“Isolated” Amelogenesis Imperfecta Associated with DLX3 Mutation: A Clinical Case

Amelogenesis imperfecta (AI) represents rare tooth anomalies that affect the quality and/or quantity of the enamel. Clinical phenotypes display a wide spectrum, ranging from mild color changes to severe structural alterations with daily pain. However, all affect the quality of life because of mechanical, psychological, esthetic, and/or social repercussions. Several gene mutations have been linked to AI as a nonsyndromic (isolated) phenotype or a wider syndrome. This case report aimed to present a family with dental structure anomalies followed up in the dental department of the Louis Mourier Hospital (APHP, France) for their extremely poor dental condition. The proband and his mother were clinically diagnosed with AI, and genetic analysis revealed an already described variant in DLX3. Then, the family was further examined for tricho-dento-osseous syndrome. This report illustrates the challenge of diagnosing dental structure anomalies, specifically AI, in adults and highlights the need for an accurate and accessible molecular diagnosis for those anomalies to discriminate between isolated and syndromic pathologies.


Introduction
Amelogenesis imperfecta (AI) is a rare dental disease affecting the enamel structure, with an estimated prevalence of 1/700 to 1/14000 (ORPHA disorder 88661), a large range due to a mis-or underevaluation of tooth structure anomalies and lack of large clinical studies on this topic. AI has been described for years following the Witkop classification [1], aiming to specify the different aspects, forms, and colors of enamel anomalies. e bedrock of this classification is the AI "macro" pathophysiology, namely, impairment of enamel secretion or maturation stages during amelogenesis. However, presently, it is recognized that (1) clinical evaluation of the dysmorphic enamel is biased as soon as a tooth erupts, due to attrition, erosion, and other intraoral phenomena and (2) phenotypes are heterogeneous and can overlap [2,3].
Another crucial aspect when considering AI is that this anomaly can be inherited alone or with other anomalies within a wider syndrome, such as enamel renal syndrome associated with FAM20A mutations [4] and AI associated with vitamin D-resistant rickets linked to VDR or CYP27BA mutations [5]. Since 1998, DLX3 mutations have been associated with tricho-dento-osseous syndrome (TDO, OMIM#190320) [6,7]. TDO is described as a dominant inherited syndrome combining hypoplastic AI with taurodontism, head and neck skeletal anomalies (increased thickness, higher bone density, and obliteration of facial sinuses), and curly or kinky hair. Nail and skin anomalies are often associated with TDO [8][9][10]. Other dental defects, such as pulp obliteration, shorter and dysmorphic roots, and dental eruption disorders (early or delayed), have sometimes been reported [11]. Recent publications have highlighted the possibility of "attenuated" TDO [12][13][14], relaunching the debate on a common or distinct diagnosis between TDO and the hypoplastic-hypomature type with the taurodontism form of AI (OMIM #104510), both linked to DLX3 mutations [15].
In direct line with this debate, here, we describe the unexpected diagnosis of the known DLX3 c.398G > C, p. Arg133 Pro, variant in a family followed for global dental rehabilitation by the dental department of Louis Mourier Hospital (Paris, France).

Case Presentation
A 10-year-old male African child presented at the emergency department of Louis Mourier Hospital with cellulitis related to necrosis of the maxillary first right permanent incisor. After 1 year, he returned for general consultation. He had no history of medical conditions, and no hair, skin, or nail defects had been noted by the clinicians or reported by the parents. Oral examination revealed a healthy mucosa without bone abnormalities. However, the teeth showed significant structural and eruption abnormalities. e enamel was dark yellow, streaked, and severely worn (Figure 1(a)). It appeared translucent in the areas of wear. e enamel radio-opacity was similar to that of dentin ( Figure 1(b)). On orthopantomography (Figure 1(b)), taurodontism was observed on the first permanent molars. e tooth eruption was severely disturbed. e first permanent molars and mandibulary first left deciduous molar were impacted. Roots appeared shorter than normal. After several dental infections, dental extractions and complete denture removal were performed (Figure 1(c)). e proband's mother had a reported diagnosis of "dentinogenesis imperfecta" associated with painless mandibular bone exostosis from a previous visit. Clinically, the enamel was orange-brown and appeared severely worn out (Figure 2(a)). Panoramic radiography confirmed multifocal areas of high density in the mandibular bone ( Figure 2 is clinical picture suggested osseous dysplasia. More specifically, the multifocal fibro-osseous dysplastic areas characterized by radio-opaque cementumlike masses allowed the clinician to make a diagnosis of florid cemento-osseous dysplasia [16]. Dental examination revealed teeth with thin or absent enamel, numerous endodontic treatments and restorations, and missing teeth (Figure 2(b)). Whole-body scintigraphy performed to further explore the bone status revealed only one point of hyperossification on the right ankle. Radiographic examination of the left hand performed after a single painful episode revealed only a faint modification of the fifth phalanx. Furthermore, there were no hair, skin, or nail defects, so no formal diagnosis of TDO was established [16]. Residual teeth (except for two mandibular molars, left and right) were extracted, and rehabilitation was performed using a complete maxillary and partial mandibulary removable denture. ese oral conditions were not reported for the patient's brothers, aged 2 and 5 years (considering the limitation of the clinical examination regarding their ages), or father (Figure 3(a)).
A clinical diagnosis of AI, hypoplastic/hypocalcified, was established for the proband according to the criteria defined by Witkop [17]. Because of the proband's teeth features, along with the mother's contradictory former diagnosis of dentinogenesis imperfecta, genetic exploration was indicated.

Genetic Material and
Methods. DNA from patients I.1 and II.1 were analyzed at the Department of Molecular Biology and Genetics of Cochin Hospital (APHP, Paris) ( Figure 3(a)). High-throughput targeted sequencing was performed on an ion PGM TM system from amplicon libraries ( ermo Fisher Scientific) ( Table 1). Bioinformatics analysis relied on ermo Fisher Scientific tools and the homemade pipeline Polydiag of the Paris Descartes University, Imagine Institute bioinformatics platform. A pathogenic variant in DLX3, exon 2, c.398G > C, p. Arg133Pro, was identified in both patients I.1 and II.1, but not in I.2 (DNAs for II.2 and II.3 were not available) (Figure 3(b)), neither in a positive control cohort of patients with AI nor in the database of the Department of Molecular Biology and Genetics [15]. e substitution c.398G> C was reported by Nieminen et al. in 2011 but, then, associated with a typical TDO family. It encodes for a missense at the protein level, replacing the arginine in position 133 with a proline at the beginning of the homeodomain of the DLX3 protein. Predictors related to consensus sequence conservation interspecies (Polyphen2) and 3D structure/missense disruption (SIFT, Mutation Taster) annotate this substitution as deleterious or pathogenic. Hence, this variant was considered a class V variant (pathogenic), following the American College of Medical Genetics and Genomics recommendations [18]. Considering all these clinical, radiographic, and molecular elements, a diagnosis of "attenuated" TDO was proposed for the proband and his mother.

Discussion
AI is a highly variable tooth disorder that involves the enamel structure. Its management requires a multidisciplinary team; specialists in medical genetics, mainly in rare dental diseases; and restorative and prosthodontics experts. Presently, the diagnosis of AI still appears difficult for dentists. Clinical descriptions and the classification established by Witkop are complicated and not well applicable because of the great variability of the AI phenotype [17].
Classically, clinicians use hypocalcified, hypomature, or hypoplastic terms to describe AI. Hypoplastic forms are used when the enamel shows reduced thickness. e tooth surfaces can be smooth or rough. e hypoplastic form is mainly associated with an anomaly of enamel matrix secretion [3]. A hypomature enamel is described as opaque, white to yellow-brown, hard, and easily detachable from the dentin and has a normal thickness. ese signs are linked to a lack of protein elimination in the extracellular matrix during the enamel maturation phase. Less mechanical strength is observed. e hypocalcified form is characterized by impaired mineralization of enamel crystallites during the secretion phase. It results in a creamy-white to yellow-brown rough enamel surface. e enamel is generally normal in thickness on newly erupted teeth but rapidly tends to be chipped away or scraped from the dentin. Radiographically, enamel radio-opacity appears similar to that of the dentin [17].
Over the years, all associations of clinical AI forms have been reported. Recent studies have highlighted the cellular and molecular levels of pathophysiology knowledge, and the common and global term of AI is proposed for all clinical forms [19,20]. e modern approach to AI tends to regroup causative proteins depending on their function and pathway. Different protein groups can be mentioned, even though all the AI causative genes have not yet been discovered. e first described were the enamel matrix proteins (AMELX, ENAM), enamel matrix proteases (MMP20, KLK4), and proteins involved in cell-cell and cell-matrix adhesion (AMTN, COL17A1, LAMA3, LAMB3, ITGB6, and FAM83H), transport (WDR72, SLC24A4), and control (even if not yet well defined) of amelogenesis (GPR68, ODAPH, ACPT, FAM20A, and DLX3) [3]. erefore, this new AI approach is no longer only clinical but also biological. is requires a strong network between geneticists and specialists of rare dental diseases to improve the correlation between genotype and phenotype and close collaboration with restorative treatment experts.
is network could enhance treatment options, notably an individual therapeutic approach, such as an adaptive adhesive technique, in case of MMP20 mutation [21] or new protocols to systematically and preventively recover all cuspid teeth with LAMA3 or FAM83H mutation because of the rapid and widespread loss of enamel associated with malfunction of these proteins.
Another important aspect is to consider the isolated character or not of AI. In the described family, dental problems, even though the first proposed diagnosis based on the clinical observation of the proband's mother was inaccurate, were very well known by all family members. e proband had been supported and accompanied by his parents since his early childhood, and dental care, even if not fully appropriate, had been provided. However, a complete and definite diagnosis without a molecular approach remains difficult. When dental structure anomalies are observed for the first time in adulthood, diagnosis is almost  impossible because of posteruptive modifications and sequelae. e second aspect agrees with recent studies on AI topics, highlighting that the frontier between syndromic and nonsyndromic forms of AI is not so watertight. In this report, except for the osseous mandibular densities in the proband's mother (which, when observed in middle-aged African women, could be diagnosed as florid cemento-osseous dysplasia), only remarkably faint signs usually linked to TDO had been noted by the family or pediatricians or were even noticeable by dental surgeon specialists of rare diseases. When the DLX3 mutation diagnosis was made, the whole family had received a complete explanation on TDO and had been advised to undergo extensive testing for DLX3 mutation holders and their offspring. Specifically, the clinical diagnosis of patient II.2 should be reassessed over time. Indeed, when comparing our family with the Finnish family first described as carrying the c.398G > C frameshift [11], the clinical features of the Finnish patients were stronger for a TDO diagnosis (association of dental, facial bone, and hair phenotypes). However, one cannot ignore the mandibular ossification in patient I.1 and the severely disturbed eruption associated with shortened roots and taurodontism in patient II.2 that resembled those of Finnish patients and supported an "attenuated" TDO diagnosis.

Conclusions
Our report illustrates the challenge of diagnosing dental structure anomalies and, specifically, AI in adults and highlights the need for an accurate and accessible molecular diagnosis for these anomalies to (1) ameliorate the patient's course and provide appropriate care and (2) provide a better understanding of the underlying pathophysiology. It is essential to obtain this knowledge to assure the patient that he or she will not pass on a syndromic trait to the offspring. In France, this process also helps patients receive financial support for dental or medical care. Alongside the molecular aspect, self-esteem, wellness, and social aspects of the pathology should be better considered.

Consent
Written informed consent was obtained from the patient's parents for publication of this case report, and any potentially identifying information was removed. Gene sequencing was performed after written informed consent was provided by the patient's parents for themselves and their child.

Disclosure
Anne-Margaux Collignon and Celine Gaucher are co-last authors.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this article.

Authors' Contributions
KS, AMC, and AW were in charge of the clinical follow-up of the patient. ALB, AMC, and CG drafted the manuscript. CS and CG performed gene sequencing and molecular analysis. AMC and CG reviewed the manuscript and made substantial contributions to data interpretation. All authors approved the final version of the manuscript. Anne-Margaux Collignon and Céline Gaucher contributed equally to the work.