An integrated in silico and bioactivity-based assessment of the potential thyroid-disrupting activity of UV filters

The identification of potential endocrine-disrupting chemicals requires integrated approaches combining structural information, existing biological evidence, and targeted experimental testing. ToxTree, VEGA and thyroid-related ToxCast/ToxCast21 bioactivity data were used as complementary first-line in silico tools to evaluate ten selected UV filters: 2-ethylhexyl-4-methoxycinnamate (EHMC), 4-eminobenzoic acid (PABA), benzophenone, benzophenone-2 (BP-2), benzophenone-3 (BP-3), bezophenone-4 (BP-4), 4-methylbenzylidene camphor (4-MBC), avobenzone, homosalate and octocrylene. Propylthiouracil (PTU), resorcinol, and mefenamic acid were included as reference compounds with documented effects on thyroid-related processes, whereas 3,4-dichloroaniline was included as a presumed negative compound. ToxTree identified several not specific structural alerts in both the reference chemicals and the UV filters. VEGA did not identify reliable activity of any UV filter in the TRα or TRβ models while ToxCast/Tox21 data indicated potential TRα/TRβ antagonist activity for six UV filters, being octocrylene the most active. The strongest findings were obtained for BP-2, for which experimental data indicated DIO1 activity and a high-selectivity effect (HSE) on TPO, and for PABA, which was also experimentally classified as HSE for TPO. TTR-binding predictions were model-dependent: BP-3, avobenzone, and homosalate were predicted to be active with good reliability by the ONTOX model but inactive by the ED-Screen TTR model. Moreover, 3,4-DCA was classified as HSE for TPO with moderate reliability and associated with cytotoxicity according to ToxCast/Tox21, indicating that it should not be considered a universal negative control. Conversely, the ToxCast/Tox21 receptor signals were not corroborated by reliable VEGA TR predictions. Based on the combined bioactivity profiles, BP-2, BP-3, octocrylene and PABA emerged as the most priority UV filters for follow-up experimental testing.