Amelogenin Peptide TRAP Against Fluoride: A 2025 In Situ Test on Early Enamel Lesions (Bai et al.)

Most of the material in this dossier describes what protein-templated enamel repair might do: patent abstracts, a paper abstract and a press summary. This paper, by Qinghua Bai, Yaru Li, Mingzhu Wen, Dandan Ma, Yiwei Li and Jinpu Chu (received September 2024, accepted 11 February 2025), is different in kind. It puts a small amelogenin-derived peptide into the mouths of volunteers, measures how much softened enamel hardens again, and compares the result directly with plain water and with fluoride. The answer it reports is useful precisely because it is modest: the peptide helped, but fluoride helped considerably more.

What is TRAP, and why test it?

Amelogenin is the dominant protein of the developing enamel matrix (the authors put it at about 90% of enamel matrix proteins), and it is the same family of molecules that the patents US2018236130A1 — Amelogenin-Chitosan Hydrogel for Dentin Hypersensitivity and CN109288685A — Amelogenin-Induced Fluoro-Calcium Silicate Biomimetic Mineralization build on. As enamel matures, enzymes cut amelogenin into fragments. The paper describes the tyrosine-rich amelogenin peptide, TRAP, as the main such fragment: the 45 amino acids at the protein's N-terminus. According to the authors, this stretch carries amelogenin's only phosphorylation site (ser-16), a sequence the paper links directly to amelogenin's self-assembly, and a lectin-binding motif. Their reasoning is that if TRAP holds the functional core of amelogenin, a synthetic copy might steer calcium and phosphate back into early, pre-cavity lesions without fluoride. The group had earlier reported this in vitro; the present study is the step up to an in situ model, in which test enamel sits inside a real mouth but is not part of anyone's teeth. The peptide was made by solid-phase synthesis to match the sequence of porcine amelogenin P173.

How the experiment was built

The work has two parts.

Binding tests. In a test tube, TRAP adsorbed onto hydroxyapatite powder in a way the authors fit to a Langmuir isotherm (R² = 0.9926), giving an affinity K of 259.67 mL/µmol and a maximum of 11.12 × 10⁻² µmol of peptide per square metre of mineral. Fluorescently labelled TRAP placed on slices of bovine enamel left sparse spots on intact enamel but broad, distinct fluorescence on acid-etched enamel, which the authors read as preferential binding to demineralised surfaces.

The in situ trial. Twelve healthy volunteers aged 18–42 each wore a removable acrylic appliance on the cheek side of the lower teeth holding one block of bovine enamel. The blocks had first been softened in an acid solution for three days to mimic early caries, then screened for a narrow hardness range. In a randomised, blinded crossover design, every volunteer went through three 14-day periods, separated by 7-day washouts, receiving a different treatment each time: deionised water (the negative control), TRAP at 100 µg/mL, or sodium fluoride at 2 ppm. The reagent was applied to the block outside the mouth after meals. Volunteers used fluoride-free toothpaste during the study, though they lived in a city whose water supply was fluoridated at 0.17–0.49 mg/L.

Two outcomes were measured: surface microhardness recovery (how much of the hardness lost to acid came back) and integrated subsurface hardness, built from Knoop hardness readings on cut cross-sections at depths from 10 to 220 µm.

What was measured

All 12 volunteers completed the study and no adverse events were reported. Every group hardened significantly compared with its softened baseline, including the water group, which the authors attribute to saliva's own remineralising action.

The three groups differed significantly from one another on both measures, and the paper states that fluoride "had a significantly greater effect than all other treatments." Below the surface, the peptide and fluoride groups were harder than the water group at 30, 50, 70 and 90 µm.

What the evidence does and does not show

The study supports a narrow claim: under these conditions, TRAP produced more rehardening of artificially softened bovine enamel than water did, and less than a low-concentration fluoride control did. On subsurface repair, fluoride's advantage over water was more than twice the peptide's. The authors' conclusion keeps this ordering, describing TRAP as "not as effective as fluoride" while still promising as a non-fluoride agent. Elsewhere, though, the discussion treats the gap between the peptide's surface result and fluoride's as a small one, which undersells a gap of roughly 30% against 49%.

Several limits are visible in the document itself. The authors list three: bovine rather than human enamel, the complexity of the factors that drive demineralisation and remineralisation, and a sample of only 12 people. Beyond those, the model deliberately left the enamel uncovered by gauze so that little plaque accumulated, which the authors argue suits screening but which also removes much of the bacterial challenge that drives real caries. The lesions were laboratory-made, the outcomes are hardness proxies for mineral content rather than clinical endpoints, and the paper's clinical trial number field is marked not applicable. Hardness was not restored to the original level in any group.

Some internal inconsistencies are also worth noting. The abstract describes the subsurface result as an increase relative to water, while the results section describes the same figures as a reduction in lesion size. The surface-recovery formula as printed would give negative values unless its denominator is read the other way round. The discussion describes TRAP as composed of N- and C-terminal segments, whereas the methods define it as the N-terminal 45 residues. The ethics section says participants signed consent forms and also that written consent was waived.

Where it fits in the dossier

This is the first item in the dossier that tests an amelogenin-based approach in people's mouths against a fluoride benchmark. It does not test the elastin-like recombinamer matrix discussed in Elastin-like Recombinamers (ELRs), nor the chitosan hydrogel or fluoro-calcium silicate systems of the patents, so its numbers should not be transferred to them. What it offers is a calibration point: one published in situ comparison in which a biomimetic peptide beat water but clearly trailed fluoride. Set beside the press and patent claims gathered in Enamel Regrowth Claims vs. Evidence: The Press Summary, the Abstract and the Patent Text, it illustrates how much smaller measured effects tend to be than talk of regrowing enamel suggests. For the underlying idea, see Biomimetic Enamel Mineralization and Dental Enamel.

Reference

Bai Q, Li Y, Wen M, Ma D, Li Y, Chu J. "Effect of the recombinant amelogenin peptide TRAP on remineralization of early enamel caries: an in-situ study." Open access, 2025. https://pmc.ncbi.nlm.nih.gov/articles/PMC11866885/