, with an ion spray voltage of 2.0 kV and ion transfer tube temperature of 290 °C. Data were acquired in DIA mode with a full scan range of m/z 380–980 at a resolution of 240,000 (at 200 m/z). MS/MS spectra were collected over 150–2000 m/z with a resolution of 80,000 and a maximum injection time of 3ms. The raw files were searched and analysed using DIA-NN software, with the homo_sapiens_uniprot_2024_07_26_Swissprot—fasta protein database (20436 sequences).

Statistical and bioinformatics analyses

To assess the significance of differences (POPH vs. HC and POPH vs. PHTN), unpaired t-tests were performed on the relative quantification values of each protein, and corresponding p-values were calculated. Proteins with a fold change (FC) > 1.2 and P < 0.05 were considered upregulated, while those with FC < 0.83 and P < 0.05 were considered downregulated. The DEPs were subjected to enrichment analysis using Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG)¹¹, and InterPro databases. Consistent with KEGG citation standards (KANEHISA et al., 2023¹⁵), we systematically incorporated KEGG pathway diagrams to visualize key biological mechanisms. A Venn diagram was used to identify the DEPs between POPH vs. HC and POPH vs. PHTN. All analyses were performed in R (version 4.3.1). Intersecting proteins were further analyzed for protein-protein interactions (PPI) using the String DB protein-protein interaction database (http://string-db.org/), and the results were visualized in Cytoscape. Network topology and node centrality were assessed using the cytoHubba plugin, with hub proteins identified based on maximum clique centrality (MCC) scores. The non-parametric Spearman correlation test was used to analyze the associations between serum candidate biomarkers and RDW and PLT.

Data were collected, organized, and summarized using Excel, and statistical analyses were performed using SPSS version 26.0. Continuous variables with a normal distribution are presented as mean ± standard deviation (SD), while non-normally distributed data are expressed as median ± interquartile range (IQR). Between-group comparisons were conducted using analysis of variance (ANOVA) for normally distributed data and the Kruskal-Wallis H test for non-normally distributed data. Categorical variables were analyzed using the chi-square test. A P value < 0.05 was considered statistically significant. For multiple groups, P values were adjusted accordingly. Serum experiments included parallel controls, and patients were excluded from the study if the coefficient of variation (CV) for the concentration of the same serum sample between groups exceeded 20%. When the serum concentration fell below the assay’s detection limit, a value of zero was assigned.

Ethics approval

This study was approved by the ethics committee of the First Hospital of Jilin University (Approval ID: 2025 − 114). All patient records were anonymized prior to analysis to ensure confidentiality. This study adhered to the ethical principles of the Declaration of Helsinki. Informed consent was obtained from all participants or their legal guardians.

Results

Patient baseline characteristics

The baseline characteristics of the study population, stratified by group (PoPH, PHTN, and HC), are summarized in Table 1. A subsequent analysis of variance revealed that age did not vary significantly among the three groups (p = 0.646). Among the liver function-related indexes, AST, ALT, CHE, ALB, TBIL, and TBA demonstrated significant differences among the three groups (p < 0.05). Specifically, CHE and ALB exhibited significantly higher levels in the HC group compared to the PoPH and PHTN groups, while TBIL and TBA levels were significantly lower in the latter two groups. Furthermore. Among the hematological indicators, HGB, RDW, and PLT demonstrated significant differences among the three groups (p < 0.05). In particular, HGB and PLT levels were significantly higher in the HC group compared to the PoPH and PHTN groups, while RDW levels were significantly lower in the latter two groups. To provide further insight into PoPH-specific changes, pairwise comparisons were performed between PoPH vs. HC and PoPH vs. PHTN for all significantly different indicators (Table 2). These comparisons are crucial for identifying PoPH-specific trends, as our study aims to discover biomarkers that can differentiate PoPH from both healthy individuals and other portal hypertensive conditions.

Table 1.

Analysis of laboratory results in PoPH group, PHTN group and HC group.

Characteristic	group			p-value2
	PoPH, N = 121	PHTN, N = 121	HC, N = 61
age	60.3 ± 11.5	56 ± 8.9	56.7 ± 9.8	0.646
AST	31.5 (27.9)	54.1 (82.6)	20.0 ± 2.0	0.002
ALT	21.3 ± 15.2	33.6 (31.0)	17.4 ± 8.5	0.021
GGT	40.7 (157.0)	69.4 ± 111.1	25.4 ± 15.5	0.133
ALP	93.6 (94.1)	120.4 (54.8)	71.9 (27.5)	0.194
CHE	3225.3 ± 75.9	3248.8 ± 1767.2	8,289.0 ± 3140.0	0.002
ALB	30.7 ± 5.3	30.3 ± 4.9	45.2 ± 1.6	< 0.001
TBIL	36.6 (56.8)	58.3 (114.8)	14.2 ± 3.0	0.015
TBA	33.8 (76.1)	40.1 (176.5)	2.2 ± 0.9	< 0.001
WBC	4.4 (5.8)	5.0 ± 2.6	7.3 ± 1.5	0.241
HGB	84.4 ± 24.9	102.4 ± 29.7	148.2 ± 9.7	0.001
RDW	17.0 ± 2.7	16.6 ± 2.8	12.9 ± 0.7	0.002
PLT	88.5 ± 48.4	77.9 ± 37.5	230.2 ± 32.7	< 0.001
sex				> 0.999
Female	6 (50%)	6 (50%)	3 (50%)
Male	6 (50%)	6 (50%)	3 (50%)
Child-Pugh score				> 0.999
A	2 (16.7%)	2 (16.7%)	0 (NA%)
B	7 (58.3%)	6 (50.0%)	0 (NA%)
C	3 (25.0%)	4 (33.3%)	0 (NA%)

¹For normally distributed data, values are presented as mean ± SD or number (%), For non-normally distributed data, values are presented as median (IQR).

²Kruskal-Wallis rank sum test; Fisher’s exact test

Table 2.

Comparison between the two groups.

Characteristics	POPH (n = 12)1	PHTN (n = 12)1	HC (n = 6)1	P12	P22
RDW (%)	17.0 ± 2.7	16.6 ± 2.8	12.9 ± 0.7	< 0.001	0.66
PLT (10^9/L) ALB (g/L) HGB (g/L)	88.5 ± 48.4 30.7 ± 5.3 84.4 ± 24.9	77.9 ± 37.5 30.3 ± 4.9 102.4 ± 29.7	230.2 ± 32.7 45.2 ± 1.6 148.2 ± 9.7	< 0.0001 0.002 0.002	0.56 0.963 0.318

¹Data are presented as mean ± SD, The p-values were calculated using parametric tests (e.g., t-test or ANOVA) to compare groups.

²P1 and P2 represent the p-values between the POPH and HC groups and between the POPH and PHTN groups, respectively.

Identification of differentially expressed proteins

1,875 proteins were identified in 30 samples, including 1,875 quantitative proteins. The screening of proteins was conducted for significant differentially expressed proteins according to the criterion of a 1.2-fold or more remarkable fold change in expression (up-regulation more significant than 1.2-fold or down-regulation less than 0.83-fold) with a P value < 0.05.

In the POPH vs. HC comparison, 374 upregulated proteins and 115 downregulated proteins were identified, while in the POPH vs. PHTN comparison, 18 upregulated proteins and 38 downregulated proteins were identified (Fig. 1A and B). InterPro analysis revealed that the DEPs (POPH vs. HC) were primarily enriched in enzymatic activity, immune response, protein inhibition, and cell signaling pathways (Fig. 2A). In contrast, the DEPs (POPH vs. PHTN) played key roles in biological processes such as metabolism, protein folding, and hormone synthesis (Fig. 2B). In the KEGG database, the DEPs in both the POPH vs. HC and POPH vs. PHTN groups were involved in biological processes such as immune response, metabolism, and cell signalling (Fig. 3A and B). Similarly, in the GO analysis, the DEPs (POPH vs. HC) were necessary in cellular and molecular functions, metabolic regulation, and immune response. At the same time, the DEPs (POPH vs. PHTN) were mainly involved in multiple biological processes and molecular functions related to cell signalling, ion transport, and regulation of cell growth (Fig. 4A and B).

Fig. 1.

(A) Volcano plot of DEPs between POPH and HC. (B)Volcano plot of DEPs between POPH and PHTN.

Fig. 2.

(A) InterPro analysis (POPH vs. HC). (B) InterPro analysis (POPH vs. PHTN).

Fig. 3.

(A) KEGG analysis (POPH vs. HC). (B) KEGG analysis (POPH vs. PHTN).

Fig. 4.

(A)GO analysis (POPH vs. HC). (B) GO analysis (POPH vs. PHTN).

Identification of candidate markers

By intersecting the two comparisons, we identified 35 genes that distinguish PoPH from HC and differentiate PoPH from PHTN (Fig. 5A). The PPI network of these proteins is shown in Fig. 5B. According to the MCC ranking, the top five proteins were P04004, P02765, P02647, P02763 and P02652. Their protein quantification results and correlations with RDW and PLT are shown in Fig. 6 and Supplementary Fig. 1–2. The correlation results with RDW indicated that P04004 (R = -0.56, P < 0.01), P02647 (R = -0.64, P < 0.001) (Fig. 6A), and P02652 (R = -0.64, P < 0.001) exhibited significant negative correlations, while P02763 (R = 0.54, P < 0.01) (Fig. 6B) demonstrated a positive correlation (Fig. 6C). Similarly, P04004 (R = 0.52, P < 0.01), P02647 (R = 0.44, P = 0.015), and P02652 (R = 0.44, P = 0.015) exhibited significant positive correlations with PLT (Fig. 6A and B, and 6C). Consequently, P04004 (Vitronectin, VTN), P02647 (Apolipoprotein A-I, APOA1), and P02652 (Apolipoprotein A-II, APOA2) were selected as a candidate biomarker for subsequent validation tests.

Fig. 5.

Identification of candidate markers. (A) Venn diagram showing common DEPs. (B) PPI network of the common DEPs. The darker the colour, the larger the MCC value.

Fig. 6.

(A) The quantification of serum P04004 in the POPH, PHTN, and HC groups and its correlation with RDW and PLT. (B) P02647. (C) P02652. (*P < 0.05; **P < 0.01; ***P < 0.001).

Validation of candidate biomarkers

To validate the correlation between the aforementioned candidate biomarkers and PoPH, serum samples from 10 Patients with PoPH, 10 PHTN patients, and 10 healthy controls (HC) were collected and analysed using ELISA to assess the levels of P04004, P02647, and P02652. The results showed that the level of P04004 (vitronectin, VTN) was significantly downregulated in Patients with PoPH compared to the control groups (PHTN and HC) (P < 0.05) (Fig. 7A). Additionally, P02647 (APOA1) exhibited a significant difference between PoPH and PHTN (P < 0.05), but no significant difference was found when compared to HC (Fig. 7B). P02652 (APOA2) showed no significant differences between any of the groups (Fig. 7C).

Fig. 7.

Differences in serum P04004 (VTN), P02647 (APOA1), and P02652 (APOA2) levels between PoPH, PHTN and HC groups. (A) VTN (B) APOA1 (C) APOA2 (*P < 0.05; **P < 0.01; ***P < 0.001).

To further evaluate the diagnostic potential of VTN, receiver operating characteristic (ROC) curve analysis was performed to distinguish patients with PoPH from controls (PHTN and HC). ROC analyses revealed distinct performance profiles among candidate biomarkers. VTN demonstrated robust discriminatory capacity, achieving an AUC of 0.86 (95% CI: 0.696–1.00) for distinguishing PoPH from PHTN and a maximal AUC of 1.00 (95% CI: 1.00–1.00) for PoPH vs. HC. In contrast, APOA1 and APOA2 exhibited more modest performance with AUCs of 0.844 and 0.72 (PoPH vs. PHTN) and 0.517 and 0.675 (PoPH vs. HC), respectively, where confidence intervals for some comparisons encompassed 0.5, indicating limited statistical reliability. (Supplementary Figs. 3–5).

Discussion

Portal pulmonary hypertension (PoPH) is a serious complication of end-stage liver disease with poor survival despite its low prevalence. Its pathophysiology is characterised by its hyperdynamic circulation and reduced peripheral vascular resistance, leading to increased blood flow in the pulmonary circulation and increased shear stress on the pulmonary arteries. This, in turn, results in damage to the endothelium of the pulmonary vasculature, which in turn leads to in situ thrombosis, proliferation of vascular endothelial cells, smooth muscle cells, and fibroblasts, leading to stenosis and occlusion of small pulmonary arteries, and a progressive elevation of pulmonary arterial pressure^5,16. Since the pathogenesis of primary pulmonary hypertension (PoPH) remains to be elucidated, particularly in identifying biomarkers, establishing effective diagnostic methods and interventions has become a significant research challenge in recent years. The present study has been undertaken to explore the potential of biomarkers associated with PoPH to reveal their role in disease development and to provide a theoretical basis for clinical diagnosis and treatment.

In this study, we employed a quantitative proteomics approach, leveraging the DIA technology, to analyze serum samples from patients with PoPH. Our analysis revealed significant differences in the expression levels of 35 proteins compared to those observed in PHTN and HC subjects. Subsequent studies, encompassing GO, KEGG, data enrichment, and protein interaction network analyses, identified candidate proteins from these proteins that strongly correlate with PoPH. Using the Maximal Clique Centrality (MCC) algorithm, we identified five hub proteins as candidate biomarkers potentially involved in PoPH pathogenesis. To preliminarily assess their clinical significance, we analysed correlations between candidate protein levels and key clinical indicators, including platelet count (PLT) and red blood cell distribution width (RDW). The observed associations support their potential relevance in PoPH and provide a foundation for future mechanistic and validation studies.

PLT has been observed in patients with PoPH and may be associated with hepatic insufficiency and hypersplenism, et al. Le noted that the incidence of decreased PLT is higher in PoPH than in idiopathic pulmonary arterial hypertension (IPAH) and connective tissue disease-associated pulmonary hypertension (CTD-PH)¹⁷. Reduced platelets may not only lead to coagulation dysfunction but also increase the risk of thrombosis, which can further exacerbate pulmonary hypertension^18–20. Notably, platelets and their role in the coagulation process have been observed to release a range of inflammatory mediators, including platelet factor 4 (PF4) and platelet-derived growth factor (PDGF). These mediators have been shown to play a pivotal role in promoting vascular remodeling and inflammatory responses, particularly relevant to the pathological progression of PoPH²¹.

In addition to PLT, RDW is an essential clinical indicator reflecting erythrocyte volume distribution heterogeneity. In recent years, it has been shown that RDW is closely associated with various prognostic correlation studies of cardiovascular disease occurrence and liver disease^22–24. Notably, in patients with pulmonary hypertension induced by portal hypertension, RDW levels were found to be significantly higher compared to those observed in patients with idiopathic pulmonary arterial hypertension (IPAH)²⁵. In various diseases, increased RDW may be closely related to inflammation, oxidative stress, and endothelial dysfunction, which go a long way in the pathogenesis of PoPH^26–28.

Following a comprehensive literature review, three candidate biomarkers (APOA1, APOA2, and VTN) were selected for further investigation. ELISA experiments then validated these. Following a comprehensive analysis of the differences between the groups, it was determined that Vitronectin (VTN) has the potential to serve as a significant candidate biomarker in the context of PoPH. Prior studies have demonstrated that, in patients with chronic liver disease, VTN levels are reduced in plasma and increased in liver tissue²⁹. Montaldo et al. further confirmed the significant increase of VTN in cirrhotic liver stroma, thus classifying VTN as an immune biomarker of liver cirrhosis³⁰. Furthermore, in patients with IPAH, Yu et al. found that the expression of VTN was down-regulated using proteomic techniques³¹. Our ELISA-based findings are consistent with these previous observations and suggest a potential association between reduced VTN levels and the pathophysiology of PoPH.

In addition to the ELISA results, we performed ROC curve analysis to evaluate the diagnostic performance of VTN and other candidate biomarkers. VTN showed excellent discriminatory power, with an AUC of 1.00 (95% CI: 1.00–1.00) between PoPH and healthy controls, and an AUC of 0.86 (95% CI: 0.696–1.00) between PoPH and PHTN patients. These values suggest that VTN holds strong potential as a non-invasive biomarker for distinguishing PoPH not only from HC but also from patients with PHTN. In contrast, APOA1 and APOA2 demonstrated lower AUCs and wider confidence intervals, indicating more limited clinical utility. While these findings suggest a strong potential for VTN as a diagnostic biomarker for PoPH, the complete separation may be influenced by the small sample size or possible sampling bias. Therefore, further validation in larger, independent cohorts is necessary.

VTN is a hepatic-synthesized glycoprotein that is widely present in plasma and is involved in some critical biological processes by binding to a variety of ligands, including integrins, plasminogen activator inhibitor-1 (PAI-1) and urokinase plasminogen activator receptor^32,33. VTN plays a vital role in hemostasis and thrombosis, wound healing, and remodeling of blood vessels^32,34. Furthermore, the protein in question has been demonstrated to promote extracellular matrix degradation, a property that has been shown to play a significant role in tumourigenesis^35,36. It has been demonstrated that VTN promotes adhesion and migration of endothelial cells (ECs) and vascular smooth muscle cells (VSMCs) through its RGD sequence binding to specific cell surface receptors, such as integrins αVβ3 and αVβ5, a process that is essential for vascular remodeling³⁷. Garg et al. demonstrated that VTN-deficient mice with vascular smooth muscle cells (VSMCs) migrated faster than wild-type mice³⁸. Furthermore, VTN has been shown to promote cell proliferation and survival by activating multiple signal transduction pathways, including the PI3K/Akt and MAPK pathways, through binding to the corresponding ligands^39,40. These signalling pathways regulate the proliferation and migration of vascular smooth muscle cells and endothelial cells during vascular remodelling, affecting pathological changes in pulmonary arteries. Consequently, VTN dysfunction may contribute to the development of various pathological conditions.

Among the pathological features of PoPH, in situ, thrombus formation within the pulmonary arteries is a key factor, and this process is closely related to the function of VTN. Mathew et al. found that fluid shear stress and fibroblast growth factor-2 (FGF-2) increase the cell-associated VTN through the activation of integrins αVβ5, which promotes the activity of the fibrinolytic protease system⁴¹. Conversely, it has been documented that VTN modulates coagulation by interacting with integrin αIIbβ3 on the platelet surface, thereby facilitating platelet adhesion and aggregation³⁷. An imbalance between the coagulation and fibrinolytic systems determines the eventual formation of a thrombus. Fay et al. observed that VTN-deficient mice formed thrombi more rapidly at the site of arterial injury than wild-type mice, suggesting that VTN plays a critical protective role in antithrombotic effects⁴². This finding was validated by Rehman et al.⁴³. Notably, the in vitro assay of this study showed that VTN in plasma could inhibit platelet aggregation, thereby reducing the risk of thrombosis. Consequently, we hypothesise that the reduced levels of VTN in the plasma of patients with PoPH may contribute to increased platelet aggregation and the development of thrombosis.

In the context of pulmonary hypertension, inflammatory responses have been demonstrated to induce inflammatory cell infiltration within the vessel wall, promoting vascular remodelling and exacerbating pulmonary hypertension. VTN has been shown to activate inflammatory signalling pathways by binding to integrin receptors, thus promoting inflammatory cell infiltration and the release of inflammatory factors. This, in turn, results in the exacerbation of vascular endothelial damage and the promotion of thrombosis. Hayashida et al. found that mice with a VTN gene knockout exhibited lower pro-inflammatory cytokine mRNA expression and less inflammatory cell infiltration in a non-alcoholic fatty liver disease (NASH) model, which attenuated hepatic fibrosis⁴⁴, suggesting that VTN promotes hepatic fibrosis and inflammatory response. Subsequently, in 2021, Chakravarty et al. further demonstrated that levels of inflammatory factors, such as interleukin-1β (IL-1β) and monocyte chemoattractant protein-1 (MCP-1), were elevated in both the aortic wall and plasma of VTN-deficient mice, indicating their involvement in the vascular inflammatory response⁴⁵. These studies suggest that VTN plays a pivotal role in promoting the inflammatory response. These functions may implicate VTN in the complex interplay between liver dysfunction and pulmonary vascular changes observed in PoPH. Nevertheless, the exact role of VTN in this disease context remains to be elucidated. Future mechanistic studies, such as functional assays and animal models, will be necessary to determine whether VTN plays a causal role in PoPH pathogenesis.

Despite the novel insights provided by this study on the potential use of biomarkers for PoPH screening, several limitations must be acknowledged. First and foremost, while the echocardiography-based diagnostic framework aligns with the 2022 ESC/ERS non-invasive pulmonary hypertension criteria, it inherently lacks the hemodynamic precision of right heart catheterization. Second, the relatively small sample size may limit the statistical power and generalizability of the findings. Future studies with larger cohorts are needed to confirm these results. Third, although participants were age and sex-matched across groups to minimize confounding, the influence of additional clinical variables, such as liver disease severity, on protein expression could not be fully evaluated due to limited sample size. Future studies with larger populations are warranted to systematically assess these covariates. Additionally, our findings warrant further validation using complementary experimental approaches, such as Western blotting (WB) or quantitative PCR (qPCR), to confirm the observed protein expression patterns.

To validate the findings of this study, future research should adopt a comprehensive approach encompassing a range of experimental methodologies. Specifically, understanding the mechanism by which VTN contributes to the development of PoPH is a critical area for further exploration. Consequently, subsequent studies should incorporate larger sample sizes and multicentre designs to enhance the robustness of the results. Furthermore, combining in vitro and in vivo experiments will be essential to elucidate the precise mechanism of action of VTN in PoPH and to assess its potential applications in clinical diagnosis and therapeutic interventions.

Conclusion

While this study provides novel proteomic insights, we explicitly acknowledge that PoPH diagnosis relied on echocardiographic criteria rather than right heart catheterization. Our large-scale proteomic analysis was conducted to investigate protein expression differences among PoPH, HC, and PHTN groups. Differentially expressed proteins were identified and found to be involved in immune response, metabolism, and cell signaling. P04004, P02647, and P02652 were significantly correlated with PLT and RDW, indicating their potential as PoPH indicators. ELISA confirmed that P04004 (boswellian protein, VTN) was significantly reduced in the PoPH group, suggesting it as a critical PoPH biomarker. These findings illuminate PoPH pathogenesis and suggest targets for early diagnosis and therapeutic intervention.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

Data collection, and analysis were performed by Jialin Du, Yifu Zhang, Diandian Hao, Hui Wang, Yulin Ren, and Yuze Song. The first draft of the manuscript was written by Jialin Du. Data analysis: Jialin Du. Xiaoyu Wen was responsible for revising and finalizing the manuscript. All authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

The study were sponsored by the Foundation of Science and Technology Commission of Jilin Province (Grants No. 20220203126SF).

Data availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.