TOXI 1. Geisen, Susanne ETH Zürich
Chemical mechanism of O6-carboxymethyldeoxyguanine formation from azaserine and abundance in cells
Azaserine, an antibiotic and anticancer agent, is widely used as a model chemical for inducing DNA adducts, including the pro-mutagenic adducts O⁶-carboxymethyldeoxyguanine (O⁶-CMdG) and O⁶-methyldeoxyguanine (O⁶-MedG). These adducts are relevant as potential biomarkers associated with carcinogenesis resulting from exposure to N-nitroso compounds. Despite its use since 1977 in studies concerning DNA damage, knowledge of the chemical basis of DNA adduct formation is limited. It has been hypothesized that azaserine is converted to diazoacetate, which is responsible for carboxymethylation of DNA. Two potential mechanisms for this process have been suggested: β-elimination of azaserine to diazoacetate and pyruvate, or direct hydrolysis of azaserine to diazoacetate and L-serine. To elucidate the chemical mechanism of azaserine-induced O⁶-CMdG formation we studied the stability of azaserine and its reactivity with deoxyguanine (dG) as a function of pH. Further, to quantify levels and persistence of azaserine-induced DNA damage, we established a nanoflow liquid chromatography tandem mass spectrometry approach for simultaneously quantifying O⁶-MedG and O⁶-CMdG. The data suggest that azaserine is highly stable from pH 2 to 11. Furthermore, we have gained, for the first time, evidence of the formation of a dG-azaserine intermediate, which appears to spontaneously decompose to form O⁶-CMdG. This observation suggests that a direct reaction of azaserine with dG is involved in O⁶-CMdG-formation, rather than the spontaneous breakdown of azaserine to diazoactate followed by reaction with DNA. Finally, we evaluated the persistence of O⁶-CMdG and O⁶-MedG in human colon epithelial cells (HCEC) exposed to azaserine under simulated chronic exposure conditions, and found that maximal O⁶-CMdG adduct levels are reached after 48 h exposure. These results provide fundamental chemical insight concerning the basis of how azaserine gives rise to DNA damage and supports its use as a model compound in carcinogenesis research, particularly concerning the identification of factors that mitigate adverse effects of O⁶-CMdG.

TOXI 2. Li, Yupeng University of Minnesota
Investigation of 2′-Deoxyadenosine-derived adducts specifically formed in rat liver by N’-nitrosonornicotine (NNN) metabolism
The International Agency for Research on Cancer has classified the tobacco-specific nitrosamine N‘-nitrosonornicotine (NNN) as “carcinogenic to humans” (Group 1). To exert its carcinogenicity, NNN requires metabolic activation to form reactive intermediates which alkylate DNA. Previous studies have identified P450s-catalyzed 2′-hydroxylation and 5′-hydroxylation of NNN as major metabolic pathways, with preferential activation through the 5′-hydroxylation pathway in some cultured human tissues and in the patas monkey. So far, the only adducts identified from NNN 5′-hydroxylation in rat tissues were Py-Py-dI and Py-Py-dN. To expand the DNA adduct panel formed by NNN 5′-hydroxylation and identify possible activation biomarkers of NNN metabolism, we investigated the formation of dAdo-derived adducts using a new highly sensitive and specific LC-NSI-HRMS/MS method. Two types of NNN-specific dAdo-derived adducts Py-THF-dAdo and Py-Py(OH)-dN were observed for the first time in calf thymus DNA incubated with 5’-acetoxyNNN. More importantly, Py-Py(OH)-dN was also observed in relatively high abundance in the liver DNA of rats treated with different doses of racemic NNN in the drinking water for 3 weeks. These new adducts were characterized using authentic synthesized chemical standards. Both NMR and mass spectrometry data agreed well with the proposed structures of Py-THF-dAdo and Py-Py(OH)-dN. Reduction of both adducts formed the corresponding products N⁶-PHB-dAdo and Py-Py-dN, which were readily observed by mass spectrometry. Optimization of the LC method successfully resolved the four diastereomers of Py-THF-dAdo. However, the broad peak corresponding to Py-Py(OH)-dN seemed to contain multiple isomers in equilibration with its open chain form N⁶-OPB-dAdo, preventing accurate quantitation of this adduct. The quantitation of its reduced form Py-Py-dN is currently in progress. The NNN-specific dAdo adducts Py-THF-dAdo and Py-Py(OH)-dN formed by NNN 5′-hydroxyaltion provide a more comprehensive understanding of the mechanism of DNA adduct formation by NNN.

TOXI 3. Terrell, John Vanderbilt University
Covalent binding of anthracyclines to AP sites in DNA suggest an unstudied mechanism of cytotoxicity from chemotherapeutic regimens
Abasic (AP) sites are a common form of DNA damage and a base excision repair intermediate. AP sites are found in high abundance in cells, and levels can be increased by toxicant exposure and chemotherapeutic regiments. Previous studies have shown that the anthracycline Doxorubicin (DOX) can covalently modify DNA in the presence of formaldehyde, which forms a methylene bridge between the N2-position of guanine and the amino group of DOX. The combination of DOX and formaldehyde shows enhanced cytotoxicity in cell culture, even against DOX resistant cells. We report here that Mitoxantrone (MTX), a related anthracycline anticancer agent, can form a covalent conjugate with AP-sites in model DNA systems. The MTX-AP site conjugate was characterized by ESI LC-MS/MS analysis after NaB(CN)H₃ reduction of the imine intermediate. We hypothesize that this previously unrecognized reactivity contributes to the cytotoxic mechanism of MTX, DOX, and related anthracyclines.

TOXI 4. Ahmed, C. M. Sabbir University of California, Riverside
Integrative analysis of lncRNA-mRNA co-expression in human lung epithelial cells exposed to dimethyl selenide (DMSe)-derived secondary organic aerosols
Dimethyl selenide (DMSe) is one of the major volatile organoselenium compounds released in the air from both aquatic and terrestrial environments through plant metabolism and microbial methylation. We have recently demonstrated that the atmospheric oxidation of DMSe leads to the formation of secondary organic aerosol (SOA). The resultant DMSe-derived SOA is a potent stressor in human airway epithelial cells (BEAS-2B) that can perturb several major biological pathways, such as genotoxicity, DNA damage, and p53-mediated stress responses. Mounting evidence has suggested that long non-coding RNAs (lncRNAs, over 200 nt in length) are involved in regulating diverse biological processes including lung carcinogenesis. However, the roles of the lncRNA mediated regulation of DNA damage response and repair in response to the environmental stimuli remain to be elucidated. In this study, we utilized bioinformatics tools to perform integrative analysis of lncRNA-mRNA co-expression in response to DMSe-derived SOA on the human airway epithelial cell line BEAS-2B. In the present genome-wide study we showed that known lncRNAs are differentially expressed in response to the O₃ (270 up- and 33 down-regulated) and OH (322 up- and 48 down-regulated) oxidation of DMSe. From integrated lncRNA-mRNA co-expression network analysis, we found known lncRNAs including PINCR, LINC01629, LINC02014, and AC010624.1. Results from the co-expression analysis will be further validated to investigate the functional significance of identified lncRNA by knocking down selective lncRNA in BEAS-2B using synthetic small interfering RNAs (siRNAs). Thus, this study provides a mechanistic insight into the epigenetic regulation of gene expression in response to the environmental stimuli which ultimately leads to better understanding of the role of environmental stimuli in lung carcinogenesis.

TOXI 5. Li, Miao FDA
Development of the generic pregnancy physiologically based pharmacokinetic (PBPK) model to support regulatory evaluations of the safety and efficacy of therapeutic agents
Due to the ethical concerns, pregnant women are rarely included for the clinical pharmacology studies to test the safety and efficacy of therapeutic agents. The physiological changes during pregnancy significantly affects the pharmacokinetics of these drugs. The regulatory agencies need to provide guidance for the use of these therapeutic agents to protect and promote perinatal health. The physiologically based pharmacokinetic (PBPK) model is a useful computational tool to predict the absorption, distribution, metabolism and excretion of drugs during pregnancy by incorporating both the physiological changes and drug-specific properties. An open-source generic pregnancy PBPK model is needed as a transparent and flexible tool to overcome the sparsity in pregnancy pharmacokinetic data and support regulatory decisions in these life-stages. In the current project, we have reviewed recently published repositories of physiological changes during pregnancy and established a generic pregnancy PBPK model that captures the dynamics of maternal and fetal pharmacokinetics over all three trimesters. The model has been evaluated for two commonly used antiretroviral drugs, Acyclovir and Darunavir, through intravenous or oral administration. Acyclovir is a drug with low protein binding and eliminated mainly through renal clearance, and Darunavir is the one with moderate protein binding and eliminated through liver metabolism. The developed model well predicted the pharmacokinetics for both drugs in non-pregnant and pregnant adults, and simulated the changes in pharmacokinetics with different gestational ages. A user-friendly interface is under development to facilitate the usage of the adaptable PBPK model by reviewers or non-modelers. Furthermore, by incorporating the population analysis, the current model can be expanded to simulate drug pharmacokinetics based on population variability.

TOXI 6. Demir, Merve Univ of California
Fishing for OG lesions: Structural and biochemical insights into OG detection by MutY
Oxidation of guanine (G) to 8-oxo-guanine (OG) by reactive oxygen species can result in G:C to T:A transversion mutations in the genome due to the ability of OG to mispair with adenine (A). The DNA glycosylase MutY is responsible for recognition of the OG:A mispair and removal of the undamaged, but contextually miscoding A. The role of this enzyme in maintaining genomic integrity is highlighted by the disease association of functionally compromised variants of the human MutY homolog (MUTYH); failure to repair the OG:A mispair by inherited variants of MUTYH is correlated with increased risk of colorectal cancer, a syndrome known as MUTYH-Associated Polyposis (MAP). An understanding of the mechanism of OG recognition and A excision by MutY can provide insights into how some MAP variants are deficient in their function and how the enzyme can specifically recognize and process the OG:A mismatches. Through 2D-NMR and crystallographic studies of Geobacillus stearothermophilus (Gs) MutY bound to a transition state analog (TSAC-OG:1N), the David Laboratory has recently proposed a revised two-step mechanism for A excision that includes formation of a covalent intermediate with the active site residue D144. Herein we report a new crystal structure of Gs MutY in complex with DNA containing G:1N (TSAC-G:1N) that highlights a serine residue in OG versus G specificity. The significance of Ser308 and its neighboring residues is further explored by mutational analysis to unravel the importance of this region for OG specific recognition in cells and in vitro.

TOXI 7. Hurley, Katherine ETH Zürich
Biosynthesis of acrolein with a coupled enzyme system to emulate continuous acrolein production by gut microbiota
Human exposure to acrolein results from its presence in car exhaust and frying oil vapors, due to lipid peroxidation and polyamine catabolism, and it has recently been identified to be a product of glycerol metabolism by human gut microbes. While acrolein toxicity is well characterized, there is little understanding of the chemical-biomolecular interactions and their biological consequences resulting from intestinal acrolein formation, especially at concentrations and rates that are relevant for gut bacteria production. Acrolein reacts with DNA forming potentially mutagenic guanine adducts (Acr-dG) and with amino acids, such as lysine (FDP-lysine), leading to possible loss of protein function and premature proteolysis. Additionally, acrolein is a highly reactive and volatile molecule, thus making it difficult to study the effects of its exposure in an in vitro experimental setting. We established a cell culture-compatible coupled-enzyme system, using a polyamine catabolism enzyme and catalase to scavenge hydrogen peroxide side product, to produce acrolein continuously for up to 20 hours. Thus, emulating the constant production of acrolein by the gut bacteria and circumventing the volatile nature of acrolein. We performed cell-culture compatibility studies with the coupled enzyme system and observed enzymatic turnover in media and in as little as 6 hours a change in cell morphology of colon epithelial cells was observed indicating cytotoxicity. The production of acrolein was monitored by a fluorescent conjugate and the formation of Acr-dG adducts and FDP-lysine protein adducts utilizing capLC-ESI-MS and ELISA, respectively. To recapitulate the production of acrolein by gut microbiota, the components in the coupled enzyme system were systematically adjusted and related to levels of corresponding adducts. This system allows us to regulate the continuous production of acrolein and eliminate the use of bacterial culture to assess the biological relevance of glycerol dehydration in the intestine using in vitro models.

TOXI 8. Pujari, Suresh University of Minnesota
Synthesis, structural confirmation, and detection of EB-FAPy-dG and DEB-FAPy-dG adducts in human cells
1,3-Butadiene (BD) is a human carcinogen which undergoes metabolic activation by cytochrome P450 monooxygenases to 1,2-epoxy-3-butene (EB) and 1,2,3,4-diepoxy butane (DEB). These reactive electrophiles alkylate DNA at the N7-position of guanine to give N7-alkylated-dG adducts. These N7-EB-dG and N7-DEB-dG adducts are hydrolytically unstable and undergo 5-membered ring rupture to give corresponding ring opened EB and DEB-formamidopyrimidine (EB-FAPy-dG and DEB-FAPy-dG) adducts. Here, we report the first synthesis of both EB-FAPy-dG and DEB-FAPy-dG adducts. These lesions were fully characterized by UV spectroscopy, NMR, and liquid chromatography tandem mass spectrometry. Further, EB-FAPy-dG lesions were detected in EB-treated calf thymus DNA and in EB-treated mammalian cells using quantitative isotope dilution nano HPLC-NSI-MS/MS. EB-FAPy-dG adduct formation in EB-treated calf thymus DNA was found to be concentration dependent. Repair of EB-FAPy-dG adducts was also investigated. Adduct amounts were 2-fold higher in BER deficient NEIL1^-/- mouse embryonic fibroblasts (MEF) as compared to isogenic controls (NEIL1^+/+), suggesting that this lesion may be a substrate for NEIL1. Overall, our results indicate that EB-FAPy-dG and DEB-FAPy-dG adducts form under physiological conditions, prompting future studies to determine their contributions to the genotoxicity and mutagenicity of 1,3-butadiene. To elucidate the effects of these FAPy lesions, our efforts are directed towards the synthesis of both EB and DEB-FAPy-dG phosphoramidites and their site-specific incorporation into DNA for biological evaluation.

TOXI 9. Cranford, Matthew Baylor University
Locating the second polymerase switch: Where resumption of replication limits the length of translesion synthesis
Upon encountering damaged bases in the DNA template, a replicative DNA polymerase (Pol) is obstructed for continued synthesis. One response to replication stalling includes “on-the-fly” translesion synthesis (TLS), where a specialized, yet error-prone, Pol is directly recruited to the damage site to bypass the lesion. In response to 8-oxo-G DNA lesions, TLS PolY from S. solfataricus is uniquely efficient and accurate for incorporation of deoxycytidine. Prior investigations have shown that PolY alone is capable of extension synthesis downstream of the lesion. However, regulation of its catalytic role beyond the lesion has not been fully evaluated in the presence of other replication components. Building from prior kinetic frameworks, we characterized lesion bypass intermediates formed by different archaeal replication subassemblies in response to 8-oxo-G. After TLS by PolY, a novel intermediate was identified at the template position three base pairs beyond an 8-oxo-G lesion, where PolY becomes catalytically inefficient for additional extension. Accumulation of this intermediate is reduced in the presence of replicative PolB1, which is capable of rapid extension synthesis from the intermediate position. The complementation of Pol specificities and activities at this intermediate reveals a precise position where Pols are handed-off after bypass of 8-oxo-G. Altogether, these findings suggest that mediation of TLS and replicative Pol activities through a processivity clamp can control the position where replication resumes and limits the extent of error-prone TLS extension downstream of a DNA lesion.

TOXI 10. Wu, Junzhou Massachusetts Institute of Technology
Derivatization of Phosphorothioate DNA modifications by direct chemical conjugation
Phosphorothioate (PT) modifications of DNA were found to occur naturally in bacteria and archaea, and more recently on RNA in both prokaryotes and eukaryotes. PTs involve the substitution of a non-bonding oxygen with sulfur in the phosphate in the DNA or RNA backbone in an Rp configuration and, in DNA, are catalyzed by the five-protein DndA-E family. While the dnd genes constitute a restriction-modification (R-M) system with restriction genes dndF-H, the redox and nucleophilic properties of sulfur in PTs leads to DNA strand breaks, genomic instability, and reduced fitness during oxidative stress. However, the impact of PT oxidation on the ecological niche of bacteria under biologically relevant conditions has not been studied, such as under the redox imbalance occurring in chronic intestinal inflammation. Despite the success sequencing of PT in purified bacterial genomes, it is still a challenge to use the current PT sequencing methods for a complex bacterial population such as occurs in the gut microbiome. Therefore, we are developing a new strategy to enrich for PT-containing DNA and to site-specifically map the PTs. The principle of the method involved in PT detection is that the phosphorothioate triester arising from labelling the reaction of PTs with alkylating agents tends to rearrange to a more stable phosphoramidate structure. In addition to the increased stability, the rearrangement step also helps to eliminate nonspecific byproducts from DNA alkylation. Using an alkylating agent containing a biotin ligand, PT-containing DNA fragments can be enriched for further analysis, such as gene enrichment by qPCR, site-specific localization by sequencing, and analysis of associated proteins by proteomic analysis. This work would provide a new strategy for defining the biological and toxicological impacts of PT modifications on a genomic scale.

TOXI 11. Jaffrey, Samie Weill Cornell Medicine
Chemical methods for detecting and quantifying nucleotide modifications in mRNA
An emerging concept is that an “epitranscriptomic code” of nucleotide modifications is found within mRNA and long noncoding RNA, and control their fate and function in cells. This concept was initiated by the transcriptome-wide mapping technology of N⁶-methyladenosine (m⁶A) which revealed that m⁶A levels change during development, are differentially expressed in different tissues, located in at least a fourth of all mRNAs, and enriched in specific regions of the transcript body. More recent studies show that other nucleotide modifications are present in mRNA, including N⁶,2’-O-dimethyladenosine (m⁶A_m), which is exclusively located at adjacent to the 7-methylguanosine-cap at the first encoded nucleotide in up to 40% of mRNAs. The evidence for other nucleotide modifications in mRNA is limited and there has been considerable controversy regarding their existence. Here we will describe new methods for site-specific detection of nucleotide modifications at essentially any position in mRNA, as well as novel mass spectrometry methods for selective detection of nucleotide modications in mRNA without contamination from tRNA, rRNA or other sources. We will discuss how these methods can validate or invalidate the the location of 2′-O-methyl, 7-methylguanosine, and N¹-methyladenosine at previously mapped positions in the transcriptome. Overall, these methods help to reveal the extent of nucleotide modifications that are present in mammalian mRNA.

TOXI 12. Begley, Thomas University at Albany, SUNY
Epitranscriptomic marks translationally regulate stress response programs to protect against environmental insults
The epitranscriptome – the dozens of chemical modifications in all types of RNA – is emerging as a key regulator of gene expression. There are over 120 different enzyme catalyzed modifications that can be found on RNA. These epitranscriptomic signals can regulate gene expression by altering RNA structure and stability, with some tRNA modifications having direct links to translational regulation. We have used targeted and systems-based approaches to demonstrate that cells and tissues respond to stress and toxic insults by enzymatically reprogramming the levels of many modified nucleosides in tRNA, to generate unique epitranscriptomic signatures representing responses to specific toxicants. Further we have demonstrated that stress induced changes in the levels of specific modified uridines, guanines and cytidines in the anticodon of tRNA is used to translationally regulate transcripts with specific codon characterisitcs. We have developed a transgenic mouse that is deficient in epitranscriptomic writer Alkbh8 and shown that cells and tissues from these mice are sensitized to damage by external agents (H₂O₂, rotenone and naphthalene) and fail to develop tolerance to specific toxicants. The broad impact of epitranscriptomic systems in toxicology is only beginning to be realized as they have emerged as key regulators of gene expression that control the synthesis of many stress response proteins. Further, as epitranscriptomic responses are used to prevent exposure induced cell death, they have the potential to be used as biomarkers of response or exploited for pharmaceutical applications.

TOXI 13. Gregory, Richard Boston Children’s Hospital
Role of the ‘Epitranscriptome’ in gene regulation and cancer
Cellular RNAs are subject to a myriad of different chemical modifications that play important roles in controlling RNA expression and function. The emerging field of ‘epitranscriptomics’ is offering new insights into the role of RNA posttranscriptional modifications in different human pathologies and the relevance of certain RNA methyltransferases (MTase) as possible cancer therapeutic targets is just beginning to be appreciated. For example, we and others found that the N6-methyladenosine (m⁶A) mRNA modifying enzyme METTL3, can act as an oncogene by promoting translation of certain oncogenic mRNAs to initiate tumorigenesis.

The RNA MTase METTL1 catalyzes the N7-methylguanosine (m⁷G) modification of certain tRNAs, mRNAs, and miRNA precursors. However, the role of METTL1 and its co-factor WDR4 in cancer remains largely unexplored. Here we reveal the oncogenic role of METTL1. METTL1 is frequently amplified and overexpressed in tumors and correlates with poor patient survival. METTL1 depletion in human cancer cells causes decreased abundance of m⁷G-modified tRNAs, altered cell cycle, and inhibits tumorigenicity. Strikingly, METTL1/WDR4 overexpression induces oncogenic transformation and tumorigenesis. Mechanistically, we find increased abundance of a subset of m⁷G-modified tRNAs, and increased translation of mRNAs enriched in the corresponding codons including cell cycle regulators. Accordingly, we find expression of a particular m⁷G-modified tRNA is significantly elevated in many tumor types, correlates with patient survival, and overexpression of this tRNA enhances reporter gene expression and cell transformation. Thus, METTL1/WDR4-mediated m⁷G tRNA modification drives oncogenic transformation, thereby highlighting METTL1 as a potential cancer therapeutic target.

TOXI 14. Dedon, Peter Massachusetts Institute of Technology; Singapore-MIT Alliance for Research and Technology
Old becomes new in the epi-verse of DNA and RNA modifications
The Central Dogma defines the “what” of biology: genes are transcribed into messenger RNAs that are translated into proteins. But it says little about the “when” or “how much” of gene expression, which gets messy with the 20,000 genes in humans. Convergent technologies have uncovered information-rich scheduling systems for gene expression involving the dozens of chemical modifications of DNA, RNA, and histone proteins in every cell — the epigenome and epitranscriptome. Many of these modifications have been known for decades, but their function is only now emerging with the application of new technologies. This symposium has brought together leaders in the newest of these – the epitranscriptome – a universal “system of systems” in which messenger, transfer, and ribosomal RNA modifications interact with alternative genetic information to regulate translation of proteins. Here I will discuss the discoveries we have made over the past decade in epigenetics and epitranscriptomics in viruses, bacteria, yeast, parasites, mice, and humans, with an eye on key technologies that have advanced the fields. The components of these systems are just now being recognized as critical regulators of cell biology and as drivers of disease, with significant practical implications for developing new therapeutics and industrial processes.

TOXI 15. Bertozzi, Carolyn Stanford University
Multifunctional biologics for extracellular targeted degradation
The antibody-drug conjugate field stimulated great progress in protein engineering strategies for site-specific chemical conjugation. These methods are now being extended to other chemically modified protein modalities beyond ADCs. This presentation will focus on two such modalities – antibody-enzyme conjugates (AECs) and lysosome-targeting chimeras (LYTACs), both being multifunctional biologics that act by targeted degradation of extracellular molecules. Examples AECs and LYTACs will be presented, covering a range of therapeutic applications.

TOXI 16. Senter, Peter Seattle Genetics Inc
Potent antibody-based conjugates for cancer therapy: From early stage research to clinically approved drugs and beyond

TOXI 17. Schutten, Melissa Genentech
Investigating nonclinical strategies to modify PBD Antibody Drug Conjugate-related toxcities
The promise of Antibody Drug Conjugates (ADCs) as an effective cancer therapy lies in the technology as to specifically target the cytotoxicity of a highly potent chemotherapeutic small molecule to an antigen of interest using an antibody directed at a tumor antigen. Though many of these molecules have been successful at initially entering clinical trials for solid and hematologic malignancies, and several of these ADCs have been approved, many of these ADCs have failed due to systemic toxicities. Over the last several years, companies have been focused on developing ADCs containing novel DNA damaging agent payloads. These ADCs carry inherent antigen-independent toxicities, including vascular leak, renal, skin and hematologic toxicities. Nonclinical approaches and investigative studies to potentially decrease/modify the severity of target organ toxicities related to PBD ADCs will be discussed in this presentation.

TOXI 18. Solis, Willy Sutro Biopharma
Nonclinical safety evaluation of site-specific antibody-drug conjugates

TOXI 19. Reddy, Jagath ModMab Therapeutics, Inc
Insights to developing nextgen antibody drug conjugates

TOXI 20. Meanwell, Nicholas Bristol Myers Squibb
Applications of fluorine in drug design
The judicious incorporation of fluorine into drug molecules can confer beneficial effects on a range of developability properties including membrane permeability, metabolism and off-target activity, stimulating its widespread application in drug discovery. The increasing deployment of fluorine in drug design and the discovery of complex fluorinated motifs with bespoke attributes is producing new insights into the properties of this unique element. Consequently, the applications of fluorine in drug design continue to grow, fueled by the parallel developments and advances in synthetic methodology. In this presentation, we will provide a synopsis of contemporary applications of fluorine in drug design with an emphasis on how the unique physicochemical attributes of fluorine are exploited to address developability challenges.

TOXI 21. Johnson, Benjamin Bristol Myers Squibb
Metabolic and pharmaceutical aspects of fluorinated compounds
The use of fluorine in drug discovery is on the rise, a trend that is driven by an improved understanding of its properties and the development of new synthetic methods for its installation within drug candidates. In turn, studies of fluorinated compounds have highlighted potential problems that can arise during metabolism. While the C–F bond (98–115 kcal/mol) is strong relative to other carbon–heteroatom bonds, the loss of fluoride from metabolic intermediates is a common occurrence. The resulting metabolites may be sufficiently electrophilic to react further with biological nucleophiles, an outcome that has been exploited in the design of mechanism-based enzyme inhibitors but has also resulted in unintended outcomes including off-target pharmacology and toxicity. The metabolism of select compounds may also involve the formation of one of several low-molecular-weight fluorinated toxins. Among the best characterized and most problematic of these is fluoroacetic acid, a compound that is converted further to a reversible but potent aconitase inhibitor that disrupts the tricarboxylic acid cycle and is lethal at low doses. This presentation will summarize the literature associated with the metabolism of fluorinated molecules, focusing on representative examples where the presence of fluorine had unintended or otherwise notable consequences. These examples are instructive for the drug-metabolism scientist and have refined our understanding of how fluorine should be deployed in drug design.

TOXI 22. Xu, Qiuwei Merck & Co., Inc.
Assessment of defluorination of fluorinated drugs in animal models
Many drugs contain fluorine substitutions as it can improve efficacy by modifying drug pharmacokinetics, metabolism, or pharmacodynamics. Although the fluorine-carbon bond is the strongest in organic chemistry, it is still susceptible to defluorination during biotransformation in vivo as well as in vitro, leading to direct exposure of the released fluoride to animals or humans. Low level of fluoride exposure is safe, and its prescription has been recommended for decades in oral health to reduce dental cavities. However, excessive and/or prolonged high level of exposure can lead to multiple types of toxicities (e.g., bone and tooth) in both animals and humans. While defluorinated metabolites can sometimes be seen in vitro, we designed an approach to evaluate the extent of compound defluorination due to drug metabolism in preclinical species. Fluoride release from in vivo defluorination usually ends up in two different pools. One portion of fluoride is absorbed by hard tissues (i.e., bone and tooth), and the remaining is mainly cleared from the body through urinary excretion. We take a weight-of-evidence approach considering in vitro and in vivo drug metabolism data in addition to direct measurement of total urine fluoride excretion in nonclinical species. This oral presentation will cover our approach of urine collection, calculation of defluorination, and correction for fluoride retained in hard tissues based on studies using oral sodium fluoride treatment. Although defluorination assessment in animals can identify potential in vivo fluoride release, ultimately measurement of the extent of defluorination in humans needs to be assessed in the clinic.

TOXI 23. Jeschke, Peter Bayer AG
Latest generation of fluorine-containing agrochemicals
In the search for new agrochemicals which match modern requirements in terms of efficacy, environmental safety, user friendliness, and economic viability, the substitution of molecules with fluorine and fluorinated motifs is an important approach.
Besides impacting biological activity, the introduction of fluorine atoms or fluorine-containing substituents in combination with further structural fragments into an active ingredient can influence some specific properties such as degradation in soil, metabolism in plants and target pests, and can result in a more favorable compatibility profile to non-target arthropods or beneficial insects.
In the past decade, there has been a remarkable rise in the number of commercial products containing fluorine atoms and/or fluorinated motifs. This enormous increase in fluorine-substituted products (Br << Cl << F) demonstrates the importance of halogenation and in particular fluorination of active ingredients. At present, these include for example trifluoro- and difluoromethyl, or trifluoro- and difluoromethoxy groups at aryl or heterocyclic aryl moieties and fragments like 2,2-difluoroethyl, difluoromethylsulfonyl or (F₃C)₂CF- and (F₃C)₂C(OMe)-groups.
A critical analysis has been carried out with the latest generation of fluorine-containing agrochemicals launched on the global crop protection market in the past 10 years and a few representative herbicides, fungicides and insecticides/acaricides have been selected to illustrate in detail which contribution the fluorine atom or special fluorine-containing substituents have made to current agricultural chemistry.

TOXI 24. Fleming, Aaron University of Utah
Mechanistic Studies of 8-Oxoguanine in Non-canonical Structures Impacting Gene Expression
In mammalian genomes, the 2′-deoxyguanosine (G) content increases in gene promoters near transcription start sites (TSS) rendering these regions sensitive to oxidative modification by reactive oxygen species (ROS) and capable of non-canonical G-quadruplex (G4) folding. Oxidation of G to 8-oxo-7,8-dihydro-2′-deoxyguanosine (OG) occurs at an elevated frequency in mammalian gene promoters that can impact gene expression. We used reporter plasmids with site-specifically incorporated OG and its repair intermediates in a model promoter to conduct mechanistic studies in mammalian cells. Gene activation occurs when OG was at an oxidation-prone site in a G4 sequence in the non-template strand near the TSS. We propose the promoter G-rich sequence in the duplex context is an antenna for oxidation events that home the DNA repair process to release OG by the glycosylase OGG1 yielding an abasic site (AP). The AP triggers a structural switch to a G4 fold where the baseless site stably exists in a loop region for APE1 recognition. We found APE1 is a conduit for interaction with transcriptional activation proteins, in addition to its role in facilitating DNA repair. In vitro studies allowed us to understand the role of APE1 binding and catalytic activity toward an AP in a G4 context for gene regulation. The plasmid-based system also allowed us to map the promoter location dependency and strand bias (non-template vs. template) for oxidatively modified promoter G4 folds to regulate transcription up or down. A comparison of different organism’s genomes identifies not all organisms harness OG as an epigenetic-like DNA modification for gene regulation.

TOXI 25. Püllen, Nikolai ETH Zurich
Nucleotide resolution mapping of DNA oxidation in the human genome
DNA oxidation arising from cellular oxidative stress is a major endogenous damage type with significant toxicological implications in disease development. 8-oxoG is the major type of oxidative DNA damage. Owing to the lack of high-resolution detection methods, the precise location and mechanisms influencing 8-oxoG prevalence in the human genome remained largely unidentified. We recently developed a nucleotide-resolution genome-wide sequencing method for oxidative damage named click-code-seq and applied this method to yeast genome. Herein, we report the genome-wide distribution of DNA oxidation in human haploid cells employing an enhanced version of click-code-seq. Compared to yeast, the much larger size of the human genome exacerbates damage detection dramatically. We sought to overcome major challenges by chemical masking of background lesions, the use of highly specific enzymes and a rapid low-artefact library preparation. Using these novel conditions, we unraveled a widely varied distribution of oxidative DNA damage across the genome, with distinct patterns related to chromatin architecture, epigenetic modification and DNA-protein interaction. Furthermore, our data imply the active modulation of damage reduction in regions of high functional importance along with damage hot spot formation in low-significance areas under oxidative stress. Sequencing of 8-oxoG paves a powerful approach for studying biological and toxicological questions surrounding DNA oxidation and its implication in downstream mutagenesis processes of medical significance.

TOXI 26. McKeague, Maureen McGill University; McGill University
Genome-wide distribution profiles of DNA damage
DNA adducts are critical risk factors in cancer, aging and disease, and are the underlying basis of most frontline cancer therapies. Determining the precise location of DNA adducts in the genome is an analytical challenge due to the low abundance of events and our inability to specifically amplify DNA damage without losing the chemical information. One solution to this problem involves the use of specific enzymes that recognize DNA damage coupled to efficient ligation of next generation sequencing compatible barcodes to mark locations of DNA damage. Using this strategy, our studies suggest that DNA damage location formation and repair is dependent on genome location. Specifically, different chromatin states and transcription factors binding sites play a role in the distribution of DNA damage. Identifying the patterns of DNA damage, the chemical basis of repair, and how these processes may be modulated, supports our understanding of mechanisms of carcinogenesis and strategies for improving cancer therapeutics.

TOXI 27. Delaney, Sarah Brown University
Global repair fingerprinting of glycosylase activity on nucleosomes
Damage to genomic DNA can lead to mutagenesis and disease. Repair of single base damage is initiated by DNA glycosylases, the first enzymes in the base excision repair (BER) pathway. Most experiments that led to our current understanding of BER were conducted using DNA oligomers. There is a fundamental gap in knowledge of how BER occurs in the context of chromatin, where eukaryotic DNA is compacted in a complex hierarchy of DNA-protein interactions. In this research we developed techniques to study the global initiation of repair on DNA packaged in nucleosomes. Using these techniques, we describe the global repair fingerprint of several glycosylases and identify locations in packaged DNA that are refractory to repair and those that are repaired efficiently. We also examined the role of histone variant proteins in modulating the initiation of repair and demonstrate that some variants dramatically increase the ability of glycosylases to remove modified bases from DNA and initiate repair. The results will be discussed in terms of structural features and dynamics of nucleosomes, and well as in the context of known mutational hotspots.

TOXI 28. Essigmann, John MIT
Mutational spectra provide insight into the mechanisms bridging DNA damage to genetic disease
Mutations are a permanent record of past chemical and biochemical insults experienced by a cell. Each time the genome replicates, it accumulates mutations in patterns that reflect: (a) the DNA sequence context-dependent formation of damage; (b) the differential activity of DNA repair proteins, which, depending on the type of lesion, can eliminate, ignore or intensify the mutagenic consequences of the lesion; and (c) the choice of replication machinery that synthesizes the nascent genomic copy. These three factors result in a richly contoured mutational spectrum that, from appearances, is distinct for most individual forms of DNA damage. This mutational pattern, if appropriately decoded can reveal the history of genome-altering events such as chemical exposures, pathogen-stimulated base deaminations, metabolic stress, and inflammation, which in turn can provide an indication of the underlying causes and mechanisms of genetic disease. New tools help us glean a deep mechanistic understanding of the cellular factors and pathways that modulate a mutational process and, in turn, provide opportunities for better diagnostic and prognostic biomarkers, better exposure risk assessment and even actionable therapeutic targets. The goal of this talk is to present a bottom-up, lesion-centric framework of mutagenesis that integrates the contributions of lesion replication, lesion repair and lesion formation to explain the complex mutational spectra that emerge in the genome following exposure to mutagens. The mutational spectra of the well-studied hepatocarcinogen aflatoxin B₁ and several alkylating agents serve as specific examples, but the implications are meant to be generalizable.

TOXI 29. Wyrick, John Washington State University
Mutational signatures induced by atypical UV photoproducts are associated with driver mutations in melanoma
Somatic mutations in skin cancers and other UV-exposed cells are typified by C-to-T (i.e., C>T) substitutions in dipyrimidine sequences. However, many oncogenic ‘driver’ mutations in melanoma do not fit this UV signature; whether such driver mutations are caused by UV lesions is unclear. In order to more thoroughly characterize the full spectrum of UV-induced mutations, we sequenced the genomes of yeast strains repeatedly exposed to high doses of UV light. Canonical UV signature mutations are evident in UV-exposed yeast, particularly in strains deficient for DNA polymerase eta (pol η). However, we also identify novel mutation signatures associated with UV exposure. We show that T>C mutations are highly enriched in UV-exposed yeast, likely due to error-prone bypass of (6-4) photoproducts by pol η. We also detect a novel A>T mutation signature associated with an atypical thymidine-adenine (TA) photoproduct, which we are able to map across the yeast genome using a new high-throughput sequencing method known as UVDE-seq. Finally, we detect new classes of tandem mutations that are induced by UV exposure. We show that these novel mutation signatures are indeed caused by bulky UV photoproducts, as they show a striking transcriptional asymmetry that is modulated by nucleotide excision repair activity in both UV-exposed yeast and human skin cancers. We propose that mutations arising from atypical UV photoproducts may explain the origin of many driver mutations in skin cancer, including the recurrent BRAF V600E and V600K mutations in melanoma.

TOXI 30. Mingard, Cécile ETH Zürich
Polymerase ζ mutational spectrum following alkylating agent exposure
DNA Polymerase ζ (Pol ζ) is a low fidelity B-family polymerase involved in DNA damage tolerance by mediating DNA adduct bypass. However, Pol ζ is in most cases inefficient at inserting a base opposite the adduct, rather it may contribute to mutagenesis due to its high efficiency in extending from base mismatches. Additionally, human Pol ζ involvement in bypass has been characterized for bulky and oxidative DNA adducts but little is known about the potential mutagenic role of Pol ζ with small alkyl DNA adducts such as O⁶-methylguanine (O⁶-MeG). O⁶-MeG is a common DNA lesion arising from nitrosamines giving rise to G > A transitions found in human cancer genomes. Using human Pol ζ and a DNA template containing O⁶-MeG, we found that Pol ζ inserts dTTP opposite O⁶-MeG with low efficiency but could extend from the resulting mismatch efficiently. We additionally characterized how flanking base identity influences extension. To understand the contribution of mismatch extension catalyzed by Pol ζ to mutagenicity in cells, we exposed mouse embryonic fibroblasts deficient in REV3 (Pol ζ catalytic subunit) with an alkylating agent that induces O⁶-MeG formation and subsequently analyzed by whole-genome sequencing the number and type of mutations in resulting clonally expanded cells. Surprisingly Pol ζ was found to reduce overall number of mutations, especially G > A mutations after alkylating agent exposure, suggesting a contribution in the error-free bypass of O⁶-MeG. Furthermore, the mutation spectrum in cells lacking Pol ζ had a high correlation with mutational signature 11 observed in patients treated with the alkylating drug temozolomide. Our results suggest a protective role of Pol ζ in methylation-induced mutagenesis. These findings have potential implications for cancer patients treated with alkylating drugs as they might be more resistant if they have high expression of Pol ζ and more fundamentally to trace origins of mutations found in cancer genomes.

TOXI 31. Xia, Menghang NIH
Profiling environmental chemicals using in vitro qHTS methods
The toxicology in the 21st century (Tox21) program, a collaboration of the National Institutes of Health, Environmental Protection Agency, and Food and Drug Administration, has utilized a quantitative high throughput screening (qHTS) approach to profile thousands of environmental chemicals against a battery of in vitro assays. Millions of data points generated from Tox21 screening were made publicly accessible. These rich datasets provide researchers with opportunities for further data mining, understanding compound action, and prioritizing compound for further in-depth studies. In order to generate high quality and reproducible data, the assay selection, optimization, validation, and screening performance play a critical role in toxicological testing. This presentation will highlight these critical components in assay development and screening process. Several examples of in vitro assays that are available for compound toxicity assessment will be evaluated in detail. The screening results from these in vitro assays will be useful for prioritizing chemicals for further in-depth mechanism-based toxicity testing.

TOXI 32. Sharma, Monita PETA International Science Consortium
In Vitro approaches for assessing respiratory toxicity
Risk assessment and management relies on approaches that can accurately and efficiently predict the toxicity of chemicals in humans. Inhalation is a major route by which exposure to substances can occur, and is an area where resources have been dedicated to optimizing relevant in vitro approaches based on human biology. In vitro approaches that employ proper controls, are reproducible, and simulate physiological conditions have been developed and can be used to test substances and concentrations that are relevant to human exposures. Presented here is an example of an in vitro testing strategy using BEAS-2B cells (a human bronchial epithelial cell line) as a model to assess the ability of silanes to cause portal-of-entry effects on the human respiratory tract. The cells were exposed to various concentrations (0.72ppm, 25ppm, and 85ppm) of triethoxysilane vapor—a GHS category 2 inhalation toxicant based on rat LC50 (lethal concentration, 50%) data—at the air-liquid interface using a capillary dosage unit coupled to a VITROCELL 6/4 exposure module. A significant concentration-dependent decrease in cell viability (resazurin-based assay) and increase in cytotoxicity (lactate dehydrogenase assay) was observed after exposure to the triethoxysilane and nitrogen dioxide (positive control) as compared to clean air (negative control). A significant increase in expression of inflammatory markers [interleukin (IL) 4, IL-6, IL-8, IL-10, IL-12p70, IL-13, and tumor necrosis factor-alpha (TNF-α)], determined by Meso Scale Discovery technology, was observed at 25ppm. Work is underway to test additional controls (Triton X-100 and sodium chloride) and silanes (with varying carbon-chain length) to assess the advantages of using a 2D cell line (BEAS-2B cell) versus a 3D human reconstructed tissue model (MucilAir™, Epithelix). This project has included regular discussions and information-sharing with experts from industry and regulatory agencies regarding choice of in vitro system, test materials, biological end points, and appropriate controls relevant to assays and test materials. These multi-sector discussions have helped facilitate the uptake of robust in vitro methods and reduce the use of animals in testing.

TOXI 33. Reipa, Vytas National Institute of Standards and Technology
Measurement issues in the phototoxicity of nanomaterials
Toxic potential of nanomaterials (NMs) to humans and environment stems from their physicochemical properties that are highly dependent on high surface to volume ratio and enhanced transport across biological barriers. Many NMs, currently used as photocatalysts, UV light blockers in sunscreens, display materials as well as in nanotherapeutics are produced in large quantities (TiO2, ZnO, WO3, CeO2, quantum dots, etc.) and could become toxic under light exposure, resulting in photo irritation, photo allergy, and photo genotoxicity. Due to the unique physicochemical features most assays developed for soluble chemicals are not automatically applicable to NMs. Methods available for macro-powders require substantial material quantities, lengthy measurements and separation of the particles from the suspension prior to analysis. As the generation of reactive oxygen species is a key determinant for causing phototoxic reactions, their detection can predict the phototoxicity potency. We have developed a rapid and inexpensive procedure to assess the NM photocatalytic activity under physiological conditions, a handy tool to screen for NM phototoxicity. This assay was tested and validated using NIST reference material SRM1898 (Titanium Dioxide Nanomaterial) during the interlaboratory comparison study and was recently published as an international standard. Problematic measurement aspects in NM phototoxicity and ongoing standardization efforts will be presented.

TOXI 34. Petersen, Elijah Elijah Petersen
Advances in the reliability of nanotoxicity assays
With the increasing usage of nanotechnology in consumer products, it is important to develop accurate methods to assess the potential adverse effects of engineered nanomaterials (ENMs) on humans or environmentally important organisms. However, many of these measurements are challenged by the unique behaviors of ENMs which differ from those of dissolved metals or hydrophobic organic chemicals. Thus, modifications to typical toxicity assays are needed to account for these behaviors to avoid artifacts and ensure accurate measurements. In this talk, I will first discuss the different artifacts that can occur during measurements on the toxicological effects of nanomaterials and potential control experiments to assess these artifacts. Then, I will discuss the usage of measurement science tools, such as cause-and-effect analysis and incorporation of process control measurements into the assay plate design, to support the successful interlaboratory comparison of the MTS nanocytotoxicity assay, and the evaluation of an ecotoxicity assay, an International Standardization Organization C. elegans growth and reproduction inhibition assay, for use with nanomaterials. All opinions are my own.

TOXI 35. Burrows, Cynthia University of Utah
Free radical oxidation of guanine in DNA—mutagenic or epigenetic?
Guanine-rich sequences in DNA exhibit two interesting chemistries, a sensitivity to one-electron oxidation, and the ability to fold into non-canonical structures such as G-quadruplexes (G4s). Oxidation of dG residues in DNA commonly leads to the mutagenic lesion 8-oxo-dG (or OG), and this damage can occur over long distances via charge transport in the duplex. Because G tracks of 3 or more contiguous Gs have the lowest redox potential compared to other sequences, DNA oxidation by a one-electron oxidant is focused on G-rich sequences such as those that form G4s. The abundance of G4 sequences at regulatory regions of the genome, such as gene promoters and intron splice sites, presents a dichotomy—are these oxidation-sensitive sites prone to mutagenesis or does oxidation of a G4 play a regulatory epigenetic role?

TOXI 36. Jokipii Krueger, Caitlin University of Minnesota
Endogenous versus exogenous sources of N7-(2,3,4-trihydroxybut-1-yl) guanine and N7-(1-hydroxyl-3-buten-1-yl) guanine DNA adducts

TOXI 37. Fukuhara, Kiyoshi Showa University
Generation of reactive oxygen species following photoirradiation of nitromusks
Nitrobenzene derivatives such as musk xylene, musk ketone, and musk ambrette, known as nitromusks, are typically used in fragrances. Among these, musk ambrette is known to cause photoallergic contact dermatitis, as well as mutations via metabolic activation of its nitro group. However, the detailed mechanism of the photoallergy remains unknown. In general, phototoxicity is mainly caused by generation of singlet oxygen (¹O₂). ESR spin-trapping experiments using TEMP were performed to evaluate the ¹O₂-generating ability of nitromusks. However, formation of a paramagnetic TEMPO as a spin adduct with ¹O₂ was not observed for any of the investigated compounds. On the other hand, photoirradiation of the investigated nitromusk compounds in the presence of NADH, with DMPO as a spin trap, resulted in the formation of DMPO/hydroxyl radical adducts. Musk ambrette exhibited the most intense DMPO/hydroxyl radical adduct. Addition of SOD and catalase attenuated the DMPO/hydroxyl radical signal, suggesting generation of a hydroxyl radical via superoxide produced by reduction of molecular oxygen in the reaction system. ESR experiments were carried out following photoirradiation in the absence of a spin trap under anaerobic conditions. Results demonstrated formation of radical anions of musk xylene and musk ambrette by the reduction of NADH. These results suggest that a one-electron transfer reaction from NADH to molecular oxygen proceeded via photoirradiated nitromusk, resulting in the production of a hydroxyl radical as a consequence of superoxide disproportionation. Photoirradiatied DNA-cleaving activities of nitromusks were also evaluated using pBR322, with the strongest cleavage observed for musk ambrette. This result provides support that nitromusk photoallergy is caused by reductive activation of molecular oxygen by NADH.

TOXI 38. Ohno, Akiko National Institute of Health Sciences
Application to toxicity evaluation of titanium dioxide nanoparticles based on physicochemical properties using multivariate analysis method
Nanomaterials (NMs) have the same component as a constituent unit but have unique characteristics such that they exhibit different functions and activities depending on the form. Therefore, it is expected to be applied to new materials having superior properties that have not been seen before, and active research and development are being promoted internationally. However, it is difficult to predict toxicity based on structure-activity relationships from constituents such as low-molecular-weight chemicals, and a comprehensive approach to toxicity evaluation and safety that uses the morphological characteristics of nanoparticles as descriptors is required. In this study, we collected test data of physicochemical properties and harmfulness information for six types of titanium dioxide nanoparticles (TiO2 NPs), and searched and scrutinized data necessary for statistical analysis. Furthermore, we analyzed the similarity of physical property values and examined the usefulness of this method for toxicity evaluation by using multivariate analysis for the relationship between physicochemical properties and hazard information. Sources of information on the physicochemical properties and hazard test data on repeated dose toxicity by inhalation and intratracheal administration routes for six types of TiO2 NPs were provided by the OECD NM safety assessment test document (Dossier) and eNanoMapper DB. SIMCA-P was used for statistical analysis. Physicochemical properties of about 168 items for which was collected information were classified by hierarchical clustering analysis, and their similarities were shown by Dendrogram. Combinations of items showing similarity in physicochemical properties were indicated by PCA, and the relevance of hazard information could be linked to property items showing toxicity by OPLS. From the results obtained in this analysis, it was possible to clarify the slight differences in physicochemical properties associated with toxicity using a vast data collection on the physicochemical properties of NM. Therefore, it is suggested that the various statistical methods used in this analysis can be useful tools for evaluating the properties of NMs, and are expected to be applied to toxicity evaluation methods based on the properties of NM.

TOXI 39. Muller, Fabrice ETH Zurich
In vitro and in silico testing strategies for predicting human liver toxicity from foodborne chemicals
Currently, evaluating toxicological hazards of new chemicals with animal testing approaches are outdated and their ethical and financial limitations are widely acknowledged. At the same time, non-alcoholic fatty liver disease (NAFLD) is a rising global burden affecting around 30% of the adult population. NAFLD encompasses a complex disease spectrum ranging from benign steatosis to non-alcoholic steatohepatitis that can progress to cirrhosis and liver failure. The prevalence of NAFLD and its link to obesity and metabolic disorders suggests that environmental exposures could contribute to NAFLD. Systematic assessments of chemicals as NAFLD risk factors have not been carried out in part due to a lack of validated in vitro and in silico testing strategies. In order to address this gap, we developed a high content imaging approach for simultaneously quantifying mechanism-based NAFLD markers in metabolically competent human liver cells and extrapolating human oral equivalent doses for molecular responses. Various exposures with known NAFLD-relevant pathophysiology were evaluated. Cells were fluorescently stained for nuclei, lipids, mitochondrial dysfunction and oxidative stress and imaged on a high content imaging system. An algorithm was developed for automated image processing and cell death, lipid droplet number and size, mitochondrial dysfunction and oxidative stress were quantified on a population as well as single cell level. Oral equivalent doses were extrapolated using physiologically based pharmacokinetic modelling along with reverse dosimetry. Chemicals induced a concentration-dependent increase in lipid accumulation, disruption of mitochondrial function and integrity, and increase in oxidative stress, including heterogeneous responses of single cells in the population. Extrapolated oral equivalent doses for the key events addressed were within biologically relevant exposure ranges for the test chemicals. Future research will aim to establish a predictive model and screen many food-related chemicals for their risk to induce NAFLD.

TOXI 40. Brown, Brandon California Environmental Protection Agency
Systematic review for DPR human health risk assessments: best practices as applied to case studies with allyl isothiocyanate, methomyl, and linuron
The California Department of Pesticide Regulation (DPR) characterizes the risks of pesticides to human health using a science-based, weight-of-evidence approach. This approach includes a rigorous systematic literature review to identify and evaluate all relevant data from any verifiable source. DPR has developed best practices for conducting systematic reviews as a key part of risk assessment. DPR’s systematic review is conducted using PECO criteria as a standard for assessing the relevance of published data for assessing potential risks, posed by pesticides, to human health. PECO criteria describe the populations (P), exposures (E), comparators (C), and outcomes (O) of interest for a specific chemical. With each criterion established, the systematic review is conducted in four phases: search, screen, review, and optimize (SSRO). Specific search terms are used to query databases that identify all possible publications with relevant data. Each publication is screened with relevant publications being used to identify or derive points of departure (PoD). The process is optimized as information gaps are identified. The systematic literature review results for three pesticides (allyl isothiocyanate (AITC), methomyl, and linuron) are reported here. Publications were screened for relevance to human health risk assessment and later stratified into categories. Datasets suitable for dose-response modeling were analyzed using Benchmark Dose Software. This formalized systematic review process supports human health risk assessment at DPR. Our systematic reviews for AITC, methomyl, and linuron provided the weight of evidence considerations for likely modes of action and identified sensitive toxicological effects and potential PoDs. This streamlined process will improve the rigor of human health risk assessments.

TOXI 41. Kyzer, Jillian University of Wisconsin
Investigation of the mechanism of immunologic response to chronic opioid use
Long-term opioid use has not been shown to be more effective than NSAIDS for chronic pain, however one study showed that 61% of patients with chronic back pain were prescribed opioids at least once in the previous year, and 31% of patients were categorized as long-term opiate users. In addition to the risk of developing opioid use disorder (OUD), long-term opioid use has been shown to cause chronic inflammation and hyperalgesia, suggesting that opioids may induce an immune response upon chronic use.

To investigate this hypothesis, plasma samples were obtained from individuals with chronic back pain who had been prescribed opioids as well as negative controls. The plasma samples were analyzed by enzyme-linked immunosorbent assay (ELISA), screening against bovine serum albumin (BSA) conjugated to either oxycodone or hydrocodone to quantify the levels of opioid-specific antibodies in the samples. Varying levels of antibodies were found, and we thus investigated the potential mechanism by which these antibodies were produced. As other drugs containing secondary amines have been shown to form advanced glycation endproducts (AGEs) which can generate self-reactive antibodies, we hypothesized that norhydrocodone and noroxycodone would as well. We investigated the potential for norhydrocodone to form AGEs through monitoring in vitro formation of reactive intermediates.

TOXI 42. Geier, Mitra California Environmental Protection Agency
Allyl isothiocyanate (AITC) hazard identification
Allyl isothiocyanate (AITC) is a naturally occurring fungicide, herbicide, insecticide, bacteriocide, nematicide, and animal repellent. It acts as a chemical irritant and is used as a model inflammatory and nociceptive pain-inducing agent. Risk assessment of AITC was prompted by its proposed registration in California as a fumigant. Acute, subchronic and chronic toxicological studies of AITC were compiled by conducting extensive systematic review of publically available data, and data required for federal and state registration. The dose responses of endpoints were modeled using U.S. EPA’s Benchmark Dose Software (BMDS 3.1.2), and evaluated for use as critical points of departure. Clinical symptoms of exposure to AITC in humans include irritation, lacrymation, and acute pain and inflammation. Toxicokinetic data demonstrates that AITC is readily completely absorbed (>90%) after oral ingestion. It is primarily metabolized via glutathione conjugation in rats and humans and by hydrolysis in mice, and all metabolites are excreted in urine within 24 h. The studies reviewed for this assessment demonstrated that AITC is capable of inducing both portal of entry (including skin irritation and corrosion, nasal epithelial degeneration, and thickened stomach mucosa) and systemic effects. In rats, systemic effects include cataracts, retinopathy, urinary bladder hyperplasia, increased liver weight, glutathione induction, decreased motor activity, and increased thyroid weight. Tumors were observed in rats, but not in mice, with chronic oral exposure. While the results for in vitro and in vivo mutagenicity tests of AITC were largely negative, results reported for several genotoxicity assays, including DNA damage and chromosomal aberrations, were considered to be “weakly positive” or “marginal” by the study authors because they were observed at concentrations that also resulted in cytotoxicity. The systemic effect of urinary bladder hyperplasia appears to be the most sensitive effect in chronic oral exposure studies.

TOXI 43. Blake, Timothy Lawrence Technological University
Synthesis and reproductive toxicity of bisphenol analogs
Bisphenols are a class of endocrine-disrupting chemicals that can adversely affect human health. Despite their well-known association with a variety of health-related disorders, they are ubiquitous in everyday life and are found in many plastic products. Bisphenol A (BPA), perhaps the most common bisphenol, has been closely associated with developmental and reproductive disorders, for example. Several analogs of BPA, bisphenols S and Z (BPS and BPZ), have been studied in the search for safer chemicals to potentially replace BPA in consumer products; however, preliminary studies have shown that neither BPS nor BPZ are less toxic than BPA. In this study, a small library of BPA and BPZ analogs was prepared for toxicity testing in the model organism C. elegans. Most published synthetic routes employed to prepare BPA and related analogs result in low yields of approximately 30%. Through a systematic study of reaction time, temperature, stoichiometry, and acid catalyst, an effective methodology was developed to synthesize bisphenols in much higher yields of approximately 70% following purification via C18 reversed-phase chromatography. Through toxicity testing, it was found that BPZ and three related analogs caused dramatically lower reproductive ability in C. elegans. Future work will focus on preparing additional bisphenol analogs and further testing their toxicity to develop a more complete understanding of how bisphenol structure relates to reproductive toxicity.

TOXI 44. Jiang, Yang ETH Zürich
Genome-wide sequencing of 5’-aldehyde lesions resulting from DNA oxidation
DNA single strand breaks are the most abundant forms of DNA oxidative damage resulting from reactive oxygen species. Persistent SSBs could lead to replication folk collapse and double strand break formation upon interaction with the replisome, which may further lead to genome rearrangements, cell death and disease. Numerous chemical forms of single strand breaks and corresponding repair pathways have been identified, such as 3′-phosphate and 5’-aldehyde. However, the genome-wide landscape of these lesions remains largely unknown because of the lack of methods for sequencing this form of damage with high specificity and resolution. We developed a new strategy to locate the 5’-aldehyde at single nucleotide resolution. The principle of the method involves labelling the 5’-aldehyde terminus with a hydroxylamine-functionalized oligonucleotide, giving rise to a biocompatible altered DNA linkage and allowing labelled sites to be amplified by polymerase chain reaction. In this study, the specificity of labelling and bypass of other aldehyde modifications, such as an abasic site and 5-formylcytosine were characterized, supporting that various sites are chemically labelled but only those derived from 5’-aldehyde precursors could be bypassed and amplified for sequencing. The results of this work provide a new strategy for studies aiming to understand the biological and toxicological impacts of 5’-aldehyde terminus in a genome-wide scale.

TOXI 45. Chen, Jin University of California, Riverside
Identification and toxicity evaluation of carbonyl emissions from vaping of alcohol-containing flavored e-liquids
E-cigarette liquids (e-liquids) have diverse flavorings such as menthol and cinnamaldehyde, which are considered as safe for ingestion but their safety for inhalation is uncertain. Studies have found an association between e-cigarette vaping and lung injuries, which in part may be linked to carbonyls (e.g., formaldehyde and acrolein) emitted from thermal degradation of flavoring compounds. However, identification and toxicity characterization of carbonyls emitted during vaping of flavored e-liquids have not been fully assessed. We hypothesize alcohol flavoring chemicals in e-liquids during vaping processes could be an important source of carbonyl emissions, which may enhance toxicity of e-cigarette usage. E-liquids flavored with menthol, geraniol, linalool and 2-hexenol are selected to investigate composition and toxicity of carbonyls (i.e., menthone, cis-/trans-geranial, 6-hydroxy-2,6-dimethylocta-2,7-dienal, and cis-/trans-2-hexenal) emitted from vaping. Density function theory (DFT) and condensed Fukui functions are used to calculate electrophilicity of emitted carbonyls, which will be compared with experimental thiol reactivity using glutathione assays. E-cigarette aerosols will be generated in an indoor Teflon chamber and collected on filters for cytotoxicity testing in human airway epithelial lung cells (BEAS-2B). Carbonyls emitted in gas and particle phase from vaping will be analyzed using high-resolution time-of-flight chemical ionization mass spectrometer (HR-ToF-CIMS) and solid-phase microextraction (SPME) coupled to gas chromatography and mass spectrometry (GC/MS). DFT results show that carbonyl products are reactive with electrophilicities ranged from 1.26 eV (menthone) to 2.12 eV (cis-2-hexenal). Condensed Fukui functions indicate local reactivities are dominated by C_C=O, C_α,C=C and C_β,C=C, which favor nucleophilic attacks for α,β-unsaturated carbonyls (geranial, 2-hexenal and 6-hydroxy-2,6-dimethylocta-2,7-dienal). Our previous study evidenced a good correlation between experimental dithiothreitol reactivity and DFT-calculated electrophilicity for carbonyls. Emitted aerosols and carbonyls from vaping are expected to elicit cytotoxic effects in BEAS-2B. Results from this study will provide insights about toxicity of carbonyls emitted from vaping and associated e-cigarette health risks.

TOXI 46. Liu, Jia | Jacob, Peyton University of California, San Francisco | University of California, San Francisco
1-(3-Pyridinyl)-1,2-dihydroxyethane (3-pyridylethylene glycol) is excreted in smokers’ urine. Evidence for metabolic epoxidation of 3-ethenylpyridine from inhaled tobacco smoke
3-Ethenylpyridine (3-EP) is a major nicotine pyrolysis product present in the gas phase of tobacco smoke. For a number of years it has been used as an environmental tracer for tobacco smoke. It would be expected to be a useful tobacco smoke biomarker as well, but nothing has been published about its metabolism and excretion. As part of our studies to develop 3-EP as a biomarker, we developed analytical methods for likely metabolites. One likely metabolite, 3-pyridylethylene glycol (3-EP Diol) was detected in smokers’ urine and in rat liver homogenates incubated with 3-EP. In addition to being a candidate biomarker, the likely metabolic pathway for its formation would involve the intermediacy of 3-(2-oxiranyl)-pyridine (3-EP Epoxide), followed by conversion to the diol catalyzed by epoxide hydrolase. This may have toxicological significance. Epoxides formed metabolically from alkenes can be carcinogenic or cytotoxic. Butadiene, a human carcinogen, is metabolically activated to epoxides thought to be responsible for its carcinogenicity. Styrene is metabolically activated to an epoxide, styrene-7,8-oxide. Both styrene and its epoxide are considered probable human carcinogens by the International Agency for Research on Cancer (IARC). 3-EP is structurally similar to styrene suggesting possible toxicity. In this presentation we will describe our studies to develop 3-EP as a biomarker and discuss evidence for metabolic formation of 3-EP epoxide.

TOXI 47. Fairman, Kiara National Center for Toxicological Research
Contribution of parameter sensitivities to the predictive potential of pregnancy-induced pharmacokinetic changes for antihypertensives using physiologically based pharmacokinetic (PBPK) modeling
Background
Antihypertensives are the most commonly prescribed medications in pregnancy. However, pregnant women are often excluded from clinical trials due to ethical concerns, making it difficult to accurately dose, access efficacy and safety outcomes. Many physiological changes occur in pregnancy, for example, fluctuations in tissue volumes and organ blood flows, necessitating alternative or adaptive dosing regimens to ensure efficacious and safe drug delivery.
PBPK modeling is a series of mathematical equations that estimate the pharmacokinetics of drugs in real time for dynamically changing life-stages such as pregnancy. These organ compartment concentration estimations help determine if the desired internal concentration is achieved with the external dosage. Characterizing drug interactions with the body’s physiological parameters is key to accurately describing the pharmacokinetics. We implemented drug and model-based sensitivity analysis of the PBPK model to determine key contributing factors to overall drug exposure.
Results
Nifedipine was used as an initial case study and represents a highly protein bound, single cytochrome-P450 metabolized drug. Physiological equations capturing pregnancy related changes were gathered from the literature and implemented in our PBPK pregnancy model. The normalized sensitivity coefficient (NSC) was calculated for various model input parameters using area under the curve (AUC) as the observed model output. Then parameters with the highest NSCs were used to estimate their individual contributions to the cumulative change in AUC during each pregnancy trimester. Intrinsic clearance consistently contributed to over 70% of changes; whereas body weight and hepatic blood flow contributions increased over trimesters (7%, 26%, 41% and 7%, 12%, 27% respectively).
Conclusions
Our method allows for the estimation of the extent each parameter contributes to pregnancy related pharmacokinetic changes. This could help prioritize parameterization of future PBPK models in data-sparse life-stages, thus improving prediction capabilities and sensitivity assessments of PBPK modeling for use as a regulatory tool.

TOXI 48. DeMott, Michael MIT
Phosphorothioate DNA modification in human gut microbiome
Phosphorothioate (PT) is the only known DNA backbone modification, where a non-bridging oxygen located in the phosphate has been replaced with sulfur. As an epigenetic mark in both restriction-modification systems and regulation of gene expression, PTs are inserted by the dnd and ssp gene families. The widespread observation of these gene families in bacteria and archaea raised the question of the presence of PT modifications in the mammalian and human microbiome. Here we present findings that point to an abundance of PT-containing microbes in the mouse and human gut microbiome. Chromatography-coupled mass spectrometric analysis of fecal DNA for PT-containing dinucleotides in a limit digest with nuclease P1 revealed 11 of 16 possible dinucleotides in mice and varying dinucleotide patterns in 4 human donors. The stability of these PT patterns in the human gut over time was assessed in two donors at 0, 17 and 32 weeks, with almost identical patterns across this timeframe. This stable pattern implies that these PT containing organisms are not transient in the gut and likely provide an important contribution to the overall microbiome composition.

We next used a novel deep sequencing method to identify PT-containing microbes in the gut microbiome and map locations for PTs in the microbial genomes. Sequence alignments were performed using the 9000 sequenced genomes of gut microbiome isolates in the Broad Institute Microbiome Library (BML) and the Global Microbiome Conservancy. A genomic analysis of these sequences revealed 722 strains (8%) showing evidence of synteny of dnd and ssp genes. The fecal DNA metagenomics revealed the majority of PT+ strains were from the classes Bacteroidia and Clostridia, four individual PT-containing B. vulgatus strains isolated from the BML matching the sequencing data from the human fecal donors. These donors also showed high levels of GA and GT PT dinucleotides in the fecal DNA, suggesting that these strains are present and contribute at least some portion of these PT dinucleotides found in all of the sequenced donors.

The correlations between the sequenced bacterial isolates, the fecal DNA sequencing metagenomics, and the PT dinucleotide analyses show remarkable epigenetic correlations in the human gut microbiome. More analysis across a larger population will be necessary to confirm and to potentially explore other possible contributing factors, including diet, geographic location, recurring pharmaceutical intake, age, and gender.

TOXI 49. Stornetta, Alessia University of Minnesota
Analysis of double-stranded oligodeoxynucleotides crosslinked by the bacterial genotoxin colibactin using HILIC chromatography and high-resolution mass spectrometry
The E. coli-produced genotoxin colibactin is implicated in colon cancer development, most likely through induction of DNA modifications leading to cancer-initiating mutations. We have previously detected several colibactin-derived DNA modifications, including a crosslinked DNA adduct. To better understand its biological significance and sequence-specificity, we developed a liquid chromatography-mass spectrometry (LC-MS) method to analyze and fragment intact crosslinked double-stranded oligodeoxynucleotides (dsODNs). Traditionally, analysis of oligodeoxynucleotides involves ion-pair reversed-phase LC-MS. However, this approach typically requires significant optimization and instrument dedication to negative mode due to the difficulty in removing the ion-pairing agents and their extreme adverse effect on positive mode operation. Recently, a new diol-based HILIC column has been proven to be a suitable alternative. Our goal is to use this column and negative ionization high-resolution accurate-mass MSⁿ to characterize dsODNs (14-, 25- and 50mers) containing the putative colibactin-binding sequence “AAATT”. Colibactin-producing E. coli grown to log phase were incubated with dsODNs for 5 h at 37 °C. Chromatography was performed using the HILICpak VN-50 2.0 x 150 mm column (Shodex) under gradient conditions using acetonitrile/water with ammonium acetate. Interestingly, a column temperature-dependent melting of the 25mer dsODN into single stranded (ss) ODN was observed, suggesting the possibility to use this phenomenon as a strategy for detecting crosslinked dsODNs, which are expected to be more thermally stable. MS/MS analysis using an Orbitrap Lumos Tribrid found that CID of 30 was qualitatively similar to HCD of 15, whereas an increase in the HCD energy from 15 to 30 resulted in a more ion-rich fragmentation spectrum. Fragment ions common to ssODN and dsODN were identified, indicating possible sequence characterization of unmodified and crosslinked dsODNs by product ion identification using Mongo Oligo Mass Calculator software (v2.08). This method, currently being used for the analysis of crosslinked dsODN samples, has the potential to characterize unmodified and crosslinked dsODNs without ion-pairing reagent requirements and dedicated instrumentation.

TOXI 50. Chen, Yitzai | Li, Deyu URI | URI
Sequence effects of 4-aminobiphenyl DNA lesion-induced mutational spectra
Many agents in environment can result in DNA damage directly or through metabolic processes. 4-aminobiphenyl, a well-known aromatic amine carcinogen present in smoke, can be activated by N-acetyltransferase to from C8-substituted dG-ABP (major), N2-substituted dG-ABP and N6-substituted dA-ABP adducts. These ABP lesions are bulky lesions which have been demonstrated to provide conformational heterogeneities and to affect DNA replication and repair.
In this study, we firstly synthesized and identified the C8-dG ABP adduct containing oligonucleotides. Following the Competitive Replication and Adduct Bypass (CRAB) and Restriction Endonuclease and Post-labeling (REAP) Assays, we examined the replication block and mutagenicity of the lesion and correlated with the conformations of the lesion in ds-DNA. The results show that ABP caused the change of mutational spectra is sequence dependent.

TOXI 51. Wang, Jie | Li, Deyu URI | URI
Structure and mutagenicity of cinnamaldehyde-induced DNA adducts
E-cigarette devices aerosolize flavored e-liquids, which are typically composed of chemical flavorings, humectants, and nicotine. There are more than 7,700 commercially available flavored e-liquids which have not been evaluated for the effect on human health. Cinnamaldehyde, a flavoring additive commonly used in e-liquid, contains α,β-unsaturated carbonyl structure. Due to its reactivity with cellular nucleophiles, it may cause a potential health risk. Our preliminary data showed that cinnamaldehyde could induce DNA damage under physiological condition. We characterize the structures of the DNA adducts by LC-MS and NMR. Furthermore, the genotoxicity and mutagenicity of the DNA lesion are studied in cell though the competitive replication of adduct bypass (CRAB) and the restriction endonuclease and postlabeling analysis (REAP) assays.

TOXI 52. Sun, Jingjing MIT; ETH Zurich
Molecular beacon reporters of substrate specificity of DNA 8-oxoguanine glycosylase mutants
DNA glycosylases play a key role in the recognition and excision of damaged bases, protecting genomic DNA from endogenous and exogenous damaging agents. Moreover, DNA glycosylases are useful tools for evaluating the distribution of damage in genomic DNA. However, these strategies have limitations as many DNA glycosylases can recognize and remove similar adducts on different bases. We are focusing on human 8-oxoguanine DNA glycosylase (hOGG1) which removes 8-oxoguanine (8-oxoG) and 2, 6-diamino-4-oxo-5-formamidopyrimidine (FapyG). To screen out a hOGG1 mutant with higher substrate specificity, we have developed a fluorescent DNA probe based strategy for rapid profiling of glycosylase variants. Nucleotides with oxidative damage (8-oxoG and FapyG) were synthetically incorporated into molecular beacons. A one-pot assay was developed using hOGG1 overexpression E.coli directly without protein purification, exhibiting good sensitivity and specificity towards hOGG1. The signal intensity of color-matched fluorophores indicated the hOGG1 mutant activity targeting 8-oxoG and FapyG. Using this approach, several reported hOGG1 polymorphic variants were studied, showing consistent 8-oxoG/FapyG selectivity as gel-based assay. In addition, a library of Asp322 random mutants was generated and screened. Several mutants with higher activity on FapyG than 8-oxoG were found. The results suggested that Asp322 was essential for 8-oxoG cleavage but not for FapyG. Thus, we developed a method that can be used for fast substrate profiling of glycosylase, benefiting to mutants screening and mechanism studies.

TOXI 53. Petersen, Elijah NIST
Using a measurement science approach to Increase confidence in the EASA skin sensitization assay
NICEATM is coordinating a multi-laboratory validation study to characterize the Electrophilic Allergen Screening Assay (EASA) for risk and hazard assessment for skin sensitization classification. The EASA assay functions by assessing potential skin sensitizers by measuring depletion of one of two probe molecules. The use ofthis assay in a single cuvette format resulted in multiple measurement challenges, such as low throughput and inability to include sufficient control measurements. We redesigned this assay protocol to work with a 96-well plate format. Process control measurements were incorporated into a 96-well platedesign to quantify key sources of variability each time the assay is run. Examples of process controls added are: positive control standard curve, bubble/precipitate interference, and QC charting. We also conducted considerable robustness testing (i.e. stability, pipetting, disposables, etc.) Additional control experiments were performed to evaluate photodegradation of probe molecules, bias from bubbles caused by pipetting, and statistical power of the assay system. Sixty compounds from NTP were tested in the 96-well plate design. One key insight revealed by this process was the interference from test compounds, namely producing anabsorbance or fluorescence signal similar to that of the probe molecules, which would not have been previously detected using the single cuvette assay. Nottaking the interference into account has been shown to lead to potential false negative identification. We also replaced the traditional static call line (i.e. 3 x SD) with a modified t-test to perform custom tailored “call lines” for each compound tested taking into account the uncertainty around its measurements. The data from the 60 compounds were analyzed using a Bayesian analysis as well as the student t-test statistic and has approximately 70 – 75% concordance with LLNA data. Overall, the measurement science approach described here provides steps that can be taken to increase confidence for in vitro assays by fully characterizingsources of variability and potential biases in the assay that will facilitate interlaboratory testing and standardization.

TOXI 54. Bellamri, Medjda University of Minnesota
Cytotoxicity and genotoxicity of tobacco smoke carcinogens in human bladder cells
Bladder cancer (BC) is the fourth most common cancer in men in the United States and the most lethal cancer malignancy of the urinary system. Chronic infections, genetic susceptibilities, environmental, and occupational exposure are risk factors for BC. Smoking remains a primary risk factor for BC and accounts for up to 50% of the cases. Tobacco combustion produces more than 70 carcinogens. However, despite many years of study, the principal chemicals in tobacco smoke and their mechanisms of action leading to BC remain unknown. Aromatic amines, (AAs) such as 4-aminobiphenyl (4-ABP) and 2-naphthylamine (2-NA), formed in tobacco smoke, are implicated in BC of smokers. This association is based on high occupational exposures of textile dye and rubber factory workers. However, 4-ABP and 2-NA only occur at several nanograms per cigarette. These levels may be insufficient to induce BC. In contrast, chemicals such as heterocyclic aromatic amines (HAAs), including 2-amino-9H-pyrido[4,3-b]indole (AaC), and aldehydes, such as acrolein, occur at levels that are 100-fold or higher than the levels of 4-ABP and 2-NA in tobacco smoke. HAAs are multisite rodent carcinogens and potential human carcinogens, and acrolein is a DNA-damaging agent of the bladder.

Thus, we have begun to characterize the cytotoxicity and genotoxicity of tobacco smoke condensate (TSC) of the 1R6F reference cigarette in RT4 cells, a well-characterized human epithelial bladder cell line. We observed a dose and time-dependent cytotoxicity of TSC using the MTT assay. Employing liquid-liquid extraction, we separated TSC into a neutral fraction containing polycyclic aromatic hydrocarbons (PAHs), nitro-PAHs, and hydrophobic aldehydes, and a basic fraction containing HAAs, AAs, and N-nitroso compounds. The TSC cytotoxicity was primarily attributed to the neutral fraction. Moreover, the reactive oxygen species (ROS) scavengers glutathione and N-acetyl-cysteine strongly reduced the TSC neutral fraction-induced cytotoxicity, implicating ROS formation, and oxidative stress as a mechanism of action.

Our ongoing work is devoted to the characterization of bulky, oxidative, and lipid peroxidation DNA adducts formed by chemicals in the TSC basic and neutral fractions using high-resolution mass spectrometry with targeted and untargeted approaches. We plan to identify the major chemicals responsible for DNA damage and cytotoxicity and investigate the role of hepatic metabolism in the cytotoxic and DNA damage of TSC.

TOXI 55. Li, Yupeng University of Minnesota
Identification of N’-Nitrosonornicotine (NNN)-specific DNA Adduct N6-[4-Hydroxy-1-(pyridine-3-yl)butyl]-2′-deoxyadenosine in Rat Liver
The tobacco specific nitrosamines N’-nitrosonornicotine (NNN) and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) are the two most abundant strong carcinogens present in unburned tobacco. They both readily cause tumors in laboratory animals and are classified as “carcinogenic to humans” by the International Agency for Research on Cancer. DNA adduct formation caused by these two carcinogens is believed to play a critical role in tobacco carcinogenesis. Among all the DNA adducts formed by NNN and NNK, 2′-deoxyadenosine-derived adducts have not been fully characterized. In the study reported here, we for the first time characterized a new dAdo adduct N6-[4-hydroxy-1-(pyridine-3-yl)butyl]-2′-deoxyadenosine (N6-HPB-dAdo) formed both in vitro in calf thymus DNA from 5′-acetoxyNNN and in vivo in rat liver. This adduct was specifically formed by NNN 5′-hydroxyaltion. Chemical standards of N6-HPB-dAdo and the corresponding isotopically labelled internal standard [pyridine-D4]N6-HPB-dAdo were synthesized using a four-step method. Both NMR and mass spectrometry data agreed well with the proposed structure of N6-HPB-dAdo. The new adduct co-eluted with the synthesized internal standard under different LC conditions. Its product ion patterns of MS2and MS3 transitions were also consistent with the proposed fragmentation patterns. Chromatographic resolution of the two diastereomers of N6-HPB-dAdo was also successfully achieved. Quantitation suggested a dose-dependent response of the level of this new adduct in the liver of rats treated with NNN. The formation of N6-HPB-dAdo observed in this study leads to new insights in understanding the mechanism of carcinogenesis by NNN.

TOXI 56. Guo, Jingshu University of Minnesota
Development of DNA Adductome Mass Spectral Database
Environmental and dietary genotoxicants, endogenous electrophiles, and ionizing radiation damage the human genome through covalent modifications of DNA, known as DNA adducts. DNA adductomic analysis is a newly established field for the characterization of DNA damage. There is currently no DNA adduct mass spectral database, which greatly limits the power of DNA adductomics analysis. Over 200 authentic DNA adduct standards are being collected from 36 domestic and international collaborators and commercial sources. MSn spectra of the already-received 90 standards are being acquired on Orbitrap and Q-TOF instruments at the MS2 scan stage with multiple collision energies (CE). MS3 fragmentation is also performed on the Orbitrap instrument using data-dependent-MSn data collection, allowing MS2 and MS3 scans at multiple collision energies to be collected in a single injection. A custom bioinformatics workflow has been established, which uses a number of data processing steps to select high quality Orbitrap and Q-TOF MS spectra for inclusion into the spectral library. These steps include selection of spectra arising from precursors with a verified low ppm mass error and sufficient abundance (>50% of the total ion current for the scan), and subsequent data processing using MS-FINDER (v 3.2) for product ion annotation. Consensus spectra are then created for incorporation into MassBank of North America (MoNA), PubChem, and mzCloud (Thermo Fisher Scientific). Related adduct metadata such as the common and IUPAC names of the adducts, canonical and isomeric SMILES notations, and MS scanning parameters will be uploaded to PubChem, from where data will be automatically propagated to other public databases such as Human Metabolome Database (HMDB). MoNA, PubChem, and mzCloud have agreed to provide technical assistance for the integration of our annotated mass spectra and adduct information into their databases.

TOXI 57. Oketola, Adebola University of Ibadan; University of Ibadan
Human exposure risk to dioxin-like PCBs in Nigeria

Dioxins-like polychlorinated biphenyls (DL-PCBs) are highly toxic substances, which belongs to the group of chemically and structurally related halogenated aromatic hydrocarbons. They have been shown to pose severe adverse effects on human health. This study assesses the levels and health risk of DL-PCBs in foods commonly consumed by adult population in Nigeria using southwest as a case study. Eight different food categories were randomly collected and extracted using standard QuEChERS protocol. Clean extracts were analysed by gas chromatography micro-electron capture detector. Aquatic foods had the highest average total DL-PCBs concentration (1.67±1.1pg/g). This was followed by meat products (0.88±0.3 pg/g), dairy products (0.86±0.8 pg/g), edible oils (0.22±0.2 pg/g), fruits (0.14 ±0.1pg/g), vegetables (0.14±0.1pg/g), eggs (0.09±0.1pg/g) and cereals (0.08±0.1pg/g). These concentrations were below the action level of European Commission for DL-PCBs in foods. The estimated average dietary intake of DL-PCBs by an adult was 1.24 pgTEQ/kg bw/day. Health risk assessment of DL-PCBs in foods suggested unlikely harmful effects via exposure through the food intakes with respect to the current concentrations found in the foods commonly consumed in Southwest Nigeria. Routine monitoring of DL-PCBs is recommended to avoid public exposure via the food intakes.

TOXI 58. Narang, Ajit Genentech
Difficult path to simple: enabling discovery to development transition
Toxicology and clinical testing of parenteral products needs limited change and deeper understanding of underlying mechanistic pathways. This talk will discuss on toxicology versus clinical paradigms for product testing, and how timelines come to play a dominating role that sometimes becomes counter-productive. Using case studies, cross-functional discussion of how the relationship between toxicology/TK and product design is understood contemporarily and ways to dissect these mechanistically.

TOXI 59. Thorp, Clare Creme Global
Development of a framework for evidence-based risk assessment of potential carcinogens in human diets
Directly DNA reactive carcinogens are considered to be of most concern to public health because it is assumed that these substances do not demonstrate a carcinogenic threshold below which they are not carcinogenic. More specifically, any exposure is theoretically assumed to be associated with some risk of producing a carcinogenic effect outcome in proportion to the dose. As technological advances have made it possible to detect extremely low levels of chemicals in food, current perceptions of risk and risk tolerance must also evolve because zero risk is not a realistic target for the post-market management of potential carcinogens in food (Gillespie et al., 2011). Additionally, current practices do not effectively incorporate the concept of prioritizing higher risk scenarios for management action, necessary to ensure the most effective use of resources as chemical detection in food continues to evolve and improve.

This presentation describes the implementation of a tiered risk assessment framework, articulated in the form of a mathematical model, to facilitate prioritization and risk management decisions for potential human dietary carcinogens. It is based on the framework published by M. E. Meek et. al (2011) the application of which is not confined to any particular type of chemical or effect. This framework prioritizes contaminants for risk management action based first on exposure to ensure the effective use of resources. A similar tiered risk assessment approach was adopted by the Research Institute for Fragrance Materials, published by Api et al. (2015). It uses Thresholds of Toxicological Concern (TTC) and aggregate exposure assessments to derive Margins of Exposure. The benefit of this approach is that the Margin of Exposure (MoE) concept is more readily understood than estimations of upper bound, excess lifetime cancer risks and it enables the role of the linear no-threshold model to be contrasted with the MoE concept in a more easily understood manner.

This presentation will elaborate on the approach and methodologies used to combine dietary intake data, advanced probabilistic exposure models and toxicological data into an evidence-based risk assessment framework to enable the prioritization of carcinogens in food for further assessment and/or appropriate risk management.

TOXI 60. Bacurio, Jan Henric University of Connecticut
Translesion synthesis of Fapy-dG in human cells is dependent on DNA sequence contexts
Oxidative stress generates Fapy-dG and 8-OxodGuo from dG in comparable yields. These two lesions are believed to arise from a common radical intermediate. 8-OxodGuo is one of the most extensively studied DNA lesions. In contrast, much less is known about the consequences of Fapy-dG in mammalian cells, due to the lack of synthetic approaches to prepare Fapy-dG containing oligonucleotides in a desired sequence. We have recently developed a novel synthetic method to prepare oligonucleotides containing Fapy-dG in any DNA sequence context, which enabled us to synthesize it in three mutational hotspots, namely codons 248 (CG*G), 249 (AG*G), and 273 (CG*T) of the p53 tumor suppressor gene. Plasmid DNA containing Fapy-dG in each of these three sequence contexts was constructed and replicated in human embryonic kidney (HEK) 293T cells. The mutation frequency (MF) induced by Fapy-dG in codons 248, 249, and 273 were 16%, 5%, and 11%, respectively. To determine the roles of the translesion synthesis (TLS) DNA polymerases in Fapy-dG bypass and mutagenesis, these constructs were also replicated in TLS polymerase knockout HEK 293T cell lines. TLS efficiency was reduced in each of the three sites in pol η- and pol κ-deficient cells, suggesting that these two polymerases are important for bypassing Fapy-dG. In codons 248 and 249, a significant increase in MF was observed in pol ι-deficient cells suggesting that pol ι participates in part of the error-free bypass of Fapy-dG. In contrast, in codon 273 MF was reduced upon knockout of pol κ, pol ι, or pol ζ suggesting that these TLS polymerases bypass the lesion in an error-prone manner. The predominant mutation in codons 248 and 249 was G→A transitions, while G→T transversions were dominant in codon 273. The results of this study suggest that Fapy-dG induced mutations in codons 248, 249, and 273 of the p53 tumor suppressor gene in human cells not only display different types and frequencies but the roles of the TLS polymerases are also different in the three different sequence contexts.

TOXI 61. Pennell, Kurt Brown University
Application of high-resolution mass spectrometry for assessment of environmental exposures and metabolic responses
Although humans are exposed to a wide range of chemicals from multiple sources, most environmental exposure studies rely on analytical methods that consider only a limited number of toxicants or biomarkers (typically < 30). However, the recent development of affordable high-resolution mass spectrometers (HRMS) allows for detection of a broad spectrum of environmental, dietary, endogenous, and drug-derived chemicals, allowing for unbiased discovery of relationships between toxicant exposures and metabolic responses. In this work, we present examples of targeted and untargeted methods to investigate associations between exposure to environmental toxicants and alterations in metabolic profiles and pathways using gas and liquid chromatography (GC and LC) HRMS platforms. In this first study, untargeted LC-HRMS analysis of serum samples from 114 8-year old children identified 16,097 metabolic features that were correlated with independently measured serum concentrations of four perfluoroalkyl substances (PFAS) using linear regression adjusted for child age, sex, and race. Serum PFAS concentrations were associated with metabolic features annotated primarily as lipids and dietary factors and alteration of metabolic pathways associated with arginine, proline, aspartate, asparagine, and butanoate metabolism. In the same samples, we also annotated 178 PFAS, of which 137 were categorized into families based on their chemical structures. In the second study, untargeted GC-HRMS was used to analyze full-term amniotic fluid samples in order to identify compounds with potential endocrine disrupting activity. Targeted and untargeted analysis was performed in a single injection. With targeted analysis, we quantified 36 known endocrine disrupting compounds (EDCs), while 3,243 compounds were annotated with untargeted analysis. Annotation was achieved by retention time alignment and mass spectra matching with the unit-mass resolution NIST 17 Mass Spectra Library and an in-house library of certified reference standards. Compounds with high potential endocrine disrupting activity were then profiled for toxicity using an effect-directed analysis approach.

TOXI 62. DeWitt, Jamie East Carolina University
Mechanisms of toxicity for per- and polyfluoroalkyl substances: Are we there yet?
Per- and polyfluoroalkyl substances (PFAS) are a class of over 5,000 individual synthetic fluorinated chemicals used extensively in numerous products and processes. Sub-groups of “long-chain” PFAS that contain more than six or eight carbons and sulfonic or carboxylic acid functional groups have been phased out of production and use in the U.S. and several European countries due to persistence, bioaccumulation, and toxicity (PBT). Replacements for phased out PFAS are being detected in environmental media and biota, and many replacement compounds are understudied with respect to PBT. One perceived limitation associated with the management of PFAS is that a unifying mechanism of toxicological action for adverse health effects in humans has not been agreed upon, even for the well-studied PFAS. However, numerous mechanisms have been posited, ranging from cytotoxicity (necrosis and apoptosis), activation of peroxisome proliferator-activated receptors, and other nuclear receptors, induction of oxidative stress, endocrine disruption, alteration of inflammatory pathways, and modification of cell-cell communication. Some of these mechanisms of toxicological action have been dismissed as not being relevant to human health, not relevant to more than just a few individual PFAS, or only relevant to a limited number of adverse health outcomes. Regardless, these mechanisms demonstrate that PFAS interfere with biological molecules to disrupt physiology and produce toxicity. This presentation will provide an overview of these different mechanisms of PFAS toxicological action, discuss what is known about their operability across cell types, tissues, and organisms, and address their relevance to adverse health effects associated with PFAS exposure in human populations. Supported by the National Institute of Environmental Health Sciences (NIEHS grant P42 ES031009-01).

TOXI 63. Sobus, Jon US EPA
EPA’s research initiatives on non-targeted analyses of environmental chemicals
Thousands of chemicals are present in the air we breathe, water we drink, food we eat, and materials we encounter. Routine environmental and biological monitoring has been performed on <1% of these known compounds, owing to the exclusive use of targeted analytical methods, particularly in state and federal laboratories where emphasis is on regulatory monitoring. The immense paucity of human and ecological exposure data has underscored a need for higher-throughput monitoring methods, capable of identifying emerging contaminants and mapping chemical exposure trends on local, regional, national, and global scales. To address this need, the U.S. EPA’s Office of Research and Development has developed a research program oriented towards advancing non-targeted analysis (NTA) methods. The program includes three primary research components. The first component focuses on developing and integrating models and informatics tools to support NTA applications. The second component focuses on the specific applications of models and tools towards charactering emerging contaminants in high-priority media. The third component focuses on the evaluation of existing or novel NTA methods via an international collaborative study, known as “EPA’s Non-Targeted Analysis Collaborative Trial” (ENTACT). This presentation will highlight: 1) results of a cross-laboratory performance evaluation using the latest findings from ENTACT (including true positive and false negative rates for individual labs); 2) novel modeling approaches being used by ORD to prioritize tentatively identified emerging contaminants based on anticipated health risk; and 3) a web application (i.e., the “EPA NTA WebApp”) developed by ORD to enable widespread utilization of EPA data, models, and workflows. Example WebApp output will be shown, with input data based on recent analyses of environmental (e.g., drinking water) and human biological (e.g., placenta) specimens. The collective efforts described here are moving the exposure and health science communities towards improved speed, expanded coverage, and reduced uncertainty in chemical safety assessments.

TOXI 64. Lein, Pamela University of California, Davis
Sulfated persistent organic pollutants (POPs): mechanistic studies of an emerging class of neurotoxic chemicals
Human exposure to polychlorinated biphenyls (PCBs) remains a significant public health risk, because of not only the persistence of legacy PCBs, but also the unintentional production of contemporary PCBs as byproducts of current pigment production processes. The developing human brain is a vulnerable target of PCBs, but research on PCB developmental neurotoxicity has focused primarily on the legacy higher chlorinated PCBs. By comparison, there is negligible data regarding the developmental neurotoxicity of contemporary PCBs, many of which are lower chlorinated congeners not found in the legacy commercial mixtures. These lower chlorinated PCB congeners have recently emerged as ubiquitous contaminants in various environmental media throughout the world, including air in homes and schools. Of concern, one of the more prevalent contemporary congeners, PCB 11, has been detected in the serum of women and their children, and of pregnant women at risk for having a child with a neurodevelopmental disorder. This presentation will discuss recent data demonstrating that PCB 11 and its sulfated metabolite, which is the most common metabolite found in human serum, disrupt normal patterns of neuronal connectivity in primary rat neuron-glia co-cultures. These effects are observed at concentrations relevant to those found in the human gestational environment. Air-borne sulfated organic compounds have also recently been detected in traffic-related air pollution (TRAP). Emerging mechanistic data suggests that at least a subset of these TRAP-associated organosulfates similarly disrupt normal patterns of neuronal connectivity in vitro. These data sets will be discussed in the context of their relevance to epidemiologic studies implicating PCBs and TRAP as risk factors for neurodevelopmental disorders and neurodegenerative disease.

TOXI 65. Balshaw, David National Institute of Environmental Health Sciences
NIEHS efforts in exposure science and the exposome
The chemical environment that humans live in is a very complex environment, with extensive variation in multiple dimensions, not only in terms of the chemicals we are exposed to but also their variation in time, space, the way we contact them, and the way they interact with our biology. The totality of these multiple dimensions on a comprehensive scale is a concept called the exposome, and efforts to advance our ability to measure the exposome fall into advancing exposure science. The National Institute of Environmental Health Sciences has invested considerable effort over more than a decade to advance exposure science by characterizing the exposome. This presentation will highlight these efforts including both wearable technologies for measuring personal exposures to airborne pollutants as well as efforts to characterize human exposures through the analysis of biological samples through both targeted and untargeted analyses.

TOXI 66. Turesky, Robert University of Minnesota
Targeted and untargeted screening of DNA adducts in the genome of prostate cancer patients
The continuous exposure to exogenous and endogenous genotoxicants results in covalent modifications of DNA, leading to increased risk of diseases, including cancer. Prostate cancer (PC) is the second leading cause of cancer-related death in men. Two major risk factors for PC are lifestyle/dietary factors and inflammation, which contain many types of DNA-damaging agents.

We established liquid chromatography-mass spectrometry (LC-MS)-based targeted methods to quantify 18 DNA adducts of known exogenous and endogenous genotoxicants in prostate tissues of PC patients. This targeted-MS² method frequently detected isomeric lipid peroxidation (LPO) adducts 6/8-hydroxy-1,N²-propano-2′-deoxyguanosine (6/8-OH-PdG). LPO adducts of etheno-2′-deoxyadenosine (edA), 2′-deoxyguanosine adduct of malondialdehyde (M1dG), and a DNA adduct of the cooked meat carcinogen, 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP), N-(2′-deoxyguanosin-8-yl)-PhIP (dG-C8-PhIP), were also detected in about 5% of the prostate specimens.

The totality of DNA adduct formation in PC patients is under study using our recently developed untargeted data-independent acquisition (DIA) method, wide-Selected Ion Monitoring/Tandem MS (wide-SIM/MS²), which screens for the nearly-universal constant neutral loss (CNL) of 2′-deoxyribose (dR) observed upon MS² fragmentation of modified and unmodified 2′-deoxyribonucleosides ions. DIA data was processed using a recently developed automated algorithm, wSIM-City. A list of over 30 adducts (known and unknown) was detected in PC DNA samples through the use of wide-SIM/MS² and wSIM-City. The LPO adducts 6/8-OH-PdG and dG-C8-PhIP were detected as high probability scoring results (P > 0.90), thus demonstrating the ability of wSIM-City to detect DNA adducts in human tissues. We are in the process of characterizing the structures of the other putative adducts by targeted-MS³ analysis.

The combination of targeted-MS² and wide-SIM/MS² data acquisition, with wSIM-City data analysis, can significantly advance our understanding of the relationships between external and internal exposures of harmful chemicals and PC risk.

TOXI 67. Boysen, Gunnar University of Arkansas for Medical Sciences; University of Arkansas for Medical Sciences
Detection and discrimination of DNA adducts differing in size, regiochemistry and functional group by nanopore sequencing
Chemically induced DNA adducts can lead to mutations and cancer. Unfortunately, because common analytical methods (e.g., liquid chromatography-mass spectrometry) require adducts to be digested or liberated the from DNA before quantification, information about their positions within the DNA sequence is lost. Advances in nanopores sequencing technologies allow individual DNA molecules to be analyzed at single-nucleobase resolution, enabling us to study the dynamic epigenetic modifications and exposure-induced DNA adducts in their native forms on the DNA strand. We applied and evaluated the commercially available Oxford Nanopore Technology (ONT) sequencing platform for site-specific detection of DNA adducts and for distinguishing individual alkylated DNA adducts. Using ONT and the publicly available ELIGOS software, we analyzed a library of 15 plasmids containing site-specifically inserted O⁶– or N²-alkyl-2’-deoxyguanosine lesions differing sizes and regiochemistries. Positions of DNA adducts were correctly located and individual DNA adducts were clearly distinguished from each other.

TOXI 68. Peterson, Lisa University of Minnesota
Acrolein increases the pulmonary tumorigenic activity of the tobacco specific nitrosamine, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK)
Tobacco smoke is a mixture of more than 7000 chemicals. Animal studies with purified compounds indicate that at least 70 are toxic and/or carcinogenic. Little is known about how these tobacco chemicals work together to cause disease. Two of the tobacco smoke chemicals are the tobacco specific nitrosamine 4-methylnitrosamino-1-(3-pyridyl)-1-butanone (NNK) and acrolein. NNK is a lung specific carcinogen and is a known human carcinogen. Acrolein is toxic to humans and is mutagenic in model systems. It is thought to contribute significantly to the harmful effects of tobacco smoke. To test our hypothesis that co-exposure to NNK and acrolein is more carcinogenic than either chemical by itself, we exposed A/J mice to NNK (i.p, 0, 2.5, or 7.5 µmol in saline) in the presence or absence of acrolein (0 or 15 ppmv) for 3 hours in a nose-only inhalation chamber. Lung tumors were counted 16 weeks later. While acrolein alone under these exposure conditions had no impact on the number of lung adenomas (Control: 0.2 ± 0.4; acrolein alone: 0.3 ± 0.6 adenomas/mouse), it enhanced the number of lung adenomas observed in NNK-treated mice. This interaction was highly significant (p value = 1.8 x 10^-14). Acrolein more than doubled the adenomas observed in mice treated with 2.5 µmol NNK (NNK alone: 2.3 ± 1.8; NNK and acrolein: 5.9 ± 3.5 adenomas/mouse). The co-exposure effect was less with 7.5 µmol NNK (NNK alone: 12.9 ± 5.2; NNK and acrolein: 18.8 ± 4.9 adenomas/mouse). Acrolein had no impact on pulmonary levels of NNK-derived methyl DNA adducts. This study supports the hypothesis that tobacco smoke chemicals combine to contribute to the carcinogenic potency of the tobacco smoke [funded by CA-184987].

TOXI 69. Stone, Michael Vanderbilt University
Structure of a stable interstrand DNA crosslink involving a β-N-Glycosyl linkage between an N6-dA amino group and abasic (AP) site
Delineating the chemistry and biology of complex DNA damage arising subsequent to N7-alkylation of dG is of considerable interest, since high levels of these alkylation products occur. These N7-dG adducts may yield apurinic sites (AP sites) in DNA. AP sites, once formed, epimerize via a reactive aldehyde intermediate. In sequences in which an adenine nucleotide is located on the opposing strand and offset 1 nt to the 3′-side of the AP site, the N⁶-dA amino group can react with the AP aldehyde intermediate to form an interstrand crosslink (ICL). We constructed this ICL in an oligodeoxynucleotide for NMR studies. The spectra afforded detailed resonance assignments, indicating an ordered structure for the ICL. Structural refinement, using molecular dynamics calculations restrained by NOE data (rMD calculations), afforded a refined structure. The ICL formed with the 2′-deoxyribosyl ring of the AP site in the β configuration. Juxtapositioning the N⁶-dA amine and C1 of the AP site within bonding distance while simultaneously maintaining the two flanking unpaired A⁹ and T²⁹ bases stacked within the DNA was accomplished by unwinding of the DNA at the ICL. The structure will be discussed in the context of studies describing the replication-dependent unhooking of the dA-AP ICL by the base excision repair glycosylase NEIL3.

TOXI 70. Scharer, Orlando Institute for Basic Science; Ulsan National Institute of Science and Technology (UNIST)
Replicative and translesion synthesis DNA polymerases in ICL repair
DNA interstrand crosslinks (ICLs) are formed by antitumor agents such as cisplatin and nitrogen mustards and inhibit DNA metabolism by covalently linking two strands of DNA. ICLs are processed by multiple complex repair pathways in humans, primarily in the context of replication. All of these pathways share a polymerase step past a partially processed or unhooked ICL as a common element, but the questions of which polymerases are involved and the mechanisms by which they synthesize DNA past ICLs remain poorly understood. We have developed synthetic methods for the generation of site-specific ICLs formed by cisplatin and nitrogen mustards and studied the ability of various polymerases to bypass ICLs. We found that two main factors influenced the efficiency of ICL bypass: the length of the dsDNA flanking the ICL and the structure of the crosslink bridging two bases. Interestingly, we found that the replicative DNA polymerases d and e were able to bypass many of the nitrogen mustard, but not the cisplatin ICLs. Our studies reveal important differences in how DNA polymerases process ICL formed by various antitumor agents and may help inform the use of crosslinking agents in personalized medicine.

TOXI 71. Geacintov, Nicholas New York University
Impact of molecular structure and DNA adduct conformation on the repair and helicase-driven DNA unwinding step in human cell-free assays
In human cells, environmental contaminants like polycyclic aromatic hydrocarbons (PAH) or aromatic amines (AA), are metabolized to highly reactive intermediates that bind chemically to nucleotides to generate DNA lesions. Unrepaired, error-prone translesion bypass can give rise to mutations in critical tumor suppressor genes and oncogenes that can play a role in cancer etiology and cancer initiation. A major repair mechanism that removes these genotoxic lesions is the multiple-protein Nucleotide Excision Repair (NER) apparatus that employs a two-stage system for recognizing and excising DNA lesions. The first is the DNA damage-sensing factor XPC-RAD23B factor, and the second is a helicase-driven unwinding mechanism that distinguishes between true chemical DNA damage and aberrant but nontoxic DNA conformations. However, there is significant evidence that some DNA lesions are slowly repaired and some are even fully NER-resistant. While significant progress has been made in understanding the damage-sensing and DNA lesion recognition step, little is known about the helicase-driven verification mechanism and how it depends on the molecular structure and conformation of different DNA lesions. We have developed a new helicase activity assay in human cell extracts that allows us to examine the impact of the structural features and physical sizes of DNA lesions on helicase unwinding rates. This presentation will focus on molecular structure-helicase activity relationships and insights based on Modeling and Molecular Dynamic Simulation methods.

TOXI 72. Suo, Zucai The Ohio State University
Identification and investigation of a novel function of the RuvC nuclease of CRISPR/Cas9
Since the initial characterization of Streptococcus pyogenes CRISPR/Cas9 as a powerful gene-editing tool, it has been widely accepted that Cas9 generates blunt-ended DNA products by concerted cleavage of the target (tDNA) and non-target (ntDNA) strands three nucleotides away from the protospacer adjacent motif (PAM) by HNH and RuvC nuclease active sites, respectively. Following initial DNA cleavage, RuvC catalyzes 3’→5′ degradation of the ntDNA resulting in DNA products of various lengths. Here, we found that Cas9 selects multiple sites for initial ntDNA cleavage and preferentially generates staggered-ended DNA products containing single-nucleotide 5’-overhangs. We also quantitatively evaluated 3’→5′ post-cleavage trimming (PCT) activity of RuvC to find that ntDNA degradation continues up to the -10 position on the PAM distal DNA product and is kinetically significant when compared to extremely slow DNA product release. We also discovered a previously unidentified 5′-3′ PCT activity of RuvC which can shorten the PAM proximal ntDNA product by precisely one nucleotide with a comparable rate as the 3’→5′ PCT activity. Together, our results demonstrate that RuvC-catalyzed post-cleavage trimming ultimately generates DNA fragments with heterogeneous ends following initial DNA cleavage including a PAM proximal fragment with a blunt end and a PAM distal fragment with a staggered-end, 3’-recessed on the ntDNA strand. These kinetic and biochemical findings underline the importance of temporal control of Cas9 during gene-editing experiments and help explain the patterns of nucleotide insertions at sites of Cas9-catalyzed gene modification in vivo.

TOXI 73. Zhao, Linlin University of California, Riverside
Mitochondrial DNA degradation: The potential moonlighting function of mitochondrial transcription factor A
In higher eukaryotic cells, mitochondria are essential subcellular organelles for energy production, cell signaling, and the biosynthesis of biomolecules. The mitochondrial DNA (mtDNA) genome is indispensable for mitochondrial function because it encodes protein subunits of the electron transport chain and a full set of transfer and ribosomal RNAs. MtDNA degradation has emerged as an essential quality control measure to maintain mtDNA and to cope with mtDNA damage resulting from endogenous and environmental factors. Among all types of DNA damage known, abasic (AP) sites, sourced from base excision repair and spontaneous base loss, are the most abundant endogenous DNA lesions in cells. In mitochondria, AP sites trigger rapid DNA loss; however, the mechanism and molecular factors involved in the process remain elusive. Herein, we demonstrate that the stability of AP sites is reduced dramatically upon binding to a major mtDNA packaging protein, mitochondrial transcription factor A (TFAM). The half-life of AP lesions within TFAM-DNA complexes is 2- to 3 orders of magnitude shorter than that in free DNA, depending on their position. The TFAM-catalyzed AP-DNA destabilization occurs with non-specific DNA or mitochondrial light-strand promoter sequence, yielding DNA single-strand breaks and DNA-TFAM cross-links. TFAM-DNA cross-link intermediates prior to the strand scission were also observed upon treating AP-DNA with mitochondrial extracts of human cells. In situ trapping of the reaction intermediates (DNA-TFAM cross-links) revealed that the reaction proceeds via Schiff base chemistry facilitated by lysine residues. Collectively, our data suggest a novel role of TFAM in facilitating the turnover of abasic DNA.

TOXI 74. Shafirovich, Vladimir New York University
Interplay of nucleotide excision repair and base excision repair pathway in plasmid DNA substrates harboring oxidatively generated guanine lesions
The interplay between nucleotide excision repair (NER) and base excision repair (BER) of non-bulky, oxidatively generated DNA lesions has long been a subject of significant interest. We discovered that the repair of oxidatively generated guanine lesions by both NER and BER pathways is more efficient with the lesions embedded in covalently closed circular DNA plasmids than in the same, but linearized plasmid DNA duplexes. The plasmid substrates were constructed from a pUC19-derived pUC19NN (2686 bp) by gapped-vector technology. The gap was filled by radiolabeled 5′-³²pCCATCHCTACC oligonucleotides containing the site-specifically positioned hydantoin (H) oxidation products of 8-oxoguanine, spiroiminodihydantoin (Sp), and 5-guanidinohydantoin (Gh). After incubation of these plasmids in HeLa cell extracts, restriction enzymes were employed to produce shorter fragments for quantifying the yields of NER products (the characteristic 24-32 nt ladder of dual excision fragments) using polyacrylamide gel electrophoresis. In addition, BER single incision products were also identified. In the case of the circular plasmids, the NER dual incision yields were astonishingly high ( ~ 50%), that was about ~4 times greater than in the same, but linearized plasmid. In turn, the yields of BER incision products remained at similar levels of 30-40% in the circular, as well as the linearized DNA plasmids. These results suggest that the modes of binding and/or dissociation of the NER DNA damage sensing factor XPC-RAD23B are remarkably different in circular and linearized DNA.

TOXI 75. Yan, James US Environmental Protection Agency
Evaluation of persistent, bioaccumulative, toxic chemicals under the Frank R. Lautenberg chemical safety for the 21st century act
Under the Frank R. Lautenberg Chemical Safety for the 21^st Century Act, the Environmental Protection Agency (EPA) is required to make an affirmative determination on the safety of all new chemicals. Under this new law, there is particular interest in chemicals that are Persistent, Bioaccumulative, and Toxic (PBT), as these can pose significant risks to human health through the environment. Because of the hazards PBT chemicals can pose in surface water, identifying and regulating PBT chemicals has become a prominent focus under the new law, and the EPA continues to strive for accurate and pertinent chemical review of these new chemicals to ensure the safety of the environment and general population while allowing for continued chemical innovation.

TOXI 76. Kim, Sunghwan National Institutes of Health
Toxicological information in PubChem
PubChem (https://pubchem.ncbi.nlm.nih.gov) is a public chemical information resource at the U.S. National Institutes of Health. It collects chemical information from 700+ data sources and disseminates the collected data to the public free of charge. Arguably, PubChem contains the largest amount of chemical information available in the public domain, with more than 250 million depositor-provided substance descriptions, 100 million unique chemical structures, and 265 million bioactivity outcomes from one million assays covering around twenty thousand unique protein target sequences.
Included in the many types of content in PubChem is toxicological information about chemicals, e.g., human and animal toxicity, ecotoxicity, exposure limits, exposure symptoms, and antidote & emergency treatment. Notably, a substantial amount of toxicological information from resources formerly offered by the TOXicology data NETwork (TOXNET) is now integrated into PubChem, e.g., the Hazardous Substances Data Bank (HSDB), LactMed, and LiverTox. In addition, PubChem contains a large amount of bioactivity and toxicity screening data that can be used to build toxicity prediction models based on statistical and machine-learning approaches. This presentation provides an overview of PubChem’s toxicological information as well as tools and services that help users exploit this information. It also describes how open data in PubChem can be used to develop prediction models for chemical toxicity.