Amelogenesis Imperfecta in Two Families with Defined AMELX Deletions in ARHGAP6

Amelogenesis imperfecta (AI) is a group of inherited conditions featuring isolated enamel malformations. About 5% of AI cases show an X-linked pattern of inheritance, which are caused by mutations in AMELX. In humans there are two, non-allelic amelogenin genes: AMELX (Xp22.3) and AMELY (Yp11.2). About 90% of amelogenin expression is from AMELX, which is nested within intron 1 of the gene encoding Rho GTPase activating protein 6 (ARHGAP6). We recruited two AI families and determined that their disease-causing mutations were partial deletions in ARHGAP6 that completely deleted AMELX. Affected males in both families had a distinctive enamel phenotype resembling “snow-capped” teeth. The 96,240 bp deletion in family 1 was confined to intron 1 of ARHGAP6 (g.302534_398773del96240), but removed alternative ARHGAP6 promoters 1c and 1d. Analyses of developing teeth in mice showed that ARHGAP6 is not expressed from these promoters in ameloblasts. The 52,654 bp deletion in family 2 (g.363924_416577del52654insA) removed ARHGAP6 promoter 1d and exon 2, precluding normal expression of ARHGAP6. The male proband of family 2 had slightly thinner enamel with greater surface roughness, but exhibited the same pattern of enamel malformations characteristic of males in family 1, which themselves showed minor variations in their enamel phenotypes. We conclude that the enamel defects in both families were caused by amelogenin insufficiency, that deletion of AMELX results in males with a characteristic snow-capped enamel phenotype, and failed ARHGAP6 expression did not appreciably alter the severity of enamel defects when AMELX was absent.


Introduction
Non-syndromic amelogenesis imperfecta (AI) is a heterogeneous collection of inherited defects in dental enamel formation that includes a multiplicity of enamel phenotypes, patterns of inheritance, and causative genes [1,2]. Non-syndromic forms of AI are divided into 14 subtypes [3]. To date, defects in 7 genes have been show to cause non-syndromic AI. Autosomal dominant forms can be caused by defects in enamelin (ENAM, 4q21) [4] and family with sequence similarity 83 member H (FAM83H, 8q24.3) [5]. Autosomal recessive AI can be caused by defects in enamelysin (MMP20, 11q22.3-q23) [6], kallikrein-related peptidase 4 (KLK4, 19q13.4) [7], WD repeat containing domain 72 (WDR72, 15q21.3) [8] and chromosome 4 open reading frame 26 (C4orf26, 4q21.1) [9]. X-linked AI is caused by amelogenin defects (AMELX, Xp22.31-p22.1) [10]. Together, defects in these genes account for about half of all AI cases [11,12]. X-linked AI accounts for about 5% of all AI cases [13]. There are two amelogenin genes (AMELX and AMELY), but the gene on the Y-chromosome (Yp11.2) is expressed at low levels [14]. AMELX and AMELY do not undergo homologous recombination, have diverged [15], and because of this divergence, are able to be used in forensics for sex determination [16]. The wide scale application of PCR tests for sex determination provides data on the frequency of AMELY deletions. DNA repeat sequences on the Y chromosome apparently create a structural instability that leads to recurrent 3-4 Mb deletions inclusive of AMELY, PRKY and TBL1Y, and sometimes PCDH11Y, with no apparent negative selection [17]. The frequency of AMELY deletions varies by population and is generally low, perhaps about 0.6% [18]. It is ,0.02% in Australia [19] and ,0.04% in China [20], but higher among particular ethnic groups (,3.6%) in Malaysia and India [21]. AMELY deletions have been reported for ''normal populations'', with the extrapolation that the tooth enamel must therefore be normal in these people. To our knowledge no oral photos or radiographs have been published of person lacking AMELY, although two individuals with AMELY deletions were specifically reported to have normal teeth [22].
Amelogenin (AMEL) belongs to the secretory calcium-binding phosphoprotein (SCPP) family of genes [23]. With the exception of AMEL, all human SCPP genes are clustered on chromosome 4, which include 5 genes encoding acidic proteins and 16 genes encoding proteins enriched proline and glutamine [24]. Amel apparently arose in the proline-glutamine-rich gene cluster and later transposed into intron 1 of Arhgap6 (Rho GTPase activating protein 6) in the opposite orientation, resulting in a nested gene structure. This may have occurred in fish, as Amel is already nested in Arhgap6 in African clawed toads, suggesting that such was the localization of Amel in early tetrapods [25]. In eutherian mammals the pair of homologous autosomes that carried these nested genes (Amel and Arhgap6) translocated to the sex chromosomes [25].
ARHGAP6 is a GTPase-activating protein of the Rho-GAP family. Mice deleted for a 1.1-Mb genomic region spanning from the first exon of Arhgap6 (inclusive of Amelx) to the 59end of Mid1 did not cause any detectable phenotypic or behavioral abnormalities beyond dental enamel defects. The enamel malformations were largely the same as those observed in the Amelx null mice (thin enamel, ,20 mm versus ,115 mm, comprised of parallel-oriented Figure 1. Family 1. A: Pedigree of family 1 consistent with an autosomal dominant or X-linked pattern of inheritance. DNA was obtained from 8 persons, each indicated by a black dot. The arrow identifies the proband. A question mark indicates that the enamel phenotype was unknown. B: The proband's (X*Y; III:4) DNA sequencing chromatogram spanning the deleted segment of ARHGAP6 (g.302534_398773del96240) containing all of AMELX. C: Oral photographs of the proband, a 10 year old Caucasian male in the mixed dentition stage of dental development. Most of the occlusal enamel had abraded from the primary molars, revealing the thinness of retained enamel on the lateral and mesial/distal surfaces. D: Panoramic radiograph showing a thin layer of developing enamel on the unerupted bicuspids and second molars. The enamel is more radioopaque than dentin, but less so than normal enamel. doi:10.1371/journal.pone.0052052.g001 crystallites) with the possible exception of a ''flattened layer'' on the enamel surface [26]. Despite this potential variation of the enamel phenotype in the Arhgap6/Amelx double null mice relative to the Amelx single nulls, two other Arhgap6 knockouts that retained Amelx showed no enamel phenotype. A knockout deleting only exons 6-8 of Arhgap6 resulted in mutant male and female mice ''indistinguishable from their wild-type littermates at birth and thereafter'' [27]. ''Deletion from exon 1a of Arhgap6 to Mid1 led to enamel abnormalities but deletion from exon 6 of Arhgap6 to Mid1 did not'' [26]. These findings suggest that Arhgap6 is not essential for normal enamel formation, but might modify the enamel malformation phenotype when Amelx is absent.
In humans, larger X-chromosome deletions inclusive of AMELX cause syndromes that include AI as a phenotype [28]. Microphthalmia with linear skin defects syndrome (MLS; OMIM 309801) is caused, at least in part, by inactivating mitochondrial holocytochrome synthetase (HCCS) adjacent to ARHGAP6 and AMELX on Xp. To date, 16 AMELX mutations have been reported that cause non-syndromic amelogenesis imperfecta ( Fig. S1). A partial AMELX deletion (g.2525_7247del4723) starting in intron 2 and ending in exon 7 retained only the coding region for the amelogenin signal peptide plus 2 amino acids [29,30]. The enamel phenotype of this family ''B'' [10] or ''pedigree 41'' [13,31] was described as X-linked recessive hypomineralization AI. Affected women had vertically ridged teeth, particularly in the anteriors, which were attributed to alternating bands of normal and hypoplastic enamel deposited by ameloblasts that had randomly inactivated either the normal or defective X-chromosome during development [32]. Affected males had a more severe enamel phenotype.
In this study we report the characterization of two kindreds with partial ARHGAP6 deletions that remove all of AMELX. In family 1 the deletion is confined to intron 1 and is not predicted to alter ARHGAP6 expression. In family 2 the deletion includes exon 2 of ARHGAP6 and is predicted to eliminate normal ARHGAP6 expression. These naturally occurring gene knockouts are analyzed to see what information can be gained concerning the human enamel phenotype in the absence of AMELX expression and the The enamel is whitest and thickest along the incisal edges and cusp tips. Enamel on the lateral surfaces is thin and rough and retains plaque. B: Panoramic radiograph reveals a thin layer of enamel on erupted and unerupted teeth that is more radioopaque than dentin, but less than normal enamel. doi:10.1371/journal.pone.0052052.g002 potential importance of ARHGAP6 in the process of dental enamel formation.

Materials and Methods
The human study protocol and subject consents were reviewed and approved by the Institutional Review Board at the University of Michigan and the Ethics Committee of Istanbul University, Turkey, where one of the families was recruited. Study participants signed appropriate written consents after an explanation of their contents and after their questions about the study were answered. Any minors age 8 or older signed a written assent form after their parent completed a written parental consent for participation of the minor. The animal protocol was reviewed and approved by the University Committee on Use and Care of Animals (UCUCA) at the University of Michigan.

Characterization of Dental Phenotypes
The proband (III:4) of Family 1 was a 10-year-old dental patient at the University of Istanbul, Turkey. Oral photographs and radiographs were obtained for the proband and his 11-year-old cousin (III:1). The proband of Family 2 (III:7) was an 11-year-old dental patient at the Children's Clinic, University of Michigan School of Dentistry. His family was Caucasian, of Eastern European decent. Oral photographs and radiographs were  Fig. 1A). All teeth are permanent (the secondary dentition). The enamel of the posterior teeth was thickest along the cusp tips and thinner on the lateral surfaces of the crowns, which were stained and showed the underlying dentin. Attrition is evident, particularly on the first molars and the enamel is chipped in many locations. B: Oral photographs of the proband's younger (4 year old) brother (III:5, Fig. 1A). All teeth are from the primary dentition. Attrition of the molar occlusal surfaces shows the thinness of the enamel. The incisal edges of the anterior teeth are worn and chipped. Some teeth are secondarily affected by dental caries. doi:10.1371/journal.pone.0052052.g003 obtained for the proband (III:7) and his mother (II:4). Tooth H of the proband was extracted for orthodontic considerations, photographed, and examined by scanning electron microscopy.

DNA isolation, amplification, and sequencing
Peripheral whole blood (5 cc) or buccal swabs were obtained from the 8 members of Family 1 and 5 members of Family 2. Genomic DNA was isolated using the QIAamp DNA Blood Maxi Kit and protocol (Qiagen Inc, Valencia, CA) and its quality and quantity were determined by spectrophotometry at OD 260 and OD 280 . Genomic DNA (50 ng) was amplified using the Platinum PCR Supermix (Invitrogen, Carlsbad, CA), and the amplification products were purified using the QIAquick PCR Purification Kit and protocol (Invitrogen, Carlsbad, CA). The concentration of purified amplicon was estimated by the intensity of its ethidium bromide-stained band on a 1% agarose gel. The DNA sequencing year old Caucasian male. The color of dentin showed through the thin enamel surfaces of the permanent central incisors, which had pitted, rough surfaces and whiter incisal edges. The enamel of the permanent first molars was thickest along the cusp tips and marginal ridges, and thinner on the occlusal and lateral surfaces. There were spaces between most teeth. D: Bitewing radiographs and E: a panoramic radiograph reveal a thin layer of enamel that is more radioopaque than dentin, but not as radiodense as normal enamel. doi:10.1371/journal.pone.0052052.g004 reactions used 3 ng/mL for each 1000 bp of amplification product size and 1.0 pMol/mL of oligonucleotide primer and were analyzed using an ABI Model 3700 DNA sequencer (Applied Biosystems, Foster City, CA) at the University of Michigan DNA sequencing core.

Polymerase Chain Reaction and DNA Sequence Analyses
The seven AMELX exons along with adjoining intron sequences were amplified by polymerase chain reaction (PCR); however, amplification products were only observed for male and female controls-not from either proband (Fig. S2). We concluded AMELX was deleted in both probands. Because the families reported no notable medical histories and the clinical phenotype was limited to the enamel layer, we expected the deletions to be relatively small. PCR primer pairs were designed to survey upstream and downstream regions. Eight primer pairs sampled intron 1 of ARHGAP6 from nucleotides 50,924 to 291,746 (NCBI genomic reference sequence NG_012494.1) (Fig. S3). PCR amplification products were observed from both probands in all reactions. ARHGAP6 exons 2 through 7 were sampled on the other side of AMELX. Only exon 2 of ARHGAP6 in family 2 failed to amplify (Fig. S3). Thus the deleted regions were confined to the 59 region of ARHGAP6. Additional amplifications narrowed down the locations of the deletions in family 1 (Fig. S4) and family 2 (Figs. S5 and S6).
To sequence across the deletion in Family 1 genomic DNA was amplified with F: GCTAATTATTGGTGGAAAAG and Fa1-11R: GAACAGAGGCAGGCTGTGTC, producing a 2200 bp product. The amplification reactions included a 2 min denaturation at 94uC followed by 35 cycles at 94uC for 30 s, annealing at 57uC for 1 min, extension at 72uC for 6 min followed by a final extension for 20 min. The products were characterized by DNA sequencing priming with the primers used for amplification as well as Fa1-BPF: TTGCCAATCTGCTTTTTACAG, and Fa1-BPR: TATCCTCTTTGTGGGACAGC.
To sequence across the deletion in family 2, genomic DNA was amplified with Fa2-5F: TGAAAGCTAGAGGGGAAACC and Fa2-10R: TGCCAAGAGTAGCCATTTGA, producing a 1969 bp product. The amplifcation reactions included a 5 min denaturation at 94uC followed by 35 cycles at 94uC for 90 s, annealing at 58uC for 1 min, extension at 72uC for 3 min, followed by a final extension for 7 min.

Scanning electron microscopy (SEM)
The primary tooth (H) from the proband of Family 2 was sputter coated with gold for 75 s and then imaged using a Field Emission Gun Scanning Electron Microscope (FEG-SEM; Amray 1910 Field Emission Scanning Electron Microscope) at the Microscopy and Image Analysis Laboratory at the University of Michigan.

Arhgap6 RT-PCR in mouse molars
To determine which of the four Arhgap6 promoters (1A-1D) are used during tooth development, day 5 and day 11 developing first mandibular molars were collected from C57BL6 wild-type mice and processed according to the published protocol [33]. Secretory and maturation stage ameloblasts were obtained by laser capture microdissection (LCM) from the cusp slopes of ,3068 mM sections of day 5 and day 12 maxillary first molars, respectively. This was accomplished using a Leica AS LMD (Leica Micro- systems, Wetzlar, DE). The cuttings dropped by gravity into PCR test tubes containing RNA extraction buffer. RNA was extracted using an Arcturus PicoPure RNA Isolation Kit (Life Technologies, Carlsbad CA, USA). Then 16.5 ng of RNA was converted into cDNA using Invitrogen's VILO cDNA Synthesis Kit (Life Technologies). RNA was extracted from spleen, lung, and enamel organ epithelium (EOE) of day 5 and day 12 mouse mandibular first molars for positive PCR controls. The PCR reactions used promoter-specific forward primers (Ex1aF: GCAAGCATCCT-CAGTTCCTC; Ex1bF: GCAGTGAAGTAAGGGGACCA; Ex1cF: GTGACTCCTAGGGGACCACA; Ex1dF: AAGACAG-CAAAGACACCGAGA) paired with an exon 4-specific reverse primer (Ex4R: GGGATAAGGGCATTCCAAAT). The PCR amplifications included a 5 min denaturation at 95uC followed by 35 cycles at 94uC for 30 s, annealing at 58uC for 30 s, extension at 72uC for 1 min. In the last cycle the extension was for 7 min. The products were analyzed on a 1% agarose gel stained with ethidium bromide.

Results
The proband (III:4) of family 1 was a 10-year-old Caucasian male from Turkey. His mixed dentition displayed thin enamel that was only slightly more radiopaque than dentin and in some locations secondarily affected by dental caries (Fig. 1). The occlusal surfaces of the primary molars, particularly the primary first molars, were completely worn through to dentin. The permanent mandibular central incisors had recently erupted and exhibited unusually thin enamel with a rough and pitted surface. The newly erupted first molars showed more enamel on the cusp tips than on the lateral and occlusal surfaces, giving them a ''snow capped'' appearance. The incisal edges of the maxillary permanent central incisors were also visibly whiter than the rest of the crown. The pedigree was consistent with an X-linked or autosomal dominant pattern of inheritance.
The proband's affected male cousin (III:1) at age 12 exhibited enamel defects similar to those of the proband (Fig. 2), although he had an anterior open bite. His dentition showed small spacing between the teeth, and exhibited thin, rough, pitted enamel with extensive calculus build-up, particularly on the mandibular teeth. The enamel covering the cusp tips and incisal edges appeared to be whiter and less severely affected than the enamel on the lateral and occlusal surfaces. Radiographically the enamel layer was thin, but contrasted with dentin. The proband's 15 year old (III:3) and 4 year old (III:5) brothers were also affected (Fig. 3). The enamel of the primary dentition was thin, chipped easily, and showed pronounce attrition on working surfaces. The enamel of the secondary dentition was better formed on the incisal edges of the anterior teeth and the cusp tips and ridges of the posterior teeth, with progressively less enamel cervically. Based upon the darker color of the lateral surfaces of the posterior teeth of the 15 year old relative to those of his 10 and 12 year old relations, it appears that the enamel deteriorates relatively rapidly following eruption.
The proband of family 2 (III:7) was an 11-year-old Caucasian male of Eastern European ancestry with an anterior cross-bite. His late, mixed dentition displayed thin, rough enamel that barely contrasted with dentin on radiographs (Fig. 4). The incisal edges of the anterior teeth were whiter than the rest of the crowns and the posterior teeth had more enamel on the cusp tips than on the lateral surfaces. Thus, the distinctive enamel phenotype in the male subjects was a ''snow-capped'' appearance. Overall the character of the enamel defects was remarkably similar for the male members of the two families, although the proband of family 2 seemed to have greater surface roughness.
A primary cuspid ( Fig. S7; tooth H) extracted from the proband for orthodontic considerations was examined by scanning electron microscopy (Fig. 5). The enamel layer was readily distinguished from dentin and measured from 177 mm thick at the cingulum to 314 mm where the enamel still remained on the cusp slope in the direction of the cusp tip. This is well below the average thickness (1140 mm) of enamel on primary teeth [34]. The scanned surface Figure 7. Correlating genotypes and affection status. A: Pedigrees for family 1 (left) and family 2 (right). B: PCR primers used to amplify across the deletion junctions. The numbers associated with the primers are the positions of the annealing sites in the ARHGAP6 genomic reference sequence (NG_012494.1). The same forward primer was paired first with a reverse primer within the deleted region to specifically detect wild-type, and then with a primer that annealed on the other side of the deletion to specifically detect the defective allele. (The wild-type wouldn't amplify because the product is too large.). C: PCR amplification for wild-type X-chromosome. D: PCR amplification for X-chromosome with deletion. Normal males (XY) and females (XX) only show a product in the first amplification. Affected females (X*X) show a product in both amplifications. Affected males (X*Y) show only a product in the second amplification. doi:10.1371/journal.pone.0052052.g007 of the enamel appeared smooth, but marked by craters roughly 10 mm to 50 mm in diameter that were not noticable without magnification. The dentin appears to be completely normal.
The mother (II:4) of the proband was less severely affected than any of the male subjects (Fig. 6). Her enamel was thin and contrasted well with dentin on radiographs, but was not as radioopaque as normal. White flecks, similar to those associated with mild fluorisis, were observed in the anterior teeth, and were sometimes arranged in vertical columns.
Mutation analyses using genomic DNA from the probands of families 1 and 2 were initiated, but none of the seven AMELX exons amplified with the appropriate primers. PCR primer pairs were designed to survey surrounding DNA sequence within ARHGAP6, the results of which narrowly defined the deleted regions in both families (Figs. S3, S4, S5, S6 and S7). Ultimately, we were able to amplify across the deletions in both families and determine the DNA sequences at the break points (Fig. 7). The deletions correlated with the enamel malformations in both families. All affected females were heterozygous for the AMELX deletion (X*X). The 96,240 bp deletion in ARHGAP6 in family 1 (g.302534_398773del96240) was confined to intron 1 of that gene, but deleted all of AMELX. The 52,654 bp deletion in ARHGAP6 in family 2 (g.363924_416577del52654insA) deleted the 39 end of intron 1 including all of AMELX, exon 2, and a small part of the 59 end of intron 2. To better assess the possible effects of these deletions on ARHGAP6 expression we analyzed the DNA sequences 59 to exon 2 on human and mouse ARHGAP6 expressed sequence tags (ESTs). This analysis indicated that ARHGAP6 expression is directed by at least 4 different promoters, each associated with a different exon 1 (exons 1a, 1b, 1c, and 1d). We identified one human EST and 4 mouse ARHGAP6 ESTs that contained exon 1a; 5 human and 44 mouse ESTs with exon 1b; one human and 9 mouse ESTs with exon 1c; and one human and no mouse ESTs with exon 1d (Fig. S8). The structure of the ARHGAP6 gene showing the positions of the alternative promoters, AMELX, and the deletions in families 1 and 2 is shown in Fig. 8.
To identify which Arhgap6 promoters are used during enamel formation, we isolated mRNA from ameloblasts (by laser capture) and enamel organ epithelia (EOE) of mouse first molars at days 5 (secretory stage) and 12 (maturation stage), amplified them with primers specific for exons 1a through 1d, and analyzed the products on agarose gels stained with ethidium bromide (Fig. 9). No Arhgap6 expression from promoters 1a and 1c was detected in secretory stage ameloblasts or EOE or maturation stage ameloblasts. Only trace expression was detected from these promoters in maturation stage EOE. Arhgap6 expression from promoter 1b was low, but detectable, in secretory stage ameloblasts and EOE, and was noticably stronger in maturation stage ameloblasts and EOE. These results indicate that in developing mouse molars Arhgap6 is expressed from promoter 1b, and mostly during the maturation stage.
Arhgap6 exons 1a and 1b both contain predicted translation initiation sites in frame with the downstream coding sequence. Neither of these promoters were deleted in our families. Exon 2 however, was deleted in family 2. The skipping of exon 2 (160 nucleotides), which is entirely coding, would shift the reading frame and likely cause degradation of mRNA expressed from the truncated Arhgap6 gene in family 2. Thus while both of our families have an X-chromosome lacking the amelogenin gene, the shortened X-chromosome in family 1 likely has a functional ARHGAP6 gene, while the one in family 2 does not.

Discussion
Amelogenin is a tooth-specific protein that accounts for almost 90% of the protein in secretory stage enamel [35]. The human expressed sequence tag database (Hs.654436), which does not include developing teeth, shows zero amelogenin transcripts out of 3,328,058 for normal tissues. The amelogenin gene is pseudogenized in vertebrates that have lost the ability to make teeth or enamel during evolution [36,37], and tooth defects are the phenotype observed in families with AMELX mutations. Multiple amelogenins are expressed due to alternative splicing [14,38,39]. Following their secretion, amelogenins are processed by matrix metalloproteinase 20 (MMP20) into assorted cleavage products that accumulate throughout the matrix [40,41]. Removal of amelogenin cleavage products from the matrix occurs primarily during the maturation stage [42], and requires the activity of kallikrein 4 [43,44]. Human AMELX mutations produce a diversity of enamel phenotypes that seem to correlate with the position of the mutation [45]; however, a Tyr to His substitution in the Nterminal domain coding region of mouse Amelx caused extensive cytotoxicity because the ameloblasts failed to secrete the mutant protein, engorging the endoplasmic reticulum/Golgi apparatus. [46]. The AI phenotype in some cases must relate to secondary effects like cell pathology, rather than how enamel forms due to amelogenin insufficiency. Complete deletion of AMELX avoids this ambiguity and gives a clearer picture of how enamel forms when only trace amounts of amelogenin (from AMELY) are secreted. A remarkable observation in both of our families with AMELX deletions is that a thicker layer of enamel forms on the cusp tips and marginal ridges relative to the lateral tooth surfaces. The presence of this enamel cannot readily be explained by expression from AMELY as mice lack an amelogenin gene on the Ychromosome, and yet Amelx null mice produce a thin enamel layer [47]. In contrast, Enam and Ambn null mice fail to make an enamel layer [48,49]. These observations force a reconsideration of the possible mechanisms of dental enamel formation and favor a model in which enamelin and ameloblastin are essential components of a mineralization front apparatus along the secretory surfaces of the ameloblast cell membrane [50], whereas amelogenin serves other roles.
AMELX is nested within the large (.400 kb) first intron of ARHGAP6 [51]. ARHGAP6 is expressed in many tissues, but at low levels. There are 28 ARHGAP6 ESTs (Hs.435291) out of a total of 3,328,058 human ESTs for normal tissues. ARHGAP6 activates the GTPase activity of RhoA [27], which inhibits RhoA, a small G protein that regulates actin polymerization. RhoA activity appears to be essential for secretory stage enamel formation in that transgenic expression of enhanced green fluorescent protein (EGFP) fused to a dominant negative variant of RhoA (EGFP-RhoA DN ) in secretory stage ameloblasts resulted in the production of hypoplastic, pitted enamel [52]. ARHGAP6 may act in ways not mediated by RhoA by directly interacting with actin [27] or phospholipase C delta 1 (PLC-d1) [53].
Both of our AI families had complete AMELX deletions, but the missing parts of ARHGAP6 varied. The deletion in Family 1 removed ARHGAP6 alternative promoters 1c and 1d. Our RT-PCR analyses showed that these promoters are not active in ameloblasts, so it is unlikely that this deletion affects ARHGAP6 expression in developing teeth. The deletion in Family 2 removed exon 2 of ARHGAP6, which should have precluded expression of the ARHGAP6 protein. It seems likely then that Family 1 had an AMELX defect only, while Family 2 had a combined AMELX and ARHGAP6 defect. The enamel phenotypes in both families were similar with a characteristic snow-capped appearance caused by the deposition of a relatively thick layer of enamel on the cusp tips, the buccal-occlusal and lingual-occlusal cusp slopes, and marginal ridges. The enamel defects were severe, but there was not a complete absence of enamel. The proband of Family 2 seemed to have less contrasting enamel on radiographs and greater surface roughness than the affected members of Family 1. The enamel phenotypes in these two families are remarkably similar given the broader phenotypic variation observed among the many persons with defined AMELX mutations (Fig. S1). We suspect that the minor variations in enamel phenotype between our two families reflects the natural range of phenotypic variation in persons lacking AMELX, but the suspicion that an absence of ARHGAP6 is the cause of the increased surface roughness cannot be excluded given the minor differences in enamel phenotype observed in Amelx null and Amelx/Arhgap6 double null mice [26]. Despite the possibility that defects in ARHGAP6 could modify the AMELX null phenotype, ARHGAP6 defects alone apparently do not cause enamel malformations. Arhgap6 null mice showed no enamel defects [26]. AMELX mutations are consistently found in human kindreds with X-linked AI, suggesting there is no second X-linked gene that causes AI, as this would lead to the accumulation of unsolved X-linked AI cases without AMELX mutations, which is not observed. Figure S1 AMELX disease-causing mutations. (DOC) Figure S2 Tooth H from the proband of family 2 before it was processed for SEM analyses.