What was tested
Yasmin Mohtadi Hamadani and Laura Evers, co-first authors in Anamaria Balic’s group at the University of Zurich Institute of Oral Biology, with Satu-Marja Myllymäki, Emma Juuri, Maria Jussila and Irma Thesleff in Helsinki, Paul Gueguen at the Functional Genomics Center Zurich and Mina Mina at the University of Connecticut, published in Stem Cells (online 26 November 2025; 29 January 2026 issue, volume 44; open access, CC BY 4.0). The question: when during incisor development does Sox2 mark a true epithelial stem cell, and when does the cervical loop become a functional niche rather than just a shape? They combined sparse genetic lineage tracing in Sox2CreERT2/+; R26RmT/mG and Sox2CreERT2/+; R26RtdT/+ embryos (single tamoxifen injection, 75 mg/kg, with EdU at 25 mg/kg where proliferation was tested), single-cell RNA sequencing of dental epithelium from E11.5 mandibles (8 mandibles, 3 litters), E14.5 incisor buds (4 mandibles, 3 litters) and E16.5 buds (2 mandibles, 2 litters), E14.5 explant culture with the myosin II inhibitor blebbistatin, and two conditional knockouts in epithelium, K14-Cre; Myh9fl/fl and K14-Cre; Foxi3fl/fl. Animal work ran under Finnish license ESAVI/26019/2020; sequencing data sit at GEO GSE299463; funding came from the Jane and Aatos Erkko Foundation, Swiss Dental Association grant 357-24 and Swiss National Science Foundation grant 310030_215594; no conflicts declared.
What they found
First, the stem cells predate the niche. Sox2-expressing cells in the newly formed cervical loops at E14.5 are quiescent (EdU label retention, no clonal expansion into the inner enamel epithelium), and clusters of Sox2 progeny in that differentiating layer first appear at E16.5, not at E15.5: the niche becomes functional between E15.5 and E16.5, matching the appearance of Sfrp5-positive early progeny in their data and the Lgr5, Bmi1 and Oct4 marker domains reported at that stage. Transcriptionally, Sox2-positive cells at E14.5 share 40.5% of genes with their E16.5 counterparts, versus 17.7% with Sox2-positive cells at E11.5, so it is the E14.5 population, not the earlier dental lamina, that feeds the lifelong incisor stem cell pool.
Second, contractile force holds these cells in place until the niche can. Acta2 (alpha-smooth muscle actin) positive cells form a columnar array through the core of the E13.5 bud and extend into the cervical loops as they form; phosphorylated myosin light chain, the activated form that drives myosin II contraction, sits in a continuous band at the leading edge oriented perpendicular to the direction the epithelium is extending. Blebbistatin-treated E14.5 explants lose that confinement: Sox2-positive cells spread across the labial side, leading-edge cells flatten (lower aspect ratio), and the Acta2-positive population expands, while overall bud size and shape are not significantly changed. Myh9 conditional knockout incisors show the same Sox2 spread at E17, milder, which the authors attribute to compensation by other myosin heavy chains. Foxi3 conditional knockout incisors show Sox2 spread plus precocious Sfrp5 activation, that is, premature differentiation, together with significantly reduced myosin light chain phosphorylation at the leading edge and almost no Acta2-positive cells; the result is small, misshapen incisors.
What it does not show
Everything is mouse and embryonic: no human cells, no postnatal or adult niche maintenance, no functional enamel output measured. Force itself is never measured; contractility is inferred from myosin light chain phosphorylation staining plus inhibitor and mutant phenotypes, and the Foxi3-contractility link is correlational. The proposed Acta2 compensation loop is explicitly speculative. One detail a reviewer should flag: the Methods print the blebbistatin concentration as “15 µmol/ml,” which read literally is 15 mmol/L, far above the micromolar range standard in organ culture, so the in vitro dose should be treated with caution even though the genetic data point the same way. Note also what the title compresses: the perturbations show actomyosin maintains and confines already-specified Sox2 stem cells; they do not show that force creates the stem cell state, which is dated by lineage tracing alone.
Where it sits in the field
For the bioengineered tooth germ program, this is a design constraint, not a tier input: a dissociated and reaggregated germ, or a synthetic one, must rebuild not only the right signaling map but the right tissue tension, or its epithelial stem cells drift out of confinement and differentiate early, exactly the failure mode shown here in two mutants. For the dental epithelial organoid program, cervical-loop organoid models now have a mechanical checkpoint to reproduce, alongside the Wnt, Shh and Notch pathways. The result also extends this site’s September mechanobiology record: the YAP-centered machinery this pass traced under orthodontic load in incisor epithelium and in odontoblast polarization now reaches back to niche assembly itself. Nothing here is an animal-jaw or human result, so no tier moves; the current field assessment stands at /field/.
Where we differ from the coverage
We found no press or popular coverage of this paper to differ from. Against the title’s compression, noted above: “require actomyosin forces” is established for niche assembly and stem cell confinement in the embryo, not for stem cell specification itself, and nothing here speaks to maintaining an adult niche or to engineered germs; any application to bioengineering is a design inference, and we have labeled it as one.
Provenance: grounded in the full open-access text (CC BY 4.0) of Mohtadi Hamadani, Evers et al., Stem Cells 44(2), 2026 (online 26 November 2025), DOI 10.1093/stmcls/sxaf074, read in full from the PubMed Central version PMC12855154; every number above was checked against that text. Method and sourcing standard at /method/.