Enam expression is regulated by Msx2

Abstract

Background:

The precise formation of mineralized dental tissues such as enamel and/or dentin require tight transcriptional control of the secretion of matrix proteins. Here we have investigated the transcriptional regulation of the second most prominent enamel matrix protein, enamelin, and its regulation through the major odontogenic transcription factor, MSX2.

Results:

Using in vitro and in vivo approaches, we identified that (i) Enam expression is reduced in the Msx2 mouse mutant pre-secretory and secretory ameloblasts, (ii) Enam is an early response gene whose expression is under the control of Msx2, (iii) Msx2 binds to Enam promoter in vitro, suggesting that enam is a direct target for Msx2 and that (iv) Msx2 alone represses Enam gene expression.

Conclusions:

Collectively, these results illustrate that Enam gene expression is controlled by Msx2 in a spatio-temporal manner. They also suggest that Msx2 may interact with other transcription factors to control spatial and temporal expression of Enam and hence amelogenesis and enamel biomineralization.

INTRODUCTION

Amelogenesis is the process of enamel formation. It takes place in three, well-defined stages known as the secretory, transition and maturation phases and the cell mediating this process is the ameloblast. The ameloblast starts its differentiation process as an epithelial proliferative cell, separated from adjacent mesenchyme by a basement membrane. The initial differentiation, orientation of ameloblasts and their coordinated function are critical to amelogenesis and involves the secretion of a proteinaceous matrix in which immature enamel hydroxyapatite (HA) crystallites are deposited. The matrix is then degraded and replaced, almost entirely, with HA mineral. The functional ameloblasts express numerous proteins, including enzymes, signaling molecules, cell-cell adhesion molecules, transcription factors and several secreted proteins, such as ameloblastin, amelogenin, enamelin, tuftelin, DSPP, apin, amelotin ^1,2,3. Of all secreted proteins, amelogenin, enamelin and ameloblastin are the major secretory products of ameloblasts that contribute to enamel formation ^4,5. Amelogenin is the most abundant one, while ameloblastin and enamelin are less abundant and degrade rapidly during enamel formation ⁶.

Enam is uniquely expressed by the ameloblasts and is expressed during the secretory, transition and early maturation stages of ameloblast life cycle ^7,8. In silico analysis of mammalian and non-mammalian tetrapods indicates that the (i) ENAM gene has originated very early in vertebrate evolution with 25 amino acids of its sequence to be conserved for 350 million years of tetrapod evolution, and (ii) its regulation during evolution is critical for attributing correct ENAM functions to different species ^9,10.

Many genes when mutated have been demonstrated to result in non-syndromic and syndromic AI. In humans, mutations in ENAM gene are associated with non-syndromic amelogenesis imperfecta (AI), a heterogeneous group of genetic conditions characterized by defects in the formation of enamel and are found in non-syndromic conditions that affect only the enamel formation of teeth, or are part of congenital disorders, such as ectodermal dysplasias affecting more than one ectodermal organ. The first mutation identified in ENAM gene resulted in a dominant-negative effect of aberrant splicing causing an autosomal dominant AI with a severe, smooth hypoplastic phenotype (MIM #104500) ¹¹. Another milder, local hypoplastic phenotype (MIM #204650) is caused by missense mutations in ENAM ¹², while other autosomal recessive inheritance mutations have also been documented ^{13, 14, 15, 16}; http://dna2.leeds.ac.uk/LOVD/). On the other hand, syndromic AI are part of congenital disorders and the genes responsible for these disorders control the development and/or maintenance of many other organs.

Animal and human data indicate that transcription factors are critical for the formation of several organs and structures including the tooth enamel ². Among them is the transcription factor Msx2. Mice lacking the homeobox gene Msx2 exhibit defects in several ectodermal organs including the tooth and the process of amelogenesis. In the Msx2 knock out (KO) mice the ameloblasts secrete sparse amounts of enamel matrix ^{2, 17, 18, 19, 20}. Detailed morphometric analysis revealed that the amount of enamel deposited by the mutant ameloblasts is < 2% the amount of enamel deposited by the wild type ameloblasts ^{2, 17, 19}. These results show that depletion of Msx2 function causes abnormalities in amelogenesis, by controlling the ameloblast differentiation process and enamel production. Several reports suggest a possible role of Msx2 in regulating enamel formation through the control of downstream genes such as, laminin 5 alpha 3, follistatin, amelogenin, DSPP, cytokeratin 5, MMP20, KLK4, all of which are equally important for enamel formation ^{2, 17–27}. The role of transcription factors such as Msx2 in directly regulating any of the enamel proteins, however, is not extensively studied except for amelogenin ^{8, 18, 21, 28, 29, 30}.

In the present study, we show that (i) Enam expression is reduced in the Msx2 mouse mutant pre-secretoty and secretory stage ameloblasts, (ii) Enam is an early response gene whose expression is under the control of Msx2, (iii) Msx2 binds to Enam promoter in vitro, (iv) Msx2 alone represses Enam gene expression. These results suggest that Enam gene expression is partially under the control of Msx2 regulation in a spatio-temporal manner. They also suggest that Msx2 may interact with other transcription factors and that these interactions could relieve repression allowing thus Enam to be expressed, or Enam is under the control of multiple transcription factors that coordinately control its spatial and temporal expression.

RESULTS

Msx2 is necessary for Enam gene expression during late tooth development

To determine whether Msx2 is required for Enam regulation and whether this requirement is associated with the defect in amelogenesis, in situ hybridization was performed in wild type and Msx2 deficient mouse molar and incisor tooth germs at postnatal day 1 (P1), postnatal day 3 (P3), postnatal day 6 (P6) and postnatal day 9 (P9) (Fig. 1A–P). Msx2 is expressed by pre-secretory ameloblasts (P1), secretory ameloblasts (P3-P6) and is not expressed by (P9) when ameloblasts are at the maturation stage ^{2,3, 17, 18, 26, 31, 32}. A dramatic reduction of Enam expression is observed in Msx2 deficient molars and incisors compared to wild type ones at P1 (Fig. 1A–D), P3 (Fig. 1E–H) and P6 (Fig. 1I–L) when ameloblasts are at their pre-secretory and secretory stage. This result indicates that Enam requires Msx2 for its expression in the pre-secretory and secretory stage ameloblasts. At P9, Msx2 is not expressed by mature ameloblasts, while Enam is restricted to the lower buccal and lingual sides of the crown around the cervical area (Fig. 1M). The expression of Enam showed no considerable difference between wild-type and Msx2-deficient tooth germs at this stage, considering the qualitative nature of the ISH (Fig. 1O). In the Msx2 deficient incisors even a slight increase in Enam expression is observed compared to wild type ones, at P9 (Fig. 1N, P).

Fig. 1: Msx2 is essential for enamelin gene expression during amelogenesis:

In situ hybridization analyses of transcripts: At postnatal day P1: in wild type (A, B) and Msx2 deficient (C, D) first lower molar teeth (A, C) and Incisors (B, D) At postnatal day P3: in wild type (E, F) and Msx2 deficient (G, H) first lower molar teeth (E, F) and Incisors (G, H) At postnatal day P6: in wild type (I, J) and Msx2 deficient (K, L) first lower molar teeth (I, K) and Incisors (J, LAt postnatal day P9: in wild type (M, N) and Msx2 deficient (O, P) first lower molar teeth (M, O) and Incisors (N, P) Expression of Enamelin is reduced in Msx2 deficient ameloblasts compared to wild type at P1-P6 but not in P9. Abbreviations: LM: lower molars; IN: incisors. Scale: X100 (N=4).

Lower Panel: Total RNA from P1, P3, P6 and P9 tooth germs was extracted, reverse transcribed and qPCR was performed. Enamelin is downregulated in the Msx2 deficient P1, P3 and P6 tooth germs but not at P9, further confirming our in-situ results. The higher expression of enamelin in the Msx2 deficient teeth compared to wild type ones at P9 qPCR is the result of the cumulative expression of enamelin from the buccal and lingual sides of the crown of the molars and from the proximal region of the crown of the incisors.

To test whether Msx2 regulates the expression of Enam, in a quantitative manner, we performed real time quantitative PCR. Total RNA from P1, P3, P6 and P9 first molar tooth germs was extracted, reverse transcribed and qPCR was performed. We show that the expression of Enam is downregulated in the Msx2 deficient P1, P3 and P6 tooth germs, but not at P9, further confirming our in vivo results (Fig. 1, lower panel). The higher expression of Enam in the Msx2 deficient teeth compared to wild type ones at P9 qPCR is the result of the cumulative expression of Enam from the buccal and lingual sides of the crown of the molars.

We also performed loss-of and gain-of function studies in LS8 cells because this cell line is an ideal model to test gene regulation during tooth development ^{19, 27}. Specifically, we tested the effects of acute knockdown of Msx2 in LS8 ameloblast-derived cells and compared to what happens in development where Msx2 is permanently absent in the Msx2 deficient mice (Fig. 2A–D). For the knock down experiment, we used siRNA technology in LS8 cells (Fig. 2A, B). After 48 and 72 hrs transfection the cells were subjected to real time quantitative PCR. We found that upon transient silencing of Msx2, Enam was downregulated, further suggesting that Enam requires Msx2 for its expression (Fig. 2A, B). In addition, we used lentiviral shRNA mediated approach to assess the effects of permantly silencing Msx2 gene. Specifically, the LS8 cells were infected with mouse Msx2shRNA lentiviral transduction particles. qPCR shows that Msx2shRNA lentiviral transduction particles effectively reduce Enam gene expression in LS8 cells compared to control shRNA treated cells (Fig. 2C, D). For the gain-of function assays, induction of Msx2 was achieved through transient transfection of Msx2 expression plasmid into LS8 cell lines. Overexpression Msx2 in both cell lines results in significant increase of Enam expression (Fig. 3A).

Fig. 2: Loss of function of Msx2 in LS8 ameloblast-derived cells:

Expression of Enamelin after Msx2 knockdown using 2 different methods (A) siRNA mediated and (C) shRNA lentiviral-mediated to confirm downregulation of Msx2 transiently as well as long-term, respectively. (B) and (D) is the q-PCR to confirm the results as seen in (A) and (C). Gapdh is the normalizing gene. Experiments were done in triplicates. ** p=<0.005; ***p=<0.0001. Dotted line demarcates representative sample.

Fig. 3: The expression of enamelin is upregulated and modulated early in response to Msx2:

(A) Both LS8 and G5 cells were overexpressed with Msx2 over-expression plasmid. Representative RT-PCR showing enamelin is upregulated in both cell lines after Msx2 overexpression. (B) Time dependent Assay: LS8 cells were transfected with Msx2 over-expression plasmid for several time points, 4, 8, 16, 24 and 48 hours. Total RNA was isolated from the cells and subjected to qPCR analysis. Msx2 and Enamelin could be detected as early as 4h after transfection by real time qPCR. By 16h, there was a significant increase in the expression of Msx2 with a corresponding increase of Enamelin expression. Gapdh is the normalizing gene. Bottom panel shows the expression of Msx2 and with corresponding upregulation of Enamelin in a logarithmic scale, after overexpression with Msx2 overexpression vector; Control: cells transfected with control vector only. The experiment was conducted 3 times in replicates of 3. Dotted line demarcates representative sample.

The expression of enam is modulated early in response to Msx2

To determine the kinetics of enam gene expression, we performed a time dependent assay to ascertain whether its expression is modulated early in response to Msx2 upregulation (Fig. 3B). LS8 cells were transfected with Msx2 over-expression plasmid for several time points, 4, 8, 16, 24 and 48 hours. Total RNA was isolated from the cells and subjected to qPCR analysis. Enam could be detected as early as 4h after transfection by real time qPCR (Fig. 3B). Enam levels increased immediately within 4 hours of Msx2 expression and reached a maximum peak at 16 hours post transfection, following which there was a gradual decline around 24h and subsequently in 48h (Fig. 3B). We did not see any significant response earlier than 4 hours; thus, these results indicate that Enam is secondary early response gene to Msx2 (response after 4 hours of Msx2 overexpression).

Msx2 directly binds to Msx2 recognition sites on the Enam promoter

The loss-of and gain-of-function studies along with the in vivo experiments using the Msx2 deficient mice show that Msx2 is required for the expression of Enam in the pre-secretory and secretory ameloblasts, during amelogenesis. Computational sequence analysis of the nucleotides in the proximal 5.2kb of the murine Enam promoter region revealed the presence of 3 putative homeodomain binding sites upstream from the transcription initiation site in the mouse (Fig. 4A). To determine whether Msx2 binds to any of these sites and therefore, directly regulates Enam, chromatin immuno-precipitation was performed with exogenously expressed Msx2-FLAG in LS8 cells and P3 wild type mouse M1 molar tooth germs. Immunoprecipitated chromatin fragments (IP samples) and non-immunoprecipitated samples (1% input) were subjected to PCR analysis using specific primers spanning the three binding sites. PCR amplifications showed that Msx2 binds directly to all three putative sites carrying the conserved motif (TAAT) in the endogenous promoter of the mouse Enam gene, −500TAATta, −2000TAATtta (weak, not shown) and −2900TAAttc (Fig. 4B, C). This result demonstrates that Msx2 binds directly to the proximal Enam promoter in vitro and in vivo.

Fig. 4: Msx2 directly binds to Msx2 recognition sites on the Enam promoter:

(A) In silico model showing three potential homeodomain binding sites (−500bp, −2000bp, −2900bp) on the Enamelin promoter using MatInspector. The different binding sites are represented by red ellipses. Primers were designed from different promoter regions (black arrows). (B&C) After chromatin immunoprecipitation, samples from LS8 cells transfected with pCMV-Msx2-FLAG and P3 wild type mouse M1 molar tooth germs were PCR amplified and the binding region was directly amplified prior to immunoprecipitation (1% Input) and specifically amplified in the immunoprecipitated sample (anti-FLAG) and (anti-Msx2). No amplification was detected in the normal mouse serum IgG-immunoprecipitated sample (IgG; negative control; DW: distilled water, negative control).). Pol A is the positive control. Band in the input lane shows endogenous binding while band in the sample lane shows binding after specific immunoprecipitation with FLAG tagged antibody and Msx2 antibody after overexpression. The results confirm the binding of Msx2 to the predicted regions of the Enam promoter, both in vitro and in vivo. (D) LS8 cells were co-transfected with enamelin-luciferase reporter plasmids and Msx2 expression plasmid. Cells were harvested 24h after transfection for reporter gene assays. Transcription efficiency was determined using Renilla luciferase plasmid. For this and subsequent experiments, the levels of luciferase activity were normalized to Renilla luciferase activity and expressed as fold luciferase activity relative to the level of luciferase activity from cells transfected with the reporter construct and empty expression plasmid. All the transfection experiments were performed three times, and results are shown as means +/− standard deviations. There was almost half-fold repression of enamelin activity in all the 3 regions that contained the putative binding sites. Right panel shows the quantification of the expression level of luciferase for the vector relative to Msx2. The 500bp region shows a significant downregulation of Msx2. ***p<0.005; RLU: relative luminescence unit

To determine if Enam promoter is repressed or activated by Msx2, LS8 cells were co-transfected with Enam-luciferase reporter plasmid and Msx2 overexpression plasmid (Fig. 4D) and luciferase activity measured. We found that transfection of Msx2 with 3 putative binding sites resulted in a repression of the Enam-luciferase reporter plasmids. This shows that Msx2 acts as a repressor, suppressing but not alleviating the expression of Enam, acting as a dosage regulator.

In sum, we show that (i) Enam expression is reduced in the Msx2 mouse mutant pre-secretory and secretory ameloblasts, (ii) Enam is an early response gene whose expression is under the control of Msx2, (iii) Msx2 binds to Enam promoter in vitro, suggesting that Enam is a direct target for Msx2 and that (iv) Msx2 alone represses Enam gene expression. Collectively, these results show that Enam gene expression is partially under the control of Msx2 regulation in a spatio-temporal manner through a remote enhancer. They also suggest that Msx2 may interact with other transcription factors that coordinately control its spatial and temporal expression.

DISCUSSION

Msx2 is required for Enam expression

Ameloblast differentiation program is impaired in Msx2^−/− mice leading to enamel dysplasia ²⁰. Msx2 is expressed by preameloblasts, early secretory, secretory ameloblasts and ceases to be expressed by maturation stage ameloblasts ^{17–19, 26, 31, 32}. On the other hand, Enam expression is initiated early during the preameloblast stage and continues through the secretory and early maturation stages of ameloblast life cycle¹. Our in vivo data indicate that at P1, P3 and P6 when Msx2 and Enam are co-expresssed, Enam requires Msx2 for its expression. At P9, when the ameloblasts are at the maturation stage and Msx2 is no longer expressed, the Msx2 mutant ameloblasts continue to express Enam (Fig. 1). In addition to our in vivo data, our gain of function, loss of function and time dependent assay further confirm that Enam requires Msx2 for its expression. Our results are also consistent with RT-qPCR analysis that revealed reduced expression of both amelogenin and Enam in Msx2^−/− mouse dental epithelium ¹⁸. Moreover, our characterization of Enam promoter for Msx2 binding sites revealed three putative Msx2 binding sites and our ChIP experiments provided evidence that Msx2 binds to Enam promoter, suggesting that Enam is a target for Msx2. Based on the above, we could potentially draw the conclusion that Msx2 may promote Enam expression directly acting, as an activator of Enam expression. Our luciferase experiments, on the other hand, indicate that Msx2 alone represses Enam expression. How results apparently contradictory may find an explanation?

Msx2 partially regulates Enam potentially in concert with other repressors and activators.

Msx2 transcription factor is known to act as, both, a repressor ^{33, 34} and activator ³⁵ and, like most transcription factors, Msx2 does not act alone but rather in concert with other transcription factors to regulate the final dosage and the onset of expression of downstream genes ^{2, 21, 25}. In that context, Enam’s expression during the presecretory and secretory stages may be partially regulated by Msx2, in concert with additional repressors and/or activators. These transcription factors may interact with Msx2 physically and/or in vivo via a protein-protein interaction mechanism to control Enam expression level, like what it has been shown for another secreted protein, the amelogenin. Msx2 is shown to interact with C/EBPα to repress the promoter activity of amelogenin-promoter reporter constructs independent of its intrinsic DNA binding activity. In transient co-transfection assays, Msx2 and C/EBPα antagonize each other in regulating the expression of the mouse amelogenin gene ²¹.

For Enam we know that co-transfection analysis and mutation assays revealed two conserved LEF1 responsive elements located at −1002 and −597bp upstream of the Enam translation initiation site that could augment transcriptional activity of the Enam, suggesting that the beta-Catenin/LEF1 is a key transcriptional complex regulating transcriptional activity of the Enam ³⁶. Runt-related transcription factor 2 (Runx2) is also involved in amelogenesis. In the Runx2 conditional knockout (cKO) mouse, qRT-PCR analysis revealed that the expression of Enam was increased suggesting that Runx2 may act as a repressor of Enam gene ^{37, 38}. These results were further confirmed by in vitro studies showing that Enam expression levels were downregulated in Runx2 over-expressing cells ³⁹. Athanassiou-Papafthymiou and colleagues also showed that Enam expression levels were subject of Dlx3 transcription factor regulation. Enam expression was up-regulated in Dlx3 over-expressing cells ^{39, 40}, whereas knockdown of Dlx3 down-regulated its expression ⁴¹. More importantly, chIP and luciferase assays have shown that DLX3 transactivated Enam ⁴¹, most probaby through a potential cis-regulatory element for Enam located 5.2kb upstream of the enamelin translation inititation site. This cis-regulatory element found to be sufficient to drive endogenous Enam in ameloblast cells using transgenic mice ^{8, 40}. This indicated that DLX3 participates in the tissue-specific expression of Enam in ameloblasts. Interestingly, when we analyzed this 5.2kb region using MatInspector software, we found 3 putative Msx2 binding sites between −3900~−500bp region (Fig.4A), that Msx2 binds strongly and directly to −500bp and −2900bp regions (Fig. 4B) and that Msx2 alone represses Enam (Fig. 4C).

Considering the forementioned studies and in the context of our findings, we propose that except for Msx2, other transcription factors, such as Lef1, Runx2 or Dlx3, may keep on regulating and fine-tuning Enam’s expression level during the presecretory and secretory stages ^36–42. It is also quite possible that Msx2 may recruit unknown or uncharacterized factors, that may work in concert as transcriptional repressors or activators in a time and context dependent manner to regulate onset and right dose of Enam gene expression. Indeed, in addition to Msx2, we have identified multiple binding sites for several other transcription factors like Sox9, Dlx1, Isl1, Lhx6, Pax6, Nfy, and Sp3 in proximity with the Msx2 binding sites in the Enam’s regulatory region (data not shown). This suggests that besides Msx2, Dlx3 and Runx2 binding sites, the regulatory regions of the Enam promoter may also contain several binding sites for other transcription factors that could act as negative or positive regulators of its overall expression.

In sum, it is obvious that Msx2 is playing an important role in regulating amelogenesis but not alone, but rather in concert with other transcription factors. Although additional studies will help to better understand the relationship between Enam and Msx2, it seems that the right dose of enamelin is essential, and it is critical for amelogenesis in general as it was further demonstrated in transgenic mouse lines over-expressing enamelin. Hu and colleagues have shown that by introducing enamelin transgene at various expression levels into the Enam^−/− background did not fully recover enamel formation while a medium expresser in the Enam^+/− background did”⁴³. “Thus, too much or too little enamelin is essential for ameloblast integrity and enamel formation” ^{18, 43}. If Enam acts in a dose-dependent manner, our data indicates that its biological function is dictated by a network of transcription factors including Msx2 that fine-tunes its optimum dosage and onset of full expression in a spatial and temporal manner.

EXPERIMENTAL PROCEDURES

Mice and genotyping

All animal studies and experimental procedures were conducted in accordance to the guidelines for the care and use of laboratory animals by the Massahusetts General Hospital, Boston, MA and the Forsyth Institute, Cambridge, MA. Postnatal pups (P1, P3, P6 and P9) were collected from matings of Msx2 heterozygous animals maintained in BALB/c background. Genotyping was performed as previously described ^{17, 19}. Age matched wildtype pups and/or embryos served as the appropriate controls.

In situ hybridization (ISH)

Postnatal animals (P1, P3, P6, P9) were collected and heads decapitated for making coronal and sagittal sections. P1 and P3 samples were immediately fixed in freshly made 4% paraformadehyde while P6 and P9 samples were decalcified in 12.5% EDTA+2.5% PFA/PBS-DEPC at 4°C for 1 week. All samples were then dehydrated through graded ethanol series, embedded in paraffin, sectioned at 8µm and processed for ISH, as previously described ^{17, 19}. Sense (5’CCAGACTTCCTGCCTCAAAG 3’) and antisense primers (5’AGGACTTTCAGTGGGTGTGG 3’) were used to synthesize the enamelin probe in a PCR reaction. T7 primer sequence sites were added to the antisense sequence to generate the antisense probe by PCR method. The PCR products were gel purified (Qiagen Inc, Valencia, CA), labeled with DIG-UTP (Roche Biochemica, Mannhein, Germany) and used directly for hybridization. The sense probe was used as a negative control.

Cell culture

Two different dental epithelial cell lines were used in the present study – the rat dental epithelial cell line (G5) and the mouse dental epithelial cell line (LS8), shown ¹⁹. Both cell lines were maintained in high-glucose Dulbecco’s modified Eagle’s medium (Gibco, Grand Island, NY, USA), containing 100 U/ml penicillin, 100 mg/ml streptomycin, and 10% fetal bovine serum (Gibco, Grand Island, NY, USA) at 37 °C in 5% CO2 humidified atmosphere following the standard protocols ²⁷.

Gain-of-Function and Loss-of-Function Experiments

For overexpression of Msx2, LS8 cells were transfected with pCMVtag2-Flag-Msx2, and then cultured for 48h-72h following which total RNA was isolated from the cells using Trizol (Qiagen, MD, USA). An empty vector (pCMVtag2) served as a negative control for gain-of-function studies. For loss of function of Msx2, commercially available small interfering RNA for Msx2 (Msx2-siRNA) was used (Santa Cruz, CA, USA). We used the following oligonucliotides, sense sequence 5’CAGCUCUCUGAACCUUAC 3’ (sc-43947). As negative control we used a scramble sequence that will not lead to the specific degradation of any known cellular mRNA: sense scramble control 5’UUCUCCGAACGUGUCACG 3’ (sc-37007). Overexpression and gene knockdown studies were performed following the protocol as described ¹⁹.

Time dependent assay

LS8 cells were transfected with pCMVtag2-Flag-Msx2 (Invitrogen, USA) and then cultured for up to 48h. The cells were harvested at different time points (4h, 8h, 16h, 24h and 48h) for RNA isolation and subjected to real time qPCR analysis to check for time-dependent expression of Msx2 and enamelin. These experiments have not been performed in the presence of the protein synthesis inhibitor, cycloheximide and thus, we do not know whether the secondary response of Enam gene require de novo protein synthesis for transcription.

Real-time quantitative PCR

Total RNA from LS8 cells and mice P1, P3, P6 and P9 molar tooth germs (M1) was extracted with Trizol according to manufacturer’s instructions and reverse transcription was performed using qScript cDNA synthesis kit (Quanta Biosciences, Gaithersburg, MD). Quantitative PCR was carried out in LightCycler and LightCycler-Faststart DNA Master SYBR Green I (Roche Diagnostics, Switzerland). The expression level of each sample was normalized to Gapdh (glyceraldehyde-3phosphate dehydrogenase) mRNA expression. The primers used are as follows:

Msx2: F 5’AGACATATGAGCCCCACCAC 3’/R 5’CAAGGCTAGAAGCTGGGATG 3’

Enamelin: F 5’ TCCAGGAAACCCAACTTACG 3’/R 5’TTTCTTCCGAAATGGACTGG 3’

GAPDH: F5’GCAAAGTGGAGATTGTTGCCAT3’/R5’CCTTGACTGTGCCGTTGAATTT3’

Reporter Construct

The 5’-flanking region of the mouse enamelin gene was generated by PCR using AccuStart taq DNA polymerase HiFi (Quanta, MD, USA), according to the manufacturer’s instructions. Three different constructs were made using the following PCR primers:

pGL3-4578: F 5’CCCGGGCTCGAGATCTGTAACTACTACCTTTGAGGGC 3’

R 5’CCGGAATGCCAAGCTTAGAGAGAGCCAAGGAGCAAGA 3’

pGL3-500: F 5’CCCGGGCTCGAGATCTCCTAACAACGAAGCTACATCTG 3’

R 5’CCGGAATGCCAAGCTTTTATTACCATCAACCATACCCTTA 3’

pGL3-2000: F 5’CCCGGGCTCGAGATCTTATGTCAATGTAAACAGTGTTATGC 3’

pGL3-2900: F 5’ CCCGGGCTCGAGATCTGGTCCCAGACTAAGAAGGCT 3’

The reverse primer was common for the pGL3-500, pGL3-2000 and pGL3-2900 constructs.

The amplified products were extracted and purified with NucleoSpin Extract II and cloned into pGL3-Basic vector (Promega, Madison, WI) using Fusion HD Cloning kit (Clontech, Mountain View, CA, USA). All constructs were confirmed by DNA sequencing.

Luciferase Assay

The enamelin-luciferase reporter plasmids were constructed using Enamelin 5’flanking regions encompassing the putative Msx2 binding sites (as predicted from UCSC Genome browser and MATINSPECTOR), and cloned into pGL3 basic luciferase vector (Promega, USA). The Reporter vector was transfected into LS8 cells together with pCMV-FLAG-Msx2 or vector only (negative control) and phRLTK (as normalizing internal control) with Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) following the manufacturer’s instructions for 48h, following which the cells were harvested, the luciferase activity recorded using Promega kit, after normalizing with firefly/renilla luciferase activity. The data was obtained from three independent experiments, and each experiment was done in triplicates.

Chromatin Immunoprecipitation

Chromatin immunoprecipitation (chIP) was performed using the EZ-Magna chip kit (Millipore, Billerica, USA) according to the manufacturer’s instructions. Forty-eight hours after transfection with pCMV-FLAG-Msx2 expression plasmid, LS8 cells and P3 wild type mouse M1 molar tooth germs were fixed and crosslinked with 1% (v/v) formaldehyde at 37°C for 10 min. Crosslinking was stopped by adding glycine to a final concentration of 125mM, followed by washing with cold PBS. After sonication chromatin was incubated with magnetic beads conjugated to either 1µg of monoclonal anti-Flag antibody (F3165, Sigma) or 1µg of normal mouse IgG (EZ-Magna chip kit) or anti-Msx2. Immunoprecipitated chromatin was reverse crosslinked and washed before DNA extraction. Finally, the immunoprecipitated DNA and the corresponding non-immunoprecipitated DNA (input) was subjected to PCR using the following enamelin primers:

F1: 5’ TTGGCCAGCTCCTCTAAAAG 3’/ R1 5’ CACTGGCCACCATCAAAAG 3’

F2: 5’ TATGCTCACTACTCAATTAC 3’/ R2 5’ CGTAGTTCCAAAGTTTAGTG 3’

F3: 5’ GGGAGGCAAGTGGATATTT 3’/ R3 5’ CGGACGTGACTTTTCTCCAT 3’

Control-F: 5’TCCATTCCCTGGTATCCTGA 3’/ R 5’ CCAAAATTCACCCATCCATT 3’

In silico analysis of promoter binding sites

UCSC MatInspector software was used to predict the putative promoter binding regions for Msx2. Primers were designed from these predicted regions using Primer 3 database for chIP followed by PCR amplification.

Imaging

The imaging for ISH was done using Olympus microscope.

Statistics

Each cell culture experiment was replicated 3 times. For ISH, a minimum of 3–4 mice pups were used. Statistics was done using one-way ANOVA or students t-tailed test, wherever applicable using GraphPad prism (version 7, CA). P value of <0.05 was considered statistically significant.

Data Availability Statement:

All datasets generated for this study are included in the article.