Transepithelial resistance and inulin permeability as endpoints in in vitro nephrotoxicity testing

The protective effect and molecular mechanism of glycyrrhizic acid glycosides against Tripterygium glycosides induced nephrotoxicity based on the RhoA/ROCK1 signalling pathway

ETHNOPHARMACOLOGICAL RELEVANCE

Tripterygium glycosides (TG), which are extracted from the traditional Chinese medicine, Tripterygium wilfordii Hook F. (TwHF), has promising applications in the treatment of renal diseases; however, since its active components exerts bidirectional kidney toxicity, its clinical application is severely restricted.

AIM OF THE STUDY

Recent investigations have demonstrated definite toxicity-reducing effects from glycyrrhizic acid glycosides (GA) when combined with TG; however, the mechanism remains unclear. To our knowledge, this is the first study to investigate the specific molecular mechanism by which GA alleviates TG-induced renal toxicity from the perspective of tight junctions.

MATERIALS AND METHODS

Dynamic analyses, which investigated the changes in kidney toxicity biomarkers for different combinations and concentrations of TG and GA, were conducted for three weeks on SD rats and renal tissue structural changes were examined after three weeks of administration. Additionally, the transcription and translation levels of the relevant tight junctions and RhoA/ROCK1/MLC signalling proteins were analysed in HK-2 cells.

RESULTS

Our study showed that TG can cause transient tubulotoxicity at certain doses, and that the combined application of GA and TG can repair tight junction structures by regulating the key factors in the RhoA/ROCK1/MLC signalling pathway, thus reducing TG-induced nephrotoxicity.

CONCLUSIONS

Overall, this study provides a new strategy to reduce TG-induced toxicity by protecting renal tight junctions.

Progress and Challenges in the Use of a Liver-on-a-Chip for Hepatotropic Infectious Diseases

The liver is a target organ of life-threatening pathogens and prominently contributes to the variation in drug responses and drug-induced liver injury among patients. Currently available drugs significantly decrease the morbidity and mortality of liver-dwelling pathogens worldwide; however, emerging clinical evidence reveals the importance of host factors in the design of safe and effective therapies for individuals, known as personalized medicine. Given the primary adherence of cells in conventional two-dimensional culture, the use of these one-size-fit-to-all models in preclinical drug development can lead to substantial failures in assessing therapeutic safety and efficacy. Advances in stem cell biology, bioengineering and material sciences allow us to develop a more physiologically relevant model that is capable of recapitulating the human liver. This report reviews the current use of liver-on-a-chip models of hepatotropic infectious diseases in the context of precision medicine including hepatitis virus and malaria parasites, assesses patient-specific responses to antiviral drugs, and designs personalized therapeutic treatments to address the need for a personalized liver-like model. Second, most organs-on-chips lack a monitoring system for cell functions in real time; thus, the review discusses recent advances and challenges in combining liver-on-a-chip technology with biosensors for assessing hepatocyte viability and functions. Prospectively, the biosensor-integrated liver-on-a-chip device would provide novel biological insights that could accelerate the development of novel therapeutic compounds.

Protein Substrate Alters Cell Physiology in Primary Culture of Vocal Fold Epithelial Cells

The basement membrane interacts directly with the vocal fold epithelium. Signaling between the basement membrane and the epithelium modulates gene regulation, differentiation, and proliferation. The purpose of this study was to identify an appropriate simple single-protein substrate for growth of rabbit vocal fold epithelial cells. Vocal folds from 3 New Zealand white rabbits (Oryctolagus cuniculus) were treated to isolate epithelial cells, and cells were seeded onto cell culture inserts coated with collagen I, collagen IV, laminin, or fibronectin. Transepithelial electrical resistance (TEER) was measured, and phase contrast microscopy, PanCK, CK14, and E-cadherin immunofluorescence were utilized to assess for epithelial cell-type characteristics. Further investigation via immunofluorescence labeling was conducted to assess proliferation (Ki67) and differentiation (Vimentin). There was a significant main effect of substrate on TEER, with collagen IV eliciting the highest, and laminin the lowest resistance. Assessment of relative TEER across cell lines identified a larger range of TEER in collagen I and laminin. Phase contrast imaging identified altered morphology in the laminin condition, but cell layer depth did not appear to be related to TEER, differentiation, or morphology. Ki67 staining additionally showed no significant difference in proliferation. All conditions had confluent epithelial cells and dispersed mesenchymal cells, with increased mesenchymal cell numbers over time; however, a higher proportion of mesenchymal cells was observed in the laminin condition. The results suggest collagen IV is a preferable basement membrane substrate for in vitro vocal fold epithelial primary cell culture, providing consistent TEER and characteristic cell morphology, and that laminin is an unsuitable substrate for vocal fold epithelial cells and may promote mesenchymal cell proliferation.

Comparison of Various Cell Lines and Three-Dimensional Mucociliary Tissue Model Systems to Estimate Drug Permeability Using an In Vitro Transport Study to Predict Nasal Drug Absorption in Rats

Recently, various types of cultured cells have been used to research the mechanisms of transport and metabolism of drugs. Although many studies using cultured cell systems have been published, a comparison of different cultured cell systems has never been reported. In this study, Caco-2, Calu-3, Madin–Darby canine kidney (MDCK), EpiAirway and MucilAir were used as popular in vitro cell culture systems, and the permeability of model compounds across these cell systems was evaluated to compare barrier characteristics and to clarify their usefulness as an estimation system for nasal drug absorption in rats. MDCK unexpectedly showed the best correlation (r = 0.949) with the fractional absorption (F_n) in rats. Secondly, a high correlation was observed in Calu-3 (r = 0.898). Also, Caco-2 (r = 0.787) and MucilAir (r = 0.750) showed a relatively good correlation with F_n. The correlation between F_n and permeability to EpiAirway was the poorest (r = 0.550). Because EpiAirway forms leakier tight junctions than other cell culture systems, the paracellular permeability was likely overestimated with this system. On the other hand, because MDCK formed such tight cellular junctions that compounds of paracellular model were less likely permeated, the paracellular permeability could be underestimated. Calu-3, Caco-2 and MucilAir form suitable cellular junctions and barriers, indicating that those cell systems enable the precise estimation of nasal drug absorption.

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.

Comparison of transepithelial resistance measurement techniques: Chopsticks vs. Endohm

Background

TER measurements across confluent cellular monolayers provide a useful indication of TJ strength between epithelial and endothelial cells in culture. Having a reliable and accurate method of measuring cell-to-cell adhesion is critical to studies in pathophysiology and cancer metastasis. However, the use of different technical approaches to measure TER has reportedly yielded inconsistent measurements within the same cell lines.

Methods

In the current study, we compared the peak TER values for the MDCK (canine kidney) and MCF-7 (human breast cancer) epithelial cell lines using two common approaches (Chopstick and Endohm) and two types of polymer inserts (PC and PET).

Results

Both cell lines demonstrated a statistically significant difference in the peak TERs obtained using the two different approaches. Further, the MDCK (but not the MCF-7) cells demonstrated a statistically significant difference between the peak TERs when using the same approach but different inserts.

Conclusion

Our study indicates the importance of using a single approach when seeking to measure and compare the TER values of cultured cell lines.

Drug-induced nephrotoxicity: clinical impact and preclinical in vitro models

The kidney is a major target for drug-induced toxicity. Drug-induced nephrotoxicity remains a major problem in the clinical setting, where the use of nephrotoxic drugs is often unavoidable. This leads frequently to acute kidney injury, and current problems are discussed. One strategy to avoid such problems would be the development of drugs with decreased nephrotoxic potential. However, the prediction of nephrotoxicity during preclinical drug development is difficult and nephrotoxicity is typically detected only late. Also, the nephrotoxic potential of newly approved drugs is often underestimated. Regulatory approved or validated in vitro models for the prediction of nephrotoxicity are currently not available. Here, we will review current approaches on the development of such models. This includes a discussion of three-dimensional and microfluidic models and recently developed stem cell based approaches. Most in vitro models have been tested with a limited number of compounds and are of unclear predictivity. However, some studies have tested larger numbers of compounds and the predictivity of the respective in vitro model had been determined. The results showed that high predictivity can be obtained by using primary or stem cell derived human renal cells in combination with appropriate end points.

A screen for apoptotic synergism between clinical relevant nephrotoxicant and the cytokine TNF-α

Nephrotoxicity remains one of the main reasons for post-market drug withdrawal. Tumour necrosis factor α (TNF-α) secretion has been shown to underlie the nephrotoxicity induced by some of these drugs. Yet, there is currently no reliable and sensitive in vitro assay available to screen for nephrotoxicants of which toxicity largely depends on TNF-α secretion. Therefore, we developed and applied a sensitive fluorescence-based in vitro assay for TNF-α-mediated nephrotoxicity screening using mouse immortalized proximal tubular epithelial cells (IM-PTECs). Our assay allows rapid evaluation of TNF-α-mediated toxicant-induced apoptosis and necrosis using fixed endpoint and live cell measurements. To evaluate our assay, sixteen nephrotoxicants and two control non-nephrotoxicants were used. Out of the sixteen nephrotoxicants, eight induced cell death, of which five induced apoptosis as well as necrosis. Moreover, TNF-α significantly enhanced apoptotic cell death induced by cisplatin, cyclosporine A, tacrolimus and azidothymidine. These nephrotoxicants are known to induce inflammation in vivo which has been linked to an enhancement of nephrotoxicity for cisplatin, cyclosporine A and tacrolimus, confirming the functionality of our assay. Overall, our assay allows rapid and sensitive measurement of apoptosis and necrosis induced by a combination of nephrotoxicants and inflammatory components such as TNF-α and can be used as an alternative assay for nephrotoxicity prediction in vitro.

Triptolide alters barrier function in renal proximal tubular cells in rats

Alteration of the tight junction complex in renal epithelial cells can affect renal barrier function and perturb normal kidney homeostasis. The objective of the present study was to determine whether triptolide could affect tight junctions in the proximal tubule epithelial cells both in vivo and in vitro. Wistar rats were gavaged with triptolide at 0, 100, 200 or 400 μg/kg/day for 28 days. Pathologic examination of the kidney showed that triptolide primarily affected the proximal tubules. The nephrotoxicity of triptolide is morphologically characterized by the detachment of the proximal tubular epithelial cells from each other. Immunohistochemical analysis showed that there was marked alteration in the localization of Zonula Occludens 1 protein (ZO-1) in the proximal tubule epithelium. Additionally, the uptake of FITC-dextran, a marker of fluid phase endocytosis in the proximal tubule, was considerably lower in triptolide-treated animals than in normal rats. Supported by these results, we detected significant increases in blood urea nitrogen (BUN) but not of creatinine (Cr) in rats treated with triptolide, indicating damage to the proximal tubules. Furthermore, triptolide treatment caused an alteration of the tight junction complex, resulting in changes in paracellular permeability in NRK-52E cells in vitro. Taken together, these results suggest that triptolide induced renal toxicity in rats and that the mechanism of toxicity was related to the disruption of cell-cell junctions and alterations of the paracellular permeability in the proximal tubule.

Selection of test methods to be included in a testing strategy to predict acute oral toxicity: an approach based on statistical analysis of data collected in phase 1 of the ACuteTox project

More than 50 different in vitro and in silico methods assessing specific organ- and system-toxicity, such as haemato-, neuro-, nephro- and hepatotoxicity, as well as intestinal absorption, distribution and metabolism, have been used in the first phase of the ACuteTox project to test a common set of 57 chemicals. This paper describes the methods used for statistical evaluation of concentration-response data collected for each of the endpoint assays, and for the development of a testing strategy applicable for acute toxicity classification of chemicals based on the achieved results of the concentration-response analysis. A final list of in vitro test methods considered to be promising candidates for building blocks of the testing strategy is presented. Only these selected test methods were further investigated in the prevalidation phase of the project. The test methods were chosen according to their reproducibility and reliability and most importantly, according to their potential to classify chemicals into the official acute oral toxicity categories of the EU Classification, Labelling and Packaging (CLP) Regulation. The potential of the test methods to correctly classify the chemicals was assessed by Classification and Regression Trees (CART) analysis.

Tissue-Specific stem cell differentiation in an in vitro airway model

The respiratory tract is the primary site of exposure to airborne compounds, with the bronchial epithelium providing one of the first lines of defence. A growing need exists for an accurate in vitro model of the bronchial epithelium. Here, normal human bronchial epithelial (NHBE) cells cultured at an air/liquid interface create a fully differentiated, in-vivo-like model of the human bronchial epithelium. Developmental characterisation includes (i) trans-epithelial electrical resistance, (ii) morphology and (iii) bronchial cell specific stains/markers. It is concluded that the basal/progenitor cells create a pseudo-stratified, mucociliary NHBE model containing basal, serous, Clara, goblet and ciliated cells, reflective of the normal human bronchial epithelium (days 24-33 ALI culture).

An in vitro approach to assess the toxicity of inhaled tobacco smoke components: nicotine, cadmium, formaldehyde and urethane

One of the first lines of defence to inhaled toxins is the barrier formed by the tracheobronchial epithelium, making this the ideal region for studying the toxicity of inhaled substances. This study utilises a highly differentiated, three-dimensional, in vitro model of human upper respiratory tract epithelium (EpiAirway-100) to measure the acute toxicological responses to well-characterised tobacco smoke components. To determine the suitability of this model for screening inhaled toxicants, the EpiAirway tissue model (ETM) was treated apically with tobacco smoke components (nicotine, formaldehyde, cadmium, urethane) which are known to induce a variety of toxic effects (e.g. cytotoxic, thrombogenic, carcinogenic). A range of concentrations were used to model different mechanisms and severity of toxicity which were then compared to known in vivo responses. Similar trends in stress response occurred, with distinct alterations to the tissue in response to all four toxins. At high concentrations, cell viability decreased and tight junctions were degraded, but at sub-toxic concentrations epithelial resistance (indicating tissue integrity) increased 20-60% from control. This peak in resistance coincided with an increase in secreted protein levels, elevated cytokine release and goblet cell hyperplasia and hypertrophy. In conclusion, acute exposure to tobacco smoke components induces measurable toxic responses within human respiratory epithelium. Sub-toxic concentrations appear to illicit a protective response by increasing mucus secretion and mediating immune responses via cytokine release. These responses are comparable to human in vivo responses, indicating potential for the ETM as a tool for screening the toxicity of inhaled compounds.

Epithelial barrier characteristics and expression of cell adhesion molecules in proximal tubule-derived cell lines commonly used for in vitro toxicity studies

Recent studies indicate that the actions of several nephrotoxic substances involve alterations in the function of cell adhesion molecules and changes in the paracellular permeability of the proximal tubule. In light of these findings, there is a need for appropriate in vitro model systems to study these phenomenae in greater detail. In the present study, the transepithelial resistance (TER), paracellular permeability of 14C-mannitol and immunofluorescent labeling of cell adhesion molecules (E-cadherin, N-cadherin, ZO-1, occludin, and claudins-2 and -7) were evaluated in several proximal tubule-derived cell lines that have been commonly used as model systems for in vitro toxicity studies. The cell lines studied included: LLC-PK1, OK, NRK-52E and HK-2, along with commercially available primary cultures of human renal proximal tubule epithelial cells (HRPTE). LLC-PK1 cells developed the highest TER followed by the OK cells and NRK-52E cells. The other cell lines failed to develop a TER even after 2 weeks in culture. There was a direct correlation between TER and ability to restrict paracellular movement of 14C-mannitol. Labeling studies showed that the LLC-PK1 and NRK-52E cells expressed high levels of E-cadherin while the HRPTE cells expressed lower levels. OK cells expressed moderate levels of N-cadherin while LLC-PK1 and NRK-52E cells expressed lower levels in isolated patches of cells. All cell lines expressed moderate-high levels of ZO-1. LLC-PK1 also expressed the tight-junction proteins occludin and claudin-7; OK cells also expressed moderate levels of occludin. All other cell lines had weak claudin-7 and occludin labeling. None of the cell lines expressed claudin-2. These results show that the LLC-PK1, OK and NRK-52E cell lines exhibit characteristics that most closely resembled those of the proximal tubule in vivo, and they indicate that these cell lines would be appropriate models for studying the effects of toxicants on cell-cell junctions and cell adhesion molecules.