MatTek scientists will be attending and presenting posters at the Society of Toxicology Annual Meeting in Salt Lake City, Utah. Our scientists are presenting 10 NEW POSTERS with new research, new applications, and two new tissue models. Read more to see what we’ve been working on and request copies of our posters. We can’t wait to see you at SOT 2024!
Meet our team at Booth #1527
Reconstruction of Novel 3D Models of the Three Parts of the Human Small Intestine for Drug Toxicity (P133)
Zachary Stevens1, Jonathan Cheong2, Mitchell Klausner1, Alex Armento1, Seyoum Ayehunie1
1MatTek Life Sciences, Ashland, MA and 2Genentech, San Francisco, CA
Session: Safety Assessment: Pharmaceutical-Drug Development II | March 11, 2024 | 2:15 – 4:15pm
Background: Gastrointestinal (GI) toxicity is a common adverse event that limits pharmaceutical development across diverse therapy areas. The mechanisms of drug-induced gastrointestinal (GI) toxicity are often poorly understood, in part due to the lack of physiologically relevant in vitro models that recapitulate the role of the three parts/segments of the small intestine. Traditional in vitro cell culture approaches utilize immortalized human colorectal Caco-2 cells cultured for about 21 days for assessing ADME properties. Despite being in use for more than 5 decades, this approach has limitations in that it is cell-based (colon-derived), lacks major drug transporters and drug metabolizing enzymes, does not have fully polarized structural features, and is not predictive of GI toxicity. To mimic the physiology and functionality of the human gut, we previously developed the small intestinal model utilizing cells derived from the ileum. However, the pharma industry has a need for in vitro 3D models representing the other parts of the small intestine (duodenum, and jejunum). Having models of the different parts of the small intestine will allow identification of the primary site of absorption for each drug candidate, to predict efficacy or identify compounds with lower risk and/or aid to develop dosing schedules that mitigate risks. Methods: State-of-the-art tissue culture and tissue engineering methods were used to reconstruct 3D GI models representing the duodenum, jejunum, and ileum of the human small intestine. The 3D tissue models use primary cells and were characterized morphologically by histology and functionally by performing drug permeation studies. Gene expression levels of drug transporters and drug metabolizing enzymes were quantified by qPCR & protein quantification analyses were performed by LC/MS. The activities of the transporters were evaluated using Vinblastine as substrate for P-gP and MRP. Results: The 3D tissues developed structural features resembling villi, with barrier function. The mRNA data from the 3D tissue models agreed well with published data showing high expressions of SLCO2B1, SLC16A1 and SLC15A1 by the three segments of the small intestine. Some of the genes transcribed to proteins include: 1) drug transporters such as P-glycoprotein (PgP, MDR1), multi-drug resistant protein (MRP-3), breast cancer resistant protein (BCRP, ABCG2). Further analysis of the mRNA data reveals P-gP, MRP-2, and ABCG2 expressions in the ileum model were lower compared to the duodenum and Jejunum models. On the other hand, SLC15A1 Peptide transporter 1 (PepT-1) expression was greater in the ileum model compared to the others (Ileum > jejunum > duodenum). Expression of drug metabolizing enzymes such as CYP3A4, CYP2C9, UDP glucuronosyltransferase 1 family, polypeptide A1 (UGT1A1), and Carboxylesterase 1 (CES-1) was maintained in each segment at a comparative level. However, CYP2C9 expression was more pronounced in the jejunum and the rank order was jejunum > duodenum > ileum. The efflux ratio for Vinblastine decreased when the MK571 (MRP inhibitor) and Elacridar (P-gP inhibitor) were added with Vinblastine. Treatment of the tissues with Raloxifene (an anti-osteoporotic drug, CYP3A4 substrate) showed formation of raloxifene-6-glucuronide demonstrating the presence of the phase II glucuronidase enzymes in the tissue models. When Zafirlukast, an inhibitor of the glucuronidase enzymes, was added with Raloxifene, raloxifene-6-glucuronide formation was reduced. Conclusion: These results suggest that the reconstructed tissues from the three segments of the small intestine will serve as useful tools to predict both investigational and traditional GI drug safety and absorption in the GI tract. In addition, use of these models will reduce animal use and improve the pre-clinical drug development process.
Identification of Compounds with Weak Skin Irritation Potential Using in vitro Methods Based on a 3D Reconstructed Human Epidermis Model (P304)
Jana Halajova1, Lenka Hudecova1, Jan Markus1, Christian Pellevoisin2, Marek Puskar1, Mitchell Klausner2, Silvia Letasiova1
1MatTek Europe, Bratislava, Slovakia, 2MatTek Life Sciences, Ashland, MA, USA
Session: Skin | March 11, 2024 | 9:15 – 11:15
Background: The replacement of animal testing with in vitro methods has reduced the incidences of false positive results, reduced the number of animal tests, and increased the efficiency in terms of time and cost. The replacement for alternative methods is supported by the concept of the “three Rs,” which includes replacement, reduction and refinement of individual original methods carried out on animals. In the chemical, pharmaceutical and cosmetic industries, in vitro models have replaced many methods used in safety and efficiency testing, including tests for skin sensitization, skin corrosion, skin absorption and skin irritation. Methods: This work was focused on the utilization of an in vitro reconstituted human skin model to identify substances with a weak irritation potential. The in vitro method for determining skin irritation according to OECD test guideline 439 and the method for determining skin irritation for extracts from medical devices according to ISO standard – ISO 10993-23:2021, were used to detect the irritation of 15 test articles (TAs) in 5 concentrations (0.1%, 0.5%, 1%, 5%, 10%) in non-polar solvent (sesame oil) and in polar solvent (saline). If no effects were observed at these concentrations, TAs were also tested neat. The concentration that reduced tissue viability to 50% (EC-50) was calculated. ELISA assays were also used to measure the release of pro-inflammatory cytokines, specifically interleukins IL-1α, IL-6 and IL-8, and the release of IL-18, indicating the potential for skin sensitization. Results: 5 of the TAs (allyl heptanoate, heptyl salicylate, linalyl acetate, methyl laurate, hexyl salicylate) had viability comparable with negative control tissues in all concentrations in both solvents as well as in undiluted form. On the other hand, 7 TAs (10-undecenoic acid, lactic acid, 2-ethoxyehtyl methacrylate, 1-decanol, methyl methacrylate, 2-bromobutane, 50% sodium carbonate) did not cause the irritation effect in all tested concentrations in both solvents, but irritation effect was observed using the TAs. The remaining 3 TAs (heptanoic acid, SDS and 10% sodium hypochlorite) decreased tissue viability below 50%, so the EC-50s could be calculated. When using the skin irritation test for medical device extracts according to ISO 10993-23:2021 we found that 3 compounds (heptyl butyrate, hexyl salicylate, methyl laurate) did not decrease viability at any concentration in both solvents and the viability of the undiluted form was 100%. On the other hand, 2 compounds (methyl methacrylate, allyl heptanoate) caused decreased tissue viability only in the undiluted form. All other compounds caused decreases in viability in at least one solvent as well as in the undiluted form. These results were confirmed by release of 1α, IL-6, and IL-8 and by the release of IL-18 when looking for skin sensitization potential. Although the relationship between these in vitro results and known in vivo results remains unclear, we anticipate that the materials that caused in vitro cytotoxicity or cytokine release will have some level of in vivo response. Conclusion: This approach represents a promising in vitro method that may be utilized to distinguish non-irritants from compounds with weak but existing skin irritation potential.
Validation of Vaginal Tissue Model for Toxicity and Microbiome Studies (P422)
Kalyani Guntur, Jon Oldach, Mitchell Klausner, Alex Armento, Silvia Letasiova, Seyoum Ayehunie
MatTek Life Sciences, Ashland, MA
Session: Reproductive Toxicology II | March 12, 2024 | 2:15 – 4:15
Background: Personal lubricants delivered to the vaginal canal can cause damage to the vaginal epithelium, induce inflammation, and alter the microbiome of the vaginal ecosystem. Such epithelial barrier damage by chemicals/therapeutics and microbial dysbiosis can make women vulnerable to sexually transmitted and other infections. Therefore, manufacturers of personal lubricants and medical devices are required to show biocompatibility and safety assessment data to support regulatory decision-making within a specified context of use. The goal of the study is to validate the use of the 3D human in vitro vaginal tissue model as an alternate to the rabbit vaginal irritation (RVI) assay requested for the regulatory evaluation of medical devices in contact with the vaginal mucosa. Additionally, we examined the utility of the model for microbiome study in the vaginal microenvironment. Methods: A double-blinded study was conducted in vitro and in vivo with N=14 coded test articles (TAs) including materials intended to be in contact with vaginal tissues in the form of preservatives, contraceptives, solvents, viscosity enhancers, antiseptics, and cleansing agents (surfactants). The TAs were topically applied in vitro and in vivo at 2% dose with 5 repeat exposures over 6 days. Dose volumes were proportionally adjusted based on an estimated vaginal surface area. N=5 rabbits and N=3 EpiVaginal tissues were used per TA. RVI score was used to monitor in vivo irritation; for in vitro vaginal irritation assay, MTT, TEER, and histological analysis were used as endpoints. The variability of the reconstructed 3D vaginal tissues was further evaluated using tissues reconstructed from cells obtained from 4 donors. A total of 55 test articles (including those tested in RVI assay) commonly used in feminine care products were assessed using tissues from the four donors. For the in vitro assay triplicate vaginal tissues/per donor were topically exposed to TAs at a single concentration (2%) and incubated for 24 hr. After the 24 hour exposure, tissues were rinsed with PBS and examined for tissue viability (MTT assay) and for barrier integrity (TEER measurements). To study the vaginal microbiome, the effect of lactobacillus iners (normal flora in the vagina) alone or in the presence of Gardnerella vaginalis (associated with the disease bacterial vaginosis) on tissue viability (MTT), barrier integrity (TEER), and structural features (histology) were assessed. Results: In the RVI assay, benzalkonium chloride (BZK) was identified as a mild/severe irritant and the effect of nonoxynol-9 (N9) in rabbits was highly variable. In vitro results showed that four TAs including the two known irritants, BZK and N9 were correctly predicted as irritants by MTT viability and TEER (<50% reduction at time 24 hrs). These results were reproducible in vaginal tissues reconstructed from cells of the 4 donors. The results showed that 5 of the 55 test articles were found to be irritant in both the MTT and TEER assays in all donors. These TAs were Gynol containing 2% N9, BZK, sodium dodecyl sulfate, acetic acid, and Cremophor. These TAs resulted in >50% reduction in tissue viability and TEER values versus the saline controls. Only 1 test article, copper sulfate, showed low MTT viability (9.5%) but the TEER values for 3 of the 4 donors were non-irritating (>50%). The remaining test articles were found to be non-irritants in both MTT and TEER assays in all donors. In the microbiome study, exposure of the 3D vaginal tissue for 24 hrs to Lactobacillus iners followed by Gardnerella for 44 h results in no significant difference in MTT viability and TEER measurements. However, tissues exposed to Gardnerella showed perturbation of glycogen filled layer which was thinner in thickness compared to the no-bacteria tissues and tissues exposed only to lactobacillus. Conclusion: Tissues from the four donors were highly reproducible and a decrease in MTT and TEER appears to be useful endpoints for preclinical toxicity screening of chemicals and feminine care products. The reconstructed 3D vaginal tissue model is also a useful tool to study microbial dysbiosis and symbiosis. The use of this in vitro system to assess the safety of topically applied vaginal care products, drugs, vaginally inserted medical devices is cost effective and could replace the use of animals for experimentation.
Toxicological Aspects of in vitro Oral Mucosa Wound Healing Models (P792)
Jennifer Molignano1, Marek Puskar2, George Robert Jackson1, Silvia Letasiova2, Alex Armento1, Seyoum Ayehunie1, Mitch Klausner1
1MatTek Life Sciences, Ashland, MA and 2MatTek Europe, Bratislava, Slovakia
Session: New Approach Methods: In Vitro II | March 12, 2024 | 2:15 – 4:15
Background: Wounds in the oral cavity tissues can result from surgical procedures or accidents. Oral wounds present a site for potential infection and increased sensitivity to oral irritants. Untreated wounds can lead to pathogen invasion, chronic pain, poor cosmetic outcomes, and in some cases, extended hospitalization. Thus, damaged oral mucosal tissue needs to be treated and healed as soon as possible. The objective of this project is to develop new in vitro oral wound healing models and determine their toxicological profile using common dentifrice materials. Methods: Two tissue models were used in this study (both produced by MatTek Corporation and MatTek Europe): a) EpiOral (ORL), which consists of normal, human-derived oral epithelial cells cultured to form a highly differentiated model of human (non-cornified) buccal tissue, and b) EpiDerm (EPI), a cornified skin model which was used as a surrogate for gingival tissue. Tissues were wounded using a 3 mm biopsy punch (representing 11.6% of total tissue area) and wound closure was followed using brightfield microscopy, histological cross sections, and transepithelial electrical resistance (TEER) measurements. The toxicological profile of the tissues was probed using the common dentifrice additive, sodium dodecyl sulfate (SDS), at concentrations ranging from 0.2-2.0% and for doses from 25 to 400 μL. The exposure time for SDS which decreased the tissue viability to 50% (ET-50) was determined. Results: Histological cross sections of the wounded tissues (WT) showed that the wounds extend down to the underlying inert microporous membrane. Over a 4-day period post-wound (PW), the basal cells migrate into the wound to form a continuous monolayer as shown in brightfield images and histology cross-sections. Immediately following wounding, TEER for WT decreases to <20% of the non-wounded tissue (NWT) controls, however, TEER increases as wound healing proceeds. For the ORL model, no difference in tissue viability between WT and NWT was observed over the range of SDS concentrations and doses tested. However, for the cornified EPI model, which has a barrier similar to gingival tissue, the ET-50 for 1% SDS, 100 uL dose was 32.2 and 15.3 mins for the NWT and WT, respectively, and 234 and 107 mins when dosed with 25 uL of 1% SDS for the NWT and WT. Our conclusion is that the wound does not further decrease the weak barrier in the ORL tissue and hence no enhanced SDS toxicity is observed following wounding. For the EPI tissue which has a more significant barrier, the wound decreases the composite barrier function of the tissue and leads to significant differences in toxicity between NWT and WT. Conclusions: These results demonstrate successful development of wound models for tissues of the oral cavity. The wounds can be followed with brightfield microscopy, TEER, histology, and sensitivity to a common dentifrice additive. These models will be useful for testing new therapeutic compounds designed to hasten wound closure in the oral cavity and for determining toxicity profiles in barrier-compromised oral cavity tissues.
Prevalidation of the EpiDerm RhE Model in the EpiSensA Assay for in vitro Skin Sensitization (P311)
C. Pellevoisin1, F. Carriero2, B. De Servi2, K. Guntur3, M. Klausner3, and M. Meloni2
1Urbilateria, Saint Cyr sur Loire, France; 2VitroScreen, Milano,Italy; and 3MatTek Life Sciences, Ashland, MA, USA
Session: Skin Sensitization | March 11, 2024 | 9:15 – 11:15
Background and Purpose: Since the publication of the “The Adverse Outcome Pathway (AOP) for Skin Sensitization Initiated by Covalent Binding to Proteins” in 2012 (1), several methods based on the modelling of AOP Key events have been validated and integrated into TG 442C, D and E. In 2023, JacVam has scientifically validated the epidermal sensitization assay (EpiSensA) developed by Kao Corporation for integration into the OECD TG 442D for in vitro skin sensitization. This method addressing the key event 2 of the AOP is based on the use of a reconstructed human epidermis (RhE) model as an experimental system. The validation showed the performance of this method for pre/pro-haptens, thanks to metabolic activity of RhE models (2), as well as for lipophilic substances which can overcome some of the limitations of existing in chemico and in vitro 2D cells culture methods. In view of conducting an OECD me-too validation, we carried out a prevalidation for a similar method using the EpiDerm RhE model. The EpiDerm RhE model is already validated in the OECD TG431, TG439 and TG498 which are implemented in a large number of OECD member countries. The fact that the EpiSensA method was originally developed using this model was also an important factor in the selection of this model for a me-too validation (3). Methods: EpiSensA is based on gene expression quantification of four biomarkers related to the induction of skin sensitization: activating transcription factor 3(ATF3) and interleukin 8 (IL‐8) which reflect the inflammatory response of keratinocytes; glutamate-cysteine ligase, modifier subunit (GCLM) and DnaJ (Hsp40)homolog, subfamily B, member 4 (DNAJB4) which reflect the induction of cytoprotective gene pathways. The prediction model of the assays is based on the modulation of the expression of the four target genes quantified by quantitative real-time PCR analysis 6 hours after topical exposure of test chemicals. The chemical is classified as skin sensitizer if the fold induction of the expression of at least one of the genes exceeded the respective cut‐off value: 15‐fold for ATF3,2‐fold for GCLM or DNAJB4, and 4 fold for IL‐8. Results: For the prevalidation study, 10 chemicals covering the whole range from weak to extreme sensitizing compounds were tested in two laboratories, MatTek Life Sciences in the US and VitroScreen in Italy. Compared to the validated reference method (VRM), the protocol was optimized. To take account of the surface area of the EpiDerm model the volume of test chemical applied was increased from 5μl to 10μl. Regarding the exposure length, 6h was used as proposed in the VRM and 2 shorter times were evaluated, 30’ and 60’ followed by a post-incubation to perform gene expression 6h after application of the tested chemicals. Using a shorter exposure time could reduce the risk of cytotoxicity induced by certain compounds, while maintaining a post-incubation period allow the response to develop. This is useful in the case of weak skin sensitizers with cytotoxic effects. It is then possible to test higher concentrations while respecting the acceptance criteria of 80% cell viability. To measure the cell viability of the RhE, we also evaluated the use of the Alamar blue kit alongside to the lactatedehydrogenase (LDH) quantification method recommended in the VRM. Conclusions: Based on the set of chemicals tested, the accuracy of the method and its reproducibility with the EpiDerm RhE model meet the requirements of the VRM performance standards. The next step is now to transfer the method to two additional laboratories to carry out the multicenter validation study this year. The expected extension of the applicability domain of the OECD TG442D supported by the use of RhE models could enable this method to meet regulatory requirements for product categories other than chemicals, such as agrochemicals (4) or medical devices.
Subcategorization of Eye Irritants Using the EpiOcular Time-to-Toxicity Test Method (P639)
Silvia Letasiova1, Lenka Hudecova1, Jan Markus1, Yulia Kaluzhny2, Els Adriaens3, Mitch Klausner2
1 MatTek Europe, Bratislava, Slovakia, 2MatTek Life Sciences Ashland, MA, USA, 3Adriaens Consulting BVBA, Aalter, Belgium
Session: Ocular Toxicology | March 11, 2024 | 11:45 – 1:45
Background: As per OECD TG 405 “Acute Eye Irritation/ Corrosion”, albino rabbits were traditionally used to assess the eye damage/eye irritation of test materials. However, in 2015, OECD TG 492 “Reconstructed human Cornea-like Epithelium (RhCE) test method for identifying chemicals not requiring classification and labelling for eye irritation or serious eye damage” was accepted and validated for the use of in vitro ocular tissue models. Initially, this TG allowed for distinguishing between substances and mixtures not requiring classification and those that must be labeled for eye irritation or serious eye damage. Differentiation between materials causing serious eye damage and less-severe eye irritation was not included in the TG. Recently, OECD TG 492B was accepted which allows for distinguishing between chemicals that: a) do not require labeling for serious eye damage or eye irritancy (No Category or No Cat), b) cause serious eye damage (Category 1 or Cat 1), and c) are eye irritants (Category 2 or Cat 2) according to the UN GHS ocular hazard categories. Methods: Results from 2 studies, the CON4EI project (2017) and the ALT4EI project (2022), were combined and re-analyzed. Prediction models for liquids and solids were developed as part of the CON4EI project which involved a set of 80 chemicals (38 liquids and 42 solids). An additional 64 chemicals were tested within the ALT4EI project. When combined, a robust final set of 144 reference chemicals – 78 liquids and 66 solids, was used to confirm the new testing strategy. Results: The performance criteria, established by the OECD expert group overseeing OECD TG 492B, were met for all 144 chemicals. This data set was used to develop the EpiOcular time-to-toxicity test method for eye hazard identification of liquid and solid chemicals according to the three UN GHS. Based on the new testing strategy for liquids, 78.7% of Cat 1 (N=27), 63.5% of Cat 2 (N=26) and 82.0% of No Cat (N=25) were correctly identified. Using the new testing strategy for solids, 75.0% of Cat 1 (N=28), 59.4% of Cat 2 (N=16) and 80.3% of No Cat (N=22) materials were correctly predicted. Overall, the new test method correctly predicted 76.8% of Cat 1 (N=55), 61.9% of Cat 2 (N=42), and 81.2% of No Cat (N=47) test articles. Conclusion: The EpiOcular™ time-to-toxicity test method is a novel approach for subcategorizing both liquid and solid compounds. The prediction models that were developed for liquids and solids are capable of distinguishing substances and mixtures into the 3 UN GHS ocular hazard categories: No Cat, Cat 2, and Cat 1. The new test method will further reduce the need to perform animal tests for determination of the eye irritation potential of materials.
Human 3D Colon Tissue Model for Toxicity and Microbiome Studies (P132)
Jon Oldach, Camden Holm, Michell Klausner, Alex Armento, Seyoum Ayehunie
MatTek Life Sciences, Ashland, MA, USA
Session: Safety Assessment: Pharmaceutical-Drug Development II | March 11, 2024 | 2:15 – 4:15
Background: The microbiome of the colon is an ecosystem of microbes which is important to the immune system of the human body. Any damage or injury to the colonic epithelium disturbs this ecosystem can lead to inflammation and disease such as colorectal cancer. The colonic epithelium is a dynamic structure with a self-renewing capacity and serves as an organ for reabsorption of water, electrolytes, and bacterial metabolites. Developing a polarized and well-differentiated colon epithelial model with distinct luminal and basolateral sides has been a challenge. The aim of this study is to reconstruct and characterize a human 3D colon tissue model generated using epithelial cells from the human ascending colon to study 1) toxicants, 2) microbiome, 3) safety and efficacy of colorectal care products and anti-microbial agents, and 4) inflammation. Methods: This study characterizes the structural features of a novel in vitro tissue model reconstructed from normal human primary colon epithelial cells (CEC) with or without fibroblasts. Primary cells were isolated from a healthy human ascending colon, expanded in monolayer culture, and seeded onto microporous membrane inserts under air-liquid interface (ALI) conditions to form a well stratified 3D organotypic CEC tissues. The tissues were characterized for polarity (H&E staining), barrier integrity (transepithelial electrical resistance, TEER) measurement, presence of mucin producing goblet cells by Periodic acid–Schiff (PAS) staining, brush border formation (Villin staining), drug transporters and drug metabolizing enzymes (qPCR), and functionality in toxicological studies using Indomethacin (0.01-0.5mg/mL) and SN38 (20 uM) following apical exposure for 24 hr. Finally, the immune response of the tissue was probed by inducing it with cytokines and Tll-like receptor ligands. Results: Analysis of the 3D colon tissue model revealed: 1) wall-to-wall tissue growth in the cell culture insert, 2) an epithelial layer with in vivo like morphology, 3) a physiological TEER value of >200 Ώ*cm2 mimicking the colon microenvironment, and 4) surface marker expression of CK19 (epithelial cell marker), Alician blue PAS staining (mucous producing goblet cells), villin for brush border formation. Gene expression analysis revealed expression of drug transporters such as efflux transporters (PgP, MRP-1, 3, 5), sodium/glucose cotransporter (SLC5A1), and organic anion transporters (OAT) 2A1, 2B1, 3A1, and 4A1. Additionally, the tissue model also showed strong gene expression of both Phase I and Phase II drug metabolizing enzymes such as CYP2C18, CYP2C19, CYP 1A1, CYP 2B6, CYP2R1, CYP2S1, CYP3A4, CYP3A5, CYPs4F2, Carboxylesterases 1-3, glutathione S-transferases, and UDP gluconosyltransferases. Exposure to indomethacin and SN38 showed a 40% and a 96% decrease in TEER values, respectively, compared to untreated controls which indicate toxicity to the CEC tissue. Induction of the colon tissue model with cytokines (TNF-α and IL-β) and a Toll-like receptor ligand (LPS) induced release of inflammatory cytokines such as IL-6 and IL-8. Conclusions: This novel human cell-based CEC tissue model will likely be a useful tool for: 1) pre-clinical assessment of chemicals/candidate drugs for their toxicity potential, 2) studying the colon microbiome and bacterial metabolites, and 3) monitoring mucosal inflammation induced by chemical/compound insults. The CEC model can be added to the toolbox of 3D gastrointestinal tract tissue models to predict drug/chemical safety and efficacy, disease modeling, and microbiome studies. Ultimately this model will play a role in reducing animal testing and improving the preclinical drug development process.
3D Primary Human Kidney Tissue Model for Nephrotoxicity (P515)
Joseph Finelli, Yulia Kaluzhny, Michell Klausner, Alex Armento, Seyoum Ayehunie
MatTek Life Sciences, Ashland, MA
Session: Kidney | March 13, 2024 | 11:45 – 1:45
Background: Currently renal toxicity accounts for only 2% of drug failure during the preclinical phase but nearly 20% of drug withdrawal is associated with kidney damage during human clinical trials. This discrepancy emphasizes the need for a more physiological and functional alternative assay system that can predict and recapitulate drug induced kidney injury/toxicity in humans. The proximal tubular (PT) region is the most common site for compound and nutrient reabsorption and is highly susceptible to drug and toxin damage. PT region is responsible for essential kidney functions, including reabsorption of low molecular weight proteins, solutes, and glucose; secretion of acids; and clearance of administered medications. The goal of this study is to develop a novel physiologically relevant and functional primary human kidney-cell-based three-dimensional (3D) organotypic tissue model that can accurately predict drug-induced human nephrotoxicity. Methods: Human primary proximal tubular epithelial cells (PTEC) were isolated and expanded in a monolayer culture prior to seeding onto microporous membrane inserts to reconstruct a 3D organotypic tissue model. 3D tissues were analyzed by histology, barrier integrity (transepithelial electrical resistance, TEER), immunostaining, and qPCR on days 9-23. Receptor mediated FITC-albumin uptake and transpeptidase hydrolytic activity of glutamyl transpeptidase (GGT1) and leucine aminopeptidase (LAP) were assayed on days 9-16. Results: The PTEC organotypic tissues organize into characteristic tubular structures, develop a barrier with mean TEER values of 169±33.4 W×cm2 between days 9 to 23 of the culture period. Histological features show epithelial layer with tubular structure. Immunohistochemical analysis of the 3D kidney model stained positive for tight junction proteins ZO-1, and clauidn-1. At a gene level, the polarized organotypic tissues expressing brush border proteins megalin, cubilin, and villin, together with water channel AQP1 and GGT1 on the apical side and sodium-potassium ATPase pump on the basolateral side. Transmission electron microscopy confirms brush border formation. Real-time qPCR also confirmed that tissues express a panel of PTEC-specific markers that are necessary for renal clearance, secretion, and reabsorption: aminopeptidase CD13, p-glycoprotein (PgP; MDR1), multidrug resistance proteins MRP 1, 3, 4, and 5), CYP450 enzymes, glucose transporters SGLT1/2, multidrug and toxin extrusion transporter MATE1, organic cation and anion transporters OCT1/2, OCTN1/2, and OATP4C1, urate transporter URAT1, and sodium phosphate co-transporter NP2. Specific concentration and time dependent receptor mediated uptake of FITC-albumin by the PTEC tissues was observed by fluorescent microscopy. Kidney uptake of albumin was inhibited by addition of BSA (competitive binding utilizing a common receptor for albumin) or the drug chlorpromazine, an inhibitor of the calthrin-dependent endocytosis. Hydrolytic activity was monitored by the conversion of γ-Glutamyl-p-nitroanilide (GPNA) and L-leucine-p-nitroanilide (LLNA), substrates for GGT1 and LAP using spectrophotometric assays of p-nitroaniline (PNA) following 30 min incubation. Specific transpeptidase hydrolytic activity was inhibited in the presence of an irreversible inhibitor acivicin (1.2mM) by 88.8% (GPNA) and 35.0% (LLNA). Glucose uptake was also enhanced by the addition of sodium chloride. Furthermore, treatment of the 3D kidney model with Cisplatin, a known nephrotoxin that causes acute and chronic kidney injury, shows compromise in PTEC barrier integrity and reduced viability of tissues in a time and concentration dependent manner. Conclusions: The reconstructed in vitro 3D PTEC organotypic tissue is physiological in terms of structure, barrier properties, gene expression, and tissue functionality mimicking the in vivo human PT region. This model is anticipated to be a useful tool to evaluate human nephrotoxicity and to perform mechanistic studies that can improve the predictivity of human responses to pharmacological compounds. This model will help establish confidence in modeling drug induced kidney damage/injury and reduce animal use for experimentation.
Novel Organotypic 3D Human Liver Tissue Model for Drug Screening and Disease Modeling (P474)
Camden Holm, Mateo Frare, Michell Klausner, Alex Armento, and Seyoum Ayehunie
MatTek Life Sciences, Ashland, MA USA
Session: Liver: In Vitro | March 13, 2024 | 11:45 – 1:45
Background: Liver is one of the common sites of injury to toxicants or drugs and liver failure results in about 30% of post marketing withdrawals of pharmaceutical drugs. There is a need to develop a human primary cell-based 3D organotypic hepatocyte/liver tissue model that exhibits well polarized hepatic morphology, maintains high level expression of major liver associated drug transporters, and expresses metabolizing enzymes with metabolic competence. Models that can be cultured for weeks without phenotypic changes and that allow for repeat exposure and long-term dosing schedules are needed. Such models will recapitulate and predict the important aspects of human response to drugs/toxicants. Methods: We developed a human 3D liver tissue model on inert membrane using adult primary human hepatocytes. The hepatocytes were seeded onto Transwell inserts and fed with specialized medium to form well differentiated liver tissue with distinct apical and basolateral surfaces. This new model is distinct from liver commonly used liver spheroids. Tissues were characterized morphologically by histology. Albumin production was determine using ELISA and tight junction formation (ZO-1) was monitored by immunohistochemistry. qPCR was performed on tissues cultured for up to 23 days to monitor phenotypic and functional changes associated with tissue structure, drug transporters, and Phase I and Phase II drug metabolizing enzymes. Utility of the tissue model for drug toxicity studies was evaluated by dosing the human liver construct with two different concentrations of five model drugs (SN38, Bosentan, Diclofenac, Fialuridine, and Tolcapone) and the levels of liver injury were determined. Measurements of liver injury include compromised barrier integrity (TEER), reduction in albumin production, increased alanine aminotransferase (ALT) and Aspartate aminotransferase (AST) release, which are two of the biomarkers with clinical relevance in liver function test. To determine functionality of the liver model, tissues were exposed to the drug Midazolam, a substrate for CYP3A4 enzyme for two hrs. Culture supernatants from apical and basolateral sides were harvested and analyzed by LC/MS for 1-hydroxy midazolam, one of the metabolites of Midazolam. Results: Characterization of the tissue model showed 3D columnar hepatocyte tissue formation (histology), hexagonal cellular structure (topical view imaging), albumin production (immunohistochemistry and ELISA), and tight junction formation. qPCR results demonstrated high level expression of enzymes involved in drug metabolism such as CYP2C8, CYP2C9, CYP2C18, CYP2C19, CYP2D6, CYP2E1, CYP3A4, CYP3A5, CYP3A7, and CYP4A11. A significant accumulation of 1-hydroxy midazolam, the drug metabolite of Midazolam, was found in the culture supernatant. Repeated application of Fialuridine, a drug intended for hepatitis B treatment that was abruptly terminated due to induction of liver failure or causing of severe liver toxicity during human clinical trials, showed compromised barrier, reduced albumin release, and an increase in ALT and AST levels in a time-dependent manner, indicative of drug induced liver injury (DILI) predicting human responses. The positive control SN38, a metabolite of the cancer drug Irinotecan, also showed an increase in ALT and AST levels. This novel 3D human liver tissue model creates an opportunity to study liver physiology in an in vitro tissue microenvironment as a stand-alone platform or can be incorporated into organ on a chip device in microphysiological system (MPS) for organ-organ interaction simulations. Conclusions: In summary, the primary cell derived adult hepatocytes liver tissue model can 1) be cultured for a relatively longer time compared to existing practices without changing its functionality, 2) be used for infectious disease modeling, and 3) play a key role in screening drug candidates during drug development phase or for mechanism of action studies for investigational drugs. Such a model fits well with the FDA modernization Act 2.0 guidelines and will have impact in the development of new approach methodologies (NAMs) intended to identify adverse effects of therapeutic candidates and reduce animal use for experimentation.
Use of 3D Airway Tissue Models Cultured from Rat, Primate, or Human Cells for Translational Inhalation Studies (P826)
Seyoum Ayehunie1, George R Jackson1, Kaitlyn Coen1, Timothy Landry1, Jan Markus2, Silvia Letasiova2, Mitch Klausner1, Alex Armento1
1MatTek Life Sciences, Ashland, MA and 2MatTek Europe, Bratislava, Slovakia
Session: Respiratory Toxicology I | March 12, 2024 | 11:45 – 1:45
In vitro models of the respiratory tract are highly differentiated and have been commercially available since 2000. These models have been widely used for toxicological, respiratory infection, tobacco safety, and inhaled drug delivery studies. Availability of tissue models representing the different segments of the respiratory tract were instrumental in gaining insight into the underlying mechanisms of SARS-CoV-2 infection and for screening of anti-viral compounds. Despite these useful applications, there are large databases of animal toxicity data which are not directly translatable to data obtained from the human in vitro airway tissue models due to species differences. To close this translational gap, cells harvested from both rat and non-human primate (rhesus monkey) tissues were utilized to develop models similar to EpiAirway, the tracheobronchial tissue model offered by MatTek that is cultured with normal human cells. The tissues were characterized for structure (histology), epithelial cell markers (IHC), barrier integrity (transepithelial electrical resistance, TEER measurement), and functionality (inhalation toxicological studies). The animal cell-derived 3D tissues exhibited similar characteristics to human tissues including: well polarized epithelia with physiological TEER values of >300 Ώ*cm2, cilia formation on the apical surfaces, and mucin production mimicking the airway microenvironment. Acute exposure to 4 chemical toxicants (CT) showed species-specific changes in tissue viability and membrane integrity as measured by MTT and TEER assays, respectively. The effective dose concentration that reduces tissue viability by 50% (ED-50) for vinyl acetate (VA) and chloroacetaldehyde (CA) were both <2 mg/tissue and the ED-50 for propylene glycol (PG) was > 20 mg/tissue for all species. However, the ED-50 values for toluene (T) showed differences between the species: human >20 mg, primate 16.2±1.7 mg, and rat 13.8±0.1mg. Based on the MTT viability and TEER values, the test chemicals were rank ordered from high to minimal toxicity: CA > VA > T > PG and the vehicle controls (water and corn oil). In order to verify the reliability of these models worldwide, quality control data for tissue lots produced in the US and Europe were compared. TEER values in the US averaged 1094 ± 325 (n=141 lots) vs. 913 ± 238 (n=64 lots) in Europe. Results were not statistically different (p < 0.001). Although more chemicals need to be tested, the multispecies 3D airway tissue models will be vital translational tools to predict airway inhalation toxicity and to bridge the in vitro – in vivo knowledge gap to reliably predict human responses, while providing a worldwide alternative approach to animal experimentation.