Novel LRP6 Mutations Causing Non-Syndromic Oligodontia

The process of tooth formation is a series of reciprocal interactions between the ectoderm and mesoderm, and it is believed that many genetic factors are involved in this complex process. More than a dozen genes have been identified in non-syndromic tooth agenesis; however, the genetic etiology underlying tooth agenesis is not fully understood yet. In this study, we identified two novel LRP6 mutations in two non-syndromic oligodontia families. Both probands had 16 and 17 missing teeth in their permanent dentition. Mutational analysis identified a de novo frameshift mutation by a 1-bp insertion in exon 9 (NM_002336.2: c.1870dupA, p.(Met624Asnfs*29)) and a splicing donor site mutation in intron 8 (c.1762+2T>C). An in vitro splicing assay confirmed the deletion of exon 8, and the deletion would result in a frameshift. Due to the premature termination codons introduced by the frameshift, both mutant transcripts would be degraded by nonsense-mediated mRNA decay, resulting in haploinsufficiency.


Introduction
Tooth formation is a result of a series of reciprocal ectodermal mesenchymal interactions [1]. Genetic defects or environmental assaults can cause tooth agenesis (TA) or a malformed tooth, such as peg lateralis, cone-shaped tooth, or taurodontism [2]. Due to the shared developmental origin, failure of tooth formation sometimes accompanies alterations in other ectodermally derived structures, such as the nails, hair, and sweat glands [3]. Therefore, TA can be classified as syndromic or non-syndromic TA, but the defects in the same gene can cause both TAs without a clear-cut boundary [4].
Oligodontia is a term that indicates a rare genetic condition in which six or more teeth (excluding wisdom teeth) are congenitally missing [5]. Anodontia, in which all teeth are missing, is extremely rare, but hypodontia, in which five or less teeth are missing, is relatively common. The pattern of TA, the locations of the conserved and missing teeth, could suggest candidate disease-causing genes [6]. Genetic heterogeneity and variable expressivity, however, make it hard to predict an exact culprit that causes selective tooth agenesis [7].
In this study, we recruited families with oligodontia and performed mutational analysis by whole-exome sequencing. The mutational analyses revealed two novel LRP6 mutations, and an in vitro splicing assay verified the effect of the mutation. This report expands the mutational spectrum of the LRP6 gene. A literature review with the identified mutations compared the effect of the LRP6 mutations to advance our understanding of the molecular pathogenesis of TA.

Human Subject Enrollment
Protocols of the study and patient consent were reviewed and approved by the institutional review board of Seoul National University Dental Hospital (CRI05003G and 9 December 2021). Informed consent was obtained from all participating individuals after explaining the nature of the study. The pedigree was drawn with the family histories taken. Clinical examination and sample collection were performed according to the principles in the Declaration of Helsinki.

Genomic DNA Isolation and Whole-Exome Sequencing
Genomic DNA was isolated from 2 mL of peripheral blood by a conventional method with the NucleoSpin Blood L kit (Macherey-Nagel GmbH & Co., Düren, Germany) according to the manufacturer's instructions. After measuring the quality and quantity using the NanoDrop1000 spectrophotometer (Thermo Fisher Scientific, Wilmington, DE, USA), the DNA samples of the probands were submitted for whole-exome sequencing (Yale Center for Mendelian Genomics, West Haven, CT, USA, and Theragen Etex Bio Institute, Suwon-si, Gyeonggi-do, Korea). After exome capturing, paired-end sequencing reads were generated.

Bioinformatic Analysis
Obtained sequencing reads (Table 1) were processed using a series of bioinformatic analyses as previously described [22]. Briefly, after trimming to remove the adapter sequences, the reads were aligned to the reference human genome assembly (hg38). Bioinformatics analysis programs, such as Samtools and Genome Analysis Tool Kit, were used to get a list of sequence variants [23,24]. The dbSNP build 147 database was used for the annotation of the sequence variants, and the annotated variants were filtered with a minor allele frequency of 0.01.

Human Identification and Paternity Test
DNA samples of the trio of family 1 (the proband and parents) were submitted for human identification (Macrogen). A short tandem repeat multiplex assay was performed with the AmpFISTR ® Identifiler ® KIT (Applied Biosystems, Foster City, CA, USA) using 15 tetranucleotide repeat loci and the Amelogenin gender-determining marker: all thirteen of the required loci for the Combined DNA Index System and two additional loci (D2S1338 and D19S433). Genotyping data were obtained with the Applied Biosystems ® 3730/3730xl DNA Analyzer and analyzed with GeneMapper ID v3.2 (Applied Biosystems).

In Vitro Splicing Assay
The wild-type and mutant genomic fragments of the LRP6 gene were amplified from a DNA sample of the proband of family 2 using the Pfu Plus 5x PCR Master mix (Elpis Biotech, Daejeon, Korea) and cloned into the TOPcloner Blunt V2 vector (Enzynomics, Seoul, Korea) using the same primers as in the exon 8 sequencing (including exons 8 and 9). After the sequences were confirmed, the wild-type and mutant fragments were subcloned into the pSPL3 splicing vector after double digestion with NotI and BamHI restriction endonucleases. COS-7 cells were transiently transfected with the wild-type and mutant pSPL3 vectors, and total RNA was isolated after 36 hours, and the cDNA was synthesized. RT-PCR (sense 5 -TCTGAGTCACCTGGACAACC-3 and antisense 5 -AGGAAAGCCTCTGGGACAAT-3 ) was performed with the EzPCR™ Basic 5x Master Mix (Elpis Biotech). Amplification bands were excised from an agarose gel following electrophoresis, purified, and characterized by direct DNA sequencing.

Clinical Phenotype and Mutational Analysis of Family 1
The proband of family 1 was a 5-year-11-month-old female from a non-consanguineous Korean family ( Figure 1). Pregnancy and delivery were uneventful, and she had no remarkable past medical history. Her deciduous dentition was normal without malformed or missing teeth; therefore, her family did not know about the missing multiple teeth in her permanent dentition. They were told about it when an examination was performed at a local dental clinic. The right maxillary second deciduous molar was exfoliated, and the distal-shoe space maintainer was placed. The left maxillary first molar exhibited an ectopic eruption with the resorption of the distal root of the left second deciduous molar. The maxillary first molars had taurodontism. She had 16 permanent teeth missing excluding the third molars (which were also missing), but no other syndromic features including oral exostosis. Her parents were normal without any missing teeth. Therefore, a recessive or de novo dominant mutation was suspected.
Mutational analysis of the whole-exome data resulted in three variants in genes related to TA: heterozygous WNT10A (NM_025216: c.637G>A, p.(Gly213Ser)), homozygous EDAR (NM_022336: c.1138A>C, p.(Ser380Arg)) and heterozygous LRP6 (NM_002336.2: c.1870dupA, p.(Met624Asnfs*29)) variants. The WNT10A variant was inherited from the healthy mother and previously reported to cause autosomal recessive oligodontia [25]. Therefore, this heterozygous variant is not considered as a disease-causing mutation. The parents were heterozygous for the EDAR variant, and the proband was homozygous. This variant was previously reported as a disease-causing mutation to cause autosomal dominant inheritance with high conservation and in silico prediction values. However, it has been shown this variant exhibits an increased EDAR signaling output to a similar level of the variant p.(Val370Ala). Therefore, this variant also is not considered as a diseasecausing mutation, even though it is homogeneous. The heterozygous LRP6 variant was a 1-bp insertion in exon 9 of 23 exons and predicted to cause a change in the reading frame (p. (Met624Asnfs*29)). This frameshift would result in a premature termination codon (PTC) in the early exon; therefore, it would be degraded by the nonsense-mediated mRNA decay system (NMDS) [26], resulting in the haploinsufficiency of LRP6. This mutation was not found in the parents, and the paternity test confirmed that this mutation occurred spontaneously. Therefore, this frameshift mutation caused by a de novo 1-bp nucleotide insertion is believed to cause severe oligodontia in the proband.
EDAR (NM_022336: c.1138A>C, p.(Ser380Arg)) and heterozygous LRP6 (NM_002336.2: c.1870dupA, p.(Met624Asnfs*29)) variants. The WNT10A variant was inherited from the healthy mother and previously reported to cause autosomal recessive oligodontia [25]. Therefore, this heterozygous variant is not considered as a disease-causing mutation. The parents were heterozygous for the EDAR variant, and the proband was homozygous. This variant was previously reported as a disease-causing mutation to cause autosomal dominant inheritance with high conservation and in silico prediction values. However, it has been shown this variant exhibits an increased EDAR signaling output to a similar level of the variant p.(Val370Ala). Therefore, this variant also is not considered as a disease-causing mutation, even though it is homogeneous. The heterozygous LRP6 variant was a 1-bp insertion in exon 9 of 23 exons and predicted to cause a change in the reading frame (p. (Met624Asnfs*29)). This frameshift would result in a premature termination codon (PTC) in the early exon; therefore, it would be degraded by the nonsense-mediated mRNA decay system (NMDS) [26], resulting in the haploinsufficiency of LRP6. This mutation was not found in the parents, and the paternity test confirmed that this mutation occurred spontaneously. Therefore, this frameshift mutation caused by a de novo 1-bp nucleotide insertion is believed to cause severe oligodontia in the proband.

Clinical Phenotype and Mutational Analysis of Family 2
The proband of family 2 was a 14-year-5-month-old male from a non-consanguineous Korean family (Figure 2). He had no other remarkable past medical history. His mother and sister were reported to have missing teeth; however, examination and confirmation were not available. He had 17 missing permanent teeth excluding his missing third molars. Maxillary lateral incisors were peg lateralis bilaterally, and the first molars showed the characteristic feature of taurodontism. Whether the deciduous teeth were missing could not be confirmed. There were no other syndromic features including oral exostosis.

Discussion
In this study, two novel LRP6 mutations were identified, and the mutational effe was confirmed by an in vitro splicing assay. Both mutations would result in the haploi sufficiency caused by mutant mRNA degradation by the NMDS. Both probands exhibit severe TA (16 and 17 missing) in their permanent dentition. Usually, the TA phenoty caused by LRP6 mutations does not, minimal if any (lateral incisors), exhibit any missin deciduous teeth [27]. Even though missing deciduous teeth could not be confirmed in t proband of family 2, the proband of family 1 does not have any missing deciduous teet A previous study reported taurodontism in about one-third of affected individuals [2 and the probands in our study also showed taurodontism: mandibular second molars the proband of family 1 and all four first permanent molars in the proband of family 2.

Discussion
In this study, two novel LRP6 mutations were identified, and the mutational effect was confirmed by an in vitro splicing assay. Both mutations would result in the haploinsufficiency caused by mutant mRNA degradation by the NMDS. Both probands exhibited severe TA (16 and 17 missing) in their permanent dentition. Usually, the TA phenotype caused by LRP6 mutations does not, minimal if any (lateral incisors), exhibit any missing deciduous teeth [27]. Even though missing deciduous teeth could not be confirmed in the proband of family 2, the proband of family 1 does not have any missing deciduous teeth. A previous study reported taurodontism in about one-third of affected individuals [20], and the probands in our study also showed taurodontism: mandibular second molars in the proband of family 1 and all four first permanent molars in the proband of family 2. LRP6 mutations related to tooth agenesis is listed in Table 2. Interstitial loss of 290 kb in 12p13.2 [41] Expressivity is variable, especially wide in some missense mutations [33,38]; furthermore, an individual with incomplete penetrance was reported in a familial case with a frameshift mutation (c.1144_1145dupAG), which is supposed to be degraded by the NMDS [20]. A recent study demonstrated that there could be many more unknown genetic players involved in TA because mutations were identified in only about 25% of the study participants (12 out of 49 subjects) [36]. Additionally, a possible synergistic effect of the LRP6 and WNT10A mutations has been also suggested [30]; therefore, it would be possible that multigenic effects among genes, even unknown and yet to be identified, are involved in TA.
Without proper functional studies, interpretation of an identified potential candidate variant sometimes misleads us to misunderstand the molecular pathogenesis of TA (including other complex disease entities), especially if the variant is rare and the amino acid encoded is highly conserved among species, because in silico predictions would favor the pathologic effect if so [42,43]. Due to the limitation of the repertoire of the known genes, it is possible that it might not be disease-causing but just very rare, and the disease is caused by other unknown gene(s).
The domain structure of the LRP6 protein shows interesting features (Figure 4). LRP6 is a single-pass transmembrane protein with an N-terminal ectodomain with four characteristic β-propeller/EGF-like domain repeats (E1, E2, E3, and E4) and a C-terminus cytoplasmic domain [20]. There is a hinge-like flexibility between the E1E2 and E3E4 structures, and this flexibility enables conformational freedom if there is no ligand bound [44]. This free movement hampers the unwanted formation of homo or heterobinding of LRP6 to form an uncontrolled signaling platform [45]. With a ligand bound such as Wnt, restriction to free movement and conformational change enables the formation of oligomerization to convey controlled signals [46]. Interestingly, a conformational change caused by Dkk1 binding prevents oligomerization [44].  Without proper functional studies, interpretation of an identified potential candidate variant sometimes misleads us to misunderstand the molecular pathogenesis of TA (including other complex disease entities), especially if the variant is rare and the amino acid encoded is highly conserved among species, because in silico predictions would favor the pathologic effect if so [42,43]. Due to the limitation of the repertoire of the known genes, it is possible that it might not be disease-causing but just very rare, and the disease is LPR6 mutations sometimes accompany syndromic features such as sparse hair like in ectodermal dysplasia, a cleft lip and palate, hand preaxial polydactyly, and an orofacial cleft (OFC) ( Table 2) [27,29,31,32,35,38]. The mutation in OFC is a deletion causing a frameshift in the last exon; however, due to the PTC location in the last exon, the mutant transcript would escape from the NMDS and produce a truncated protein with novel amino acids in the cytoplasmic C-terminal region [27]. This would be more harmful than the other frameshift mutations degraded by the NMDS by a dominant negative effect. Ligand binding would be normal with the intact ectodomain, but intracellular signaling would be disturbed or trigger a malfunction. Some mutations show interesting features, such as a high bone mass (with or without torus palatinus) with minimal TA (maxillary only or all four lateral incisors missing) [31,32]. The location of the mutations is in the E1 domain and may suggest involvement of certain factors to cause the high bone mass. Likewise, expanding the mutational spectrum of LRP6 and related phenotypes may provide insight into the complex regulation and function of LRP6-mediated signaling pathways.

Conclusions
In this study, we recruited families with non-syndromic oligodontia and identified two novel LRP6 mutations: a de novo frameshift mutation by a 1-bp insertion in exon 9 (c.1870dupA, p.(Met624Asnfs*29)) and a splicing donor site mutation in intron 8 (c.1762+2T>C). A minigene splicing assay confirmed that the effect of the splicing mutation and the mRNAs of both mutations are predicted to be degraded by the NMDS due to PTC. Further studies including genetic and functional studies are warranted to understand and characterize the complex mechanism of tooth formation.
Author Contributions: Y.L., W.C. and Y.J.K. contributed to analysis and interpretation, data acquisition, and critically revised the manuscript; J.-W.K. contributed to the conception, design, data acquisition, analysis, and interpretation, drafted the manuscript, and critically revised the manuscript. All authors have read and agreed to the published version of the manuscript. Informed Consent Statement: Informed consent was obtained from all individual participants included in the study.