What was tested
Kantaputra and colleagues, led from the Faculty of Dentistry at Chiang Mai University with coauthors at King’s College London, Niigata University, Cincinnati Children’s Hospital, and institutions in Thailand and Turkey, report in PNAS (published online October 28, 2026; DOI 10.1073/pnas.2603057123; PMID 42949964) on the dental side of the ciliopathy gene C2CD3. The gene was previously linked to Orofaciodigital syndrome XIV; here the team assembled 24 patients carrying nine rare and seven novel C2CD3 variants, documented their dental defects, and then modeled the gene in two embryos: a conditional mouse knockout (Ap2Cre;C2cd3 exons 4 to 5 floxed) and talpid2, a classic chicken mutant that carries a C2CD3 variant.
What they found
The patients presented dental defects across number, shape, and timing of tooth development, spanning the keyword phenotypes of tooth agenesis, supernumerary teeth, and natal teeth. In the mouse, mutants showed a range of prenatal dental defects that mimicked several patient phenotypes. Hedgehog signaling diverged by jaw: Shh was upregulated in the developing upper jaw, where the midline widened, while Shh expression was reduced in the lower jaw, a context-dependent effect of the cilium on hedgehog output. In talpid2 chicken embryos, tooth buds were still evident with SHH upregulated early in both upper and lower jaws, a different pattern from the mouse, and cartilage defects appeared in both the mouse knockouts and the chicken embryos. The authors argue that C2CD3-dependent ciliogenesis coordinates the morphogenetic potential of the jaw and the patterning of the forming dentition, conserved across vertebrates.
What it does not show
The evolutionary claim is a hypothesis, not a mechanism demonstrated in the paper. The abstract’s own wording is that alterations in C2CD3-dependent ciliogenesis “may have played a role” in modulating Shh signaling to alter the timing and position of tooth germs across vertebrate evolution. No experiment here reverses tooth loss in a bird lineage, and tooth buds in a mutant chicken embryo are not a reacquisition of a dentition. The 24 patients are rare-variant cases assembled for phenotype description, not a cohort sized for genotype-phenotype penetrance, and the abstract gives no per-phenotype breakdown. All findings are developmental anatomy and expression data: no rescue experiment, no quantified tooth-germ counts in the abstract, and no clinical application. The full text is paywalled, so this piece stays strictly at the level of the abstract and indexing record.
Where it sits in the field
For the bioengineered tooth germ route (/programs/bioengineered-tooth-germ/), the relevance is the developmental-control gap, not a therapeutic result. Whole-tooth engineering has produced teeth in mice, but human protocols need control of tooth number, position, and timing, and this paper maps one upstream control point: a cilia gene that shifts hedgehog output differently in upper and lower jaws, changing how many teeth form, where, and when. Positional memory in reaggregated dental mesenchyme is already on the program’s watch list, and C2CD3 joins that same layer of the problem, the signaling system that patterns tooth germs rather than the cells that build them. The program tier is unchanged; the 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 any headline form of “scientists find the gene that made birds lose their teeth,” we would restate the paper as what it is: a human genetics study of a ciliopathy with dental anomalies, plus embryology consistent across mouse and chicken, from which the authors propose a plausible evolutionary role for cilia-hedgehog tuning and label it as such. The tooth buds in talpid2 embryos are the interesting control data, not evidence that teeth are one mutation away in birds.
Provenance: grounded in the PubMed abstract and the Europe PMC index record (PMID 42949964), retrieved 2026-10-31; the PNAS full text is paywalled, and every number above (24 patients, nine rare and seven novel variants, exons 4 to 5 floxed, PMID 42949964) was read from the abstract or the index record, not from secondary coverage. Method and sourcing standard at /method/.