High-throughput mass spectrometry maps thesepsisplasma proteome and the variation in response between patients 

     Anti-PD-1 cancer immunotherapy induces central nervous system immune-related adverse events by activating microglia

     Pan-cancer proteomics reveals a panoramic landscape of novel therapeutic targets

     Proteomic study of candidate proteins behind a new mechanism of 5-fluorouracil resistance in colorectal cancer

     Plasma proteomics identifies biomarkers of Parkinson's disease seven years before symptom onset

     Identifying proteomic risk factors for cancer through prospective and exome-wide analysis of 1,463 circulating proteins and the risk of 19 cancers in UK Biobank

     Proteomic profiling of cardiorespiratory fitness for predicting mortality and multisystem disease risk

     Proteomic analysis of extracellular vesicles derived from colon cancer tumour tissue

     Proteomic analysis of cancer-associated alterations in the human unfoldome

      AlphaDIA enables end-to-end transfer learning for feature-free proteomics

Sepsis is caused by an extreme immune response to infection that can damage organs and lead to death. It is a global health challenge, and there is an urgent need to better understand its pathophysiology and to find new treatments. The teams of Julian C. Knight and Roman Fischer at the University of Oxford, UK, applied high-throughput tandem mass spectrometry to chart the plasma proteome in sepsis and in comparison groups (non-infectious critical illness, post-operative inflammation and healthy volunteers), covering 2,612 samples in total. They described the proteins, pathways and co-expression modules in sepsis and integrated this with paired leukocyte transcriptome data. They mapped the plasma proteomic response of the host over time in sepsis and identified features associated with aetiology, clinical phenotype (including organ failure) and severity. This work reveals subphenotype information on sepsis response states, disease course and outcomes, identifies potential biomarkers, and provides information to support precision treatment of sepsis.

Anti-PD-1 immunotherapy has achieved remarkable results in cancer treatment, but it may be accompanied by central nervous system immune-related adverse events (CNS-irAEs). The specific role of microglia in CNS-irAEs induced by anti-PD-1 therapy is still not clearly understood. A study by the team of Janaki Manoja Vinnakota at the Medical Center – University of Freiburg, Germany, has uncovered a key part of this puzzle. In experimental mice, the team observed that anti-PD-1 treatment produced signs of morphological activation in microglia together with up-regulated expression of major histocompatibility complex class II (MHC II), pointing to changes in microglial function. Notably, the neurocognitive impairment caused by anti-PD-1 treatment was not mediated directly by T cells, B cells or natural killer cells; instead, microglia played a central role. Using single-cell RNA sequencing, the researchers mapped in detail the transcriptomic changes in microglia after treatment. They found that the anti-PD-1 antibody interacts directly with microglia to trigger this cascade of effects, rather than acting indirectly through peripheral T cell activation. Proteomic analysis further identified spleen tyrosine kinase (Syk) as a key regulatory factor in microglial activation under anti-PD-1. Inhibiting Syk not only effectively reduced excessive microglial activation but also improved neurocognitive function, without compromising anti-melanoma efficacy, demonstrating the potential for treating such complications. The study also involved the analysis of human samples: using cell imaging techniques, the researchers found increased expression of treatment-related surface markers in central nervous system tissue from patients who had received anti-PD-1 therapy, consistent with a state of microglial activation, further validating the findings from the animal model. The study not only identifies the mechanism by which Syk-driven microglia act in CNS-irAEs triggered by anti-PD-1 immunotherapy, it also proposes a strategy based on Syk inhibitors that holds promise for effectively preventing or alleviating such complications while preserving anti-tumour efficacy, bringing new therapeutic hope to patients.

Although next-generation sequencing has driven a deeper understanding of the cancer genome and transcriptome, most patients with cancer still need radiotherapy and chemotherapy, treatments associated with a significant risk of recurrence and toxicity. Targeted therapies — including small-molecule drugs, monoclonal antibodies, antibody-drug conjugates (ADCs), proteolysis-targeting chimeras (PROTACs), antibody-directed enzyme prodrug therapy (ADEPTs), cancer treatment vaccines, immune checkpoint inhibitors and T-cell therapies — hold promise for more effective and more precise cancer treatment. Because proteins are the main targets of these therapies and the functional effectors of cancer driver genes and epigenetic aberrations, proteogenomics provides a powerful framework for exploring existing and future cancer therapeutic targets. By integrating proteomic data from 1,043 patients in CPTAC (Clinical Proteomic Tumor Analysis Consortium) with other datasets, the teams of Bing Zhang at Baylor College of Medicine, Qiang Gao at Fudan University and Daming Gao at the Center for Excellence in Molecular Cell Science, Chinese Academy of Sciences identified 2,863 pan-cancer druggable target proteins. Analysis revealed a broad abundance range and identified the biological factors that influence the correlation between mRNA and protein. In addition, integrating tumour proteomic data with genetic screening data from cell lines made it possible to identify druggable dependencies driven by protein overexpression or overactivation, making the prediction of effective drug targets more precise. These results not only expand the range of therapeutic targets, they also refine precision treatment strategies for cancer.

Scan the QR code to read the full text

Proteomic study of candidate proteins behind a new mechanism of 5-fluorouracil resistance in colorectal cancer

A major challenge in the treatment of colorectal cancer (CRC) is that patients develop acquired resistance to key drugs such as 5-fluorouracil (5-FU). Although some resistance mechanisms have been elucidated, response rates to treatment have stalled and many unknown resistance mechanisms remain to be explored. To this end, the team of Mario Ortega Duran at the Institute of Cancer Therapeutics, University of Bradford, UK, used subpopulations of colorectal cancer cell lines with laboratory-induced resistance to 5-FU, aiming to identify potential key proteins by proteomic means. The study applied stable isotope labelling by amino acids in cell culture (SILAC) combined with cell culture, and analysed protein expression using two-dimensional liquid chromatography coupled with an Orbitrap Fusion Tribrid mass spectrometer, comparing the protein expression of 5-FU-resistant cell lines grown in heavy-isotope-enriched medium with that of the corresponding sensitive cell lines and the original parental cell lines. A total of 3,003 proteins were quantified; six proteins (CD44, APP, NAGLU, CORO7, AGR2, PLSCR1) were found to be up-regulated and another six (VPS45, RBMS2, RIOK1, RAP1GDS1, POLR3D, CD55) down-regulated. Eleven of these 12 differentially expressed proteins have already been shown to be associated with resistance mechanisms or to be involved in the development and progression of colorectal cancer. In addition, techniques such as immunoblotting further confirmed elevated levels of two known resistance-related proteins, CD44 and CD63, consistent with the results of the proteomic analysis. The high sensitivity demonstrated by this method provides a powerful tool for dissecting the complex mechanisms of acquired resistance in colorectal cancer.

Parkinson's disease (PD) is a complex and increasingly prevalent neurodegenerative disorder of the central nervous system. It progresses from a premotor stage, characterised by non-motor symptoms such as rapid eye movement sleep behaviour disorder, to a stage of motor disability. There is an urgent need for objective biomarkers at an early stage in order to intervene in and slow the underlying neurodegenerative process. Research teams from University College London and the University Medical Center Göttingen, Germany, performed targeted mass spectrometry analysis on blood samples from patients with recently diagnosed motor Parkinson's disease (n=99), patients with rapid eye movement sleep behaviour disorder (two cohorts: n=18 and n=54) and healthy controls (n=36). A machine learning model analysing the expression of eight proteins (progranulin, mannan-binding lectin serine peptidase 2, endoplasmic reticulum chaperone BiP, prostaglandin-H2 D-isomerase, intercellular adhesion molecule 1, complement C3, Dickkopf WNT signalling pathway inhibitor 3 and plasma protease C1 inhibitor) identified all patients with Parkinson's disease accurately and classified 79% of individuals as early as seven years before onset. Some of these biomarkers correlate with symptom severity. This particular blood test panel reflects molecular events at an early stage and may help identify at-risk participants for clinical trials aimed at slowing or preventing motor Parkinson's disease.

The team of Keren Papier from the Cancer Epidemiology Unit of the Nuffield Department of Population Health, University of Oxford, UK, carried out a study using UK Biobank resources to investigate the association between 1,463 plasma proteins and the risk of developing 19 cancers and 9 cancer subtypes, with participants followed up for a mean of 12 years. To analyse in depth the genetic basis of newly identified protein–cancer relationships, the team applied two genetic approaches: cis-pQTL analysis and an exome-wide protein genetic score (exGS). The results showed that 618 proteins were confirmed to be associated with cancer risk; among them, 107 showed abnormal levels that were still observable more than 7 years after the blood draw and were linked to a subsequent cancer diagnosis. Among the 618 proteins, the associations of 29 were supported by genetic analysis, further strengthening their potential as cancer biomarkers. Four proteins (CD74, TNFRSF1B, ADAM8 and SFTPA2) showed associations that not only withstood the test of long-term (more than 7 years) diagnosis, but were also confirmed in the cis-pQTL and exGS analyses: CD74 and TNFRSF1B were associated with non-Hodgkin lymphoma (NHL), ADAM8 with leukaemia, and SFTPA2 with lung cancer. The study by Keren Papier and colleagues reveals significant associations between multiple blood proteins and cancer risk, particularly changes that can be detected in blood many years before the clinical diagnosis of cancer. These findings are consistently supported by genetic analysis, suggesting that such protein changes may play a role in the early development of cancer. The study therefore both deepens our understanding of cancer biology and provides an important scientific basis for developing blood tests to achieve early warning and intervention in cancer.

Scan the QR code to read the full text

Proteomic profiling of cardiorespiratory fitness for predicting mortality and multisystem disease risk

Although cardiorespiratory fitness (CRF) has wide-ranging effects on metabolic, cardiovascular, pulmonary and neurological health, challenges in the feasibility and reproducibility of CRF measurement have hindered its use in clinical decision-making. The team of Ravi V. Shah at Vanderbilt University, Tennessee, USA, linked the proteomic profiles of 14,145 individuals from four international cohorts with CRF, using different CRF determination methods to establish, validate and characterise a proteomic CRF score. In a UK Biobank cohort of approximately 22,000 people, the proteomic CRF score was associated with reduced all-cause mortality (unadjusted hazard ratio 0.50, 95% confidence interval 0.48-0.52). The proteomic CRF score was also associated with multisystem disease risk, and provided risk reclassification and discrimination beyond clinical risk factors, as well as modifying a high polygenic risk for certain diseases. Finally, the team observed the dynamic behaviour of the proteomic CRF score in individuals undergoing a 20-week exercise training programme, and its correlation with the extent of the training effect on CRF, suggesting that the score may be used for personalised exercise recommendations. These results show that population-based proteomics provides biologically relevant molecular data on CRF that add to genetic risk and are potentially modifiable and clinically translatable.

Scan the QR code to read the full text

Proteomic analysis of extracellular vesicles derived from colon cancer tumour tissue

Colon cancer is a common tumour whose incidence has continued to climb in recent years, a trend that has intensified the urgency of identifying biomarkers and gaining a deeper understanding of disease mechanisms. Extracellular vesicles, as key mediators of intercellular communication, open up new avenues for investigating the mechanisms of disease progression and searching for potential biomarkers. The team of Aleksander Cvjetkovic at the Krefting Research Centre, Institute of Medicine, Sahlgrenska Academy, University of Gothenburg, Sweden, has carried out frontier research in this field. The team collected colon tumour tissue and normal mucosal tissue from 10 patients, isolated tissue-resident vesicles from them, and then carried out comprehensive proteomic analysis using LC-MS/MS. More than 2,000 proteins were identified in total, most of them known vesicle marker proteins. Bioinformatics analysis revealed that in tumour-derived vesicles, proteins associated with energy metabolism and cell adhesion showed decreased expression, while proteins involved in protein synthesis showed increased expression. Particularly noteworthy is that 53 membrane proteins were significantly overexpressed in tumour-derived vesicles, several of which possess enzymatic activity capable of degrading the extracellular matrix, such as fibroblast activation protein (FAP), cell surface hyaluronidase (CEMIP2) and ephrin type-B receptor 3 (EPHB3). The overexpression of these proteins is consistent with the quantitative results at the whole-proteome level. The marked differences in proteomic expression between healthy and cancerous tissue highlight the important role the interstitial vesicle secretome plays in the development of colon cancer, providing important clues for further research into disease mechanisms and the exploration of novel diagnostic markers.

Scan the QR code to read the full text

Proteomic analysis of cancer-associated alterations in the human unfoldome

Many proteins lack a stable three-dimensional structure. These intrinsically disordered proteins (IDPs), or hybrid proteins containing ordered domains together with intrinsically disordered protein regions (IDPRs), often carry out regulatory functions related to molecular recognition and signal transduction. IDPs/IDPRs make up a large part of the human proteome and are termed the "unfoldome". The team of Victor Paromov at Meharry Medical College, USA, studied the human breast cancer unfoldome and the relationship between IDPs and key disease genes and pathways. The team used a bottom-up proteomics approach — MudPIT (multidimensional protein identification technology) — to profile differentially expressed IDPs in normal (MCF-10A) and breast cancer (BT-549) cell lines. Overall, the team identified 2,271 protein groups in the unfoldome of the normal and cancerous proteomes, of which 148 IDPs were found to be significantly differentially expressed in cancer cells. Further analysis produced annotations for 140 IDPs, which were then classified into GO (Gene Ontology) categories and pathways. In total, 65% (91/140) of the IDPs were related to various diseases, and 20% (28/140) mapped to cancer terms. A substantial portion of the differentially expressed IDPs contained disordered regions, a conclusion confirmed by in silico characterisation. Overall, the analyses suggest high levels of interactivity in the human cancer unfoldome and a prevalence of moderately to highly disordered proteins in the network.

Scan the QR code to read the full text

AlphaDIA enables end-to-end transfer learning for feature-free proteomics

With the rapid development of mass spectrometry, proteomics continues to widen the boundaries of its applications. At the same time, the vast volumes of data produced by new instruments and advanced acquisition techniques pose major challenges for bioinformatics analysis. In order to mine these treasure troves of biological data efficiently and rigorously, the team of Georg Wallmann at the Max Planck Institute of Biochemistry has launched "alphaDIA" — an open-source, modular search framework designed for data-independent acquisition (DIA) proteomics. The team's innovation lies in developing a feature-free identification algorithm that excels at capturing patterns in the complex data generated by high-sensitivity time-of-flight mass spectrometers, and is particularly well suited to the novel, highly efficient scan modes that conventional algorithms struggle to handle. Through rigorous benchmarking, alphaDIA has demonstrated excellent peptide identification and quantification capabilities, proving its competitiveness in technical performance. alphaDIA is not limited to existing resources: it also introduces a novel search strategy that applies end-to-end transfer learning combined with fully predicted spectral libraries. This approach relies on continuously optimised deep neural networks designed to predict performance under specific instrument and experimental conditions, making universal DIA analysis of any type of post-translational modification (PTM) a reality. This strategy not only strengthens support for known spectral libraries, it also greatly extends the depth and breadth of proteomics research. To ensure that the results are widely accessible and applied, alphaDIA has been designed as a high-performance, easy-to-access platform that runs smoothly both in local environments and deployed in the cloud, and is fully open to the public — advancing the democratisation of DIA data analysis and marking proteomics research's entry into a more open and collaborative new era.