Prediction of drug-induced nephrotoxicity and injury mechanisms with human induced pluripotent stem cell-derived cells and machine learning methods. Sci Rep. 2015;5:12337. [PMID: 26212763 PMCID: PMC4515747 DOI: 10.1038/srep12337]

The use of artificial intelligence in induced pluripotent stem cell-based technology over 10-year period: A systematic scoping review

BACKGROUND

Stem cell research, particularly in the domain of induced pluripotent stem cell (iPSC) technology, has shown significant progress. The integration of artificial intelligence (AI), especially machine learning (ML) and deep learning (DL), has played a pivotal role in refining iPSC classification, monitoring cell functionality, and conducting genetic analysis. These enhancements are broadening the applications of iPSC technology in disease modelling, drug screening, and regenerative medicine. This review aims to explore the role of AI in the advancement of iPSC research.

METHODS

In December 2023, data were collected from three electronic databases (PubMed, Web of Science, and Science Direct) to investigate the application of AI technology in iPSC processing.

RESULTS

This systematic scoping review encompassed 79 studies that met the inclusion criteria. The number of research studies in this area has increased over time, with the United States emerging as a leading contributor in this field. AI technologies have been diversely applied in iPSC technology, encompassing the classification of cell types, assessment of disease-specific phenotypes in iPSC-derived cells, and the facilitation of drug screening using iPSC. The precision of AI methodologies has improved significantly in recent years, creating a foundation for future advancements in iPSC-based technologies.

CONCLUSIONS

Our review offers insights into the role of AI in regenerative and personalized medicine, highlighting both challenges and opportunities. Although still in its early stages, AI technologies show significant promise in advancing our understanding of disease progression and development, paving the way for future clinical applications.

Assessment of osteopontin as an early nephrotoxicity indicator in human renal proximal tubule cells and its application in evaluating lanthanum-induced nephrotoxicity

Nephrotoxicity is a common adverse effect induced by various chemicals, necessitating the development of reliable toxicity screening models for nephrotoxicity assessment. In this study, we assessed a group of nephrotoxicity indicators derived from different toxicity pathways, including conventional endpoints and kidney tubular injury biomarkers such as clusterin (CLU), kidney injury molecule-I (KIM-1), osteopontin (OPN), and neutrophil gelatinase-associated lipocalin (NGAL), using HK-2 and induced pluripotent stem cells (iPSCs)-derived renal proximal tubular epithelial-like cells (PTLs). Among the biomarkers tested, OPN emerged as the most discerning and precise marker. The predictive potential of OPN was tested using a panel of 10 nephrotoxic and 5 non-nephrotoxic compounds. The results demonstrated that combining OPN with the half-maximal inhibitory concentration (IC50) enhanced the diagnostic accuracy in both cellular models. Additionally, PTLs cells showed superior predictive efficacy for nephrotoxicity compared to HK-2 cells in this investigation. The two cellular models were utilized to evaluate the nephrotoxicity of lanthanum. The findings indicated that lanthanum possesses nephrotoxic properties; however, the degree of nephrotoxicity was relatively low, consistent with the outcomes of in vivo experiments.

Sensitivity of Human Induced Pluripotent Stem Cells and Thereof Differentiated Kidney Proximal Tubular Cells towards Selected Nephrotoxins

Proximal tubular epithelial cells (PTEC) are constantly exposed to potentially toxic metabolites and xenobiotics. The regenerative potential of the kidney enables the replacement of damaged cells either via the differentiation of stem cells or the re-acquisition of proliferative properties of the PTEC. Nevertheless, it is known that renal function declines, suggesting that the deteriorated cells are not replaced by fully functional cells. To understand the possible causes of this loss of kidney cell function, it is crucial to understand the role of toxins during the regeneration process. Therefore, we investigated the sensitivity and function of human induced pluripotent stem cells (hiPSC), hiPSC differentiating, and hiPSC differentiated into proximal tubular epithelial-like cells (PTELC) to known nephrotoxins. hiPSC were differentiated into PTELC, which exhibited similar morphology to PTEC, expressed prototypical PTEC markers, and were able to undergo albumin endocytosis. When treated with two nephrotoxins, hiPSC and differentiating hiPSC were more sensitive to cisplatin than differentiated PTELC, whereas all stages were equally sensitive to cyclosporin A. Both toxins also had an inhibitory effect on albumin uptake. Our results suggest a high sensitivity of differentiating cells towards toxins, which could have an unfavorable effect on regenerative processes. To study this, our model of hiPSC differentiating into PTELC appears suitable.

Generation of proximal tubule-enhanced kidney organoids from human pluripotent stem cells

Kidney organoids derived from human pluripotent stem cells (hPSCs) are now being used as models of renal disease and nephrotoxicity screening. However, the proximal tubules (PTs), which are responsible for most kidney reabsorption functions, remain immature in kidney organoids with limited expression of critical transporters essential for nephron functionality. Here, we describe a protocol for improved specification of nephron progenitors from hPSCs that results in kidney organoids with elongated proximalized nephrons displaying improved PT maturity compared with those generated using standard kidney organoid protocols. We also describe a methodology for assessing the functionality of the PTs within the organoids and visualizing maturation markers via immunofluorescence. Using these assays, PT-enhanced organoids display increased expression of a range of critical transporters, translating to improved functionality measured by substrate uptake and transport. This protocol consists of an extended (13 d) monolayer differentiation phase, during which time hPSCs are exposed to nephron progenitor maintenance media (CDBLY2), better emulating human metanephric progenitor specification in vivo. Following nephron progenitor specification, the cells are aggregated and cultured as a three-dimensional micromass on an air-liquid interface to facilitate further differentiation and segmentation into proximalized nephrons. Experience in culturing hPSCs is required to conduct this protocol and expertise in kidney organoid generation is advantageous.

Cells sorted off hiPSC-derived kidney organoids coupled with immortalized cells reliably model the proximal tubule

Of late, numerous microphysiological systems have been employed to model the renal proximal tubule. Yet there is lack of research on refining the functions of the proximal tubule epithelial layer-selective filtration and reabsorption. In this report, pseudo proximal tubule cells extracted from human-induced pluripotent stem cell-derived kidney organoids are combined and cultured with immortalized proximal tubule cells. It is shown that the cocultured tissue is an impervious epithelium that offers improved levels of certain transporters, extracellular matrix proteins collagen and laminin, and superior glucose transport and P-glycoprotein activity. mRNA expression levels higher than those obtained from each cell type were detected, suggesting an anomalous synergistic crosstalk between the two. Alongside, the improvements in morphological characteristics and performance of the immortalized proximal tubule tissue layer exposed, upon maturation, to human umbilical vein endothelial cells are thoroughly quantified and compared. Glucose and albumin reabsorption, as well as xenobiotic efflux rates through P-glycoprotein were all improved. The data presented abreast highlight the advantages of the cocultured epithelial layer and the non-iPSC-based bilayer. The in vitro models presented herein can be helpful in personalized nephrotoxicity studies.

The Synergy between Deep Learning and Organs-on-Chips for High-Throughput Drug Screening: A Review

Organs-on-chips (OoCs) are miniature microfluidic systems that have arguably become a class of advanced in vitro models. Deep learning, as an emerging topic in machine learning, has the ability to extract a hidden statistical relationship from the input data. Recently, these two areas have become integrated to achieve synergy for accelerating drug screening. This review provides a brief description of the basic concepts of deep learning used in OoCs and exemplifies the successful use cases for different types of OoCs. These microfluidic chips are of potential to be assembled as highly potent human-on-chips with complex physiological or pathological functions. Finally, we discuss the future supply with perspectives and potential challenges in terms of combining OoCs and deep learning for image processing and automation designs.

Machine Learning in Tissue Engineering

Machine learning (ML) and artificial intelligence have accelerated scientific discovery, augmented clinical practice, and deepened fundamental understanding of many biological phenomena. ML technologies have now been applied to diverse areas of tissue engineering research, including biomaterial design, scaffold fabrication, and cell/tissue modeling. Emerging ML-empowered strategies include machine-optimized polymer synthesis, predictive modeling of scaffold fabrication processes, complex analyses of structure-function relationships, and deep learning of spatialized cell phenotypes and tissue composition. The emergence of ML in tissue engineering, while relatively recent, has already enabled increasingly complex and multivariate analyses of the relationships between biological, chemical, and physical factors in driving tissue regenerative outcomes. This review highlights the novel methodologies, emerging strategies, and areas of potential growth within this rapidly evolving area of research. Impact statement Machine learning (ML) has accelerated scientific discovery and augmented clinical practice across multiple fields. Now, ML has driven exciting new paradigms in tissue engineering research, including machine-optimized biomaterial design, predictive modeling of scaffold fabrication, and spatiotemporal analysis of cell and tissue systems. The emergence of ML in tissue engineering, while relatively recent, has already enabled increasingly complex analyses of the relationships between biological, chemical, and physical factors in driving tissue regenerative outcomes. This review highlights the novel methodologies, emerging strategies, and areas of potential growth within this rapidly evolving area of research.

Examination of the emerging role of transporters in the assessment of nephrotoxicity

INTRODUCTION

The kidney is vulnerable to various injuries based on its function in the elimination of many xenobiotics, endogenous substances and metabolites. Since transporters are critical for the renal elimination of those substances, it is urgent to understand the emerging role of transporters in nephrotoxicity.

AREAS COVERED

This review summarizes the contribution of major renal transporters to nephrotoxicity induced by some drugs or toxins; addresses the role of transporter-mediated endogenous metabolic disturbances in nephrotoxicity; and discusses the advantages and disadvantages of in vitro models based on transporter expression and function.

EXPERT OPINION

Due to the crucial role of transporters in the renal disposition of xenobiotics and endogenous substances, it is necessary to further elucidate their renal transport mechanisms and pay more attention to the underlying relationship between the transport of endogenous substances and nephrotoxicity. Considering the species differences in the expression and function of transporters, and the low expression of transporters in general cell models, in vitro humanized models, such as humanized 3D organoids, shows significant promise in nephrotoxicity prediction and mechanism study.

Tissue Culture Models of AKI: From Tubule Cells to Human Kidney Organoids

AKI affects approximately 13.3 million people around the world each year, causing CKD and/or mortality. The mammalian kidney cannot generate new nephrons after postnatal renal damage and regenerative therapies for AKI are not available. Human kidney tissue culture systems can complement animal models of AKI and/or address some of their limitations. Donor-derived somatic cells, such as renal tubule epithelial cells or cell lines (RPTEC/hTERT, ciPTEC, HK-2, Nki-2, and CIHP-1), have been used for decades to permit drug toxicity screening and studies into potential AKI mechanisms. However, tubule cell lines do not fully recapitulate tubular epithelial cell properties in situ when grown under classic tissue culture conditions. Improving tissue culture models of AKI would increase our understanding of the mechanisms, leading to new therapeutics. Human pluripotent stem cells (hPSCs) can be differentiated into kidney organoids and various renal cell types. Injury to human kidney organoids results in renal cell-type crosstalk and upregulation of kidney injury biomarkers that are difficult to induce in primary tubule cell cultures. However, current protocols produce kidney organoids that are not mature and contain off-target cell types. Promising bioengineering techniques, such as bioprinting and "kidney-on-a-chip" methods, as applied to kidney nephrotoxicity modeling advantages and limitations are discussed. This review explores the mechanisms and detection of AKI in tissue culture, with an emphasis on bioengineered approaches such as human kidney organoid models.

Organotypic and Microphysiological Human Tissue Models for Drug Discovery and Development-Current State-of-the-Art and Future Perspectives

The number of successful drug development projects has been stagnant for decades despite major breakthroughs in chemistry, molecular biology, and genetics. Unreliable target identification and poor translatability of preclinical models have been identified as major causes of failure. To improve predictions of clinical efficacy and safety, interest has shifted to three-dimensional culture methods in which human cells can retain many physiologically and functionally relevant phenotypes for extended periods of time. Here, we review the state of the art of available organotypic culture techniques and critically review emerging models of human tissues with key importance for pharmacokinetics, pharmacodynamics, and toxicity. In addition, developments in bioprinting and microfluidic multiorgan cultures to emulate systemic drug disposition are summarized. We close by highlighting important trends regarding the fabrication of organotypic culture platforms and the choice of platform material to limit drug absorption and polymer leaching while supporting the phenotypic maintenance of cultured cells and allowing for scalable device fabrication. We conclude that organotypic and microphysiological human tissue models constitute promising systems to promote drug discovery and development by facilitating drug target identification and improving the preclinical evaluation of drug toxicity and pharmacokinetics. There is, however, a critical need for further validation, benchmarking, and consolidation efforts ideally conducted in intersectoral multicenter settings to accelerate acceptance of these novel models as reliable tools for translational pharmacology and toxicology. SIGNIFICANCE STATEMENT: Organotypic and microphysiological culture of human cells has emerged as a promising tool for preclinical drug discovery and development that might be able to narrow the translation gap. This review discusses recent technological and methodological advancements and the use of these systems for hit discovery and the evaluation of toxicity, clearance, and absorption of lead compounds.

Moving Towards Induced Pluripotent Stem Cell-based Therapies with Artificial Intelligence and Machine Learning

The advent of induced pluripotent stem cell (iPSC) technology, which allows to transform one cell type into another, holds the promise to produce therapeutic cells and organs on demand. Realization of this objective is contingent on the ability to demonstrate quality and safety of the cellular product for its intended use. Bottlenecks and backlogs to the clinical use of iPSCs have been fully outlined and a need has emerged for safer and standardized protocols to trigger cell reprogramming and functional differentiation. Amidst great challenges, in particular associated with lengthy culture time and laborious cell characterization, a demand for faster and more accurate methods for the validation of cell identity and function at different stages of the iPSC manufacturing process has risen. Artificial intelligence-based methods are proving helpful for these complex tasks and might revolutionize the way iPSCs are managed to create surrogate cells and organs. Here, we briefly review recent progress in artificial intelligence approaches for evaluation of iPSCs and their derivatives in experimental studies.

In silico approaches in organ toxicity hazard assessment: Current status and future needs for predicting heart, kidney and lung toxicities

The kidneys, heart and lungs are vital organ systems evaluated as part of acute or chronic toxicity assessments. New methodologies are being developed to predict these adverse effects based on in vitro and in silico approaches. This paper reviews the current state of the art in predicting these organ toxicities. It outlines the biological basis, processes and endpoints for kidney toxicity, pulmonary toxicity, respiratory irritation and sensitization as well as functional and structural cardiac toxicities. The review also covers current experimental approaches, including off-target panels from secondary pharmacology batteries. Current in silico approaches for prediction of these effects and mechanisms are described as well as obstacles to the use of in silico methods. Ultimately, a commonly accepted protocol for performing such assessment would be a valuable resource to expand the use of such approaches across different regulatory and industrial applications. However, a number of factors impede their widespread deployment including a lack of a comprehensive mechanistic understanding, limited in vitro testing approaches and limited in vivo databases suitable for modeling, a limited understanding of how to incorporate absorption, distribution, metabolism, and excretion (ADME) considerations into the overall process, a lack of in silico models designed to predict a safe dose and an accepted framework for organizing the key characteristics of these organ toxicants.

Generation and characterization of iPSC-derived renal proximal tubule-like cells with extended stability

The renal proximal tubule is responsible for re-absorption of the majority of the glomerular filtrate and its proper function is necessary for whole-body homeostasis. Aging, certain diseases and chemical-induced toxicity are factors that contribute to proximal tubule injury and chronic kidney disease progression. To better understand these processes, it would be advantageous to generate renal tissues from human induced pluripotent stem cells (iPSC). Here, we report the differentiation and characterization of iPSC lines into proximal tubular-like cells (PTL). The protocol is a step wise exposure of small molecules and growth factors, including the GSK3 inhibitor (CHIR99021), the retinoic acid receptor activator (TTNPB), FGF9 and EGF, to drive iPSC to PTL via cell stages representing characteristics of early stages of renal development. Genome-wide RNA sequencing showed that PTL clustered within a kidney phenotype. PTL expressed proximal tubular-specific markers, including megalin (LRP2), showed a polarized phenotype, and were responsive to parathyroid hormone. PTL could take up albumin and exhibited ABCB1 transport activity. The phenotype was stable for up to 7 days and was maintained after passaging. This protocol will form the basis of an optimized strategy for molecular investigations using iPSC derived PTL.

Human Pluripotent Stem-Cell-Derived Models as a Missing Link in Drug Discovery and Development

Human pluripotent stem cells (hPSCs), including human embryonic stem cells (hESCs) and human-induced pluripotent stem cells (hiPSCs), have the potential to accelerate the drug discovery and development process. In this review, by analyzing each stage of the drug discovery and development process, we identified the active role of hPSC-derived in vitro models in phenotypic screening, target-based screening, target validation, toxicology evaluation, precision medicine, clinical trial in a dish, and post-clinical studies. Patient-derived or genome-edited PSCs can generate valid in vitro models for dissecting disease mechanisms, discovering novel drug targets, screening drug candidates, and preclinically and post-clinically evaluating drug safety and efficacy. With the advances in modern biotechnologies and developmental biology, hPSC-derived in vitro models will hopefully improve the cost-effectiveness and the success rate of drug discovery and development.

Prediction of drug-induced kidney injury in drug discovery

Drug induced kidney injury is one of the leading causes of failure of drug development programs in the clinic. Early prediction of renal toxicity potential of drugs is crucial to the success of drug candidates in the clinic. The dynamic nature of the functioning of the kidney and the presence of drug uptake proteins introduce additional challenges in the prediction of renal injury caused by drugs. Renal injury due to drugs can be caused by a wide variety of mechanisms and can be broadly classified as toxic or obstructive. Several biomarkers are available for in vitro and in vivo detection of renal injury. In vitro static and dynamic (microfluidic) cellular models and preclinical models can provide valuable information regarding the toxicity potential of drugs. Differences in pharmacology and subsequent disconnect in biomarker response, differences in the expression of transporter and enzyme proteins between in vitro to in vivo systems and between preclinical species and humans are some of the limitations of current experimental models. The progress in microfluidic (kidney-on-chip) platforms in combination with the ability of 3-dimensional cell culture can help in addressing some of these issues in the future. Finally, newer in silico and computational techniques like physiologically based pharmacokinetic modeling and machine learning have demonstrated potential in assisting prediction of drug induced kidney injury.

Stem cells based in vitro models: trends and prospects in biomaterials cytotoxicity studies

Advanced biomaterials are increasingly used for numerous medical applications from the delivery of cancer-targeted therapeutics to the treatment of cardiovascular diseases. The issues of foreign body reactions induced by biomaterials must be controlled for preventing treatment failure. Therefore, it is important to assess the biocompatibility and cytotoxicity of biomaterials on cell culture systems before proceeding to in vivo studies in animal models and subsequent clinical trials. Direct use of biomaterials on animals create technical challenges and ethical issues and therefore, the use of non-animal models such as stem cell cultures could be useful for determination of their safety. However, failure to recapitulate the complex in vivo microenvironment have largely restricted stem cell cultures for testing the cytotoxicity of biomaterials. Nevertheless, properties of stem cells such as their self-renewal and ability to differentiate into various cell lineages make them an ideal candidate for in vitro screening studies. Furthermore, the application of stem cells in biomaterials screening studies may overcome the challenges associated with the inability to develop a complex heterogeneous tissue using primary cells. Currently, embryonic stem cells, adult stem cells, and induced pluripotent stem cells are being used as in vitro preliminary biomaterials testing models with demonstrated advantages over mature primary cell or cell line based in vitro models. This review discusses the status and future directions of in vitro stem cell-based cultures and their derivatives such as spheroids and organoids for the screening of their safety before their application to animal models and human in translational research.

The Implementation of the Three Rs in Regulatory Toxicity and Biosafety Assessment: The Indian Perspective

Animal models have long served as a basis for scientific experimentation, biomedical research, drug development and testing, disease modelling and toxicity studies, as they are widely thought to provide meaningful, human-relevant predictions. However, many of these systems are resource intensive and time-consuming, have low predictive value and are associated with great social and ethical dilemmas. Often drugs appear to be effective and safe in these classical animal models, but later prove to be ineffective and/or unsafe in clinical trials. These issues have paved the way for a paradigm shift from the use of in vivo approaches, toward the 'science of alternatives'. This has fuelled several research and regulatory initiatives, including the ban on the testing of cosmetics on animals. The new paradigm has been shifted toward increasing the relevance of the models for human predictivity and translational efficacy, and this has resulted in the recent development of many new methodologies, from 3-D bio-organoids to bioengineered 'human-on-a-chip' models. These improvements have the potential to significantly advance medical research globally. This paper offers a stance on the existing strategies and practices that utilise alternatives to animals, and outlines progress on the incorporation of these models into basic and applied research and education, specifically in India. It also seeks to provide a strategic roadmap to streamline the future directions for the country's policy changes and investments. This strategic roadmap could be a useful resource to guide research institutions, industries, regulatory agencies, contract research organisations and other stakeholders in transitioning toward modern approaches to safety and risk assessment that could replace or reduce the use of animals without compromising the safety of humans or the environment.

hiPSCs for predictive modelling of neurodegenerative diseases: dreaming the possible

Human induced pluripotent stem cells (hiPSCs) were first generated in 2007, but the full translational potential of this valuable tool has yet to be realized. The potential applications of hiPSCs are especially relevant to neurology, as brain cells from patients are rarely available for research. hiPSCs from individuals with neuropsychiatric or neurodegenerative diseases have facilitated biological and multi-omics studies as well as large-scale screening of chemical libraries. However, researchers are struggling to improve the scalability, reproducibility and quality of this descriptive disease modelling. Addressing these limitations will be the first step towards a new era in hiPSC research - that of predictive disease modelling - involving the correlation and integration of in vitro experimental data with longitudinal clinical data. This approach is a key element of the emerging precision medicine paradigm, in which hiPSCs could become a powerful diagnostic and prognostic tool. Here, we consider the steps necessary to achieve predictive modelling of neurodegenerative disease with hiPSCs, using Huntington disease as an example.

Machine Learning in Predictive Toxicology: Recent Applications and Future Directions for Classification Models

In recent times, machine learning has become increasingly prominent in predictive toxicology as it has shifted from in vivo studies toward in silico studies. Currently, in vitro methods together with other computational methods such as quantitative structure-activity relationship modeling and absorption, distribution, metabolism, and excretion calculations are being used. An overview of machine learning and its applications in predictive toxicology is presented here, including support vector machines (SVMs), random forest (RF) and decision trees (DTs), neural networks, regression models, naïve Bayes, k-nearest neighbors, and ensemble learning. The recent successes of these machine learning methods in predictive toxicology are summarized, and a comparison of some models used in predictive toxicology is presented. In predictive toxicology, SVMs, RF, and DTs are the dominant machine learning methods due to the characteristics of the data available. Lastly, this review describes the current challenges facing the use of machine learning in predictive toxicology and offers insights into the possible areas of improvement in the field.

3D Cell Printing of Tissue/Organ-Mimicking Constructs for Therapeutic and Drug Testing Applications

The development of artificial tissue/organs with the functional maturity of their native equivalents is one of the long-awaited panaceas for the medical and pharmaceutical industries. Advanced 3D cell-printing technology and various functional bioinks are promising technologies in the field of tissue engineering that have enabled the fabrication of complex 3D living tissue/organs. Various requirements for these tissues, including a complex and large-volume structure, tissue-specific microenvironments, and functional vasculatures, have been addressed to develop engineered tissue/organs with native relevance. Functional tissue/organ constructs have been developed that satisfy such criteria and may facilitate both in vivo replenishment of damaged tissue and the development of reliable in vitro testing platforms for drug development. This review describes key developments in technologies and materials for engineering 3D cell-printed constructs for therapeutic and drug testing applications.

Differentiating Induced Pluripotent Stem Cells into Renal Cells: A New Approach to Treat Kidney Diseases

Renal disease is a major issue for global public health. Despite some progress in supportive care, the mortality rates among patients with this condition remain alarmingly high. Studies in pursuit of innovative strategies to treat renal diseases, especially stimulating kidney regeneration, have been developed. In this field, stem cell-based therapy has been a promising area. Induced pluripotent stem cell-derived renal cells (iPSC-RCs) represent an interesting source of cells for treating kidney diseases. Advances in regenerative medicine using iPSC-RCs and their application to the kidney are discussed in this review. Furthermore, the way differentiation protocols of induced pluripotent stem cells into renal cells may also be applied for the generation of kidney organoids is also described, contributing to studies in renal development, kidney diseases, and drug toxicity tests. The translation of the differentiation methodologies into animal model studies and the safety and feasibility of renal differentiated cells as a treatment for kidney injury are also highlighted. Although only few studies were published in this field, the results seem promising and support the use of iPSC-RCs as a potential therapy in the future.

Freshly isolated primary human proximal tubule cells as an in vitro model for the detection of renal tubular toxicity

Drug induced kidney injury (DIKI) is a common reason for compound attrition in drug development pipelines with proximal tubule epithelial cells (PTECs) most commonly associated with DIKI. Here, we investigated freshly isolated human (hPTECs) as an in vitro model for assessing renal tubular toxicity. The freshly isolated hPTECs were first characterized to confirm gene expression of important renal transporters involved in drug handling which was further corroborated by confirming the functional activity of organic cation transporter 2 and organic anion transporter 1 by using transporter specific inhibitors. Additionally, functionality of megalin/cubilin endocytic receptors was also confirmed. A training set of 36 compounds was used to test the ability of the model to classify them using six different endpoints which included three biomarkers (Kidney Injury Molecule-1, Neutrophil gelatinase-associated lipocalin, and Clusterin) and three non-specific injury endpoints (ATP depletion, LDH leakage, and barrier permeability via transepithelial electrical resistance) in a dose-dependent manner across two independent kidney donors. In general, biomarkers showed higher predictivity than non-specific endpoints, with Clusterin showing the highest predictivity (Sensitivity/Specificity - 65.0/93.8 %). By using the thresholds generated from the training set, nine candidate internal Takeda compounds were screened where PTEC toxicity was identified as one of the findings in preclinical animal studies. The model correctly classified four of six true positives and two of three true negatives, showing validation of the in vitro model for detection of tubular toxicants. This work thus shows the potential application of freshly isolated primary hPTECs using translational biomarkers in assessment of tubular toxicity within the drug discovery pipeline.

Predicting direct hepatocyte toxicity in humans by combining high-throughput imaging of HepaRG cells and machine learning-based phenotypic profiling

Accurate prediction of drug- and chemical-induced hepatotoxicity remains to be a problem for pharmaceutical companies as well as other industries and regulators. The goal of the current study was to develop an in vitro/in silico method for the rapid and accurate prediction of drug- and chemical-induced hepatocyte injury in humans. HepaRG cells were employed for high-throughput imaging in combination with phenotypic profiling. A reference set of 69 drugs and chemicals was screened at a range of 7 concentrations, and the cellular response values were used for training a supervised classifier and for determining assay performance by using tenfold cross-validation. The results showed that the best performing phenotypic features were related to nuclear translocation of RELA (RELA proto-oncogene, NF-kB subunit; also known as NF-kappa B p65), DNA organization, and the F-actin cytoskeleton. Using a subset of 30 phenotypic features, direct hepatocyte toxicity in humans could be predicted with a test sensitivity, specificity and balanced accuracy of 73%, 92%, and 83%, respectively. The method was applied to another set of 26 drugs and chemicals with unclear annotation and their hepatocyte toxicity in humans was predicted. The results also revealed that the identified discriminative phenotypic changes were related to cell death and cellular senescence. Whereas cell death-related endpoints are widely applied in in vitro toxicology, cellular senescence-related endpoints are not, although cellular senescence can be induced by various drugs and other small molecule compounds and plays an important role in liver injury and disease. These findings show how phenotypic profiling can reveal unexpected chemical-induced mechanisms in toxicology.

Genetically Engineered Human Kidney Cells for Real-Time Cytotoxicity Testing In Vitro

Classic toxicology studies often utilize in vivo animal models. Newer approaches employing in vitro organ-specific cellular models have been developed in recent years to help accelerate the speed and reduce the cost of traditional toxicology testing. Toward the goal of supporting in vitro cellular model research with a regulatory application in mind, we have developed a 'designer' human kidney cell line called HK2-Vi that can fluorescently measure the cytotoxicity of potential toxins on proximal tubule cell viability in a direct exposure in vitro model. HK2-Vi was designed to be a reagent-less kinetic assay that can yield data on short- or long-term cell viability after toxin exposure. To generate HK2-Vi, we used monocistronic lentiviral transduction methods to genetically engineer a human kidney cell line called HK-2 to stably co-express two transgenes. The first is Perceval HR, which encodes a fluorescent biosensor of both cytosolic ATP and ADP and the second is pHRed, which encodes a biosensor of cytosolic pH. Relative levels of cellular ATP and ADP effectively serve as a reliable and robust indicator of cell viability. Because the fluorescence Perceval HR is pH-dependent, we co-expressed the pHRed genetic biosensor to correct for variations in pH if necessary. Heterogenous populations of transduced renal cells were enriched by flow cytometry before monoclonal cellular populations were isolated by cell culture methods. A single clonal population of co-transduced cells expressing both Perceval HR and pHRed was selected to be HK2-Vi. This established cell line can now serve as a tool for in vitro toxicology testing and the methods described herein serve as a model for developing designer cell lines derived from other organs.

Assessing Toxicity with Human Cell-Based In Vitro Methods

In toxicology, there is a strong push towards replacing animal experiments with alternative methods, which include cell-based in vitro methods for the assessment of adverse health effects in humans. High-throughput methods are of central interest due to the large and steadily growing numbers of compounds that require assessment. Tremendous progress has been made during the last decade in developing and applying such methods. Innovative technologies for addressing complex biological interactions include induced pluripotent stem cell- and organoid-based approaches, organotypic coculture systems, and microfluidic 'multiorgan' chips. Combining in vitro methods with bioinformatics and in silico modeling generates new powerful tools for toxicity assessment, and the rapid progress in the field is expected to continue.

Machine Learning for Identifying Medication-Associated Acute Kidney Injury

One of the prominent problems in clinical medicine is medication-induced acute kidney injury (AKI). Avoiding this problem can prevent patient harm and reduce healthcare expenditures. Several researches have been conducted to identify AKI-associated medications using statistical, data mining, and machine learning techniques. However, these studies are limited to assessing the impact of known nephrotoxic medications and do not comprehensively explore the relationship between medication combinations and AKI. In this paper, we present a population-based retrospective cohort study that employs automated data analysis techniques to identify medications and medication combinations that are associated with a higher risk of AKI. By integrating multivariable logistic regression, frequent itemset mining, and stratified analysis, this study is designed to explore the complex relationships between medications and AKI in such a way that has never been attempted before. Through an analysis of prescription records of one million older patients stored in the healthcare administrative dataset at ICES (an independent, non-profit, world-leading research organization that uses population-based health and social data to produce knowledge on a broad range of healthcare issues), we identified 55 AKI-associated medications among 595 distinct medications and 78 AKI-associated medication combinations among 7748 frequent medication combinations. In addition, through a stratified analysis, we identified 37 cases where a particular medication was associated with increasing the risk of AKI when used with another medication. We have shown that our results are consistent with previous studies through consultation with a nephrologist and an electronic literature search. This research demonstrates how automated analysis techniques can be used to accomplish data-driven tasks using massive clinical datasets.

Generating Kidney from Stem Cells

Human kidney tissue can now be generated via the directed differentiation of human pluripotent stem cells. This advance is anticipated to facilitate the modeling of human kidney diseases, provide platforms for nephrotoxicity screening, enable cellular therapy, and potentially generate tissue for renal replacement. All such applications will rely upon the accuracy and reliability of the model and the capacity for stem cell-derived kidney tissue to recapitulate both normal and diseased states. In this review, we discuss the models available, how well they recapitulate the human kidney, and how far we are from application of these cells for use in cellular therapies.

Development of a Novel Hanging Drop Platform for Engineering Controllable 3D Microenvironments

Conventional biomedical research is mostly performed by utilizing a two-dimensional monolayer culture, which fails to recapitulate the three-dimensional (3D) organization and microenvironment of native tissues. To overcome this limitation, several methods are developed to fabricate microtissues with the desired 3D microenvironment. However, they tend to be time-consuming, labor-intensive, or costly, thus hindering the application of 3D microtissues as models in a wide variety of research fields. In the present study, we have developed a pressure-assisted network for droplet accumulation (PANDA) system, an easy-to-use chip that comprises a multichannel fluidic system and a hanging drop cell culture module for uniform 3D microtissue formation. This system can control the desired artificial niches for modulating the fate of the stem cells to form the different sizes of microtissue by adjusting the seeding density. Furthermore, a large number of highly consistent 3D glomerulus-like heterogeneous microtissues that are composed of kidney glomerular podocytes and mesenchymal stem cells have been formed successfully. These data suggest that the developed PANDA system can be employed as a rapid and economical platform to fabricate microtissues with tunable 3D microenvironment and cellular heterogeneity, thus can be employed as tissue-mimicking models in various biomedical research.

Comparison of the predictive outcomes for anti-tuberculosis drug-induced hepatotoxicity by different machine learning techniques

BACKGROUND

The study compared the predictive outcomes of artificial neural network, support vector machine and random forest on the occurrence of anti-tuberculosis drug-induced hepatotoxicity.

METHODS

The clinical and genomic data of patients treated with anti-tuberculosis drugs at Taipei Medical University-Wanfang Hospital were used as training sets, and those at Taipei Medical University-Shuang Ho Hospital served as test sets. Features were selected through a univariate risk factor analysis and literature evaluation. The accuracy, sensitivity, specificity, and the area under the receiver operating characteristic curve were calculated to compare the traditional, genomic, and combined models of the three techniques.

RESULTS

Nine models were created with 7 clinical factors and 4 genotypes. Artificial neural network with clinical and genomic factors exhibited the best performance, with an accuracy of 88.67%, a sensitivity of 80%, and a specificity of 90.4% for the test set. The area under the receiver operating characteristic curve of this best model reached 0.894 for training set and 0.898 for test set, which was significantly better than 0.801 for training set and 0.728 for test set by support vector machine and 0.724 for training set and 0.718 for test set by random forest.

CONCLUSIONS

Artificial neural network with clinical and genomic data can become a clinical useful tool in predicting anti-tuberculosis drug-induced hepatotoxicity. The machine learning technique can be an innovation to predict and prevent adverse drug reaction.

Induced pluripotent stem cells for therapy personalization in pediatric patients: Focus on drug-induced adverse events

Adverse drug reactions (ADRs) are major clinical problems, particularly in special populations such as pediatric patients. Indeed, ADRs may be caused by a plethora of different drugs leading, in some cases, to hospitalization, disability or even death. In addition, pediatric patients may respond differently to drugs with respect to adults and may be prone to developing different kinds of ADRs, leading, in some cases, to more severe consequences. To improve the comprehension, and thus the prevention, of ADRs, the set-up of sensitive and personalized assays is urgently needed. Important progress is represented by the possibility of setting up groundbreaking patient-specific assays. This goal has been powerfully achieved using induced pluripotent stem cells (iPSCs). Due to their genetic and physiological species-specific differences and their ability to be differentiated ideally into all tissues of the human body, this model may be accurate in predicting drug toxicity, especially when this toxicity is related to individual genetic differences. This review is an up-to-date summary of the employment of iPSCs as a model to study ADRs, with particular attention to drugs used in the pediatric field. We especially focused on the intestinal, hepatic, pancreatic, renal, cardiac, and neuronal levels, also discussing progress in organoids creation. The latter are three-dimensional in vitro culture systems derived from pluripotent or adult stem cells simulating the architecture and functionality of native organs such as the intestine, liver, pancreas, kidney, heart, and brain. Based on the existing knowledge, these models are powerful and promising tools in multiple clinical applications including toxicity screening, disease modeling, personalized and regenerative medicine.

Emerging In Vitro Systems to Screen and Predict Drug-Induced Kidney Toxicity

Drug attrition related to kidney toxicity remains a challenge in drug discovery and development. In vitro models established over the past 2 decades to supplement in vivo studies have improved the throughput capacity of toxicity evaluation, but usually suffer from low predictive value. To achieve a paradigm shift in the prediction of drug-induced kidney toxicity, two aspects are fundamental: increased physiological relevance of the kidney model, and use of appropriate toxicity end points. Recent studies have suggested that increasing the physiological relevance of kidney models can improve their sensitivity to drug-induced damage. Here, we discuss how advanced culture models, including modified cell lines, induced pluripotent stem cells, kidney organoid cultures, and microfluidic devices enhance in vivo similarity. To this end, culture models aim to increase the proximal tubule epithelial phenotype, reconstitute multiple tissue compartments and extracellular matrix, allow exposure to fluid shear stress, and enable interaction between multiple cell types. Applying computation-aided end points and novel biomarkers to advanced culture models will further improve sensitivity and clinical relevance of in vitro drug-induced toxicity prediction. Implemented at the right stage of drug discovery and development and coupled to high-content evaluation techniques, these models have the potential to reduce attrition and aid the selection of candidate drugs with an appropriate safety profile.

Kidney-based in vitro models for drug-induced toxicity testing

The kidney is frequently involved in adverse effects caused by exposure to foreign compounds, including drugs. An early prediction of those effects is crucial for allowing novel, safe drugs entering the market. Yet, in current pharmacotherapy, drug-induced nephrotoxicity accounts for up to 25% of the reported serious adverse effects, of which one-third is attributed to antimicrobials use. Adverse drug effects can be due to direct toxicity, for instance as a result of kidney-specific determinants, or indirectly by, e.g., vascular effects or crystals deposition. Currently used in vitro assays do not adequately predict in vivo observed effects, predominantly due to an inadequate preservation of the organs' microenvironment in the models applied. The kidney is highly complex, composed of a filter unit and a tubular segment, together containing over 20 different cell types. The tubular epithelium is highly polarized, and the maintenance of this polarity is critical for optimal functioning and response to environmental signals. Cell polarity is dependent on communication between cells, which includes paracrine and autocrine signals, as well as biomechanic and chemotactic processes. These processes all influence kidney cell proliferation, migration, and differentiation. For drug disposition studies, this microenvironment is essential for prediction of toxic responses. This review provides an overview of drug-induced injuries to the kidney, details on relevant and translational biomarkers, and advances in 3D cultures of human renal cells, including organoids and kidney-on-a-chip platforms.

Vascular Cells and Tissue Constructs Derived from Human Pluripotent Stem Cells for Toxicological Screening

The ability of human stem cells to generate somatic cell lineages makes them ideal candidates for use in toxicological testing and eventually, preclinical drug development. Such resources would support an evolution away from human primary cells or research animal models, which suffer from variability and poor predictability, toward off-the-shelf assays of chemical toxicity and drug efficacy using human cells and tissues. To this end, we generated vascular cell populations (smooth muscle cells and endothelial cells) from human pluripotent stem cells (hPSCs), arranged them into 3D co-cultures within supportive gel matrices, and directed their propensity for self-organization resembling microvasculature. The resulting vascular cell populations and co-cultured constructs were then arrayed in high throughput and used for screening a library of environmental and clinical chemical agents for immunological and toxicological responses. The screen effectively stratified the chemicals into various levels of toxicity, with both cell type-specific and co-culture-dependent responses observed. Thus, hPSC-derived vascular cells and constructs could be progressed further toward use in toxicant and drug screening.

Current State of Renal Regenerative Therapies

The worldwide increase in the number of patients with end-stage renal disease leads to a growing waiting list for kidney transplantation resulting from the scarcity of kidney donors. Therefore, alternative treatment options for patients with end-stage renal disease are being sought. In vitro differentiation of stem cells into renal tissue is a promising approach to repair nonfunctional kidney tissue. Impressive headway has been made in the use of stem cells with the use of adult renal progenitor cells, embryonic stem cells, and induced pluripotent stem cells for the development toward primitive kidney structures. Currently, efforts are directed at improving long-term maintenance and stability of the cells. This review aims to provide a comprehensive overview of the cell sources used for the generation of kidney cells and strategies used for transplantation in in vivo models. Furthermore, it provides a perspective on stability and safety during future clinical application of in vitro generated kidney cells.

Road Map for Development of Stem Cell-Based Alternative Test Methods

Much progress has been made in establishing strategies for differentiation of induced human pluripotent stem cells (hiPSCs). However, differentiated hiPSCs are not yet routinely used for prediction of toxicity. Here, limiting factors are summarised and possibilities for improvement are discussed, with a focus on hepatocytes, cardiomyocytes, tubular epithelial cells, and developmental toxicity. Moreover, we make recommendations for further fine-tuning of differentiation protocols for hiPSCs to hepatocyte-like cells by comparing individual steps of currently available protocols to the mechanisms occurring during embryonic development. A road map is proposed to facilitate test system development, including a description of the most useful performance metrics.

Parallel generation of easily selectable multiple nephronal cell types from human pluripotent stem cells

Human pluripotent stem cells (hPSCs) provide a source for the generation of defined kidney cells and renal organoids applicable in regenerative medicine, disease modeling, and drug screening. These applications require the provision of hPSC-derived renal cells by reproducible, scalable, and efficient methods. We established a chemically defined protocol by application of Activin A, BMP4, and Retinoic acid followed by GDNF, which steered hPSCs to the renal lineage and resulted in populations of SIX2+/CITED1+ metanephric mesenchyme- (MM) and of HOXB7+/GRHL2+ ureteric bud (UB)-like cells already by 6 days. Transcriptome analysis corroborated that the PSC-derived cell types at day 8 resemble their renal vesicle and ureteric epithelial counterpart in vivo, forming tubular and glomerular renal cells 6 days later. We demonstrate that starting from hPSCs, our in vitro protocol generates a pool of nephrogenic progenitors at the renal vesicle stage, which can be further directed into specialized nephronal cell types including mesangial-, proximal tubular-, distal tubular, collecting duct epithelial cells, and podocyte precursors after 14 days. This simple and rapid method to produce renal cells from a common precursor pool in 2D culture provides the basis for scaled-up production of tailored renal cell types, which are applicable for drug testing or cell-based regenerative therapies.

Application of human pluripotent stem cells and pluripotent stem cell-derived cellular models for assessing drug toxicity

Introduction: Human pluripotent stem cells (hPSCs) are capable of differentiating into all types of cells in the body and so provide suitable toxicology screening systems even for hard-to-obtain human tissues. Since hPSCs can also be generated from differentiated cells and current gene editing technologies allow targeted genome modifications, hPSCs can be applied for drug toxicity screening both in normal and disease-specific models. Targeted hPSC differentiation is still a challenge but cardiac, neuronal or liver cells, and complex cellular models are already available for practical applications. Areas covered: The authors review new gene-editing and cell-biology technologies to generate sensitive toxicity screening systems based on hPSCs. Then the authors present the use of undifferentiated hPSCs for examining embryonic toxicity and discuss drug screening possibilities in hPSC-derived models. The authors focus on the application of human cardiomyocytes, hepatocytes, and neural cultures in toxicity testing, and discuss the recent possibilities for drug screening in a 'body-on-a-chip' model system. Expert opinion: hPSCs and their genetically engineered derivatives provide new possibilities to investigate drug toxicity in human tissues. The key issues in this regard are still the selection and generation of proper model systems, and the interpretation of the results in understanding in vivo drug effects.

Bioengineering Organs for Blood Detoxification

For patients with severe kidney or liver failure the best solution is currently organ transplantation. However, not all patients are eligible for transplantation and due to limited organ availability, most patients are currently treated with therapies using artificial kidney and artificial liver devices. These therapies, despite their relative success in preserving the patients' life, have important limitations since they can only replace part of the natural kidney or liver functions. As blood detoxification (and other functions) in these highly perfused organs is achieved by specialized cells, it seems relevant to review the approaches leading to bioengineered organs fulfilling most of the native organ functions. There, the culture of cells of specific phenotypes on adapted scaffolds that can be perfused takes place. In this review paper, first the functions of kidney and liver organs are briefly described. Then artificial kidney/liver devices, bioartificial kidney devices, and bioartificial liver devices are focused on, as well as biohybrid constructs obtained by decellularization and recellularization of animal organs. For all organs, a thorough overview of the literature is given and the perspectives for their application in the clinic are discussed.

Comment on Sjögren et al. (2018) A novel multi-parametric high-content screening assay in ciPTEC-OAT1 to predict drug-induced nephrotoxicity in drug discovery. Arch Toxicol 92(10):3175-3190

Rapid progress is made in the development of high-content screening assays for the prediction of nephrotoxicity. The findings from different laboratories are consistent with respect to endpoints and concentration ranges screened. Discrepancies regarding compound annotation and the predictive performance analysis are discussed.

Human iPS Cell-Derived Patient Tissues and 3D Cell Culture Part 1: Target Identification and Lead Optimization

Human-induced pluripotent stem cells (HiPSCs), and new technologies to culture them into functional cell types and tissues, are now aiding drug discovery. Patient-derived HiPSCs can provide disease models that are more clinically relevant and so more predictive than the currently available animal-derived or tumor cell-derived cells. These cells, consequently, exhibit disease phenotypes close to the human pathology, particularly when cultured under conditions that allow them to recapitulate the tissue architecture in three-dimensional (3D) systems. A key feature of HiPSCs is that they can be cultured under conditions that favor formation of multicellular spheroids or organoids. By culturing and differentiating in systems mimicking the human tissue in vivo, the HiPSC microenvironment further reflects patient in vivo physiology, pathophysiology, and ultimately pharmacological responsiveness. We assess the rationale for using HiPSCs in several phases of preclinical drug discovery, specifically in disease modeling, target identification, and lead optimization. We also discuss the growing use of HiPSCs in compound lead optimization, particularly in profiling compounds for their potential metabolic liability and off-target toxicities. Collectively, we contend that both approaches, HiPSCs and 3D cell culture, when used in concert, have exciting potential for the development of novel medicines.