Identification of DSPP novel variants and phenotype analysis in dentinogenesis dysplasia Shields type II patients

To investigate the genetic causes and teeth characteristics of dentin dysplasia Shields type II(DD-II) in three Chinese families. Data from three Chinese families affected with DD-II were collected. Whole-exome sequencing (WES) and whole-genome sequencing (WGS) were conducted to screen for variations, and Sanger sequencing was used to verify mutation sites. The physical and chemical characteristics of the affected teeth including tooth structure, hardness, mineral content, and ultrastructure were investigated. A novel frameshift deletion mutation c.1871_1874del(p.Ser624fs) in DSPP was found in families A and B, while no pathogenic mutation was found in family C. The affected teeth’s pulp cavities were obliterated, and the root canals were smaller than normal teeth and irregularly distributed comprising a network. The patients’ teeth also had reduced dentin hardness and highly irregular dentinal tubules. The Mg content of the teeth was significantly lower than that of the controls, but the Na content was obviously higher than that of the controls. A novel frameshift deletion mutation, c.1871_1874del (p.Ser624fs), in the DPP region of the DSPP gene causes DD-II. The DD-II teeth demonstrated compromised mechanical properties and changed ultrastructure, suggesting an impaired function of DPP. Our findings expand the mutational spectrum of the DSPP gene and strengthen the understanding of clinical phenotypes related to the frameshift deletion in the DPP region of the DSPP gene. A DSPP mutation can alter the characteristics of the affected teeth, including tooth structure, hardness, mineral content, and ultrastructure.


Introduction
Hereditary dentin defects (HDD) show autosomal dominant transmission patterns and are characterized by abnormal dentin development. The affected teeth are typically grey or yellow-brown in color, and the pulp chamber is either obliterated early or experiences delayed development. The enamel is easy to exfoliate, leading to severe attrition of the dentin. Due to the abnormal development of dentin, caries is more likely to occur than in normal teeth, which seriously affects the patient's aesthetics and chewing function. Shields et al. [1] classified hereditary dentin dysplasia into dentinogenesis imperfecta (DGI) and dentin dysplasia (DD) in 1973 based on clinical and radiographic features. Among them, DGI is divided into three types, and DD is classified into two types.
The dentin sialophosphoprotein (DSPP) gene plays an important role in dentin development and mineralization. Studies have found that the DSPP gene is the cause of DGI, Shields type II(DGI-II) (OMIM 125490) [2,3], DGI, Shields type III(DGI-III) (OMIM 125500) [4], DD, Shields type II (DD-II) (OMIM 125420) [5], and deafness with dentinogenesis (OMIM 605594) [2]. In DGI-II patients, both primary and permanent teeth are typically amber and translucent, showing significant attrition. A bulbous crown with marked cervical constriction is a typical feature. Radiographically, the teeth show short, constricted roots and premature pulpal obliteration. Sensorineural hearing loss has been reported as a rare feature. In DGI-III patients, besides the symptoms similar to DGI-II, the primary teeth often exhibit early pulp exposure. Radiographically, the pulp cavity and root canal of the affected tooth are wide, while the dentin is thin and shell-shaped. In DD-II patients, the features in the primary dentition resemble those observed in DGI-II; however, the permanent dentition is either unaffected or shows mild radiographic abnormalities such as thistle-tube deformity of the pulp chamber or partial obliteration [2]. DGI-II, DGI-III, and DD-II are now considered as a spectrum of diseases [6]. Among them, DGI-III has the most severe clinical manifestations, followed by DGI-II, and DD-II is the mildest.
DSPP (NM_014208) is located on chromosome 4q21.3 and includes 5 exons. Exons 1-4 and the 5' end of exon 5 encode dentin sialoprotein (DSP), and the 3' end of exon 5 encodes dentin phosphoprotein (DPP) (Fig. 1) [7]. Until now, more than 60 heterozygous mutations have been identified in DSPP, and these mutations are mainly found in exons 2, 3, 4, and 5 ( Fig. 1) [8]. Previous studies [4,[8][9][10] indicated that mutation types in the DSP region included missense mutations, nonsense mutations, and splicing mutations, and the mutations in the DSP region mostly correlated with DGI-II and DGI-III. Mutations in the DPP region are mainly frameshift mutations; mutations near the N-terminal of DPP were associated with DD-II, while mutations in the C-terminal of DPP correlated with DGI-II and DGI-III [4,10] (Fig. 1, Table S1). However, the underlying reason for the correlation between the mutation locations and the phenotypes remains to be elucidated.
In this study, three Chinese families with the DD-II phenotype of DD-II were recruited. By thorough sequence analysis, a novel frameshift deletion mutation c.1871_1874del(p. Ser624fs) in the DPP region of the DSPP gene was identified in families A and B. This mutation impairs the function of DPP. The primary teeth of the patients were collected, and the physical and chemical properties of the teeth were examined. The study provides a detailed understanding of the genotype-phenotype correlation in DD-II affected by a frameshift deletion in the DPP region of the DSPP gene.

Participants and clinical assessments
All procedures involving human participants in this study were performed in accordance with the Declaration of Helsinki. Written informed consent was obtained from each participant. This study was approved by the Medical Ethics Committee of the Sichuan Provincial People's Hospital [No. 2020(02)].
Three Chinese families with phenotype of DD-II were recruited for this study. The probands and their relatives were recruited, and the detailed information of the participants was listed in Table S2. No affected individuals had symptoms of hearing loss or bone defects.

DNA extraction
Blood samples were collected from all participants, then genomic DNA was isolated from peripheral blood using TIAN-GEN Blood DNA Kit (TIANGEN, Beijing, China), according to standard procedures. Nanodrop was used to detect the concentration of extracted DNA. The DNA samples were stored at −20 °C until use, and DNA integrity was assessed by 1% agarose gel electrophoresis as previously described [11].

Next-generation sequencing for mutation analysis
We performed WES on the proband's genomic DNA from three families. A total of 200 ng genomic DNA of each participant was sheared by Biorupter (Diagenode, Belgium) to acquire 150~200 bp fragments. The ends of the DNA fragment were repaired, and Illumina Adaptor was added (Fast Library Prep Kit, iGeneTech, Beijing, China). After sequencing, the library was constructed, and the whole exons were captured with AIExome Enrichment Kit V1 (iGeneTech, Beijing, China) and sequenced on the Illumina platform (Illumina, San Diego, CA, USA) with 150 base paired-end reads. The mean depth was 302×, with > 99.27% bases covered at 10×, with > 99.27% bases covered at 10×. Due to no pathogenic variants shown by WES, we additionally performed WGS on the proband's genomic DNA from family C to identify the genetic cause. A total of 200 ng genomic DNA of each individual was used to construct a sequencing library with Enzyme Plus Library Prep Kit (iGeneTech, Beijing, China) according to the instructions. The WGS was performed on a DNB-SEQ-T7 sequencing instrument with 150 base paired-end reads. The mean depth was 35×, with > 91.78% bases covered at 10×. Raw reads were filtered to remove low quality reads by using FastQC. Then, clean reads were mapped to the reference genome GRCh37 by using Burrows-Wheeler Aligner (BWA). Variant calling was performed using GATK with Haplotype Caller, then SNV and InDel were annotated by dbSNP138 (https:// www. ncbi. nlm. nih. gov/ proje cts/ SNP/), 1000 Genomes Project (ftp:// ftp. 1000g enomes. ebi. ac. uk/ vol1/ ftp), ClinVar (https:// www. ncbi. nlm. nih. gov/ clinv ar) and Gnomad (https:// gnomad. broad insti tute. org/) and in-house database. Common variants (minor allele frequency, > 0.1%) were excluded. Variants were classified in accordance with the interpretation guidelines of the American College of Medical Genetics and Genomics (ACMG), as potentially pathogenic variants, variants of unknown clinical significance, or benign variants. Deleterious variants of unknown clinical significance were further classified as associated or not associated to the proband phenotype. In this study, we focused on disease-causing genes for HDD, osteogenesis imperfecta, and amelogenesis imperfecta in the Online Mendelian Inheritance in Man (OMIM) database (https:// omim. org/) as described before [11,12].

Sanger sequencing
Candidate pathogenic variants were first validated using sanger sequencing. Specific primers were designed to amplify the corresponding regions. PCR products were sequenced using an ABI 3730XL Genetic Analyzer (Applied Biosystems, Foster City, CA, USA) according to manuals for the BigDyeTM Terminator v3.1 Cycle Sequencing Kits. The primers are listed in Table S3.

Micro-CT
Two primary molars extracted due to severe apical infection from the probands of families A and B, respectively, were collected for Micro-CT. Teeth were scanned using a micro-CT system (μCT 100; SCANCO Medical AG, Bruttisellen, Switzerland) with an isotropic resolution of 22.5 μm at 70 kV/200 uA with an exposure time of 400 ms.

Microhardness
A total of four primary teeth from the patients were collected for the microhardness test: one primary anterior tooth and one primary molar from family A, one primary molar from family B, and one primary anterior tooth of from the proband of a DGI-II family we have reported [11]. Four primary teeth from age-matched normal individuals were used as control. The periodontal membrane and other debris were removed, and the tooth samples were immersed in a 10% formaldehyde solution at 4 °C and crosscut into 2 parts by dental diamond burs and inlaid with epoxy resin. Teeth were polished with 600, 1000, 2000, 3000, 5000, 7000, and 10,000 mesh water sandpaper (Starcke, Melle, Germany) in sequence and ultrasonically cleaned with distilled water. Three points of the enamel and five points from the dentin were selected from each tooth, and the measurement of the microhardness was performed three times and averaged at each point. Vickers microhardness was measured by using a microhardness tester (QNESS Q10A+, Austria), for which a 0.2 kgf load was applied for 12 s to obtain the measurement.

Scanning electron microscopy (SEM) investigation and energy dispersive spectroscopy (EDS) investigation
After polishing, teeth were vacuum dried, the surfaces were coated with gold in a vacuum, and the cross-section surfaces were observed using SEM as previously described [11] (Inspect F50, FEI Co., Hillsboro, O, USA). EDS results were analyzed using an SEM (FEI Co.) equipped with an EDS INCA system (Oxford Instruments Analytical, Abingdon, UK). The EDS was conducted at a high accelerating voltage of 20 kV. Three points of the enamel and five points from the dentin were selected from each tooth, which is the same as for microhardness (Fig. 4A). Elemental measurements were repeated 3 times at each point, and the Ca/P ratio was calculated.

Statistical analysis
All data were analyzed by SPSS 22.00. All data were expressed as mean ± SD. The student's t-test was used to compare the mean ± SD between two groups, and the oneway ANOVA analysis was used to compare the mean ± SD of multiple groups. 0.05 was seen to have a significant difference.

Clinical and radiographic features
Family A was a 3-generation Chinese family; a total of four family members were collected including two affected individuals (II:1, III:1) and two unaffected members (I:2, II:2) ( Fig. 2A). The proband (III:1) was a 4-year-old boy with a typical yellow-brownish translucent appearance of all teeth (Fig. 2F). The panoramic radiograph also showed a typical feature of obliterated pulp cavities (Fig. 2G). The mother of the proband (II:1) showed normal permanent teeth in both color and shape (Fig. 2H). However, the panoramic radiograph also revealed thistle-shaped pulp cavities (Fig. 2I).
Family B was a 3-generation Chinese family; three family members were collected including two affected ones (II:3, III:2) and one normal phenotypic individual (II:4), and the mother of the proband denied family history (Fig. 2C). The proband (III:2) was a 7-year-old boy with a typical yellowbrownish translucent appearance of primary teeth, but all the permanent teeth of the proband were normal both in color and shape (Fig. 2J). The periapical radiographs revealed obliterated pulp cavities in the primary teeth (Fig. 2K). The mother of the proband (II:3) showed normal permanent teeth in both color and shape (Fig. 2L). But the intraoral panoramic radiograph showed thistle-shaped pulp cavities in the permanent teeth of the entire mouth (Fig. 2M). Families A and B were from different provinces of China, and both of them denied knowing each other.
Family C was a 4-generation Chinese family, and four family members were collected including three affected ones (III:23, IV:16, IV:17) and one unaffected member (III:24) (Fig. 2E). Both the proband (IV:16) and her younger brother (IV:17) showed the typical yellow-brownish translucent appearance of primary teeth, and the newly erupted permanent teeth of the proband were normal both in color and shape (Fig. 2N). The panoramic radiograph of both (IV: 16) and (IV:17) revealed obliterated pulp cavities in the primary teeth (Fig. 2O). The father of the proband (III:23) had relatively normal permanent teeth, but the canines showed a more brownish color than normal teeth (Fig. 2P). The panoramic radiograph also revealed thistle-shaped pulp cavities, especially for the anterior teeth (Fig. 2Q).
According to the clinical manifestations, families A, B, and C were diagnosed as dentin DD-II.

Genetic findings
A novel frameshift deletion mutation c.1871_1874del(p. Ser624fs) on exon 5 of the DSPP gene was identified by WES in the probands in both families A and B. The affected individuals were heterozygous for the variant, and the variant was not found in the unaffected family members. The variant was completely co-segregated with the phenotype in the family (Fig. 2B, D). The novel mutation was classified as a pathogenic variant according to the ACMG interpretation guidelines [12]. This variant caused a 4-bp deletion near the N-terminus and a frameshift mutation starting from the serine position on codon position 624.
No pathogenic variant in the DSPP or other possible genes related to HDD, osteogenesis imperfecta, and amelogenesis imperfecta was found by WES and WGS in family C. Four polymorphisms were found in DSPP, and no polymorphism was found in other possible genes related to HDD, osteogenesis imperfecta, and amelogenesis imperfecta (Table S4).

Micro-CT
Micro-CT results revealed that the teeth suffered from severe root obliteration, and root canals were smaller than normal teeth and irregularly distributed comprising of a network as reported before [13]. The macro photo of the tooth also showed that the surface of the affected teeth is milky white, the dentin is a typical yellow-brownish, and translucent, most of the enamel still exists, and the root canal is very thin and almost invisible (Fig. 3).

SEM and EDS results
The SEM showed that the dentinal tubules of the patient teeth were highly irregular in both size and shape and reduced in number (Fig. 4D, E), while the normal dentinal tubules were highly regular in shape and size (Fig. 4F, G). The EDS results showed that the Mg content of the DD dentin was significantly lower than that of the normal dentin (0.36 ± 0.16 vs. 0.79 ± 0.11, P = 0.000). The Na content of the DD dentin Fig. 3 The images of the DD-II teeth. A-C Micro-CT of the primary molar of family A, III:1. The red portion showed that the root canals were smaller than normal teeth and irregularly distributed. D The macro photo of the tooth showed obliterated root canals. E-G Micro-CT of the primary molar of family B, III:2. The red portion showed that the root canals were also smaller than normal teeth and irregularly distributed. H The Macro photo of the tooth showed obliterated root canals was significantly higher than that of the normal dentin (0.64 ± 0.09 vs. 0.50 ± 0.07, P = 0.007). However, there were no significant differences in other elements of the DD teeth and controls (Table S7, Fig. 4H). The EDS result of the enamel also showed that the mineral content of the DD-II's enamel was similar to controls (Table S8, Fig. 4I).

Discussion
DSPP belongs to the small integrin-binding ligand N-linked glycoprotein (SIBLING) family of secreted phosphoproteins, which are involved in bone mineralization. Full-length nascent DSPP is cleaved to yield dentin sialoprotein (DSP) and dentin phosphoprotein (DPP) by bone morphogenetic protein 1(BMP1) [14]. DSP and DPP, two principal noncollagenous proteins of mature dentin matrices, play essential but distinct roles in dentinogenesis. DSP regulates the initiation of dentin mineralization, and DPP is involved in the maturation of mineralized dentin [15,16].
Veis and Perry reported DPP in 1967 for the first time [17]. DPP, a highly acidic protein, is the major noncollagenous protein in the dentin matrix. It is synthesized and secreted by mature odontoblasts [18][19][20] and then is transported to the mineralization front, where it binds calcium ions with high affinity [21,22], aggregates collagen fibrils [23], and assumes a structure promoting the formation of initial apatite crystals. As predentin is converted to dentin in the mineralization process, the mineral crystals grow in an ordered direction. DPP and other proteins bind to the growing hydroxyapatite faces to modulate crystal growth [24].
In this study, three DD-II families in total were recruited. In families A and B, through the WES sequencing method, a novel frameshift deletion mutation of DSPP, c.1871-1874del was identified, which shows complete cosegregation with the disease phenotype in both families. In family C, no pathogenic mutations were identified through WES and WGS sequencing methods. However, four polymorphisms were found in exon 5 in DSPP gene, which encodes the DPP protein. The DPP region contains highly repetitive DS and DSS motifs; indels could easily occur in this region, and previous studies have shown that indels may have no apparent effect on protein function and accumulate in the population [4,25,26]. Also, the frequencies of the four polymorphisms of DPP among the normal population are high (Table S4). So, we believe that the polymorphisms in family C had no functional effects. Previous studies have shown that there are no pathogenic mutations in DSPP or Col1A2 in some patients diagnosed with DGI-III or DD-II, which is consistent with our results. Song et al. [10] reported eight families with mutations in the DSP region. They found DPP mutations in five families, while no pathogenic mutations were confirmed in the remaining three families. Wang et al. [27] found a mutation in COL1A2 in a DGI family without bone defects, and they suggested that COL1A1 and COL1A2 should be considered as candidate genes in isolated DGI cases without a DSPP mutation. Li et al. [8] reported seven families diagnosed with DGI/ DD, one of which had no pathogenic variants and only several polymorphisms. We speculated that there were other pathogenic genes for the dentin defects, or polymorphisms may mitigate a defect when they appear together, and copy number variation may lead to the occurrence and development of disease to some degree [8,28].
In previous studies, mutations related to DD-II were found only in the signal peptide region and DPP region [8,9,29], but not in the DSP region, which may be related to the different functions of DSP and DPP. However, the underlying mechanism for the correlation between the mutation locations and the phenotypes is still unclear. In 1973, Shields et al. classified inherited dentin malformations into three types of dentinogenesis imperfecta (DGI) and two types of dentin dysplasia (DD) according to clinical manifestation [1]. Later, studies have shown that DD-II, DGI-II, and DGI-III are all autosomal dominant diseases caused by DSPP gene [2,3], and DGI-II, DGI-III, and DD-II are suggested to be considered as a spectrum of diseases [6]. Recently, Simmer et al. [30] proposed a modified Shields classification. They suggested to classify the disease based on the causative mutation rather than phenotypic severity, and patients identified with 5′-DSPP defects be diagnosed as DGI-III, while those with 3′-DSPP defects be diagnosed as DGI-II. Further research is still required to investigate the genotype-phenotype correlation in HDD.
Micro-and macro-structure photos showed that the primary molars had severe pulp obliteration, and the root canals were fine and reticular, indicating early-stage pulp obliteration. Taleb et al. [31] studied three Danish DGI-II families, whose mutations were located in the DSP region. CT of the primary molars showed an enlarged pulp cavity and root canal, which was different from the primary molars in this study. Park et al. [32] analyzed a premolar with a mutation in the DPP region (c.2688delT) and found that the morphology of the pulp cavity did not change significantly. In the DGI-II family (c.53T > G) studied by Lee et al. (c.53T>A, p.Val18Asp) [33] and in our previous study [11], the teeth were characterized by congenital enamel defects. These studies suggest that the corresponding mutation mechanisms of different DSPP mutation sites are different, and the molecular mechanism remains to be further studied.
The microhardness results revealed that the dentin hardness of DD-II and DGI-II teeth was significantly lower than that of normal teeth. This result is consistent with previous studies. For example, Wieczorek et al. [34] found that the hardness of DI teeth was significantly lower than that of normal teeth. Nutchoey et al. [35] found that the hardness of OI patients' teeth was significantly lower than that of control teeth. The reason may be that DSPP mutation leads to disordered dentin formation, resulting in dysplastic dentin with defects in both structure and hardness. In this study, the enamel hardness of DD-II was not significantly different from that of the normal teeth, indicating that the mutation of c.1871-1874del mutation did not affect the formation of enamel. SEM results showed that the teeth had a reduced number of dentin tubules, and the dentin tubules were irregular. Large gaps and ectopic calcification masses were observed in the dentin. The EDS results of dentin showed that Ca, P, C, and O had no significant change, while Mg decreased significantly, and Na increased significantly. The decrease in Mg is consistent with some previous studies [32,36,37]. Mg 2+ is essential for maintaining the physiological homeostasis of tissues and organs and promotes the proliferation of osteoblasts and osteogenesis [38]. The decrease in Mg in DD-II teeth may be related to the structural disorder and reduced hardness of dentin. The Na content was significantly increased, which is consistent with the research of Park et al. [32] and Sabel et al. [37]. The high Na content may reflect impaired mineralization of DD-II dentine [39]. The EDS results of enamel showed no significant difference compared with controls, which is consistent with the results of a previous study by Intarak et al. on enamel analysis of OI patients [40]. This result also suggests that the mutation of the DPP region does not affect enamel development, and the early enamel exfoliation is the result of a reduction in dentin hardness under long-term chewing pressure [41].
Analysis of tooth hard tissue showed that DD-II teeth had abnormal dentin structure and element content, which often led to early tooth fraction and periapical infections. Abnormal pulp morphology further increases the difficulty of dental treatment [42,43]. Thus, whole-life-cycle dental health management is of great importance to these patients.
In this study, we studied three DD-II families in total and identified a novel DSPP mutation in families A and B. Morphological observation showed that the enamel of DD-II primary teeth was relatively intact, the pulp cavity was partially obliterated, and the root canal was fine and reticular. The microhardness of dentin was significantly lower than that of normal teeth. The element test showed that the content of Mg was significantly decreased, and the content of Na was significantly increased. Our study expands the phenotypic and genotypic spectra of DD-II.