Clinical functional proteomics of intercellular signalling in pancreatic cancer

    Small molecule modulation of the protein corona for deep plasma proteome profiling

    A plasma proteome atlas in 53,026 healthy and diseased adults

    Decreased lipidated ApoE–receptor interactions confer protection against pathogenicity of ApoE and its lipid cargoes in lysosomes

   Cancer-associated fibroblasts maintain critical pancreatic cancer cell lipid homeostasis in the tumor microenvironment

    Exploring structural diversity across the protein universe with The Encyclopedia of Domains

    Proteogenomic analysis reveals non-small cell lung cancer subtypes predicting chromosome instability and tumor microenvironment

   A proteogenomic analysis of cervical cancer reveals therapeutic and biological insights

    Proteogenomic analysis of human cerebrospinal fluid identifies neurologically relevant regulation and implicates causal proteins for Alzheimer's disease

      A non-invasive lipid panel of MASLD fibrosis transition underscores the role of lipoprotein sulfatides in hepatic immunomodulation

Pancreatic ductal adenocarcinoma (PDAC) is one of the most aggressive and lethal cancers, with a five-year survival rate below 10%. Its distinctive tumor microenvironment (TME) contains abundant non-malignant stromal cells and extracellular matrix, and these features drive cancer progression through complex intercellular signalling. The team of Professor Tian Ruijun at the Southern University of Science and Technology developed and applied a multidimensional clinical functional proteomics strategy, TMEPro, to dissect the dynamic signalling events between cancer cells and stromal cells in PDAC. Combining mass spectrometry, the study systematically analyzed 100 human PDAC samples and genetically engineered mouse models, focusing on the expression patterns of glycosylated secreted proteins and plasma membrane proteins, and used bioinformatics to reveal the key role of the PDGFR–PTPN11–FOS signalling axis in intercellular signal transmission. In addition, the study found that matrix metalloproteinases (MMPs) regulate signalling between tumor cells by mediating the cleavage of the AXL receptor tyrosine kinase, a cleavage process closely associated with lymph node metastasis in PDAC patients. Functional validation experiments showed that combined blockade of AXL cleavage and kinase activity significantly suppressed the proliferative and metastatic capacity of cancer cells. In summary, using TMEPro, this study constructed for the first time a map of the interaction signalling network within the PDAC tumor microenvironment, revealing novel interaction mechanisms and providing a rich data resource, as well as new theoretical support and clinical strategic directions for the development of PDAC diagnostics and therapeutic targets.

Plasma proteomic analysis is of great importance for disease diagnosis, monitoring and biomarker discovery, but the presence of high-abundance proteins such as albumin poses a major challenge to the detection of low-abundance proteins. The Mahmoudi group at Michigan State University proposed an approach that modulates the nanoparticle protein corona with small molecules to greatly increase the depth of plasma proteome profiling. The study showed that adding a variety of small molecules—including metabolites, lipids, vitamins and nutrients—to plasma, where they interact with polystyrene nanoparticles, significantly changes the composition of the protein corona, enabling the detection of 1,793 proteins compared with only 218 proteins from plasma analysis alone, an 8.25-fold improvement. Moreover, the addition of phosphatidylcholines (PtdChos) allowed the detection of 897 proteins, demonstrating that PtdChos markedly reduce the proportion of high-abundance proteins (albumin, for example, dropped from 81% in raw plasma to 17%) through hydrophobic interactions and hydrogen bonding with albumin. Using an optimized data-independent acquisition method, the addition of PtdChos enabled the detection of 1,436 proteins in a single plasma sample. In addition, top-down proteomic analysis found that the addition of PtdChos detected 337 additional proteoforms compared with the untreated protein corona, including disease-related protein variants. The study shows that small-molecule modulation of the protein corona not only substantially expands coverage of low-abundance proteins, but also opens new possibilities for discovering disease-related biomarkers through specific protein enrichment. This work provides an efficient and economical alternative for deep plasma proteome profiling, with broad application potential in biomarker development and clinical translation.

A plasma proteome atlas in 53,026 healthy and diseased adults

As precision medicine advances, research on the plasma proteome has become increasingly important for revealing the biological mechanisms of health and disease. Based on 2,920 plasma proteins measured in 53,026 participants, the team of Professor Yu Jintai at Huashan Hospital, Fudan University, performed a comprehensive analysis of 406 common diseases, 660 incident diseases and 986 health-related traits, constructing an association atlas of the human plasma proteome and disease phenotypes. Through machine learning and the integration of genetic data, the study established 168,100 protein–disease associations and 554,488 protein–trait associations, finding that more than 650 proteins play shared roles in at least 50 diseases, and identifying 474 potential causal proteins together with related drug repurposing opportunities. The study found that many plasma proteins show high discriminative power for disease diagnosis and prediction. For example, protein-based prediction models achieved excellent performance with an AUC above 0.8 across 183 diseases, while the integration of protein quantitative trait locus (pQTL) data revealed 37 drug repurposing opportunities and 26 potential therapeutic targets with favorable safety profiles. In addition, Mendelian randomization analysis of the proteome and disease identified several proteins with causal relationships, includingGDF15and EGFR, which are not only important disease markers but also provide potential targets for the common treatment of multiple diseases. This study lays a solid foundation for the clinical application of plasma proteomics and, through an open-access interactive resource, offers a new perspective for the systematic study of disease and the dissection of biological mechanisms. The findings underscore the important role of proteomics in early disease detection, multimorbidity risk assessment and personalized treatment.

Decreased lipidated ApoE–receptor interactions protect ApoE and its lipid cargoes in lysosomes from pathogenicity

The occurrence of late-onset Alzheimer's disease (LOAD) is closely related to the genotype of apolipoprotein E (ApoE): ApoE4 significantly increases disease risk, whereas ApoE2 is protective. However, the specific mechanism by which ApoE contributes to lysosomal dysfunction is not fully understood. A research team in California, USA, focused on the lipid cargo of ApoE and its interaction with the low-density lipoprotein receptor (LDLR), exploring how these factors influence lysosomal pathology and neurodegenerative changes. The study found that ApoE4 carrying polyunsaturated fatty acid cholesteryl esters (PUFA-CE) aggravates lipofuscin deposition, lipid peroxidation and tau accumulation through LDLR-mediated lysosomal targeting, thereby exacerbating neuronal dysfunction. In contrast, ApoE2 and ApoE carrying the Christchurch protective mutation markedly reduced these pathogenic effects by decreasing binding to LDLR. Furthermore, in a mouse model of Alzheimer's disease, blocking the ApoE–LDLR interaction effectively relieved lysosomal burden and attenuated disease-related pathological features. This study reveals that the lipidation state of ApoE and LDLR-mediated lysosomal targeting are important regulatory factors in LOAD pathology, and highlights that reducing ApoE–LDLR interaction can significantly lower lipid-related neurotoxicity, providing a new theoretical basis and potential targets for LOAD treatment strategies.

Pancreatic ductal adenocarcinoma (PDAC) is a highly aggressive malignancy whose microenvironment (TME) is rich in cancer-associated fibroblasts (CAFs) and dense stroma, which restricts the supply of oxygen and nutrients and causes severe tissue hypoxia. A research team at the Abramson Family Cancer Research Institute of the University of Pennsylvania revealed that CAFs help PDAC cells maintain lipid homeostasis under the oxygen- and nutrient-poor conditions of the tumor by secreting unsaturated lipids, particularly lysophosphatidylcholines (LPCs). The study found that LPCs secreted by CAFs are taken up by PDAC cells and converted into phosphatidylcholine (PC) to maintain cell membrane integrity and relieve endoplasmic reticulum (ER) stress. When the pathway converting LPC to PC was inhibited, the survival of PDAC cells dropped significantly and ER stress and apoptosis increased markedly. Further work showed that, being relatively less hypoxic, CAFs have a stronger capacity for unsaturated lipid synthesis than neighboring cancer cells, making them important lipid donors to PDAC cells. Strategies that block LPC metabolism, such as the use of LPCAT (LPC acyltransferase) inhibitors, markedly weakened the ability of PDAC cells to survive under tumor conditions. Both animal experiments and in vitro organoid studies verified the effectiveness of blocking LPC metabolism in suppressing tumors. In summary, this study reveals that CAFs support the survival of PDAC cells in the extreme tumor microenvironment through a mechanism of unsaturated LPC-mediated lipid cross-feeding, offering new ideas and potential strategies for metabolism-targeted therapy in pancreatic cancer.

Exploring structural diversity across the protein universe with The Encyclopedia of Domains

With the advent of theAlphaFoldProtein Structure Database (AFDB), protein structural information has grown enormously, but traditional sequence-based methods struggle to fully resolve its domains. To address this problem, the present study developed The Encyclopedia of Domains (TED), which combines deep learning with advanced structural comparison techniques to identify and classify domains across more than 214 million protein sequences in AFDB. TED identified approximately 365 million domains, of which more than 100 million could not be detected by traditional methods, and found that 77% of domains are similar to known CATH superfamilies. Through TED, the study revealed more than 10,000 novel superfamily interactions and thousands of new fold structures, greatly expanding the protein fold space. The analysis also showed that some folds are highly conserved across all superkingdoms and carry out essential cellular functions, while others display specific evolutionary adaptations. In addition, TED detected highly symmetric structures and repeat domains, providing new perspectives for functional studies and evolutionary analysis. Overall, TED provides an innovative tool for resolving protein structure–function relationships, significantly improves the utility of AFDB, and lays a foundation for drug development and functional annotation research. This resource will continue to improve as the database is updated, offering important support for structural biology and evolutionary research.

Non-small cell lung cancer (NSCLC) is the leading cause of cancer-related death worldwide, and its high heterogeneity poses a challenge to precision treatment. By integrating genomic, transcriptomic, proteomic and post-translational modification data, the research group of Kwang Pyo Kim at Kyung Hee University in South Korea analyzed 229 Korean NSCLC patients and, combined with validation data from another 462 patients, identified five molecular subtypes with clinical and prognostic significance: metabolic, alveolar-like, proliferative, hypoxic and immune. The proliferative subtype was significantly associated with whole-genome doubling and cell cycle pathway activity, indicating increased chromosome instability; the hypoxic subtype showed activation of the PI3K-Akt pathway, a poorer prognosis and a close association with tumor metastasis. The study also found marked differences in tumor microenvironment between subtypes. The immune subtype, for example, is rich in tumor-infiltrating lymphocytes with high immune activity and shows greater sensitivity to adjuvant immunotherapy. In addition, proteomic and post-translational modification analyses revealed potential therapeutic targets, such as specific modifications of SLK and LRRFIP1, features directly linked to poor prognosis and treatment response. Through multi-omics integration, this study provides an in-depth dissection of NSCLC molecular subtypes and their associated mechanisms, offering new perspectives for stratified diagnosis, prognostic assessment and personalized treatment of NSCLC patients. The findings lay an important foundation for developing precision treatment strategies targeting specific subtypes and point the way for future research.

Cervical cancer (CC) is the fourth most common malignancy among women worldwide. Although HPV vaccination and screening strategies have reduced the incidence of CC in high-income countries, it remains a major public health threat to women's health in low-income countries. The team of Professor Gao Daming at the Chinese Academy of Sciences performed an integrated multi-omics analysis—genome, transcriptome, proteome, phosphoproteome and acetylome—on samples from 139 Chinese CC patients, revealing the features of HPV-driven cancer and their regulatory mechanisms. The study found that EP300-mediated acetylation of FOSL2-K222 promotes cancer cell proliferation, and classified patients into three subtypes with distinct prognoses, genetic alterations and immune infiltration characteristics. In addition, PRKCB was identified as a potential biomarker of radiotherapy response, offering a new avenue for improving outcomes in locally advanced patients. The study revealed that HPV infection significantly affects phenotypic perturbation in CC, with viral gene expression higher in tumors than in normal tissue. Chromosomal copy number alterations regulate protein expression through cis and trans effects, and post-translational modifications play a key role in tumor progression. Specific proteins such as B4GALT1 and HK2 were significantly negatively correlated with overall patient survival, and metabolic pathways were further validated as potential therapeutic targets. Through proteogenomics, this study reveals the molecular mechanisms of cervical cancer and HPV-related oncogenic features, providing an important basis and data support for precision diagnosis, personalized treatment and the development of new diagnostic and therapeutic strategies.

Proteogenomic analysis of human cerebrospinal fluid identifies neurologically relevant regulation and implicates causal proteins for Alzheimer's disease

Alzheimer's disease (AD) is the most common neurodegenerative disease. Its genetic component is substantial, yet many disease-related genes remain unknown. The teams of Yun Ju Sung and Carlos Cruchaga at Washington University School of Medicine in St. Louis, Missouri, USA, performed multi-omics analysis of cerebrospinal fluid (CSF) samples from 3,506 individuals and created a CSF protein quantitative trait locus (pQTL) atlas covering 6,361 proteins, revealing the distinctive genetic regulation of CSF proteins. The study identified 3,885 independent pQTL associations, of which 2,885 were CSF-specific, and further clarified the pleiotropic regulatory roles of chromosomal regions such as 3q28 and 19q13.32 (near APOE) in neurodevelopmental and AD-related pathways. By integrating pQTL data with genome-wide association studies of AD, the study identified 38 potential causal proteins highly associated with AD, 15 of which can be targeted by existing drugs. These proteins are mainly enriched in immune pathways and microglial function, and are significantly associated with key AD biomarkers such as phosphorylated tau. In addition, the researchers developed a proteomics-based AD prediction model that significantly outperformed conventional genetic models, showing higher accuracy in detecting AD status and predicting preclinical markers. By revealing the genetic regulation and pathogenic mechanisms of CSF proteins, this study provides new targets and directions for the precision diagnosis and treatment of neurological diseases, and lays a foundation for molecular research and drug development in AD.

The diagnosis of metabolic dysfunction-associated steatotic liver disease (MASLD) relies on liver biopsy, an invasive approach that raises issues of cost and risk. The teams of Professor Shui Guanghou at the Institute of Genetics and Developmental Biology of the Chinese Academy of Sciences, Director Zheng Minghua at the First Affiliated Hospital of Wenzhou Medical University, and Academician Wang Fusheng at the Fifth Medical Center of the PLA General Hospital used lipidomics and steroidomics to perform an integrated analysis of serum and liver tissue samples from 519 patients, and developed a non-invasive lipid panel to distinguish mild from moderate fibrosis in MASLD. The study trained a diagnostic tool based on a lasso regression model combining 15 lipids and 5 clinical variables, which showed good diagnostic performance in discriminating mild fibrosis (F1) from no fibrosis (F0) (AUROC = 0.775), significantly outperforming the conventional FIB-4 score. The study further revealed that circulating sulfatides (SLs) play a key role in the pathological mechanism of MASLD fibrosis. Lipidomic analysis showed that the redistribution of sulfatides from high-density lipoproteins (HDLs) to low-density lipoproteins (LDLs) is closely associated with the occurrence of fibrosis. Reduced sulfatide levels in patient LDLs significantly weakened the activation of type II natural killer T cells, implying an important role for SLs in regulating hepatic immune responses. In addition, integrated data analysis indicated that disordered lipid metabolism and the immunomodulatory role of sulfatides jointly drive the progression of MASLD fibrosis. The non-invasive lipid panel developed in this study provides an effective tool for the early diagnosis and treatment of MASLD, while revealing the key role of lipoprotein sulfatides in hepatic immune regulation and opening a new direction for precision medicine in metabolic liver disease.