The model and the exposures
Hasevoets and colleagues built human epithelial tooth organoids (ETO) from dental follicle tissue surrounding unerupted third molars collected at Ziekenhuis Oost-Limburg, with ethics approval 13/0104U from Hasselt University. The organoids were grown in Matrigel drops, passaged every 7 to 10 days, and then switched from proliferation medium (TOM) to mineralization-inducing medium (MIM) for ameloblast-like differentiation. The team tested four exposures linked in epidemiological work to developmental enamel defects: amoxicillin (AMX), bisphenol A (BPA), vinblastine (VB), and vincristine (VC). They analyzed samples after four and seven days of differentiation using AlamarBlue viability, qRT-PCR, RNA-seq, Western blot, and transmission electron microscopy. The study is entirely in vitro.
Amoxicillin suppresses maturation-stage markers
AMX concentrations from 0.1 to 4 mg/mL did not change organoid morphology or reduce metabolic activity relative to MIM controls. After seven days, however, 4 mg/mL AMX reduced expression of the maturation-stage marker amelotin (AMTN) by approximately 2.8-fold compared with MIM controls (qRT-PCR ΔCt: -4.21 ± 1.41 versus -5.65 ± 1.82), while odontogenic ameloblast-associated protein (ODAM) showed a downward trend. Pre-secretory markers DSPP and VWDE did not change. RNA-seq on four biological replicates per condition found only modest global changes, but gene set enrichment analysis showed significant negative enrichment for oxidative phosphorylation, glycolysis, protein secretion, and the unfolded protein response. At the protein level, desmoglein-1 (DSG1) trended downward and HSPA5 protein was nonsignificantly reduced (p = 0.1920) despite increased HSPA5 mRNA, a pattern the authors interpret as possible proteostatic stress. These findings are reported in Figures 1 to 3 of the primary report.
Vinca alkaloids shift differentiation toward an immature state
VB and VC, which bind tubulin and disrupt microtubule dynamics, also left viability above the 50 percent exclusion threshold but altered marker expression. After four days, 250 nM vinblastine increased the pre-secretory marker VWDE 5.6-fold (ΔCt: -16.68 ± 1.34 versus -14.20 ± 1.56 for MIM controls). By seven days, 250 nM and 500 nM vinblastine significantly reduced both AMTN and ODAM in a dose-dependent pattern. Vincristine followed a similar pattern: AMTN was reduced at 500 nM after four days, and both AMTN and ODAM were reduced at 250 nM and 500 nM after seven days. Western blot showed dose-dependent reduction of α-tubulin protein, and electron microscopy revealed multilamellar, myelin-like vesicular structures not seen in controls. The authors read this as cytoskeletal disruption trapping cells in a pre-secretory-like state. These data are in Figures 5 to 7 of the primary report.
What it does not show
BPA behaved differently: only 30 µg/mL was cytotoxic, and non-toxic concentrations (0.2 to 10 µg/mL) did not significantly change the amelogenesis markers tested. The authors explicitly note that the AMX, BPA, and vinca alkaloid concentrations were chosen to bracket non-lethal ranges in the organoid system, not to reproduce pediatric pharmacokinetics. The study does not quantify mineral content, enamel thickness, prism organization, or hardness. There is no animal or patient data. The organoids model exposure-induced disruption of amelogenesis; they do not regenerate enamel or repair a tooth.
Where it sits
For the dental epithelial organoid program, the paper adds a disease-modeling use case: a human-derived system in which environmental and therapeutic exposures can be tested for direct effects on ameloblast maturation. It does not advance the program toward its next milestone of native-thickness, native-oriented enamel in a living jaw, and it does not change the program’s tier. The current field assessment remains at /field/. Three authors disclose that they are inventors on a patent (EP2023054422) related to the generation of human epithelial tooth organoids.
Provenance: the analysis was grounded in the open-access Research Square preprint (doi:10.21203/rs.3.rs-9160520/v1), which the authors identify as the version later published in Stem Cell Research & Therapy (doi:10.1186/s13287-026-05206-y), including the abstract, methods, results, and figure captions, per our method at /method/.