Wolf-Hirschhorn syndrome candidate 1 (Whsc1) methyltransferase signals via a Pitx2-miR-23/24 axis to effect tooth development

Wolf-Hirschhorn syndrome (WHS) is a developmental disorder attributed to a partial deletion on the short arm of chromosome 4. WHS patients suffer from oral manifestations including cleft lip and palate, hypodontia, and taurodontism. WHS candidate 1 (WHSC1) gene is a H3K36-specific methyltransferase that is deleted in every reported case of WHS. Mutation in this gene also results in tooth anomalies in patients. However, the correlation between genetic abnormalities and the tooth anomalies has remained controversial. In our study, we aimed to clarify the role of WHSC1 in tooth development. We profiled the Whsc1 expression pattern during mouse incisor and molar development by immunofluorescence staining and found Whsc1 expression is reduced as tooth development proceeds. Using real-time quantitative reverse transcription PCR, Western blot, chromatin immunoprecipitation, and luciferase assays, we determined that Whsc1 and Pitx2, the initial transcription factor involved in tooth development, positively and reciprocally regulate each other through their gene promoters. miRNAs are known to regulate gene expression posttranscriptionally during development. We previously reported miR-23a/b and miR-24-1/2 were highly expressed in the mature tooth germ. Interestingly, we demonstrate here that these two miRs directly target Whsc1 and repress its expression. Additionally, this miR cluster is also negatively regulated by Pitx2. We show the expression of these two miRs and Whsc1 are inversely correlated during mouse mandibular development. Taken together, our results provide new insights into the potential role of Whsc1 in regulating tooth development and a possible molecular mechanism underlying the dental defects in WHS.

Wolf-Hirschhorn syndrome (WHS) is a developmental disorder attributed to a partial deletion on the short arm of chromosome 4. WHS patients suffer from oral manifestations including cleft lip and palate, hypodontia, and taurodontism.WHS candidate 1 (WHSC1) gene is a H3K36-specific methyltransferase that is deleted in every reported case of WHS.Mutation in this gene also results in tooth anomalies in patients.However, the correlation between genetic abnormalities and the tooth anomalies has remained controversial.In our study, we aimed to clarify the role of WHSC1 in tooth development.We profiled the Whsc1 expression pattern during mouse incisor and molar development by immunofluorescence staining and found Whsc1 expression is reduced as tooth development proceeds.Using real-time quantitative reverse transcription PCR, Western blot, chromatin immunoprecipitation, and luciferase assays, we determined that Whsc1 and Pitx2, the initial transcription factor involved in tooth development, positively and reciprocally regulate each other through their gene promoters.miRNAs are known to regulate gene expression posttranscriptionally during development.We previously reported miR-23a/b and miR-24-1/2 were highly expressed in the mature tooth germ.Interestingly, we demonstrate here that these two miRs directly target Whsc1 and repress its expression.Additionally, this miR cluster is also negatively regulated by Pitx2.We show the expression of these two miRs and Whsc1 are inversely correlated during mouse mandibular development.Taken together, our results provide new insights into the potential role of Whsc1 in regulating tooth development and a possible molecular mechanism underlying the dental defects in WHS.
Wolf-Hirschhorn syndrome (WHS) is a genetic disorder that is estimated to affect one in 50,000 births.WHS patients show different levels of symptoms which consist of a characteristic facial appearance, delayed growth and development, intellectual disabilities, and seizures (1)(2)(3)(4)(5).WHS is caused by subtelomeric deletions of the short arm of chromosome 4 (4p) and the variety and severity of the clinical features largely depend on the number and roles of the genes in the deletion (6)(7)(8).In addition to these common features, delayed tooth development and tooth anomalies have also been considered as underestimated traits with a prevalence of 50% in WHS patients (9)(10)(11).WHS candidate 1(WHSC1) or nuclear receptor binding SET domain protein 2 (NSD2), is a H3K36specific methyltransferase (12)(13)(14)(15).This gene resides in the "critical region" of WHS and is deleted in every case of WHS (4,16).Interestingly, it has been reported that patients with nonsense or loss-of-function variant of WHSC1 also exhibit a subset of WHS features, including intrauterine growth retardation and global developmental delay (17,18).Interestingly, a patient with de novo nonsense mutation in the WHSC1 gene (c.3412C> T, p.Arg1138Ter, NM_001042424.2) also shows tooth enamel dystrophy (19).Whsc1-deficient mice have WHS features including growth retardation, cleft palate, bone defects, congenital heart disease, and malocclusion (15,20).These reports suggest a potential role of WHSC1 in regulating tooth development.
Tooth development is a complex process requiring reciprocal interactions between the dental epithelium and mesenchyme, involving bone morphogenetic protein, wingless-related integration site, fibroblast growth factor, and Sonic Hedgehog signaling pathways (21).This cell-cell communication requires rigid spatiotemporal regulation of transcription factors (TFs) to ensure a staged morphogenesis of individual tooth germs (initiation, placode, bud, cap, and bell), as well as odontoblast and ameloblast differentiation, the two unique dental cell types producing dental hard tissues (22).PITX2 is the earliest TF observed in tooth development, which has long been considered to regulate the transcriptional hierarchy in early stages of tooth development, as well as the stem cell niche (23)(24)(25)(26)(27)(28).It controls dental epithelial stem cell activity by activating several genes, including Lef-1 and Sox2 and thus initializes embryonic tooth development and enamel formation (29)(30)(31)(32)(33)(34).
In this report, we aimed to determine if Whsc1 has a role in regulating tooth development.We hypothesized that Whsc1 participates in the regulatory network involving TFs and miR-NAs during mouse lower incisor development.In this work, we describe a new molecular mechanism for Whsc1/Nsd2 in regulating mouse tooth development.Whsc1, as a methyltransferase, positively regulates the Pitx2 promoter.The Whsc1 promoter is also regulated by Pitx2.Furthermore, Whsc1 is directly targeted by miR-23a/b and miR-24-1/2.The expression of Whsc1 and miR-23-27-24 clusters are negatively correlated during development.Interestingly, miR-23-27-24 clusters are repressed by Pitx2, through interaction of the Pitx2 protein on a distal binding element upstream of the miR clusters.

Significance
While the developmental and protein function of Whsc1 has been studied, little is known about its specific expression pattern and regulatory role in causing WHS.Our research defines a specific expression and regulatory network between Whsc1, Pitx2, and miRs encoded by the miR-23-27-24 cluster in tooth development, which also indicates a potential mechanism underlying tooth anomalies in WHS.

Whsc1 is developmentally regulated and associated with mesenchymal and epithelial progenitor cells in the mouse incisor
We first examined the expression pattern of Whsc1 during mouse incisor development (Fig. 1).A schematic of the embryonic stages of mouse lower incisor development is shown for reference (Fig. 1A).At the placode stage (E11.5),Whsc1 is extensively expressed in the mandible and maxilla.It is more restricted to the dental mesenchyme (DM) and dental epithelium in the mandible, while still expressed ubiquitously in the maxilla at the bud stage (E13.5).At the cap stage (E14.5),when the enamel knot (EK) is formed, Whsc1 expression is limited to the lingual cervical loop, labial cervical loop (LaCL) and the DM in the lower incisor tooth germ.Later at bell stages (E16.5 and E18.5), Whsc1 expression decreases over time.At P1, when the mature tooth germ is formed, Whsc1 expression is further reduced and limited to the transit amplifying zone and some preameloblasts in the LaCL, as well as some dental mesenchymal cells (Fig. 1B).By costaining with Lef-1, which marks the EK in the lower incisor tooth germ, we found that Whsc1 was excluded from the EK (Fig. 1C).EK is generally recognized as a signaling center that controls the growth of the surrounding epithelium and mesenchyme, while it itself contains a group of nondividing cells (51,52).Previous studies have reported several cell cycle proteins are expressed in the developing lower incisor tooth germ.The expression pattern of Ki-67 and cyclin A is similar to Whsc1 (53,54).We also performed immunofluorescence staining of Whsc1 in the developing first molar and found a similar expression pattern compared to the incisor (Fig. S1, A and B).This specific expression pattern suggests that Whsc1 may be linked to proliferation.To test this hypothesis, we performed cell proliferation assays with LS-8 and LS-8-NSD2 cells.Both the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) and cell counting assays demonstrate a role of Whsc1 in activating proliferation in dental epithelial cells (Fig. S1, C and D).
We next costained Whsc1 with progenitor cell markers in the P0 (at birth) tooth germ (Fig. 2).We first costained Whsc1 with Pitx2, the first transcription marker observed during tooth development, and found that they colocalize in the LaCL or dental epithelial stem cells (DESCs) (Fig. 2A).Another DESC marker, Sox2 shows a similar colocalization with Whsc1 in the DESCs (Fig. 2B).Gli-1, a Sonic Hedgehog signaling mediator and a dental mesenchymal stem cell marker (55,56), colocalizes with Whsc1 in the DESCs and DM (Fig. 2C).These data together demonstrate that Whsc1 is associated with the epithelia and mesenchyme progenitor cells.

Whsc1 and Pitx2 positively and reciprocally regulate their expression
Because Pitx2 is an early developmental TF and Whsc1 is expressed in early development, we asked if Pitx2 regulates Whsc1expression.We first established Pitx2 and Whsc1 overexpression stable dental epithelial LS-8 cell lines and performed real-time quantitative reverse transcription PCR (RT-qPCR) (Fig. 3A).Whsc1 mRNA was increased in the LS-8-Pitx2 stable cell line and surprisingly, Pitx2 mRNA was increased in the Whsc1 stable cell line.We confirmed this regulation by transfection of Pitx2 and Whsc1 in HEK 293 cells.Thus, Pitx2 transfection increased Whsc1 (Multiple myeloma SET domain [MMSET] II and MMSETI) protein and Whsc1 transfection increased Pitx2 protein expression, compared to untransformed cells (Fig. 3B).
We next screened the promoter region of the murine Whsc1 gene and found a Pitx2 binding site (TAATCC) at −2.7 kb, which is also conserved in humans (Fig. 3C).The 3.7 kb Whsc1 promoter was cloned in a luciferase reporter, and the overexpression of Pitx2 expression increased luciferase (luc) Whsc1 regulates gene expression during development activity 6-fold in cells (Fig. 3C).While Whsc1 did not increase Whsc1 promoter luc activity, cotransfection of Pitx2 and Whsc1 increased Whsc1 promoter luc activity at 9-fold compared to control (Fig. 3C).We have previously shown that Pitx2 autoregulates its expression by binding to a Pitx2 cis-element in the Pitx2 promoter (57).We next asked if Whsc1 could regulate the Pitx2 promoter.By performing the luciferase assay with the 3 kb Pitx2 promoter reporter, we show that Pitx2 or Whsc1 can upregulate luc activity by 11-fold and 4-fold, respectively (Fig. 3D).Together Whsc1 and Pitx2 show an increase in Pitx2 promoter activity to 14-fold (Fig. 3D).Because Whsc1 is a methyltransferase that methylates H3K36 in the proximal promoters of genes (12)(13)(14)(15)20), Whsc1 may activate the Pitx2 promoter by depositing active chromatin marks onto its promoter.A chromatin immunoprecipitation (ChIP) assay using a Pitx2 antibody indicated that Pitx2 binds to its cis-element in the Whsc1 promoter (Fig. 3E).Interestingly, the binding of Pitx2 and Whsc1 at the Pitx2 promoter were detected by the ChIP assay (Fig. 3F), suggesting that Whsc1 is binding to chromatin at the Pitx2 binding element in the Pitx2 promoter.Moreover, H3K36me2 is also associated with the Pitx2 binding sites in the Pitx2 and Whsc1 promoters demonstrated by ChIP assays (Fig. S2).In addition, Whsc1 as well as the H3K36me2 modifications were detected at the Pitx2 binding element in the Amelogenin and Sox2 promoters (Fig. S3).
Taken together, these data indicate that Pitx2 upregulates Whsc1 expression by binding to and activating its promoter; Whsc1 regulates gene expression during development performed miRNA arrays in P0 mouse tooth germs and revealed the miR expression profile in the late-stage development of mouse incisors and molars (37).We also performed a miR array in the P0 incisor epithelium of WT and Pitx2c Tg mice to detect miRs regulated by Pitx2.Interestingly, we found that miRs from the miR-23-27-24 cluster (miR-23a/b, miR-24-1/2 and miR-27a/b) are highly expressed in both incisors and molars.Moreover, their expressions are downregulated in the P0 incisor epithelium of Pitx2c Tg mice (Fig. 4A).We then confirmed that miR-23a and miR-23b expressions were decreased by Pitx2 in transfected LS-8 cells (Fig. 4B).We found a highly conserved Pitx2 binding site approximately 8.7 kb upstream of the transcription start site (TSS) of pre-miR-23a-27a-24-2 (Fig. 4, C and D).We performed a ChIP assay with the anti-PITX2 antibody and confirmed the binding of Pitx2 at this specific genomic region (Fig. 4E, see asterisk, and F control).We then cloned a 1 kb region containing the Pitx2 binding element into a luciferase reporter plasmid and performed luciferase assay.The result showed that overexpression of Pitx2 can repress the luciferase activity from the miR promoter by 50% (Fig. 4G).There is also a highly conserved Pitx2 binding element at 73 kb upstream of pre-miR-23b-27b-24-1 (Fig. S4, A and B).The binding of Pitx2 at this element was confirmed by ChIP assay (Fig. S4, C and D).These data  only bind to the WT 3 0 UTR to reduce luciferase activity (50%) in the PMIS-EV LS-8 cell line (Fig. 5D).However, this inhibition can be recovered when endogenous miR-23-a/b was inhibited by PMIS-miR-23 in the LS-8 cell line (Fig. 5D).

Discussion
Epigenetic regulation of gene expression during development is fundamental to tissue and organ specific development.These processes have been well-documented and provide a level of gene expression that supports cell proliferation and differentiation (58)(59)(60)(61).Epigenetic dysfunction is associated with cancer and developmental anomalies such as (MMSET/NSD2) (62).miRNAs regulate gene expression posttranscriptionally and combined with epigenetic factors they offer a mechanism to spatially and temporarily modulate gene expression during different cell processes such as transcription and translation (63).In this report, we have identified a new gene expression regulatory mechanism where Whsc1, a methyltransferase, modulates H3K36 at proximal promoters.However, Whsc1 expression is also regulated by miR-23a/b and miR-24-1/2.We demonstrate new mechanisms for Whsc1 in regulating murine tooth development (Fig. 8).

Whsc1 regulates tooth development and underlying mechanisms of tooth anomalies in WHS
Tooth anomalies are an underrepresented feature of WHS, with a prevalence of 50% among WHS patients.A candidate gene Msh Homeobox 1 (MSX1), which is located 4.9 Mb from the telomere of 4p short arm, is highly associated with tooth agenesis in WHS, including oligodontia (10,64).However, patients with small deletions (less than 3 Mb) were also found exhibiting delayed tooth eruption and oligodontia, indicating alternative mechanisms underlying the tooth agenesis in WHS (65,66).WHSC1 is deleted from every case of WHS.We have demonstrated that Whsc1 is expressed in the mouse developing tooth germ, and its expression becomes more restricted during embryonic development.At birth (P0) Whsc1 expression is localized within the LaCL or dental epithelial stem cells and the adjacent DM stem cell region and transient amplifying cells of the lower incisor.These expression profiles suggest that Whsc1 regulates a progenitor cell niche at later stages of tooth development and associated with undifferentiated cell types.
PITX2 is associated with Axenfeld-Rieger syndrome (ARS), a rare autosomal dominant disorder characterized by a series of craniofacial malfunctions primarily affecting the eye and tooth (23,24).This novel mechanism of Whsc1 and Pitx2 reported in this article has not only broadened our understanding of WHS but has also extended its implications to ARS.Moreover, it also indicates that a potential shared pathway or regulatory network that may contribute to the manifestation of WHS, giving further insights to the etiology of both WHS and ARS.

Whsc1 is associated with proliferating cells in the tooth germ
We show that at E14.5 Whsc1 is expressed in both LaCL and lingual cervical loop epithelial cells but excluded from the EK, which contains cells that are nonproliferative and apoptotic (67,68).This expression pattern is identical to that of several cell cycle markers, including Ki67 and cyclin A (53,54).Our cell proliferation assay indicates that Whsc1 activates dental epithelial cell proliferation.Whsc1 has also been reported to positively regulate cell proliferation (69)(70)(71).These findings demonstrate a potential role of Whsc1 in controlling cell proliferation during tooth development.

Whsc1 expression is controlled by miRNAs
Whsc1 expression begins at the early stage of mouse embryogenesis and decreases during development.In the mature incisor tooth germ at P0, Whsc1 expression is restricted to the Sox2 + cells in the LaCL and Gli-1 + cells in the surrounding mesenchyme.The spatial-temporal regulation of gene expression during tooth development is also known to be regulated by miRs.We have shown that Whsc1 is directly targeted by miR-23a/b and miR-24-1/2.The expression of miR-23a/b and miR-24-1/2 increases as tooth development proceeds, which negatively correlated with the expression of Whsc1.According to the previous microarray, miR-23b is highly expressed in both the mouse P0 incisor epithelium and mesenchyme, while miR-23a and miR-24 are more enriched in the epithelium (37).We recently reported that miR-23a/b regulates ameloblast differentiation by targeting Hmgn2 to inhibit its function (50).Hmgn2 is a chromatin-associated high mobility group protein that binds chromatin and modulates transcriptional activity (72).Hmgn2 inhibits several TFs from binding DNA and this prevents the activation of gene expression.These data demonstrate a new mechanism where miR-23a/b and miR-24 are expressed in the more differentiated cell types in the tooth germ, which controls the ameloblast and odontoblast differentiation through epigenetic factors.

The role of Whsc1 in regulating gene expression
Whsc1, also known as Nsd2, encodes a H3K36-specific methyltransferase.This chromatin marker is highly associated with open chromatin, which is mostly enriched in the proximity to the TSS of a gene and recruits RNA Pol II machinery to activate transcription (12,13).It has been reported that Whsc1 collaborates with TFs to fine-tune the expression of the downstream targets, in regulating heart and bone development (15,20).We show new roles for Whsc1 in regulating the Pitx2 promoter and its downstream targets Sox2 and Amelogenin.Interestingly, we did not observe a direct interaction between Whsc1 and Pitx2 (data not shown), which indicates an indirect interaction between these two proteins or alternatively, an indirect regulatory relationship between these two genes.There are genes that have been reported to interact with the Pitx2 protein and activate its DNA binding and transcriptional activity (31,32,(73)(74)(75).Whsc1 could activate Pitx2 through chromatin modulation at the Pitx2 proximal promoter, or Whsc1 can activate the expression of gene(s) that can increase Pitx2 transactivation activity.
In summary, in the postnatal mouse lower incisor, Whsc1 is coexpressed with Pitx2, a dental epithelial stem cell marker, in the transit amplifying zone of inner enamel epithelium, stellate reticulum, as well as some pre-ameloblasts in the LaCL.In the Whsc1+/Pitx2+ epithelial cells, Pitx2 activates Whsc1 expression by binding to its promoter, but inhibits miR-23/27/24 cluster expression by binding to its distal regulatory element; Whsc1 activates the Pitx2 promoter by depositing the active chromatin mark H3K36me1/2.In the dental mesenchyme, Whsc1 is also expressed in the Gli-1+ dental mesenchymal stem cells.miR-23a/b and miR-24-1/2 are expressed in the more differentiated DM and epithelial cells, with miR-23b expressed in the mesenchyme and all of the miR-23/27/24 cluster is expressed in the epithelium (37).In those cells miR-23a/b and miR-24-1/2 bind to the 3 0 -UTR regions of Whsc1 and inhibit its expression and function.Thus, we show how Whsc1 is required for early progenitor cell propagation and the later expression of miR-23a/b and miR-24-1/2 is required for cell differentiation by reducing Whsc1 expression.

Experimental procedures
Mouse strain breeding and embryonic staging All animals were housed, and all procedures performed in accordance with the guidelines approved by the University of Iowa Office of Animal Care.All experimental procedures were approved in accordance with the University of Iowa IACUC guidelines.For embryonic staging experiments, the observed vaginal plug date of the female was designated as E0.5.Embryos were collected at the required time point.Embryos were subjected to tissue fixation or mandibular RNA extraction right after harvesting.The Krt14-PITX2C transgenic line has been described previously (76).In the postnatal mouse lower incisor, Whsc1 is coexpressed with Pitx2, in the LaCL, the transit amplifying zone of inner enamel epithelium, and stellate reticulum.In the Whsc1+/Pitx2+ epithelial cells, Pitx2 activates Whsc1 expression by binding to its promoter, but inhibits miR-23/27/24 cluster expression by binding to its enhancer; Whsc1 activates the Pitx2 promoter by depositing the active chromatin mark H3K36me1.In the dental mesenchyme, Whsc1 is also expressed in the Gli-1+ dental mesenchymal stem cells.miR-23a/b and miR-24-1/2 are expressed in the more differentiated dental mesenchyme and epithelial cells, with miR-23b expressed in the mesenchyme and all of them expressed in the epithelium (37).In those cells, miR-23a/b and miR-24-1/2 bind to the 3 0 -UTR regions of Whsc1 and inhibit its expression and function.LaCL, labial cervical loop; Whsc1, Wolf-Hirschhorn syndrome candidate 1.

Immunofluorescence staining and histology
Murine embryos and postnatal pups were fixed in 4% paraformaldehyde (ChemCruz, Santa Cruz Biotechnology) and taken through a standard dehydration before being embedded in the paraffin.Samples were sectioned at 7 μm with a Thermo (HM325) microtome as previously reported (33).Sections were subjected to standard H&E staining protocols.Slides with paraffin sections were subjected to a series of dewaxing and rehydration steps and then followed by citric acid antigen retrieval in a 100 C water bath for 12 min.The slides were then blocked with 20% donkey serum and incubated with primary antibodies overnight at 4 C. Slides were washed with 1xPBS, incubated with Alexa Fluor 488 and 594 secondary antibodies (Thermo Fisher Scientific), and stained with 4 0 ,6-diamidino-2-phenylindole.Confocal pictures were taken with a ZEISS 700 confocal microscope and Zen imaging software.The primary antibodies are listed in Table 1.
Cell culture, transfections, and reporter assays LS-8 (82), and HEK 293 cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin. PITX2, WHSC1/NSD2, PMIS-EV, PMIS-miR-23, and PMIS-miR-24 expression plasmids were transfected into LS-8 or HEK 293 cells by either PEI or Lipofectamine 2000 (Invitrogen) reagents, followed by RT-qPCR and Western blot assays.For luciferase reporter assays, cells were seeded 24 h before transfection in 60 mm petri dishes and transfected with 2.5 μg of expression plasmid, 5 μg of reporter plasmid and 0.2 μg of SV-40 β-galactosidase plasmid.Cell transfections were performed by either PEI or Lipofectamine 2000 (Invitrogen) reagents with a DNA:PEI/Lipofectamine 2000 ratio of 1:3 or 1:2.Transfected cells were incubated in 60 mm culture dishes for 48 h and fed with 10% FBS and DMEM.Following lysis with either Reporter Lysis 5X Buffer (Promega) or radioimmunoprecipitation assay buffer, assays for reporter activity (luciferase assay, Promega) as well as for protein concentration (Bradford assay, Bio-Rad) were carried out.β-galactosidase was measured using the Galacto-Light Plus reagents (Tropix Inc) as an internal normalizer.All luciferase activities were normalized to βgalactosidase activity and are shown as mean-fold differences relative to empty luciferase plasmids and are shown as mean ± SEM.
Lentiviral production and stable cell line establishment HEK 293T cells were seeded in a 100 mm cell culture dish followed by PEI transfection with pMD2.G, psPAX2 and PMIS-EV, PMIS-miR-23a, PMIS-miR-24, PITX2, or NSD2 expression plasmids.The medium was changed 24 h post transfection and lentivirus-containing medium was collected at 48 h, 72 h, and 96 h post transfection and filtered through a 0.45 μm polyvinylidene fluoride filter.LS-8 cells were seeded in 100 mm cell culture dishes.After 24 h, the medium was aspirated and replaced by lentiviral-containing medium with 8 μg/ml polybrene.Medium was changed to normal culture medium after 24 h.Transduced cells were either selected through puromycin (1 μg/ml) selection, G418 (5 mg/ml) selection or cell sorting.For cell sorting, the cells were trypsinized, washed and resuspended with PBS, and filtered through a 70 μm nylon mesh strainer.The GFP + cell populations were sorted through either Becton Dickinson FACS Aria II or FACS Fusion cell sorters.The sorted cells were then cultured in normal cell culture medium for stable cell line expansion.

MTT assay
LS8 or LS8-NSD2 cells were seeded in triplicates for each assay time points in 96-well plates at 20,000 cells/well and

Cell counting assay
LS8 or LS8-NSD2 cells were seeded in quadruplicates for each harvesting time points in 60 mm dishes at 10 5 cells/dish and cultured with DMEM (10% FBS and 1% penicillin/ streptomycin).Cells were trypsinized and suspended with 1.5 to 2.5 ml culture medium, followed by cell counting using a hemocytometer at 24 h, 48 h, 72 h, 96 h, and 120 h post seeding.

Real-time quantitative reverse transcription PCR
Total RNA was isolated from cells or mouse mandible tissues using miRNeasy Mini Kit (Qiagen) or standard RNA preparation protocol.Reverse transcription and quantitative real-time PCR for microRNA detection were carried out with miScript PCR system (Qiagen) according to the manufacturer's protocol.Reverse transcription and quantitative real-time PCR for mRNAs were performed using a TaKaRa kit (TaKaRa, RR036A, RR420L).All Ct numbers were below 35 cycles.PCR products were examined by melting curve analysis and the sequences were confirmed.Fold change was calculated using the 2 -ΔΔCT method.The primers used for qPCR are listed in Table 2.The primers used to detect miR-23a/b-3p and miR-24-1/2 were purchased from Qiagen.

Western blot assays
Cell lysates from LS-8, HEK 293 cells and stable cell lines were analyzed on 4% stacking gel and 10 to 15% SDS-PAGE separating gels.Following electrophoresis, the protein was transferred to polyvinylidene fluoride membranes (Millipore), immunoblotted, and detected with a horseradish peroxidaseconjugated secondary antibody and Clarity Western ECL Blotting Substrate (Bio-Rad).The following antibodies were used to detect the proteins (Table 3).

Chromatin immunoprecipitation assay
ChIP assays were performed as previously described (33) using the ChIP Assay Kit (Zymo research).LS-8 cells were cross-linked in 1% formaldehyde at room temp for 7 min.Cross-linked cells were sonicated three times (6 s duration for each round, 25% of maximum amplitude) to shear the genomic DNA in to 200 to 1000 bp fragments.Then the DNA/protein complexes were immunoprecipitated with 5 μg PITX2 antibody (Capra Science), WHSC1 and H3K36me2 (Abcam) antibodies or 5 μg rabbit IgG as control.Precipitated DNAs were subjected to PCR to evaluate the enrichment of Pitx2 binding.The primers used for PCR are listed in Table 4.All the PCR products were analyzed on a 1.5% agarose gel for the correct size and confirmed by sequencing.

miRNA microarray
Incisor and molar tooth germs were dissected from P0 and P10 mice using a dissection microscope.To separate epithelium and mesenchyme, the tooth germs were treated with Dispase II and Collagenase I (Worthington) for 30 min at 37 C.This procedure separates the epithelium from the mesenchyme and allows for specific RNA extraction of the two tissue types (37).Total RNAs including microRNA were prepared using miRNeasy Mini Kit from Qiagen.LC Sciences performed the miRNA microarray analyses.

Statistical analysis
All quantified results are presented as mean ± SEM, and with an N value indicating the number of biological replicates.A two-tailed unpaired Student's t test and either oneor two-way ANOVA were used to determine statistical significance.Whsc1 regulates gene expression during development

Figure 3 .
Figure 3. Whsc1 and Pitx2 positively and reciprocally regulate one another.A, RNA isolated from the LS-8, LS-8-Vector, LS-8-Pitx2, and LS-8-Whsc1 overexpressing cells were analyzed by qPCR for Whsc1 and Pitx2 transcript levels.Relative levels of each transcript are shown (N = 3).B, HEK 293T cells were transfected with control, Pitx2 or Whsc1/Nsd2 expression plasmids and proteins were isolated 48 h after transfection.Isolated proteins were then resolved on a 10% SDS-polyacrylamide gel and probed for Pitx2 and two isoforms of Whsc1 (MMSET I and II).Gapdh was used as internal control.The proteins were visualized using ECL reagents.C and D, LS-8 cells were cotransfected with the Whsc1 or Pitx2 luciferase promoter (5 μg) construct, either Pitx2, Whsc1 or both expression constructs (2.5 μg).The luciferase activities are shown as mean fold activation compared with the normalized luciferase activity in empty vector (pcDNA 3.1) with the Whsc1 or Pitx2 promoter reporter (N = 3-5).E, schematic of the Whsc1 promoter chromatin region and ChIP primers.ChIP assay using anti-PITX2 ABC Ab for chromatin immunoprecipitations (IP).Pitx2 bound to the Whsc1 distal element in the Whsc1 promoter.IgG alone did not IP the chromatin.PCR and no AB control groups did not produce a band.Control primers to an upstream region of the Whsc1 promoter did not detect an IP product in any group except the input.F, schematic of the Pitx2 promoter chromatin region and ChIP primers.ChIP assay using either anti-PITX2 ABC or anti-WHSC1 Ab for chromatin immunoprecipitations. Pitx2 and Whsc1 bound to the Pitx2 promoter.IgG alone did not IP the chromatin.PCR and no AB control groups did not produce a band.Control primers to an upstream region of the Pitx2 promoter did not detect an IP product in any group except the input.#, PITX2 binding site.*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.ChIP, chromatin immunoprecipitation; IgG, immunoglobulin G; IP, immunoprecipitation; MMSET, multiple myeloma SET domain; qPCR, quantitative PCR; Whsc1, Wolf-Hirschhorn syndrome candidate 1.

Figure 5 .
Figure 5. miR-23a/b-3p directly targets and inhibits Whsc1.A, schematics showing the potential miR-23a/b-3p binding site in the Whsc1 3 0 UTR, which is highly conserved among different species.Red ribonucleotides represent the seed sequence of miR-23a/b-3p; blue ribonucleotides represent the potential binding sites; and gold ribonucleotides represent the mutated binding site used in the luciferase reporter.B, RT-qPCR shows Whsc1 is upregulated in the LS-8-PMIS-miR-23 stable cell line.The expression of Whsc1 mRNA is normalized to β-actin.Data are represented as fold change against LS-8-PMIS-EV.C, the two main protein isoforms of WHSC1: MMSET I and MMSET II, are upregulated in the LS-8-PMIS-miR-23 stable cell line.Western blots were quantified using Fiji ImageJ.The band intensity of MMSET I and II were normalized to Gapdh.Data represent the fold change against LS-8-PMIS-EV.D, miR-23a/b-3p directly targets the WT but not the mutated mouse Whsc1 3 0 UTR.The psiCheck2 reporter with either no binding site control, WT or MT Whsc1 3 0 UTR (5 μg) were transfected to both LS-8-PMIS-EV and LS-8-PMIS-miR-23.Cells were incubated for 48 h and then assayed for Firefly and Renilla luciferase activity (N = 5-7).Data were shown as fold change against no binding site control group.*p < 0.05; **p < 0.01.MMSET, multiple myeloma SET domain; Mut, mutant; PMIS, plasmid-based miRNA inhibition system; RT-qPCR, real-time quantitative reverse transcription PCR; Whsc1, Wolf-Hirschhorn syndrome candidate 1.

Figure 6 .
Figure 6.miR-24-1/2-3p directly targets and inhibits Whsc1.A, schematics showing the potential miR-24-1/2-3p binding sites in the Whsc1 3 0 UTR, which are highly conserved among different species.Red ribonucleotides represent the seed sequence of miR-24-1/2-3p; blue ribonucleotides represent the potential binding sites; and gold ribonucleotides represent the mutated binding site used in the luciferase reporter.B, RT-qPCR shows Whsc1 is upregulated in the LS-8-PMIS-miR-24 stable cell line.The expression of Whsc1 mRNA is normalized to β-actin.Data are represented as fold change against LS-8-PMIS-EV.C, the two main protein isoforms of WHSC1; MMSET I, and MMSET II are upregulated in the LS-8-PMIS-miR-24 stable cell line.Western blots were quantified using Fiji ImageJ.The band intensity of MMSET I and II were normalized to Gapdh.Data represent the fold change against LS-8-PMIS-EV.D, miR-24-1/2-3p directly targets the WT but not the mutated mouse Whsc1 3 0 UTR.The psiCheck2 reporter with either no binding site control, WT or Mut Whsc1 3 0 UTR (5 μg) were transfected to both LS-8-PMIS-EV and LS-8-PMIS-miR-24.Cells were incubated for 48 h and then assayed for Firefly and Renilla luciferase activity (N = 5).Data were shown as fold change against no binding site control group.*p < 0.05.BS, binding site; MMSET, multiple myeloma SET domain; Mut, mutant; PMIS, plasmid-based miRNA inhibition system; RT-qPCR, real-time quantitative reverse transcription PCR; Whsc1, Wolf-Hirschhorn syndrome candidate 1.

Figure 7 .
Figure 7. Whsc1 negatively correlates with miR-23-3p and miR-24-3p expression during tooth development.Mouse embryos were harvested at different stages during tooth development.The mandibular RNAs were collected and used for qRT-PCR.The miR-23a/b or miR-24-1/2 expression levels were normalized to U6 and Whsc1 expression was normalized to β-actin.Data at later time points were represented as fold change against E11.5 (N = 3).qRT-PCR, quantitative real-time PCR; Whsc1, Wolf-Hirschhorn syndrome candidate 1.

Figure 8 .
Figure 8. Model for the role of Whsc1, Pitx2, miR-23a/b, and miR-24 in the regulation of tooth development.In the postnatal mouse lower incisor, Whsc1 is coexpressed with Pitx2, in the LaCL, the transit amplifying zone of inner enamel epithelium, and stellate reticulum.In the Whsc1+/Pitx2+ epithelial cells, Pitx2 activates Whsc1 expression by binding to its promoter, but inhibits miR-23/27/24 cluster expression by binding to its enhancer; Whsc1 activates the Pitx2 promoter by depositing the active chromatin mark H3K36me1.In the dental mesenchyme, Whsc1 is also expressed in the Gli-1+ dental mesenchymal stem cells.miR-23a/b and miR-24-1/2 are expressed in the more differentiated dental mesenchyme and epithelial cells, with miR-23b expressed in the mesenchyme and all of them expressed in the epithelium(37).In those cells, miR-23a/b and miR-24-1/2 bind to the 3 0 -UTR regions of Whsc1 and inhibit its expression and function.LaCL, labial cervical loop; Whsc1, Wolf-Hirschhorn syndrome candidate 1.

Table 1
List of antibodies for immunofluorescence The plate was then wrapped with foil and shaken on an orbital shaker for 15 min after which the absorbance was read at 620 nm.

Table 3
List of antibodies used for Western blot assays