American College of Toxicology

43rd Annual Meeting

Poster Abstracts

Gaylord Rockies, Denver, Colorado

The abstracts accepted for presentation at the ACT Annual Meeting were reviewed and approved by the ACT Program Committee. Every effort has been made to reproduce the abstracts as submitted. The author(s) of the abstracts are solely responsible for the data and content provided. The American College of Toxicology does not claim to validate, promote, or share in the representation of the content provided by the submitting author.

100s—General Toxicology

P101: North American Travel Grant Recipient: Role of Metalloproteases in Mediating Skin Penetration in Parasitic Nematodes

Courtney R. McClure¹, Michelle L. Castelletto¹, and Elissa A. Hallem¹

¹University of California, Los Angeles, Los Angeles, CA, USA

The skin-penetrating parasitic nematode Strongyloides stercoralis infects 610 million people worldwide. There are currently no available prophylactics to prevent infection with S. stercoralis or any other nematode. To identify potential targets for nematode prophylactics, we are investigating the role of the astacin family of metalloproteases in mediating skin penetration using the rat parasite Strongyloides ratti as a model. Astacins have long been hypothesized to play a role in skin penetration; however, these proteins remain functionally uncharacterized. We analyzed publicly available RNA-seq data and identified six astacin genes that are both highly expressed in infective larvae and upregulated in infective larvae relative to other life stages: SRAE_2000454000, SRAE_2000469400, SRAE_0000046000, SRAE_0000047800, SRAE_0000048100, and SRAE_0000057400. For example, SRAE_2000454000 is highly expressed in infective larvae (log2CPM: 10.174) and highly upregulated in infective larvae relative to parasitic females (logFC: 12.05, p-value: 1.92e-07). We used WormBase ParaSite to retrieve astacin gene sequences based on their gene IDs. We then used a MUSCLE alignment to align amino acid sequences and IQ-Tree to conduct a phylogenetic analysis of the astacin gene families in S. stercoralis, S. ratti, and the free-living nematode C. elegans. We found that all six of these astacin genes belong to parasite-specific expansions of the astacin gene family, with no close homologs in C. elegans. We are now testing the requirement for these six genes in mediating skin penetration using CRISPR. In the long run, compounds that specifically inhibit Strongyloides astacins could form the basis for novel topical anthelmintics.

P102: International Travel Grant Recipient: Assessment of Heavy Metal Levels in Fruits and Vegetables Sold in Gwagwalada Market, Abuja, Nigeria

Onakpa Monday¹ and Bolanle Akanbi¹

¹University of Abuja. Nigeria, Abuja, FCT, Nigeria

Fruits and Vegetables such as Banana (Musa), Cucumber (Amaranthus cruentus), Green Peas (Pisum sativum), Carrot (Dactus carota subsp. Sativus), Onion (Allium cepa), Lettuce (Latuca sativa), and Cabbage (Brassica oleracea var. capitata) randomly collected from the Gwagwalada market was carried out. The samples were assessed for the heavy metal presence and concentrations using Atomic Absorption Spectrophotometer (AAS). The presence of Cd, Cr, Ni, and Pb was measured using granite- furnace AAS (detection limits of 0.002 µgL - ¹, 0.004 µgL - ¹, 0.07 µgL - ¹, and 0.05 µgL - ¹ respectively) and the presence of Cu, Mn, and Zn by air- acetylene flame AAS (detection limit of 0.0015 µgL - ¹ for all the three heavy metals), both using the atomic absorption spectrophotometer. Hollow cathode lamps were used as sources of radiation from detection. The level of concentration of Copper in, Cucumber was found to be12.45 ± 1.2, Cabbage had 10.68 ± 3.1, Banana had 16.14 ± 1.1, Lettuce had 17.89 ± 1.2, Green peas had 17.28 ± 1.1, Carrot had 19.78 ± 1.2 Onion has 12.94 ± 1.2. The level of concentration of Lead, Cucumber had 8.65 ± 1.1, Cabbage had 0, Banana had 100.40 ± 10.0, Lettuce 68.00 ± 7.0, Green peas 64.76 ± 5.8, Carrot had 61.28 ± 5.6, Onion had 8.92 ± 1.1. It is necessary to monitor levels of these heavy metals in food and in the body. It is recommended that farmlands be located away from mining sites. Also, regular monitoring of foods for these heavy metals is essential to prevent their excessive buildup in the food chain.

P103: North American Graduate Fellowship Recipient: Neuroinflammation Oxylipins and Arachidonic Acid Levels in an Atropinized Rat Model Following Acute Sarin Surrogate Exposure and Novel Oxime Therapy

Chiquita Y. Price¹, Matthew K. Ross¹, Brian S. Backer¹, and Janice E. Chambers¹

¹Center for Environmental Health Sciences, College of Veterinary Medicine, Mississippi State University, Mississippi State, Mississippi, USA

The sarin surrogate nitrophenyl isopropyl methylphosphonate (NIMP), a potent organophosphorus compound (OP), causes acute neurotoxicity by inhibiting acetylcholinesterase (AChE) in the brain. Acute OP exposure causes toxic cholinergic signs and is known to induce neuroinflammation, leading to long-term neurologic and cognitive deficits. Our laboratory has invented a series of oxime AChE reactivators (US Patent 9,227,937) that enter the brain, reduce time to cessation of seizure-like activities, and attenuate neuropathology after NIMP exposure. In this study, critical brain regions of seizure initiation were examined for neuroinflammatory oxylipins and arachidonic acid (AA) 1-, 4-, and 7- days post lethal NIMP (0.6 mg/kg, s.c.) challenge and novel oxime therapy (146 μmol/kg, i.m.). Novel oxime therapy dose is molar equivalent to the human dose of the U.S. approved oxime, pralidoxime (2-PAM). When compared to vehicle controls, NIMP challenge decreased the most abundant pro-inflammatory oxylipins to be at day 1, increased at day 4, and their levels were similar at day 7 indicating possible resolution of neuroinflammation. By day 4, NIMP-challenged animals given oxime therapy exhibited higher levels of pro-inflammatory oxylipins than those in the vehicle control group; however, they were lower than those animals challenged with NIMP only, suggesting an oxime-mediated anti-inflammatory effect. AA levels mimicked the increased and decreased in oxylipin levels. These results suggested that NIMP can induce neuroinflammatory lipids by day 4 and that novel oxime therapy could attenuate this effect. Thus, reducing OP-induced neuroinflammation might contribute to the neuroprotection previously observed with these novel oximes. (Supported by NIH U01NS107127, U01NS123255)

P104: International Travel Grant Recipient: Assessment of Stress Hormone, Antioxidant Status, and Histopathology as models for Physiological Stressors induced by Radiofrequency Radiation on Mus Musculus

Funmilayo V. Doherty¹, Idowu A. Aneyo², and Soladoye B. Iwajomo²

¹Yaba College of Technology, Lagos, Lagos, Nigeria

²University of Lagos, Lagos, Lagos, Nigeria

Physiological regulatory hormones aid vertebrates to maintain stable conditions internally in a toxic environment. Radiofrequency radiation (RFR) has been identified as a potential threat that can alter the stability of a living system. This study aims to investigate the toxicological effects of RFR induced on endogenic stress hormone, oxidative stress, and histopathology of mice after 60 days of exposure under a telecommunication mast. Forty mice, Mus Musculus (five weeks old) were selected in four exposure cages with eight mice per cage and the control group. Levels of cortisol, thyrotropin (TSH), Triiodothyronine (T₃), and thyroxine (T₄) were assessed in the brain using enzyme-linked immunosorbent assay (Elisa) kits. Levels of oxidative stress; (GSH: glutathione, CAT: catalase, SOD: superoxide dismutase, & MDA: malondialdehyde), acetylcholinesterase & histopathology were carried out in the brain using standard protocols. The levels of GSH, CAT, SOD, and acetylcholinesterase were determined and showed a significant decrease (P < 0.05) in exposed when compared with control. The MDA level significantly increased as the exposure days increased. The stress (cortisol, TSH, T3 &T4) results do not show a significant change in the exposed group compared to the control. The histological micrograph does not show any alteration in the sections in the brain, however, histologic sections of the observed tumor in exposed mice showed tubules lined by distorted spermatogenic cells. In conclusion, oxidative stress levels and histopathology can be used as models for physiological stressors induced by RFR in vertebrates and prolonged exposure to RFR should be prevented.

P105: Decreases in Circulating T Cells, Natural Killer (NK) Cells and Lymphoid Organ Cellularity in Cynomolgus Monkeys by REGN7257, an Antagonistic Antibody to the Human Interleukin 2 Receptor Gamma Subunit (IL2RG)

Chandrashekhar K. Korgaonkar¹, Prachi Sharma¹, Hao Pan¹, Audrey Lefloch¹, Jamie Orengo¹, Adelekan Oyejide¹, and Jon Popke¹

¹Regeneron Pharmaceuticals, Tarrytown, NY, USA

REGN7257 is an antagonistic monoclonal antibody to common gamma chain (g_c), a cytokine receptor subunit shared by the receptor complexes of six different cytokines (IL2, IL4, IL7, IL9, IL15 and IL21). REGN7257 is being investigated for the treatment of refractory severe aplastic anemia. In a toxicology study in cynomolgus monkeys, REGN7257 was administered once weekly for five weeks by intravenous infusion at 0 (control), 10, 30 and 100 mg/kg followed by an eight-week recovery period. Continuous exposure to REGN7257 was observed at all dose levels during the dosing period with no observed impact of anti-drug antibodies. REGN7257 was well tolerated at all dose levels. Consistent with pharmacology, decreases in absolute lymphocyte count and different lymphocyte populations (NK cells and/or T cells) were noted at all doses. Decreases in certain T-cell populations persisted during the recovery period and were consistent with detectable serum REGN7257 levels. The clinical pathology changes included minimal to mild decreases in serum albumin and increases in globulin at all doses and minimal increases in fibrinogen at 100 mg/kg. The histopathology findings were considered non-adverse and included minimal to mild increases in mononuclear and/or mixed cell infiltrates in several non-lymphoid tissues and renal intratubular hemoglobin casts (terminal necropsy only) at all doses and decreased lymphocytic cellularity of lymphoid organs with correlating decreased thymus weights at ≥ 30 mg/kg. In conclusion, REGN7257-related pharmacological effects were noted at ≥ 10 mg/kg without any apparent adverse effects up to 100 mg/kg in cynomolgus monkeys following once weekly administration for five weeks.

P106: North American Graduate Fellowship Recipient: Photochemical Targeting of PFAS-Induced Platinum Resistance in Ovarian Cancer

Brittany P. Rickard¹, Suzanne E. Fenton², and Imran Rizvi¹

¹University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, USA

²National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina, USA

Per- and polyfluoroalkyl substances (PFAS) are known disruptors of mitochondrial function and have been linked to increased ovarian cancer (OvCa) risk. OvCa is the most lethal gynecologic malignancy, mainly because most patients become resistant to platinum-based chemotherapy, highlighting the need for the development of mechanism-based treatments. The purpose of this study was to evaluate 1) the contribution of PFAS to mitochondrial disruption and platinum resistance in OvCa, and 2) the ability of mitochondrial-targeted photodynamic priming (mt-PDP), a photochemistry-based treatment, to overcome PFAS-induced platinum resistance. OVCAR-3 cells were exposed to PFAS at sub-cytotoxic concentrations for 48 hours prior to mt-PDP and/or treatment with 25-400μM carboplatin, a platinum-based chemotherapeutic agent. Mitochondrial membrane potential (ΔΨ_m) and cell survival fraction post-mt-PDP and/or carboplatin treatment were measured using JC-1 dye and the CellTiter Glo assay, respectively. Carboplatin resistance, defined as an increased survival fraction post-carboplatin treatment compared to controls (normalized survival fraction = 1.0), was observed in OVCAR-3 cells exposed to perfluoroheptanoic acid (PFHpA; 1.492 ± 0.445, n = 4, P < 0.05) and perfluoropentanoic acid (PFPA; 1.264 ± 0.153, n = 4, P < 0.05). Additionally, ΔΨ_m increased in OVCAR-3 cells exposed to PFHpA (1.208 ± 0.256, n = 4, P < 0.05) or PFPA (1.386 ± 0.298, n = 4, P < 0.05) and remained elevated post-carboplatin treatment, indicative of improved cellular health. When PFAS-exposed cells received mt-PDP prior to carboplatin administration, survival fraction decreased in PFHpA (0.0825 ± 0.029, n = 4, P < 0.05) or PFPA-exposed (0.0875 ± 0.033, n = 4, P < 0.05) OVCAR-3 cells by ∼90%. These findings suggest PFAS contribute to platinum resistance in OvCa by altering mitochondrial function. On a larger scale, mt-PDP may prove effective for the treatment of platinum-resistant OvCa.

P107: North American Travel Grant Recipient: RNA Binding Protein ZFP36L1 Protects Against Acetaminophen-Induced Acute Liver Injury

Rahul Kumar¹, Yogesh Saini¹, and Sonika Patial¹

¹Department of Comparative Biomedical Sciences, School of Veterinary Medicine, Louisiana State University, Baton Rouge, Louisiana, USA

Zinc protein 36 like 1 (ZFP36L1), an RNA binding protein with mRNA destabilizing properties, is abundantly present in the liver but its role in liver physiopathology remains elusive. Our initial experiments showed a significant downregulation of Zfp36l1 in the liver after acetaminophen overdose. Here, we tested the hypothesis that the deletion of ZFP36L1 in the liver exacerbates acetaminophen-induced acute liver injury. Adult C57BL/6J liver-specific ZFP36L1 knockout (L1^LKO; AlbCre⁺/Zfp36l1^flox/flox) and flox-only control (L1^FLX; AlbCre^-/ Zfp36l1^flox/flox) mice were fasted overnight and were intraperitoneally injected with acetaminophen dissolved in saline (300 mg/Kg body weight). Liver injury and associated inflammation endpoints were assessed at 1, 6, 24, and 72-hours post-administration. Acetaminophen overdose resulted in increased centrilobular necrosis with prominent neutrophil infiltration in the liver of L1^LKO mice compared to L1^FLX mice. Further, the serum levels of liver function enzymes, ALT (p = 0.0256), and AST (p = 0.0151) were also significantly elevated in L1^LKO than in the L1^FLX group. Finally, the expression levels of pro-inflammatory mediators, particularly Tnfa and Cxcl2, were upregulated while the levels of anti-inflammatory mediator, Il10, were reduced in L1^LKO mice but not in LI^FLX mice. Interestingly, the liver glutathione levels and cytochrome P-450 2E1 expression at basal and 6-hours post-acetaminophen dosage along with c-Jun N-terminal kinase (JNK) activation in the liver were comparable between L1^LKO and L1^FLX mice. In conclusion, RNA binding protein, ZFP36L1, protects against acetaminophen-induced acute liver injury by preventing the development of a proinflammatory phenotype. ZFP36L1 could be exploited as a therapeutic target against xenobiotic stress in the liver.

P108: Removed

P109: International Travel Grant Recipient: Biochemical and Pharmacological Properties of Diphenyl diselenide against DMBA-induced Mammary Tumorigenesis in Wistar Rats

Adeleye A. Edema¹, Chijioke N. Onah¹, Anuoluwa A. Oloyede¹, and Oluwatosin A. Adaramoye¹

¹Molecular Drug Metabolism and Toxicology Laboratories, Department of Biochemistry, Faculty of Basic Medical Sciences, College of Medicine, University of Ibadan, Ibadan, Oyo State, Nigeria

Background: Breast cancer is the leading cause of death in women worldwide. The use of natural or synthetic dietary or pharmacological substances to prevent breast cancer has emerged as an appealing option for combating the disease rising prevalence. Objective: In this study, we investigated the effect of diphenyl diselenide (PhSe)₂ against 7, 12 dimethylbenz[a]anthracene (DMBA) induced mammary gland tumor. Materials and Methods: Forty female Wistar rats were assigned into five equal groups: Group 1 served as control, group 2 received DMBA (80mg/kg), group 3 and 4 received DMBA and treated with (PhSe)₂ at 5 and 10 mg/kg, respectively while group 5 received DMBA and treated with vincristine at 0.5 mg/kg. DMBA (oral) was dissolved in corn oil, (PhSe)₂ (oral) and vincristine (intraperitoneal) were given thrice and once per week, respectively for 15 weeks. Results: DMBA administration significantly decreased body weight gain by 37% while organo-somatic weight of mammary gland increased by 23%. DMBA decreased the activities of mammary reduced glutathione, glutathione-s-transferase, and glutathione peroxidase by 16%, 68%, and 30%, respectively. DMBA increased oxidative stress markers; nitrite, malondialdehyde and myeloperoxidase by 24%, 75%, 67%, respectively. Furthermore, DMBA increased tumor necrosis factor-alpha and decreased Bcl-2 associated X protein. Immunohistochemistry revealed expression of estrogen receptors, epidermal growth factor receptors in DMBA-administered rats. Histology revealed glands with hyperchromatic nuclei glandular epithelial cells and infiltrating inflammatory cells in DMBA rats. Conclusion: Treatment with diphenyl diselenide reduces DMBA-induced oxidative stress, inflammation, and restores mammary gland cytoarchitecture by inhibiting tumorigenesis via the inflammation and apoptosis pathways.

P110: Withdrawn

P111: North American Travel Grant Recipient: Evaluation of 1,4-Dinitrobenzene Male Reproductive Toxicity: In Vivo and In Vitro Findings Using Wistar Rats and Cell Lines Models

Janet O. Olugbodi¹, Thomas K. Monsees², Mary Olaleye³, and Afolabi C. Akinmoladun³

¹Bingham University, Karu, Nasarawa, Nigeria

²University of The Western Cape, Cape Town, Western Cape, South Africa

³Federal University of Technology, Akure, Ondo, Nigeria

1,4-Dinitrobenzene (1,4-DNB) widely used in certain dye, plastic and explosives manufacturing industries has been identified as an environmental toxicant and industrial pollutant causing great health concern. This study aimed at providing in vivo and in vitro findings regarding 1,4-DNB-induced reproductive dysfunction in male Wistar rats and TM3 Leydig cell lines. For in vivo studies, twenty-one male Wistar was randomly divided into three groups (n = 7), and orally administered with vehicle, 1,4-DNB (25, and 75 mg/kg) for twenty-one days. Thereafter, sperm count, viability and morphology, oxidative stress marker, body weight gain, organ weight and histology were evaluated. Furthermore, TM3 Leydig cells exposed to 1,4-DNB (10, 30 and 50 µg/ml), and cell viability, epithelial cell barriers, mitochondrial membrane potential, apoptosis and reactive oxygen species were determined. Results showed significant decrease (p < 0.05) in sperm count, sperm viability, sperm morphology, body weight gain, testis weights and increased testicular malondialdehyde (MDA) level in all treated groups. Additionally, 1,4-DNB treated group’s revealed moderate congestion of interstitial vessels, fibrotic seminiferous tubules, degenerated germinal cells, abnormal Leydig cells and oedema in the testes. Results from in vitro studies indicated that 1,4-DNB had cytotoxic impact on cell viability and disrupt epithelial cell barriers in treated Leydig cells. Also, findings from this study showed that 1,4-DNB exposure significantly increase (p < 0.05) apoptosis and reactive oxygen species levels, caused mitochondrial membrane injury in TM3 Leydig cells. Overall, 1,4-DNB induce toxicity in the testes, spermatozoa and Leydig cells and could lead to male-related infertility in people who are unduly exposed to the toxicants.

P112: North American Travel Grant Recipient: Decreasing Cholesterol Efflux from Macrophages Prevents Crystalline Silica (cSiO₂)-Induced Lysosomal Membrane Permeability (LMP)

Rebekah L. Kendall¹ and Andrij Holian¹

¹Center of Environmental Health Sciences, University of Montana, Missoula, MT, USA

Exposure to cSiO₂ can result in progressive lung fibrosis and no effective treatments are available. LMP within macrophages taking up cSiO₂ is the proposed rate-limiting step in the development of NLRP3 inflammasome assembly, activation, and resulting inflammation. Increased lysosomal cholesterol reduces LMP in a variety of cell models. Cationic amphiphilic drugs that functionally inhibit acid sphingomyelinase activity (FIASMAs) accumulate in the lysosome and reduce cholesterol efflux. We propose FIASMAs prevent LMP and IL-1β release by reducing cholesterol efflux, thereby increasing lysosomal cholesterol. Here we show that the decrease in cholesterol efflux is protective of cSiO₂ caused LMP and IL-1β release from C57Bl/6 bone marrow derived macrophages (BMdM) while increasing cholesterol efflux exaggerates cSiO₂ induced inflammation. cSiO₂ (50 µg/mL; 24 hr) caused a statistically significant increase in LMP and IL-1β compared to control cells; however, pretreatment with imipramine (IMP), hydroxychloroquine (HCQ), fluvoxamine (FLV), or fluoxetine (FLX) before cSiO₂ exposure significantly reduced LMP and IL-1β. Total & free cholesterol levels were slightly increased with all FIASMA treatments (except FLV), while cholesterol esters were increased with all treatments. Cholesterol efflux from FIASMA treated cells was significantly decreased from control cells. In contrast to the effect of the FIASMAs, a cholesterol efflux promotor, phosphatidylglycerol (PG; 50 µM) resulted in a significant increase in LMP and IL-1β compared to cSiO₂ treatment alone and diminished the efficacy of FIASMA drugs on preventing IL-1β release. These findings support the suggestion that lysosomal cholesterol content can regulate cSiO₂-induced LMP. This work is supported by NIEHS award F31ES033562

P113: North American Graduate Fellowship Recipient: Systemic Overexpression of Tristetraprolin (TTP) Mitigates Inflammatory Responses in Allergic Asthmatic Mice

Richa Lamichhane¹, Yogesh Saini², and Sonika Patial¹

¹Department of Comparative Biomedical Sciences, School of Veterinary Medicine, Louisiana State University, Baton Rouge, Louisiana, USA

²Louisiana State University, Baton Rouge, LA, USA

Allergic asthma is a chronic inflammatory disease characterized by immune cell infiltration, airway remodeling, mucus cell metaplasia (MCM), and airway hyper-responsiveness (AHR). Tristetraprolin (TTP) is an mRNA binding protein that binds to AU-rich elements within the 3′ untranslated regions of certain transcripts for inflammatory genes and increases their rate of decay. Here, we tested the hypothesis that systemic TTP overexpression mitigates mixed allergen (MA)-induced allergic asthma. TTP-overexpression (TTP^ΔARE) and TTP-WT (TTP^WT) adult mice from the same line were intranasally exposed to MA for four weeks. Lung injury, inflammation, and airway remodeling were assessed after 48 hours post-last dose of MA. Additionally, bone-marrow transplantations were performed followed by administration of MA 8-weeks post-bone marrow reconstitution. As compared to MA-challenged TTP^WT mice, the MA-challenged TTP^ΔARE mice showed significantly decreased infiltration of immune cells in BALF (P < 0.0001), mitigated consolidation, and fibrosis in the peribronchial region, and inhibited MCM. Interestingly, irradiated TTP^WT mice reconstituted with TTP^ΔARE hematopoietic progenitor cells (HPCs) exhibited significantly decreased inflammation(p < 0.05), their reconstitution with TTP^KO HPCs exaggerated inflammation(p < 0.05). Furthermore, the reconstitution of irradiated TTP^ΔARE mice with either TTP^WT or TTP^KO HPCs significantly increased inflammation (p < 0.05). However, the inflammation was more robust when TTP^KO HPCs were given to TTP^WT recipients compared to TTP^ΔARE recipients. These data together suggest that the systemic overexpression of TTP protects mice against MA-induced asthma and that this effect is largely contributed by TTP overexpression in the hematopoietic lineage cells of the lung. Our findings emphasize the role of TTP in the amelioration of MA-induced asthma.

P114: North American Travel Grant Recipient: A Mouse Study Investigating the Role of CYP2E1 in 1,4-dioxane-induced Liver Cytotoxicity and Genotoxicity

Yewei Wang¹, Georgia Charkoftaki¹, David Orlicky², Emily Davidson¹, Gary Ginsberg¹, David C. Thompson¹, Vasilis Vasiliou¹, and Ying Chen¹

¹Yale University, New Haven, Connecticut, USA

²University of Colorado Anschutz Medical Center, University of Colorado, Aurora, CO, USA

1,4-Dioxane (DX), an emerging water contaminant, has been classified as a group 2B carcinogen with the primary target organ being the liver in animal studies. A clear understanding of the DX carcinogenic mechanism is important for the development and evaluation of effective mitigation of this public health hazard. Studies from our group revealed that high dose DX exposure to mice in the drinking water up to 3 months caused mild cytotoxicity, but oxidative DNA damage in the liver, a process correlating with hepatic CYP2E1 induction and elevation of lipid peroxidation. To test the hypothesis that CYP2E1 is required for DX-induced liver cytotoxicity and genotoxicity, in the current study, we treated Cyp2e1-null (Cyp2e1 ^KO ) male mice with 5000 ppm DX in the drinking water for 1 week and 3 months. Both short-term and sub-chronic DX exposure failed to cause liver cytotoxicity in Cyp2e1 ^KO mice. Hepatic oxidative stress and NRF2 antioxidant response, which were observed in DX-treated WT mice, were also blunted in Cyp2e1 ^KO mice. Unexpectedly, subchronic DX exposure induced a trend of elevated oxidative DNA damage and suppression of DNA damage repair response in Cyp2e1 ^KO livers. Our findings suggest that CYP2E1 is responsible for DX-induced liver cytotoxicity and oxidative stress, may be partially involved in DX-induced liver genotoxicity.

P115: Withdrawn

P116: North American Graduate Fellowship Recipient: Structure Activity Relationship and Gliotoxicity of Lower Chlorinated Biphenyls

Neha N. Paranjape¹, Brianna C. Cagle¹, Andres Martinez¹, Ronald B. Tjalkens², Hans J. Lehmler¹, and Jonathan A. Doorn¹

¹University of Iowa, Iowa City, IA, USA

²Colorado State University, Fort Collins, Colorado, USA

Exposure to polychlorinated biphenyls (PCBs) is associated with developmental neurotoxicity and neurodegenerative disorders; however, the underlying mechanisms of pathogenesis are unknown. Because normal brain function is largely glial cell-dependent, we hypothesize that glia play an important role in PCB-mediated neurotoxicity. We assessed the toxicity of four lower chlorinated PCB congeners, 4-chlorobiphenyl (PCB3), 3,3’-dichlorobiphenyl (PCB11), 2,3’,4-trichlorobiphenyl (PCB25), and 2,2’,5,5’-tetrachlorobiphenyl (PCB52) and, their corresponding human-relevant hydroxylated and sulfated metabolites. Cell viability of C6 cells (rat glioma cell line) or primary glial cells (from C57BL/6 mice or Sprague-Dawley rats) exposed to test compounds at a concentration range of 0.5 to 50 μM for 24 h was assessed by MTT assay. PCB52 and both its metabolites were the most toxic in C6 cells at LC₅₀ concentrations of 8.4 μM (PCB52), 2.2 μM (4-OH-PCB52), and 8.8 μM (4-PCB52-sulfate). Hence, further studies focused on these compounds. PCB52 and its metabolites were similarly cytotoxic to primary glia as in C6 cells. PCB52 metabolites generate mitochondria-targeted oxidative stress and subsequent studies using a Seahorse analyzer demonstrate that they cause proton leak, suggestive of energetic crisis in C6 cells. PCB partitioning in different abiotic and biotic compartments like plastic-air-cells and lipid-water-protein is structure dependent and is useful in explaining observed cytotoxicity. These findings suggest that mechanisms of cytotoxicity by parent congener and metabolites are distinct, and that PCB cytotoxicity is structure dependent. Future studies will assess mechanisms of cytotoxicity in glial cells and functional implications of these findings in PCB-mediated neurotoxicity. This work was funded by [ES005605, ES013661, ES029035].

P117: Removed

P118: North American Graduate Fellowship Recipient: Alcohol and HIV-induced Hepatocyte Apoptotic Bodies Program Hepatic Stellate Cells for Profibrotic Activation: The role of Reactive Oxygen Species

Moses New-Aaron¹, Murali Ganesan², Siva Koganti¹, Weimin Wang¹, Edward Makarov¹, Mojisola Ogunnaike¹, Kusum K. Kharbanda¹, Larisa Y. Poluektova¹, and Natalia A. Osna¹

¹University of Nebraska Medical Center, Omaha, Nebraska, USA

²University of Nebraska Medical Center, Omaha, NE - Nebraska, USA

Fifty percent of liver diseases among people living with HIV are due to alcohol. Recently, we found that both alcohol metabolites, acetaldehyde, and HIV, induce hepatocyte apoptosis resulting in the release of large extracellular vesicles called apoptotic bodies (ABs), and the engulfment of these hepatocyte ABs by hepatic stellate cells (HSC) leads to profibrotic activation. This study aimed to establish the mechanisms of HSC activation after engulfment of ABs from acetaldehyde and HIV-exposed hepatocyte (AB_AGS+HIV). Huh7.5-CYP (RLW) cells were used to generate hepatocyte ABs and LX2 cell-line as HSC. To generate ABs, RLW cells were pretreated for 24 h with acetaldehyde, then exposed overnight to HIV1_ADA and to acetaldehyde for 96 h. After that, ABs were isolated from cell suspension by differential centrifugation method and incubated with LX2 cells (3:1 ratio) for pro-fibrotic genes and protein analyses. We found that HSC internalized ABs via the tyrosine kinase receptor, Axl. While the HIV gag RNA/HIV proteins were deposited in LX2 cells after AB internalization, no productive HIV infection was elicited in LX2 and immune cells exposed to AB_AGS+HIV; however, they triggered ROS and IL6 generation, which, in turn, activated pro-fibrotic genes via the JNK-ERK1/2 and JAK-STAT3 pathways. Similarly, ongoing pro-fibrotic activation was observed in immunodeficient NSG mice fed ethanol and injected with HIV-derived RLW ABs. We conclude that HSC activation by hepatocyte AB_AGS+HIV engulfment is mediated by ROS-dependent JNK-ERK1/2 and IL6 –triggering of JAK-STAT3 pathways. This can explain the mechanisms of liver fibrosis development frequently observed among alcohol abusing PLWH.

P119: Identification of MicroRNAs as Biomarkers for Drug-Induced Testicular Toxicity in Rats

Edwin F. Custodio¹, Dany Gitnick¹, Odalie L. Mamayson², and James E. McDuffie¹

¹Neurocrine, San Diego, California, USA

²Crinetics Pharmaceuticals, San Diego, California, USA

MicroRNAs (miRNAs) regulate gene expression during various pathological processes, potentially making them important biomarkers for early identification of testicular toxicity which remains a challenging issue in drug development. Nitrofurazone (NF) was used as a tool compound to evaluate optimal timepoints for monitoring acute testicular injury. Sprague-Dawley rats (n = 4/group) received a single oral dose of vehicle or NF at 300 mg/kg. NF testicular toxicity was monitored using anatomic pathology, serum inhibin B, and miRNAs at 24 and 96 hours postdose. At 24 hours, NF induced testicular degeneration, nuclear smudging, cytoplasmic hyperchromasia of pachytene spermatocytes, and cytoplasmic microvacuolation in Sertoli cells. At 96 hours, degenerative changes progressed to loss of round and elongated spermatids, variable reduction in Sertoli cells and increased tubular macrovacuoles. In testis, 38 of 816 miRNAs were modulated at 24 and/or 96 hours, correlating well to temporal microscopic testicular lesions caused by NF in rats. Furthermore, serum inhibin B progressively decreased in NF-treated rats when compared to controls at 24 (-0.5x; p < 0.01) and 96 ( < LLOQ) hours. RNA-Sequence profiling identified multiple testis-enriched miRNAs, providing miRNA candidates that warrant further evaluation as potential safety biomarkers when monitoring for acute drug-induced acute testicular injury in rats.

P120: Investigating Relationships between Occupational Jet Fuel Exposure in the Military and the Pursuit and Success Rate of Disability Compensation Claims after Separation

Terra D. Vincent¹, Immanuel Samuel², Felicia Denteh², Abigail Werner³, Aaron Schneiderman⁴, Keith Hancock³, Warren S. Monks⁵, Raul Mirza⁵, and Gregory Wolff⁶

¹Exposure Science, Health Outcomes Military Exposures, Department of Veterans Affairs, Washington, DC, USA

²Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, MD, USA

³Compensation Service, Department of Veterans Affairs, Washington, DC, USA

⁴Epidemiology Program, Health Outcomes Military Exposures, Department of Veterans Affairs, Washington, DC, USA

⁵Occupational and Environmental Medicine Branch, Army Public Health Center, Aberdeen Proving Ground, MD, USA

⁶Epidemiology Consult Service, US Air Force School of Aerospace Medicine, Wright-Patterson Air Force Base, OH, USA

Exposure to jet fuels is a common occupational hazard in military service; however, existing literature exploring the impact of these exposures on long-term health outcomes remains limited. In the present study, data from the Departments of Veterans Affairs (VA) and Defense (DoD) were linked to evaluate the health of Veterans who were exposed to jet fuels during their service. Relationships between Military Occupation Specialty codes and disability compensation claim variables were examined by logistic regression, and interactions with exposure duration and deployment history were assessed, with significance set at an alpha level of p < 0.05. Veterans with probable occupational exposure to jet fuel were more likely to file a claim for compensation than unexposed Veterans, regardless of exposure duration. Likewise, compensation claims were granted at a higher rate for exposed Veterans compared to unexposed Veterans, regardless of exposure duration. Our findings suggest that occupational exposure to jet fuels is associated with compensation claim trends for chronic health outcomes that occur in Veterans, even if their service in fuel-exposed occupations was short (e.g., three years or less). Further, jet-fuel exposed Veterans that filed successful compensation claims were observed to have higher levels of service-connected disability than unexposed Veterans with successful claims. Future studies will incorporate additional data sources, such as healthcare encounters and mortality, with the compensation claims data to provide a better understanding of the impact of occupational jet fuel exposure in this Veteran population.

P121: A Novel PD-L1xCD40 Bispecific Antibody (HBM9027) Demonstrated a Better Nonclinical Safety Profile than Selicrelumab in Humanized Transgenic Mouse and Cynomolgus Monkey

Jimmy Rong¹, Haishan Luo¹, Zihong Meng¹, Yuandong Wang¹, Yunxing Yang¹, Fei Chen¹, Xin Gan¹, Yiping Rong¹, and Victor Chen¹

¹Harbour BioMed, Shanghai, Shanghai, China

CD40 agonistic mAb is emerging as a front-line immunotherapy for cancers. However, serious immune-related adverse reactions and liver toxicity limit its clinical application such as Selicrelumab. We report a novel PD-L1xCD40 bispecific antibody designed to have crosslinking enhanced CD40 activity by using heavy chain only antibody-based building blocks. The optimized format warrants the specific activation of CD40 in PD-L1 positive tumors with a better safety profile. hPDL1×hCD40 knock-in mice were inoculated with MC38-hPDL1 tumor cells and treated with 12 mg/kg HBM9027 or 1, 2.5, 5.0 mg/kg Selicrelumab by intraperitoneal route, BIW for 2 weeks. HBM9027 demonstrated a better nonclinical safety profile than Selicrelumab in cytokine release (TNF-α, IL-6, IFN-γ, and IL12p40), liver enzyme (AST and ALT) and histopathology. Cynomolgus monkeys, a relevant pharmacology species, were administered with HBM9027 at 10/50 or 100/200 mg/kg or Selicrelumab at 5/20 mg/kg weekly for 29 or 36 days. All monkeys survived to the scheduled necropsy without moribund conditions. No HBM9027-related changes were noted in cytokine release, hematology, coagulation, organ weight and histopathology. Selicrelumab-related findings included decreases in lymphocytes, platelet count as well as mildly prolonged APTT and markedly decreased in fibrinogen. Histopathology examinations show HBM9027, up to 200 mg/kg had minimal perivascular mononuclear infiltration in liver. While monkeys treated with Selicrelumab showed significant toxicology responses including neutrophilic infiltration in heart, hyperplasia of adventitia in gallbladder and spleen enlargement. Other noteworthy findings include tubulointerstitial nephritis, microgranuloma, hyperplasia of hepatocytes and oval cells in liver. Overall, HBM9027 demonstrates a better safety profile than Selicrelumab.

P122: In Vitro Interactions of Novel Oxime Acetylcholinesterase Reactivators with Blood Brain Barrier Efflux Transporters P-Glycoprotein and Breast Cancer Resistance Protein

Darby C. Stanford¹, Mary Beth Dail¹, and Janice E. Chambers¹

¹Center for Environmental Health Sciences, College of Veterinary Medicine, Mississippi State University, Mississippi State, Mississippi, USA

Novel phenoxyalkyl pyridinium acetylcholinesterase (AChE) reactivators (US patent 9,227,937) that showed convincing evidence of penetration into the brains of intact rats were developed by our laboratories. The oximes were sorted into two groups (24–35% and 10–17%) based on percent brain AChE reactivation following exposure of rats to the sarin surrogate nitrophenyl isopropyl methylphosphonate (NIMP). ATP-binding cassette (ABC) efflux transporters such as p-glycoprotein (P-gp) and breast cancer resistance protein (BCRP) provide defense along the blood brain barrier. More efficacious oximes were expected to be poor efflux transporter substrates, thus remaining in the brain longer and having more time to reactivate AChE. In a previous in vitro study using rat P-gp, it was found, however, that the 24–35% efficacy oximes had higher ATPase activity (i.e., were better P-gp substrates) than the 10-17% efficacy oximes so the reactivation difference was not explained by P-gp export. Suspecting another transporter may be involved, additional novel oximes and pralidoxime (2-PAM) were screened using in vitro human P-gp and BCRP membrane assays. The amount of Pi from ATP hydrolysis measured in the presence of an oxime is directly proportional to the transporter’s activity. Total ATPase activity was calculated. Oxime 55 was the only oxime transported by BCRP. Human P-gp transport was again variable among oximes. Apparently, the lower efficacy of the 10-17% reactivation oxime group is not due to greater P-gp or BCRP efflux. The explanation may lie with another transporter or an altogether different mechanism. (Supported by NIH U01NS107127)

P123: Development of hTERT Immortalized Neonatal Melanocytes for Toxicity Studies and Melanogenesis Regulation

Diana Douglas¹, Michael Maddox¹, Joshua Franklin¹, Kevin Grady¹, and Weiguo Shu¹

¹ATCC Cell Systems, Gaithersburg, MD, USA

Melanocytes are the cells that produce melanin and they are involved in the process of epidermal pigmentation, which involves two main steps. First, melanocytes perform a variety of complex biochemical and physiological steps to produce, package, and exocytose melanin-containing melanosomes. Secondly, the melanosomes are taken up by neighboring keratinocytes where the stored melanin protects the underlying tissues from damaging UV radiation. In additional to playing an integral role in skin pigmentation, the dysfunction of melanocytes is associated with skin disease and skin cancer. Therefore, it is important to have a reliable model system to study melanocytes and associated processes. Primary cells offer one model system to study pigmentation and dermal agents that may disrupt the melanogenesis; however, they are limited by lifespan, and donor-to-donor variability. Here, we created an immortalized melanocyte cell model—hTERT neonatal melanocytes—by retroviral transduction of human telomerase (hTERT) into primary cells. In addition to enhanced longevity (up to 35 doublings), physiologic marker expression (tyrosinase positive, fibroblast marker negative), and ability to create melanosomes in 3D organotypic co-cultures, the cell line also showed expected levels of responses to several stimulators and inhibitors of melanogenesis. In summary, immortalized melanocytes provide a versatile in vitro cell model for the study of skin toxicology and melanogenesis regulation.

P124: Pharmacokinetic Profile of anti-TfR1 Antibody-Oligonucleotide Conjugates (AOC) Developed for the Treatment of Rare Muscle Disease

Laura Leung¹, Philip Kovach¹, Eileen Blasi¹, Maral Llewellyn¹, Aaron Anderson¹, Karla Schramm¹, Aidan Kim¹, Aaron Yu¹, and Husam Younis¹

¹Avidity Biosciences, San Diego, CA, USA

Antibody oligonucleotide conjugate (AOC) therapeutics combine the tissue specificity of monoclonal antibodies with the precision and potency of oligonucleotides to enable the targeted delivery of oligonucleotides to previously untreatable cell types. Using the transferrin receptor 1 (TfR1) antibody to target muscle tissue, we developed AOCs conjugated to siRNA or phosphorodiamidate morpholino oligomers (PMOs) to advance treatment options for myotonic dystrophy type 1 (DM1) and Duchenne muscular dystrophy (DMD), respectively. The objective of this research was to characterize the pharmacokinetics of these anti-TfR1 AOCs in plasma and tissues in the cynomolgus monkey. Bioanalytical assays were developed to determine concentrations of the intact AOC molecule and its individual components consisting of the monoclonal antibody and the oligonucleotide. Following intravenous (IV) administration of AOC in cynomolgus monkeys, plasma exposure based on C_max and AUC increased in a dose-dependent manner that correlated with muscle tissue concentration. Detection of the intact AOC, monoclonal antibody, or the oligonucleotide yielded similar results, suggesting that the AOC was predominantly intact in systemic circulation. Distribution to muscle was antibody-dependent given the consistency in tissue exposure across the siRNA and PMO oligonucleotides, with a tissue half-life of approximately 10 days. Importantly, post-dose changes in pharmacologic activity correlated in a dose-dependent manner with muscle oligonucleotide exposure. No pharmacologic activity was observed in non-muscle tissues despite broad tissue biodistribution, confirming targeted delivery using our AOC platform. Our data herein demonstrates several robust and sensitive bioanalytical assays that support the pharmacokinetic characterization of AOCs.

P125: Evaluation of Cannabidiol (CBD) and Cannabinol (CBN) Toxicity in the Human Quad-Culture Liver-Chip

Jacob D. Larson¹, Sushma Jadalannagari², Jake T. Chaff², James Velez², Josiah Sliz², Lorna Ewart³, and Donna Webster¹, Jiang Hu¹

¹Herbalife Nutrition, Torrance, CA, USA

²Emulate Inc., Boston, MA, USA

³Emulate, Inc., Boston, MA, USA

As cannabinoid use expands there is a need to evaluate for toxicity potential. Considering the rising need to reduce animal use in toxicology testing, this study was conducted to compare CBD and CBN hepatotoxicity using the Human-Quad Culture Liver-Chip, consisting of upper and lower channels seeded with primary human hepatocytes and non-parenchymal cells, respectively. Dosing concentrations were adjusted to account for compound distribution in the Chip (∼80% for CBD and 70% for CBN), which was assessed using a specially designed assay to achieve target dosing concentrations of 0.24, 3, or 4.7µM CBD and CBN. Both channels of Liver-Chips were continuously dosed for 7 days with CBD or CBN. Effluent samples were collected and brightfield images were taken on days 1, 3, 5, and 7. CBN showed greater cytotoxicity to primary human hepatocytes through a decline in healthy morphology compared to CBD. No major changes were observed in albumin, ALT and AST release for CBD and CBN relative to the control. At 4.7µM, CBD significantly increased IL-6, IL-8 and MCP-1 cytokine and LDH levels in top-channel effluents on day 7. No changes in IL-6 and MCP-1 were observed for CBN, a decrease of IL-8 was observed at 3 and 4.7µM CBN on day 7, and an increase of LDH at 4.7µM on day 1 and 0.24µM on day 5. These results indicated differences in CBD and CBN toxicity profiles while demonstrating the sensitivity and specificity of the Liver-Chip model as an alternative tool for liver toxicity screening.

P126: Withdrawn

P127: NADPH Oxidase Inhibitor Diphenyleneiodonium Prevents Arsenic-Induced Downregulation of the ATP Binding Cassette Transporter G1 in Mouse Aortic Endothelial Cells

Ekhtear Hossain¹, Homyra Tasnim², Bryan T. Rogers¹, Raphyel Rosby¹, Mary Beals¹, and Oswald D'Auvergne¹

¹Department of Biological Sciences and Chemistry, Southern University and A&M College, Baton Rouge, Louisiana, USA

²Department of Biological Sciences, Louisiana State University, Baton Rouge, Louisiana, USA

Arsenic is a naturally occurring toxic element present in the air, food, soil, and water in several countries, including the U.S. The U.S. Geological Survey (USGS) reported that over 2 million Americans are estimated to be drinking from wells with high concentrations (>10 ug/L) of arsenic. Epidemiological studies reported that chronic arsenic exposure is a well-known risk factor for cardiovascular diseases, including atherosclerosis. However, the cellular and molecular mechanism underlying arsenic-induced atherosclerosis remains obscure. ABCG1, an ATP-binding cassette transporter, highly expressed by vascular endothelial cells, promotes cholesterol efflux to high-density lipoprotein (HDL) particles in reverse cholesterol transport (RCT). Earlier research indicates that the increased atherosclerosis lesions were attributable at least in part due to endothelial deficiency of ABCG1. To delineate the mechanism of arsenic-induced atherosclerosis, we first examined whether sodium arsenite (SA) affects ABCG1 mRNA expressions in mouse aortic endothelial cells (END-D) using qRT-PCR. Interestingly, ABCG1 mRNA expressions were found to decrease in a time-dependent manner after SA exposure. We next examined whether SA-induced oxidative stress was implicated in the downregulation of ABCG1 mRNA expressions. We found that both antioxidants, n-acetyl cysteine (NAC) and diphenyleneiodonium (DPI), effectively prevented the downregulation of SA-induced ABCG1 levels. All experimental evidence so far obtained suggests that arsenic-induced oxidative stress is implicated in the downregulation of ABCG1 in endothelial cells, thus raising the possibility that arsenic-induced decrease of ABCG1 may contribute to the pathogenesis of atherosclerosis. Additional in vivo and in vitro studies are warranted to gain a deeper understanding of these effects.

P128: Survey of Rodent Motor Disturbances on Preclinical Studies

Tim Streit ¹

¹Labcorp Early Development Laboratories Inc., Madison, WI, USA

Rodent motor disturbances observed on preclinical toxicology and carcinogenicity studies may lead to challenging assessments when determining if observations are drug-related or background. A common tool to support these conclusions is to compare the frequency and incidence of a treatment group finding to a historical review of similar data collected from control animals. A review of historical data across 10 years of rodent preclinical studies at the Labcorp-Madison site (2011 to 2021) including more than fifty 2-year carcinogenicity studies of various strains of rodent was conducted to survey the frequency and incidence of various motor disturbances to provide a reliable source on the prevalence in recent studies. The majority of motor disturbances in chronic studies were described as tonic convulsions, clonic convulsions, or myoclonic jerking (∼87%), although other terms such as tremors, twitching, rigid stance, circling or barrel rolling were also documented. A slight sex difference was appreciated with males accounting for 55% of the observations compared to females. Overall incidence across studies was variable ranging from 0 to 16% of control animals that had at least one motor disturbance observation during the course of the study. Additional assessments such as strain comparisons are pending. It is clear that motor activity disturbances are a common, but variable background finding in subchronic and chronic preclinical studies. Low frequency or sporadic observations drug-treated group should be expected on long-terms studies, and consistent terminology is necessary to ensure the proper conclusions are ultimately made.

P129: Pharmacokinetic Evaluation of Colchicine and Hydrochlorothiazide in Sinclair and Gottingen Miniature Swine

Guy F. Bouchard¹, Miri Pannu¹, Tina Rogers¹, Yafei Chen¹, and Derek F. Brocksmith²

¹Sinclair Research, An Altasciences Company, Auxvasse, Missouri, USA ²Sinclair Bio Resources, LLC, Auxvasse, Missouri, USA

The objective of this study was to determine and compare plasma pharmacokinetic profiles of two compounds with very different elimination pathways, colchicine and hydrochlorothiazide, in Sinclair and Gottingen minipig. Hydrochlorothiazide not metabolized, mostly excreted in urine. Three males and three females Sinclair or Gottingen minipig were dosed intravenously with colchicine (20µg/kgIV;Day1) and orally administered hydrochlorothiazide (HCTZ;25mg/kgPO;Day4) in a cross over design. Serial blood samples were collected (colchicine: Predose,5min,15min,30min,1,2,4,7,24hr post; HCTZ: Predose,30 min,1,2,4,7, and 24hr post) and non-compartmental pharmacokinetics analysis was performed (Phoenix WinNonlin8.2; Cmax, Tmax, t1/2, AUClast and AUC0-∞). All animals demonstrated quantifiable plasma colchicine and HCTZ concentrations throughout the entire 24-hour blood sampling period. Colchicine had similar exposure profile between Sinclair and Gottingen (Cmax:51,290 vs 47,620pg/mL in males and 57,920 vs 53,550 pg/mL in females; t1/2:12.9-15.1 vs 12.8-14.3 hr;AUClast: 35,440 vs 26,180 hr*pg/mL in males and 33,370 vs 27,010 hr*pg/mL in females). Sinclair showed more consistent and similar HCTZ data between males and females than Gottingen (Cmax: 910.7 vs 818.4 ng/mL in males and 879.0 vs 474.3 ng/mL in females; t1/2: 3.8-13.0 vs 6.5-12.1 hr; AUClast: 9,109 vs 8,416 hr*ng/mL in males and 5,672 vs 4,831 hr*ng/mL in females; median Tmax:4.0 vs 2.0hr in males and 2.0 vs 24hr in females). In this study, we have successfully demonstrated the utility of minipig to determine the systemic exposure of two compounds with very different metabolism and elimination pathways. Both strains of minipig show comparable PK parameters of colchicine, with more consistent exposure data of HCTZ in Sinclair minipig.

P130: Genetic Selection and Nutritional Management for a Downsized Sinclair Miniature Swine

Derek F. Brocksmith¹, Guy F. Bouchard¹, Miri Pannu², and Bill Miller¹

¹Sinclair Bio Resources, LLC, Auxvasse, Missouri, USA ²Sinclair Research, An Altasciences Company, Auxvasse, Missouri, USA

With a larger size test animal, more test material is required to complete the toxicology study thus increasing cost of the study. The objective of this study was to utilize genomic selection, phenotype, body weight, and nutritional management to reduce the total body mass in adult Sinclair S-1 MS. A large baseline single nucleotide polymorphism (SNP) bank was established for each breeder of the Sinclair MS colony. The SNPs were then used to calculate genomic estimated breeding values for body weight of the breeding animals. Finally, a computer algorithm manipulates a large dataset of phenotypic and genomic data to determine the optimal mating combinations to result in a reduction in adult body mass. Utilizing these selection criteria, we have observed an 8% decline in body size per year in the Sinclair S-1 MS. After 10 years of this genetic selection, the body weight reduction outcome of the 6^th generation at 6 and 12months of age, is a significant (p < 0.001) reduction in body weight in both (34.5%) and (45.8%) respectively. Further concurrent studies are being conducted to standardize the clinical pathology, biometric and reproductive variables, as well as validation studies for compound kinetics, dermal irritation, and feeding regimen assessment. We have developed a genomic selection tool based on body mass and ADG to down size the Sinclair S-1 MS over the course of the next several generations.

P131: Comparison of Respiratory Parameters in Beagle Dog and Cynomolgus Monkeys of Stand-alone Safety Pharmacology and Repeat Dose Toxicology Studies

Sue McPherson¹, Rui Wu¹, An Shanshan¹, Yangming Chen¹, and Xinru Ji¹

¹WuXi AppTec, Cranbury, New Jersey, USA

The dog and primate are considered to be valid species to non-rodent species for cardiovascular (CV) and respiratory safety pharmacology studies. The respiratory parameters are evaluated in stand-alone CV safety pharmacology studies following single doses, for new chemical entity. For biologics and cancer drugs, incorporating respiratory system endpoints into repeat-dose toxicology studies (combined-studies) is encouraged by scientific justification and regulatory requirements. The respiratory data collection with a face mask (minimum 10 minutes/time point) in conscious dogs or monkeys are regularly used in stand-alone studies or combined-studies utilizing the DSI/Ponemah Physiology Platform devices. To establish background data, a review on historical data including respiratory rate, tidal volume and derived minute volume, was performed for IND-enabling toxicity and safety pharmacology studies for dogs and monkeys. When comparing the respiratory parameters data in stand-alone studies, lower tidal volume in both sexes (-31% in males and -28% in females), with lower bodyweight in males and females (-29% and -19%), and lower derived minute volume (-31%) in males were noted in the monkeys for the combined-studies. For the dogs, lower tidal volume in females (-16%) correlating with lower bodyweight in females (-22%) were seen in the combined-studies when compared to the data in stand-alone studies. Other parameters were comparable for those two types of studies in monkeys and dogs. Incorporating respiratory endpoints into repeat-dose toxicology studies can provide reliable respiratory data as stand-alone safety pharmacology studies in dogs and monkeys and is considered to be an appropriate option for preclinical respiratory assessment.

200s—Regulatory Toxicology

P201: Setting Impurity Limits for Endogenous Substances: Recommendations for a Harmonized Procedure and an Example using Fatty Acids

Ernest Fung¹, Suren Bandara², Joel Bercu³, Courtney Callis⁴, Dave Dolan⁵, Jessica Graham⁶, Diem HaMai⁶, Ester Lovsin Barle⁷, Andrew Maier⁸, Melisa Masuda-herrera³, Chandrika Moudgal⁶, Jillian Parker¹, John Reichard⁹, Reena Sandhu¹⁰, and Robert Jolly⁴

¹Cardno ChemRisk, Aliso Viejo, CA, USA

²Boehringer Ingelheim, San Francisco, CA, USA

³Gilead Sciences, Foster City, CA, USA

⁴Eli Lilly & Company, Indianapolis, IN, USA

⁵Amgen, Thousand Oaks, CA, USA

⁶Genentech, South San Francisco, CA, USA

⁷Takeda, Zurich, Switzerland, Switzerland

⁸Cardno ChemRisk, Cincinnati, OH, USA

⁹University of Cincinnati, Cincinnati, OH, USA

¹⁰SafeDose Ltd, Toronto, Ontario, Canada

Endogenous substances, such as fatty acids (FAs), are often purposefully used in parenterally administered pharmaceuticals, but may also be present as impurities. Currently, no consensus guidance exists on how to set impurity limits for endogenous substances. Specific procedures are needed, as the toxicity data (amount/type) available for endogenous substances typically differs from data available for other types of impurities. This presents a challenge for setting health-based exposure limits (HBELs) to protect patient health. An additional complicating factor is that these impurities are often present in parenterally delivered products, whereas the available data for endogenous substances is typically oral. We propose as a systematic model to facilitate the development of HBELs for endogenous substances, which we call Risk Assessment Process Maps (RAPMAPs), that accounts for these differences. The RAPMAP approach yields a harmonized risk assessment framework that was applied to derive HBELs for eight FAs commonly used in parenteral drug products. This tiered approach was used to derive HBELs by evaluating different lines of evidence with further vetting based on anticipated perturbations in ranges of physiological serum levels, impacts of dose-rate, and consideration of intermittent dosing. Parenteral HBELs of 100 to 500 mg/day were generated for these eight FAs and a class-based limit of 50 mg/day was proposed in the absence of chemical-specific data. This default limit is consistent with the low toxicity of this chemical class and the ICH Q3C value for Class 3 (low toxic potential) solvents.

P202: Effective Waiver Strategies for Environmental Risk Assessment of Pharmaceuticals in the European Union

Francis EH. Crawley¹ and David B. Mayfield²

¹Labcorp, Eye, Suffolk, United Kingdom ²Labcorp, Greenfield, Indiana, USA

An environmental risk assessment (ERA) is mandatory for marketing authorisation of pharmaceuticals in the European Union (EU). If the predicted environmental concentration of the pharmaceutical in surface water (PECsw) is >0.01 µg/L, then a costly program of physico-chemical, environmental fate and effects testing is required. In this review, we explore strategies for reducing the need for laboratory testing while maintaining environmental safety. Various methods are available for lowering the estimated PECsw by refining the prevalence of the disease to be treated using reliable epidemiological data and by considering the annual dosage regimen. For generic drugs, it may be possible to demonstrate negligible increases in environmental exposure once marketed in the EU. For established drugs, for which a new authorisation is required, and no ERA is available, there may be sufficient information in the public domain to assess the environmental risk. Such studies are unlikely to be compliant with Good Laboratory Practice (GLP) or following OECD Guidelines. Therefore, a critical science-based assessment is made on the reliability and relevance of these studies. A reliability category is assigned according to available assessment methods and a decision made on their acceptability for an ERA. This approach will be presented for ibuprofen, a universally marketed anti-inflammatory drug. The options explored herein may allow testing to be waived, while also providing necessary information to complete the mandatory ERA and gain regulatory approval.

P203: A Retrospective Evaluation of Non GLP Exploratory Dog Study Designs and their Robustness in Predicting Toxicity for Planning First-In-Human Studies

Shubhra Chaudhuri¹ and Lorenz Kempeneers²

¹Preclinical Safety and Translational Sciences, Janssen Research and Development, LLC, Spring House, PA, USA

²Preclinical Safety and Translational Sciences, Janssen Research and Development, LLC, Beerse, Beerse, Belgium

The objective of this evaluation was to assess past/current study designs used at Janssen for conducting repeat dose nGLP toxicity studies in dogs and their predictability towards designing “first-in-human” (FIH) enabling GLP studies. This evaluation aims to provide a recommendation on a refined study design in terms of animals used (3Rs) while addressing scientific goals. 11 study pairs (FIH and preceding nGLP) were identified for this exercise. GLP study designs were noted to be consistent; but variations existed in nGLP study designs with respect to number of animals used per group, dosing duration, inclusion of a vehicle group or not and including single vs. both sexes. Toxicology end points were compared across these study pairs to identify parameters that contributed to suboptimal predictions. Dosing duration of 2 weeks seemed critical and shorter duration led to mispredictions of tolerability on two occasions. A minimum of 2 dogs per group was considered necessary for these studies while inclusion of a vehicle group did not provide additional scientific value for standardly-used vehicles. Systemic exposure differences between sexes was noted only on one occasion suggesting inclusion of both sexes to be of limited value generally (except prior knowledge of differential exposure). Overall, while this was a limited dataset, -this evaluation was still sufficient for a review of the different variations in nGLP study designs and recommend an optimum study design for nGLP studies with additions on a case-by-case scientifically driven rationale.

P204: A Statistically-Robust Method for Nitrosamine Potency Assessment

David J. Ponting¹, Rachael E. Tennant¹, and Rob Thomas¹

¹Lhasa Limited, Leeds, West Yorkshire, United Kingdom

The recent crisis surrounding nitrosamine impurities in a wide variety of pharmaceuticals has led to a requirement for the urgent evaluation of many different N-nitroso compounds for potential mutagenic hazard and carcinogenic risk. Since the dataset available for both mutagenicity and carcinogenicity is relatively limited in size and chemical diversity, yet observed carcinogenic potencies cover a range of four orders of magnitude [1], conventional (Q)SAR methods have not thus far proven sufficient for evaluating which structural features are significant. We present a method using Bayesian multiple-linear-regression on a set of expert-determined structural features [2] to evaluate which of these features are truly statistically significant, such as methyl groups increasing potency and isopropyl groups decreasing it. Blinded comparison of the statistical results with expert knowledge indicates excellent concordance. The lists of significant features thus derived can be used to accurately categorise nitrosamine-containing compounds into binned potency categories (at the order of magnitude level), which can be used in several ways. Firstly, this can indicate which compounds within a portfolio are most concerning and should be prioritised for investigation. Secondly, having a list of statistically-significant features can be used to support the selection of read-across analogues for compounds with insufficient carcinogenicity data to determine an acceptable daily intake. Thirdly, absent a suitable read-across analogue, the lower bound of each category could be used as a proposed limit. 1: Thresher et al (2020), Regul Toxicol Pharmacol, 116, 104749. 2: Cross and Ponting (2021), Comput Toxicol, 20, 100186

P205: MRI Guided Intraparenchymal Gene Therapy: Dosing Strategies, Historical Data and Image Analysis

Simon Authier¹, Nehla Bennamoune¹, Morgan K. Foret¹, Erik Larson¹, Clotilde Lecrux-Leblond¹, Archival Santos Galicia¹, Raluca Kubaszky², and Mohamed Said Maghezzi¹

¹Charles River Laboratories, Laval, Quebec, Canada ²Otsuka, Rockville, Maryland, USA

Convection enhanced delivery (CED) enables targeted therapies to neuronal structures for central nervous system diseases. Real-time iMRI using stereotactic drug delivery was developed in non-human primates (NHPs) for gene therapy safety/efficacy and biodistribution studies. Cynomolgus monkeys were dosed with control or therapeutics to putamen (n = 140), globus pallidus (n = 56), thalamus (n = 74), caudate nuclei (n = 54) and/or cerebellar nuclei (n = 24). Dose volumes ranged from 50 to 200 μL at 120 to 300 μL/hr using a clinical cannula (Clearpoint). Procedure related clinical signs included superficial changes at the catheter insertion site, signs of emesis, decreased appetence, and mild physiological tremors lasting up to 7 days post-surgery. Procedure-related histopathological results showed mild local changes along the cannula tract, as expected for this type of procedure. Overall, real-time iMRI-guided CED was safe in this preclinical toxicology model, enabling the use of clinical methods and devices during preclinical development. MRI images at completion of the drug delivery were analyzed with a reference 7T MRI Cynomolgus monkey atlas and targeted structures were segmented for volume quantification; Putamen (628.3mm3), Globus Pallidus (181.9mm3) and Thalamus (695.4mm3). The second part of the analysis delineated Gadolinium-contrast in T1-weighted scans. The manual segmentation was performed for each slice; e.g. (putamen, 15 slices at 1mm spacing). Precise brain structure coverage quantification by drug formulation is of particular relevance when aiming to assess drug effect within a targeted brain region. The methodology for target coverage quantification can contribute to the interpretation of neuropathology and biodistribution data from individual animals.

P206: Removed

P207: Using BMD Analysis to Identify Strain Selectivity in the Ames Assay

Kevin Cross¹ and George Johnson²

¹Instem, Columbus, Ohio, USA

²Swansea University Medical School, Swansea, Whales, United Kingdom

Investigations to select the most sensitive Ames assay experimental conditions for detecting N-nitrosamines have suggested using benchmark dose (BMD) analysis of mutagenic potency to identify strain selectivity (Heflich GTA 2022). At the 2017 International Workshop on Genotoxicity Testing (IWGT), Salmonella TA1535 and TA100 strain selectivity was determined (Williams Mutation Research 2019) using statistical analysis (>10,000 compounds). Linear regressions calculated mutagenic potencies over the linear portion of dose response curves. Of 1,151 compounds tested in both TA1535 and TA100, 2% (27/1,151) were determined as TA1535 positive-only by authors, with 10 confirmed by statistical analysis (9 low-potency mutagens < 0.05 rev/ug). This investigation employed PROAST BMD analysis using a critical effect size (CES) of 50% to determine if BMD models identified comparable strain selectivity using the same dataset of low potency mutagens. BMD analysis has the advantage of non-linear model fitting of all dose response data, is less dependent on dose spacing, number of doses measured and uses covariate analysis to increase precision. Overlap of BMD uncertainty ranges between the two strains was used to determine selectivity. All IWGT strain-selective compounds were identified via BMD analysis, and some previously identified with similar responses were now found to be strain-selective. CES selection is critical in defining a selective response, while convergence and goodness-of-fit also must be considered. Additional investigations will consider other CES values (200%, 300%) and automatic CES generation supporting a data-driven analysis. While here eight different compound classes were analyzed, the BMD approach should also be appropriate for analysis of nitrosamines.

300s—Safety Evaluation Nonpharmaceuticals

P301: Comparative In Vitro Assessment of Heated Tobacco Aerosols to Combustible Cigarettes

Matthew Stevenson¹, Sarah Jean Pour², Edgar Trelles-Sticken², Fiona Chapman¹, Liam Simms¹, and Roman Wieczorek²

¹Imperial Brands PLC, Bristol, Avon, United Kingdom ²Reemtsma Cigarettenfabriken GmbH, An Imperial Brands PLC Company, Hamburg, Hamburg, Germany

Heated tobacco products (HTP) are an expanding category of non-combustible tobacco products, which are designed to provide adults smokers a potentially reduced harm alternative to combustible cigarettes. In the current study we assessed the aerosol from two HTPs (2 stick variants each) and compared the biological response to 1R6F reference cigarette smoke in the Ames, In Vitro Micronucleus (IVM) and Neutral Red Uptake (NRU) assays. Cells were exposed to smoke or aerosol at the air liquid interface using an internal puffing machine ‘smoke aerosol exposure in vitro system’ (Burghart Tabaktechnik, Germany) for IVM and NRU. For the treatment of bacteria suspensions with fresh smoke/ aerosol from the test product, a Smoking Robot VC10® S-Type (Vitrocell Systems GmbH, Germany) was used. The Ames test was performed in compliance with OECD 471 (TA98 &TA100 ± S9) with the IVM performed in compliance with OECD 487 (V79 cells ± S9). The NRU cytotoxicity assay used BEAS-2B cells, and followed standard assay protocols in accordance with ISO 17025. There were clear cytotoxic (EC50: 0.203 puffs), mutagenic and genotoxic (ECMN3: 0.7-0.8 puffs) effects observed for 1R6F cigarette smoke. Whilst, for the HTPs a marked reduction in cytotoxicity, mutagenic and genotoxicity was observed (10 – 56-fold reduction compared to cigarettes), under the conditions of the tests. The presented results show that HTP emissions are less toxicologically active than combustible cigarettes, and are consistent with the available published literature reporting reduced in vitro/in vivo toxicity of a wide variety of HTPs when compared to cigarette smoke.

P302: North American Travel Grant Recipient: Carnosine Treated Mice Exhibit DNA Methylation Changes in Parkinson’s Disease-Related Genes and Biological Processes

JT Toebbe¹ and Mary Beth Genter¹

¹University of Cincinnati, Cincinnati, Ohio, USA

Parkinson’s disease (PD) is an increasingly prevalent disease in our society. A single cause of PD has not yet been elucidated, but the most supported culprit is the aggregation of α-synuclein (aSyn), a protein found throughout the nervous system. Aggregation results from oxidative stress and post-translational modifications, such as phosphorylation of Ser129. Previous findings in our lab revealed that intranasally-administered (IN) carnosine mitigated motor dysfunction and decreased aSyn accumulation in the olfactory epithelium (OE) and substantia nigra in mice. To understand the basis for these beneficial effects, specifically reduced aSyn aggregation, reduced representation bisulfite sequencing (RRBS) DNA methylation analysis was performed to identify epigenetic changes resulting from IN carnosine treatment in mouse OE. Methylation status of the aSyn promoter was unaffected by carnosine, indicating that decreased accumulation is not a result of hypermethylation and subsequent silencing of the aSyn gene. Other genes related to PD with differentially methylated promoters included discoidin domain receptor tyrosine kinase 1 (Ddr1), biglycan (Bgn), and G-protein coupled receptor 4 (Gpr4). The Ddr1 promoter was hypermethylated, while Bgn and Gpr4 promoters were hypomethylated after carnosine treatment. Functional enrichment analysis using ToppGene showed that several neuro-based biological processes were repressed following IN carnosine treatment, such as neuron development and neurogenesis. Genes associated with negatively regulating these processes, including Bmp7, Foxa2, and RhoA, are hypermethylated and presumably reduced in expression, supporting the potential epigenetic benefit that carnosine treatment contributes to improving neurological function.

P303: International Travel Grant Recipient: Evaluation of Bioactive Metabolites of Indigenous Plants Used as Birth Control

Jackson M. Kenya Njoka¹

¹Jomo Kenyatta University of Agriculture and Technology, Nairobi, Juja, Kiambu County, Kenya

Modern contraceptives have been associated with unwanted mild to severe adverse side effects. Unwanted pregnancies and abortions have also been reported mainly in the developing nations. These negative effects such as hormonal imbalance due to combined oral contraceptives has led to alternative such for safer birth control such as natural products. Medicinal plants have long been in use in regulating the number of children and spacing in various cultures of the world way before the emergence of the modern conventional medicine. However, efficacy of these extracts or decoctions as well as their safety has not been verified or proven scientifically. Under this background, the current study involves phytochemical screening, isolation and characterization of secondary metabolites of Kenyan plants used to control births in humans. Safety, antifertility and reversibility investigations are undertaken in rat model after obtaining ethical clearance approval on animal use in research at the Kenya Medical Research Institute, KEMRI. Preliminary sub-acute toxicity studies show a dose-related activity with high dosages found to be toxic. At a dose-rate of 800 mg/kg body weight, the lungs and liver have been observed to have tumor-like growths in the current on-going toxicity evaluation. This supports the already known indigenous practices on the use of the particular plants with cautions. At a lower dosage the extract is found to be safe with some potentiality in preventing conception in rats, an observation that already validate the claimed efficacy in controlling birth in female dogs using the extract of the plant extract under investigation.

P304: North American Travel Grant Recipient: Potential Postnatal Neurotoxicity Mediated by Maternal E-Cigarette Exposure

Sabrina Rahman Archie¹, Ali Ehsan Sifat¹, Heidi Villalba¹, Sejal Sharma¹, Yong Zhang¹, and Thomas Abbruscato¹

¹Texas Tech University Health Sciences Center, Amarillo, TX, USA

Despite the prevalence perception about electronic cigarette (e-Cig) as a safe alternative of smoking for pregnant women, growing concern related to its potential toxic impact on neonatal health warrants adequate investigation. Due to the noticeable growth in e-Cig usage during pregnancy (∼15%), it has become critical to study the long-term impact of maternal e-Cig smoking on postnatal health outcomes. Hence, in this study, we have evaluated the consequences of prenatal e-Cig use on postnatal blood-brain barrier (BBB) integrity, neuro-inflammation, and ischemic stroke outcomes. Pregnant CD1 mice (E5) were exposed to e‐Cig (2.4% nicotine) till postnatal day (PD) 7. The expression level of structural elements of the BBB including, tight junction proteins (ZO-1, claudin-5, occludin), astrocyte (GFAP), pericyte (PDGFRβ), basement membrane (Laminin α1, Laminin α4), NeuN, AQP4 and GLUT-1 were analyzed in offspring using western blot and immunohistochemistry at PD 7, PD 23, PD 45 and PD 90. Relevant neuro-inflammatory cytokines were quantified in postnatal brain. Ischemic stroke outcomes were evaluated by middle cerebral artery occlusion (MCAO) method. In our study, significantly reduced expression of tight junction proteins and astrocyte marker were observed in male and female offspring till PD 90 (P < 0.05). Moreover, higher level of cytokines was found in offspring brain at PD 7. Additionally, prenatally e-Cig exposed female adult offspring had worsened ischemic stroke injury and neurological score compared to control (P < 0.05). Our findings suggest that prenatal e-Cig exposure induces long-term neurotoxic effects on neonates by disrupting BBB integrity, inducing neuro-inflammation, and worsening ischemic stroke outcomes.

P305: The Effects of a Single or Bilateral Intramuscular Vaccine Injection on IgG Binding and Pseudovirus Neutralization (pVNT) Antibody Titers in BALB/c Female Mice

Deborah McCoy¹, Leonard Azzarano², Isabelle Carletti³, Douglas Donahue⁴, and Corey Mallett⁴

¹GSK, Collegeville, PA, USA

²GSK, Collegeville, Pennsylvania, USA

³GSK, Rixensart, Belgium, Belgium

⁴GSK, Rockville, Maryland, USA

Previous preclinical in-vivo imaging studies using Magnetic Resonance Imaging showed that a volume larger than 25µL delivered intramuscularly to mice would leak into the subcutaneous space leading to the potential to confound data interpretation. In GSK vaccine immunogenicity studies, a 50µL dose volume is utilized to deliver vaccines for immunogenicity assessment. To determine if subcutaneous leakage affected vaccine serum titers, we assessed the effect of a single 50µL dose in one hindlimb compared to two doses (25µL /hindlimb) on levels of serum IgG binding antibody titers and pVNT titers. Two groups of female BALB/c mice (n = 4/group, 7-8 weeks old) were treated with the same mRNA vaccine with either a single IM injection of 50µL in one hindlimb or two IM injections of 25 µL in the right and left hindlimbs. A concurrent control group (n = 4, 7–8 weeks old) was administered physiological saline as two 25µL injections. The mRNA vaccine was administered on days 1 (prime dose) and 21 (booster dose). Serum was collected on days 14, 21 (pre-booster injections), and 35 to determine circulating antibody levels. IM injections were tolerated in both groups with no noticeable effects. Data appears to indicate that no appreciable differences in IgG binding antibodies or pVNT were observed between groups after first dose; but a second, more robust study will be conducted. Therefore, a single IM injection of 50 µL may be an acceptable choice for vaccine administration, which could reduce potential pain or distress to the animal and increase efficiency by reducing animal manipulation.

P306: A Dermal Sensitization Threshold (DST) for Device Extractables

Rebecca Ticknor¹, Isaac Mohar¹, Joel Cohen², and Tom Lewandowski¹

¹Gradient, Seattle, WA, USA ²Gradient, Boston, MA, USA

Evaluation of skin sensitization risk is required for essentially all medical devices and is traditionally evaluated using the guinea pig maximization test (GPMT) on device extracts (polar and non-polar). In an effort to reduce animal testing, multiple dermal sensitization thresholds (DSTs) have been proposed including 5 μg/day (PQRI, 2013), 0.8 μg/cm² (SCCS, 2012), and 0.55 μg/cm² (Safford, 2008). However, each of these DSTs have been developed using databases of chemicals that may not reflect those extracted from devices. In this poster, we use toxicology-based acceptable exposure limits (AELs) for chemicals identified as skin sensitizers through polar and semi-polar extractions of consumer products and devices to derive a DST that can be used to screen device extracts. AELs were derived using chemical-specific experimental data and applying assessment factors to account for interindividual variability, use, exposure conditions, and other uncertainties. Based on an analysis of AELs for 375 organic skin sensitizers, a DST of 4.2 μg/cm² was derived. This DST would be expected to be protective of 95% of extractable skin sensitizers in a device or consumer product but may not be protective against strong skin sensitizers such as certain isocyanates or acrylates. For devices where materials may contain strong sensitizers, a more conservative DST of 0.43 μg/cm² was derived. In lieu of animal testing, the proposed DSTs are data-derived thresholds that can be used for screening device extracted chemicals and will be particularly useful for chemicals that lack experimental data.

400s—Toxicology Methods

P401: A Case Study Comparison of Two Pre-Clinical Studies Utilizing Traditional Versus Modified Methods for Surgical Closing and Post-Operative Treatment for Muscle Biopsy Sites in Nonhuman Primates

Katherine Irby¹, David Benedict¹, Brett Megrath¹, Julie Forget¹, Andrew Gorman, and DVM, DACLAM¹

¹Altasciences Preclinical Seattle LLC, Seattle, Washington, USA

Muscle biopsy in nonhuman primates is a specialized collection procedure for samples that can be utilized to diagnose or monitor certain diseases and/or assess biodistribution in tissue. The procedure carries an inherent risk of postoperative complications including dehiscence and infections due to animal movement, grooming, and environmental conditions of group housing. The traditional method utilizes absorbable sutures to close muscle and skin, with optional surgical adhesive for skin. Standard postoperative care included administration of a non-steroidal anti-inflammatory, limited social contact, daily surgical site observations, and potential antibiotic administration if indications of postoperative infection developed. However, evolving study designs including multiple sample collections and housing animals in corrals compared to cages led to a higher rate of post-surgical complications and introduced confounding variables to toxicology studies. Modifications were made to the post-surgical procedures to reduce the duration of single housing, allow the surgical site to withstand greater activity levels (including grooming), and reduce postoperative complications. In the modified method, the muscle was closed using absorbable sutures and the skin was closed using absorbable sutures, surgical staples, and optional surgical adhesive to reduce dehiscence. After completion of the procedure, a prophylactic, long-acting antibiotic and non-steroidal anti-inflammatory were uniformly administered. With these modifications, animals could be returned to social housing immediately and postoperative complications including dehiscence, discharge, and site swelling decreased from 7% of biopsy sites to less than 1% of biopsy sites. Moreover, it eliminated the need for revision surgery and reduced confounding variables with inconsistent antibiotic administration.

P402: North American Travel Grant Recipient: Unbiased Approach for the Identification of Disease Sensitive to Multi-Chemical Exposures- Using a Computational Approach

Olatunbosun Arowolo¹ and Alexander Suvorov¹

¹Department of Environmental Health Sciences, University of Massachusetts, Amherst, Massachusetts, USA

Chemical exposures from diverse sources merge on a limited number of molecular pathways described as toxicity pathways. Changes of the same set of molecular pathways in different cell and tissue types may generate seemingly unrelated health conditions. Today, healthcare system will likely treat these conditions as unrelated, while recognition of these conditions as syndromes with common environmental etiology may facilitate identification of causal routes of major health problems to inform efficient interventions. We propose an in-silico approach to identify diseases sensitive to multi-chemical exposures. First, sensitivity of genes to multi-chemical exposures was identified using a total of 591,084 individual chemical-gene interactions from the Comparative Toxicogenomic Database (CTD) and expressed as a number of chemical-gene interactions per gene. Next, molecular pathways enriched with genes sensitive to chemical exposures were identified using Gene Set Enrichment Analysis (GSEA) against KEGG and Reactome datasets. Top sensitive pathways were resubmitted to the CTD to identify disease categories associated with these pathways. Finally, diseases sharing common pathways, were identified using hierarchical clustering. Health conditions predicted as the most sensitive to cumulative chemical exposures include major public health problems: multiple forms of cancer, metabolic, cardiovascular disease, neurodevelopmental and neurodegenerative diseases, and autoimmune conditions. This analysis predictably identified clusters of diseases with known shared pathology (e.g. asthma and pneumonia), but also it identified some clusters which are difficult to explain (e.g. liver cirrhosis and cleft palate). Further, research is needed to identify major health conditions and syndromes sensitive to cumulative chemical exposures.

P403: Imaging-Enabled Simultaneous Investigation of Tumor Kill and Collateral Toxicity-Induced Tissue Damage Using Venetoclax-Based Mono and Combination Therapy in a Multiple Myeloma Model

Steven E. Johnson¹, Sarah R. Mudd², Erwin R. Boghaert², Joel D. Leverson², Magali R. Guffroy², Bruce A. Trela², Chad R. Haney¹, Alisha N. Spann¹, Jungwa Lee¹, and Ming Zhao¹

¹Northwestern University, Chicago, IL, USA

²AbbVie Inc., North Chicago, IL, USA

Introduction: Successful oncologic therapies are as much about preventing toxicity-induced cell death in collateral tissues as they are about killing malignancies. We applied an in vivo imaging technique utilizing ^99mTc-duramycin which confers a simultaneous readout for the pro-apoptotic action of chemotherapeutic agents in tumors and off-target tissue damage to examine the anti-tumor efficacy and systemic toxicity-induced tissue damage of venetoclax and bortezomib in comparison as well as in combination in a preclinical multiple myeloma model. Methods: Baseline scans were acquired in OPM-2 multiple myeloma xenograft tumor-bearing SCID-beige mice then drug treatment was started with venetoclax (AbbVie, 100mg/kg, i.p., once daily), bortezomib (Bortezomib, 1mg/kg, i.v., once every 4 days) or combination. For each scan at baseline, days 1, 3, 5 and 8, ^99mTc-duramycin was injected intravenously, and SPECT/CT were acquired. Radioactivity uptake, a surrogate marker for tumor kill and collateral tissue damage, was quantified. Results in key tissues were compared with histopathology. Results: Significant elevation in radioactivity uptake was detected in tissues in treated animals versus control. Key findings were: 1) venetoclax and bortezomib target different spatial aspects of the tumor; 2) the efficacy of the combination therapy is dominated by bortezomib; 3) at the dosages tested, Bortezomib showed more extensive effect compared to venetoclax; and 4) the combination therapy resulted in greater effect on susceptible tissues. Conclusion: The outcome demonstrated potential utilities of the technology for characterizing the impact of drug dosing and combination in vivo, thus for determining the effectiveness of anticancer treatment and collateral tissue damage.

P404: Intra-Articular Dose Administration in Rats, Dogs, and Nonhuman Primates to Support Nonclinical Safety Testing of Novel Therapeutics

Scott Williams¹ and Jessica Schroud¹

¹Labcorp Early Development Laboratories Inc., Madison, WI, USA

There has been an increased focus on the development of novel therapies for direct injection into the intra-articular space for the treatment of rheumatoid and osteoarthritis. Degenerative joint disease is a progressive disease involving the gradual degeneration of cartilage leading to disability. Current nonsurgical treatments focus on managing pain and swelling and increasing joint flexibility. Cartilage damage is generally accepted to be irreversible and there are currently no approved disease-modifying treatments. Intra-articular injection is an attractive approach because it delivers the therapeutic directly into the synovium, reducing systemic circulation and quantity of drug product needed. The nonclinical development of a small molecule new chemical entity generally requires a rodent and non-rodent species for toxicology testing. In this study, intra-articular dose administration was characterized in the stifle joints of rats, dogs, and nonhuman primates. Suitable anesthesia regimens for each species were developed for the injection procedure. Injection procedures and successful injection volumes were established for rats (≤0.1 mL), dogs (≤2 mL), and nonhuman primates (≤1 mL). Injection success was confirmed by visual inspection of contrasting agent within the articular space, with minimal leakage and no rupturing of the synovial membrane. A post-procedural analgesia regimen was developed for each species to alleviate pain and discomfort. Additionally, successful synovial fluid recovery was demonstrated in each species. In summary, intra-articular dose administration by injection into the stifle joint of rats, dogs, and nonhuman primates is a successful procedure for use in preclinical safety evaluations of novel therapeutics.

P405: High Impact of 3D InSight^TM Liver Model DILI Assays on Pharmaceutical Drug Development through High Throughput and High Predictive Power

Bruno Filippi¹, Armin Wolf¹, Natalia Zapiorkowska-Blumer¹, Monika Tu¹, and Friederike Wenz¹

¹InSphero AG, Schlieren, ZH, Switzerland

Hepatotoxicity is an important safety liability of new drug candidates potentially causing the discontinuation of preclinical and clinical development. Micro-physiological in vitro systems raised expectations of their use for safety assessment, but few made it into the regulatory and industrial workflow. The 3D InSight^TM liver model is a standardized high throughput in vitro model based on human primary cell with essential structural and functional features of the native liver. Previous studies demonstrated that this model predicts hepatotoxicity more accurately than 2D models. In this study, the performance of this model was evaluated by testing 63 drugs chosen from the DILIrank, a Food & Drugs Administration dataset consisting of drugs classified according to their hepatotoxicity potential. The cellular ATP IC₅₀ values of the compounds generated in this model were compared to human exposure data. Using this approach, 90.3% of the drugs annotated by the FDA as “Most-DILI-Concern” were accurately predicted as hepatotoxic whereas 81.3% of the “No-DILI-Concern” drugs were accurately predicted as non-hepatotoxic. Complex methods using transcriptomics and sensitizers increased the sensitivity of the model further. In future studies, more DILIrank compounds will be assayed using this approach in collaboration with pharmaceutical partners and regulatory bodies. The sensitivity, specificity, easy-of-use, and cost-effectiveness makes of the 3D human InSight^TM liver spheroid model a productive in vitro tool for liver safety assessment. It enables the generation of high-quality hepatotoxicity datasets for the drug development process, thus supporting critical internal go/no-go decision-making.

P406: Intra-Cisterna Magna Delivery in the New Zealand White Rabbit

Catherine Parisien¹, Tania Robillard¹, Frederick Emond¹, Claudine Tremblay¹, Ryan Schafbuch¹, and Stefania Cinquino¹

¹Charles River Laboratories, Senneville, Quebec, Canada

Intrathecal injection into the cisterna magna space of rats and non-human primates was previously validated in our laboratory as routes of cerebrospinal drug delivery of potential therapeutic agents. In order to develop this method in an alternative non-rodent species, our laboratory assessed the feasibility of intra-cisterna magna (ICM) administration in the New Zealand White rabbit. Briefly, 10 animals received an analgesic prior to and following the procedures and were anesthetized using a cocktail of ketamine, dexmedetomidine, and butorphanol. The puncture area was shaved and prepared, and animals were placed in a lateral recumbency position with the muzzle in ventral flexion. A spinal needle was slowly inserted perpendicular to the skin between C1 and the skull. Proper positioning of the needle and placement in the cisterna magna was confirmed by cerebrospinal fluid (CSF) flow and fluoroscopy imaging, at which point the dosing syringe was connected and a 0.3-mL dose volume of artificial CSF was slowly injected over a targeted rate of 1 mL/min. Clinical condition and food consumption were monitored daily, body weight once weekly, and necropsies performed on Day 29. Following a single ICM administration, transient clinical signs were limited to eyes partly closed and decreased activity; absent by 24 hours postdose. There were no apparent changes in body weights, food consumption, no abnormal findings in CSF clinical chemistry and cell count parameters, and minimal background changes in histopathology. In conclusion, a single ICM injection in the New Zealand White rabbit was well tolerated and presented minimal experimental background changes.

P407: Overview of Potential Mechanistic Evidences of Perfluorooctane Sulfonic Acid (PFOS) Organized by Key Characteristic of Carcinogens (KCs) Approach and Identification/Grouping of PFOS Precursors Using Computational Tools

Jennifer ChingYi Hsieh¹, Meng Sun¹, and Martha S. Sandy²

¹Office of Environmental Health Hazard Assessment, California Environmental Protection Agency, Sacramento, California, USA

²Office of Environmental Health Hazard Assessment, California Environmental Protection Agency, Oakland, California, USA

PFOS is a persistent and ubiquitous per- and poly-fluoroalkyl substance (PFAS). Exposures to PFOS, which is bioaccumulative and toxic, pose public health risks, including cancers. The KCs provide a methodical approach to identifying and organizing mechanistic evidence in cancer hazard identification. We applied this approach to PFOS and identified evidence relevant to 8 of the 10 KCs. Overall, the available mechanistic evidence for PFOS is strongest for the following 5 KCs: modulates receptor-mediated effects, induces oxidative stress, is genotoxic, is immunosuppressive, and induces epigenetic alterations. We also identified numerous PFASs that can transform or degrade to PFOS. We compiled a list of potential PFOS precursors from multiple data sources, including government reports (i.e., Environment Canada (2006), Organisation for Economic Co-operation and Development (OECD) (2007), Australia (NICNAS 2019)), the PFAS Master List from the EPA CompTox Chemical Dashboard, and peer-reviewed publications. The ability of these chemicals to form PFOS was validated by quantitative structure-activity relationship (QSAR)-based metabolic simulators embedded in the OECD Toolbox or by knowledge-based expert judgment. A list of more than 200 PFOS precursors was identified. Several computational tools, i.e., ChemDoodle 2D, Maestro suite, ChemoTyper and ChemoType Editor, were applied to assign these chemicals to 14 major subgroups. The structure-activity relationships between or within these subgroups may provide some insight for prioritization of chemicals for evaluation, based on the characteristics of each subgroup (e.g., degradation half-life or other physical-chemical properties). The KCs concept and this list may assist in designing read-across approaches for health risk assessment of PFOS precursors.

P408: Time Course of Adeno-Associated Virus Immune Response in NHPs

Rajeev K. Dhawan¹, Connor J. Balzer¹, Lachelle CP. Campbell¹, Brittany D. Egan¹, Donald T. Nicholson¹, and Steve M. Jennings¹

¹Charles River Laboratories, Wilmington, MA, USA

Most of 200 current gene therapy trials employing viral vectors are based on AAV vectors. Presence of AAV neutralizing antibodies (NAb) as pre-existing immunity can undermine the potential efficacy of the drug and biases outcome of the studies both in humans and NHPs. Prescreening of NHPs for AAV NAbs prior to studies is routinely done but it’s not clear how close to a study start date it should be performed. We conducted a time course study of AAV NAb prevalence to find out how much time in advance NHPs can be screened. In this study, 100 cynomolgus NHPs were sourced from a single Asian supplier and housed in separate rooms. These NHPs were not always kept with the same roommates and occasionally were grouped with cohorts from other suppliers. Sera was collected from NHPs at three time points starting with 100 animals at 0 months, falling to 87 at 4 months and 65 at 7 months post arrival. Out of these twenty serum samples (12 males and 8 females) were tested by luminescence based AAV2, AAV8, AAV9 and AAVrh74 NAb screening assays at 1/10, 1/20 and 1/40 sample dilutions. While the majority of NHPs showed no significant change in AAV NAb titer over the course of the study, two animals showed an increase in NAb titer greater than 4-fold. Results from these studies suggests that routine prescreening for AAV NAbs in study animals can be done 3-4 months in advance without any significant seroconversion but with proper housing conditions.

P409: Immunophenotyping of Cynomolgus Monkey Immune Cell Phenotypes in Whole Blood Using Central/Effector Memory, Activation, Proliferation, and Regulatory T-Cell Markers

Michael Shea¹, Nicole Shepherd², Kathryn MacPherson², Joseph Hollenbaugh², Amber Rowse³, and Amanda Lucchini¹

¹Labcorp Early Development Laboratories Inc., Madison, WI, USA

²Labcorp Early Development Laboratories Inc., Greenfield, IN, USA

³Labcorp Early Development Laboratories Inc., Ann Arbor, MI, USA

Flow cytometry is a preferred method when performing immunophenotyping for safety assessment and clinical diagnostics. To best serve these needs, we validated a 5-panel flow cytometry approach to quantify whole-blood relative frequencies and absolute cell counts of T-cells, B-cells, NK-cells (together abbreviated as TBNK), and a wide array of subpopulations including; central and effector memory cells, activated cells, proliferating cells, and Regulatory T-cells. All panels shared a common TBNK backbone of markers including CD45, CD3, CD4, CD8, CD20, and CD159a. The first panel validated Regulatory T-cells using the combination of CD25 and the intracellular maker, FoxP3 (CD25+FoxP3+). The second panel validated Regulatory T-cells using CD25 and CD127 (CD25+CD127-). The third panel used CD95 and CD28 to validate Naïve T cells (CD95-CD28+), Central Memory T cells (CD95+CD28+), and Effector Memory T cells (CD95+CD28-). The fourth panel included Ki67 to quantify proliferating cells. The fifth panel validated either CD25+ or CD69+ activated cell phenotypes. A total of 32 animals were used to establish a variety of validation parameters including Assay Precision, Whole Blood Stability, Fixed Sample Stability, Scalability, Reproducibility, Analytical Sensitivity, and Lower Limit of Quantitation. %CV’s were calculated for each reported phenotype.

P410: Cross Study Analyses of SEND Data: Toxicity Profile Classification

Mark Carfagna¹, William Houser², R. Thompson³, Kevin Snyder⁴, Md Yousuf Ali⁵, Brianna Paisley¹, Stephanie Leuenroth-Quinn⁴, Wenxian Wang², Saurabh Vispute⁶, Hiroyuki Hanafusa⁷, Susan Butler⁵, Jesse Anderson⁴, Cm Sabbir Ahmed⁵, Nikolai Jensen⁸, and Tamio Fukushima⁷

¹Eli Lilly & Company, Indianapolis, IN, USA

²Bristol Myers Squibb, New Brunswick, New Jersey, USA

³Janssen, Spring House, PA, USA

⁴US Food & Drug Administration, Silver Spring, Maryland, USA

⁵Oak Ridge Institute for Science and Education, Oak Ridge, TN, USA

⁶Pfizer, Groton, CT, USA

⁷Shionogi & Co., Ltd., Osaka, Japan, Japan

⁸NovoNordisk, Copenhagen, Denmark, Denmark

The objective of this work, a collaborative effort between BioCelerate and FDA, is to develop SEND data harmonization/transformation strategies and apply analytic techniques to enable an understanding of the similarities and differences between two or more data sets. Example use cases include understanding a single compound’s toxicity profile across all studies performed or evaluating on- versus off-target toxicity for multiple compounds intended for the same pharmacological target. A subset of de-identified studies from the BioCelerate Toxicology Data Sharing database were used for the analyses. The analyses involved transformation and visualization of SEND data variables to integrate both numerical and categorical data. Toxicity profiles for key organ systems were developed by selecting pertinent data from the body weight (BW), organ weight (OW), microscopic (MI), and laboratory (LB) domains. Profiles for liver, kidney, brain/neuro, GI tract and lungs were developed. In addition, a dashboard and user defined scoring system were created to facilitate custom analyses for data of interest. The output of the analyses includes a series of radar plots that enable the user to visualize and evaluate data from the organ system level to individual animal data points. This evaluation of data provides the tools for scientists to compare and contrast toxicity profiles across multiple studies using SEND data. Toxicity profile analysis of different compounds intended for the same pharmacological target will be presented in the poster. Cross-study comparisons using SEND data are expected to improve the identification of unique findings related to the intended target, species and duration of dosing.

P411: Comparison of Maintenance Regimens Between Dosing Sessions on Patency Rates and Animal Welfare in a 3-Month Subcutaneous Infusion Study in Sprague Dawley Rats

JoAnna Bultman¹, Teresa R. Gleason¹, Alexis J. Burgess¹, Heather R. Greier¹, and Joseph W. Yuengert¹

¹Charles River Labs, Ashland, OH, USA

Subcutaneous (SC) infusion has indications in human medicine for hydration, nutrition, and administration of medications and has advantages over intravenous infusion due to ease of application and fewer complications. Our lab was requested to dose rats for 96 hours via continuous SC infusion once monthly for 3 months (4 doses). A pilot study was conducted to evaluate catheter patency maintenance regimens and animal welfare. Forty male and 40 female rats, implanted with SC catheters attached to transcutaneous buttons in the scapular area, were procured. Group 1 (20 M/20 F) dosed continuously with 0.9% sterile saline (NaCl) for 96 hours at a rate of 0.06 mL/hour followed by a 3-week maintenance period where 0.5 mL NaCl was bolus dosed twice weekly; animals were socially housed during the maintenance period. Group 2 (20 M/20F) was dosed at the same rate followed by a maintenance period where NaCl was administered continuously at a rate of 0.02 mL/hr; animals were single housed during the maintenance period due to tethers. Clinical observations, body weights, patency rates, macroscopic and microscopic findings were evaluated. Clinical observations included mild, transient irritation at the exteriorization site did not impact overall animal health. All animals gained weight and there were no dosing/maintenance regimen-related effects on survival, gross or histologic observations. Patency was maintained for both groups. Damage to tethers occurred at a greater rate in Group 2 (50% vs 15%). The Group 1 maintenance routine was preferred for animal welfare because animals could be socially housed between dosing sessions.

P412: Evaluating Short and Long-Term toxicity Response of Models Comprised of Fully Differentiated Primary Bronchial Tracheal Epithelial Cells to Either Cadmium Chloride or Pentamidine

Kevin M. Tyo¹ and Fang Tian¹

¹ATCC, Gaithersburg, MD United States, USA

Respiratory tract diseases stemming from toxic compound exposure significantly contribute to the global health burden. Traditional in vitro airway models, due to their lack of physiological relevance, are often unable to provide meaningful and accurate toxicological assessments. Advanced in vitro airway models, however, promise to provide more predictive information for use in human airway health. Here, we constructed mature airway models comprising fully differentiated primary bronchial tracheal epithelial cells incubated in 24-well plate inserts and cultured under air-liquid interface for 4 weeks. The toxicological response to short-term (24 hours) exposure to either cadmium chloride (CdCl₂) or pentamidine were evaluated and compared in both differentiated and undifferentiated cells. The toxicological response to long-term exposure (1, 2 weeks) to either compound in differentiated airway models was also explored. Changes in viability and cytokine expression was quantified and compared in both models. Additionally, histological imaging (H&E, alcian blue, IHC) was conducted on mature airway models to visually assess model disruption, inflammation, and tight junction disruption. We observed that all airway models expressed dose-dependent response to both CdCl₂ and pentamidine exposure, with increased cell death corresponding with increased compound concentrations. Additionally, differentiated models demonstrated higher resistivity to cell death compared to undifferentiated counterparts. Moreover, exposure to low concentrations of the compound resulted in increased cytokine expression relative to untreated controls. Finally, long-term exposure to CdCl₂ resulted in model disruption and death, whereas pentamidine exposure demonstrated limited model disruption. These results suggest that these airway models may serve as useful tools future airway toxicity research.

P413: Update on the Ongoing OECD Validation of the ToxTracker Assay for Genotoxic Mode of Action Assessment

Giel Hendriks¹, Els Adriaens², Ashley Allemang³, Jan van Benthem⁴, Julie Clements⁵, Gabrielle Cole⁶, Maria Engel⁷, Annie Hamel⁸, Darren Kidd⁵, Stephanie Kellum⁹, David Kirkland¹⁰, Tomomi Kiyota⁶, Abby Myhre⁹, Valerie Naëssens¹¹, Stefan Pfuhler³, Marise Roy⁸, Raja Settivari⁹, Maik Schuler⁷, Philippe Vanparys¹², and Andreas Zeller¹¹

¹Toxys, Oegstgeest, ZH, The Netherlands

²Adriaens Consulting, Aalter, OV, Belgium

³Procter & Gamble, Mason, OH, USA

⁴RIVM, Bilthoven, UT, The Netherlands

⁵Labcorp, Harrogate, Yorkshire, United Kingdom

⁶Genentech, South San Francisco, CA, USA

⁷Pfizer, Groton, CT, USA

⁸Charles River Laboratories, Senneville, Quebec, Canada

⁹Corteva Agriscience, Newark, DE, USA

¹⁰Kirkland Consulting, Tadcaster, Yorkshire, United Kingdom

¹¹Roche, Basel, Basel, Switzerland

¹²Gentoxicon, Vosselaar, Antwerpen, Belgium

ToxTracker is a mammalian stem cell-based reporter assay that detects activation of specific cellular signaling pathways upon chemical exposure. ToxTracker contains six different GFP-tagged reporter cell lines that together allow the accurate identification of genotoxic substances and discrimination between induction of DNA damage, oxidative stress and/or protein damage in a single test. More recently, the assay was extended to allow the discrimination between clastogenic and aneugenic compounds. The ToxTracker assay is currently being evaluated in a large international inter-laboratory validation study, approved by the OECD. The goal of this prospective validation study is to explore the applicability of ToxTracker for regulatory applications, establish the transferability and reproducibility of the assay and to explore how it can be applied to improve the in vitro genotoxicity testing strategies. The validation has been conducted strictly following OECD guidance document 34. ToxTracker was transferred to seven laboratories. The validation labs were trained to perform the assay and tested a training set of compounds to show their proficiency to run ToxTracker. Next, the labs evaluated a selection of 64 coded, well-established genotoxic and non-genotoxic chemicals with each compound being tested in three labs independently. All the experimental work has been completed and data have been analyzed. The accuracy to predict genotoxicity, as well as the intra- and inter-laboratory reproducibility were determined. In this poster, we will give an overview how the ToxTracker validation was performed and the first results from this OECD validation.

P414: ReproTracker: Next Generation In Vitro Developmental Toxicity Testing

Luke Flatt¹, Sabine Hartvelt¹, Marleen Feliksik¹, Tom Zwetsloot¹, Giel Hendriks¹, Torben Osterlund¹, Amer Jamalpoor¹, and Paula van Rossum¹

¹Toxys B.V., Oegstgeest, South Holland, The Netherlands

Birth defects due to chemical exposure remain a significant problem worldwide. Currently, assessing potential developmental toxicants relies mainly on murine-based models not necessarily representing the human situation. Several alternative in vitro assays have been developed, that often suffered from low predictability and not providing mechanistic understanding of developmental toxicity. ReproTracker mimic the differentiation of early embryonic development. Human iPSCs are directed to differentiate toward germ layer-specific cell types; hepatocytes, cardiomyocytes, and neuroectoderm. Cell fate commitment is investigated by morphological profiling and assessment of temporal expression patterns of lineage-specific biomarkers. Here, a decrease in the expression of biomarker genes and morphology disruption following compound treatment is indicative of teratogenicity. To further extend the offering of the assay, efforts have been made to explore the effect of chemical-induced biomarker perturbation on the expression of lineage-specific proteins. The work herein describes the progress to date in using high content imaging to correlate the influence of chemical exposure between transcript and protein effects during the commitment from neuroectoderm to neural progenitor cells. We have found that teratogenic chemicals down-regulate the expression of PAX6, and dampened PAX6 and Nestin protein levels at the progenitor stage, dose-dependently. Combining the tri-lineage results and comparing to the in vivo classification, the ReproTracker provides high accuracy (85%), sensitivity (85%), and specificity (84%). ReproTracker is a state-of-the-art in vitro assay that can identify the teratogenic potential of chemicals at both transcript and protein levels with high accuracy and, provide evidence as to the likely outcome of in vivo test systems.

P415: In Vitro Predictive Toxicity Screening Assays during Early Stage Drug Development: Case Study Data for the Validity of 2D and 3D Models

Sebastien Breche¹, Elise Esneault¹, Sonia Goineau-Brissieux¹, Florian Simon¹, Christophe Legrand¹, Mikeal Paquet¹, and Guillaume Froget¹

¹Porsolt S.A.S., Le Genest St Isle, Mayenne, France

Drug induced toxicity is responsible for a significant proportion of drug attrition and costly withdrawal during late-stage development. In vitro predictive toxicity assays offer a rapid and cost-effective option for identifying potential toxicity concerns at the early lead optimization stage. Key points to consider when developing in vitro predictive toxicity assays include cell type, readouts employed, throughput, and biological relevance. Potential in vitro models include traditional 2D assays or more complex 3D spheroid assays. Here we first confirm the predictive utility of a 2D hepatoxicity panel developed for screening large compound libraries at the early stages of the discovery process. This assay utilizes primary rat hepatocytes using flow cytometry and microscopy platforms, with a panel of toxicity readouts: Cytolysis, MMP, Cholestasis, Lipid accumulation, and GSH depletion. This assay displays a ∼83% correct predictive hit rate for a battery of 30 compounds previously evaluated in acute toxicity assays in rodents. These data demonstrate that rat hepatocytes are a valid and cost-effective alternative to human hepatocytes for predictive hepatoxicity screening. We also demonstrate the advantages of using complex human 3D spheroid cultures for early-stage predictive toxicity screening. 3D spheroids, generated from human neuron and astrocyte iPSC co-cultures were used to screen more than 20 neurotoxic substances in synchronized calcium oscillation assays. Known inhibitors and activators of basal calcium flux were successfully identified using a high-throughput approach. This approach can also be relevant for predictive toxicity screening in other therapeutic domains, using cell types such as cardiomyocytes and kidney cells.

P416: Advanced Primary Human Liver Cell-Based In Vitro Model for the Evaluation of Therapeutic Antisense Oligonucleotide Hepatotoxicity Liabilities

Bruno Filippi¹, Sue Grepper², Friederike Wenz¹, Monika Tu¹, Natalia Zapiorkowska-Blumer¹, and Armin Wolf¹

¹InSphero AG, Schlieren, ZH, Switzerland

²InSphero Inc., Brunswick, Maine, USA

Antisense oligonucleotides (ASOs) form a promising therapeutic modality, but their development is hampered by hepatotoxicity. Previous studies showed ASOs concentrate in the liver in which they can trigger apoptosis, off-target effects, chemistry-dependent toxicity, and the saturation of RNA processing machinery. 3D primary cell models are well-established tools for the prediction of small molecule hepatotoxicity, however their systematic evaluation as a predictive tool for ASOs hepatotoxicity is still lacking. In this work, the performance of the human 3D InSight^TM liver model in predicting ASOs hepatotoxicity is evaluated using six ASOs with well-documented in vivo liver data. ASOs’ effects on the viability of the model, cytosolic leakage and apoptosis induction were assessed. Additionally, uptake of a tool ASO and the resulting target-specific gene silencing were also investigated. The results show that the in vitro cytotoxicity of the ASOs aligns with their reported hepatotoxicity. Also, all ASOs led to an increased caspase 3/7 activity, albeit at very different magnitudes, suggesting that the induction of apoptosis is a hazard common to ASOs. Histochemistry results show that ASOs are readily and uniformly taken up by the model, even reaching the inner core of the cell spheroid. Finally, qPCR analyses show that ASO treatment leads to a specific and strong downregulation of the target mRNA. Altogether, these results suggest that the human 3D InSight^TM liver spheroid model is a promising in vitro tool for the prediction of ASO hepatotoxicity. In future investigations, more clinical ASOs will be assayed using the same procedure.

P417: Challenges and Approaches for Nonclinical Inhalation Delivery of Genetic Material

Conor A. Ruzycki¹, Jason Cox¹, Hammad Irshad¹, Derek Montoya¹, Kent Morgan¹, Jacob D. McDonald¹, Yue Zhou¹, and Philip J. Kuehl¹

¹Lovelace Biomedical Research Institute, Albuquerque, New Mexico, USA

Gene therapies provide great potential for developing novel treatments of disease. Targeted gene delivery through the inhalation route maximizes efficacy and reduces systemic exposure. Non-clinical models are critical in evaluating the efficacy and safety of inhaled gene therapies and incurs a number of unique challenges. This abstract will present multiple case studies on the development of fit for purpose methods with a focus on identifying optimal approaches and areas for improvement. Methods for the delivery of naked siRNAs, lipid nanoparticle encased mRNAs, lentiviral and adeno-associated viral vectors for siRNA, and antisense oligonucleotides are discussed. Each program requires a fit for purpose combination of aerosol generation technique (compared to jet nebulizers, vibrating mesh nebulizers or novel methods), collection methods (filter, condensate, impactor, and/or impinger) and analytical technique (qPCR, ddPCR, HPLC, etc.) based on the method development results. These methods are applied to rodent and non-rodent exposure systems to characterize viability, integrity, aerosol concentration and dose. Ensuring that these methods and aerosol generation systems are fit for purpose ensure the quality/ compliance of the work and reduces the material requirements for the precious test articles. An example of fit for purpose methods would include a viral vector RNA test article for oral inhalation. This case study utilized a compressed air jet nebulizer, with a qPCR assay extracted off of a filter media, a cell-based assay for viability from aerosol condensate and particle size a cascade impactor. Together these methods allowed calculation of dose with standard methods and conduct of non-clinical toxicology studies.

P418: Characterization of Pain Modulators on Nerve Conduction Following Electric, Tactile and Heat Stimulation in Dogs

Nataliya Sadekova¹, Adam Werts², Andrew Brown¹, Kevin Norton¹, and Brian Roche³

¹Charles River Laboratories, Senneville, Quebec, Canada

²Zoetis, Kalamazoo, Michigan, USA

³Charles River Laboratories, Ashland, Ohio, USA

Peripheral neuropathy can be an important side effect of certain medications. In preclinical settings, nerve conduction velocity (NCV) evaluation provides a sensitive and valid index of induced neuropathies. However, standard NCV typically uses electrical stimulation which preferentially activates large diameter fibers and limits detection of changes in small diameter fibers. Hence, the use of tactile and heat stimulation should help isolate effects on small diameter afferent fibers, such as Aδ and C, which primarily convey mechanical and thermal pain sensation. In this study, onset latency and NCV were measured in the purely sensory sural nerve in dogs along with somatosensory evoked potentials (SEP at L6-L5 and L5-L4) following electrical, tactile and heat stimulation. To evaluate the sensitivity of these measurements, the dogs were first administered known analgesics, tramadol and gabapentin. Both compounds increased onset latency of the sural nerve response and SEPs following tactile stimulation confirming their analgesic effect, with no significant changes noted in response to heat. Additionally, to explore whether this technique can detect potential increased pain sensitivity, dogs were sensitized with an injection of nerve growth factor (NGF). The threshold force (N) required to induce a response to tactile stimulation was then measured. NGF administration significantly decreased response thresholds. In conclusion, NCV and tactile stimulation threshold evaluation of the sural nerve provides a sensitive method to study effects of compounds on fibers responsible for pain sensation.

500s—Safety Evaluation Pharmaceuticals

P501: Evaluation of the Use of Glatiramer Acetate during Breast Feeding

Nurit Ashkenazi¹ and Hussein Hallak²

¹Research and Development Teva Pharmaceutical Industries, Ltd, Netanya, North Central District, Israel

²Research and Development Teva Pharmaceutical Industries Ltd, Netanya, North Central District, Israel

Background. Glatiramer acetate (GA) is treatment for multiple sclerosis, which occurs mostly in women during childbearing years. GA is a polypeptide mixture of varying sequences and sizes, making it difficult to develop pharmacokinetic methods to accurately account for concentration of all possible peptides. Present report describes methods used to examine the likelihood of GA transfer into breastmilk. Methods. Protein binding of [¹²⁵I]-GA was measured by gel-filtration. Pharmacokinetic exposure was assessed using polyclonal ELISA method following single 60mg SC injection of GA in 17 healthy volunteers (HV), SC 30mg/kg injection and oral 600mg/kg in 4 cynomolgus monkeys, and 3mg/kg SC injection in 3 dogs. Human relative infant dose (RID) was calculated assuming C_max in breastmilk = drug serum C_max values and was factored for oral absorption. Results. [¹²⁵I]-GA has high plasma proteins binding (98%). Limited pharmacokinetic data from animals and humans demonstrated short exposure duration. For HV, GA serum concentrations < level of quantification (LOQ, 50ng/mL) in eight subjects by one hour. For monkeys and dogs, GA ≤ LOQ at 4-6 hours postdose. Oral vs SC-administration C_max and AUC were < 2.6% and < 0.5%, respectively. RID = 9%; when adjusted for oral absorption, RID = 0.2%. Conclusion. GA is unlikely to be transferred to breastmilk due to its high protein binding. The short exposure duration demonstrated that there is limited time in which GA could transfer to breastmilk. GA has low oral absorption thus unlikely to be absorbed from the infant GI. The RID for GA is 0.2%, a value considered 'safe' by the World Health Organization.

P502: Electrophysiological Changes in the Rabbit Ventricular Wedge Assay and Human-Induced Pluripotent Stem-Cell Derived (iPS) Cardiomyocytes Correlated with Severe Arrhythmia Observed In Vivo, Whereas Standard In Vitro Ion Channel Assays Were Not Predictive of Effects Seen

Alan Brown¹, Greg Friedrichs², Hai-Ming Tan², Martin Traebert³, Valerie Weber³, Gan-Xin Yan⁴, and Nancy Yao²

¹Novartis Institutes for Biomedical Research, Cambridge, MA, USA

²Novartis Institutes for Biomedical Research, East Hanover, NJ, USA

³Novartis Institutes for Biomedical Research, Basel, CH, Switzerland

⁴Lankenau Institute for Medical Research, Wynnewood, PA, USA

Compound A is a low molecular weight irreversible inhibitor of myeloperoxidase investigated as a potential therapy for peripheral artery disease. A single oral dose in dogs at ≥5 mg/kg resulted in severe and prolonged arrhythmias as determined using jacketed ECG telemetry equipment. Premature ventricular contractions, bigeminy, ventricular tachycardia and atrial-ventricular block were observed in a dose-dependent fashion (at Cmax, free ≥1.67 µM) that progressed in severity over time. Nevertheless, a panel of 13 ion channel (K, Na, Ca) assays, including HERG, did not identify pharmacologic risks of the molecule. Compound A and related Compound B were subsequently evaluated for electrophysiological effects in the isolated rabbit ventricular wedge assay. Compounds A and B prolonged QT and Tp-e intervals at ≥1 and ≥0.3 µM, respectively, and both prolonged QRS at ≥5 µM. Compound A produced early after depolarizations and premature ventricular complexes at ≥5 µM. These data indicate both compounds can inhibit HERG (Ikr) and Nav1.5 ion channels. In iPS cardiomyocytes, Compounds A and B prolonged field potential duration at ≥3 µM and induced cellular dysrhythmia at ≥10 and ≥3 µM, respectively, further supporting pro-arrhythmic liability. In a repeat-dose rat toxicology study, heart tissue:plasma concentration ratios for A were ≥19X at 24 hours post-dose. In conclusion, in vitro ion channel assays may not identify cardiovascular risks observed in vivo, which can be affected by tissue drug distribution. Risk for arrhythmia may increase with a “trappable” ion channel inhibitor, particularly if cardiac tissue drug levels achieve a critical threshold for effect.

P503: A Combined In Vitro Approach for Early Seizure Prediction Utilising Human Derived Induced Pluripotent Stem Cells and Human Ion Channel Assays

Kim Rockley¹, Michael Morton¹, and Ruth Roberts¹

¹ApconiX, Alderley Edge, Cheshire, United Kingdom

Seizure liability remains a significant cause of attrition throughout drug development. Advances in stem cell biology coupled with an increased understanding of the role of ion channels in seizure offer an opportunity to improve identification of potential seizure risks preclinically. Human derived induced pluripotent stem cells (hiPSCs) representative of cellular subtypes in the brain can be incorporated into physiologically relevant in vitro models to predict seizure risk using high-throughput microelectrode array (MEA). hiPSC iCell GlutaNeurons containing 80% glutamatergic/20% GABAergic neurons were plated with astrocytes and monitored using the Axion Maestro Edge MEA system. Compounds being studied in the Health and Environmental Sciences Institute (HESI) NeuTox project (4-AP, amoxapine, chlorpromazine, linopirdine, pentylenetetrazole, picrotoxin, pilocarpine, phenytoin, strychnine) and other seizurogenic drugs (bupropion, clozapine, diphenhydramine, paroxetine, quetiapine) caused characteristic changes to electrical activity (changes to network bursting, firing rate etc.) indicative of seizure. Alongside this, the same compounds were screened against a panel of 15 ion channels with strong links to seizure (Nav1.1, Nav1.2, Nav1.6, Kv7.2/7.3, Kv7.3/7.5, Kv1.1, Kv4.2, KCa4.1, Kv2.1, Kv3.1, KCa1.1, Cav2.1, GABA α₁β₂γ₂, nicotinic α₄β₂, NMDA 1/2A) using automated electrophysiology. Of the ion channels tested, clozapine, diphenhydramine, chlorpromazine, amoxapine, paroxetine, bupropion and linopirdine inhibited two or more ion channels, and all seizure causing compounds demonstrated at least one “hit” against our seizure panel. These studies highlight the potential utility of a combined in vitro approach for early seizure prediction to provide mechanistic information, and support optimal drug design in early development to reduce animal usage and save time and resource

P504: An Integrated, Human-Relevant, and Evidence-Based Translational Toxicology Paradigm for Cardiovascular Safety Assessment

Rachel A. Dee¹, Scott S. Auerbach¹, Brian R. Berridge¹, Kevin L. Dreher², David L. Gerhold³, Nicole C. Kleinstreuer¹, Shagun Krishna¹, Kelly A. Shipkowski¹, Sreenivasa C. Ramaiahgari¹, Xian Wu¹, and Brandy E. Beverly¹

¹National Institute of Environmental Health Sciences, Durham, NC, USA

²North Carolina State University, Raleigh, NC, USA

³National Center for Advancing Translational Sciences, Bethesda, MD, USA

Cardiovascular (CV) disease remains the most significant global cause of morbidity and mortality in humans. Lifestyle and genetics are key contributors but they cannot alone or in combination account for all the risk associated with the development of CV disease. Environmental exposures from chemicals, including pharmaceuticals, are recognized to be important contributors, but cardiovascular safety liabilities remain a common cause of drug development attrition. The NIEHS Division of the National Toxicology Program has initiated a Cardiovascular Health Effects Innovation initiative that is developing and implementing a translational toxicology pipeline of safety assessment capabilities to rapidly and accurately identify agents that can cause or contribute to CV disease risk. Our approach is evidence-based, beginning with in silico QSAR modeling and medium to high throughput bioactivity screening, including simple and complex in vitro confirmatory assays, and culminating with targeted in vivo validation assessments for adverse structural and functional cardiovascular biomarkers. Early predictions based on in silico models and in vitro bioactivity will be qualified in progressively complex assay systems, allowing high predictive confidence in early pipeline steps, model applicability, and identification of capability development needs. The assay systems used will be aligned to known human CV failure modes to reflect human biology as much as the complexity of the system permits with a goal of optimizing the translational relevance of adverse outcomes.

P505: Nonclinical Safety Profile of Darovasertib, a Potent and Selective PKC Inhibitor under Evaluation for the Treatment of Metastatic Uveal Melanoma

Claire L. Neilan¹, Lisa Bell², and Mick O'Quigley¹

¹IDEAYA Biosciences, South San Francisco, CA, USA

²Time To Approval Consulting, San Carlos, CA, USA

Darovasertib is a potent and selective inhibitor of the PKC family of kinases. IDEAYA Biosciences is evaluating the combination of darovasertib and crizotinib, a cMET inhibitor, in patients with GNAQ/GNA11 mutant solid tumors, including metastatic uveal melanoma. The nonclinical safety profile of darovasertib was evaluated to support dosing in patients and in clinical pharmacology studies in healthy volunteers. Repeat oral dose toxicity studies of up to 13 weeks in duration, with a 4-week recovery period, were conducted in the rat and dog. 13-week study findings at exposures equivalent to those associated with the therapeutic clinical dose included decreased red cell mass parameters, erythrophagocytosis in the mesenteric lymph nodes that correlated with hemorrhage of the fundic portion of the glandular stomach (rats) and GALT (dogs), and increased liver weights that lacked a histological correlate. All findings were non-adverse, reversible and consistent with those observed in 4-week toxicity studies. Darovasertib tested Ames-negative, but positive for clastogenicity at high, non-clinically relevant concentrations in human peripheral blood lymphocytes. In a combined in vivo peripheral blood micronucleus/liver comet assay, there was no evidence of direct DNA damage following darovasertib treatment. Moreover, no increase in peripheral blood micronuclei was observed up to and including the high dose of 300 mg/kg/day, a dose associated with a 4-fold exposure margin. Safety pharmacology studies did not reveal any findings of clinical concern. Taken together, this nonclinical safety profile of darovasertib supports the continued dosing in cancer patients and clinical pharmacology studies in healthy volunteers.

P506: First-in-Human Dose Selection for a Novel Fusion Protein Using Animal Safety Data, Pharmacodynamics, and Pharmacokinetic Modeling

Marie C. Fortin¹, Dino Manca¹, Hemamalini Gursahani¹, Tong Lin¹, and Svetlana Markova¹

¹Jazz Pharmaceuticals, Palo Alto, CA, USA

JZP341 is an engineered fusion protein designed to provide long-lasting serum asparaginase activity in patients with tumors potentially responsive to metabolic deprivation. It is comprised of an active moiety which metabolizes asparagine and, to a lesser extent, glutamine to their acid, and a half-life extending moiety. The nonclinical development program included evaluation of the in vitro and in vivo pharmacology, single- and repeat-dose PK in rodent and non-rodent species, dose-range finding and GLP toxicology studies in rats and dogs, and stage-appropriate safety pharmacology studies. The GLP toxicology studies included an assessment of the pharmacodynamic, pharmacokinetics, and antidrug antibody formation. The assessment of the respiratory, cardiovascular, and central nervous systems did not reveal any effect of JZP341 on these functions at exposures greater than those anticipated clinically. JZP341 effects were limited to dose-dependent body weight loss, reduced body weight gains, and changes in the liver, kidney, and lymphoid organs consistent with those reported with other asparaginases. The first-in-human (FIH) dose was selected based on an integrated evaluation of JZP341 pharmacodynamics, pharmacokinetic modeling, and animal safety data to determine a safe starting dose anticipated to provide acceptable levels of serum asparaginase activity and ample margins of safety in a FIH clinical study in adults patients.

P507: Lack of Translation of Favorable Nonclinical Toxicology Profile of Antisense Oligonucleotide (ASO) ALG-020572 to Chronic Hepatitis B (CHB) Patients

Sandra Chang¹, Dinah Misner¹, Kusum Gupta¹, Gina Gratz², Jin Hong¹, Kha Le¹, Vikrant Gohil¹, Vivek K. Rajwanshi¹, Megan Fitzgerald¹, Felix Lai¹, Meenakshi Venkatraman¹, Min Wu¹, Matthew McClure¹, John Fry¹, David B. Smith¹, Julian A. Symons¹, Lawrence Blatt¹, Leonid Beigelman¹, Tse-I Lin¹, and Sushmita Chanda¹

¹Aligos Therapeutics, Inc., South San Francisco, CA, USA

²Inotiv, Evansville, IN, USA

ALG-020572 is a 5’ GalNAc-conjugated ASO that was being developed for the treatment of CHB by targeting the HBsAg coding region of the HBV genome. Once ALG-020572 enters hepatocytes, the GalNAc moiety is rapidly cleaved to release the active ASO, ALG-020579. ALG-020572 was administered twice weekly by subcutaneous injection for 2 weeks then weekly thereafter up to the highest dose of 50 mg/kg/dose (NOAEL) for up to 4 weeks in CD-1 mice and monkeys. Non-adverse and reversible findings consistent with oligonucleotide uptake were noted in the liver, kidney, lymph nodes, and injection site for both species. Rodent specific oligonucleotide-related inflammatory response, that to date have not translated into safety risks for humans, were noted in mice studies with sporadic, non-dose responsive ALT elevations (up to 6-fold) and non-adverse minimal single cell necrosis. Dosing for up to 20-weeks in chronic mouse studies yielded similar results with no worsening over time. ALG-020572 was well tolerated following single SC doses in heathy volunteers up to 480 mg, the highest dose tested. Overlapping or higher tissue and plasma concentration of the oligonucleotides were achieved in animals compared to humans. Despite a favorable non-clinical and healthy volunteer safety profile, ALG-020572 was discontinued in Phase 1 due to ALT flares (>10x upper limit of normal) seen in 4/6 patients after 2-7 SC doses of 210 mg at the same frequency as in the toxicology studies. Mechanistic studies are ongoing, notably in humanized mice +/-HBV infection, in 3D human liver models and in vitro cellular RNAseq evaluation.

P508: Safety Evaluation of Dydrogesterone Up to Exposure Saturation Level in Female Sprague Dawley Rats

Krishna H¹, Sushant K¹, Vichare AS¹, Yogesh B¹, Nikhilesh P¹, and Srinivasa H¹

¹Global Preclinical and Product Safety, Abbott Healthcare Pvt. Ltd., Mumbai, Maharashtra, India

Dydrogesterone is a potent, orally active progestogen, with a chemical structure closely resembling with endogenous progesterone. A 90-day repeated dose oral toxicity study along with systemic exposure assessment was performed in female Sprague Dawley rats to understand the on- and off-target adverse effects if any, up to exposure saturation level as there is limited information on the effects at higher exposure. A total of 4 doses were evaluated in this study. These doses were selected based on a pre-study wherein dydrogesterone exposure saturation levels were identified. The concentrations of Dydrogesterone and its metabolite, 20α-Dihydrodydrogesterone in rat plasma was analyzed using a validated LC-MS/MS method. Dydrogesterone was administered twice daily via oral gavage for 90-consecutive days and the study was conducted in accordance with OECD guidelines. The study included standard 90-day parameters such as clinical signs, mortality, ophthalmological examination, body weights, food consumption, toxicokinetics, clinical pathology, organ weights, gross pathology, and histopathology. Results from the study did not reveal any adverse findings other than reversible reproductive effects which are well-known pharmacological/progestogenic activity of dydrogesterone. The anticipated pharmacological effects were prominent at higher doses. Maximum plasma concentration and plasma exposure was dose proportional at lower doses and less than dose proportional at higher doses. Both, dydrogesterone and its metabolite, 20α-dihydrodydrogesterone did not show any propensity of accumulation. No additional off-target adverse effects were noted. Overall, the results of the present study confirm the safety of dydrogesterone at higher exposures, providing further support to safe clinical use.

P509: Hepatitis A Virus (HAV) Infection in Cynomolgus Monkeys Confounds the Safety Evaluation of a potential new Medicine

Chris J. Powell¹, John C. Kapeghian², John C. Bernal³, John R. Foster⁴, and Robert E. Lanford⁵

¹University of Cambridge - MRC Toxicology Unit, Cambridge, Cambridgeshire, United Kingdom

²Peclinical Safety Associates, The Woodlands, TX, USA

³CORES Veterinary Consulting, Walnut Grove, TX, USA

⁴Toxpath Sciences, Cheshire, Cheshire, United Kingdom

⁵Texas Biomedical Research Institute, San Antonio, TX, USA

In a 13-week toxicity study in Cynomolgus macaques (Mauritian) conducted at a non-US-based contract laboratory, 5-10 fold elevated serum ALT/AST and/or GLDH (200-1,000 U/L) were noted in individual animals from all groups including controls, with no apparent dose, exposure, or time-related response. Liver histopathology revealed minimal to slight inflammatory cell accumulation in periportal zones of most animals, and minimal to slight hepatocyte degeneration/necrosis in 10/42 animals from all groups. As these findings were more notable in 6 drug-treated animals, including some in the low dose group, the draft report concluded: treatment-related hepatotoxicity at all dose levels precluded determination of a NOAEL. Due to the unusual pattern of hepatotoxicity, an alternative hypothesis was proposed: that a factor other than drug exposure might be responsible for the hepatic effects. Accordingly, snap-frozen liver was tested for hepatitis viruses using a highly specific PCR method. Tests for hepatitis B, C and E virus were negative; however, 20/42 samples were positive for hepatitis A virus (HAV). Infection was strongly associated with increased serum ALT/GLDH, and/or hepatocyte degeneration/necrosis. Re-evaluation of data in this light, concluded that the hepatic changes were not drug-related. A subsequent 6-month toxicology study in HAV-vaccinated Cynomolgus macaques (Mauritian), confirmed the absence of drug-related hepatotoxicity. Although rarely investigated, subclinical HAV infection has occasionally been reported in laboratory primates including those used for toxicology studies and may be more prevalent than the literature indicates. In this case, identification of HAV infection prevented liver findings from being incorrectly attributed to treatment with the test compound.

P510: Using AOPs to Aid Expert Review and Decision-Making in a Weight-of-Evidence Assessment for ICH S1B

Susanne A. Stalford¹, Alex Cayley¹, Emma Hill¹, Steven Kane¹, and Antonio Anax F. de Oliveira¹

¹Lhasa Limited, Leeds, West Yorkshire, United Kingdom

The use of rodent studies for predicting human carcinogenicity has been questioned with respect to relevance in many domains, prompting moves to investigate alternative methods. A weight-of-evidence approach utilising multiple pieces of data to reach a conclusion has been suggested. In the ICH S1B(R1) addendum, using evidence gathered throughout the pharmaceutical development process and framing the data in terms of six factors has been proposed in an effort to reduce the animals required for this guidance. Knowledge of modes-of-action and data gaps are also required to determine human relevance and certainty, hence a framework to support this assessment will be useful. Adverse outcome pathway (AOP) provide the perfect means of supplying this knowledge. Using the structure this concept provides, an approach to assess the evidence for ICH S1B(R1) to aid decision-making has been developed, based on 60 assays and 351 predictions organised around 37 AOPs relating to cancer. This method has the advantage of being able to give consistent results as knowledge and data are logically organised, decreasing uncertainty in outcomes. Carcinogenicity calls can be made at three levels – 1) overall outcome based on totality of evidence in the AOP network; 2) individual factor results based on relevant AOP network subsets; and 3) individual AOP calls. All three levels can aid the decision-making process in the S1 assessment. This provides transparency in decisions and allows for the probing of evidence in expert review to ensure outcomes are scientifically robust.

P511: De-risk Antibody-Based Therapies by Identifying Off-target Reactivity

Diana M. Norden¹, Rachel H. Fong¹, Jonathan T. Sullivan¹, Daniel B. Harmon¹, Michael A. Phelan¹, Talia Marano¹, Carmen T. Navia¹, and Benjamin J. Doranz¹

¹Integral Molecular, Philadelphia, PA, USA

Rigorous specificity analysis is critical for the development of antibody-based therapies, as even minimal off-target binding can lead to toxicity and clinical failures. Long believed to be exquisitely specific, recent preclinical data and our own work indicate that monoclonal antibodies (MAbs) frequently (∼25%) display cross-reactivity. In many cases, off-target interactions occur with unrelated proteins that cannot be predicted by protein sequence homology. Cross-reactivty can lead to serious or even life-threatening consequences especially when MAbs are configured as bispecifics, antibody-drug conjugates (ADC) or CAR-T cell therapies. Tissue cross-reactivity (TCR) studies have traditionally been used to screen for off-target binding, however, with poor predictive value for in vivo safety and toxicity. We developed the Membrane Proteome Array (MPA) platform to de-risk MAb-based therapeutics by testing for specificity across 6,000 human membrane proteins expressed in live cells. In contrast to TCR studies, proteins in the MPA exist in their native conformations and are not altered by fixation. The MPA assesses binding interactions by high-throughput flow cytometry allowing for high sensitivity detection and rapid analysis. We will discuss the importance of early-stage specificity testing to expose possible toxicities and present case studies for antibody and cell therapy profiling. MPA data have been used in numerous successful regulatory applications and may be used to replace or complement other cross-reactivity studies.

P512: Inclusion of Histopathology in Dose Range-Finding Nonclinical Studies for Inhaled Drug Products

Emily A. Resseguie¹ and Helen Palmer²

¹Labcorp Early Development Laboratories Inc., Somerset, NJ, USA

²Labcorp Early Development Laboratories Limited, Huntingdon, Cambridgeshire, United Kingdom

Drug development is a lengthy process involving steps that ensure the health and safety of future patients. Nonclinical study designs are based on informed decisions utilizing known information of the mechanism of action and class of molecule and are highly influenced by the route of administration. Clinical observations, clinical pathology, and macroscopic pathology in dose range-finding (DRF) studies generally provide sufficient information to select doses for pivotal studies with most delivery routes. Inhaled drug candidates are recognized for producing adverse effects at the microscopic level that are otherwise not predictable; therefore, unlike other routes of administration, inhalation DRF studies typically include histopathology of the respiratory tract. Although histopathology evaluations can add several weeks to the IND application timeline along with additional costs, it supports the proper dose selection required to provide an adequate safety margin and avoid additional studies and animal usage by ensuring achievement of a NOAEL in the pivotal studies. Therefore, DRF inhalation studies initiated from 2018 to 2020 at Labcorp were reviewed to determine whether inclusion of histopathology on preliminary inhalation studies added value for subsequent dose selection. Histopathology findings in the DRF impacted dose selection in pivotal inhalation studies for approximately 35% of rat studies and 25% of dog studies. This review identified histopathology findings in rat, dog, and monkey that support continued inclusion of respiratory tract histopathology in DRF studies. Future investigations will evaluate potential surrogate endpoints for these findings, which would reduce nonclinical drug development timelines by several weeks.

P513: Evaluation of Non-Clinical Toxicity Studies of COVID-19 Vaccines

Natalia Schilder¹, Gitte Tiesjema², Peter Theunissen², Katrien Oude Rengerink³, and Jan Willem van der Laan²

¹Division of Toxicology, Leiden Academic Center for Drug Research, Leiden, South-Holland, The Netherlands

²Section on Pharmacology, Toxicology and Kinetics. Medicines Evaluation Board, Utrecht, Utrecht, The Netherlands

³Section of Methodology. Medicines Evaluation Board, Utrecht, Utrecht, Utrecht, The Netherlands

Absence of a relevant animal disease model and lack of the use of supportive studies have limited the process of rapid development of SARS-CoV-2 vaccines. This study has evaluated how the choice of species selection and study design affected the outcomes of the non-clinical toxicity studies of the various SARS-CoV-2 vaccine concepts. The study has been focused on safety aspects i.e. on the Developmental and Reproductive Toxicity (DART) and Repeated Dose Toxicity (RDT) studies. Rat and rabbit were the most commonly used species in preclinical toxicity testing. Some COVID-19 vaccine candidates show a DART study design suboptimal for antibody transfer during lactation. The toxicity studies showed no serious adverse effects. Minor effects were expected, i.e. local reactogenicity, immune response and macroscopic findings at the injection site. In addition, comments received during the EMA assessment of the vaccines have been evaluated and consisted most frequently of commentary on study design, species selection and missing data regardless of the utilized vaccine concept. Use of supportive platform studies often substantiated the commentary on these main three categories. Animal model-based toxicity testing has shown limited value in establishing safety of the vaccines, and, more importantly, low translational value in supporting clinical development. From a 3R perspective sponsors are encouraged to focus on products from the supportive platform, both with respect to DART and RDT studies. Regulatory emphasis on data obtained from vaccines with the same platform technology data can be used to support marketing approval of new vaccines.

P514: Histopathology and Biodistribution in NCG Mice Engrafted with Mobilized CD34⁺ Human Hematopoietic Stem and Progenitor Cells

Viktorija Smutova¹, Angela Holguin Moreno¹, Malcolm Gains¹, Camila Pará¹, Catherine Spickler¹, Renée Riffon¹, Selly Hung¹, Jenny Rowe², Steve Festin², Simon Authier¹, and Alan M. Hoberman³

¹Charles River Laboratories, Laval, Quebec, Canada

²Charles River Laboratories, Wilmington, MA, USA

³Charles River Laboratories, Horsham, PA, USA

NCG triple immunodeficient mice on a NOD/Nju background, in which Prkdc and Il2rg genes have been edited by CRISPR/Cas9, lack functional/mature T, B, and NK cells, and have reduced macrophage and dendritic cell function making them an ideal model for cell therapy studies. Herein, we report the biodistribution of mobilized human CD34⁺ hematopoietic and progenitor cells (hHSPCs) by qPCR using human Alu sequences and the histopathological profile in the engrafted NCG mice. Mice were myeloablated (irradiation or Busulfex®) and 1.0 x 10⁶CD34⁺ hHSPCs/mouse, PBS (negative control) or 2.0 x 10⁶HL-60 cells/mouse (positive control) were injected intravenously and followed up for 20 weeks. No Alu (human) sequences were detected in the control animals, while in treated mice it was detected in all tested tissues. High levels were detected in spleen, liver, lung and bone marrow and lower levels in brain, heart, testes, ovaries and blood. The highest concentration was in spleen (1.54 x 10⁵ ± 8.84 x 10⁴ pg/500ng matrix) and the lowest in testes (1.87 x 10² ± 3.80 x 10² pg/500ng matrix). In mice administered HL-60 cells, tumorigenesis in various tissues was associated with mortalities. There were no neoplasms associated with mobilized CD34⁺ hHSPCs. In the spleen, mice engrafted with mobilized CD34⁺ hHSPCs had round cells in the white pulp, and in the bone marrow and decreased cellularity of the hematopoietic compartment. The biodistribution of Alu sequences along with the absence of tumorigenesis supports the relevance of the NCG mice for cell therapy efficacy and safety assessments.

P515: Association of Circulating Cytokine Levels and Clinical Signs in Cynomolgus Monkeys (Macaca fascicularis) Administered with Novel Therapeutics

Lu Zhao¹, Xiefan Fang¹, and Bradley M. Saville¹

¹Charles River Laboratories, Reno, Nevada, USA

The liability of cytokine release syndrome is commonly assessed in cynomolgus monkeys by observing clinical signs and measuring circulating cytokines. However, the relationship of cytokine changes and clinical observations is largely unknown in monkeys. We analyzed data from 101 nonclinical studies conducted from 2019 to 2021. A total of 319 cases had animals presented clinical signs along with cytokine release after intravenous administration of novel pharmaceuticals. The incidences were 43%, 30%, 30%, 22%, 9%, and 9% for emesis, decreased activity, red skin, hunched posture, dehydration, and liquid feces, respectively, and ranged from 0.3% to 6% for other clinical signs. These findings were associated with elevations of IFN-γ, IL-10, IL-1RA, IL-6, IL-8, IP-10, MCP-1, and/or TNF-α within 8 hours postdose. Among these cases, 85%, 4%, 8%, and 3% were categorized as mild, moderate (required diphenhydramine or corticosteroid treatments), marked (required euthanasia), and severe (found dead). There were no clear thresholds for each cytokine to correlate with the clinical severity, but the extreme cases generally had higher magnitude of increases. For the marked cases, peak changes averaged 412, 488, 240, 419, 1269, and 172-fold above baseline for IL-10, IL-1RA, IL-6, IL-8, MCP-1, and TNF-α, respectively. For the found dead cases, peak changes averaged 35, 224, 127, 67, 386, and 129-fold above baseline for IFN-γ, IL-10, IL-1RA, IL-6, MCP-1, and TNF-α, respectively. Minor increases ( < 10-fold) in limited cytokines (mostly MCP-1 and IL-1RA) were not sufficient to induce clinical signs. The results can benefit cytokine data interpretation and provide guidance on designing nonclinical studies.

P516: Omics Approach to Cross-Species Drug Testing for Rapid and Reliable Screening and Prediction of Drug-Induced Liver Injury

Catherine Nury¹, Matthieu Porchet¹, Sophie Dijon¹, Oksana Lavrynenko¹, Emmanuel Guedj¹, Yvan Eb-Levadoux¹, Mohamed Amin Choukrallah¹, Donna Busler², Bruno Filippi², Friederike Wenz², Armin Wolf², Nikolai V. Ivanov¹, Kasper Renggli¹, and Bjoern Titz¹

¹PMI R&D, Philip Morris Products S.A., Neuchâtel, Switzerland, Switzerland

²InSphero AG, Schlieren, ZH, Switzerland

The X-Species DILI Validation Consortium is a precompetitive consortium of pharmaceutical companies and technology providers. It aims to develop cross-species drug testing and validation strategies for rapid and reliable screening and prediction of drug-induced liver injury (DILI). Active pharmaceutical ingredients with well-documented cross-species in vivo data are evaluated using a multi-tier test strategy to then compare the in vivo and in vitro reactions in three-dimensional (3D) liver microtissues from the corresponding preclinical animal models. Where relevant, the primary in vitro toxicology endpoints (e.g., cytotoxicity, histological changes, and DILI-specific phenotypes) are supplemented by comprehensive molecular profiling with transcriptomics, proteomics, and lipidomics methods. Here, we present how these omics measurements are integrated into our multi-stage assessment strategy and the initial data from the application of these methods in 3D liver microtissues. Protein changes induced in 3D liver microtissues were quantified by mass spectrometry (MS)-based proteomics analyses using a data-independent acquisition approach. Lipid changes were quantified by MS-based lipidomics, both by direct injection MS (shotgun lipidomics) and a comprehensive liquid chromatography MS/MS panel for lipid mediators. For example, inflammatory activation of 3D liver microtissues (lipopolysaccharide stimulation) led to upregulation of acute-phase proteins (e.g., CRP, SAA1, and SAA2) and altered levels of various lipid mediators. As part of the X-species DILI validation, these high-resolution molecular measurements will likely provide further insights into the common and unique mechanisms of toxicity across different species.

P517: The Prodrug of a Novel Capsid Assembly Modulator (CAM), ALG-000184, Demonstrates a Favorable Nonclinical Toxicokinetic and Toxicology Profile for the Treatment of Chronic Hepatitis B (CHB)

Sucheta Mukherjee¹, Qingling Zhang¹, Dinah Misner¹, Gina Gratz², Zbigniew Wojcinski³, Kha Le¹, Felix Lai¹, Meenakshi Venkatraman¹, Lawrence Blatt¹, Leonid Beigelman¹, Tse-I Lin⁴, and Sushmita Chanda¹

¹Aligos Therapeutics, Inc., South San Francisco, CA, USA

²Inotiv, Mt. Vernon, IN, USA

³Tox & Path Consulting, Hillsborough, NC, USA

⁴Aligos Belgium, Leuven, Belgium, Belgium

ALG-000184 is a prodrug of ALG-001075, a novel potent class-II CAM, which demonstrated excellent physicochemical properties with high aqueous solubility, good oral bioavailability, and efficient conversion to ALG-001075 in all species studied to date. Repeat dose toxicology studies with ALG-000184 were conducted in rats (4-, 13-, 26-weeks) and dogs (4-, 13-, 39-weeks) following daily PO doses and these species were selected due to similarities in metabolic profile compared with humans. Reproductive toxicology studies included embryo-fetal development studies in rats and rabbits and male and female fertility studies in rats. ALG-000184 demonstrated a favorable safety profile in rats and dogs with monitorable and reversible effects, no genotoxic potential nor adverse CNS or cardiovascular effects in safety pharmacology studies. In repeat dose studies, ALG-000184 was generally well tolerated with key histopathological findings in kidneys in both rats (cortical tubular degeneration/dilatation/basophilia; papillary neutrophilic infiltration/ hyperplasia/fibroplasia) and dogs (calculi/pyelitis consistent with the route of elimination of ALG-001075) and systemic inflammation consistent with reactive histiocytosis in dogs. These target organ toxicities occurred at plasma exposures of ALG-001075 that were > 27-fold above clinically efficacious exposures where significant inhibition of HBV DNA is noted in CHB patients. ALG-000184 did not have effects in either rat or rabbit embryofetal development studies up to the highest doses tested. High and linear increases in ALG-001075 plasma exposure were noted over the dose ranges studied, providing a significant exposure margin from clinical doses. Overall, a favorable toxicology profile has supported ALG-000184 to advance into longer-duration clinical trials in CHB patients.

P518: Best-in-Class Antisense Oligonucleotides Against Hepatitis B Virus Using Next Generation Bridged Nucleic Acid Chemistries Significantly Increases In Vivo Efficacy and Reduces Hepatotoxicity in Mice

Dinah Misner¹, Jin Hong¹, Bruno Filippi², Sandra Chang¹, Monika Tu², Gwendoline Wiki², Friederike Wenz², Kusum Gupta¹, Vikrant Gohil¹, Saul Martinez Montero¹, Vivek K. Rajwanshi¹, Tse-I Lin³, David B. Smith¹, Julian A. Symons¹, Lawrence Blatt¹, Leonid Beigelman¹, and Sushmita Chanda¹

¹Aligos Therapeutics, Inc., South San Francisco, CA, USA

²InSphero AG, Schlieren, ZH, Switzerland

³Aligos Belgium, Leuven, Belgium, Belgium

Reducing hepatitis B surface antigen (HBsAg) is key to achieving functional cure in chronic hepatitis B (CHB) patients. Antisense oligonucleotides (ASOs) efficiently reduce HBsAg in both animal models and CHB patients. Hepatotoxicity is a major side effect of ASOs, which is exacerbated with potent, high affinity Locked-Nucleic Acid (LNA) modified ASOs. Next generation Bridged Nucleic Acid (BNA) monomers can reduce hepatotoxicity, while maintaining potency and affinity. We have demonstrated that applying next generation BNA and nucleobase chemistries in LNA ASO gapmers can significantly reduce in vivo hepatotoxicity and improve the therapeutic index in repeat-dose toxicity studies in mice. To further characterize and extend the translation from mice to humans, we explored various in vitro models of toxicity across several HBV ASOs, including cytotoxicity assessment in HepG2.2.15 cells and 3D InSight^TM human liver models (hLiMTs). hLiMTs made with primary human liver cells (hepatocytes and non-parenchymal cell fractions) are well-established to predict clinical hepatotoxicity of small molecules. hLiMTs, similar to the intact liver, freely take up oligonucleotides in the absence of any transfection reagents that results in specific target downregulation and eventual off-target effects that may result in cytotoxicity. Here we report this in vitro model to evaluate ASOs cytotoxicity for the first time, in a transfection-free system, thus providing a more relevant model to assess hepatotoxicity in humans. Data extending the utility of the model will be presented, along with an assessment of additional ASOs to better build the translation between in vitro cytotoxicity with existing clinical data.

P519: Readiness for Safety and PK/BD Studies in Tumor Bearing Models for VSV-GP Oncolytic Virus

Audrey Brenot¹, Amy Mikolaichik¹, Jaewon Kim¹, Richard Dambra¹, Kaitlynn Graca¹, Saeed Akhand¹, Joseph Ashour¹, Adam Vigil¹, and Birgit Fogal¹

¹Boehringer Ingelheim, Ridgefield, CT, USA

Boehringer Ingelheim is expanding into the oncolytic virus space with the vesicular stomatitis virus pseudotyped with the glycoprotein of the lymphocytic choriomeningitis virus (VSV-GP). From the nonclinical safety perspective, the peculiar mode of action of an oncolytic virus requires that toxicology studies are conducted in tumor-bearing animals to capture Pharmacokinetics/Toxicology signals arising from secondary replication of the virus. To ensure a leaner and more agile entry into Phase 1, we have initiated in-house development of rodent tumor model(s) to support safety and PK/BD studies for VSV-GP projects. This poster will present an update on the learnings, tools developed, and agility gained in multiple areas (tumor models, support of viral work, in vitro assays, flexible processes) towards readiness for support of Phase-1 enabling safety and Pharmacokinetics/Biodistribution studies.

P520: Clinical Manifestations of Elevated Cytokine Levels in Nonhuman Primates in Toxicology Studies

Kurt M. Stahl¹, Hua W. Li¹, Linda X. Jin¹, Stephanie L. Phelan¹, Bhaswati Barat¹, Xiaoyu Zhang¹, Jeffrey L. Nordstrom¹, and Jennifer G. Brown¹

¹MacroGenics, Inc., Rockville, MD, USA

The administration of immune-modulating agents to nonhuman primates (NHP) can elicit cytokine release syndrome (CRS), a systemic inflammatory response. There is a high degree of inter-animal variability in the clinical manifestations of elevated cytokines, which can range from no apparent symptoms to mortality. This makes it difficult to identify the cytokines and serum concentration levels that are most closely associated with CRS. To characterize relationships between elevated serum cytokines and effects on the health status of NHP, data were evaluated from 194 cynomolgus monkeys dosed intravenously with T-cell-engaging bispecific molecules and 163 vehicle control monkeys across 20 nonclinical safety studies. Clinical signs consistent with CRS occurred in 35 animals and were associated with mortality for 25 of these animals. The number and severity of clinical signs varied among the animals, but observations of hypoactivity, hunched posture, decreased body temperature, and elevated blood urea nitrogen level were most predictive of mortality even with veterinary intervention. The cytokines most commonly elevated in animals with CRS-related observations included interleukin 6 (IL 6), IL-10, tumor necrosis factor alpha, and interferon gamma. Clinical signs consistent with CRS did not occur in any vehicle control animal and increases in cytokine levels were not observed or were very minimal. An understanding of the cytokine-inducing potential of a test article and the clinical signs of CRS may allow for study design features like postdose IL-6 blockade or pre-determined criteria for veterinary care that can ameliorate the effects of CRS without compromising the safety assessment goals of the study.

P521: Environmental Assessment for Human Drug Approval – An Outdated Technical Framework?

Ifeoluwa A. Bamgbose¹ and Tim A. Verslycke¹

¹Gradient, Boston, MA, USA

Small molecules have been the backbone of human drug development for nearly a century. Market approval of new human drugs in the US and the EU requires an evaluation of environmental safety using a technical framework that was developed more than 20 years ago, primarily with small molecules in mind. In recent years, there has been a growing shift in drug development, including the development of biologics and new chemical modalities. The US remains, by far, the most important market for initial entry of new human drugs. The US Food and Drug Administration (FDA) has been approving a growing diversity of new chemical modalities, including small interfering RNA (siRNA) drugs, antisense oligonucleotides, antibody drug conjugates, fluorine atoms and/or nitrogen aromatic heterocycles, pegylated peptides, etc. Just last year, FDA approved 14 new biologics and 27 first-in-class drugs which employ a mode of action that is different from existing therapies. As a consequence of this shift in drug development, the existing technical framework for environmental safety assessment has become outdated and – in some instances – technically inappropriate. Environmental safety assessments increasingly require a modified approach tailored to the unique chemistries and modalities of present-day drugs. In this presentation, we will highlight limitations of the existing framework for drug environmental safety assessment using a number of case examples, including an insoluble substance, a structural analog to NADH and ATP, and an siRNA. On the basis of these examples, a number of recommendations are put forth.

P522: Preclinical Safety Evaluation of Etripamil Nasal Spray in Cynomolgus Macaques (Macaca fascicularis) Supports its Safe Use in PSVT Patients

Johanne Pion¹, Carlos Lopez Mendez¹, Jean-Pierre Moreau², and Douglas Wight³

¹Charles River Laboratories, Laval, Quebec, USA

²Recherche Continuum Research, Rigaud, Quebec, Canada

³Milestone Pharmaceuticals, Montreal, Quebec, Canada

Etripamil is a non-dihydropyridine calcium channel blocker formulated as a nasal spray in Phase 3 trials for atrioventricular nodal-dependent Paroxysmal Supraventricular Tachycardia (PSVT). Systemic and local toxicity following once-weekly intranasal administration of etripamil was evaluated in cynomolgus macaques to support clinical development. Groups of animals (n = 4/sex) were administered etripamil into the left nostril weekly at dose levels of 0 (vehicle), 1.9, 3.8, or 5.7 mg/kg/dose for 26 doses. Persistence, reversibility, and progression of findings were examined following a 28-day recovery period. Clinical signs were transient and were related to the intranasal administration (e.g., nasal discharge, sneezing, etc.) of etripamil. There were no macroscopic or systemic microscopic findings at any dose. Etripamil-related adaptive and reactive local changes affecting the nasal cavity, larynx, and nasopharynx were observed at ≥ 1.9 mg/kg/dose. Minimal to severe dose-dependent nasal epithelial damage was observed, mainly affecting respiratory and transitional epithelium. Following the 28-day recovery period, microscopic changes were confined to the left nasal cavity and nasopharynx. These changes were significantly lower in incidence and severity, with noticeable reversal of the adaptive and reactive changes, indicating partial to complete recovery of the epithelial lining. Based on the lack of systemic toxicity and the minimal and transient nasal changes, the systemic, No Observable Adverse Effect Level (NOAEL) of etripamil in monkeys was the high dose, 5.7 mg/kg/dose. The NOAEL for local toxicity was 1.9 mg/kg/dose. Collectively, these data support the continued development of etripamil as an acute treatment of PSVT.

P523: Nonclinical Safety Assessment of SPVN06, an AAV-based Gene Therapy for the Treatment of Rod-Cone Dystrophies

Mélanie Marie¹, Lucie Churet¹, Gregg Prawdzik¹, Pierre-Axel Vinot¹, Thierry Léveillard², José-Alain Sahel³, Myriam Marussig¹, and Florence Lorget¹

¹SparingVision, Paris, Ile de France, France

²Institut de la Vision, Paris, Ile de France, France

³UPMC Eye Center, University of Pittsburgh, Pittsburgh, PA, USA

Rod-Cone dystrophies (RCD) are inherited retinal diseases characterized by progressive loss of the rod photoreceptors, followed by cone photoreceptor degeneration, eventually leading to total blindness. 71 genes are identified in RCD, mostly affecting the rods. SPVN06 aims at slowing down the degeneration of cones by restoring RdCVF trophic support normally provided by functioning rods, and by promoting RdCVFL antioxidant activity. RdCVF and RdCVFL are encoded in the same AAV-based vector. SPVN06 single subretinal administration is expected to slow down cone degeneration in RCD patients independently of the causative mutation. SPVN06 nonclinical safety was evaluated in cynomolgus monkeys (3M+3F/group) at dose levels ranging from 6E9 to 3E11 vg/eye. Animals were followed up to 3 months. A standard toxicological assessment was conducted including an exhaustive battery of ocular testing. Immunogenicity assessments included total antibodies against AAV and RdCVF/L as well as T-cell mediated toxicity. There were no drug-related systemic effects. SPVN06-related findings were noted at ≥1E11 vg/eye and were limited to the photoreceptors and RPE cells. These findings were dose-related and characterized by reduced ffERG amplitude with microscopic correlates. In absence of an immune response to the transgenes and consistently with their mechanism of action, this toxicity was deemed attributed to the supraphysiological levels of transgenes in a healthy retina. The role of AAV to the observed toxicity couldn’t be completely ruled out but was unlikely a major contributor considering that the uveitis was transient and the presence of immune infiltrate minimum. No SPVN06-related findings were noted up to 6E10 vg/eye.

P524: SPVN06, a Novel Gene-Independent AAV-Based Gene Therapy, Dramatically Slows Down Vision Loss and Protects Cone Morphology in the rd10 Mouse Model of Rod-Cone Dystrophy

Lucie Churet¹, Mélanie Marie¹, Hanen Khabou¹, Deniz Dalkara², José-Alain Sahel³, Thierry Léveillard², and Florence Lorget¹

¹SparingVision, Paris, Ile de France, France

²Institut de la Vision, Paris, Ile de France, France

³UPMC Eye Center, University of Pittsburgh, Pittsburgh, PA, USA

Rod-Cone dystrophies (RCD) are inherited neurodegenerative diseases characterized by an initial loss of rod photoreceptors followed by loss of cone photoreceptors eventually causing blindness. SPVN06 is a novel AAV-based drug candidate encoding both human RdCVF and RdCVFL within the same vector. A single subretinal administration of SPVN06 is expected to protect against cone degeneration in RCD patients independently of the mutated gene. SPVN06 pharmacology was evaluated in the rd10 fast-progressing mouse model of RCD. The animals received, at P15 or P18, a single bilateral subretinal administration of vehicle or SPVN06 (1 µL) at dose levels ranging from 1E8 to 5E9 vg/eye (≥5M+5F/group). Retinal structure and function were evaluated by histology, optokinetic (OKT) and ffERG. A highly significant protection of the visual function (OKT) was noted at all timepoints evaluated (P32, P38 and P45) when the animals received SPVN06 1E8 vg/eye at P18 but not at P15 suggesting that stage of eye development might be critical. No functional improvement was observed for doses ≥3E8 vg/eye but a dose-related retention of ONL nuclei was noted by histology at P48, reaching significance at 5E9 vg/eye along with a protection of the cone morphology seen by cone arrestin labeling. These microscopic changes were not accompanied by functional improvements (OKT or ERG), likely related to the known mouse-specific retinal toxicity in relation to high dose and the presence of a broadly active promoter that possibly prevented functional readout. This study provides strong functional proof of concept supporting SPVN06 clinical development.

P525: Safety Assessment of DPI-4452, a Cyclic Peptide Targeting Carbonic Anhydrase IX

Ines Borrego¹, Frederic Massiere¹, Antoine Attinger¹, Aileen Hoehne², Frank Osterkamp², and Ulf Andersson¹

¹Debiopharm International SA, Lausanne, Vaud, Switzerland

²3B Pharmaceuticals GmbH, Berlin, Berlin, Germany

Carbonic anhydrase IX (CAIX) is a transmembrane metalloenzyme expressed in hypoxic tumors; its expression in normal tissues is limited to the stomach and small intestine. DPI-4452 is a new CAIX-targeting cyclic peptide that can be radiolabeled for targeted radiotherapy. DPI-4452 has a low affinity for rodent CAIX, however, the binding affinity to canine CAIX was similar to humans, allowing dogs as single species for toxicology evaluation in an intravenous extended single-dose toxicity study to support a clinical regimen of once-monthly dosing. Central nervous system, cardiovascular and respiratory safety pharmacology endpoints were examined along with clinical signs, clinical pathology, histology, microscopic evaluation, and toxicokinetics. DPI-4452 was well tolerated up to the highest dose tested (400 µg/kg) without any toxicology and safety pharmacology findings. All treated animals had detectable plasma concentrations three hours after administration, and the kidney rapidly cleared the compound. There was a dose-proportional increase in area under the plasma concentration-time curve without gender difference in exposure. The NOEL was established at 200-fold higher than the predicted human total ligand mass dose of 60 µg for the diagnostic drug and 25-fold for the radiotherapeutic drug dose of 500 µg. The preclinical package also included a dog biodistribution and dosimetry study using radiolabeled DPI-4452. The highest tissue exposure was in the stomach and the small intestine, agreeing with target expression data. The latter was defined as the dose-limiting organ based on ceiling values for external beam radiation therapy. In summary, these safety data fully support the further development of DPI-4452 in clinical trials.

P526: Sarcopenia Investigational Drugs: Nonclinical Trends and Implications for Drug Development

Sree Rayavarapu¹, Daniel Minck¹, C. Lee Elmore¹, and David Carlson¹

¹Division of Pharmacology and Toxicology for Cardiology, Hematology, Endocrinology and Nephrology, Office of Cardiology, Hematology, Endocrinology and Nephrology, Office of New Drugs, Center for Drug Evaluation and Research, U.S. Food and Drug Administration, Silver Spring, MD, USA

Sarcopenia is an age-related loss of skeletal muscle mass, function, and strength that is a substantial clinical problem for millions of older adults which impacts their quality of life. No FDA-approved drugs are currently available for sarcopenia. The authors surveyed Pharmacology/Toxicology reviews of investigational new drug applications (INDs) submitted to the FDA for sarcopenia from 2004-2021. Of the INDs surveyed, 64% were commercial and 36% were research in nature. A significant proportion of commercial INDs (∼55%) were either inactive or withdrawn. Based on the information from surveyed INDs and published literature, the pharmacological classes of investigational sarcopenia drugs included: 1) selective androgen receptor modulators (SARMs), 2) myostatin inhibitors, 3) calcium modulators, 4) immunomodulators, and 5) antioxidants. The major toxicities/target organs observed for each drug class in nonclinical species included: 1) endocrine disruption and liver toxicities for SARMs; 2) gonadal and liver toxicities for myostatin inhibitors; 3) cardiac and renal toxicities for calcium modulators; 4) renal, neurological, and gastrointestinal toxicities for immunosuppressants; and 5) liver, spleen, and bone marrow toxicities for antioxidants. Investigative sarcopenia drugs typically showed anabolic effects in pharmacology models and their toxicological profiles included exaggerated pharmacology in healthy non-sarcopenic animals, yet many drugs appear to ultimately fail due to issues related to clinical efficacy. This survey suggests current nonclinical models of sarcopenia have not been sufficient to identify clinically efficacious drugs and development of more human-relevant and translatable in vitro and in vivo models may be needed to support the development of novel sarcopenia therapies.

P527: Establishment of an Experimentally Induced Hemophilia A Model in Cynomolgus Monkeys

Li Zhan¹, Narine Lalayeva¹, Carsten Dany Ley², Zhizhan Song³, Kevin Zhang³, Jilin Deng³, Drew May¹, Brett Megrath¹, and Julie Forget¹

¹Altasciences, Everett, WA, USA

²Novo Nordisk A/S, Måløv, Denmark

³Prisys Biotech, Shanghai, Shanghai, China

Hemophilia A is a bleeding disorder due to deficiency of FVIII, an essential blood-clotting protein. A tail bleeding model in cynomolgus monkeys with antibody-induced hemophilia was initially developed by Prisys Biotech, then further refined at Altasciences, to evaluate the hemostatic effect of pro-coagulant compounds. Vehicle or polyclonal anti-FVIII antibody was intravenously administered 60 minutes prior to the initiation of the tail bleeding procedure. Blood samples were collected prior to treatment, after administration of the vehicle or anti-FVIII antibody, and before euthanasia. Comparison of activated partial thromboplastin time (APTT) and thromboelastography (TEG) was performed. Bleeding time and hemoglobin loss were evaluated at four intervals from 0 through 40 minutes post initial tail bleeding. No changes in APTT were observed in the control group. Prolongation of APTT approximately doubled by the sheep anti FVIII induction treatment. R time, K value and α-angle in TEG for vehicle control animals were stable at all time points, while induction of hemophilia significantly increased the mean R time, elevated mean K value and lowered the α-angle. The mean total bleeding time and loss of hemoglobin through 40 minutes post vein puncture increased approximately 3.7- and 2.6-fold, respectively, in the hemophilia-induced animals. Induction of hemophilia A with sheep anti-FVIII at a dose level of 16.4 mg/kg resulted in expected perturbations in blood coagulation and clot formation (APTT prolongation, R and K value increase, and α-angle decrease), and substantially increased bleeding time and blood loss, confirming successful establishment of an experimentally induced hemophilia A model in cynomolgus monkeys.

P528: Retrospective Analysis of Neurological Clinical Signs in Control Animals Using Continual Video Monitoring Data

Morgan K. Foret¹, Cristina Ballesteros¹, Jerneja Stare¹, Samantha Sparapani¹, Michelle Stoute¹, Christian Li¹, and Simon Authier¹

¹Charles River, Laval, Quebec, Canada

Baseline levels of neurological clinical signs that are not drug-related can be determined with data from control groups allowing for the characterization of potential toxicological effects of a test article. Importantly, when control groups display neurological clinical signs at unexpected frequencies due to variation between species and animal strains, this can be a confounding factor during interpretation. Establishing a background incidence of neurological clinical signs in control groups is an essential tool for elucidating drug-related CNS clinical signs and discriminating them from species-specific variation in background findings. We previously assessed the incidence of spontaneous neurologically-related clinical observations in control animals using compiled data from 306 GLP-compliant toxicology studies conducted at Charles River Laboratories Laval between 2009 and 2019. Cageside observations have a lower temporal resolution compared to continuous video monitoring. To address this limitation, in the present analysis we compiled neurological clinical observations generated from video recordings in EEG studies at Charles River Laboratories Laval in similar conditions. We present data from control groups of Sprague-Dawley rats, Beagle dogs and non-human primates. Incidence and rate (%) of clinical signs including tremors and myoclonic jerks were identified and serve as a means of improving interpretations of neurologically-related clinical observations. As expected, the daily incidence of spontaneous neurological clinical signs assessed from video monitoring was substantially higher than cageside observations. The data highlights the importance of developing an historical control database under the same conditions as the experimental study to interpret baseline levels of neurological clinical signs.

P529: In Vivo and Histological Analysis of Focal Chorioretinal Defects in Dutch Belted Rabbits

Simone Iwabe¹, Stephanie M. Shrader², Jeffrey Burdick¹, and Gustavo D. Aguirre³

¹Altasciences Preclinical Scranton, Scott Township, PA, USA

²StageBio, Mount Jackson, VA, USA

³School of Veterinary Medicine, University of Pennsylvania, Philadelphia, PA, USA

The objective of this study was to examine the in vivo retinal microanatomy of findings observed during pretest ophthalmic examination in laboratory rabbits. Seven to eight months-old male Dutch Belted rabbits were evaluated after complete ophthalmic examinations that included slit lamp and indirect ophthalmoscopy, fundus photography and cSLO/AF-OCT. Eyes presenting with lesions were collected and evaluated microscopically after in vivo imaging. Three rabbits presented unilateral, oval-pinkish lesions superior to the optic nerve head. Lesions were autofluorescent under AF mode, and OCT showed focal retinal detachment, autofluorescent material, likely lipofuscin, accumulated between photoreceptor and RPE layers. The fourth rabbit presented with a unilateral, oval-light brown lesion at the superior temporal quadrant. This lesion was not autofluorescent under AF mode, and OCT showed normal retinal layers, and focal choroidal atrophy. This correlated microscopically with localized scleral and choroidal thinning. In conclusion, fundus abnormalities often observed during pretest ophthalmic examination and broadly categorized as chorioretinal scars have been further characterized to identify specific microarchitecture changes. This is the first report correlating observations from in vivo ophthalmic examinations with microanatomical findings in Dutch-Belted rabbits.

P530: Killing Two Flies with One Strike: Using Human IgG Immunetolerant Mice for Safety Testing to Overcome Immunogenicity against Human Antibodies and Supporting 3Rs

Juliana Bessa¹, Wolfgang Richter¹, Pierre Maliver¹, Hans P. Grimm¹, and Ulrike Hopfer¹

¹Roche Innovation Center Basel, Roche Pharma Research & Early Development, Basel, Basel, Switzerland

The production of anti-drug antibodies (ADAs), may affect drug pharmacokinetics, diminish the pharmacological action of a therapeutic antibody, or cause adverse reactions not only in the clinical but also in the nonclinical setting. While the elimination of non-human sequences reduces ADA mediated immune responses in patients, human(ized) sequences are more likely to induce a xenogeneic immune response in preclinical in vivo studies. We investigated whether cross species immunogenicity in safety studies could be mitigated by the use of human IgG immune tolerant transgenic mice. We used an IgG1 fused interleukin-2 cytokine molecule (IgG-IL2), to test whether immune tolerance can be induced in this mouse model for potential use in chronic toxicity- and / or pharmacology studies. IgG-IL2 is a dimeric IL-2 mutein with enhanced avidity to the high-affinity trimeric IL-2R αβγ receptor that enhances T-regular cell-expansion without broad immune-stimulatory effects. While the molecule is pharmacologically active in mice and cynomolgus monkeys, high immunogenicity was observed in both species. The use of this model revealed a greatly improved immunogenicity profile compared to wild-type mice with no ADA development against the huIgG part of the molecule. Significant IgG-IL2–related expansions in CD4+FOXP3+ Treg cells were observed in blood and lymphoid organs. Biomarkers indicative of Treg activation and proliferation were dose dependently upregulated. Histopathology findings were fully consistent with the expected pharmacologic activity of IL-2 in mice, monkeys, and humans. While ADA development was greatly reduced decreasing exposure by target mediated disposition and target upregulation was seen over time.

P531: Prediction of Chemotherapy-Induced Peripheral Neuropathy Using Neurite Outgrowth: Comparison of Sensitivity Using Rat Cortical Neurons, IPSC Derived Human Neurons, and Rat Dorsal Root Ganglion Neurons

You Feng¹, Jenifer Bradley¹, Ben MacDonald¹, Gabrielle Garza¹, and Christopher J. Strock¹

¹Cyprotex US LLC, An Evotec Company, Watertown, MA, USA

Chemotherapy-induced peripheral neuropathy is a common side effect of modern chemotherapeutics and has significant effects on long term quality of life. There are currently limited tools used to effectively evaluate this side effect such as animal models, but these are labor intensive and may be difficult to interpret. Here we compare the responses of three cell types for their ability to predict peripheral neuropathy: rat cortical neurons, IPSC derived human neurons, and Rat Dorsal Root Ganglion Neurons. We compared these cell types using a group of chemotherapeutics from different classes to determine if all mechanisms can be predicted with these assays and if the effects are concentration dependent. For this comparison, we used a standard neurite outgrowth assay with 72-hour treatment beginning on the day of plating. All of the cell types show responses to the same chemotherapeutic classes with the taxanes, epothilones, and microtubule interfering agents inducing the most sensitivity. They all showed greater sensitivity to the neurite outgrowth endpoint than cytotoxicity. The IPSC derived human neurons and the rat cortical neurons have comparable levels of neurite outgrowth sensitivity. The rat dorsal root ganglion neurons have the greatest degree of neurite outgrowth sensitivity (both number and IC₅₀) to the chemotherapeutic agents on whole. Overall, the rat DRG neurite outgrowth assay was very effective at predicting peripheral neuropathy, but additional assays will need to be developed to identify more classes of chemotherapeutics which did not respond in this assay.

P532: Safety Review of Residual CRISPR-Cas Components in Biologics – A Toxicological Perspective

Isaac Mohar¹ and Archit Rastogi²

¹Gradient, Seattle, WA, USA

²formerly Gradient, Boston, MA, USA

Gene editing technologies like CRISPR-Cas are finding broad applications in biologics. CRISPR-Cas is often used ex vivo to engineer cells using a Cas enzyme (e.g., S. pyogenes [Sp] Cas9) complexed with a specific single-guide RNA (sgRNA) to form a ribonucleoprotein (RNP). The potential risks of residual CRISPR-Cas associated with a cell-based drug product were evaluated based on a comprehensive literature review of regulatory guidance, toxicokinetic data (4 studies), and nonclinical (14 studies) and clinical (4 studies) toxicology data. Immunogenicity and unintended gene editing (e.g., mutagenicity, double-strand breaks) were identified as the highest priority hazards. Based on published data, immunogenicity was determined to be a low and controllable toxicity risk and unintended gene editing was determined to be improbable for free RNP but would require additional consideration if the RNP was encapsidated. The available data show that free CRISPR-Cas components have a short half-life in serum (Cas9-RNP ∼ 1.5 h, sgRNA ∼ 2 h) without apparent acute or chronic systemic toxicity hazards. Notably, data were lacking to evaluate potential reproductive toxicity. Available toxicology data in non-human primates for encapsidated modified sgRNA supported a NOAEL of 1.5 mg/kg-bw; this dose was used to derive an exposure limit of 0.05 mg/kg-bw sgRNA per drug product dose. The data support using the proposed limit for sgRNA as a benchmark for residual RNP, with additional consideration if the RNP was encapsidated. Overall, this review supports concluding that free residual CRISPR-Cas would not pose an unacceptable toxicological risk to patients when controlled at the proposed levels.

P533: Selective Latent TGFβ1 Inhibition Exhibits Improved Nonclinical Safety Profile

Ajai Pal¹, Marcie Wood², Mania Kavosi¹, Shaun M. Cote¹, Ashish Kalra¹, Samantha B. Nicholls¹, Justin W. Jackson¹, George Coricor¹, Thomas Schurpf¹, and Gregory J. Carven¹

¹Scholar Rock, Inc., Cambridge, MA, USA

²ToxStrategies, Inc., Katy, TX, USA

Perturbation of the TGFβ signaling pathway has been implicated in the pathogenesis of diseases such as connective tissue disorders, fibrosis, and cancer. The pan-TGFβ inhibition approach is associated with safety liabilities. We have previously shown that a selective latent TGFβ1 inhibitor (SRK-181) has an improved safety profile compared to pan-TGFβ and is associated with immune cell engagement (Welsh et al. International J. Toxicology 2021). While this profile is advantageous for therapeutics in immuno-oncology, it may not be favorable for chronic diseases like fibrosis. Here, we hypothesize that selective targeting of extracellular matrix associated TGFβ1 would minimize effects on the immune system, allowing for chronic anti-fibrotic therapy. To that end, we have developed a context-selective antibody, LTBP-49247, that selectively inhibits TGFβ1 large latent complexes, only in the context of LTBP1 and LTBP3 and does not bind to TGFβ1 presented by immune cells via GARP or LRRC33. We have also developed a context-independent TGFβ1-37000 antibody. In contrast to LTBP-49247, TGFβ1-37000 is selective for the TGFβ1 isoform across all large latent complexes. We evaluated the toxicity profile of these selective antibodies in the chronic setting in animal models. Our data suggests that selective inhibition of latent TGFβ1 provides an improved preclinical safety profile relative to pan-TGFβ inhibition. Further, we have developed molecules with different selectivity profiles for TGFβ large latent complexes which allow for the option of immune cell engagement, as might be desired in an oncology setting, or immune avoidance which may be promising for chronic diseases such as fibrosis.

P534: Translation of DILI from Regulatory Animals to Human Using 3D Liver Microtissues – Results of the International X-Species DILI Validation Consortium

Armin Wolf¹, Bjoern Titz², Jonathan P. Jackson³, Kasper Renggli⁴, Leah Norona⁵, Piyush Bajaj⁶, Philipp Hewitt⁷, Prathap KS. Mahalingaiah⁸, Samantha Faber⁹, Simon Plummer¹⁰, Richard Brennan⁶, Friederike Wenz¹, and Bruno Filippi¹

¹InSphero , Schlieren, CH, Switzerland

²Philip Morris International R&D, Neuchâtel , CH, Switzerland

³Pfizer, Groton , Connecticut , USA

⁴Philip Morris International R&D, Neuchâtel, CH, Switzerland

⁵Genentech Inc., South San Francisco, California , USA

⁶Sanofi, Framingham , Massachusetts, USA

⁷Merck KGaA, Darmstadt, D, Germany

⁸AbbVie, North Chicago, Illinois , USA

⁹Takeda Development Center Americas, Inc, San Diego , California , USA

¹⁰MicroMatrices , Dundee, Scotland , United Kingdom

Many pharmaceutical programs have been discontinued due to narrow safety margins that were not adequately predicted by animal models, leading to a loss of promising, innovative therapeutic concepts. Because species differences may skew the effective prediction of drug-induced liver injury in humans, additional translational tools are needed for internal decision-making on the fate of drug candidates. Here we propose the 3D liver microtissue MPS platform consisting of cocultures of primary liver cells, Kupffer cells, and liver endothelial cells from rat, dog, cynomolgus monkey and human to assess the translatability of liver adversities observed in preclinical animal studies. A pre-competitive industry consortium of 6 pharma and 3 technology companies was formed to qualify the 3D in vitro platform by recapitulating the corresponding species-specific effects observed in vivo. Each pharma company provided compounds to the consortium with species-specific effects, along with well documented preclinical liver pathology and DMPK properties. All compounds will be evaluated in a three-tiered process. Goal is to recapitulate the in vivo liver toxicology at one of the 3 levels. Tier 1: Evaluation of necrosis based on cellular ATP levels and LDH leakages. If in vivo necrosis cannot be recapitulated by 3D liver microtissues then Tier 2: Evaluation of species-specific liver pathological phenotypes or the DILI-specific mechanisms, like impaired bile acid secretion or reactive metabolite formation. If negative then Tier 3: DILI pathway information by transcriptomics, proteomics and metabolomics analyses and histological information through tissue microarrays. The poster will provide a status update on ongoing studies of 40 compounds.

P535: Toxicity Assessments with Adeno-Associated Virus (AAV) Vectors for Gene Therapy: A Compilation of Study Results Using Multiple Dosing Routes

Srishti Vats¹, Philippe Ancian¹, Sylvie Bonavaud¹, Steve Bulera¹, Stefania Cinquino¹, Sarah Davis¹, Barak Gunter¹, Shawna Jackman¹, Christopher Papagiannis¹, Lucie Reynaud¹, Melinda Tyner¹, Julie Vein¹, Julie Veyssiere¹, Mark Vezina¹, Scott Wilson¹, and Simon Authier¹

¹Charles River Laboratories, Wilmington, MA, USA

Adeno-associated virus (AAV) gene therapy has proven to be a landmark development for effective and long-term transgene expression. However, these AAV properties can only be transmutable to the clinical setting after careful assessment of associated short- and long-term toxicity. AAV therapy associated clinical signs, histopathological and clinical chemistry findings, effect of dose route and optimal dose level ranges necessitate a comprehensive evaluation of non-clinical data to determine conditions for safe clinical use. Compiling pre-clinical non-human primate AAV study data encompassing multiple dose routes (Intrathecal, Intracerebroventricular, Intracisternal, Intramuscular, Intravenous and Intravitreal) acquired from across five (5) test sites over a span of ten (10) years has provided insights into AAV associated toxicity. Most post dose clinical signs include, but are not limited to, tremors, reduced appetite and ataxia often resolving by study completion. Day 4 to 8 assessments exhibited an average of 2.7x fold increase in Alanine amino-transferase (ALT) and a 1.7x fold increase in Aspartate amino-transferase (AST) levels as expected given the tropism of the vectors with recovery to pre-dose comparable levels within 2-3 weeks depending on species. Body weights were generally unaffected and treatment associated pre-terminally (unscheduled) fated animals represented 0.05%. Dose levels ranging from 1.00E+11 to 1.00E+14 vg were generally evaluated with trends toward dose- and route-dependent toxicology profiles. Dorsal root ganglia (DRG) were commonly considered as a target organ of toxicity, especially with central nervous system delivery routes. Transgene expression was considered a potential putative cause for some histopathological changes with individual variability typical for immunogenic reactions.

P536: A Cross-Pharma Benchmark of the Current Preclinical Safety Assessment Practices to Support Development of Targeted Protein Degraders: Results from an IQ Survey

Katie Stamp¹, Michelle Hemkens², Lise Loberg³, Kevin Moreau⁴, Doris Zane⁵, and Tim Hart⁶

¹Bristol Myers Squibb, Summit, New Jersey, USA

²Pfizer, San Diego, California, USA

³AbbVie Inc., North Chicago, IL, USA

⁴AstraZeneca, Cambridge, Cambridgeshire, United Kingdom

⁵Gilead Sciences, Foster City, CA, USA

⁶GSK, King of Prussia, Pennsylvania, USA

The novel small molecule platform of targeted protein degraders reached a milestone in 2019 when the first Phase 1 clinical trial using the platform was initiated in advanced cancer patients. Since then, more targeted protein degraders have entered clinical trials in several indications. Using theoretical ADME and safety concerns as a framework, the IQ Consortium Protein Degrader Working Group conducted two surveys to benchmark the current preclinical practices for targeted protein degraders, to inform in a more rigorous and effective way, the nonclinical safety and ADME assessments of degraders. Responses provided insight into potential screening gaps and best practices for nonclinical safety and ADME assessments, preIND/IND regulatory feedback, and human starting dose calculations that can be applied by companies new to or with limited experience in the field. The current experience with targeted protein degraders in the nonclinical and clinical stages of development was collated, and the perceived hazards versus observed risks within in vitro and in vivo safety experiments were discussed. Modification of standard in vitro toxicity assays included assessment of target protein degradation and extended assay duration. In vivo assessments were generally modified with different approaches aiming at answering questions specific to the platform. Areas identified for further investigation include species selection and recovery periods. The survey results suggest that the safety evaluation of targeted protein degraders does not fundamentally differ from other small molecules and proposes a fit-for-purpose flow scheme for preclinical safety evaluation.

P537: Actions of Myosin Modulators on Contractility and Ca²⁺ Transient in the Adult Human Heart

Ky Truong¹, Alexa Stafford¹, Yannick Miron¹, Guy Page¹, Andre Ghetti¹, and Najah Abi-Gerges¹

¹AnaBios Corporation, SAN DIEGO, California, USA

Acceleration or inhibition of the actin-myosin complex with myosin ATPase modulators (myotropes) can lead to increased or decreased myocardial force production, respectively. Thus, identifying the human effect of myotropes during preclinical development can aid in the determination of drug-induced contractility risk. We used adult human primary cardiomyocytes from organ donors to evaluate the effects of 7 myotropes on contractility and Ca²⁺ transients at 1Hz. Treatment effects on contractility and Ca²⁺ amplitude were expressed relative to the myocyte’s specific baseline control. We report that negative myotropes exerted concentration-dependent decreases in contractility (Mavacamten; IC₅₀ = 0.2µM), N-benzyl-p-toluene sulphonamide (IC₅₀ = 16µM), Blebbistatin (BBS, IC₅₀ = 2.6µM)). Unlike BBS, derivatives of BBS, HydroxyBBS, (S)-3'-aminoBBS and Para-aminoBBS, caused biphasic effects on contractility. Omecamtiv Mecarbil, myosin ATPase activator, increased contractility with EC₅₀ value of 0.6µM. All 7 myotropes showed no potential to alter Ca²⁺ amplitude even at concentrations that had maximal effect on contractility. For example, 0.3µM Mavacamten and 2µM Omecamtiv Mecarbil changed the Ca²⁺ amplitude by -3% and -10%, respectively. In contrast to myotropes, a β-adrenoceptor agonist (Isoproterenol 0.3µM) and a Ca²⁺ channel blocker (Verapamil 10µM) increased (122%) and decreased (60%) Ca²⁺ transient amplitude, respectively. Thus, adult human cardiomyocytes provide a useful translational strategy for the early assessment of new myotrope candidates with a novel mechanism of action.

P538: A Non-clinical Animal Model to Predict Immunogenicity of Biological Drug Products

Adetunji Fayomi¹, Hatib Kaira¹, Robyn Becker¹, Zhuoming Liu¹, and Kristina Howard¹

¹U.S. Food & Drug Administration, Silver Spring, MD, USA

Immunogenicity of biological drug products remains an important safety issue for both originator and biosimilar drug products. Standard animal models, such as rodents, recognize these products as foreign as most are humanized biological proteins. Many approaches are being used to address the need to better predict immunogenicity including in silico and in vitro methods. To address this regulatory challenge, we utilized bone marrow-liver-thymus (BLT) immune humanized mice that develop a fully engrafted human immune system. Owing to a human thymus, these mice recognize human proteins as self and may improve our ability to predict immunogenicity to biological drug products. To validate the ability of these mice to detect immunogenicity we conducted an eight-week study in which mice were dosed with the biologic once weekly. We evaluated mouse EDTA-anticoagulated blood and serum every two weeks for phenotype and activation markers via flow cytometry and for antibody production, to assess responses to infliximab and two approved infliximab biosimilars. Infliximab is known to induce anti-drug antibodies in patients. We found overall increased levels of all IgG isotypes (1-4), to infliximab and its biosimilars in many treated mice, but not control mice. In addition, we identified immune activation in T- and B-cells in treated but not control mice. Of note, we found that the two biosimilars elicited differential responses in certain donors, suggesting that subtle differences in the molecules may be identified by specific HLA types. This study shows that BLT-humanized mice may be useful for immunogenicity evaluation of biological drug products.

P539: Human Donor Variability Impacts the Ability of BLT-Humanized Mice to Produce Robust Humoral Responses

Jungeun Sung¹, Robyn Becker², Hatib Kaira², Kathleen Gabrielson³, AGM Mostofa², Katherine Shea², Alan Knapton², and Kristina Howard²

¹U.S. Food and Drug Administration, Silver Spring, MD, USA

²U.S. Food and Drug Administration, Silver Spring, MD, USA

³Johns Hopkins University, Baltimore, MD, USA

Biological drugs can elicit unwanted immune responses affecting patient safety and drug efficacy. Immunogenicity has been difficult to model because human immune responses are highly variable. This study evaluated donor-specific variability in humanization and humoral responses in BLT-humanized mice. We produced bone marrow-liver-thymus (BLT) immune humanized mice using seven unique donors with four strains of mice (NCG, NOG, NOG-EXL, and NOG-IL6). Following humanization, BLT-humanized mice were injected with either keyhole limpet hemocyanin (KLH), tetanus toxoid, hepatitis A vaccine or saline to evaluate B cell maturation and antibody production. Our results show donor-dependent effects on humanization in all mouse strains prior to study. We also observed treatment effect as well as donors influence on major immune cell reconstitution during eight weeks study period. Histology shows lymph nodes of treated mice had greater numbers of plasma cells whereas saline mice had more quiescent lymph nodes when humanized mice produced with the same donor was compared. Additionally, flow cytometric analysis shows IgM secreting plasma cells were significantly reduced in tetanus toxoid, and hepatitis A vaccine groups, compared to saline mice. Human donor impacts on IgM and IgG producing memory B cells and plasma cells reflects the variable adverse immune responses in human populations. This study shows donor impact on T-cell dependent antibody responses to immunogenic compounds in humanized mice, suggesting that they may be useful for investigation of immunogenicity to biological drug products.