Enabling novel paradigms: a biological questions-based approach to human chemical hazard and drug safety assessment

Throughput needs, costs of time and resources, and concerns about the use of animals in hazard and safety assessment studies are fueling a growing interest in adopting new approach methodologies for use in product development and risk assessment. However, current efforts to define "next-generation risk assessment" vary considerably across commercial and regulatory sectors, and an a priori definition of the biological scope of data needed to assess hazards is generally lacking. We propose that the absence of clearly defined questions that can be answered during hazard assessment is the primary barrier to the generation of a paradigm flexible enough to be used across varying product development and approval decision contexts. Herein, we propose a biological questions-based approach (BQBA) for hazard and safety assessment to facilitate fit-for-purpose method selection and more efficient evidence-based decision-making. The key pillars of this novel approach are bioavailability, bioactivity, adversity, and susceptibility. This BQBA is compared with current hazard approaches and is applied in scenarios of varying pathobiological understanding and/or regulatory testing requirements. To further define the paradigm and key questions that allow better prediction and characterization of human health hazard, a multidisciplinary collaboration among stakeholder groups should be initiated.

Standardizing Extracted Data Using Automated Application of Controlled Vocabularies

BACKGROUND

Extraction of toxicological end points from primary sources is a central component of systematic reviews and human health risk assessments. To ensure optimal use of these data, consistent language should be used for end point descriptions. However, primary source language describing treatment-related end points can vary greatly, resulting in large labor efforts to manually standardize extractions before data are fit for use.

OBJECTIVES

To minimize these labor efforts, we applied an augmented intelligence approach and developed automated tools to support standardization of extracted information via application of preexisting controlled vocabularies.

METHODS

We created and applied a harmonized controlled vocabulary crosswalk, consisting of Unified Medical Language System (UMLS) codes, German Federal Institute for Risk Assessment (BfR) DevTox harmonized terms, and The Organization for Economic Co-operation and Development (OECD) end point vocabularies, to roughly 34,000 extractions from prenatal developmental toxicology studies conducted by the National Toxicology Program (NTP) and 6,400 extractions from European Chemicals Agency (ECHA) prenatal developmental toxicology studies, all recorded based on the original study report language.

RESULTS

We automatically applied standardized controlled vocabulary terms to 75% of the NTP extracted end points and 57% of the ECHA extracted end points. Of all the standardized extracted end points, about half (51%) required manual review for potential extraneous matches or inaccuracies. Extracted end points that were not mapped to standardized terms tended to be too general or required human logic to find a good match. We estimate that this augmented intelligence approach saved > 350 hours of manual effort and yielded valuable resources including a controlled vocabulary crosswalk, organized related terms lists, code for implementing an automated mapping workflow, and a computationally accessible dataset.

DISCUSSION

Augmenting manual efforts with automation tools increased the efficiency of producing a findable, accessible, interoperable, and reusable (FAIR) dataset of regulatory guideline studies. This open-source approach can be readily applied to other legacy developmental toxicology datasets, and the code design is customizable for other study types. https://doi.org/10.1289/EHP13215.

Epithelial ovarian cancer is infiltrated by activated effector T cells co-expressing CD39, PD-1, TIM-3, CD137 and interacting with cancer cells and myeloid cells

Introduction

Despite predicted efficacy, immunotherapy in epithelial ovarian cancer (EOC) has limited clinical benefit and the prognosis of patients remains poor. There is thus a strong need for better identifying local immune dynamics and immune-suppressive pathways limiting T-cell mediated anti-tumor immunity.

Methods

In this observational study we analyzed by immunohistochemistry, gene expression profiling and flow cytometry the antigenic landscape and immune composition of 48 EOC specimens, with a focus on tumor-infiltrating lymphocytes (TILs).

Results

Activated T cells showing features of partial exhaustion with a CD137+CD39+PD-1+TIM-3+CD45RA-CD62L-CD95+ surface profile were exclusively present in EOC specimens but not in corresponding peripheral blood or ascitic fluid, indicating that the tumor microenvironment might sustain this peculiar phenotype. Interestingly, while neoplastic cells expressed several tumor-associated antigens possibly able to stimulate tumor-specific TILs, macrophages provided both co-stimulatory and inhibitory signals and were more abundant in TILs-enriched specimens harboring the CD137+CD39+PD-1+TIM-3+CD45RA-CD62L-CD95+ signature.

Conclusion

These data demonstrate that EOC is enriched in CD137+CD39+PD-1+TIM-3+CD45RA-CD62L-CD95+ T lymphocytes, a phenotype possibly modulated by antigen recognition on neoplastic cells and by a combination of inhibitory and co-stimulatory signals largely provided by infiltrating myeloid cells. Furthermore, we have identified immunosuppressive pathways potentially hampering local immunity which might be targeted by immunotherapeutic approaches.

Workshop Report: Catalyzing Knowledge-Driven Discovery in Environmental Health Sciences through a Harmonized Language

Harmonized language is essential to finding, sharing, and reusing large-scale, complex data. Gaps and barriers prevent the adoption of harmonized language approaches in environmental health sciences (EHS). To address this, the National Institute of Environmental Health Sciences and partners created the Environmental Health Language Collaborative (EHLC). The purpose of EHLC is to facilitate a community-driven effort to advance the development and adoption of harmonized language approaches in EHS. EHLC is a forum to pinpoint language harmonization gaps, to facilitate the development of, raise awareness of, and encourage the use of harmonization approaches and tools, and to develop new standards and recommendations. To ensure that EHLC's focus and structure would be sustainable long-term and meet the needs of the field, EHLC launched an inaugural workshop in September 2021 focused on "Developing Sustainable Language Solutions" and "Building a Sustainable Community". When the attendees were surveyed, 91% said harmonized language solutions would be of high value/benefit, and 60% agreed to continue contributing to EHLC efforts. Based on workshop discussions, future activities will focus on targeted collaborative use-case working groups in addition to offering education and training on ontologies, metadata, and standards, and developing an EHS language resource portal.

GRADE Guidelines 30: the GRADE approach to assessing the certainty of modeled evidence-An overview in the context of health decision-making

OBJECTIVES

The objective of the study is to present the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) conceptual approach to the assessment of certainty of evidence from modeling studies (i.e., certainty associated with model outputs).

STUDY DESIGN AND SETTING

Expert consultations and an international multidisciplinary workshop informed development of a conceptual approach to assessing the certainty of evidence from models within the context of systematic reviews, health technology assessments, and health care decisions. The discussions also clarified selected concepts and terminology used in the GRADE approach and by the modeling community. Feedback from experts in a broad range of modeling and health care disciplines addressed the content validity of the approach.

RESULTS

Workshop participants agreed that the domains determining the certainty of evidence previously identified in the GRADE approach (risk of bias, indirectness, inconsistency, imprecision, reporting bias, magnitude of an effect, dose-response relation, and the direction of residual confounding) also apply when assessing the certainty of evidence from models. The assessment depends on the nature of model inputs and the model itself and on whether one is evaluating evidence from a single model or multiple models. We propose a framework for selecting the best available evidence from models: 1) developing de novo, a model specific to the situation of interest, 2) identifying an existing model, the outputs of which provide the highest certainty evidence for the situation of interest, either "off-the-shelf" or after adaptation, and 3) using outputs from multiple models. We also present a summary of preferred terminology to facilitate communication among modeling and health care disciplines.

CONCLUSION

This conceptual GRADE approach provides a framework for using evidence from models in health decision-making and the assessment of certainty of evidence from a model or models. The GRADE Working Group and the modeling community are currently developing the detailed methods and related guidance for assessing specific domains determining the certainty of evidence from models across health care-related disciplines (e.g., therapeutic decision-making, toxicology, environmental health, and health economics).

Novel text analytics approach to identify relevant literature for human health risk assessments: A pilot study with health effects of in utero exposures

Systematic reviews involve mining literature databases to identify relevant studies. Identifying potentially relevant studies can be informed by computational tools comparing text similarity between candidate studies and selected key (i.e., seed) references. Challenge Using computational approaches to identify relevant studies for risk assessments is challenging, as these assessments examine multiple chemical effects across lifestages (e.g., human health risk assessments) or specific effects of multiple chemicals (e.g., cumulative risk). The broad scope of potentially relevant literature can make selection of seed references difficult. Approach We developed a generalized computational scoping strategy to identify human health relevant studies for multiple chemicals and multiple effects. We used semi-supervised machine learning to prioritize studies to review manually with training data derived from references cited in the hazard identification sections of several US EPA Integrated Risk Information System (IRIS) assessments. These generic training data or seed studies were clustered with the unclassified corpus to group studies based on text similarity. Clusters containing a high proportion of seed studies were prioritized for manual review. Chemical names were removed from seed studies prior to clustering resulting in a generic, chemical-independent method for identifying potentially human health relevant studies. We developed a case study that focused on identifying the array of chemicals that have been studied with respect to in utero exposure to test the recall of this novel literature searching strategy. We then evaluated the general strategy of using generic, chemical-independent training data with two previous IRIS assessments by comparing studies predicted relevant to those used in the assessments (i.e., total relevant). Outcome A keyword search designed to retrieve studies that examined the in utero effects of environmental chemicals identified over 54,000 candidate references. Clustering algorithms were applied using 1456 studies from multiple IRIS assessments with chemical names removed as training data or seeds (i.e., semi-supervised learning). Using a six-algorithm ensemble approach 2602 articles, or approximately 5% of candidate references, were "voted" relevant by four or more clustering algorithms and manual review confirmed nearly 50% of these studies were relevant. Further evaluations on two IRIS assessments, using a nine-algorithm ensemble approach and a set of generic, chemical-independent, externally-derived seed studies correctly identified 77-83% of hazard identification studies published in the assessments and eliminated the need to manually screen more than 75% of search results on average. Limitations The chemical-independent approach used to build the training literature set provides a broad and unbiased picture across a variety of endpoints and environmental exposures but does not systematically identify all available data. Variance between actual and predicted relevant studies will be greater because of the external and non-random origin of seed study selection. This approach depends on access to readily available generic training data that can be used to locate relevant references in an unclassified corpus. Impact A generic approach to identifying human health relevant studies could be an important first step in literature evaluation for risk assessments. This initial scoping approach could facilitate faster literature evaluation by focusing reviewer efforts, as well as potentially minimize reviewer bias in selection of key studies. Using externally-derived training data has applicability particularly for databases with very low search precision where identifying training data may be cost-prohibitive.

Evaluating potential human health risks from modeled inhalation exposures to volatile organic compounds emitted from oil and gas operations

Some states and localities restrict siting of new oil and gas (O&G) wells relative to public areas. Colorado includes a 500-foot exception zone for building units, but it is unclear if that sufficiently protects public health from air emissions from O&G operations. To support reviews of setback requirements, this research examines potential health risks from volatile organic compounds (VOCs) released during O&G operations.We used stochastic dispersion modeling with published emissions for 47 VOCs (collected on-site during tracer experiments) to estimate outdoor air concentrations within 2,000 feet of hypothetical individual O&G facilities in Colorado. We estimated distributions of incremental acute, subchronic, and chronic inhalation non-cancer hazard quotients (HQs) and hazard indices (HIs), and inhalation lifetime cancer risks for benzene, by coupling modeled concentrations with microenvironmental penetration factors, human-activity diaries, and health-criteria levels.Estimated exposures to most VOCs were below health criteria at 500-2,000 feet. HQs were < 1 for 43 VOCs at 500 feet from facilities, with lowest values for chronic exposures during O&G production. Hazard estimates were highest for acute exposures during O&G development, with maximum acute HQs and HIs > 1 at most distances from facilities, particularly for exposures to benzene, 2- and 3-ethyltoluene, and toluene, and for hematological, neurotoxicity, and respiratory effects. Maximum acute HQs and HIs were > 10 for highest-exposed individuals 500 feet from eight of nine modeled facilities during O&G development (and 2,000 feet from one facility during O&G flowback); hematologic toxicity associated with benzene exposure was the critical toxic effect. Estimated cancer risks from benzene exposure were < 1.0 × 10^-5 at 500 feet and beyond.Implications: Our stochastic use of emissions data from O&G facilities, along with activity-pattern exposure modeling, provides new information on potential public-health impacts due to emissions from O&G operations. The results will help in evaluating the adequacy of O&G setback distances. For an assessment of human-health risks from exposures to air emissions near individual O&G sites, we have utilized a unique dataset of tracer-derived emissions of VOCs detected at such sites in two regions of intense oil-and-gas development in Colorado. We have coupled these emission stochastically with local meteorological data and population and time-activity data to estimate the potential for acute, subchronic, and chronic exposures above health-criteria levels due to air emissions near individual sites. These results, along with other pertinent health and exposure data, can be used to inform setback distances to protect public health.

Python BMDS: A Python interface library and web application for the canonical EPA dose-response modeling software

Several primary sources of publicly available, quantitative dose-response data from traditional toxicology study designs relevant to predictive toxicology applications are now available, including the redeveloped U.S. Environmental Protection Agency's Toxicity Reference Database (ToxRefDB v2.0), the Health Assessment Workspace Collaborative (HAWC), and the National Toxicology Program's Chemical Program's Chemical Effects in Biological Systems (CEBS). These resources provide effect level information but modeling these data to a curve may be more informative for predictive toxicology applications. Benchmark Dose Software (BMDS) has been recognized broadly and used for regulatory applications at multiple agencies. However, the current BMDS software was not amenable to modeling large datasets. Herein we describe development and use of a Python package that implements a wrapper around BMDS, a software that requires manual input in the dose-response modeling process (i.e., best-fitting model-selection, reporting, and dose-dropping). In the Python BMDS, users can select the BMDS version, customize model recommendation logic, and export summaries of the resultant BMDS output. Further, using the Python interface, a web-based application programming interface (API) has been developed for easy integration into other software systems, pipelines, or databases. Software utility was demonstrated via modeling nearly 28,000 datasets in ToxRefDB v2.0, re-creation of an existing, published large-scale analysis, and demonstration of usage in software such as CEBS and HAWC. Python BMDS enables rapid-batch processing of dose-response datasets using a modeling software with broad acceptance in the toxicology community, thereby providing an important tool for leveraging the publicly available quantitative toxicology data in a reproducible manner.

ToxRefDB version 2.0: Improved utility for predictive and retrospective toxicology analyses

The Toxicity Reference Database (ToxRefDB) structures information from over 5000 in vivo toxicity studies, conducted largely to guidelines or specifications from the US Environmental Protection Agency and the National Toxicology Program, into a public resource for training and validation of predictive models. Herein, ToxRefDB version 2.0 (ToxRefDBv2) development is described. Endpoints were annotated (e.g. required, not required) according to guidelines for subacute, subchronic, chronic, developmental, and multigenerational reproductive designs, distinguishing negative responses from untested. Quantitative data were extracted, and dose-response modeling for nearly 28,000 datasets from nearly 400 endpoints using Benchmark Dose (BMD) Modeling Software were generated and stored. Implementation of controlled vocabulary improved data quality; standardization to guideline requirements and cross-referencing with United Medical Language System (UMLS) connects ToxRefDBv2 observations to vocabularies linked to UMLS, including PubMed medical subject headings. ToxRefDBv2 allows for increased connections to other resources and has greatly enhanced quantitative and qualitative utility for predictive toxicology.

Reported prevalence of hepatitis C among clients of needle exchanges in southeast Queensland

The aim of the study was to estimate the prevalence of hepatitis C infection among attenders of several high-volume needle exchanges in southeast Queensland, and to compare the prevalences among the needle exchanges. Clients at four needle exchanges were surveyed over a five-day period by means of a self-administered questionnaire. A high proportion of respondents (76 per cent) reported having been tested for hepatitis C antibodies and the overall prevalence of reported hepatitis C infection was 34 per cent. Thirty-one per cent of the respondents were amphetamine injectors, among whom there was a lower prevalence of reported hepatitis C infection than among opioid injectors (odds ratio 0.18, P < 0.01). There were some differences in the respondents' characteristics and in hepatitis C prevalence between different needle exchanges. The reported prevalence of HIV infection was 2 per cent. This study highlights the importance of surveying a range of needle exchanges to obtain a representative sample of needle-exchange clients overall. The low prevalence of hepatitis C in some groups of injecting drug users suggests that it is possible to prevent hepatitis C transmission among injecting drug users, and points to the opportunity for aiming hepatitis C prevention strategies at these groups.

Characterization and pharmacokinetic parameters of recombinant soluble interleukin-4 receptor

The interleukin-4 receptor (IL-4R) is expressed as a 140-Kd membrane glycoprotein that binds IL-4 with high affinity. Recently, cDNA clones for the murine IL-4R have been isolated. One clone encodes an integral membrane protein, while another encodes a protein in which translation is terminated before the transmembrane region, thus producing a soluble form of the IL-4R (sIL-4R). HeLa cell clones overexpressing sIL-4R were isolated using a novel filter-overlay and 125I-IL-4 ligand binding technique. Quantitative analysis demonstrated that the kinetics and affinity of IL-4 binding to the recombinant sIL-4R were similar to the native membrane-bound IL-4R. As low doses of sIL-4R specifically inhibited IL-4-induced proliferative responses in vitro, sIL-4R biodistribution and elimination parameters were evaluated to assess the pharmacokinetic potential of sIL-4R as a therapeutic agent. Pharmacokinetic studies demonstrated that radiolabeled sIL-4R had a distribution half-life of 9 minutes and an elimination half-life of 2.3 hours following intravenous (IV) administration. When administered by intraperitoneal or subcutaneous (SC) injection, the elimination half-lives were prolonged to 4.2 hours and 6.2 hours, respectively. Although the initial blood level of sIL-4R was reduced if administered by SC injection, the bioavailability was comparable with IV administration. The main sites of sIL-4R elimination were the liver and kidney.

Human macrophage colony stimulating factor (M-CSF): alternate RNA splicing generates three different proteins that are expressed on the cell surface and secreted

Human macrophage colony stimulating factor (M-CSF) cDNA clones were isolated from a pancreatic carcinoma cell line. Three different classes of M-CSF precursor protein (256, 554 and 438 amino acids in length) were predicted to be encoded by these cDNAs. Two of these, that we designate M-CSF alpha and M-CSF beta have already been described. The third, M-CSF gamma represents a novel class of M-CSF cDNA. All three precursors share a 32 amino acid signal sequence and the first 149 amino acids of the mature protein. At this position, M-CSF beta and gamma have insertions of 298 and 182 amino acids relative to M-CSF alpha. The first 182 amino acids of these insertions are shared between M-CSF beta and gamma. All three precursors share the C-terminal 75 amino acids that encode the transmembrane and cytoplasmic domains. Expression of all three cDNAs in COS-7 monkey kidney cells gave rise to soluble M-CSF activity, associated with proteins of subunit molecular weight 44 Kda (beta and gamma) or 28 Kda (alpha). In addition, M-CSF proteins could be detected on the surface of the transfected cells by indirect immunofluorescence. When the transmembrane and cytoplasmic domains of M-CSF alpha were removed by introducing a stop codon after amino acid 190, no membrane-bound M-CSF could be detected, but the truncated protein was secreted efficiently and was biologically active. This suggests that all three forms of M-CSF can exist as cell surface proteins, anchored by their hydrophobic transmembrane domains, and can be processed to soluble forms by proteolytic digestion. Although all soluble forms of M-CSF were biologically active in murine bone marrow colony and proliferation assays, they showed greatly reduced or no activity in similar assays using human bone marrow.

Human macrophage-colony stimulating factor: alternative RNA and protein processing from a single gene

Macrophage-colony stimulating factor (M-CSF, CSF-1) has been reported to be required for the proliferation and differentiation of macrophages from hematopoietic progenitor cells. Recently, two human M-CSF cDNA clones were isolated encoding proteins of 256 and 554 amino acids. We report here the isolation of a third M-CSF cDNA that encodes a protein of 438 amino acids. The coding regions for the three cDNA clones share a common amino-terminus of 149 amino acids and a common carboxyl-terminus of 75 amino acids including a membrane spanning region. In addition, we isolated a genomic clone of human M-CSF. When each of the cDNA clones or the genomic clone were transfected into COS-7 monkey kidney cells, biologically active M-CSF was expressed as judged by the ability of transfected cell supernatants to stimulate proliferation and colony formation of murine bone marrow cells, as well as formation of monocytic colonies from human bone marrow cells. Surprisingly, proliferation of human bone marrow cells was not induced by recombinant human M-CSF. Analysis of the M-CSF proteins released by COS-7 cells revealed that monomer subunit proteins of 44 or 28 kDa were produced. In addition, we found that the membrane spanning region, present in all three forms of M-CSF cDNA, was not required for the synthesis of a biologically active protein. However, when the membrane spanning region was present in the three M-CSF cDNAs, cell surface associated forms of M-CSF could be readily detected.

Generation of biologically active interleukin-1 beta by proteolytic cleavage of the inactive precursor

Interleukin-1 beta (IL-1 beta) is derived from an inactive precursor by proteolytic cleavage. To study IL-1 beta processing, we expressed the precursor in Escherichia coli, partially purified it, and used it as a substrate for various potentially relevant protease preparations. The precursor alone was virtually inactive, but incubation with membranes from human monocytes or myeloid cell lines yielded a 500-fold increase in IL-1 bioactivity. Western blot analysis of the incubated material showed that the 31,000-Da precursor is broken down to three major products, ranging from 17,400 to about 19,000 Da. The most active of these products is the smallest one, and it co-migrates during electrophoresis with mature IL-1 beta. Four purified known proteases were also tested for their effect on precursor IL-1 beta, and none of these products co-migrated with the mature protein. Chymotrypsin and Staphylococcus aureus protease yielded slightly larger products, which were highly active. Elastase and trypsin yielded substantially larger products, and these had little IL-1 activity. The products of three of the known proteases were identified by NH2-terminal sequencing. These results show conclusively that proteolysis of precursor IL-1 beta generates biological activity and that the cleavage must occur close to the mature NH2 terminus.

Stimulation of B-cell progenitors by cloned murine interleukin-7

The events involved in the commitment and development of lymphoid lineage cells are poorly understood. We have used a recently described long-term culture system to establish a bioassay that can detect a novel growth factor capable of stimulating the proliferation of lymphoid progenitors. Using direct expression in mammalian cells we have isolated a complementary DNA clone encoding this novel haematopoietic growth factor, designated interleukin-7.

High level stable expression of human interleukin-2 receptors in mouse cells generates only low affinity interleukin-2 binding sites

A bovine papilloma virus-derived vector was used to direct the high level expression in mouse C127 cells of three different cDNAs encoding the human interleukin-2 receptor. These were: the previously described cDNA clone isolated from the T-cell lymphoma, HUT-102; a cDNA clone isolated from mitogen-activated, normal peripheral blood T cells; and an altered version of the HUT-102 receptor in which Ser247, believed to be the site of protein kinase C-mediated phosphorylation, has been changed to an Ala residue. Fluorescence-activated cell-sorting using a monoclonal antibody directed against the human IL-2 receptor was used to derive stable lines of C127 cells expressing from 2-6 X 10(6) IL-2 binding sites per cell. However, all of these receptors bound IL-2 with low affinity.

Phosphorylation of the human interleukin-2 receptor and a synthetic peptide identical to its C-terminal, cytoplasmic domain

The recombinant human interleukin-2 (IL-2) receptor was expressed in mouse mammary epithelial cells following the transfection of these cells with an expression vector containing the human IL-2 receptor cDNA. The recombinant IL-2 receptor in these cells was rapidly phosphorylated in response to phorbol myristate acetate (PMA), but its phosphorylation could not be detected in the absence of PMA or upon addition of human IL-2. The C-terminal, cytoplasmic peptide domain of the IL-2 receptor, Gln-Arg-Arg-Gln-Arg-Lys-Ser-Arg-Arg-Thr-Ile, was synthesized and used as a substrate for protein kinase C. The Km for phosphorylation of the peptide by protein kinase C was 23 microM. The stoichiometry of phosphorylation was 1 mol of phosphate/mol of peptide and serine was the predominant amino acid phosphorylated. Because this peptide was a good substrate for protein kinase C in vitro, it was possible that the same serine (serine 247) was also phosphorylated in the receptor in the cell. The IL-2 receptor gene in the expression vector was therefore altered by site-directed mutagenesis to code for an IL-2 receptor containing an alanine in the place of serine 247. The IL-2 receptor expressed by these cells was not phosphorylated in the presence of PMA. These data suggest that protein kinase C, in response to PMA, phosphorylates the C-terminal serine residue (serine 247) in the human IL-2 receptor.