Clinical proteomics最新文献

Patients with uremia undergoing long-term hemodialysis are prone to multi-organ complications, but the underlying molecular mechanisms remain unclear. Ferroptosis, an iron-dependent form of cell death, has been linked to inflammation and organ damage. Its role in hemodialysis-related pathology, however, has not been well characterized. In this study, we systematically profiled low-abundance plasma proteins from six hemodialysis patients and eight healthy controls using a protein corona-based enrichment technique to enhance detection sensitivity. A total of 183 differentially expressed proteins (DEPs) were defined based on a fold-change threshold (≤ 0.25 or ≥ 4), including 101 upregulated and 82 downregulated proteins. Notably, pathway enrichment analysis highlighted the ferroptosis pathway, with altered abundance of proteins including TFRC, ALOX15, PRNP, CYBB, and ACSL1, suggesting a potential association of ferroptosis-related signals with hemodialysis-related complications. To complement the proteomic analysis, enzyme-linked immunosorbent assay (ELISA) was performed in an independent cohort. ALOX15 showed a significant and reproducible increase in plasma levels (P < 0.0001), consistent with the proteomic results. Other ferroptosis-related candidates warrant further evaluation and independent validation in larger cohorts. Furthermore, drug target prediction based on DEP data identified N-oleoyldopamine, luteolin, and catechol as potential compounds targeting the five ferroptosis-related molecules. Collectively, this study provides an exploratory plasma proteomic resource and suggests that ferroptosis-associated plasma protein changes may be relevant to hemodialysis-related complications, warranting further validation. Collectively, this study provides an exploratory plasma proteomic resource and offers initial insights into ferroptosis-associated plasma protein changes in hemodialysis patients.

Background: There are known associations between cardiovascular disease (CVD)-related plasma proteins and spirometry measures. Diffusing capacity for carbon monoxide (D_LCO) measures gas exchange that can be impaired both by lung and heart diseases. We aimed to study the associations between D_LCO and CVD-linked plasma proteins in a population-based cohort without airflow obstruction.

Methods: 89 CVD-linked proteins were analysed in 427 individuals who underwent spirometry examination with D_LCO measurement. Analyses were adjusted for age, gender, height, weight, smoking status and pack years, plates, storage time and cardiovascular morbidity (carotid plaques, hypertension and cardiac medication). Furthermore, a sensitivity analysis (n = 362) was carried out after excluding subjects with an FEV₁/VC ratio < the lower limit of normal (LLN) and steps were taken to ensure a false discovery rate under 5%.

Results: We found 18 proteins negatively associated with D_LCO%predicted after full adjustments (GLI reference equations). Eleven of these proteins (Fibroblast growth factor 23 (estimated coefficients, (adjusted p-value)): -0.010 (< 0.001), Matrix metalloproteinase-12; -0.011 (< 0.001), Growth differentiation factor 15: -0.008 (< 0.001), C-C motif chemokine 20: -0.013 (0.006), Interleukin-6: -0.014 (< 0.001), Fatty acid-binding protein, adipocyte - 0.007 (0.001), Urokinase-type plasminogen activator receptor: -0.004 (< 0.001), Interleukin-1 receptor antagonist: -0.008 (< 0.001), TNF-related apoptosis-inducing ligand receptor 2: -0.005 (0.001), Renin: -0.008 (0.03) and Spondin-1: -0.003 (0.03)) remained significant after further excluding subjects with obstruction on spirometry (FEV₁/VC < LLN).

Conclusions: Several CVD-linked plasma proteins were associated with D_LCO in subjects without airflow obstruction on spirometry and after adjustments for known cardiovascular morbidity. The likely explanations may be the pro-fibrotic and pro-inflammatory nature of many of the proteins causing changes in gas exchange. These proteins could potentially signal for early disease mechanisms leading to impaired gas exchange.

Background: Polycystic ovary syndrome (PCOS) is a complex and heterogeneous metabolic disorder that affects 6-20% of women of reproductive age. However, research on the lactylation-modified proteome in PCOS remains limited.

Methods: This study included 30 patients with PCOS and 30 control subjects, all of whom underwent intracytoplasmic sperm injection or in vitro fertilization-embryo transfer treatments at the fertility center between October 2022 and May 2023. A 4-dimensional label-free proteomic quantitation method was applied to analyze enzymatically digested peptide fragments of granulosa cell proteins. Liquid chromatography-mass spectrometry was used for protein identification and quantification.

Results: Bioinformatics analysis of differentially expressed proteins (DEPs) and differentially lactylated proteins identified 1057 DEPs between the two groups. Among these, 478 proteins were upregulated, and 579 were downregulated in the PCOS group. Regarding lactylation modifications, 668 proteins exhibited increased lactylation levels, while 1059 proteins indicated decreased lactylation levels in the PCOS group. Additionally, site-level analysis revealed 1041 upregulated and 2143 downregulated lactylation sites in the PCOS group.

Conclusions: This study provides a comprehensive quantitative overview of proteomic and lactylation-modified proteomic expression profiles in granulosa cells from patients with PCOS, offering novel insights into PCOS research. Further research is needed to clarify the specific roles of protein lactylation in PCOS pathogenesis.

Background: Mass spectrometry-based proteomics enables high-throughput quantification of thousands of proteins in clinical samples, fueling biomarker discovery for disease diagnosis and prognosis. However, leveraging complex proteomic profiles for predictive modeling often requires advanced machine learning (ML) expertise that many biomedical researchers lack. User-friendly tools are needed to apply state-of-the-art ML algorithms to proteomics data. XGBoost is a powerful tree-based ML algorithm known for high accuracy in classification tasks, and has been successfully used to classify cancer subtypes from multi-omics data.

Methods: We developed ProteoBoostR, a Shiny application that streamlines supervised ML on protein abundance datasets. It allows researchers to train, evaluate and apply XGBoost classification models through an interactive web interface, without requiring coding.

Results: We demonstrate the application of ProteoBoostR for the classification of proteomic subtypes across two independent datasets of glioblastoma multiforme, and for the detection of lung adenocarcinoma in serum. These application examples illustrate how ProteoBoostR can harness proteomic patterns for the stratification of patients.

Conclusions: ProteoBoostR is an open-source application that empowers proteomics researchers to perform advanced ML classification. It can be readily applied to other proteomic datasets and disease contexts, promoting reproducible ML analyses in proteomics and accelerating the translation of omics-based classifiers into clinical research.

Background: Due to its proximity to cells of the central nervous system, cerebrospinal fluid (CSF) is an important source of novel biomarkers for neurological diseases. Membrane-bound extracellular vesicles (EVs) are enriched for proteins of intracellular and membrane origin, implicated in the pathogenesis of some neurological diseases, and secreted into CSF. Proteomic profiling of CSF-EVs, however, is limited by the large volumes required for typical EV isolation protocols.

Methods: We appraised the performance of tetraspanin (CD81, CD63, CD9)-based immunocapture for EV isolation from 200 to 1000 µL CSF sample and compared to size-exclusion chromatography (SEC). EVs were profiled by library-free data independent-acquisition (DIA) mass spectrometry to assess protein depth and abundance of specific EV markers and known co-isolates. Abundance and precursor peptide locations for potential neuronal-specific immunocapture targets described in the literature were also assessed.

Results: Immunocapture was effective using CSF volumes as low as 200 µL, consistently detecting core EV markers and reducing relative levels of non-vesicular proteins such as Apolipoprotein B (APOB) and galectin 3 binding protein (LGALS3BP) compared with size-exclusion chromatography (SEC). Proteomic depth reached 811 ± 14 protein groups in EVs from 200 µL CSF, increasing to 1285 ± 224 when using feature alignment across runs with up to 1000 µL starting volume. These included eleven candidate biomarkers of neurological diseases that were detected in all preparation methods, with additional candidates detected by immunocapture only. Increased depth was observed for both transmembrane and secreted proteins using immunocapture compared with SEC, with proportional enrichment of transmembrane proteins.

Conclusions: This work demonstrates the effectiveness of tetraspanin immunocapture for proteomic profiling of EVs in small volumes of CSF that can be adapted to use with cell-type-specific markers of choice.

Background: Genome-wide association studies (GWASs) have identified over 100 loci associated with rheumatoid arthritis (RA) risk. Nonetheless, the contribution of these loci to RA risk remains largely unknown, hampering the development of new therapeutics. As proteins are direct effectors of disease processes, we conducted the first large-scale proteome-wide association study (PWAS) to prioritize RA risk genes based on their effects on plasma protein abundance.

Methods: We integrated RA GWAS summary statistics (discovery: 22,350 cases and 74,823 controls; replication: 31,313 cases and 995,377 controls) with precomputed protein expression weights generated from the Atherosclerosis Risk in Communities (ARIC) study (N = 7,213) and INTERVAL study (N = 3,301). Causal inference was performed using Mendelian randomization (MR) and colocalization analyses. Druggable target exploration was conducted to identify potential therapeutic targets for RA.

Results: A total of 35 genetically regulated proteins associated with RA risk, including 10 potentially causal candidates, were identified. Notably, six potentially causal proteins (FCRL3, ICOSLG, MAPK3, WISP1, FAM213A, and IL1RN) were not implicated in the original GWASs. Druggable target exploration identified 160 drug-gene interactions, including a drug, AMG-557, targeting the PWAS protein ICOSLG, which possesses superior anti-inflammatory and anti-rheumatic activity in autoimmune diseases and may therefore be a candidate for RA treatment.

Conclusions: Our results provide novel insights into RA pathogenesis and suggest promising targets for further mechanistic investigation and drug development.

Background: Lung adenocarcinoma (LUAD) is a common culprit of cancer-related deaths. Recent studies have revealed that succinylation-related genes (SRGs) are pivotal in cancer. However, the comprehensive characteristics and clinical significance of SRGs in LUAD occurrence are not clear. Therefore, our goal is to dig out the succinylation-related prognostic feature genes in LUAD.

Methods: We identified differentially expressed SRGs in LUAD, and established the LUAD prognostic model using analyses of multivariate, LASSO, and univariate Cox regression. Based on clinical information and riskscore, we graphed a nomogram of the prognostic model and analyzed the independent prognostic ability of the riskscore. Analyses of immune assessment, mutation frequency, and drug sensitivity were carried out on LUAD patients.

Results: A 9-gene prognostic model was successfully set up in this project. The receiver operation characteristic (ROC) curves illustrated that the model effectively predicted the risk of LUAD patients. The levels of immune infiltration and immune scores of LUAD patients in the high-risk (HR) group were greatly lower than those in the low-risk (LR) group. Furthermore, compared to the LR group, the HR group had a significantly elevated gene mutation rate. ENPP3 and SLC22A8 may respond to targeted drugs more sensitively. The low-expression groups of ENPP3 and SLC22A8 genes may have higher drug sensitivity to Nilotinib, ARRY-614, and Megestrol acetate, with lower drug resistance.

Conclusion: The above results indicated that the prognostic model established using SRGs can be a predictive marker for LUAD prognosis, offering references for LUAD treatment and evaluation.

Background: Triple-negative breast cancer (TNBC) is a breast cancer subtype with the highest recurrence rates, for which response to neoadjuvant chemotherapy (NACT) is a critical prognostic factor. Liquid biopsy (LB) is an emerging approach in the personalization of TNBC management; among the numerous circulating molecules assessable by LB, serum proteins are gaining interest, owing to their multiple roles in cancer progression and anti-cancer immune response. Here we report findings of interim analyses in the INSTIGO trial, which aims to discover circulating protein biomarkers of TNBC response to NACT.

Patients and methods: Blood samples were collected at diagnosis and at the time of post-NACT surgery from 30 non-metastatic TNBC patients. NACT consisted of standard carboplatin-paclitaxel and epirubicin-cyclophosphamide regimens. A panel of 21 proteins was quantified in serum using high-sensitivity multiplex immunoassays (Luminex MAP^® technology).

Results: Among the 24 analysable patients, 13 had pathological complete response (pCR) and 11 were without pCR. In the pCR group, mean CX3CL1, angiopoietin-2 (ANGPOI-2), CD40 and PD-L1 levels increased significantly after NACT (p < 0.001, p < 0.001, p = 0.006, and p = 0.02, respectively). In the non-pCR group, mean CXCL5 level tended to decrease after treatment (p = 0.06). When the difference in protein levels between the end and the start of NACT was measured for each patient (Δ of a given protein), ΔCX3CL1, ΔCXCL5 and ΔANGPOI-2 differed significantly between pCR and non-pCR patients (p = 0.003, p = 0.04, p = 0.04, respectively). The baseline concentrations of CCL5, IL8, TIE2, CX3CL1 and CXCL5 tended to be associated with response to NACT, with higher levels observed among the non-pCR patients; however, statistical significance was not reached.

Conclusion: Our findings highlight the potential of circulating proteins to be biomarkers of TNBC response to NACT. Pre/post-NACT changes in CX3CL1, CXCL5, CD40, ANGPOI-2 and PD-L1 levels suggest their relevance for assessing NACT efficacy. These results will be validated at completion of the INSTIGO trial.

Trial registration: ClinicalTrials.gov, identifier NCT04438681.