PTDSS1 serves as a prognostic biomarker of macrophage infiltration in pan-cancer and promotes hepatocellular carcinoma malignant phenotypes Authors: Chenlu Lan, Zuyin Wan, Haifei Qin, Haixiang Xie, Yu Chen, Chongjiu Qin, Zhongliu Wei, Yongguang Wei, Xin Zhou, Wei Qin, Guangzhi Zhu, Tao Peng Published: 2025 ABSTRACT BackgroundPhosphatidylserine synthase 1 (PTDSS1) inhibitor has been developed and shows promising anti-tumor effect, but existing studies about PTDSS1 remain limited in cancer.MethodsThis study preliminarily analyzed the differential expression, prognostic significance, copy number variation, methylation level, co-expression network and immune correlation of PTDSS1 in pan-cancer. Furthermore, we collected tumor samples and corresponding clinical information of 74 hepatocellular carcinoma (HCC) patients from the First Hospital of Guangxi Medical University (Guangxi cohort) to validate PTDSS1 expression and prognostic value. In vitro experiments were performed to investigate the effect of PTDSS1 knockdown on the malignant phenotypes of HCC cell lines.ResultsThis study identified that PTDSS1 was significantly upregulated across most cancer types and associated with poor prognosis of seven cancer types. Copy number amplification and hypermethylation of PTDSS1 may drive its upregulation. Furthermore, PTDSS1 exhibited a positive correlation with macrophage in various cancers, co-expression network and functional enrichment analysis revealed that PTDSS1 may involve in cell division, cell cycle, DNA repair, cell proliferation, cell migration, cell growth, immune response and WNT signaling pathway. Regarding HCC, the differential expression and prognostic significance of PTDSS1 were verified by Guangxi cohort, single cell sequencing analysis revealed PTDSS1 located in tumor cell and macrophages. In vitro experiments demonstrated that PTDSS1 knockdown suppressed the proliferation, migration, invasion of HCC cells and promoted macrophage repolarization towards the M1 phenotype.ConclusionsPTDSS1 serves as a diagnostic and prognostic biomarker associated with macrophage infiltration, and may be a promising therapeutic target.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12935-025-04144-8. Background Phospholipid serves as an essential structural component of cellular membrane in eukaryotic cell, and plays vital role in molecular transport and signal transduction. Among all phospholipids, phosphatidylserine (PS) is the most abundant negatively-charged phospholipid. Its externalization induces the polarization of tumor-associated macrophage (TAM) toward the immunosuppressive M2 phenotype in the tumor microenvironment, facilitating immune suppression and tumor growth [1]. Emerging evidence demonstrates PS as a promising therapeutic target in contemporary immunotherapy [2, 3]. The biosynthesis of PS is primarily mediated by two enzymes: phosphatidylserine synthase 1 (PTDSS1) and phosphatidylserine synthase 2 (PTDSS2) [4, 5]. Genetic mutation or expression dysregulation of PTDSS1 have been reported to cause congenital diseases and promote tumor progression by disrupting PS biosynthesis [6]. For instance, PTDSS1 mutation drive pathological PS overproduction, a hallmark of Lenz–Majewski syndrome characterized by disrupted calcium homeostasis [7]. The depletion of PTDSS1 in breast cancer reduces PS exposure, thereby suppressing tumor growth and decreasing TAM infiltration [8]. PTDSS1 overexpression is correlated with poor prognosis and facilitates the proliferation of lung adenocarcinoma [9]. Notably, selective PTDSS1 inhibitor has exhibited preclinical efficacy against the PTDSS2 loss- related malignancies [10] or PS metabolism-dependent lymphomas [11]. In contrast to metabolic immunomodulators like isocitrate dehydrogenase1/2, which function through intracellular soluble metabolites [12], PTDSS1 operates via an extracellular mechanism. It controls the abundance of the membrane PS, which, upon externalization, induces immunoregulation. This distinction underscores the unique role of PTDSS1 as an extracellular immunomodulator, defining a novel therapeutic strategy distinct from targeting intracellular metabolites. These findings indicate PTDSS1 as a promising therapeutic target. However, research about PTDSS1 in cancer remains limited. Therefore, we conducted a pan-cancer investigation on PTDSS1 using multiple public databases, aiming to delineate its oncogenic functions and provide theoretical evidence for PTDSS1-targeted therapy development. Methods Differential expression analysis of PTDSS1 The Tumor Immune Estimation Resource (TIMER2.0, http://timer.cistrome.org/) platform was used to compare PTDSS1 mRNA expression between tumor and normal samples of The Cancer Genome Atlas (TCGA) project [13]. RNA-seq data from TCGA tumor samples and Genotype-Tissue Expression (GTEx) normal samples were downloaded from the UCSC Xena database (https://xena.ucsc.edu) to validate the differential expression of PTDSS1. The TCGA and GTEx datasets were merged, processed and normalized to Z-scores using the formula (x − µ)/σ in R software. The University of Alabama at Birmingham Cancer (UALCAN) data analysis portal (https://ualcan.path.uab.edu/) was used to determine the protein expression of PTDSS1 [14]. Prognosis analysis and clinicopathologic feature of PTDSS1 Prognostic information and clinicopathologic data from the TCGA pan-cancer dataset were obtained from the UCSC Xena database, patients were stratified into two groups based on median PTDSS1 RNA expression. Univariate Cox regression analysis was performed to compare overall survival (OS) between PTDSS1-low and -high expression groups. The Kaplan-Meier Plotter (https://kmplot.com/analysis/) and Gene Expression Profiling Interactive Analysis (GEPIA, http://gepia.cancer-pku.cn/) platforms were used to generate survival curves for PTDSS1 [15, 16]. The association between PTDSS1 expression and clinicopathologic variables was assessed across pan-cancer using Wilcoxon test. Genomic alteration and DNA methylation analysis The cBioPortal (https://www.cbioportal.org/) is a comprehensive portal that provides analysis on somatic mutation, copy number variation (CNV), DNA methylation, and mRNA expression profile for a single gene across pan-cancer [17]. We used this tool to explore the genomic variation of PTDSS1. Tumor Immune Dysfunction and Exclusion (TIDE, http://tide.dfci.harvard.edu/) and MethSurv (https://biit.cs.ut.ee/methsurv/) portals were employed to perform survival analysis of CNV and DNA methylation of PTDSS1 in pan-cancer [18, 19]. Co-expression and functional enrichment analysis Co-expression genes of PTDSS1 were identified from 991 pan-cancer samples using the cBioPortal platform. Genes with a correlation coefficient ≥ 0.4 and a p-value of < 0.05 were selected for functional enrichment analysis using the Database for Annotation, Visualization, and Integrated Discovery (DAVID, https://davidbioinformatics.nih.gov/home.jsp). The CancerSEA database (http://biocc.hrbmu.edu.cn/CancerSEA/home.jsp) categorizes 14 functional gene sets related to tumor cell behavior, included angiogenesis, cell cycle, inflammation and others [20]. Gene set variation analysis was performed using the “GSVA” R package to calculate the enrichment scores of these 14 functional gene sets in pan-cancer [21]. The enrichment scores were further normalized to Z-scores by the scale function in R and subsequently were adopted to implement the correlation analysis with PTDSS1 expression. Immune correlation analysis Based on genome wide RNA-seq data from TCGA pan-cancer project, immune microenvironment scores across different cancer types were evaluated by the “ESTIMATE” R package. Immune cell infiltration scores were quantified through three independent algorithms: CIBERSORT, ssGSEA, and xCell. Simple nucleotide variation data of TCGA pan-cancer project was obtained from the Genomic Data Commons database (https://portal.gdc.cancer.gov) and processed to calculate the tumor mutation burden (TMB) using “mafools” R package, while microsatellite instability (MSI) was determined for each cancer type. Subsequently, Spearman’s rank correlation test was performed between PTDSS1 RNA expression and these immunological parameters. Comprehensive analysis of PTDSS1 in HCC Validation cohort was obtained from International Cancer Genome Consortium (ICGC, https://dcc.icgc.org/projects/LIRI-JP) to confirm the differential RNA expression and prognostic significance of PTDSS1. The Human Protein Atlas (HPA, https://www.proteinatlas.org/) was applied to examine PTDSS1 protein expression in HCC. Multivariate Cox proportional hazards regression analysis was implemented to assess whether PTDSS1 expression serves as an independent prognostic factor. Furthermore, the abundance of 22 immune cells types (CIBERSORT algorithms) was compared between PTDSS1-high and -low expression group. The TIMER2.0 platform was utilized to evaluate correlations among PTDSS1 mRNA expression, CNV, and immune cell infiltration. To investigate potential mechanisms of PTDSS1 in HCC, we identified differential expression genes between PTDSS1-high and -low expression groups, and performed functional enrichment analysis using the DAVID platform. Finally, we conducted gene set enrichment analysis (GSEA, version 4.3.2) to characterize the enriched gene sets between PTDSS1-high and -low expression groups. Spatial transcriptomics analysis Spatial transcriptomics and single-cell transcriptomic data were derived from the China National Center for Bioinformation (dataset: HRA000437, https://ngdc.cncb.ac.cn/gsa-human/browse/HRA000437). The specimen sources and processing protocols have been described in the original study [22]. Two cases containing both HCC and adjacent normal liver tissue sections met our require. Using the “Cottrazm” R package, we delineated the transitional zones between HCC and adjacent normal liver tissue regions in these sections. Cellular composition was identified through deconvolution analysis, followed by reconstruction of cell type-specific gene expression profiles at sub-spot resolution [23]. Finally, we compared the PTDSS1 expression levels across cell type-specific regions of each tissue section. Single cell sequencing analysis The GSE140228 and GSE166635 datasets were downloaded from the Gene Expression Omnibus database (https://www.ncbi.nlm.nih.gov/gds/). We performed uniform manifold approximation and projection for dimensionality reduction and data visualization. Cell types were annotated according to canonical marker gene expression patterns. Differential expression of PTDSS1 across cell populations was evaluated using the Kruskal-Wallis test. Validation of PTDSS1 expression and prognostic significance Twenty pairs of HCC and adjacent para-carcinoma samples were collected from the First Hospital of Guangxi Medical University for immunohistochemistry (IHC) and quantitative polymerase chain reaction (qPCR) validation. Additionally, we collected tumor samples and corresponding clinical information of 74 HCC patients from the First Hospital of Guangxi Medical University in 2020 for qPCR and prognostic validation (Guangxi cohort). All included patients were pathologically diagnosed with HCC after hepatectomy, this study was approved by the Ethics Committee of the First Affiliated Hospital of Guangxi Medical University (Approval Number: 2025-E0891). Construction of prognostic nomogram model The “rms” R package was used to create prognostic nomogram model which combined PTDSS1 expression group with tumor stage, and calibration curve was applied to appraise the accuracy of nomogram model. Moreover, receiver operating characteristic (ROC) curve was generated to estimate the prognostic value of nomogram model, PTDSS1 expression and tumor stage respectively. Quantitative real-time polymerase chain reaction The Trizol reagent (Invitrogen, USA) was used to extract the RNA of cells and tissues, the PrimeScript™ RT reagent kit (Takara, China) was used to reversely transcribed RNA into cDNA. QPCR was performed used the FastStart Universal SYBR® Green Master Mix (Roche, Germany), the relative mRNA expression of PTDSS1 was calculated by the 2−∆∆CT way. The primer sequences used in this study are listed in Table S1. Immunohistochemistry 20 pairs of HCC and para-cancerous tissue were formalin-fixed and paraffin-embedded followed by cutting into sections. Based on the instructions of a standardized two-step IHC kit (ZSGB‐Bio, China), the tissue sections were incubated overnight with anti‐PTDSS1 antibody at 4 °C, tissue sections were subjected to antigen repair and incubated with the primary anti-PTDSS1 antibody (Abmart, PH04305, 1:100) at 4 °C overnight. Subsequently, an enhanced enzyme‐labeled goat anti‐mouse/rabbit IgG polymer was used to incubate for 30 min. Tissue sections were then stained with DAB chromogenic kit (ZSGB‐Bio, China) and counterstained with hematoxylin. Finally, tissue sections were photographed under a microscope and analyzed by Image J software. Cell culture and transfection Functional experiments were performed using Huh-7 and Hep3B cell lines, which was purchased from the Chinese Academy of Sciences (Shanghai, China). The Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (FBS) was used to culture the Huh-7 and Hep3B cell at 37 °C and 5% CO₂ atmosphere. For PTDSS1 silencing, a sequence-specific siRNA kit designed and synthesized by Sangon Biotech (Shanghai, China) and reagent was applied to perform cell transfection according to the manufacturer’s protocol. The transfection efficiency was validated by PCR and WB after 24 and 48 h respectively. Co-culture of THP-1-derived macrophages with hepatocellular carcinoma cells induces polarization THP-1 cells (NCM Biotech, China) were cultured in RPMI-1640 supplemented with 10% FBS until reaching logarithmic growth phase. Cells (5 × 105) were seeded per well in 6-well plates and treated with phorbol myristate acetate (PMA, 100 ng/ml) for 48 h to induce differentiation into adherent M0 macrophages. After induction, the PMA-containing supernatant was removed, cells were washed once with PBS, and fresh RPMI-1640 without PMA was added for a 24 h resting period to minimize residual activation and stabilize the phenotype. After resting, the medium was replaced with a 1:1 mixture of DMEM and RPMI-1640 containing 10% fetal bovine serum. Log-phase Hep-3B-NC, Hep-3B-siPTDSS1, Huh-7-NC, and Huh-7-siPTDSS1 cells were added. Tumor cells and M0 macrophages were co-cultured at ratios of 1:1 (5 × 10⁵ cells/well), with a total volume of 2 ml per well. Cultures were maintained at 37 °C with 5% CO₂ for 48 h before sample collection. No exogenous polarizing factors were added to observe the natural effects of tumor cell-secreted factors and cell contact on THP-1-M0 polarization. The control group included M0 macrophages and tumor cells cultured separately. At the end of the experiment, cells were collected and analyzed by qPCR for M1/M2 marker genes. Cell counting kit-8 assay Cell proliferation was assessed using the Cell Counting Kit-8 (CCK-8, Beyotime, China) assay. Huh-7 cell and Hep-3B cell were plated in 96-well plates with the density of 5000 cells/well, 10 µL CCK-8 reagent was added to incubate for 2 h (37 °C, 5% CO₂), and then absorbance was quantified used a Varioskan LUX enzyme labeler (Thermo Fisher Scientific, Waltham, MA, USA) at 0, 24, 48, 72, and 96 h. Colony formation assay Colony formation assay was also used to detect the effect of PTDSS1 knockdown on cell proliferation, the Hep-3B and Huh-7 cell lines were cultured into 6-well plates with 500 cells/well and 1000 cells/well respectively. After 10–14 days, the cell colonies were fixed, stained, and photographed. Wound-healing assay Linear cellular scratches were created using sterile 200 µL pipette tips, cells were cultured with serum-free DMEM medium. Initial (0 h) wound margins were marked and photographed under microscope followed by recording at 24 h and 48 h. The migration area and cell migration rate were analyzed using ImageJ software. Transwell assay The Matrigel (Corning, USA) was diluted and added to the upper chamber of 24-well transwell at 37 °C for 1 h. Subsequently, 200 µL serum-free DMEM medium contained 1 × 10⁵ cells was seeded in the upper chamber, and 800 µL 10% FBS DMEM medium was added to the lower chamber. Following incubation for 24 h, the membrane of upper chamber was fixed with 4% paraformaldehyde for 30 min, and stained with 0.1% crystal violet stain solution for 20 min. Then, cells were counted under microscope to compare the invasion ability of different groups. Transwell migration assay was performed following the invasion assay, but without the Matrigel coating of upper chamber. Statistical analysis Comparison of binary categories was performed using the non-parametric Wilcoxon rank-sum test in R software (version 4.3.2), while Student’s t-test was applied for normally distributed data in SPSS 22.0 software. Kaplan-Meier, univariate and multivariate Cox regression survival analyses were conducted using the “survival” and “survminer” packages in R. The diagnostic and prognostic performance was evaluated using the “pROC” package. Correlation analyses employed Spearman’s rank correlation test for continuous variables and chi-square test for categorical variables in R software. Statistical significance was defined as p-value < 0.05, while functional enrichment analyses required false discovery rate < 0.25. For in vitro experiments, all assays were performed with at least three independent replicates, and group comparisons were analyzed using Student’s t-test. Results Expression profile and diagnostic value of PTDSS1 Pan-cancer analyses using the TIMER2.0 and UALCAN portals revealed that PTDSS1 was frequently upregulated across multiple cancer types at the mRNA and protein levels (Fig. 1A-C). Additionally, ROC curve analysis demonstrated its strong diagnostic potential in specific cancers with the area under curve (AUC) > 0.8 (Figure S1). Fig. 1PTDSS1 expression level in pan-cancer. A The mRNA expression of PTDSS1 between tumor and normal tissues was compared via TIMER2.0 platform. B The differential mRNA expression of PTDSS1 between tumor and normal tissues was validated used TCGA pan-cancer and GTEx datasets. C The protein expression level of PTDSS1 was analyzed by UALCAN portal. *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant PTDSS1 expression level in pan-cancer. A The mRNA expression of PTDSS1 between tumor and normal tissues was compared via TIMER2.0 platform. B The differential mRNA expression of PTDSS1 between tumor and normal tissues was validated used TCGA pan-cancer and GTEx datasets. C The protein expression level of PTDSS1 was analyzed by UALCAN portal. *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant Prognostic significance and clinicopathologic features of PTDSS1 Univariate Cox analysis revealed that high PTDSS1 expression was significantly associated with shorter OS in lower-grade glioma (LGG), liver hepatocellular carcinoma (LIHC), lung adenocarcinoma (LUAD), sarcoma (SARC), skin cutaneous melanoma (SKCM), uterine corpus endometrial carcinoma (UCEC) and uveal melanoma (UVM) (Fig. 2A), a finding corroborated by Kaplan-Meier Plotter and GEPIA platforms (Fig. 2B-C). Moreover, high PTDSS1 levels correlated with advanced tumor stage and poor differentiation (Fig. 2D), and showed modest prognostic accuracy (AUC > 0.6) in LGG, LIHC, UCEC and UVM (Figure S2). Fig. 2The prognostic significance and clinicopathologic feature of PTDSS1 in pan-cancer. A Univariate Cox regression analysis of PTDSS1-low and -high expression groups was performed used TCGA pan-cancer datasets. B-C The survival curves of PTDSS1-low and -high expression groups on Kaplan Meier Plotter B and GEPIA C portals. D The mRNA expression of PTDSS1 in different tumor stage and histological differentiation in TCGA pan-cancer datasets. *P < 0.05, **P < 0.01, ***P < 0.001 The prognostic significance and clinicopathologic feature of PTDSS1 in pan-cancer. A Univariate Cox regression analysis of PTDSS1-low and -high expression groups was performed used TCGA pan-cancer datasets. B-C The survival curves of PTDSS1-low and -high expression groups on Kaplan Meier Plotter B and GEPIA C portals. D The mRNA expression of PTDSS1 in different tumor stage and histological differentiation in TCGA pan-cancer datasets. *P < 0.05, **P < 0.01, ***P < 0.001 Genomic amplification and hypomethylation of PTDSS1 drove expression upregulation and poor prognosis Genomic analysis via cBioPortal identified PTDSS1 amplification as its predominant CNV type, with frequencies > 20% in several cancers (Fig. 3A). This amplification correlated with highest mRNA levels (Fig. 3B), and high CNV burden predicted poorer OS in specific cancers (Fig. 3C). Additionally, PTDSS1 expression showed a negative correlation with promoter methylation (Fig. 3D), and this hypomethylation, in turn, was associated with adverse prognosis (Fig. 3E), suggesting both CNV and methylation drive its upregulation. Fig. 3The CNVs and DNA methylation of PTDSS1 in in pan-cancer. A The CNV and mutation frequency of PTDSS1 on cBioPortal platform. B The mRNA expression level of PTDSS1 between different CNV types in cancer was displayed via cBioPortal platform. C Kaplan Meier survival analysis of PTDSS1 CNV frequency via TIDE portal. D The correlation analysis between PTDSS1 mRNA expression and its DNA methylation via cBioPortal platform. E Kaplan Meier survival analysis of PTDSS1 DNA methylation via MethSurv portal The CNVs and DNA methylation of PTDSS1 in in pan-cancer. A The CNV and mutation frequency of PTDSS1 on cBioPortal platform. B The mRNA expression level of PTDSS1 between different CNV types in cancer was displayed via cBioPortal platform. C Kaplan Meier survival analysis of PTDSS1 CNV frequency via TIDE portal. D The correlation analysis between PTDSS1 mRNA expression and its DNA methylation via cBioPortal platform. E Kaplan Meier survival analysis of PTDSS1 DNA methylation via MethSurv portal PTDSS1 May involve in mitosis and genomic stability Co-expression analysis revealed there were positive associations between PTDSS1 and cell cycle regulators such as MCM4, PLK1, CDC45, BUB1, and MCM10 in pan-cancer (Fig. 4A-B). Functional enrichment analysis confirmed significant associations with cell division, DNA repair, and the p53 signaling pathway (Fig. 4C). Consistent with this, GSVA showed PTDSS1 expression correlated with cell cycle and DNA repair activity (Fig. 4D), implicating it in mitotic regulation and genomic stability. Fig. 4Co-expression and functional enrichment analysis for PTDSS1 in pan-cancer. A-B Top 50 positive (A) and negative (B) co-expression genes of PTDSS1 among pan-cancer via correlation analysis on cBioPortal platform. C Functional enrichment analysis of positive co-expression genes on DAVID portal. D Correlation analysis between PTDSS1 expression and the enrichment scores of cancer-related functional gene sets via TCGA pan-cancer datasets Co-expression and functional enrichment analysis for PTDSS1 in pan-cancer. A-B Top 50 positive (A) and negative (B) co-expression genes of PTDSS1 among pan-cancer via correlation analysis on cBioPortal platform. C Functional enrichment analysis of positive co-expression genes on DAVID portal. D Correlation analysis between PTDSS1 expression and the enrichment scores of cancer-related functional gene sets via TCGA pan-cancer datasets PTDSS1 was positively correlated with macrophage infiltration Immune correlation analyses revealed that PTDSS1 was generally negatively correlated with ESTIMATE immune scores, except in UVM and kidney renal clear cell carcinoma (Figure S3A). Analysis of immune cell infiltration via multiple algorithms showed PTDSS1 positively associated with macrophage and Th2 cell infiltration, but negatively with CD8 + T cells and other effector immune cells (Figure S3B-C, S4). Furthermore, PTDSS1 exhibited divergent correlation patterns with immunomodulatory and immune checkpoint genes across different cancers (Figure S5), consistent with the ESTIMATE immune score trends. Notably, in Thyroid carcinoma, PTDSS1 showed negative correlations with both MSI and TMB (Figure S3D-E), while its concordant positive correlations with immune checkpoint genes and macrophage infiltration suggest a potential role in promoting immune evasion and immunotherapy resistance. Upregulation of PTDSS1 in HCC tissues We selected HCC from OS-related cancers for validation. PTDSS1 was consistently overexpressed in HCC tumors at the mRNA (TCGA, ICGC cohorts; Fig. 5A-C) and protein levels (HPA portal; Fig. 5E-F), showing favorable diagnostic value (Fig. 5D). High PTDSS1 correlated with poor differentiation and advanced stage (Fig. 5G), and spatial transcriptomics confirmed its tumor-specific upregulation (Fig. 5H-I). Fig. 5The PTDSS1 expression in HCC. A-B Wilcoxon test compared PTDSS1 expression of whole samples (A) and paired samples (B) in TCGA-LIHC cohort. C Wilcoxon test compared PTDSS1 expression of whole samples in ICGC-LIRI-JP cohort. D The diagnostic ROC curves of PTDSS1 in TCGA and ICGC cohorts. E-F HPA database displayed the protein expression of PTDSS1 in normal live tissues E and HCC tissues F. G The mRNA expression of PTDSS1 in different tumor stage and histological differentiation in HCC. H-I The mRNA expression of PTDSS1 from two samples of the public spatial transcriptomics dataset- HRA000437. *P < 0.05, **P < 0.01, ***P < 0.001 The PTDSS1 expression in HCC. A-B Wilcoxon test compared PTDSS1 expression of whole samples (A) and paired samples (B) in TCGA-LIHC cohort. C Wilcoxon test compared PTDSS1 expression of whole samples in ICGC-LIRI-JP cohort. D The diagnostic ROC curves of PTDSS1 in TCGA and ICGC cohorts. E-F HPA database displayed the protein expression of PTDSS1 in normal live tissues E and HCC tissues F. G The mRNA expression of PTDSS1 in different tumor stage and histological differentiation in HCC. H-I The mRNA expression of PTDSS1 from two samples of the public spatial transcriptomics dataset- HRA000437. *P < 0.05, **P < 0.01, ***P < 0.001 PTDSS1 overexpression predicts poor prognosis in HCC Patients were stratified by median PTDSS1 expression into high and low groups (Figure S6A). The high-expression group showed significantly shorter OS (Figure S6B), and PTDSS1 exhibited modest prognostic accuracy (Figure S6C), which was validated in the ICGC cohort (Figure S6D-F). Notably, multivariate Cox regression confirmed PTDSS1 as an independent prognostic factor (Figure S6G-H). PTDSS1-related DEGs involve in oncogenic pathways in HCC Differential gene expression identified 163 significantly upregulated genes and 158 downregulated genes in the high-expression group (Figure S7A). DAVID analysis revealed that upregulated genes involved in cell proliferation, cell migration, cell growth, immune response, IL-17 signaling pathway, Wnt signaling pathway, and chemokine signaling pathway, whereas downregulated genes were linked to metabolic pathways (Figure S7B-C). GSEA further corroborated these findings (Figure S7D-E). Combined PTDSS1 overexpression with TP53/CTNNB1 mutations drive poor prognosis in HCC Building on the pan-cancer correlation between PTDSS1 and TMB in HCC, we analyzed the mutation landscape. The PTDSS1-high group exhibited a higher TMB (Figure S7F), primarily driven by increased mutation frequencies of TP53 (37% vs. 20%) and CTNNB1 (31% vs. 21%) (Figure S7G-H). Consequently, patients with both high PTDSS1 and high TMB or mutations in TP53/CTNNB1 had the poorest overall survival (Figure S7I-K). This discovery suggests that PTDSS1 overexpression may be positively related to genomic instability. PTDSS1 overexpression correlates with high macrophage abundance in HCC Analysis of HCC cohorts revealed that high PTDSS1 expression correlated with an immunosuppressive microenvironment, characterized by increased macrophage and decreased activated NK cell infiltration (Fig. 6A-B). This positive association with macrophages was validated by TIMER2.0 platform (Fig. 6C) and was more pronounced in cases with PTDSS1 copy number amplification (Fig. 6D). Patients with both high PTDSS1 and macrophage infiltration had the poorest survival (Fig. 6E). Single-cell RNA sequencing confirmed PTDSS1 expression in tumor cells, macrophages, and dendritic cells, suggesting its role in shaping the immunosuppressive microenvironment (Fig. 6F-K). Fig. 6The immune landscapes of PTDSS1 in HCC. A-B Differences of immune cell abundance between PTDSS1-high and -low groups in TGCA A and ICGC B cohorts. C The correlation analysis between PTDSS1 and macrophage cell via TIMER2.0 platform. D The macrophage cell abundance in different PTDSS1 CNV states. E Kaplan Meier survival curves of PTDSS1 expression combined with macrophage cell abundance. F and I Dimensionality reduction, clustering, and cell annotation of GSE140228_10X F and GSE166635 datasets I. G and J Heat maps of PTDSS1 expression in GSE140228_10X (G) and GSE166635 J datasets. H and K The differences of PTDSS1 expression between different cell types in GSE140228_10X (H) and GSE166635 K datasets The immune landscapes of PTDSS1 in HCC. A-B Differences of immune cell abundance between PTDSS1-high and -low groups in TGCA A and ICGC B cohorts. C The correlation analysis between PTDSS1 and macrophage cell via TIMER2.0 platform. D The macrophage cell abundance in different PTDSS1 CNV states. E Kaplan Meier survival curves of PTDSS1 expression combined with macrophage cell abundance. F and I Dimensionality reduction, clustering, and cell annotation of GSE140228_10X F and GSE166635 datasets I. G and J Heat maps of PTDSS1 expression in GSE140228_10X (G) and GSE166635 J datasets. H and K The differences of PTDSS1 expression between different cell types in GSE140228_10X (H) and GSE166635 K datasets PTDSS1 knockdown promoted macrophage repolarization towards the M1 phenotype Cell adhesion is a well-established hallmark of monocyte differentiation into macrophages [24]. After 48 h of PMA induction, most THP-1 cells transitioned from a suspended (Fig. 7A) to an adherent state (Fig. 7B), indicating successful macrophage differentiation. Fig. 7Effect of PTDSS1 Knockdown on Macrophage Polarization. A THP-1 cells before PMA induction, exhibiting a suspension growth state. B THP-1 cells differentiated into M0 macrophages, displaying a typical adherent morphology. C-D Microscopic images after 48 h of co-culture of siRNA-transfected Huh-7 C and Hep-3B D cells with THP-1-derived macrophages. E-H Effect of PTDSS1 knockdown on the expression of M1 macrophage markers. I-L Effect of PTDSS1 knockdown on the expression of M2 macrophage markers. All assays were performed with at least three independent replicates, and group comparisons were analyzed using Student’s t-test. *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant Effect of PTDSS1 Knockdown on Macrophage Polarization. A THP-1 cells before PMA induction, exhibiting a suspension growth state. B THP-1 cells differentiated into M0 macrophages, displaying a typical adherent morphology. C-D Microscopic images after 48 h of co-culture of siRNA-transfected Huh-7 C and Hep-3B D cells with THP-1-derived macrophages. E-H Effect of PTDSS1 knockdown on the expression of M1 macrophage markers. I-L Effect of PTDSS1 knockdown on the expression of M2 macrophage markers. All assays were performed with at least three independent replicates, and group comparisons were analyzed using Student’s t-test. *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant We then co-cultured siRNA-transfected Huh-7 (Fig. 7C) and Hep-3B (Fig. 7D) cells with the adherent M0 macrophages for 48 h. PCR analysis revealed that, compared to the co-cultured ncPTDSS1 group, the co-cultured siPTDSS1 group exhibited a significant upregulation of CD86 and NOS2 (Fig. 7E-H) with a downregulation of ARG1 (Fig. 7J and L). Notably, we found that CD206 expression was significantly upregulated in co-cultured Hep-3B-siPTDSS1 group (Fig. 7K), contrary to the downregulation trend seen in Huh-7 co-cultures (Fig. 7I). Despite this cell line-specific discrepancy regarding CD206, the overall pattern suggested that PTDSS1 knockdown promoted macrophage repolarization towards the M1 phenotype. Validation of PTDSS1 overexpression and prognostic impact in the Guangxi HCC cohort To validate bioinformatics findings, validation in the Guangxi cohort confirmed significant upregulation of PTDSS1 in HCC at both mRNA and protein levels (Fig. 8A, B, H). The high-expression group was associated with chronic hepatitis and moderate differentiation (Fig. 8C-D, Table S2), and exhibited significantly shorter overall survival (Fig. 8E-F, Table S3). ROC analysis confirmed PTDSS1 had a value for predicting OS with the AUC = 0.648 (Fig. 8G), multivariate Cox regression analysis supported PTDSS1 as an independent prognostic factor (Table S4). Fig. 8Validation of the expression and prognostic value of PTDSS1 in Guangxi cohort. A The difference of PTDSS1 mRNA expression between 74 HCC samples and 20 normal liver samples via PCR validation. B The difference of PTDSS1 mRNA expression between 20 pairs of HCC and normal liver samples via PCR validation. C-D Correlation analysis between PTDSS1 groups and chronic hepatitis C, WHO differentiation grade D in Guangxi cohort. E-F Kaplan Meier survival curves of PTDSS1 mRNA expression grouped by the median E and optimal threshold of ROC curve F. G The diagnostic and prognostic ROC curves of PTDSS1. H PTDSS1 IHC images and its protein expression difference between 20 pairs of HCC and normal liver tissues. All assays were performed with at least three independent replicates, and group comparisons were analyzed using Student’s t-test, *P < 0.05, **P < 0.01 Validation of the expression and prognostic value of PTDSS1 in Guangxi cohort. A The difference of PTDSS1 mRNA expression between 74 HCC samples and 20 normal liver samples via PCR validation. B The difference of PTDSS1 mRNA expression between 20 pairs of HCC and normal liver samples via PCR validation. C-D Correlation analysis between PTDSS1 groups and chronic hepatitis C, WHO differentiation grade D in Guangxi cohort. E-F Kaplan Meier survival curves of PTDSS1 mRNA expression grouped by the median E and optimal threshold of ROC curve F. G The diagnostic and prognostic ROC curves of PTDSS1. H PTDSS1 IHC images and its protein expression difference between 20 pairs of HCC and normal liver tissues. All assays were performed with at least three independent replicates, and group comparisons were analyzed using Student’s t-test, *P < 0.05, **P < 0.01 Nomogram models demonstrate stronger prognostic power than tumor stage alone The nomogram models integrating tumor stage and PTDSS1 expression group (Fig. 9A, D, G) exhibited superior predictive accuracy through calibration curves validation across TCGA, ICGC and Guangxi cohorts (Fig. 9B, E, H). ROC analysis confirmed enhanced prognostic performance of the integrated models compared to tumor stage alone (Fig. 9C, F, I). Fig. 9Construction of prognostic nomogram model. A-C nomogram A, calibration curve B and prognostic receiver operator curves C of TCGA cohort. D-F nomogram D, calibration curve E and prognostic receiver operator curves F of ICGC cohort. G-I nomogram G, calibration curve H and prognostic receiver operator curves I of Guangxi cohort Construction of prognostic nomogram model. A-C nomogram A, calibration curve B and prognostic receiver operator curves C of TCGA cohort. D-F nomogram D, calibration curve E and prognostic receiver operator curves F of ICGC cohort. G-I nomogram G, calibration curve H and prognostic receiver operator curves I of Guangxi cohort PTDSS1 Silencing suppresses malignant phenotypes of HCC cells To validate functional enrichment findings, we conducted a series of functional validation experiments. PCR revealed that PTDSS1-siRNA#2 and -siRNA#3 had better knockdown efficiency than PTDSS1-siRNA#1 (Fig. 10A). Subsequently, CCK-8 assay (Fig. 10B), colony formation assay (Fig. 10C-D), wound healing assay (Fig. 10E-G), transwell invasion and migration assays (Fig. 10H) demonstrated that PTDSS1 silencing significantly inhibited the proliferation, migration, and invasion of Huh-7 and Hep-3B cell lines. Fig. 10Investigation of the malignant phenotypes of PTDSS1 knockdown-HCC cells in Vitro. A Relative mRNA expression of PTDSS1 after transfection with PTDSS1-siRNA in HCC cells. B Tumor growth curves of HCC cells in CCK8 assay. C Photographs of colony formation assay. D Statistical comparison of cell counts in colony formation assay. E,G Microscopy images of wound healing assay. G Statistical comparison of cell migration area in wound healing assay. H Transwell assay showed that PTDSS1 knockdown inhibited the migration and invasion ability of HCC cells. All assays were performed with at least three independent replicates, and group comparisons were analyzed using Student’s t-test, **P < 0.01, ***P < 0.001 Investigation of the malignant phenotypes of PTDSS1 knockdown-HCC cells in Vitro. A Relative mRNA expression of PTDSS1 after transfection with PTDSS1-siRNA in HCC cells. B Tumor growth curves of HCC cells in CCK8 assay. C Photographs of colony formation assay. D Statistical comparison of cell counts in colony formation assay. E,G Microscopy images of wound healing assay. G Statistical comparison of cell migration area in wound healing assay. H Transwell assay showed that PTDSS1 knockdown inhibited the migration and invasion ability of HCC cells. All assays were performed with at least three independent replicates, and group comparisons were analyzed using Student’s t-test, **P < 0.01, ***P < 0.001 Discussion In this study, pan-cancer analysis demonstrated that PTDSS1 was significantly overexpressed in a majority of cancer types. This upregulation may be driven by copy number amplification and promoter hypomethylation of PTDSS1. Gene upregulation in cancer can be directly driven by genomic and epigenetic alterations, particularly copy number amplification and promoter hypomethylation [25, 26]. In addition, survival analysis discovered that PTDSS1 overexpression was associated with poor prognosis of some cancer types, its prognostic performance was further evaluated and verified by multiple cohorts of HCC, including TCGA, ICGC, and Guangxi cohorts. We found that PTDSS1 was an independent and potential prognostic biomarker for HCC. Consistent with previous studies, PTDSS1 overexpression was reported be associated with unfavorable outcomes in lung cancer [27], bladder urothelial carcinoma [28], and pancreatic cancer [29], implicating it as a potential pan-cancer oncogene. This study selected HCC for further validation for two reasons. First, HCC possesses a unique tumor microenvironment characterized by a rich infiltration of TAMs. The dynamic polarization of TAMs into M1 or M2 phenotypes plays a pivotal role in remodeling the immune landscape, which is closely associated with tumor progression and immunotherapy resistance [30]. The immune correlation analysis revealed a positive association between PTDSS1 and macrophage in our study, thus HCC is an ideal model to investigate this mechanism. Second, lipid metabolic reprogramming is a recognized hallmark of HCC, which is known to influence the pathogenesis, progression, and immunotherapy response of HCC [31]. As the key enzyme catalyzing PS biosynthesis, PTDSS1 sits at a critical junction between lipid metabolism and immune regulation in HCC [8]. Regarding the functional role of PTDSS1 in cancer, functional enrichment analyses revealed that PTDSS1 upregulation may drive tumor progression and immune response by influencing cell cycle, cell proliferation, p53 signaling pathway, IL-17 signaling pathway, and WNT signaling pathway. Our findings that PTDSS1 knockdown inhibited HCC cell malignancy confirm its role in promoting tumor development, consistent with a prior study in LUAD [9]. In addition to its direct role in driving tumor progression, an equally critical aspect of PTDSS1 function lies in its ability to sculpt an immunosuppressive microenvironment. Current evidence suggests that PTDSS1 overexpression enhances PS synthesis, which amplifies PS externalization during apoptosis. This increased PS exposure promotes the polarization of TAMs toward the M2 phenotype, thereby fostering an immunosuppressive tumor microenvironment [32]. Our immune correlation analysis disclosed that PTDSS1 was positively correlated with the infiltration of macrophage cell and Th2 cell, while had a negative association with CD8+ T cell, dendritic cell, activated NK cell, and Th1 cell. A recent study found that PTDSS1 deficiency in tumor cells increased expression of MHC-I, and enhanced cytotoxicity of CD8+T cells [33]. This directly aligns with our observation of a negative correlation between PTDSS1 and CD8⁺ T cell infiltration, suggesting that PTDSS1 overexpression impairs the function of cytotoxic T cells and induces immunotherapy resistance. Second, and more critically, their single-cell RNA sequencing data demonstrated that loss of PTDSS1 skewed macrophage polarization towards an M1 phenotype, characterized by increased frequencies of iNOS⁺ myeloid cells and elevated expression of CD86 and CXCL9, with reduced expression of MRC1 [33]. Our co-culture experiments confirmed that PTDSS1 knockdown in HCC cells elevated M1 but suppressed M2 macrophage markers. This provides direct experimental validation for our bioinformatic association between PTDSS1 and macrophage abundance, supporting that PTDSS1 overexpression promotes an M2-like immunosuppressive phenotype. Regrettably, our study did not measure PS synthesis following PTDSS1 knockdown, and existing evidence remains fragmented, lacking comprehensive validation of the complete PTDSS1-PS-M2 polarization-immunosuppression axis. Future studies should prioritize elucidating this mechanistic pathway. This focus is particularly warranted for HCC, as evidence from PTDSS1-deficient mouse models shows that the liver experiences the most dramatic loss of PS [34]. Bavituximab, a prominent PS-targeting drug, exerts its antitumor immune effects by blocking PS-mediated immunosuppressive signaling [35], thereby promoting M1 macrophage polarization and cytotoxicity against tumor-associated endothelial cells [36]. Clinical trial has demonstrated that Bavituximab combined with chemotherapy achieves higher objective response rates with a favorable safety in cancers such as non-small cell lung cancer [37, 38], breast cancer [39], and HCC [40]. Besides, Bavituximab can be combined with immune checkpoint inhibitors to exert synergistic antitumor effects of HCC [3, 41]. However, a randomized phase 3 trial showed that Bavituximab failed to improve OS compared to docetaxel monotherapy in non-small cell lung cancer [42], suggesting alternative approaches are necessary for targeting the PS pathway. Indeed, a recent study has demonstrated that the PTDSS1 inhibitor, DS55980254, enhances the efficacy of anti-PD-1 therapy, with significant tumor growth inhibition and prolonged survival [33]. Furthermore, PTDSS1 inhibitors can synergize with chemotherapy to mediate antitumor effects. A key mechanism of chemotherapy-induced immunosuppression is the robust externalization of PS on tumor cells, a signal exploited by the TAM receptor pathway on macrophages to establish an anti-inflammatory microenvironment [43]. The PTDSS1 inhibitor counteracts this by reducing PS levels, thereby inhibiting TAM-mediated pathways such as PI3K-AKT and JAK-STAT and sensitizing cancer cells to chemotherapy [44]. Thus, developing PTDSS1 inhibitors represents a promising alternative therapeutic strategy. Despite our integrated approach, this study has several limitations that should be acknowledged. First, our findings, particularly the expression and prognostic value of PTDSS1, are primarily derived from bioinformatics analyses and require further validation in larger, prospective multicenter cohorts. Second, although we identified a positive correlation between PTDSS1 upregulation and M2 macrophage polarization in HCC, the underlying mechanism remains unverified. We lack direct experimental evidence linking PTDSS1 overexpression to PS externalization, and thus, whether PTDSS1 directly regulates macrophage polarization via PS exposure awaits confirmation. Finally, the association between PTDSS1 and immunotherapy response was inferred from correlative analyses and necessitates prospective validation in clinical trials. Conclusions In summary, this study revealed PTDSS1 emerged an upregulated expression pattern of in cancer, which was associated with adverse clinical features and prognosis in specific malignancies. In terms of mechanism, copy number amplification and promoter hypomethylation of PTDSS1 may drive its upregulation, and mediating cell proliferation, genomic instability, and immunosuppression may be the tumor-promoting mechanisms of PTDSS1. Regarding the clinical value, PTDSS1 was identified as a potential diagnostic and prognostic biomarker for HCC, with its expression positively correlated with macrophage infiltration. However, these findings predominantly depended on public database analyses, extensive experiments are still required in the future to elucidate the detailed molecular mechanisms of PTDSS1, thereby guiding its translational therapy. Supplementary Information Supplementary Material 2 Supplementary Material 2