Synovial fibroblast gene expression is associated with sensory nerve growth and pain in rheumatoid arthritis.

It has been presumed that rheumatoid arthritis (RA) joint pain is related to inflammation in the synovium; however, recent studies reveal that pain scores in patients do not correlate with synovial inflammation. We developed a machine-learning approach (graph-based gene expression module identification or GbGMI) to identify an 815-gene expression module associated with pain in synovial biopsy samples from patients with established RA who had limited synovial inflammation at arthroplasty. We then validated this finding in an independent cohort of synovial biopsy samples from patients who had early untreated RA with little inflammation. Single-cell RNA sequencing analyses indicated that most of these 815 genes were most robustly expressed by lining layer synovial fibroblasts. Receptor-ligand interaction analysis predicted cross-talk between human lining layer fibroblasts and human dorsal root ganglion neurons expressing calcitonin gene-related peptide (CGRP+). Both RA synovial fibroblast culture supernatant and netrin-4, which is abundantly expressed by lining fibroblasts and was within the GbGMI-identified pain-associated gene module, increased the branching of pain-sensitive murine CGRP+ dorsal root ganglion neurons in vitro. Imaging of solvent-cleared synovial tissue with little inflammation from humans with RA revealed CGRP+ pain-sensing neurons encasing blood vessels growing into synovial hypertrophic papilla. Together, these findings support a model whereby synovial lining fibroblasts express genes associated with pain that enhance the growth of pain-sensing neurons into regions of synovial hypertrophy in RA.

Platelet-lymphocyte ratio predicts chemotherapy response and prognosis in patients with gastric cancer undergoing radical resection.

Background: Amounting literatures have reported the significance of systemic inflammatory markers for evaluating tumor prognosis. But few studies have systematically compared their superiority and their impact on adjuvant chemotherapy. Aims: We aimed to investigate the ability of inflammatory markers to predict the efficacy of chemotherapy in GC patients undergoing radical therapy and to identify an effective methodology based on the study's findings that would enable clinicians to differentiate between chemotherapy-responsive populations. Methods: We retrospectively enrolled 730 GC patients who underwent radical gastrectomy. Fibrinogen (FIB), platelet-lymphocyte ratio (PLR), systemic inflammation response index (SIRI), prognostic nutritional index (PNI), systemic immune-inflammation index (SII), neutrophil-lymphocyte ratio (NLR) and lymph node ratio (LNR) were grouped according to cutoff values. Their clinical significance for GC prognosis was determined by multivariate COX regression analysis in the 730 GC patients and high/low PLR status subgroups. Cases were divided into four groups according to PLR status and adjuvant chemotherapy status and survival was compared among groups. Results: Multivariate analysis showed that PLR was an independent prognostic factor for overall survival (OS) and disease-free survival (DFS) of GC patients. Adjuvant chemotherapy improved survival more significantly in patients with low PLR than that with high PLR. Among patients receiving adjuvant chemotherapy, low PLR was significantly associated with prolonged survival in TNM stage II, but not in TNM stage III. Conclusion: Preoperative high PLR is an independent risk factor for GC patients undergoing radical gastrectomy and adversely affects the postoperative chemotherapy effect.

Glucose metabolism-based signature predicts prognosis and immunotherapy strategies for colon adenocarcinoma.

BACKGROUND: The global prevalence and metastasis rates of colon adenocarcinoma (COAD) are high, and therapeutic success is limited. Although previous research has primarily explored changes in gene phenotypes, the incidence rate of COAD remains unchanged. Metabolic reprogramming is a crucial aspect of cancer research and therapy. The present study aims to develop cluster and polygenic risk prediction models for COAD based on glucose metabolism pathways to assess the survival status of patients and potentially identify novel immunotherapy strategies and related therapeutic targets. METHODS: COAD-specific data (including clinicopathological information and gene expression profiles) were sourced from The Cancer Genome Atlas (TCGA) and two Gene Expression Omnibus (GEO) datasets (GSE33113 and GSE39582). Gene sets related to glucose metabolism were obtained from the MSigDB database. The Gene Set Variation Analysis (GSVA) method was utilized to calculate pathway scores for glucose metabolism. The hclust function in R, part of the Pheatmap package, was used to establish a clustering system. The mutation characteristics of identified clusters were assessed via MOVICS software, and differentially expressed genes (DEGs) were filtered using limma software. Signature analysis was performed using the least absolute shrinkage and selection operator (LASSO) method. Survival curves, survival receiver operating characteristic (ROC) curves and multivariate Cox regression were analyzed to assess the efficacy and accuracy of the signature for prognostic prediction. The pRRophetic program was employed to predict drug sensitivity, with data sourced from the Genomics of Drug Sensitivity in Cancer (GDSC) database. RESULTS: Four COAD subgroups (i.e., C1, C2, C3 and C4) were identified based on glucose metabolism, with the C4 group having higher survival rates. These four clusters were bifurcated into a new Clust2 system (C1 + C2 + C3 and C4). In total, 2175 DEGs were obtained (C1 + C2 + C3 vs. C4), from which 139 prognosis-related genes were identified. ROC curves predicting 1-, 3- and 5-year survival based on a signature containing nine genes showed an area under the curve greater than 0.7. Meanwhile, the study also found this feature to be an important predictor of prognosis in COAD and accordingly assessed the risk score, with higher risk scores being associated with a worse prognosis. The high-risk and low-risk groups responded differently to immunotherapy and chemotherapeutic agents, and there were differences in functional enrichment pathways. CONCLUSIONS: This unique signature based on glucose metabolism may potentially provide a basis for predicting patient prognosis, biological characteristics and more effective immunotherapy strategies for COAD.

Prediction and therapeutic targeting of the tumor microenvironment-associated gene CTSK in gastric cancer.

BACKGROUND: Cathepsin-K (CTSK) is overexpressed in Gastric cancer (GC) and the mechanism of its overexpression in GC is still unclear. The present work found CTSK as a potential predictive biomarker and immunotherapeutic target for GC based on the tumor microenvironment (TME). METHODS: From public databases, gene expression profiles and clinical data of GC were downloaded to analyze the distribution of stromal and immune cells and tumor abundance in TME. Differentially expressed genes (DEGs) associated with TME were obtained by differential analysis, followed by cross-screening to obtain CTSK as a gene associated with TME. Next, a series of methods and tools were employed to explore the relationships between clinicopathological features of GC and CTSK expression as well as prognosis, tumor immune microenvironment, immune checkpoints and drug sensitivity. And GSEA was used to investigate the potential role of CTSK in the tumor microenvironment of GC. RESULTS: From the dataset, we obtained a total of 656 DEGs associated with TME and the stromal component of TME was found to be closely involved in GC prognosis. CTSK was cross-screened as the key gene associated with TME by the PPI network and univariate Cox regression analysis. Pan-cancer analysis revealed significant high expression of CTSK in a variety of cancers. Subsequently, we hypothesized that high-expressed CTSK was closely correlated with poor prognosis and lymph node metastasis of tumors, and that CTSK, a GC TME-related gene, was largely involved in a range of biological behaviors of tumors, with a significant correlation between several immune cells. CONCLUSION: CTSK was validated as a potential prognostic biomarker related to TME of GC and could be a promising next-generation immunotherapeutic target for GC.

Prognostic utility of TME-associated genes in pancreatic cancer.

Background: Pancreatic cancer (PC) is a deadly disease. The tumor microenvironment (TME) participates in PC oncogenesis. This study focuses on the assessment of the prognostic and treatment utility of TME-associated genes in PC. Methods: After obtaining the differentially expressed TME-related genes, univariate and multivariate Cox analyses and least absolute shrinkage and selection operator (LASSO) were performed to identify genes related to prognosis, and a risk model was established to evaluate risk scores, based on The Cancer Genome Atlas (TCGA) data set, and it was validated by external data sets from the Gene Expression Omnibus (GEO) and Clinical Proteomic Tumor Analysis Consortium (CPTAC). Multiomics analyses were adopted to explore the potential mechanisms, discover novel treatment targets, and assess the sensitivities of immunotherapy and chemotherapy. Results: Five TME-associated genes, namely, FERMT1, CARD9, IL20RB, MET, and MMP3, were identified and a risk score formula constructed. Next, their mRNA expressions were verified in cancer and normal pancreatic cells. Multiple algorithms confirmed that the risk model displayed a reliable ability of prognosis prediction and was an independent prognostic factor, indicating that high-risk patients had poor outcomes. Immunocyte infiltration, gene set enrichment analysis (GSEA), and single-cell analysis all showed a strong relationship between immune mechanism and low-risk samples. The risk score could predict the sensitivity of immunotherapy and some chemotherapy regimens, which included oxaliplatin and irinotecan. Various latent treatment targets (LAG3, TIGIT, and ARID1A) were addressed by mutation landscape based on the risk model. Conclusion: The risk model based on TME-related genes can reflect the prognosis of PC patients and functions as a novel set of biomarkers for PC therapy.

Machine Learning Reveals Synovial Fibroblast Genes Associated with Pain Affect Sensory Nerve Growth in Rheumatoid Arthritis.

It has been presumed that rheumatoid arthritis (RA) joint pain is related to inflammation in the synovium; however, recent studies reveal that pain scores in patients do not correlate with synovial inflammation. We identified a module of 815 genes associated with pain, using a novel machine learning approach, Graph-based Gene expression Module Identification (GbGMI), in samples from patients with longstanding RA, but limited synovial inflammation at arthroplasty, and validated this finding in an independent cohort of synovial biopsy samples from early, untreated RA patients. Single-cell RNA-seq analyses indicated these genes were most robustly expressed by lining layer fibroblasts and receptor-ligand interaction analysis predicted robust lining layer fibroblast crosstalk with pain sensitive CGRP+ dorsal root ganglion sensory neurons. Netrin-4, which is abundantly expressed by lining fibroblasts and associated with pain, significantly increased the branching of pain-sensitive CGRP+ neurons in vitro . We conclude GbGMI is a useful method for identifying a module of genes that associate with a clinical feature of interest. Using this approach, we find that Netrin-4 is produced by synovial fibroblasts in the absence of inflammation and can enhance the outgrowth of CGRP+ pain sensitive nerve fibers. One Sentence Summary: Machine Learning reveals synovial fibroblast genes related to pain affect sensory nerve growth in Rheumatoid Arthritis addresses unmet clinical need.

Application of Deep Learning on Single-cell RNA Sequencing Data Analysis: A Review.

Single-cell RNA sequencing (scRNA-seq) has become a routinely used technique to quantify the gene expression profile of thousands of single cells simultaneously. Analysis of scRNA-seq data plays an important role in the study of cell states and phenotypes, and has helped elucidate biological processes, such as those occurring during the development of complex organisms, and improved our understanding of disease states, such as cancer, diabetes, and coronavirus disease 2019 (COVID-19). Deep learning, a recent advance of artificial intelligence that has been used to address many problems involving large datasets, has also emerged as a promising tool for scRNA-seq data analysis, as it has a capacity to extract informative and compact features from noisy, heterogeneous, and high-dimensional scRNA-seq data to improve downstream analysis. The present review aims at surveying recently developed deep learning techniques in scRNA-seq data analysis, identifying key steps within the scRNA-seq data analysis pipeline that have been advanced by deep learning, and explaining the benefits of deep learning over more conventional analytic tools. Finally, we summarize the challenges in current deep learning approaches faced within scRNA-seq data and discuss potential directions for improvements in deep learning algorithms for scRNA-seq data analysis.