Impact of benzo[a]pyrene, PCB153 and sex hormones on human ESC-Derived thyroid follicles using single cell transcriptomics.

INTRODUCTION: Endocrine disruptors are compounds of manmade origin able to interfere with the endocrine system and constitute an important environmental concern. Indeed, detrimental effects on thyroid physiology and functioning have been described. Differences exist in the susceptibility of human sexes to the incidence of thyroid disorders, like autoimmune diseases or cancer. METHODS: To study how different hormonal environments impact the thyroid response to endocrine disruptors, we exposed human embryonic stem cell-derived thyroid organoids to physiological concentrations of sex hormones resembling the serum levels of human females post-ovulation or males of reproductive age for three days. Afterwards, we added 10 µM benzo[a]pyrene or PCB153 for 24 h and analyzed the transcriptome changes via single-cell RNA sequencing with differential gene expression and gene ontology analysis. RESULTS: The sex hormones receptors genes AR, ESR1, ESR2 and PGR were expressed at low levels. Among the thyroid markers, only TG resulted downregulated by benzo[a]pyrene or benzo[a]pyrene with the "male" hormones mix. Both hormone mixtures and benzo[a]pyrene alone upregulated ribosomal genes and genes involved in oxidative phosphorylation, while their combination decreased the expression compared to benzo[a]pyrene alone. The "male" mix and benzo[a]pyrene, alone or in combination, upregulated genes involved in lipid transport and metabolism (APOA1, APOC3, APOA4, FABP1, FABP2, FABP6). The combination of "male" hormones and benzo[a]pyrene induced also genes involved in inflammation and NFkB targets. Benzo[a]pyrene upregulated CYP1A1, CYP1B1 and NQO1 irrespective of the hormonal context. The induction was stronger in the "female" mix. Benzo[a]pyrene alone upregulated genes involved in cell cycle regulation, response to reactive oxygen species and apoptosis. PCB153 had a modest effect in presence of "male" hormones, while we did not observe any changes with the "female" mix. CONCLUSION: This work shows how single cell transcriptomics can be applied to selectively study the in vitro effects of endocrine disrupters and their interaction with different hormonal contexts.

Molecular insights into PCB neurotoxicity: Comparing transcriptomic responses across dopaminergic neurons, population blood cells, and Parkinson's disease pathology.

Parkinson's disease (PD) is a complex neurodegenerative disorder influenced by genetic factors and environmental exposures. Polychlorinated biphenyls (PCBs), a group of synthetic organic compounds, have been identified as potential environmental risk factors for neurodegenerative diseases, including PD. We explored PCB-induced neurotoxicity mechanisms using iPSC-derived dopaminergic neurons and assessed their transcriptomic responses to varying PCB concentrations (0.01 µM, 0.5 µM, and 10 µM). Specifically, we focused on PCB-180, a congener known for its accumulation in human brains. The exposure durations were 24 h and 74 h, allowing us to capture both short-term and more prolonged effects on gene expression patterns. We observed that PCB exposure led to the suppression of oxidative phosphorylation, synaptic function, and neurotransmitter release, implicating these pathways in PCB-induced neurotoxicity. In our comparative analysis, we noted similarities in PCB-induced changes with other PD-related compounds like MPP+ and rotenone. Our findings also aligned with gene expression changes in human blood derived from a population exposed to PCBs, highlighting broader inflammatory responses. Additionally, molecular patterns seen in iPSC-derived neurons were confirmed in postmortem PD brain tissues, validating our in vitro results. In conclusion, our study offers novel insights into the multifaceted impacts of PCB-induced perturbations on various cellular contexts relevant to PD. The use of iPSC-derived dopaminergic neurons allowed us to decipher intricate transcriptomic alterations, bridging the gap between in vitro and in vivo findings. This work underscores the potential role of PCB exposure in neurodegenerative diseases like PD, emphasizing the need to consider both systemic and cell specific effects.

Wnt7a Decreases Brain Endothelial Barrier Function Via ß-Catenin Activation.

The blood-brain barrier consists of tightly connected endothelial cells protecting the brain's microenvironment from the periphery. These endothelial cells are characterized by specific tight junction proteins such as Claudin-5 and Occludin, forming the endothelial barrier. Disrupting these cells might lead to blood-brain barrier dysfunction. The Wnt/ß-catenin signaling pathway can regulate the expression of these tight junction proteins and subsequent barrier permeability. The aim of this study was to investigate the in vitro effects of Wnt7a mediated ß-catenin signaling on endothelial barrier integrity. Mouse brain endothelial cells, bEnd.3, were treated with recombinant Wnt7a protein or XAV939, a selective inhibitor of Wnt/ß-catenin mediated transcription to modulate the Wnt signaling pathway. The involvement of Wnt/HIF1α signaling was investigated by inhibiting Hif1α signaling with Hif1α siRNA. Wnt7a stimulation led to activation and nuclear translocation of ß-catenin, which was inhibited by XAV939. Wnt7a stimulation decreased Claudin-5 expression mediated by ß-catenin and decreased endothelial barrier formation. Wnt7a increased Hif1α and Vegfa expression mediated by ß-catenin. However, Hif1α signaling pathway did not regulate tight junction proteins Claudin-5 and Occludin. Our data suggest that Wnt7a stimulation leads to a decrease in tight junction proteins mediated by the nuclear translocation of ß-catenin, which hampers proper endothelial barrier formation. This process might be crucial in initiating endothelial cell proliferation and angiogenesis. Although HIF1α did not modulate the expression of tight junction proteins, it might play a role in brain angiogenesis and underlie pathogenic mechanisms in Wnt/HIF1α signaling in diseases such as cerebral small vessel disease.

Investigation of the effects of phthalates on in vitro thyroid models with RNA-Seq and ATAC-Seq.

Introduction: Phthalates are a class of endocrine-disrupting chemicals that have been shown to negatively correlate with thyroid hormone serum levels in humans and to cause a state of hyperactivity in the thyroid. However, their mechanism of action is not well described at the molecular level. Methods: We analyzed the response of mouse thyroid organoids to the exposure to a biologically relevant dose range of the phthalates bis(2-ethylhexyl) phthalate (DEHP), di-iso-decylphthalate (DIDP), di-iso-nonylphthalate (DINP), and di-n-octylphthalate (DnOP) for 24 h and simultaneously analyzed mRNA and miRNA expression via RNA sequencing. Using the expression data, we performed differential expression and gene set enrichment analysis. We also exposed the human thyroid follicular epithelial cell line Nthy-ori 3-1 to 1 µM of DEHP or DINP for 5 days and analyzed changes in chromatin accessibility via ATAC-Seq. Results: Dose-series analysis showed how the expression of several genes increased or decreased at the highest dose tested. As expected with the low dosing scheme, the compounds induced a modest response on the transcriptome, as we identified changes in only mmu-miR-143-3p after DINP treatment and very few differentially expressed genes. No effect was observed on thyroid markers. Ing5, a component of histones H3 and H4 acetylation complexes, was consistently upregulated in three out of four conditions compared to control, and we observed a partial overlap among the genes differentially expressed by the treatments. Gene set enrichment analysis showed enrichment in the treatment samples of the fatty acid metabolism pathway and in the control of pathways related to the receptor signalling and extracellular matrix organization. ATAC-Seq analysis showed a general increase in accessibility compared to the control, however we did not identify significant changes in accessibility in the identified regions. Discussion: With this work, we showed that despite having only a few differentially expressed genes, diverse analysis methods could be applied to retrieve relevant information on phthalates, showing the potential of in vitro thyroid-relevant systems for the analysis of endocrine disruptors.

Analysis of cell-specific transcriptional responses in human colon tissue using CIBERSORTx.

Diet is an important determinant of overall health, and has been linked to the risk of various cancers. To understand the mechanisms involved, transcriptomic responses from human intervention studies are very informative. However, gene expression analysis of human biopsy material only represents the average profile of a mixture of cell types that can mask more subtle, but relevant cell-specific changes. Here, we use the CIBERSORTx algorithm to generate single-cell gene expression from human multicellular colon tissue. We applied the CIBERSORTx to microarray data from the PHYTOME study, which investigated the effects of different types of meat on transcriptional and biomarker changes relevant to colorectal cancer (CRC) risk. First, we used single-cell mRNA sequencing data from healthy colon tissue to generate a novel signature matrix in CIBERSORTx, then we determined the proportions and gene expression of each separate cell type. After comparison, cell proportion analysis showed a continuous upward trend in the abundance of goblet cells and stem cells, and a continuous downward trend in transit amplifying cells after the addition of phytochemicals in red meat products. The dietary intervention influenced the expression of genes involved in the growth and division of stem cells, the metabolism and detoxification of enterocytes, the translation and glycosylation of goblet cells, and the inflammatory response of innate lymphoid cells. These results show that our approach offers novel insights into the heterogeneous gene expression responses of different cell types in colon tissue during a dietary intervention.

Hypoxic oligodendrocyte precursor cell-derived VEGFA is associated with blood-brain barrier impairment.

Cerebral small vessel disease is characterised by decreased cerebral blood flow and blood-brain barrier impairments which play a key role in the development of white matter lesions. We hypothesised that cerebral hypoperfusion causes local hypoxia, affecting oligodendrocyte precursor cell-endothelial cell signalling leading to blood-brain barrier dysfunction as an early mechanism for the development of white matter lesions. Bilateral carotid artery stenosis was used as a mouse model for cerebral hypoperfusion. Pimonidazole, a hypoxic cell marker, was injected prior to humane sacrifice at day 7. Myelin content, vascular density, blood-brain barrier leakages, and hypoxic cell density were quantified. Primary mouse oligodendrocyte precursor cells were exposed to hypoxia and RNA sequencing was performed. Vegfa gene expression and protein secretion was examined in an oligodendrocyte precursor cell line exposed to hypoxia. Additionally, human blood plasma VEGFA levels were measured and correlated to blood-brain barrier permeability in normal-appearing white matter and white matter lesions of cerebral small vessel disease patients and controls. Cerebral blood flow was reduced in the stenosis mice, with an increase in hypoxic cell number and blood-brain barrier leakages in the cortical areas but no changes in myelin content or vascular density. Vegfa upregulation was identified in hypoxic oligodendrocyte precursor cells, which was mediated via Hif1α and Epas1. In humans, VEGFA plasma levels were increased in patients versus controls. VEGFA plasma levels were associated with increased blood-brain barrier permeability in normal appearing white matter of patients. Cerebral hypoperfusion mediates hypoxia induced VEGFA expression in oligodendrocyte precursor cells through Hif1α/Epas1 signalling. VEGFA could in turn increase BBB permeability. In humans, increased VEGFA plasma levels in cerebral small vessel disease patients were associated with increased blood-brain barrier permeability in the normal appearing white matter. Our results support a role of VEGFA expression in cerebral hypoperfusion as seen in cerebral small vessel disease.

Applying genomics in regulatory toxicology: a report of the ECETOC workshop on omics threshold on non-adversity.

In a joint effort involving scientists from academia, industry and regulatory agencies, ECETOC's activities in Omics have led to conceptual proposals for: (1) A framework that assures data quality for reporting and inclusion of Omics data in regulatory assessments; and (2) an approach to robustly quantify these data, prior to interpretation for regulatory use. In continuation of these activities this workshop explored and identified areas of need to facilitate robust interpretation of such data in the context of deriving points of departure (POD) for risk assessment and determining an adverse change from normal variation. ECETOC was amongst the first to systematically explore the application of Omics methods, now incorporated into the group of methods known as New Approach Methodologies (NAMs), to regulatory toxicology. This support has been in the form of both projects (primarily with CEFIC/LRI) and workshops. Outputs have led to projects included in the workplan of the Extended Advisory Group on Molecular Screening and Toxicogenomics (EAGMST) group of the Organisation for Economic Co-operation and Development (OECD) and to the drafting of OECD Guidance Documents for Omics data reporting, with potentially more to follow on data transformation and interpretation. The current workshop was the last in a series of technical methods development workshops, with a sub-focus on the derivation of a POD from Omics data. Workshop presentations demonstrated that Omics data developed within robust frameworks for both scientific data generation and analysis can be used to derive a POD. The issue of noise in the data was discussed as an important consideration for identifying robust Omics changes and deriving a POD. Such variability or "noise" can comprise technical or biological variation within a dataset and should clearly be distinguished from homeostatic responses. Adverse outcome pathways (AOPs) were considered a useful framework on which to assemble Omics methods, and a number of case examples were presented in illustration of this point. What is apparent is that high dimension data will always be subject to varying processing pipelines and hence interpretation, depending on the context they are used in. Yet, they can provide valuable input for regulatory toxicology, with the pre-condition being robust methods for the collection and processing of data together with a comprehensive description how the data were interpreted, and conclusions reached.

IsoTools: a flexible workflow for long-read transcriptome sequencing analysis.

MOTIVATION: Long-read transcriptome sequencing (LRTS) has the potential to enhance our understanding of alternative splicing and the complexity of this process requires the use of versatile computational tools, with the ability to accommodate various stages of the workflow with maximum flexibility. RESULTS: We introduce IsoTools, a Python-based LRTS analysis framework that offers a wide range of functionality for transcriptome reconstruction and quantification of transcripts. Furthermore, we integrate a graph-based method for identifying alternative splicing events and a statistical approach based on the beta-binomial distribution for detecting differential events. To demonstrate the effectiveness of our methods, we applied IsoTools to PacBio LRTS data of human hepatocytes treated with the histone deacetylase inhibitor valproic acid. Our results indicate that LRTS can provide valuable insights into alternative splicing, particularly in terms of complex and differential splicing patterns, in comparison to short-read RNA-seq. AVAILABILITY AND IMPLEMENTATION: IsoTools is available on GitHub and PyPI, and its documentation, including tutorials, CLI, and API references, can be found at https://isotools.readthedocs.io/.

A novel orthotopic mouse model replicates human lung cancer cachexia.

INTRODUCTION: Cancer cachexia, highly prevalent in lung cancer, is a debilitating syndrome characterized by involuntary loss of skeletal muscle mass and is associated with poor clinical outcome, decreased survival and negative impact on tumour therapy. Various lung tumour-bearing animal models have been used to explore underlying mechanisms of cancer cachexia. However, these models do not simulate anatomical and immunological features key to lung cancer and associated muscle wasting. Overcoming these shortcomings is essential to translate experimental findings into the clinic. We therefore evaluated whether a syngeneic, orthotopic lung cancer mouse model replicates systemic and muscle-specific alterations associated with human lung cancer cachexia. METHODS: Immune competent, 11 weeks old male 129S2/Sv mice, were randomly allocated to either (1) sham control group or (2) tumour-bearing group. Syngeneic lung epithelium-derived adenocarcinoma cells (K-rasG12D ; p53R172HΔG ) were inoculated intrapulmonary into the left lung lobe of the mice. Body weight and food intake were measured daily. At baseline and weekly after surgery, grip strength was measured and tumour growth and muscle volume were assessed using micro cone beam CT imaging. After reaching predefined surrogate survival endpoint, animals were euthanized, and skeletal muscles of the lower hind limbs were collected for biochemical analysis. RESULTS: Two-third of the tumour-bearing mice developed cachexia based on predefined criteria. Final body weight (-13.7 ± 5.7%; P < 0.01), muscle mass (-13.8 ± 8.1%; P < 0.01) and muscle strength (-25.5 ± 10.5%; P < 0.001) were reduced in cachectic mice compared with sham controls and median survival time post-surgery was 33.5 days until humane endpoint. Markers for proteolysis, both ubiquitin proteasome system (Fbxo32 and Trim63) and autophagy-lysosomal pathway (Gabarapl1 and Bnip3), were significantly upregulated, whereas markers for protein synthesis (relative phosphorylation of Akt, S6 and 4E-BP1) were significantly decreased in the skeletal muscle of cachectic mice compared with control. The cachectic mice exhibited increased pentraxin-2 (P < 0.001) and CXCL1/KC (P < 0.01) expression levels in blood plasma and increased mRNA expression of IκBα (P < 0.05) in skeletal muscle, indicative for the presence of systemic inflammation. Strikingly, RNA sequencing, pathway enrichment and miRNA expression analyses of mouse skeletal muscle strongly mirrored alterations observed in muscle biopsies of patients with lung cancer cachexia. CONCLUSIONS: We developed an orthotopic model of lung cancer cachexia in immune competent mice. Because this model simulates key aspects specific to cachexia in lung cancer patients, it is highly suitable to further investigate the underlying mechanisms of lung cancer cachexia and to test the efficacy of novel intervention strategies.

Thyroid-on-a-Chip: An Organoid Platform for In Vitro Assessment of Endocrine Disruption.

Thyroid is a glandular tissue in the human body in which the function can be severely affected by endocrine disrupting chemicals (EDCs). Current in vitro assays to test endocrine disruption by chemical compounds are largely based on 2D thyroid cell cultures, which often fail to precisely evaluate the safety of these compounds. New and more advanced 3D cell culture systems are urgently needed to better recapitulate the thyroid follicular architecture and functions and help to improve the predictive power of such assays. Herein, the development of a thyroid organoid-on-a-chip (OoC) device using polymeric membranous carriers is described. Mouse embryonic stem cell derived thyroid follicles are incorporated in a microfluidic chip for a 4 day experiment at a flow rate of 12 µL min-1 . A reversible seal provides a leak-tight sealing while enabling quick and easy loading/unloading of thyroid follicles. The OoC model shows a high degree of functionality, where organoids retain expression of key thyroid genes and a typical follicular structure. Finally, transcriptional changes following benzo[k]fluoranthene exposure in the OoC device demonstrate activation of the xenobiotic aryl hydrocarbon receptor pathway. Altogether, this OoC system is a physiologically relevant thyroid model, which will represent a valuable tool to test potential EDCs.

CODA: a combo-Seq data analysis workflow.

The analysis of the combined mRNA and miRNA content of a biological sample can be of interest for answering several research questions, like biomarkers discovery, or mRNA-miRNA interactions. However, the process is costly and time-consuming, separate libraries need to be prepared and sequenced on different flowcells. Combo-Seq is a library prep kit that allows us to prepare combined mRNA-miRNA libraries starting from very low total RNA. To date, no dedicated bioinformatics method exists for the processing of Combo-Seq data. In this paper, we describe CODA (Combo-seq Data Analysis), a workflow specifically developed for the processing of Combo-Seq data that employs existing free-to-use tools. We compare CODA with exceRpt, the pipeline suggested by the kit manufacturer for this purpose. We also evaluate how Combo-Seq libraries analysed with CODA perform compared with conventional poly(A) and small RNA libraries prepared from the same samples. We show that using CODA more successfully trimmed reads are recovered compared with exceRpt, and the difference is more dramatic with short sequencing reads. We demonstrate how Combo-Seq identifies as many genes and fewer miRNAs compared to the standard libraries, and how miRNA validation favours conventional small RNA libraries over Combo-Seq. The CODA code is available at https://github.com/marta-nazzari/CODA.

Multi-omics HeCaToS dataset of repeated dose toxicity for cardiotoxic & hepatotoxic compounds.

The data currently described was generated within the EU/FP7 HeCaToS project (Hepatic and Cardiac Toxicity Systems modeling). The project aimed to develop an in silico prediction system to contribute to drug safety assessment for humans. For this purpose, multi-omics data of repeated dose toxicity were obtained for 10 hepatotoxic and 10 cardiotoxic compounds. Most data were gained from in vitro experiments in which 3D microtissues (either hepatic or cardiac) were exposed to a therapeutic (physiologically relevant concentrations calculated through PBPK-modeling) or a toxic dosing profile (IC20 after 7 days). Exposures lasted for 14 days and samples were obtained at 7 time points (therapeutic doses: 2-8-24-72-168-240-336 h; toxic doses 0-2-8-24-72-168-240 h). Transcriptomics (RNA sequencing & microRNA sequencing), proteomics (LC-MS), epigenomics (MeDIP sequencing) and metabolomics (LC-MS & NMR) data were obtained from these samples. Furthermore, functional endpoints (ATP content, Caspase3/7 and O2 consumption) were measured in exposed microtissues. Additionally, multi-omics data from human biopsies from patients are available. This data is now being released to the scientific community through the BioStudies data repository ( https://www.ebi.ac.uk/biostudies/ ).

MicroRNA-4732-3p Is Dysregulated in Breast Cancer Patients with Cardiotoxicity, and Its Therapeutic Delivery Protects the Heart from Doxorubicin-Induced Oxidative Stress in Rats.

Anthracycline-induced cardiotoxicity is the most severe collateral effect of chemotherapy originated by an excess of oxidative stress in cardiomyocytes that leads to cardiac dysfunction. We assessed clinical data from patients with breast cancer receiving anthracyclines and searched for discriminating microRNAs between patients that developed cardiotoxicity (cases) and those that did not (controls), using RNA sequencing and regression analysis. Serum levels of 25 microRNAs were differentially expressed in cases versus controls within the first year after anthracycline treatment, as assessed by three different regression models (elastic net, Robinson and Smyth exact negative binomial test and random forest). MiR-4732-3p was the only microRNA identified in all regression models and was downregulated in patients that experienced cardiotoxicity. MiR-4732-3p was also present in neonatal rat cardiomyocytes and cardiac fibroblasts and was modulated by anthracycline treatment. A miR-4732-3p mimic was cardioprotective in cardiac and fibroblast cultures, following doxorubicin challenge, in terms of cell viability and ROS levels. Notably, administration of the miR-4732-3p mimic in doxorubicin-treated rats preserved cardiac function, normalized weight loss, induced angiogenesis, and decreased apoptosis, interstitial fibrosis and cardiac myofibroblasts. At the molecular level, miR-4732-3p regulated genes of TGFß and Hippo signaling pathways. Overall, the results indicate that miR-4732-3p is a novel biomarker of cardiotoxicity that has therapeutic potential against anthracycline-induced heart damage.

Quantitative Prediction of Steatosis in Patients with Non-Alcoholic Fatty Liver by Means of Hepatic MicroRNAs Present in Serum and Correlating with Hepatic Fat.

Non-alcoholic fatty liver disease (NAFLD) is the most prevalent form of chronic liver disease worldwide, but a reliable non-invasive method to quantify liver steatosis in primary healthcare is not available. Circulating microRNAs have been proposed as biomarkers of severe/advanced NAFLD (steatohepatitis and fibrosis). However, the use of circulating miRNAs to quantitatively assess the % of liver fat in suspected NAFLD patients has not been investigated. We performed global miRNA sequencing in two sets of samples: human livers from organ donors (n = 20), and human sera from biopsy-proven NAFLD patients (n = 23), both with a wide range of steatosis quantified in their liver biopsies. Partial least squares (PLS) regression combined with recursive feature elimination (RFE) was used to select miRNAs associated with steatosis. Moreover, regression models with only 2 or 3 miRNAs, with high biological relevance, were built. Comprehensive microRNA sequencing of liver and serum samples resulted in two sets of abundantly expressed miRNAs (418 in liver and 351 in serum). Pearson correlation analyses indicated that 18% of miRNAs in liver and 14.5% in serum were significantly associated with the amount of liver fat. PLS-RFE models demonstrated that 50 was the number of miRNAs providing the lowest error in both liver and serum models predicting steatosis. Comparison of the two miRNA subsets showed 19 coincident miRNAs that were ranked according to biological significance (guide/passenger strand, relative abundance in liver and serum, number of predicted lipid metabolism target genes, correlation significance, etc.). Among them, miR-10a-5p, miR-98-5p, miR-19a-3p, miR-30e-5p, miR-32-5p and miR-145-5p showed the highest biological relevance. PLS regression models with serum levels of 2−3 of these miRNAs predicted the % of liver fat with errors <5%.

Transcriptomics analysis of human iPSC-derived dopaminergic neurons reveals a novel model for sporadic Parkinson's disease.

Parkinson's disease (PD) is a progressive, neurodegenerative disease affecting over 1% of the population beyond 65 years of age. Although some PD cases are inheritable, the majority of PD cases occur in a sporadic manner. Risk factors comprise next to heredity, age, and gender also exposure to neurotoxins from for instance pesticides and herbicides. As PD is characterized by a loss of dopaminergic neurons in the substantia nigra, it is nearly impossible to access and extract these cells from patients for investigating disease mechanisms. The emergence of induced pluripotent stem (iPSC) technology allows differentiating and growing human dopaminergic neurons, which can be used for in vitro disease modeling. Here, we differentiated human iPSCs into dopaminergic neurons, and subsequently exposed the cells to increasing concentrations of the neurotoxin MPP+. Temporal transcriptomics analysis revealed a strong time- and dose-dependent response with genes over-represented across pathways involved in PD etiology such as "Parkinson's Disease", "Dopaminergic signaling" and "calcium signaling". Moreover, we validated this disease model by showing robust overlap with a meta-analysis of transcriptomics data from substantia nigra from post-mortem PD patients. The overlap included genes linked to e.g. mitochondrial dysfunction, neuron differentiation, apoptosis and inflammation. Our data shows, that MPP+-induced, human iPSC-derived dopaminergic neurons present molecular perturbations as observed in the etiology of PD. Therefore we propose iPSC-derived dopaminergic neurons as a foundation for a novel sporadic PD model to study the pathomolecular mechanisms of PD, but also to screen for novel anti-PD drugs and to develop and test new treatment strategies.

Predicting missing proteomics values using machine learning: Filling the gap using transcriptomics and other biological features.

Proteins are often considered the main biological element in charge of the different functions and structures of a cell. However, proteomics, the global study of all expressed proteins, often performed by mass spectrometry, is limited by its stochastic sampling and can only quantify a limited amount of protein per sample. Transcriptomics, which allows an exhaustive analysis of all expressed transcripts, is often used as a surrogate. However, the transcript level does not present a high level of correlation with the corresponding protein level, notably due to the existence of several post-transcriptional regulatory mechanisms. In this publication, we hypothesize that the missing protein values in proteomics could be predicted using machine learning regression methods, trained with many features extracted from transcriptomics, including known translational regulatory elements such as microRNAs and circular RNAs. After considering different machine learning algorithms applied on two different splitting strategies, we report that random forest can predict proteins in new samples out of transcriptomics data with good accuracy. The proposed pre-processing and model building scripts can be accessed on GitHub: https://github.com/jochotecoa/ml_proteomics.

R-ODAF: Omics data analysis framework for regulatory application.

Despite the widespread use of transcriptomics technologies in toxicology research, acceptance of the data by regulatory agencies to support the hazard assessment is still limited. Fundamental issues contributing to this are the lack of reproducibility in transcriptomics data analysis arising from variance in the methods used to generate data and differences in the data processing. While research applications are flexible in the way the data are generated and interpreted, this is not the case for regulatory applications where an unambiguous answer, possibly later subject to legal scrutiny, is required. A reference analysis framework would give greater credibility to the data and allow the practitioners to justify their use of an alternative bioinformatic process by referring to a standard. In this publication, we propose a method called omics data analysis framework for regulatory application (R-ODAF), which has been built as a user-friendly pipeline to analyze raw transcriptomics data from microarray and next-generation sequencing. In the R-ODAF, we also propose additional statistical steps to remove the number of false positives obtained from standard data analysis pipelines for RNA-sequencing. We illustrate the added value of R-ODAF, compared to a standard workflow, using a typical toxicogenomics dataset of hepatocytes exposed to paracetamol.

Hepatic Steatosis Contributes to the Development of Muscle Atrophy via Inter-Organ Crosstalk.

Individuals with hepatic steatosis often display several metabolic abnormalities including insulin resistance and muscle atrophy. Previously, we found that hepatic steatosis results in an altered hepatokine secretion profile, thereby inducing skeletal muscle insulin resistance via inter-organ crosstalk. In this study, we aimed to investigate whether the altered secretion profile in the state of hepatic steatosis also induces skeletal muscle atrophy via effects on muscle protein turnover. To investigate this, eight-week-old male C57BL/6J mice were fed a chow (4.5% fat) or a high-fat diet (HFD; 45% fat) for 12 weeks to induce hepatic steatosis, after which the livers were excised and cut into ~200-µm slices. Slices were cultured to collect secretion products (conditioned medium; CM). Differentiated L6-GLUT4myc myotubes were incubated with chow or HFD CM to measure glucose uptake. Differentiated C2C12 myotubes were incubated with chow or HFD CM to measure protein synthesis and breakdown, and gene expression via RNA sequencing. Furthermore, proteomics analysis was performed in chow and HFD CM. It was found that HFD CM caused insulin resistance in L6-GLUT4myc myotubes compared with chow CM, as indicated by a blunted insulin-stimulated increase in glucose uptake. Furthermore, protein breakdown was increased in C2C12 cells incubated with HFD CM, while there was no effect on protein synthesis. RNA profiling of C2C12 cells indicated that 197 genes were differentially expressed after incubation with HFD CM, compared with chow CM, and pathway analysis showed that pathways related to anatomical structure and function were enriched. Proteomics analysis of the CM showed that 32 proteins were differentially expressed in HFD CM compared with chow CM. Pathway enrichment analysis indicated that these proteins had important functions with respect to insulin-like growth factor transport and uptake, and affect post-translational processes, including protein folding, protein secretion and protein phosphorylation. In conclusion, the results of this study support the hypothesis that secretion products from the liver contribute to the development of muscle atrophy in individuals with hepatic steatosis.

Identifying novel transcript biomarkers for hepatocellular carcinoma (HCC) using RNA-Seq datasets and machine learning.

BACKGROUND: Hepatocellular carcinoma (HCC) is one of the leading causes of cancer death in the world owing to limitations in its prognosis. The current prognosis approaches include radiological examination and detection of serum biomarkers, however, both have limited efficiency and are ineffective in early prognosis. Due to such limitations, we propose to use RNA-Seq data for evaluating putative higher accuracy biomarkers at the transcript level that could help in early prognosis. METHODS: To identify such potential transcript biomarkers, RNA-Seq data for healthy liver and various HCC cell models were subjected to five different machine learning algorithms: random forest, K-nearest neighbor, Naïve Bayes, support vector machine, and neural networks. Various metrics, namely sensitivity, specificity, MCC, informedness, and AUC-ROC (except for support vector machine) were evaluated. The algorithms that produced the highest values for all metrics were chosen to extract the top features that were subjected to recursive feature elimination. Through recursive feature elimination, the least number of features were obtained to differentiate between the healthy and HCC cell models. RESULTS: From the metrics used, it is demonstrated that the efficiency of the known protein biomarkers for HCC is comparatively lower than complete transcriptomics data. Among the different machine learning algorithms, random forest and support vector machine demonstrated the best performance. Using recursive feature elimination on top features of random forest and support vector machine three transcripts were selected that had an accuracy of 0.97 and kappa of 0.93. Of the three transcripts, two were protein coding (PARP2-202 and SPON2-203) and one was a non-coding transcript (CYREN-211). Lastly, we demonstrated that these three selected transcripts outperformed randomly taken three transcripts (15,000 combinations), hence were not chance findings, and could then be an interesting candidate for new HCC biomarker development. CONCLUSION: Using RNA-Seq data combined with machine learning approaches can aid in finding novel transcript biomarkers. The three biomarkers identified: PARP2-202, SPON2-203, and CYREN-211, presented the highest accuracy among all other transcripts in differentiating the healthy and HCC cell models. The machine learning pipeline developed in this study can be used for any RNA-Seq dataset to find novel transcript biomarkers. Code: www.github.com/rajinder4489/ML_biomarkers.