The Germ Cell Gene TDRD1 as an ERG Target Gene and a Novel Prostate Cancer Biomarker

Explainable artificial intelligence to predict and identify prostate cancer tissue by gene expression

BACKGROUND AND OBJECTIVE

Prostate cancer is one of the most prevalent forms of cancer in men worldwide. Traditional screening strategies such as serum PSA levels, which are not necessarily cancer-specific, or digital rectal exams, which are often inconclusive, are still the screening methods used for the disease. Some studies have focused on identifying biomarkers of the disease but none have been reported for diagnosis in routine clinical practice and few studies have provided tools to assist the pathologist in the decision-making process when analyzing prostate tissue. Therefore, a classifier is proposed to predict the occurrence of PCa that provides physicians with accurate predictions and understandable explanations.

METHODS

A selection of 47 genes was made based on differential expression between PCa and normal tissue, GO gene ontology as well as the literature to be used as input predictors for different machine learning methods based on eXplainable Artificial Intelligence. These methods were trained using different class-balancing strategies to build accurate classifiers using gene expression data from 550 samples from 'The Cancer Genome Atlas'. Our model was validated in four external cohorts with different ancestries, totaling 463 samples. In addition, a set of SHapley Additive exPlanations was provided to help clinicians understand the underlying reasons for each decision.

RESULTS

An in-depth analysis showed that the Random Forest algorithm combined with majority class downsampling was the best performing approach with robust statistical significance. Our method achieved an average sensitivity and specificity of 0.90 and 0.8 with an AUC of 0.84 across all databases. The relevance of DLX1, MYL9 and FGFR genes for PCa screening was demonstrated in addition to the important role of novel genes such as CAV2 and MYLK.

CONCLUSIONS

This model has shown good performance in 4 independent external cohorts of different ancestries and the explanations provided are consistent with each other and with the literature, opening a horizon for its application in clinical practice. In the near future, these genes, in combination with our model, could be applied to liquid biopsy to improve PCa screening.

The cancer testis antigen TDRD1 regulates prostate cancer proliferation by associating with snRNP biogenesis machinery

Prostate cancer is the most commonly diagnosed noncutaneous cancer in American men. TDRD1, a germ cell-specific gene, is erroneously expressed in more than half of prostate tumors, but its role in prostate cancer development remains elusive. In this study, we identified a PRMT5-TDRD1 signaling axis that regulates the proliferation of prostate cancer cells. PRMT5 is a protein arginine methyltransferase essential for small nuclear ribonucleoprotein (snRNP) biogenesis. Methylation of Sm proteins by PRMT5 is a critical initiation step for assembling snRNPs in the cytoplasm, and the final snRNP assembly takes place in Cajal bodies in the nucleus. By mass spectrum analysis, we found that TDRD1 interacts with multiple subunits of the snRNP biogenesis machinery. In the cytoplasm, TDRD1 interacts with methylated Sm proteins in a PRMT5-dependent manner. In the nucleus, TDRD1 interacts with Coilin, the scaffold protein of Cajal bodies. Ablation of TDRD1 in prostate cancer cells disrupted the integrity of Cajal bodies, affected the snRNP biogenesis, and reduced cell proliferation. Taken together, this study represents the first characterization of TDRD1 functions in prostate cancer development and suggests TDRD1 as a potential therapeutic target for prostate cancer treatment.

LncRNA/miRNA/mRNA Network Introduces Novel Biomarkers in Prostate Cancer

The construction of a competing endogenous RNA (ceRNA) network is an important step in the identification of the role of differentially expressed genes in cancers. In the current research, we used a number of bioinformatics tools to construct the ceRNA network in prostate cancer and identify the importance of these modules in predicting the survival of patients with this type of cancer. An assessment of microarray data of prostate cancer and normal samples using the Limma package led to the identification of differential expressed (DE) RNAs that we stratified into mRNA, lncRNA, and miRNAs, resulting in 684 DEmRNAs, including 437 downregulated DEmRNAs (such as TGM4 and SCGB1A1) and 241 upregulated DEmRNAs (such as TDRD1 and CRISP3); 6 DElncRNAs, including 1 downregulated DElncRNA (H19) and 5 upregulated DElncRNAs (such as PCA3 and PCGEM1); and 59 DEmiRNAs, including 30 downregulated DEmiRNAs (such as hsa-miR-1274a and hsa-miR-1274b) and 29 upregulated DEmiRNAs (such as hsa-miR-1268 and hsa-miR-1207-5p). The ceRNA network contained a total of 5 miRNAs, 5 lncRNAs, and 17 mRNAs. We identified hsa-miR-17, hsa-miR-93, hsa-miR-150, hsa-miR-25, PART1, hsa-miR-125b, PCA3, H19, RND3, and ITGB8 as the 10 hub genes in the ceRNA network. According to the ROC analysis, the expression levels of 19 hub genes showed a high diagnostic value. Taken together, we introduce a number of novel promising diagnostic biomarkers for prostate cancer.

A novel prognostic model for prostate cancer based on androgen biosynthetic and catabolic pathways

Prostate cancer (PCa) is one of the most common malignancies in males globally, and its pathogenesis is significantly related to androgen. As one of the important treatments for prostate cancer, androgen deprivation therapy (ADT) inhibits tumor proliferation by controlling androgen levels, either surgically or pharmacologically. However, patients treated with ADT inevitably develop biochemical recurrence and advance to castration-resistant prostate cancer which has been reported to be associated with androgen biosynthetic and catabolic pathways. Thus, gene expression profiles and clinical information of PCa patients were collected from TCGA, MSKCC, and GEO databases for consensus clustering based on androgen biosynthetic and catabolic pathways. Subsequently, a novel prognostic model containing 13 genes (AFF3, B4GALNT4, CD38, CHRNA2, CST2, ADGRF5, KLK14, LRRC31, MT1F, MT1G, SFTPA2, SLC7A4, TDRD1) was constructed by univariate cox regression, lasso regression, and multivariate cox regression. Patients were divided into two groups based on their risk scores: high risk (HS) and low risk (LS), and survival analysis was used to determine the difference in biochemical recurrence-free time between the two. The results were validated on the MSKCC dataset and the GEO dataset. Functional enrichment analysis revealed some pivotal pathways that may have an impact on the prognosis of patients including the CDK-RB-E2F axis, G2M checkpoint, and KRAS signaling. In addition, somatic mutation, immune infiltration, and drug sensitivity analyses were performed to further explore the characteristics of HS and LS groups. Besides, two potential therapeutic targets, BIRC5 and RHOC, were identified by us in prostate cancer. These results indicate that the prognostic model may serve as a predictive tool to guide clinical treatment and provide new insight into the basic research in prostate cancer.

Benchmarking and Testing Machine Learning Approaches with BARRA:CuRDa, a Curated RNA-Seq Database for Cancer Research

RNA-seq is gradually becoming the dominating technique employed to access the global gene expression in biological samples, allowing more flexible protocols and robust analysis. However, the nature of RNA-seq results imposes new data-handling challenges when it comes to computational analysis. With the increasing employment of machine learning (ML) techniques in biomedical sciences, databases that could provide curated data sets treated with state-of-the-art approaches already adapted to ML protocols, become essential for testing new algorithms. In this study, we present the Benchmarking of ARtificial intelligence Research: Curated RNA-seq Database (BARRA:CuRDa). BARRA:CuRDa was built exclusively for cancer research and is composed of 17 handpicked RNA-seq data sets for Homo sapiens that were gathered from the Gene Expression Omnibus, using rigorous filtering criteria. All data sets were individually submitted to sample quality analysis, removal of low-quality bases and artifacts from the experimental process, removal of ribosomal RNA, and estimation of transcript-level abundance. Moreover, all data sets were tested using standard approaches in the field, which allows them to be used as benchmark to new ML approaches. A feature selection analysis was also performed on each data set to investigate the biological accuracy of basic techniques. Results include genes already related to their specific tumoral tissue a large amount of long noncoding RNA and pseudogenes. BARRA:CuRDa is available at http://sbcb.inf.ufrgs.br/barracurda.

A Comprehensive Molecular and Clinical Analysis of the piRNA Pathway Genes in Ovarian Cancer

Simple Summary

Although ovarian cancer (OC) is one of the most lethal gynecological cancers, its development and progression remain poorly understood. The piRNA pathway is important for transposon defense and genome stability. piRNA maturation and function involve a number of genes known as the piRNA pathway genes. These genes have recently been implicated in cancer development and progression but information about their role in OC is limited. Our work aimed to provide a better understanding of the roles of piRNA pathway genes in OC. Through analyzing changes in the abundance of 10 piRNA pathway genes, we discovered gene expression differences in benign vs. cancer, chemosensitive vs. chemoresistant and post hormone treatment in OC samples and cells. Furthermore, we observed the differential effects of these genes on patient survival and OC cell invasion. Overall, this work supports a role of the piRNA pathway genes in OC progression and encourages further study of their clinical relevance.

Abstract

Ovarian cancer (OC) is one of the most lethal gynecological malignancies, yet molecular mechanisms underlying its origin and progression remain poorly understood. With increasing reports of piRNA pathway deregulation in various cancers, we aimed to better understand its role in OC through a comprehensive analysis of key genes: PIWIL1-4, DDX4, HENMT1, MAEL, PLD6, TDRD1,9 and mutants of PIWIL1 (P1∆17) and PIWIL2 (PL2L60). High-throughput qRT-PCR (n = 45) and CSIOVDB (n = 3431) showed differential gene expression when comparing benign ovarian tumors, low grade OC and high grade serous OC (HGSOC). Significant correlation of disparate piRNA pathway gene expression levels with better progression free, post-progression free and overall survival suggests a complex role of this pathway in OC. We discovered PIWIL3 expression in chemosensitive but not chemoresistant primary HGSOC cells, providing a potential target against chemoresistant disease. As a first, we revealed that follicle stimulating hormone increased PIWIL2 expression in OV-90 cells. PIWIL1, P1∆17, PIWIL2, PL2L60 and MAEL overexpression in vitro and in vivo decreased motility and invasion of OVCAR-3 and OV-90 cells. Interestingly, P1∆17 and PL2L60, induced increased motility and invasion compared to PIWIL1 and PIWIL2. Our results in HGSOC highlight the intricate role piRNA pathway genes play in the development of malignant neoplasms.

Effects of RNA Binding Proteins on the Prognosis and Malignant Progression in Prostate Cancer

Prostate cancer (PCa) is a common lethal malignancy in men. RNA binding proteins (RBPs) have been proven to regulate the biological processes of various tumors, but their roles in PCa remain less defined. In the present study, we used bioinformatics analysis to identify RBP genes with prognostic and diagnostic values. A total of 59 differentially expressed RBPs in PCa were obtained, comprising 28 upregulated and 31 downregulated RBP genes, which may play important roles in PCa. Functional enrichment analyses showed that these RBPs were mainly involved in mRNA processing, RNA splicing, and regulation of RNA splicing. Additionally, we identified nine RBP genes (EXO1, PABPC1L, REXO2, MBNL2, MSI1, CTU1, MAEL, YBX2, and ESRP2) and their prognostic values by a protein-protein interaction network and Cox regression analyses. The expression of these nine RBPs was validated using immunohistochemical staining between the tumor and normal samples. Further, the associations between the expression of these nine RBPs and pathological T staging, Gleason score, and lymph node metastasis were evaluated. Moreover, these nine RBP genes showed good diagnostic values and could categorize the PCa patients into two clusters with different malignant phenotypes. Finally, we constructed a prognostic model based on these nine RBP genes and validated them using three external datasets. The model showed good efficiency in predicting patient survival and was independent of other clinical factors. Therefore, our model could be used as a supplement for clinical factors to predict patient prognosis and thereby improve patient survival.

Identification of novel oncogenic events occurring early in prostate carcinogenesis using purified autologous malignant and non-malignant prostate epithelial cells

OBJECTIVE

To interrogate enriched prostate cancer cells and autologous non-malignant prostate epithelial cells from men with localized prostate cancer, in order to identify early oncogenic pathways.

PATIENTS AND METHODS

We collected malignant and matched non-malignant prostatectomy samples from men with adenocarcinoma involving two or more contiguous areas in only one lobe of the prostate. Tissue samples from both lobes were subjected to digestion and single-cell suspensions were prepared. Epithelial cell adhesion molecule-positive cells from cancerous and contralateral non-malignant (control) samples were isolated using magnetic beads, ensuring uniform populations were obtained for each donor. Unbiased RNA sequencing analysis was used to measure gene expression and for detection of transcribed mutations or splice variants that were over- or under-represented in malignant prostate epithelial cells relative to autologous control prostate epithelial cells.

RESULTS

From five patient samples we identified 17 genes that were altered in prostate cancer epithelial cells, with 82% of genes being downregulated. Three genes, TDRD1, ANGTL4, and CLDN3, were consistently upregulated in malignant tissue. Malignant cells from three of the five patients showed evidence of upregulated ERG signalling, however, only one of these contained a TMPRSS2-ERG rearrangement. We did not identify mutations, gene rearrangements, or splice variants that were consistent amongst the patients.

CONCLUSIONS

Events occurring early in prostate cancer oncogenesis in these samples were characterized by a predominant downregulation of gene expression along with upregulation of TDRD1, ANGTL4 and CLDN3. No consistent mutations or splice variants were observed, but upregulation of ERG signalling was seen both in the presence and absence of the classic TMPRSS2-ERG rearrangement.

The co-expression networks of differentially expressed RBPs with TFs and LncRNAs related to clinical TNM stages of cancers

BACKGROUND

RNA-binding proteins (RBPs) play important roles in cellular homeostasis by regulating the expression of thousands of transcripts, which have been reported to be involved in human tumorigenesis. Despite previous reports of the dysregulation of RBPs in cancers, the degree of dysregulation of RBPs in cancers and the intrinsic relevance between dysregulated RBPs and clinical TNM information remains unknown. Furthermore, the co-expressed networks of dysregulated RBPs with transcriptional factors and lncRNAs also require further investigation.

RESULTS

Here, we firstly analyzed the deviations of expression levels of 1,542 RBPs from 20 cancer types and found that (1) RBPs are dysregulated in almost all 20 cancer types, especially in BLCA, COAD, READ, STAD, LUAD, LUSC and GBM with proportion of deviation larger than 300% compared with non-RBPs in normal tissues. (2) Up- and down-regulated RBPs also show opposed patterns of differential expression in cancers and normal tissues. In addition, down-regulated RBPs show a greater degree of dysregulated expression than up-regulated RBPs do. Secondly, we analyzed the intrinsic relevance between dysregulated RBPs and clinical TNM information and found that (3) Clinical TNM information for two cancer types-CHOL and KICH-is shown to be closely related to patterns of differentially expressed RBPs (DE RBPs) by co-expression cluster analysis. Thirdly, we identified ten key RBPs (seven down-regulated and three up-regulated) in CHOL and seven key RBPs (five down-regulated and two up-regulated) in KICH by analyzing co-expression correlation networks. Fourthly, we constructed the co-expression networks of key RBPs between 1,570 TFs and 4,147 lncRNAs for CHOL and KICH, respectively.

CONCLUSIONS

These results may provide an insight into the understanding of the functions of RBPs in human carcinogenesis. Furthermore, key RBPs and the co-expressed networks offer useful information for potential prognostic biomarkers and therapeutic targets for patients with cancers at the N and M stages in two cancer types CHOL and KICH.

Research status and progress of the RNA or protein biomarkers for prostate cancer

Prostate cancer is a kind of male malignancy. Recently, a large number of studies have reported many potential biomarkers for the diagnosis and prognosis of prostate cancer. In this literature review, we have collected a number of potential biomarkers for prostate cancer reported in the last 5 years. Among them, some are undergoing Phase III clinical trials, and others have been approved by the US Food and Drug Administration. However, most are still in the period of basic research. The review will contribute to future research to find the biomarkers to guide clinicians to make personalized treatment decisions for each prostate cancer patient.

A Mild PUM1 Mutation Is Associated with Adult-Onset Ataxia, whereas Haploinsufficiency Causes Developmental Delay and Seizures

Certain mutations can cause proteins to accumulate in neurons, leading to neurodegeneration. We recently showed, however, that upregulation of a wild-type protein, Ataxin1, caused by haploinsufficiency of its repressor, the RNA-binding protein Pumilio1 (PUM1), also causes neurodegeneration in mice. We therefore searched for human patients with PUM1 mutations. We identified eleven individuals with either PUM1 deletions or de novo missense variants who suffer a developmental syndrome (Pumilio1-associated developmental disability, ataxia, and seizure; PADDAS). We also identified a milder missense mutation in a family with adult-onset ataxia with incomplete penetrance (Pumilio1-related cerebellar ataxia, PRCA). Studies in patient-derived cells revealed that the missense mutations reduced PUM1 protein levels by ∼25% in the adult-onset cases and by ∼50% in the infantile-onset cases; levels of known PUM1 targets increased accordingly. Changes in protein levels thus track with phenotypic severity, and identifying posttranscriptional modulators of protein expression should identify new candidate disease genes.

Expression analysis of cancer-testis genes in prostate cancer reveals candidates for immunotherapy

Prostate cancer is a prevalent disorder among men with a heterogeneous etiological background. Several molecular events and signaling perturbations have been found in this disorder. Among genes whose expressions have been altered during the prostate cancer development are cancer-testis antigens (CTAs). This group of antigens has limited expression in the normal adult tissues but aberrant expression in cancers. This property provides them the possibility to be used as cancer biomarkers and immunotherapeutic targets. Several CTAs have been shown to be immunogenic in prostate cancer patients and some of the have entered clinical trials. Based on the preliminary data obtained from these trials, it is expected that CTA-based therapeutic options are beneficial for at least a subset of prostate cancer patients.

Domain retention in transcription factor fusion genes and its biological and clinical implications: a pan-cancer study

Genomic rearrangements involving transcription factors (TFs) can form fusion proteins resulting in either enhanced, weakened, or even loss of TF activity. Functional domain (FD) retention is a critical factor in the activity of transcription factor fusion genes (TFFGs). A systematic investigation of FD retention in TFFGs and their outcome (e.g. expression changes) in a pan-cancer study has not yet been completed. Here, we examined the FD retention status in 386 TFFGs across 13 major cancer types and identified 83 TFFGs involving 67 TFs that retained FDs. To measure the potential biological relevance of TFs in TFFGs, we introduced a Major Active Isofusion Index (MAII) and built a prioritized TFFG network using MAII scores and the observed frequency of fusion positive samples. Interestingly, the four TFFGs (PML-RARA, RUNX1-RUNX1T1, TMPRSS2-ERG, and SFPQ-TFE3) with the highest MAII scores showed 50 differentially expressed target genes (DETGs) in fusion-positive versus fusion-negative cancer samples. DETG analysis revealed that they were involved in tumorigenesis-related processes in each cancer type. PLAU, which encodes plasminogen activator urokinase and serves as a biomarker for tumor invasion, was found to be consistently activated in the samples with the highest MAII scores. Among the 50 DETGs, 21 were drug targetable genes. Fourteen of these 21 DETGs were expressed in acute myeloid leukemia (AML) samples. Accordingly, we constructed an AML-specific TFFG network, which included 38 DETGs in RUNX1-RUNX1T1 or PML-RARA positive samples. In summary, this study revealed several TFFGs and their potential target genes, and provided insights into the clinical implications of TFFGs.