What the case report describes
A 9-year-and-3-month-old boy came to the orthodontic clinic at Fondazione Policlinico Universitario A. Gemelli IRCCS in Rome because of gaps between his incisors. Orthopantomography showed 15 permanent teeth missing (FDI positions 1.2, 1.3, 1.4, 2.2, 2.3, 2.4, 2.5, 3.1, 3.2, 3.3, 3.4, 4.1, 4.2, 4.3, 4.4), eight of them in the mandibular arch, with mild taurodontism of the first permanent molars (Shifman index 20). His father is also affected: seven missing teeth (1.5, 2.5, 3.5, 3.1, 3.8, 4.1, 4.8), taurodontism indices of 30 in three lower molars and 20 in another, and a removable denture replacing his missing central incisors. Neither the proband nor his father shows skin, hair, nail, or sweating abnormalities, which matters because LEF1 variants reported to date usually come with ectodermal findings. The authors report the case in Dentistry Journal, published 10 October 2026.
What the sequencing found
Whole exome sequencing (NovaSeq 6000, DRAGEN v4.3, GRCh38) found no pathogenic or likely pathogenic variants in WNT10A, MSX1, PAX9, or AXIN2 and no relevant copy number changes. It did find one heterozygous missense variant in LEF1 (transcript NM_016269.5): c.903G>C, p.(Lys301Asn), in exon 8. The variant is absent from gnomAD v4.1.1 and from ClinVar as of 3 August 2026. Predictors lean deleterious: REVEL 0.81 (moderate), AlphaMissense 1 (strong), PolyPhen-2 and SIFT damaging. Lys301 sits in the HMG DNA-binding domain, is evolutionarily conserved, and swapping lysine for asparagine removes a positive charge; the structural figure was built on the mouse NMR structure of the LEF1 HMG domain bound to DNA (PDB 2LEF). Sanger segregation in the trio returned the variant in the affected father, absent in the unaffected mother. A younger sibling is too young to assess.
The classification, and an arithmetic tangle
Under the ClinGen semi-quantitative Bayesian reading of ACMG/AMP, the paper reports a cumulative score of +3.0: PM2 at supporting strength (+1) for population absence, PP3 at moderate strength (+2) for the REVEL score. That lands at variant of uncertain significance, class 3. Two things sit oddly next to each other in the results. First, the text also assigns PP1 at supporting strength (+1) for the segregation pattern, and in the same paragraph states that the single informative meiosis is insufficient to apply PP1; the published +3.0 total matches PM2 and PP3 alone, so the segregation points never actually enter the tally. Second, PM1 (location in the HMG domain) was considered and then not applied, on the honest ground that the region is not established as a mutational hotspot devoid of benign variation. Neither point changes the outcome: one affected parent-child pair is one meiosis, the mother was incompletely phenotyped, and the variant stays a VUS either way.
What it does not show
There is no functional data: nothing on expression, protein stability, DNA binding, or Wnt/beta-catenin transcriptional output for the variant protein, and the authors list those assays as the required next step. There is one family and no recurrence in unrelated cases. The sibling is untested. The taurodontism measurements are explicitly descriptive, since 2D orthopantomography cannot be calibrated reliably for a three-dimensional pulp chamber. So the accurate sentence is: a novel, ultrarare LEF1 variant co-occurs with familial tooth agenesis across one informative meiosis and is a candidate, not a cause.
Where it sits in the field
The interest for the anti-USAG-1 program is a mechanistic mirror, and it runs in both directions without touching the program’s evidence tier. TRG-035 works by blocking USAG-1 (SOSTDC1), an antagonist of Wnt and BMP signaling; in animal models, raising that signaling initiates an extra round of tooth development. LEF1 is the transcriptional effector of canonical Wnt signaling, and loss of LEF1 function in mice arrests teeth at early developmental stages. If p.(Lys301Asn) is ever shown to be pathogenic, it would be the human inverse of the antibody’s mechanism: too little Wnt output, too few teeth. That is context for reading the pathway, not evidence about TRG-035. It also sharpens the field assessment: the agenesis genetics record keeps accumulating one-family reports like this one, where segregation and in silico tools generate candidates faster than functional studies can confirm them, and the phenotype range of LEF1 itself (non-syndromic agenesis with taurodontism when heterozygous, per the cases cited in the paper, versus ectodermal dysplasia with limb malformations when biallelic) is still being drawn from a small variant count.