Formation and evaluation of mechanism-based chemical categories for regulatory read-across assessment of repeated-dose toxicity: A case of hemolytic anemia.

The aim of this study is to define chemical categories that can be applied to regulatory read-across assessments for repeated-dose toxicity, by classifying toxic substances based on their structures and mechanism of actions (MoAs). Hemolytic anemia, which often appears primarily, was examined as an example. An integrated database was constructed by collecting publicly available datasets on repeated-dose toxicity, in which 423 out of a total of 1518 chemicals were identified as capable of inducing hemolytic anemia. Subsequently, by grouping these chemicals based on their chemical structures and plausible MoAs on hemolytic substances, we identified the following categories: (i) anilines, (ii) nitrobenzenes, (iii) nitroanilines, (iv) dinitroanilines, (v) ethylene glycol alkyl ethers, (vi) hydroquinones, (vii) oximes, and (viii) hydrazines. In these categories, the toxicant and the measurable key events leading to hematotoxicity were identified, thereby allowing us to justify the categories and to discriminate the category substances. Moreover, toxicokinetics seems to critically affect the hemolytic levels of the category substances. Overall, the categories were validated through a comprehensive analysis of the collected information, while the utility was demonstrated by conducting a case study on the selected category. Further endeavors with this approach would attain categories for other organ toxicity endpoints.

Constructing a developmental and reproductive toxicity database of chemicals (DART NIHS DB) for integrated approaches to testing and assessment.

Developmental and reproductive toxicity (DART) is an important endpoint, and databases (DBs) are essential for evaluating the risk of untested substances using alternative methods. We have constructed a reliable and transparent DART DB, which we named DART NIHS DB, using the publicly available datasets of DART studies of industrial chemicals conducted by Japanese government ministries in accordance with the corresponding OECD test guidelines (OECD TG421 and TG422). This DB is unique because its dataset chemicals have little overlap with those of ToxRefDB, which compiles large-scale DART data, and it is reliable because the included datasets were created after reviewing the individual study reports. In DART NIHS DB, 171 of 404 substances exhibited signs of DART, which occurred during fertility and early embryonic development (49 substances), organogenesis (59 substances), and the perinatal period (161 substances). When the lowest-observed-adverse-effect level (LOAEL) of DART was compared with that of repeated-dose toxicity (RDT), 15 substances (12%) had a lower LOAEL for DART than for RDT. Of these, five substances displayed significant DART at doses of ≤ 50 mg/kg bw/day. The chemical and toxicity information in this DB will be useful for the development of stage-specific adverse outcome pathways (AOPs) via integration with mechanistic information. The whole datasets of the DB can be implemented in read-across support tools such as the OECD QSAR Toolbox, which will further lead to future integrated approaches to testing and assessment based on AOPs.

Determination of serum 7-ketocholesterol concentrations and their relationships with coronary multiple risks in diabetes mellitus.

Oxysterols have cytotoxic effects and contribute to the development of atherosclerosis. To examine association between 7-ketocholesterol and diabetes mellitus, and other coronary risk factors, we developed a reliable quantitative method to measure serum 7-ketocholesterol (s-7KCHO) and studied s-7KCHO in patients with type 2 diabetes mellitus (T2DM). The s-7KCHO was detected by gas chromatography-mass spectrometry assay. The s-7KCHO was significantly higher in patients with T2DM (n=137, 33.8 ng/ml) compared to non-diabetic healthy subjects (n=89, 16.1 ng/ml). Patients with T2DM were divided into two groups with two or more than two risk factors (defined as multiple risk factors group) and with zero or one risk factor (non-multiple risk factors group). The s-7KCHO was significantly higher in multiple risk factors group (39.5 ng/ml) compared to non-multiple risk factors (30.1 ng/ml). Among patients with multiple risk factors group, s-7KCHO was significantly higher in patients with high low-density lipoprotein cholesterol (LDL-C) levels (45.1+/-5.9 ng/ml) compared to those with normal LDL-C levels (35.3+/-7.0 ng/ml). Furthermore, s-7KCHO increased according to the number of concurrent coronary risk factors. These results suggest that serum 7-ketocholesterol levels may depend on the multiple risk factors and serum LDL-C levels.