ConsensusPathDB 2022: molecular interactions update as a resource for network biology

Progress in toxicogenomics to protect human health

Toxicogenomics measures molecular features, such as transcripts, proteins, metabolites and epigenomic modifications, to understand and predict the toxicological effects of environmental and pharmaceutical exposures. Transcriptomics has become an integral tool in contemporary toxicology research owing to innovations in gene expression profiling that can provide mechanistic and quantitative information at scale. These data can be used to predict toxicological hazards through the use of transcriptomic biomarkers, network inference analyses, pattern-matching approaches and artificial intelligence. Furthermore, emerging approaches, such as high-throughput dose-response modelling, can leverage toxicogenomic data for human health protection even in the absence of predicting specific hazards. Finally, single-cell transcriptomics and multi-omics provide detailed insights into toxicological mechanisms. Here, we review the progress since the inception of toxicogenomics in applying transcriptomics towards toxicology testing and highlight advances that are transforming risk assessment.

Mutant huntingtin impairs neurodevelopment in human brain organoids through CHCHD2-mediated neurometabolic failure

Expansion of the glutamine tract (poly-Q) in the protein huntingtin (HTT) causes the neurodegenerative disorder Huntington's disease (HD). Emerging evidence suggests that mutant HTT (mHTT) disrupts brain development. To gain mechanistic insights into the neurodevelopmental impact of human mHTT, we engineered male induced pluripotent stem cells to introduce a biallelic or monoallelic mutant 70Q expansion or to remove the poly-Q tract of HTT. The introduction of a 70Q mutation caused aberrant development of cerebral organoids with loss of neural progenitor organization. The early neurodevelopmental signature of mHTT highlighted the dysregulation of the protein coiled-coil-helix-coiled-coil-helix domain containing 2 (CHCHD2), a transcription factor involved in mitochondrial integrated stress response. CHCHD2 repression was associated with abnormal mitochondrial morpho-dynamics that was reverted upon overexpression of CHCHD2. Removing the poly-Q tract from HTT normalized CHCHD2 levels and corrected key mitochondrial defects. Hence, mHTT-mediated disruption of human neurodevelopment is paralleled by aberrant neurometabolic programming mediated by dysregulation of CHCHD2, which could then serve as an early interventional target for HD.

Unveiling hidden connections in omics data via pyPARAGON: an integrative hybrid approach for disease network construction

Network inference or reconstruction algorithms play an integral role in successfully analyzing and identifying causal relationships between omics hits for detecting dysregulated and altered signaling components in various contexts, encompassing disease states and drug perturbations. However, accurate representation of signaling networks and identification of context-specific interactions within sparse omics datasets in complex interactomes pose significant challenges in integrative approaches. To address these challenges, we present pyPARAGON (PAgeRAnk-flux on Graphlet-guided network for multi-Omic data integratioN), a novel tool that combines network propagation with graphlets. pyPARAGON enhances accuracy and minimizes the inclusion of nonspecific interactions in signaling networks by utilizing network rather than relying on pairwise connections among proteins. Through comprehensive evaluations on benchmark signaling pathways, we demonstrate that pyPARAGON outperforms state-of-the-art approaches in node propagation and edge inference. Furthermore, pyPARAGON exhibits promising performance in discovering cancer driver networks. Notably, we demonstrate its utility in network-based stratification of patient tumors by integrating phosphoproteomic data from 105 breast cancer tumors with the interactome and demonstrating tumor-specific signaling pathways. Overall, pyPARAGON is a novel tool for analyzing and integrating multi-omic data in the context of signaling networks. pyPARAGON is available at https://github.com/netlab-ku/pyPARAGON.

Biodegradation of PVCs through in-vitro identification of Bacillus albus and computational pathway analysis of ABH enzyme

Microplastics pose significant challenges to ecosystems and organisms. They can be ingested by marine and terrestrial species, leading to potential health risks and ecological disruptions. This study aims to address the urgent need for effective remediation strategies by focusing on the biodegradation of microplastics, specifically polyvinyl chloride (PVC) derivatives, using the bacterial strain Bacillus albus. The study provides a comprehensive background on the accumulation of noxious substances in the environment and the importance of harnessing biodegradation as an eco-friendly method for pollutant elimination. The specific objective is to investigate the enzymatic capabilities of Bacillus albus, particularly the alpha/beta hydrolases (ABH), in degrading microplastics. To achieve this, in-silico studies were conducted, including analysis of the ABH protein sequence and its interaction with potential inhibitors targeting PVC derivatives. Docking scores of - 7.2 kcal/mol were obtained to evaluate the efficacy of the interactions. The study demonstrates the promising bioremediation prospects of Bacillus albus for microplastics, highlighting its potential as a key player in addressing microplastic pollution. The findings underscore the urgent need for further experimental validation and practical implementation of Bacillus albus in environmental remediation strategies.

Time-course cross-species transcriptomics reveals conserved hepatotoxicity pathways induced by repeated administration of cyclosporine A

Cyclosporine A (CsA) has shown efficacy against immunity-related diseases despite its toxicity in various organs, including the liver, emphasizing the need to elucidate its underlying hepatotoxicity mechanism. This study aimed to capture the alterations in genome-wide expression over time and the subsequent perturbations of corresponding pathways across species. Six data from humans, mice, and rats, including animal liver tissue, human liver microtissues, and two liver cell lines exposed to CsA toxic dose, were used. The microtissue exposed to CsA for 10 d was analyzed to obtain dynamically differentially expressed genes (DEGs). Single-time points data at 1, 3, 5, 7, and 28 d of different species were used to provide additional evidence. Using liver microtissue-based longitudinal design, DEGs that were consistently up- or down-regulated over time were captured, and the well-known mechanism involved in CsA toxicity was elucidated. Thirty DEGs that consistently changed in longitudinal data were also altered in 28-d rat in-house data with concordant expression. Some genes (e.g. TUBB2A, PLIN2, APOB) showed good concordance with identified DEGs in 1-d and 7-d mouse data. Pathway analysis revealed up-regulations of protein processing, asparagine N-linked glycosylation, and cargo concentration in the endoplasmic reticulum. Furthermore, the down-regulations of pathways related to biological oxidations and metabolite and lipid metabolism were elucidated. These pathways were also enriched in single-time-point data and conserved across species, implying their biological significance and generalizability. Overall, the human organoids-based longitudinal design coupled with cross-species validation provides temporal molecular change tracking, aiding mechanistic elucidation and biologically relevant biomarker discovery.

Cancer drug sensitivity estimation using modular deep Graph Neural Networks

Computational drug sensitivity models have the potential to improve therapeutic outcomes by identifying targeted drugs components that are tailored to the transcriptomic profile of a given primary tumor. The SMILES representation of molecules that is used by state-of-the-art drug-sensitivity models is not conducive for neural networks to generalize to new drugs, in part because the distance between atoms does not generally correspond to the distance between their representation in the SMILES strings. Graph-attention networks, on the other hand, are high-capacity models that require large training-data volumes which are not available for drug-sensitivity estimation. We develop a modular drug-sensitivity graph-attentional neural network. The modular architecture allows us to separately pre-train the graph encoder and graph-attentional pooling layer on related tasks for which more data are available. We observe that this model outperforms reference models for the use cases of precision oncology and drug discovery; in particular, it is better able to predict the specific interaction between drug and cell line that is not explained by the general cytotoxicity of the drug and the overall survivability of the cell line. The complete source code is available at https://zenodo.org/doi/10.5281/zenodo.8020945. All experiments are based on the publicly available GDSC data.

Decoding the molecular interplay in the central dogma: An overview of mass spectrometry-based methods to investigate protein-metabolite interactions

With the emergence of next-generation nucleotide sequencing and mass spectrometry-based proteomics and metabolomics tools, we have comprehensive and scalable methods to analyze the genes, transcripts, proteins, and metabolites of a multitude of biological systems. Despite the fascinating new molecular insights at the genome, transcriptome, proteome and metabolome scale, we are still far from fully understanding cellular organization, cell cycles and biology at the molecular level. Significant advances in sensitivity and depth for both sequencing as well as mass spectrometry-based methods allow the analysis at the single cell and single molecule level. At the same time, new tools are emerging that enable the investigation of molecular interactions throughout the central dogma of molecular biology. In this review, we provide an overview of established and recently developed mass spectrometry-based tools to probe metabolite-protein interactions-from individual interaction pairs to interactions at the proteome-metabolome scale.

Cell-type-specific consequences of mosaic structural variants in hematopoietic stem and progenitor cells

The functional impact and cellular context of mosaic structural variants (mSVs) in normal tissues is understudied. Utilizing Strand-seq, we sequenced 1,133 single-cell genomes from 19 human donors of increasing age, and discovered the heterogeneous mSV landscapes of hematopoietic stem and progenitor cells. While mSVs are continuously acquired throughout life, expanded subclones in our cohort are confined to individuals >60. Cells already harboring mSVs are more likely to acquire additional somatic structural variants, including megabase-scale segmental aneuploidies. Capitalizing on comprehensive single-cell micrococcal nuclease digestion with sequencing reference data, we conducted high-resolution cell-typing for eight hematopoietic stem and progenitor cells. Clonally expanded mSVs disrupt normal cellular function by dysregulating diverse cellular pathways, and enriching for myeloid progenitors. Our findings underscore the contribution of mSVs to the cellular and molecular phenotypes associated with the aging hematopoietic system, and establish a foundation for deciphering the molecular links between mSVs, aging and disease susceptibility in normal tissues.

Broad proteomics analysis of seeding-induced aggregation of α-synuclein in M83 neurons reveals remodeling of proteostasis mechanisms that might contribute to Parkinson's disease pathogenesis

Aggregation of misfolded α-synuclein (α-syn) is a key characteristic feature of Parkinson's disease (PD) and related synucleinopathies. The nature of these aggregates and their contribution to cellular dysfunction is still not clearly elucidated. We employed mass spectrometry-based total and phospho-proteomics to characterize the underlying molecular and biological changes due to α-syn aggregation using the M83 mouse primary neuronal model of PD. We identified gross changes in the proteome that coincided with the formation of large Lewy body-like α-syn aggregates in these neurons. We used protein-protein interaction (PPI)-based network analysis to identify key protein clusters modulating specific biological pathways that may be dysregulated and identified several mechanisms that regulate protein homeostasis (proteostasis). The observed changes in the proteome may include both homeostatic compensation and dysregulation due to α-syn aggregation and a greater understanding of both processes and their role in α-syn-related proteostasis may lead to improved therapeutic options for patients with PD and related disorders.

A global phosphosite-correlated network map of Thousand And One Kinase 1 (TAOK1)

Thousand and one amino acid kinase 1 (TAOK1) is a sterile 20 family Serine/Threonine kinase linked to microtubule dynamics, checkpoint signaling, DNA damage response, and neurological functions. Molecular-level alterations of TAOK1 have been associated with neurodevelopment disorders and cancers. Despite their known involvement in physiological and pathophysiological processes, and as a core member of the hippo signaling pathway, the phosphoregulatory network of TAOK1 has not been visualized. Aimed to explore this network, we first analyzed the predominantly detected and differentially regulated TAOK1 phosphosites in global phosphoproteome datasets across diverse experimental conditions. Based on 709 qualitative and 210 quantitative differential cellular phosphoproteome datasets that were systematically assembled, we identified that phosphorylation at Ser421, Ser9, Ser965, and Ser445 predominantly represented TAOK1 in almost 75% of these datasets. Surprisingly, the functional role of all these phosphosites in TAOK1 remains unexplored. Hence, we employed a robust strategy to extract the phosphosites in proteins that significantly correlated in expression with predominant TAOK1 phosphosites. This led to the first categorization of the phosphosites including those in the currently known and predicted interactors, kinases, and substrates, that positively/negatively correlated with the expression status of each predominant TAOK1 phosphosites. Subsequently, we also analyzed the phosphosites in core proteins of the hippo signaling pathway. Based on the TAOK1 phosphoregulatory network analysis, we inferred the potential role of the predominant TAOK1 phosphosites. Especially, we propose pSer9 as an autophosphorylation and TAOK1 kinase activity-associated phosphosite and pS421, the most frequently detected phosphosite in TAOK1, as a significant regulatory phosphosite involved in the maintenance of genome integrity. Considering that the impact of all phosphosites that predominantly represent each kinase is essential for the efficient interpretation of global phosphoproteome datasets, we believe that the approach undertaken in this study is suitable to be extended to other kinases for accelerated research.

Exploration of functional relations among differentially co-expressed genes identifies regulators in glioblastoma

The conventional computational approaches to investigating a disease confront inherent constraints as they often need to improve in delving beyond protein functional associations and grasping their deeper contextual significance within the disease framework. Such context-specificity can be explored using clinical data by evaluating the change in interaction between the biological entities in different conditions by investigating the differential co-expression relationships. We believe that the integration and analysis of differential co-expression and the functional relationships, primarily focusing on the source nodes, will open novel insights about disease progression as the source proteins could trigger signaling cascades, mostly because they are transcription factors, cell surface receptors, or enzymes that respond instantly to a particular stimulus. A thorough contextual investigation of these nodes could lead to a helpful beginning point for identifying potential causal linkages and guiding subsequent scientific investigations to uncover mechanisms underlying observed associations. Our methodology includes functional protein-protein Interaction (PPI) data and co-expression information and filters functional linkages through a series of critical steps, culminating in the identification of a robust set of regulators. Our analysis identified eleven key regulators-AKT1, BRCA1, CAMK2G, CUL1, FGFR3, KIF3A, NUP210, PRKACB, RAB8A, RPS6KA2 and TGFB3-in glioblastoma. These regulators play a pivotal role in disease classification, cell growth control, and patient survivability and exhibit associations with immune infiltrations and disease hallmarks. This underscores the importance of assessing correlation towards causality in unraveling complex biological insights.

Comparative analysis of diisononyl phthalate and di(isononyl)cyclohexane-1,2 dicarboxylate plasticizers in regulation of lipid metabolism in 3T3-L1 cells

Diisononyl phthalate (DINP) and di(isononyl)cyclohexane-1,2-dicarboxylate (DINCH) are plasticizers introduced to replace previously used phthalate plasticizers in polymeric products. Exposure to DINP and DINCH has been shown to impact lipid metabolism. However, there are limited studies that address the mechanisms of toxicity of these two plasticizers. Here, a comparative toxicity analysis has been performed to evaluate the impacts of DINP and DINCH on 3T3-L1 cells. The preadipocyte 3T3-L1 cells were exposed to 1, 10, and 100 μM of DINP or DINCH for 10 days and assessed for lipid accumulation, gene expression, and protein analysis. Lipid staining showed that higher concentrations of DINP and DINCH can induce adipogenesis. The gene expression analysis demonstrated that both DINP and DINCH could alter the expression of lipid-related genes involved in adipogenesis. DINP and DINCH upregulated Pparγ, Pparα, C/EBPα Fabp4, and Fabp5, while both compounds significantly downregulated Fasn and Gata2. Protein analysis showed that both DINP and DINCH repressed the expression of FASN. Additionally, we analyzed an independent transcriptome dataset encompassing temporal data on lipid differentiation within 3T3-L1 cells. Subsequently, we derived a gene set that accurately portrays significant pathways involved in lipid differentiation, which we subsequently subjected to experimental validation through quantitative polymerase chain reaction. In addition, we extended our analysis to encompass a thorough assessment of the expression profiles of this identical gene set across 40 discrete transcriptome datasets that have linked to diverse pathological conditions to foreseen any potential association with DINP and DINCH exposure. Comparative analysis indicated that DINP could be more effective in regulating lipid metabolism.

Tyr352 as a Predominant Phosphosite in the Understudied Kinase and Molecular Target, HIPK1: Implications for Cancer Therapy

Homeodomain-interacting protein kinase 1 (HIPK1) is majorly found in the nucleoplasm. HIPK1 is associated with cell proliferation, tumor necrosis factor-mediated cellular apoptosis, transcription regulation, and DNA damage response, and thought to play significant roles in health and common diseases such as cancer. Despite this, HIPK1 remains an understudied molecular target. In the present study, based on a systematic screening and mapping approach, we assembled 424 qualitative and 44 quantitative phosphoproteome datasets with 15 phosphosites in HIPK1 reported across multiple studies. These HIPK1 phosphosites were not currently attributed to any functions. Among them, Tyr352 within the kinase domain was identified as the predominant phosphosite modulated in 22 differential datasets. To analyze the functional association of HIPK1 Tyr352, we first employed a stringent criterion to derive its positively and negatively correlated protein phosphosites. Subsequently, we categorized the correlated phosphosites in known interactors, known/predicted kinases, and substrates of HIPK1, for their prioritized validation. Bioinformatics analysis identified their significant association with biological processes such as the regulation of RNA splicing, DNA-templated transcription, and cellular metabolic processes. HIPK1 Tyr352 was also identified to be upregulated in Her2+ cell lines and a subset of pancreatic and cholangiocarcinoma tissues. These data and the systems biology approach undertaken in the present study serve as a platform to explore the functional role of other phosphosites in HIPK1, and by extension, inform cancer drug discovery and oncotherapy innovation. In all, this study highlights the comprehensive phosphosite map of HIPK1 kinase and the first of its kind phosphosite-centric analysis of HIPK1 kinase based on global-level phosphoproteomics datasets derived from human cellular differential experiments across distinct experimental conditions.

Cystatin from the helminth Ascaris lumbricoides upregulates mevalonate and cholesterol biosynthesis pathways and immunomodulatory genes in human monocyte-derived dendritic cells

Background

Ascaris lumbricoides cystatin (Al-CPI) prevents the development of allergic airway inflammation and dextran-induced colitis in mice models. It has been suggested that helminth-derived cystatins inhibit cathepsins in dendritic cells (DC), but their immunomodulatory mechanisms are unclear. We aimed to analyze the transcriptional profile of human monocyte-derived DC (moDC) upon stimulation with Al-CPI to elucidate target genes and pathways of parasite immunomodulation.

Methods

moDC were generated from peripheral blood monocytes from six healthy human donors of Denmark, stimulated with 1 µM of Al-CPI, and cultured for 5 hours at 37°C. RNA was sequenced using TrueSeq RNA libraries and the NextSeq 550 v2.5 (75 cycles) sequencing kit (Illumina, Inc). After QC, reads were aligned to the human GRCh38 genome using Spliced Transcripts Alignment to a Reference (STAR) software. Differential expression was calculated by DESEq2 and expressed in fold changes (FC). Cell surface markers and cytokine production by moDC were evaluated by flow cytometry.

Results

Compared to unstimulated cells, Al-CPI stimulated moDC showed differential expression of 444 transcripts (|FC| ≥1.3). The top significant differences were in Kruppel-like factor 10 (KLF10, FC 3.3, PBH = 3 x 10-136), palladin (FC 2, PBH = 3 x 10-41), and the low-density lipoprotein receptor (LDLR, FC 2.6, PBH = 5 x 10-41). Upregulated genes were enriched in regulation of cholesterol biosynthesis by sterol regulatory element-binding proteins (SREBP) signaling pathways and immune pathways. Several genes in the cholesterol biosynthetic pathway showed significantly increased expression upon Al-CPI stimulation, even in the presence of lipopolysaccharide (LPS). Regarding the pathway of negative regulation of immune response, we found a significant decrease in the cell surface expression of CD86, HLA-DR, and PD-L1 upon stimulation with 1 µM Al-CPI.

Conclusion

Al-CPI modifies the transcriptome of moDC, increasing several transcripts encoding enzymes involved in cholesterol biosynthesis and SREBP signaling. Moreover, Al-CPI target several transcripts in the TNF-alpha signaling pathway influencing cytokine release by moDC. In addition, mRNA levels of genes encoding KLF10 and other members of the TGF beta and the IL-10 families were also modified by Al-CPI stimulation. The regulation of the mevalonate pathway and cholesterol biosynthesis suggests new mechanisms involved in DC responses to helminth immunomodulatory molecules.

Longitudinal plasma proteome profiling reveals the diversity of biomarkers for diagnosis and cetuximab therapy response of colorectal cancer

Cetuximab therapy is the major treatment for colorectal cancer (CRC), but drug resistance limits its effectiveness. Here, we perform longitudinal and deep proteomic profiling of 641 plasma samples originated from 147 CRC patients (CRCs) undergoing cetuximab therapy with multi-course treatment, and 90 healthy controls (HCs). COL12A1, THBS2, S100A8, and S100A9 are screened as potential proteins to distinguish CRCs from HCs both in plasma and tissue validation cohorts. We identify the potential biomarkers (RRAS2, MMP8, FBLN1, RPTOR, and IMPDH2) for the initial response prediction. In a longitudinal setting, we identify two clusters with distinct fluctuations and construct the model with high accuracy to predict the longitudinal response, further validated in the independent cohort. This study reveals the heterogeneity of different biomarkers for tumor diagnosis, the initial and longitudinal response prediction respectively in the first course and multi-course cetuximab treatment, may ultimately be useful in monitoring and intervention strategies for CRC.

Unearthing Insights into Metabolic Syndrome by Linking Drugs, Targets, and Gene Expressions Using Similarity Measures and Graph Theory

AIMS AND OBJECTIVES

Metabolic syndrome (MetS) is a group of metabolic disorders that includes obesity in combination with at least any two of the following conditions, i.e., insulin resistance, high blood pressure, low HDL cholesterol, and high triglycerides level. Treatment of this syndrome is challenging because of the multiple interlinked factors that lead to increased risks of type-2 diabetes and cardiovascular diseases. This study aims to conduct extensive in silico analysis to (i) find central genes that play a pivotal role in MetS and (ii) propose suitable drugs for therapy. Our objective is to first create a drug-disease network and then identify novel genes in the drug-disease network with strong associations to drug targets, which can help in increasing the therapeutical effects of different drugs. In the future, these novel genes can be used to calculate drug synergy and propose new drugs for the effective treatment of MetS.

METHODS

For this purpose, we (i) investigated associated drugs and pathways for MetS, (ii) employed eight different similarity measures to construct eight gene regulatory networks, (iii) chose an optimal network, where a maximum number of drug targets were central, (iv) determined central genes exhibiting strong associations with these drug targets and associated disease-causing pathways, and lastly (v) employed these candidate genes to propose suitable drugs.

RESULTS

Our results indicated (i) a novel drug-disease network complex, with (ii) novel genes associated with MetS.

CONCLUSION

Our developed drug-disease network complex closely represents MetS with associated novel findings and markers for an improved understanding of the disease and suggested therapy.

SMG: self-supervised masked graph learning for cancer gene identification

Cancer genomics is dedicated to elucidating the genes and pathways that contribute to cancer progression and development. Identifying cancer genes (CGs) associated with the initiation and progression of cancer is critical for characterization of molecular-level mechanism in cancer research. In recent years, the growing availability of high-throughput molecular data and advancements in deep learning technologies has enabled the modelling of complex interactions and topological information within genomic data. Nevertheless, because of the limited labelled data, pinpointing CGs from a multitude of potential mutations remains an exceptionally challenging task. To address this, we propose a novel deep learning framework, termed self-supervised masked graph learning (SMG), which comprises SMG reconstruction (pretext task) and task-specific fine-tuning (downstream task). In the pretext task, the nodes of multi-omic featured protein-protein interaction (PPI) networks are randomly substituted with a defined mask token. The PPI networks are then reconstructed using the graph neural network (GNN)-based autoencoder, which explores the node correlations in a self-prediction manner. In the downstream tasks, the pre-trained GNN encoder embeds the input networks into feature graphs, whereas a task-specific layer proceeds with the final prediction. To assess the performance of the proposed SMG method, benchmarking experiments are performed on three node-level tasks (identification of CGs, essential genes and healthy driver genes) and one graph-level task (identification of disease subnetwork) across eight PPI networks. Benchmarking experiments and performance comparison with existing state-of-the-art methods demonstrate the superiority of SMG on multi-omic feature engineering.

XGDAG: explainable gene-disease associations via graph neural networks

MOTIVATION

Disease gene prioritization consists in identifying genes that are likely to be involved in the mechanisms of a given disease, providing a ranking of such genes. Recently, the research community has used computational methods to uncover unknown gene-disease associations; these methods range from combinatorial to machine learning-based approaches. In particular, during the last years, approaches based on deep learning have provided superior results compared to more traditional ones. Yet, the problem with these is their inherent black-box structure, which prevents interpretability.

RESULTS

We propose a new methodology for disease gene discovery, which leverages graph-structured data using graph neural networks (GNNs) along with an explainability phase for determining the ranking of candidate genes and understanding the model's output. Our approach is based on a positive-unlabeled learning strategy, which outperforms existing gene discovery methods by exploiting GNNs in a non-black-box fashion. Our methodology is effective even in scenarios where a large number of associated genes need to be retrieved, in which gene prioritization methods often tend to lose their reliability.

AVAILABILITY AND IMPLEMENTATION

The source code of XGDAG is available on GitHub at: https://github.com/GiDeCarlo/XGDAG. The data underlying this article are available at: https://www.disgenet.org/, https://thebiogrid.org/, https://doi.org/10.1371/journal.pcbi.1004120.s003, and https://doi.org/10.1371/journal.pcbi.1004120.s004.

Capturing time-dependent activation of genes and stress-response pathways using transcriptomics in iPSC-derived renal proximal tubule cells

Transcriptomic analysis is a powerful method in the utilization of New Approach Methods (NAMs) for identifying mechanisms of toxicity and application to hazard characterization. With this regard, mapping toxicological events to time of exposure would be helpful to characterize early events. Here, we investigated time-dependent changes in gene expression levels in iPSC-derived renal proximal tubular-like cells (PTL) treated with five diverse compounds using TempO-Seq transcriptomics with the aims to evaluate the application of PTL for toxicity prediction and to report on temporal effects for the activation of cellular stress response pathways. PTL were treated with either 50 μM amiodarone, 10 μM sodium arsenate, 5 nM rotenone, or 300 nM tunicamycin over a temporal time course between 1 and 24 h. The TGFβ-type I receptor kinase inhibitor GW788388 (1 μM) was used as a negative control. Pathway analysis revealed the induction of key stress-response pathways, including Nrf2 oxidative stress response, unfolding protein response, and metal stress response. Early response genes per pathway were identified much earlier than 24 h and included HMOX1, ATF3, DDIT3, and several MT1 isotypes. GW788388 did not induce any genes within the stress response pathways above, but showed deregulation of genes involved in TGFβ inhibition, including downregulation of CYP24A1 and SERPINE1 and upregulation of WT1. This study highlights the application of iPSC-derived renal cells for prediction of cellular toxicity and sheds new light on the temporal and early effects of key genes that are involved in cellular stress response pathways.

MicroRNA control of the myogenic cell transcriptome and proteome: the role of miR-16

MicroRNAs (miRs) control stem cell biology and fate. Ubiquitously expressed and conserved miR-16 was the first miR implicated in tumorigenesis. miR-16 is low in muscle during developmental hypertrophy and regeneration. It is enriched in proliferating myogenic progenitor cells but is repressed during differentiation. The induction of miR-16 blocks myoblast differentiation and myotube formation, whereas knockdown enhances these processes. Despite a central role for miR-16 in myogenic cell biology, how it mediates its potent effects is incompletely defined. In this investigation, global transcriptomic and proteomic analyses after miR-16 knockdown in proliferating C2C12 myoblasts revealed how miR-16 influences myogenic cell fate. Eighteen hours after miR-16 inhibition, ribosomal protein gene expression levels were higher relative to control myoblasts and p53 pathway-related gene abundance was lower. At the protein level at this same time point, miR-16 knockdown globally upregulated tricarboxylic acid (TCA) cycle proteins while downregulating RNA metabolism-related proteins. miR-16 inhibition induced specific proteins associated with myogenic differentiation such as ACTA2, EEF1A2, and OPA1. We extend prior work in hypertrophic muscle tissue and show that miR-16 is lower in mechanically overloaded muscle in vivo. Our data collectively point to how miR-16 is implicated in aspects of myogenic cell differentiation. A deeper understanding of the role of miR-16 in myogenic cells has consequences for muscle developmental growth, exercise-induced hypertrophy, and regenerative repair after injury, all of which involve myogenic progenitors.

Mutational landscape of primary spinal cord astrocytoma

Primary spinal cord astrocytoma (SCA) is a rare disease. Knowledge about the molecular profiles of SCAs mostly comes from intracranial glioma; the pattern of genetic alterations of SCAs is not well understood. Herein, we describe genome-sequencing analyses of primary SCAs, aiming to characterize the mutational landscape of primary SCAs. We utilized whole exome sequencing (WES) to analyze somatic nucleotide variants (SNVs) and copy number variants (CNVs) among 51 primary SCAs. Driver genes were searched using four algorithms. GISTIC2 was used to detect significant CNVs. Additionally, recurrently mutated pathways were also summarized. A total of 12 driver genes were identified. Of those, H3F3A (47.1%), TP53 (29.4%), NF1 (19.6%), ATRX (17.6%), and PPM1D (17.6%) were the most frequently mutated genes. Furthermore, three novel driver genes seldom reported in glioma were identified: HNRNPC, SYNE1, and RBM10. Several germline mutations, including three variants (SLC16A8 rs2235573, LMF1 rs3751667, FAM20C rs774848096) that were associated with risk of brain glioma, were frequently observed in SCAs. Moreover, 12q14.1 (13.7%) encompassing the oncogene CDK4 was recurrently amplified and negatively affected patient prognosis. Besides frequently mutated RTK/RAS pathway and PI3K pathway, the cell cycle pathway controlling the phosphorylation of retinoblastoma protein (RB) was mutated in 39.2% of patients. Overall, a considerable degree of the somatic mutation landscape is shared between SCAs and brainstem glioma. Our work provides a key insight into the molecular profiling of primary SCAs, which might represent candidate drug targets and complement the molecular atlas of glioma. © 2023 The Pathological Society of Great Britain and Ireland.

A molecular signature defining exercise adaptation with ageing and in vivo partial reprogramming in skeletal muscle

Exercise promotes functional improvements in aged tissues, but the extent to which it simulates partial molecular reprogramming is unknown. Using transcriptome profiling from (1) a skeletal muscle-specific in vivo Oct3/4, Klf4, Sox2 and Myc (OKSM) reprogramming-factor expression murine model; (2) an in vivo inducible muscle-specific Myc induction murine model; (3) a translatable high-volume hypertrophic exercise training approach in aged mice; and (4) human exercise muscle biopsies, we collectively defined exercise-induced genes that are common to partial reprogramming. Late-life exercise training lowered murine DNA methylation age according to several contemporary muscle-specific clocks. A comparison of the murine soleus transcriptome after late-life exercise training to the soleus transcriptome after OKSM induction revealed an overlapping signature that included higher JunB and Sun1. Also, within this signature, downregulation of specific mitochondrial and muscle-enriched genes was conserved in skeletal muscle of long-term exercise-trained humans; among these was muscle-specific Abra/Stars. Myc is the OKSM factor most induced by exercise in muscle and was elevated following exercise training in aged mice. A pulse of MYC rewired the global soleus muscle methylome, and the transcriptome after a MYC pulse partially recapitulated OKSM induction. A common signature also emerged in the murine MYC-controlled and exercise adaptation transcriptomes, including lower muscle-specific Melusin and reactive oxygen species-associated Romo1. With Myc, OKSM and exercise training in mice, as well habitual exercise in humans, the complex I accessory subunit Ndufb11 was lower; low Ndufb11 is linked to longevity in rodents. Collectively, exercise shares similarities with genetic in vivo partial reprogramming. KEY POINTS: Advances in the last decade related to cellular epigenetic reprogramming (e.g. DNA methylome remodelling) toward a pluripotent state via the Yamanaka transcription factors Oct3/4, Klf4, Sox2 and Myc (OKSM) provide a window into potential mechanisms for combatting the deleterious effects of cellular ageing. Using global gene expression analysis, we compared the effects of in vivo OKSM-mediated partial reprogramming in skeletal muscle fibres of mice to the effects of late-life murine exercise training in muscle. Myc is the Yamanaka factor most induced by exercise in skeletal muscle, and so we compared the MYC-controlled transcriptome in muscle to Yamanaka factor-mediated and exercise adaptation mRNA landscapes in mice and humans. A single pulse of MYC is sufficient to remodel the muscle methylome. We identify partial reprogramming-associated genes that are innately altered by exercise training and conserved in humans, and propose that MYC contributes to some of these responses.

AFB1 and OTA Promote Immune Toxicity in Human LymphoBlastic T Cells at Transcriptomic Level

Aflatoxin B1 (AFB1) and ochratoxin A (OTA) are typical contaminants of food and feed, which have serious implications for human and animal health, even at low concentrations. Therefore, a transcriptomic study was carried out to analyze gene expression changes triggered by low doses of AFB1 and OTA (100 nM; 7 days), individually and combined, in human lymphoblastic T cells. RNA-sequencing analysis showed that AFB1-exposure resulted in 99 differential gene expressions (DEGs), while 77 DEGs were obtained in OTA-exposure and 3236 DEGs in the combined one. Overall, 16% of human genome expression was altered. Gene ontology analysis revealed, for all studied conditions, biological processes and molecular functions typically associated with the immune system. PathVisio analysis pointed to ataxia telangiectasia mutated signaling as the most significantly altered pathway in AFB1-exposure, glycolysis in OTA-exposure, and ferroptosis in the mixed condition (Z-score > 1.96; adjusted p-value ≤ 0.05). Thus, the results demonstrated the potential DNA damage caused by AFB1, the possible metabolic reprogramming promoted by OTA, and the plausible cell death with oxidative stress prompted by the mixed exposure. They may be considered viable mechanisms of action to promote immune toxicity in vitro.

Host M-CSF induced gene expression drives changes in susceptible and resistant mice-derived BMdMs upon Leishmania major infection

Leishmaniases are a group of diseases with different clinical manifestations. Macrophage-Leishmania interactions are central to the course of the infection. The outcome of the disease depends not only on the pathogenicity and virulence of the parasite, but also on the activation state, the genetic background, and the underlying complex interaction networks operative in the host macrophages. Mouse models, with mice strains having contrasting behavior in response to parasite infection, have been very helpful in exploring the mechanisms underlying differences in disease progression. We here analyzed previously generated dynamic transcriptome data obtained from Leishmania major (L. major) infected bone marrow derived macrophages (BMdMs) from resistant and susceptible mouse. We first identified differentially expressed genes (DEGs) between the M-CSF differentiated macrophages derived from the two hosts, and found a differential basal transcriptome profile independent of Leishmania infection. These host signatures, in which 75% of the genes are directly or indirectly related to the immune system, may account for the differences in the immune response to infection between the two strains. To gain further insights into the underlying biological processes induced by L. major infection driven by the M-CSF DEGs, we mapped the time-resolved expression profiles onto a large protein-protein interaction (PPI) network and performed network propagation to identify modules of interacting proteins that agglomerate infection response signals for each strain. This analysis revealed profound differences in the resulting responses networks related to immune signaling and metabolism that were validated by qRT-PCR time series experiments leading to plausible and provable hypotheses for the differences in disease pathophysiology. In summary, we demonstrate that the host's gene expression background determines to a large degree its response to L. major infection, and that the gene expression analysis combined with network propagation is an effective approach to help identifying dynamically altered mouse strain-specific networks that hold mechanistic information about these contrasting responses to infection.

Analysis of 1276 Haplotype-Resolved Genomes Allows Characterization of Cis- and Trans-Abundant Genes

Many methods for haplotyping have materialized, but their application on a significant scale has been rare to date. Here we summarize analyses that were carried out in 1092 genomes from the 1000 Genomes Consortium and validated in an unprecedented number of 184 PGP genomes that have been experimentally haplotype-resolved by application of the Long-Fragment Read (LFR) technology. These analyses provided first insights into the diplotypic nature of human genomes and its potential functional implications. Thus, protein-changing variants were not randomly distributed between the two homologues of 18,121 autosomal protein-coding genes but occurred significantly more frequently in cis than in trans configurations in virtually each of the 1276 phased genomes. This resulted in global cis/trans ratios of ~60:40, establishing "cis abundance" as a universal characteristic of diploid human genomes. This phenomenon was based on two different classes of genes, a larger one exhibiting cis configurations of protein-changing variants in excess, so-called "cis-abundant" genes, and a smaller one of "trans-abundant" genes. These two gene classes, which together constitute a common diplotypic exome, were further functionally distinguished by means of gene ontology (GO) and pathway enrichment analysis. Moreover, they were distinguishable in terms of their effects on the human interactome, where they constitute distinct cis and trans modules, as shown with network propagation on a large integrated protein-protein interaction network. These analyses, recently performed with updated database and analysis tools, further consolidated the characterization of cis- and trans-abundant genes while expanding previous results. In this chapter, we present the key results along with the materials and methods to motivate readers to investigate these findings independently and gain further insights into the diplotypic nature of genes and genomes.

Functional analysis of structural variants in single cells using Strand-seq

Somatic structural variants (SVs) are widespread in cancer, but their impact on disease evolution is understudied due to a lack of methods to directly characterize their functional consequences. We present a computational method, scNOVA, which uses Strand-seq to perform haplotype-aware integration of SV discovery and molecular phenotyping in single cells by using nucleosome occupancy to infer gene expression as a readout. Application to leukemias and cell lines identifies local effects of copy-balanced rearrangements on gene deregulation, and consequences of SVs on aberrant signaling pathways in subclones. We discovered distinct SV subclones with dysregulated Wnt signaling in a chronic lymphocytic leukemia patient. We further uncovered the consequences of subclonal chromothripsis in T cell acute lymphoblastic leukemia, which revealed c-Myb activation, enrichment of a primitive cell state and informed successful targeting of the subclone in cell culture, using a Notch inhibitor. By directly linking SVs to their functional effects, scNOVA enables systematic single-cell multiomic studies of structural variation in heterogeneous cell populations.

Network Analysis of Inflammatory Bowel Disease Research: Towards the Interactome

BACKGROUND AND AIMS

Modern views accept that inflammatory bowel diseases [IBD] emerge from complex interactions among the multiple components of a biological network known as the 'IBD interactome'. These diverse components belong to different functional levels including cells, molecules, genes and biological processes. This diversity can make it difficult to integrate available empirical information from human patients into a collective view of aetiopathogenesis, a necessary step to understand the interactome. Herein, we quantitatively analyse how the representativeness of components involved in human IBD and their relationships ha ve changed over time.

METHODS

A bibliographic search in PubMed retrieved 25 971 abstracts of experimental studies on IBD in humans, published between 1990 and 2020. Abstracts were scanned automatically for 1218 IBD interactome components proposed in recent reviews. The resulting databases are freely available and were visualized as networks indicating the frequency at which different components are referenced together within each abstract.

RESULTS

As expected, over time there was an increase in components added to the IBD network and heightened connectivity within and across functional levels. However, certain components were consistently studied together, forming preserved motifs in the networks. These overrepresented and highly linked components reflect main 'hypotheses' in IBD research in humans. Interestingly, 82% of the components cited in reviews were absent or showed low frequency, suggesting that many aspects of the proposed IBD interactome still have weak experimental support in humans.

CONCLUSIONS

A reductionist and fragmented approach to the study of IBD has prevailed in previous decades, highlighting the importance of transitioning towards a more integrated interactome framework.

The STRING database in 2023: protein-protein association networks and functional enrichment analyses for any sequenced genome of interest

Much of the complexity within cells arises from functional and regulatory interactions among proteins. The core of these interactions is increasingly known, but novel interactions continue to be discovered, and the information remains scattered across different database resources, experimental modalities and levels of mechanistic detail. The STRING database (https://string-db.org/) systematically collects and integrates protein-protein interactions-both physical interactions as well as functional associations. The data originate from a number of sources: automated text mining of the scientific literature, computational interaction predictions from co-expression, conserved genomic context, databases of interaction experiments and known complexes/pathways from curated sources. All of these interactions are critically assessed, scored, and subsequently automatically transferred to less well-studied organisms using hierarchical orthology information. The data can be accessed via the website, but also programmatically and via bulk downloads. The most recent developments in STRING (version 12.0) are: (i) it is now possible to create, browse and analyze a full interaction network for any novel genome of interest, by submitting its complement of encoded proteins, (ii) the co-expression channel now uses variational auto-encoders to predict interactions, and it covers two new sources, single-cell RNA-seq and experimental proteomics data and (iii) the confidence in each experimentally derived interaction is now estimated based on the detection method used, and communicated to the user in the web-interface. Furthermore, STRING continues to enhance its facilities for functional enrichment analysis, which are now fully available also for user-submitted genomes.

Systematic Analysis of Cellular Signaling Pathways and Therapeutic Targets for SLC45A3:ERG Fusion-Positive Prostate Cancer

ETS-related gene (ERG) fusion affects prostate cancer depending on the degree of expression of ERG. Solute Carrier Family 45 Member 3 (SLC45A3) is the second-most common 5′ partner gene of ERG rearrangement. However, the molecular pathological features of SLC45A3:ERG (S:E) fusion and therapeutic methods have not been studied at all. S:E fusion-positive cancers (n = 10) were selected from the Tumor Fusion Gene Data Portal website. Fusion-negative cancers (n = 50) were selected by sorting ERG expression level in descending order and selecting the bottom to 50th sample. Totally, 1325 ERG correlated genes were identified by a Pearson correlation test using over 0.3 of absolute correlation coefficiency (|R| > 0.3). Pathway analysis was performed using over-representation analysis of correlated genes, and seven cancer-related pathways (focal adhesion kinase (FAK)/PI3K-Akt, JAK-STAT, Notch, receptor tyrosine kinase/PDGF, TGF-β, VEGFA, and Wnt signaling) were identified. In particular, focal adhesion kinase (FAK)/PI3K-Akt signaling and JAK-STAT signaling were significantly enriched in S:E fusion-positive prostate cancer. We further identified therapeutic targets and candidate drugs for S:E fusion-positive prostate cancer using gene−drug network analysis. Interestingly, PDGFRA and PDGFRB were the most frequently predicted therapeutic targets, and imatinib targeted both genes. In this study, we provide extensive information on cellular signaling pathways involved in S:E fusion-positive prostate cancer and also suggest therapeutic methods.

Identification of miRNA biomarkers for breast cancer by combining ensemble regularized multinomial logistic regression and Cox regression

BACKGROUND

Breast cancer is one of the most common cancers in women. It is necessary to classify breast cancer subtypes because different subtypes need specific treatment. Identifying biomarkers and classifying breast cancer subtypes is essential for developing appropriate treatment methods for patients. MiRNAs can be easily detected in tumor biopsy and play an inhibitory or promoting role in breast cancer, which are considered promising biomarkers for distinguishing subtypes.

RESULTS

A new method combing ensemble regularized multinomial logistic regression and Cox regression was proposed for identifying miRNA biomarkers in breast cancer. After adopting stratified sampling and bootstrap sampling, the most suitable sample subset for miRNA feature screening was determined via ensemble 100 regularized multinomial logistic regression models. 124 miRNAs that participated in the classification of at least 3 subtypes and appeared at least 50 times in 100 integrations were screened as features. 22 miRNAs from the proposed feature set were further identified as the biomarkers for breast cancer by using Cox regression based on survival analysis. The accuracy of 5 methods on the proposed feature set was significantly higher than on the other two feature sets. The results of 7 biological analyses illustrated the rationality of the identified biomarkers.

CONCLUSIONS

The screened features can better distinguish breast cancer subtypes. Notably, the genes and proteins related to the proposed 22 miRNAs were considered oncogenes or inhibitors of breast cancer. 9 of the 22 miRNAs have been proved to be markers of breast cancer. Therefore, our results can be considered in future related research.

Multi-transcriptome analysis following an acute skeletal muscle growth stimulus yields tools for discerning global and MYC regulatory networks

Myc is a powerful transcription factor implicated in epigenetic reprogramming, cellular plasticity, and rapid growth as well as tumorigenesis. Cancer in skeletal muscle is extremely rare despite marked and sustained Myc induction during loading-induced hypertrophy. Here, we investigated global, actively transcribed, stable, and myonucleus-specific transcriptomes following an acute hypertrophic stimulus in mouse plantaris. With these datasets, we define global and Myc-specific dynamics at the onset of mechanical overload-induced muscle fiber growth. Data collation across analyses reveals an under-appreciated role for the muscle fiber in extracellular matrix remodeling during adaptation, along with the contribution of mRNA stability to epigenetic-related transcript levels in muscle. We also identify Runx1 and Ankrd1 (Marp1) as abundant myonucleus-enriched loading-induced genes. We observed that a strong induction of cell cycle regulators including Myc occurs with mechanical overload in myonuclei. Additionally, in vivo Myc-controlled gene expression in the plantaris was defined using a genetic muscle fiber-specific doxycycline-inducible Myc-overexpression model. We determined Myc is implicated in numerous aspects of gene expression during early-phase muscle fiber growth. Specifically, brief induction of Myc protein in muscle represses Reverbα, Reverbβ, and Myh2 while increasing Rpl3, recapitulating gene expression in myonuclei during acute overload. Experimental, comparative, and in silico analyses place Myc at the center of a stable and actively transcribed, loading-responsive, muscle fiber-localized regulatory hub. Collectively, our experiments are a roadmap for understanding global and Myc-mediated transcriptional networks that regulate rapid remodeling in postmitotic cells. We provide open webtools for exploring the five RNA-seq datasets as a resource to the field.

Pre-Training on In Vitro and Fine-Tuning on Patient-Derived Data Improves Deep Neural Networks for Anti-Cancer Drug-Sensitivity Prediction

Large-scale databases that report the inhibitory capacities of many combinations of candidate drug compounds and cultivated cancer cell lines have driven the development of preclinical drug-sensitivity models based on machine learning. However, cultivated cell lines have devolved from human cancer cells over years or even decades under selective pressure in culture conditions. Moreover, models that have been trained on in vitro data cannot account for interactions with other types of cells. Drug-response data that are based on patient-derived cell cultures, xenografts, and organoids, on the other hand, are not available in the quantities that are needed to train high-capacity machine-learning models. We found that pre-training deep neural network models of drug sensitivity on in vitro drug-sensitivity databases before fine-tuning the model parameters on patient-derived data improves the models’ accuracy and improves the biological plausibility of the features, compared to training only on patient-derived data. From our experiments, we can conclude that pre-trained models outperform models that have been trained on the target domains in the vast majority of cases.

Gradient tree boosting and network propagation for the identification of pan-cancer survival networks

Cancer survival prediction is typically done with uninterpretable machine learning techniques, e.g., gradient tree boosting. Therefore, additional steps are needed to infer biological plausibility of the predictions. Here, we describe a protocol that combines pan-cancer survival prediction with XGBoost tree-ensemble learning and subsequent propagation of the learned feature weights on protein interaction networks. This protocol is based on TCGA transcriptome data of 8,024 patients from 25 cancer types but can easily be adapted to cancer patient data from other sources.

For complete details on the use and execution of this protocol, please refer to Thedinga and Herwig (2022).

•

Efficient pan-cancer survival prediction with XGBoost

•

Network propagation with NetCore improves biological plausibility of features

•

Combined approach identifies pan-cancer survival networks

The 2022 Nucleic Acids Research database issue and the online molecular biology database collection

The 2022 Nucleic Acids Research Database Issue contains 185 papers, including 87 papers reporting on new databases and 85 updates from resources previously published in the Issue. Thirteen additional manuscripts provide updates on databases most recently published elsewhere. Seven new databases focus specifically on COVID-19 and SARS-CoV-2, including SCoV2-MD, the first of the Issue's Breakthrough Articles. Major nucleic acid databases reporting updates include MODOMICS, JASPAR and miRTarBase. The AlphaFold Protein Structure Database, described in the second Breakthrough Article, is the stand-out in the protein section, where the Human Proteoform Atlas and GproteinDb are other notable new arrivals. Updates from DisProt, FuzDB and ELM comprehensively cover disordered proteins. Under the metabolism and signalling section Reactome, ConsensusPathDB, HMDB and CAZy are major returning resources. In microbial and viral genomes taxonomy and systematics are well covered by LPSN, TYGS and GTDB. Genomics resources include Ensembl, Ensembl Genomes and UCSC Genome Browser. Major returning pharmacology resource names include the IUPHAR/BPS guide and the Therapeutic Target Database. New plant databases include PlantGSAD for gene lists and qPTMplants for post-translational modifications. The entire Database Issue is freely available online on the Nucleic Acids Research website (https://academic.oup.com/nar). Our latest update to the NAR online Molecular Biology Database Collection brings the total number of entries to 1645. Following last year's major cleanup, we have updated 317 entries, listing 89 new resources and trimming 80 discontinued URLs. The current release is available at http://www.oxfordjournals.org/nar/database/c/.