Nucleotide variants of genes encoding components of the Wnt signalling pathway and the risk of non-syndromic tooth agenesis


  • A Mostowska,

    Corresponding author
    1. Department of Biochemistry and Molecular Biology, Poznan University of Medical Sciences, Poznan, Poland
    • Corresponding author: Dr Adrianna Mostowska, Department of Biochemistry and Molecular Biology, Poznan University of Medical Sciences, 6 Swiecickiego Str.,60-781 Poznan, Poland.

      Tel.: +48 618546513;

      fax: +48 618546510;


    Search for more papers by this author
  • B Biedziak,

    1. Department of Biochemistry and Molecular Biology, Poznan University of Medical Sciences, Poznan, Poland
    2. Private Orthodontic Practice, Poznan, Poland
    Search for more papers by this author
  • M Zadurska,

    1. Department of Orthodontics, Institute of Dentistry, The Medical University of Warsaw, Warsaw, Poland
    Search for more papers by this author
  • I Dunin-Wilczynska,

    1. Department of Jaw Orthopaedics, Medical University of Lublin, Lublin, Poland
    Search for more papers by this author
  • M Lianeri,

    1. Department of Biochemistry and Molecular Biology, Poznan University of Medical Sciences, Poznan, Poland
    Search for more papers by this author
  • PP Jagodzinski

    1. Department of Biochemistry and Molecular Biology, Poznan University of Medical Sciences, Poznan, Poland
    Search for more papers by this author

  • All authors have read and approved the manuscript. The authors declare no conflict of interest.


Tooth agenesis is one of the most common dental anomalies, with a complex and not yet fully elucidated aetiology. Given the crucial role of the Wnt signalling pathway during tooth development, the purpose of this study was to determine whether nucleotide variants of genes encoding components of this signalling pathway might be associated with hypodontia and oligodontia in the Polish population. A set of 34 single nucleotide polymorphism (SNPs) in 13 WNT and WNT-related genes were analyzed in a group of 157 patients with tooth agenesis and a properly matched control group (n = 430). In addition, direct sequencing was performed to detect mutations in the MSX1, PAX9 and WNT10A genes. Both single-marker and haplotype analyses showed highly significant association between SNPs in the WNT10A gene and the risk for tooth agenesis. Moreover, nine pathogenic mutations within the coding region of the WNT10A gene were identified in 26 out of 42 (62%) tested patients. One novel heterozygous mutation was identified in the PAX9 gene. Borderline association with the risk of non-syndromic tooth agenesis was also observed for the APC, CTNNB1, DVL2 and WNT11 polymorphisms. In conclusion, nucleotide variants of genes encoding important components of the Wnt signalling pathway might influence the risk of tooth agenesis.

Tooth agenesis is the most common developmental anomaly of human dentition [1]. According to the number of missing teeth, dental agenesis can be classified as hypodontia (lack of one to five teeth, excluding third molars), oligodontia (lack of six or more teeth, excluding third molars) and anodontia (agenesis of all teeth) [2]. The reported incidence of missing permanent teeth, excluding third molars, varies from 2.6% to 11.3% depending on demographic and geographic profiles [3]. The absence of wisdom teeth occurs in up to 30% of the general population [3]. The prevalence of missing teeth in the primary dentition is considerably lower, ranging between 0.4% and 0.9% in the European population, and about 2.4% in the Japanese population [1]. Dental agenesis may occur either as a non-syndromic (isolated) trait or as a component in congenital syndromes such as hypohidrotic ectodermal dysplasia, Ellis van Crevald syndrome or Rieger syndrome [1, 4].

Molecular studies of odontogenesis, mostly using mouse teeth as models, have led to the discovery of numerous genes and signalling pathways associated with the patterning, morphogenesis and cell differentiation in teeth [5]. Mutations in many of these genes can cause various dental defects in man [6, 7]. MSX1, PAX9, AXIN2 and EDA belong to the well known candidate genes, the mutations and/or polymorphic variants of which are responsible for agenesis of permanent teeth [8-11]. Two recent studies have revealed that WNT10A may also be a susceptibility gene for this dental anomaly [12, 13]. Van den Boogaard et al. have identified potentially damaging WNT10A mutations in 56% of analysed patients with non-syndromic hypodontia, suggesting that WNT10A might be a major gene in the aetiology of this common developmental defect [13].

The WNT10A and AXIN2 genes encode components of the Wnt/β-catenin signalling pathway (known as the canonical Wnt pathway), which plays essential roles at multiple stages of tooth morphogenesis including the activation of mesenchymal odontogenic potential during early tooth development or the induction and maintenance of primary and secondary enamel knots [14, 15]. Wnt/β-catenin signalling has been also reported to play an important role in establishing the size and morphology of the adult tooth [16]. Several Wnts and Wnt pathway mediators are broadly expressed during the development of embryonic teeth [17]. In the absence of Wnt stimulation, cytoplasmic β-catenin is associated with a multiprotein complex (destruction complex) composed of adenomatous polyposis coli (APC), axis inhibition (AXIN) proteins, glycogen synthase kinase 3β (GSK3β) and casein kinase I (CKI). Upon sequential phosphorylation by CKI and GSK3β, β-catenin is ubiquitinated and targeted for proteosomal degradation (Fig. 1a). Binding of a Wnt ligand to a Frizzled receptor and low density lipoprotein-related protein 5/6 (LRP5/6) coreceptors activates intracellular Dishevelled (DVL) proteins leading to the inhibition of the destruction complex and stabilization of cytoplasmic β-catenin. As a result, β-catenin accumulates and translocates into the nucleus where it interacts with lymphoid enhancer factor/T-cell factor family of transcription factors and activates the expression of a number of genes (Fig. 1b) [18]. β-catenin plays a dual role during tooth development, because it is involved not only in Wnt signalling, but also in the cell–cell adhesion mediated by cadherins [19]. Studies in animal models have revealed the inhibition of the canonical Wnt pathway either by deleting of Lef1 function or overexpressing the Wnt inhibitor Dkk1 arrests tooth development at an early bud stage [20-22]. Conversely, the constitutive activation of β-catenin in embryonic oral epithelium results in continuous supernumerary tooth formation as well as abnormal cusp patterning [23, 24].

Figure 1.

Schematic diagram of the Wnt/β-catenin signalling pathway. (a) In the absence of Wnt signal, cytoplasmic β-catenin is associated with a multiprotein destruction complex composed of APC, AXIN proteins, GSK3β and CKI. Upon sequential phosphorylation by CKI and GSK3β, β-catenin is ubiquitinated and targeted for proteosomal degradation. (b) Binding of a Wnt ligand to a FZD receptor and LRP5/6 coreceptors activates intracellular DVL proteins, leading to disassembly of the destruction complex and stabilization of cytoplasmic β-catenin. Consequently, β-catenin accumulates and translocates into the nucleus, where it interacts with the LEF/TCF family of transcription factors and initiates the transcription of the WNT target genes. APC, adenomatous polyposis coli; AXIN, axis inhibition protein; GSK3β, glycogen synthase kinase 3β; CKI, casein kinase I; FZD, Frizzled receptor; LRP5/6, LDL-related protein 5/6 coreceptors; DVL, Dishevelled; LEF/TCF, lymphoid enhancer factor/T-cell factor.

Given the role of the Wnt/β-catenin pathway in tooth development and recent discoveries by Kantaputra et al. and van den Boogaard et al. [12, 13], we conducted an association study to determine whether common nucleotide variations in WNT genes (WNT3, WNT3A, WNT4, WNT5A, WNT10A, WNT10B and WNT11), genes encoding components of the canonical Wnt pathway (APC, AXIN1, CTNNB1 encoding β-catenin, DVL2 and GSK3β) and E-cadherin (CDH1) may contribute to the risk of non-syndromic tooth agenesis. Moreover, we performed direct sequencing to detect mutations in the MSX1, PAX9 and WNT10A genes in 42 patients with both hypodontia and oligodontia.

Materials and methods


Peripheral blood samples were collected from 157 unrelated subjects with non-syndromic agenesis of permanent teeth. The inclusion criterion was congenital agenesis of at least one permanent tooth, excluding third molars. Case eligibility was ascertained by clinicians using dental panoramic radiographs, clinical examination and detailed diagnostic information from medical records. A tooth was considered to be developmentally missing when it was not discerned clinically or radiographically and the dental history revealed no previous extraction. Microdontia was not considered as a form of tooth agenesis in our study. All patients were recruited from the Department of Orthodontics at the Medical University of Warsaw and the Department of Jaw Orthopaedics at the Medical University of Lublin. Blood samples were also collected by one of the authors (B. B.) during her private orthodontic practise. The patient group consisted of 103 (65%) females and 54 (35%) males and included 100 (64%) individuals with hypodontia and 57 (36%) individuals with oligodontia. The median number of missing teeth per person was three (range from 1–26). Fifty four affected individuals (34%) had a positive family history for tooth agenesis. The median age of patients was 18 years (range from 10–36 years). Detailed patient group characteristics are presented in Table 1. The control group consisted of 430 healthy, age and gender matched individuals who were not affected with tooth agenesis (the third molars were not taken into account) and other major craniofacial abnormalities. All study participants were Caucasians of Polish origin. The study was approved by the local Ethics Committee. Written and oral consent was obtained from all participants.

Table 1. Characteristics of the patient group
 All patientsPatients with WNT10A mutationsaPatients without WNT10A mutationsa
  1. a

    Mutations in the WNT10A gene were found in 26 of 42 (61.9%) patients with isolated tooth agenesis.

  2. b

    The distribution of missing permanent teeth showed differences between patients with and without WNT10A mutations, X2 = 18.529 and p = 0.005.

Gender distribution
Type of tooth agenesis (total number of cases)1572616
Hypodontia (1–5 permanent teeth missing)10063.6913.85318.75
Oligodontia (≥6 permanent teeth missing)5736.312596.151381.25
Type of permanent teeth missing (total teeth missing)801323 b123 b
Second premolar25031.217322.603931.71
Lateral incisor20625.726018.583226.02
Second molar9311.615617.34129.76
First premolar8110.113912.071713.82
Central incisor779.613711.461512.20
First molar303.75175.2600.00
Number of cases missing premolars33141.3211234.675645.53
Number of cases missing incisors28335.339730.034738.21
Number of cases missing molars12315.367322.60129.76
Number of cases missing canines647.994112.6986.50
Positive family history5434.391557.691062.50

SNP selection and genotyping

Single nucleotide polymorphisms (SNPs) in WNT genes (WNT3, WNT3A, WNT4, WNT5A, WNT10A, WNT10B, WNT11) and genes encoding components of the canonical Wnt signalling pathway and E-cadherin (APC, AXIN1, CDH1, CTNNB1, DVL2, GSK3β) were identified from public sources such as the NCBI dbSNP database (http://www.ncbi.nlm.nih. gov/projects/SNP/), the HapMap Genome Browser ( and the 1000 Genomes Browser ( A final set of 32 SNPs was selected for genotyping based on functional significance, minor allele frequency (MAF) over 10% in the Caucasian population, linkage disequilibrium (LD) patterns, and association with tooth agenesis or other craniofacial malformations in previous studies. Two polymorphisms of the WNT10A gene (rs121908119 and rs121908120) were also included to this study. These pathogenic variants are very rare in the general population, with MAF of 0.002% and 0.02%, respectively (data from the 1000 Genomes Browser). Characteristics of those SNPs that were finally selected are presented in Table S1. DNA was isolated from peripheral blood lymphocytes by salt extraction method. Genotyping was carried out either by polymerase chain reaction (PCR) followed by digestion of the amplified products with the appropriate restriction enzyme [PCR-restriction fragment length polymorphism (PCR-RFLP)], or by high-resolution melting (HRM) curve analysis on the Bio-Rad CFX96 Real Time PCR system (Bio-Rad, Hercules, CA). As a quality control measure, approximately 10% of randomly selected samples were genotyped in duplicate to check for concordance. Samples that failed genotyping were removed from statistical calculations. Primer sequences and conditions for PCR-RFLP and HRM analyses are presented in Table S2.

Mutation analysis

In 42 patients with non-syndromic tooth agenesis (38 with oligodontia and 4 with hypodontia), mutation screening of the coding regions including exon–intron boundaries of MSX1, PAX9 and WNT10A was performed by direct sequencing. Cycle sequencing was carried out according to the manufacturer's instructions using ABI Prism™ BigDye™ Terminator Cycle Sequencing kit and ABI Prism 3730 capillary sequencer (Applied Biosystems; Foster City, CA). Sequencing template PCR conditions and sequencing primers are presented in Table S3. When sequencing results indicated a possible new aetiological variant, the case sample was re-sequenced or analyzed using either the PCR-RFLP or HRM method. In addition, to exclude the possibility that the novel nucleotide variant might be a rare polymorphism, 280 controls were tested for the presence of this mutation. The putative functional consequences of all identified missense and silent mutations were analysed in silico using polyphen-2 ( and esefinder 3.0 ( tools.

Statistical analysis

For each SNP, the Hardy–Weinberg equilibrium (HWE) was assessed by the chi-squared test in both patients and controls, and a p-value <0.05 was considered as significant disequilibrium. The differences in allele and genotype frequencies between cases and controls were determined using standard chi-square or Fisher tests. SNPs were tested for association with non-syndromic tooth agenesis using the Cochran–Armitage trend test. The strength of the association was measured by odds ratios (ORs) with 95% confidence intervals (CIs). The data were analyzed under recessive and dominant inheritance models. The Bonferroni correction for multiple comparisons was applied to all significant associations, and a p-value <0.0015 (0.05/34) was considered to indicate statistical significance. LD and haplotype association analyses were performed using haploview 4.2 software. Statistical significance was assessed using the 10,000-fold permutation test.


Single-marker association analysis

The concordance rate for the duplicate samples in the quality control test was 100%. All tested SNPs did not show significant deviation from HWE in tooth agenesis patients and control participants (p > 0.05). The only one exception was the WNT10 rs121908119 variant, for which deviation from HWE was observed among cases (p = 0.0317). The MAF for all but the WNT10A markers was at least 18% (Table 2). Analysis of the entire data set for 34 SNPs in 13 Wnt pathway related genes revealed significant associations between polymorphic variants of CTNNB1, DVL2, WNT10A and WNT11 and the risk of non-syndromic tooth agenesis (Table 2). Under assumption of a dominant model (dd + Dd vs DD, where d is the minor allele), the calculated ORs with CIs that did not include 1.0 were observed for the following SNPs: rs4533622 and rs2953 in CTNNB1, rs121908119 and rs121908120 in WNT10A, and rs17749202 in WNT11 (Table 2). For the WNT10A rs121908120 and DVL2 rs35594616 and rs2074222 variants, an association was detected under the recessive model (dd vs Dd + DD, where d is the minor allele; Table 2). After applying the Bonferroni correction for the number of SNPs tested, only the pathogenic variants of WNT10A, rs121908119 and rs121908120, retained statistical significance. For the remaining analyzed SNPs of WNT3, WNT3A, WNT4, WNT5A, WNT10B, APC, AXIN1, CDH1 and GSK3β there was no evidence for both allelic and genotypic association with the risk of non-syndromic tooth agenesis (Table 2).

Table 2. Association of WNT and WNT related gene SNPs with the risk of non-syndromic tooth agenesisa
Geners no.AllelesbMAFGenotypes casescGenotypes controlscptrend valueORdominant (95% CI)d; p valueORrecessive(95% CI)e; p value
  1. CI, confidence interval; MAF, minor allele frequency calculated from the control samples; OR, odds ratio.

  2. a

    Statistically significant results that survived the correction for multiple testing are highlighted in bold (p < 0.0015).

  3. b

    Uppercase denotes the more frequent allele in the control samples.

  4. c

    The order of genotypes: DD/Dd/dd (d is the minor allele).

  5. d

    Dominant model: dd + Dd vs DD (d is the minor allele).

  6. e

    Recessive model: dd vs Dd + DD (d is the minor allele).

  7. f

    Fisher exact test.

APCrs11954856G/t0.4945/67/45115/212/1030.6790.91 (0.61–1.37); 0.6441.28 (0.85–1.92); 0.245
 rs351771c/T0.4148/71/37146/212/720.1231.16 (0.78–1.72); 0.4691.55 (0.99–2.42); 0.055
 rs459552a/T0.2872/67/18215/185/300.1461.18 (0.82–1.70); 0.3741.73 (0.93–3.20); 0.079
AXIN1rs214252A/g0.2396/53/8250/164/160.7600.88 (0.61–1.28); 0.5121.39 (0.58–3.31); 0.457
 rs11649255c/T0.2783/57/16232/162/360.6661.03 (0.71–1.49); 0.8721.25 (0.67–2.33); 0.478
 rs9921222C/t0.4547/76/34129/212/890.8771.00 (0.67–1.50); 0.9881.06 (0.68–1.65); 0.801
 rs1805105C/t0.3663/74/20171/212/470.8130.99 (0.68–1.43); 0.9371.19 (0.68–2.08); 0.542
CDH1rs16260a/C0.2887/55/14224/172/340.6630.86 (0.60–1.25); 0.4311.15 (0.60–2.20); 0.677
 rs9929218a/G0.2979/66/12215/180/350.8910.99 (0.69–1.42); 0.9460.93 (0.47–1.85); 0.845
 rs1801026C/t0.18110/43/4292/121/170.4850.90 (0.61–1.35); 0.6190.64 (0.21–1.92); 0.616f
CTNNB1rs4533622A/c0.4434/86/37134/211/850.0411.64 (1.06–2.52); 0.0241.25 (0.81–1.94); 0.315
 rs2953g/T0.4435/85/37134/212/840.0481.58 (1.03–2.42); 0.0361.27 (0.82–1.97); 0.285
DVL2rs35594616C/t0.3680/66/11186/182/620.0180.73 (0.51–1.06); 0.0970.45 (0.23–0.87); 0.016
 rs2074222a/G0.3581/64/12194/173/630.0370.77 (0.53–1.11); 0.1640.48 (0.25–0.92); 0.024
 rs222836c/T0.4856/68/33122/201/1070.1030.71 (0.48–1.05); 0.0890.80 (0.52–1.25); 0.331
GSK3Brs3732361a/G0.4362/68/27136/218/760.2000.71 (0.49–1.04); 0.0750.97 (0.60–1.57); 0.893
 rs7617372A/g0.21105/42/9276/128/260.5450.87 (0.59–1.28); 0.4840.95 (0.44–2.08); 0.900
 rs334558A/g0.4065/69/22147/219/640.1940.73 (0.50–1.06); 0.0960.94 (0.56–1.59); 0.813
WNT3rs3809857G/t0.3562/72/22177/203/500.5341.06 (0.73–1.54); 0.7571.25 (0.73–2.14); 0.420
 rs12452064a/G0.4738/77/42117/223/900.1751.17 (0.77–1.79); 0.4651.38 (0.90–2.11); 0.135
 rs7207916a/G0.3869/68/19166/205/590.2590.79 (0.55–1.15); 0.2190.87 (0.50–1.52); 0.627
 rs9890413A/g0.2683/64/10226/181/230.8990.99 (0.68–1.43); 0.9471.20 (0.56–2.59); 0.635
WNT3Ars708111c/T0.4949/68/39109/223/980.5610.74 (0.50–1.11); 0.1441.13 (0.74–1.73); 0.577
 rs752107C/t0.3370/67/19193/188/490.8991.00 (0.69–1.45); 0.9981.08 (0.61–1.90); 0.793
WNT4rs3765350c/T0.22105/47/5262/151/170.2010.77 (0.53–1.14); 0.1880.80 (0.29–2.20); 0.664
 rs7526484C/t0.2684/58/14232/170/280.6611.00 (0.70–1.45); 0.9821.42 (0.72–2.77); 0.307
WNT5Ars1047898A/g0.4359/77/20138/216/760.0990.78 (0.53–1.14); 0.1950.69 (0.40–1.17); 0.161
 rs566926a/C0.2987/59/11222/168/400.3140.86 (0.59–1.24); 0.4160.73 (0.37–1.47); 0.382
WNT10Ars121908119a/C0.00143/12/2430/0/0<0.000187.00 (5.15–1468.70); < 0.0001f14.30 (0.68–299.83); 0.0679f
 rs121908120a/T0.01124/29/4418/12/0<0.00019.27 (4.65–18.49); <0.000125.24 (1.35–471.89); 0.005f
WNT10Brs3741627g/T0.4059/79/18155/202/730.2640.93 (0.63–1.35); 0.6930.64 (0.37–1.11); 0.108
 rs833840C/g0.4049/80/27158/199/730.3871.27 (0.86–1.88); 0.2331.02 (0.63–1.66); 0.925
WNT11rs17749202c/T0.3354/82/21194/185/510.0521.57 (1.07–2.29); 0.0201.15 (0.67–1.98); 0.620
 rs94111a/G0.3762/71/23180/186/640.7631.09 (0.751–1.59); 0.6460.99 (0.59–1.66); 0.966


The rs121908120 polymorphism (p.Phe228Ile) showed a significantly higher frequency in patients with isolated tooth agenesis compared with controls (MAF of 0.12 and 0.01, respectively; p < 0.0001). The calculated OR for individuals carrying the rs121908129 A allele (AA or AT genotype) compared with TT homozygotes was 9.27 (95% CI: 4.65–18.49, p = 2.018 × 10−13). An increased risk of dental agenesis was also observed for AA homozygotes compared with individuals with the T allele (AT or TT genotype), however, this result was not statistically significant after adjustment for multiple testing (OR = 25.24, 95% CI: 1.35–471.89, p = 0.005). The rs121908119 variant (p.Cys107Ter) was identified in both homozygous and heterozygous form only in patients with isolated tooth agenesis. The OR for affected individuals carrying the rs121908119 risk allele was 87.00 (95% CI: 5.15–1468.70, p = 3.670 × 10−10).

Haplotype analysis

Haplotype analysis of polymorphisms in the APC, DVL2 and WNT10A genes revealed several 2-SNP and 3-SNP haplotypes significantly associated with the risk of non-syndromic tooth agenesis (Table 3). The strongest correlation with this dental anomaly was observed for the C-T and C-A haplotypes consisting of WNT10A rs121908119 and rs12190812 pathogenic variants (p = 3.889 × 10−20 and p = 1.11 × 10−11, respectively). These results remained highly significant even after applying the permutation-based correction. In the case of APC and DVL2 variants, the haplotype analysis showed stronger evidence for association with the risk of tooth agenesis than the most significant single marker analysis (Tables 2 and 3). Haplotype analysis of the remaining tested genes did not show differences in haplotype frequencies between cases and controls (Table S4).

Table 3. Analysis of haplotypes in the APC, DVL2 and WNT10A genes and the risk of non-syndromic tooth agenesisa
GenePolymorphismsHaplotypesFrequencyCase, control ratiosX2p Valuepcorr valueb
  1. a

    Statistically significant results are highlighted in bold.

  2. b

    p-Value calculated using permutation test and a total of 10,000 permutations.

APCrs11954856_rs351771GT0.4990.483, 0.5040.4370.50870.9013
  TC0.4140.447, 0.4021.8900.16920.4382
  TT0.0740.053, 0.0822.7010.10030.2803
  GC0.0130.017, 0.0120.5430.46120.8517
 rs351771_rs459552TT0.5540.535, 0.5610.6390.42400.8333
  CA0.2780.328, 0.2605.3310.02100.0673
  CT0.1490.134, 0.1540.7310.39260.8104
  TA0.0190.003, 0.0255.9250.01490.0434
 rs11954856_rs351771_rs459552GTT0.4800.480, 0.4790.0000.98531.0000
  TCA0.2760.328, 0.2575.7480.01650.0616
  TCT0.1380.118, 0.1451.3800.24010.7341
  TTT0.0740.054, 0.0822.5730.10870.4311
  GTA0.0190.003, 0.0256.0920.01360.0492
  GCT0.0110.016, 0.0091.1510.28340.7912
DVL2rs35594616_rs2074222CG0.6440.717, 0.6189.9860.00160.0153
  TA0.3080.274, 0.3212.3280.12700.4800
  TG0.0260.002, 0.03510.3670.00130.0144
  CA0.0210.007, 0.0264.1310.04230.2177
 rs2074222_rs222836GT0.5290.564, 0.5162.2000.13800.3057
  AC0.3280.276, 0.3475.2060.02250.0463
  GC0.1420.156, 0.1360.7370.39050.6474
 rs35594616_rs2074222_rs222836CGT0.5090.564, 0.4895.1680.02300.1009
  TAC0.3080.275, 0.3212.2990.12950.4523
  CGC0.1350.154, 0.1291.2480.26390.7114
  TGT0.0200.001, 0.0277.9980.00470.0202
  CAC0.0190.004, 0.0255.5230.01880.0794
WNT10Ars121908119_rs121908120CT0.9520.853, 0.98684.4763.889E-200.000E0
  CA0.0360.100, 0.01446.1151.115E-110.000E0

Mutation analysis

Mutation screening of coding sequences of the MSX1, PAX9 and WNT10A genes in 42 patients with non-syndromic tooth agenesis revealed mutations in 27 individuals (64%). All of the mutations except one, which was found in PAX9, were identified in the WNT10A gene (Table 4). No mutation was found in the MSX1 gene.

Table 4. Phenotypes of patients with mutations identified in the WNT10A and PAX9 genesa
     Right upper jaw (q1)Left upper jaw (q2)
Patient IDGenderNumber of missing teethbFamily historyMutationRight lower jaw (q4)Left lower jaw (q3)
  1. M, male; F, female; HET, mutation identified in heterozygous form; HOM, mutation identified in homozygous form.

  2. a

    Missing teeth are indicated with X.

  3. b

    Number of missing teeth excluding third molars.

WNT10A gene
B.DF20YesAla135Asp_HET + Phe228Ile_HETX XXXX  XXXX X
A.S.M17NoCys107Ter_HET + Phe228Ile_HETX X  X  XX X X
M.KM16NoCys107Ter_HET + Phe228Ile_HETX XX X  X XX  
     X X XXXX X X X
A.P.F15YesPhe228Ile_HOMX X  X  X  X X
     X XX XXXX  X X
R.BM14YesCys107Ter_HET + Phe228Ile_HET  X XX  X  X  
     X   XXXXXX X X
K.K.F14NoCys107Ter_HET + Phe228Ile_HETX XX X  X XX X
     X XX      XX X
A.Z.F13YesPhe228Ile_HET + Met375Thr_HETX   XX  XX   X
     X X  XXXX    X
M.W.M13NoCys107Ter_HET + Phe228Ile_HET X  XX   XX X 
     X XX   X  XX X
M.BF13Yes?Phe228Ile_HOM  XXXX  XXXX  
       XX X    XX  
       X        X  
I.T.F11NoArg128Ter_HET + Phe228Ile_HET  XX      XX  
       XX  XX  XX X
K.MF11YesCys107Ter_HET + Phe228Ile_HET             X
WNT10A gene
M.ŁM10YesPhe228Ile_HOM  X XX  XX X  
       X X    X X  
A.M.F10NoPhe228Ile_HOMX XXX      X  
       XX      XX X
S.OF10YesPhe228Ile_HETX X  X     X X
      XX   X    X X
C.GM9YesAla315Ala_HETX    X  X    X
     X     XXX    X
D.NM9NoPhe228Ile_HET  XXX    X X  
       XX      XX  
J.L.F9YesCys107Ter_HET + Arg113Cys_HETX    X  X    X
     X    XXX     X
A.CF8YesArg113Cys_HET + Phe228Ile_HET  X  X  X  X  
       X    XX  X  
A.BM7YesArg171Cys_HET     X  X     
       XX  XX   X  
J.K.M6YesArg171Cys_HET     X  X     
       X   XX   X  
J.H.F6NoPhe228Ile_HET X   X  X   X 
      X          X 
K.T.F4NoCys96Arg_HET + Cys107Ter_HET              
PAX9 gene
B.MM9YesSer119AlafsX199_HET XX        XXX
     X     XX     X


Nine nucleotide substitutions within the coding region of the WNT10A gene were identified in 26 probands (Fig. 2); 7 patients were homozygous, 12 patients were compound heterozygous and 7 patients were heterozygous for a single mutation (Table 4). Two identified WNT10A mutations were predicted to cause premature termination of translation (p.Cys107Ter and p.Arg128Ter), six were missense mutations affecting highly evolutionary conserved amino acids and predicted to be potentially damaging (p.Cys96Arg, p.Arg113Cys, p.Ala135Asp, p.Phe228Ile, p.Arg171Cys and p.Met375Thr) and one was a silent mutation affecting exonic splicing enhancer elements (p.Ala315Ala). Three identified mutations, specifically p.Ala135Asp, p.Ala315Ala and p.Met375Thr, were novel, previously undescribed WNT10A variants. PCR-RFLP analyses revealed that these novel variants were absent in 560 control chromosomes of healthy individuals. Two mutations, p.Cys107Ter (rs121908119) and p.Phe228Ile (rs121908120), considered to be pathogenic mutations, were previously selected for our association study. While the p.Cys107Ter variant was not identified in healthy controls, p.Phe228Ile heterozygous carriers were found within the control population (Results section, Table 2).

Figure 2.

Distribution of mutations identified in the WNT10A gene in non-syndromic tooth agenesis patients. Previously undescribed WNT10A mutations are marked with an asterisk.

In the group of 26 probands with WNT10A mutations, the mean number of missing permanent teeth per person, excluding third molars, was 12.3 (range 4–26) compared with 7.7 (range 2–11) in the non-mutation group of patients (n = 16). The percentage of tooth agenesis per tooth type showed differences between WNT10A mutation carriers and non-carriers (p = 0.005; Table 1). In patients with mutations, the most frequently missing teeth were mandibular second premolars (12.4%), followed by maxillary lateral incisors (11.5%) mandibular first incisors (10.8%) and maxillary second premolars (10.2%). In non-mutation carriers, maxillary lateral incisors (17.9) were the most frequently missing teeth, followed by mandibular second premolars (16.3%), maxillary second premolars (15.4) and mandibular first incisors (9.8%). Analysis of detailed diagnostic information from medical records showed that patients with WNT10A mutations did not display any dysmorphic features or evident abnormalities of nails, skin, hair or sweat glands.


One novel mutation was identified in the highly conserved paired box sequence of the PAX9 gene (exon 2). This four nucleotide heterozygous insertion, c.353insTGCC, replaces serine at position 119 by alanine, causing a frameshift and premature termination of translation by 25 amino acids (Ser119AlafsX199). The novel PAX9 mutation was not identified in 280 healthy controls. The affected proband lacked nine permanent teeth, including maxillary second premolars and first molars, maxillary left second molar and mandibular central incisors and second molars (Table 4). No abnormalities of nails, skin, hair or sweat glands were observed. The proband had a positive family history of tooth agenesis.


The Wnt/β-catenin signalling pathway is involved in various aspects of embryonic development, adult tissue homeostasis and stem cell renewal [18, 25]. Abnormal Wnt signalling has been linked to many human diseases, including oral congenital anomalies and tumours derived from oral tissues [14]. Several studies have suggested that WNT genes might be candidate genes for non-syndromic cleft lip and/or palate (NSCL/P), which is the most common craniofacial anomaly and is often associated with tooth agenesis [26-28]. Nucleotide variants in the AXIN2 gene have been associated with tooth agenesis-colorectal cancer syndrome and both familial and sporadic forms of tooth agenesis [10, 29, 30]. Mutations in the WNT10A gene are a frequent cause of a broad spectrum of ectodermal dysplasias, including odonto-onycho-dermal dysplasia (OODD) and Schöpf-Schults-Passarge syndrome [31]. In addition, pathogenic variants in WNT10A have recently been identified in patients with non-syndromic tooth agenesis [12, 13]. Therefore, the purpose of the current research project was to determine the contribution of nucleotide variants in 13 WNT and WNT-related genes to the risk of non-syndromic hypodontia and oligodontia in the Polish population.

The most important finding was the confirmation that WNT10A is a key gene in the aetiology of non-syndromic tooth agenesis. We have identified pathogenic mutations located in the highly conserved coding sequence of WNT10A, including three previously undescribed variants, in 26 out of 42 (62%) patients tested. The most frequently observed mutations were p.Cys107Ter and p.Phe228Ile. In probands homozygous for the p.Cys107Ter variant almost all permanent teeth were missing. Of interest, residues Cys107 and Phe228 represent true mutational hot spots and are the most commonly affected amino acids in patients with OODD [32]. Analysis of the control group revealed that the potentially damaging p.Phe228Ile variant was present also in healthy individuals with an allele frequency of 1%. Nevertheless, this mutation was significantly associated with an over-ninefold increase in the risk of non-syndromic tooth agenesis. It has been estimated that approximately half of all individuals who are heterozygous for p.Phe228Ile missense mutation manifest some form of ectodermal defects [31]. This may indicate that in our control group composed of 430 healthy subjects where 6 individuals displaying some clinical anomalies affecting hair, teeth, nails or sweat glands are not recognized during clinical examinations.

Comparison of dental diagrams of patients with and without WNT10A mutations has revealed statistically significant differences in the distribution of congenitally missing permanent teeth. However, because of the relatively small number of patients in our study, this observation does not allow us to define the characteristic tooth phenotype associated with pathogenic variants of the WNT10A gene. Therefore, additional studies involving a large number of patients with both hypodontia and oligodontia should be undertaken to establish WNT10A genotype–phenotype relationships, as was performed in the case of the MSX1 and PAX9 genes [33].

During mouse odontogenesis, Wnt10a is specifically expressed in the primary and secondary enamel knots, which are epithelial signalling centers playing a pivotal role regulating different steps of tooth development [34, 35]. The expression of Wnt10a is also associated with odontoblast differentiation [35]. Yamashiro et al. have showed that, in secretory odontoblasts, Wnt10a is an upstream regulatory molecule for dentin sialophosphoprotein, which is an important odontoblast differentiation marker necessary for dentin mineralization [35]. The Wnt10a expression patterns suggest that the encoded signalling molecule may also have functions in other tissues of the embryo and adult [36, 37]. In embryonic tissues, strong expression of this gene is observed in the face, limbs, skin and hair follicle placodes, whereas in adult tissues it is expressed at highest levels in the brain including the pituitary gland [36]. Analysis of human phenotypes associated with the studied WNT10A mutations confirms the importance of this gene in normal formation and regeneration of teeth, hair, skin and nails [38, 39].

To date, the four genes MSX1, PAX9, EDA and AXIN2 were considered the main candidate genes for tooth genesis [33]. However, their mutations and polymorphisms could explain only a fraction of the genetic component contributing to this dental anomaly. Bergendal et al. have showed that mutations of MSX1, PAX9, EDA and AXIN2 together with mutations of the EDAR and EDARADD genes were present in approximately 11% of the probands in a population with severe isolated hypodontia [40]. In this study, mutational analysis of the coding regions of the MSX1 and PAX9 genes revealed the pathogenic variant only in one patient with oligodontia. This novel heterozygous insertion, c.353insTGCC, resulting in a frameshift and premature termination of translation was identified in the highly conserved paired box sequence of the PAX9 gene. The phenotype of the patient confirmed the consideration that PAX9 mutations are dominantly associated with molar agenesis [41].

Another finding of our study was the association of the CTNNB1, DVL2 and WNT11 polymorphisms with the risk of non-syndromic tooth agenesis. These results must be interpreted with caution, because after applying the Bonferroni adjustment for multiple testing none of these associations remained statistically significant. On the other hand, however, the Bonferroni correction is the most conservative approach to control false positives that may lead to the underestimation of weak and moderate genetic effects. In the case of DVL2, the positive results of single-SNP tests were confirmed by haplotype analysis, which provided even stronger evidence for a causative role of this gene in the aetiology of non-syndromic tooth agenesis. The Dvl2 protein is involved in both the canonical Wnt signalling pathway and in the planar cell polarity pathway [42]. The role of Dvl2 in the canonical Wnt signalling depends on its ability to form a platform for dynamic recruitment of Axin and other Wnt partners [43]. It is interesting to note that in our previous studies we have showed that the AXIN2 nucleotide variants are risk factors for both non-syndromic tooth agenesis and orofacial clefts in the Polish population [29, 44]. Moreover, we have shown for the first time that polymorphisms of DVL2 are significantly associated with the susceptibility to NSCL/P [44]. All these results may confirm the assumption that tooth agenesis can be considered as an extended phenotype for orofacial clefts and that analysis of the aetiology of NSCL/P might be a good candidate-gene strategy for dental anomalies [45].

Haplotype analysis has showed that 2-SNP and 3-SNP haplotypes composed of APC polymorphisms may also be correlated with the risk of non-syndromic tooth agenesis. The multifunctional scaffolding protein encoded by this gene is an essential interacting component of the β-catenin destruction complex [18]. Inactivating mutations in APC are primarily responsible for familial adenomatous polyposis (FAP) and sporadic colorectal cancers (CRC) [46]. Interestingly, 10–20% of patients with FAP, resulting from loss-of-function APC mutations, exhibit oral and maxillofacial symptoms including an increased risk of jaw osteomas, odontomas, and supernumerary or unerupted teeth [47]. By contrast, nonsense or frameshift loss-of-function mutations in AXIN2 are associated with dominant inherited oligodontia and colorectal neoplasia [10]. In mouse models, Dvl2 has also been implicated in the progression of this cancer [48]. These observations, together with the results of our APC, AXIN2 and DVL2 association studies, may confirm the hypothesis that a diagnosis of congenital tooth agenesis or supernumerary teeth may be considered a possible indicator of CRC susceptibility.

In summary, in this study we successfully genotyped 34 SNPs from 13 WNT and WNT-related genes in patients with non-syndromic tooth agenesis and properly matched the control group, and we analysed the coding sequence of MSX1, PAX9 and WNT10A in 42 patients. We confirmed previous findings that WNT10A is a key susceptibility gene for this common dental anomaly, and found evidence for a borderline association between polymorphic variants of APC, CTNNB1, DVL2 and WNT11 and the risk of both hypodontia and oligodontia in the Polish population. To our knowledge, this is the first association study to have examined the potential role of nucleotide variants of WNT and WNT-related genes in the aetiology of non-syndromic tooth agenesis. Our findings are of interest especially in light of observations showing that misregulation of the Wnt/β-catenin signalling pathway is associated with many human tumours, including CRC or those cancers derived from oral tissues [14]. However, several limitations need to be noted regarding this study, including the lack of segregation analysis of the identified WNT10A and PAX9 mutations with the disease phenotype in the families, the small number of analyzed patients and controls, the number of selected SNPs being too small to achieve full gene coverage, and the selection of control group participants without a detailed analysis of skin, hair, nail or sweat gland anomalies. Moreover, most of the polymorphisms tested in our study are not functional variants, and their association with the congenital lack of permanent teeth may be due to linkage disequilibrium with one or more aetiological mutations or polymorphisms of these genes or another gene/genes located in close proximity. Therefore, further studies should focus on (i) confirmation of our results in different populations, (ii) identification of functional variants responsible for the identified association signals and (iii) searching for the relationship between the Wnt signalling pathway and increased risk of CRC in patients with non-syndromic tooth agenesis.


The technical assistance of MSc Sylwia Matuszewska and MSc Joanna Jurkowlaniec-Salazar as well as Ms Longina Nowak is gratefully acknowledged. This work was supported by grant no. NN403 539240 from the Polish Ministry of Science and Higher Education.