NFI-C2 Temporal-spatial Expression and Cellular Localization Pattern during Tooth Formation

Abstract

Currently, little is known regarding critical signaling pathways during later stages of tooth development, especially those associated with root formation. Nfi-c null mice, lacking molar roots, have implicated the transcription factor NFI-C as having an essential role in root development. Previously, we identified three NFI-C isoforms expressed in dental tissues with NFI-C2 being the major transcript. However, the expression pattern of the NFI-C2 protein is not characterized. In this study we performed in situ hybridization and immunohistochemistry using isoform specific probes. We show the production of a NFI-C2 peptide antibody, its characterization, the temporal-spatial expression pattern of the NFI-C2 protein during odontogenesis and sub-cellular localization in dental cells. Moderate NFI-C2 staining, as early as bud stage, was detected mostly in the condensing dental ectomesenchyme. This staining intensified within the dental pulp at later stages culminating in high expression in the dentin producing odontoblasts. The dental epithelium showed slight staining until cytodifferentiation of enamel organ into ameloblasts and stratum intermedium. During root formation NFI-C2 expression was high in the Hertwig’s epithelial root sheath and later was found in the fully developed root and its supporting tissues. NFI-C2 cellular staining was cytosolic, associated with the Golgi, and nuclear. This data suggests a broader role for NFI-C during tooth formation than limited to root and periodontal ligament development.

Graphical abstract

Introduction

Odontogenesis is the complex process of tooth development, which results from the interplay of signaling cascades between the ectoderm of the first branchial arch and the cranial neural crest -derived ectomesenchyme. A continued progression of epithelial-mesenchymal interactions also dictates the process of root development following completion of the tooth crown structure. A bilayer extension of the inner and outer enamel organ epithelium, known as Hertwig’s epithelial root sheath (HERS), continues to grow apically at the cervical loop between the pulp and dental follicle and induces radicular or root dentin formation as well as the future periodontium (Ten Cate, 1998).

Although the morphological events associated with root formation have been extensively studied, our understanding of the biological mechanism involved is just emerging. Several signaling pathways have been implicated in regulation of tooth development, including transforming growth factor beta/bone morphogenic protein (TGFβ/BMP), sonic hedgehog (SHH), fibroblast growth factor (FGF), Wnt, as well as several transcription factors such as Msx, Dlx, Runx2 and Sox (Vainio et al., 1993; Thomadakis et al., 1999; Wilkinson et al., 1989; Kettunen et al., 2000; Gritli-Linde et al., 2002; Tummers & Thesleff, 2009; Sarkar et al., 2000; Chen et al., 1996; Stock et al., 1996). However, their role during specific events of root formation has yet to be well characterized.

Previous studies have identified a member of the nuclear factor I (NFI) family of genes that is critical for normal root formation in mice (Steele-Perkins et al., 2003). NFI genes function as transcription factors and adenovirus DNA replication factors (Mermod et al., 1989; Gounari et al., 1990). This gene family contains four members, NFI-A, NFI-B, NFI-C and NFI-X, all of which exist as multiple isoforms. NFI proteins share a highly conserved N-terminal domain that plays a role in DNA binding, dimerization and adenovirus replication. The C-terminal domain of NFI proteins is believed to act as a transcript modulator domain (Kruse & Sippel, 1994; Chaudhry et al., 1997) and shows no homology among the four gene family members, but is highly conserved across species for each member. Each member has distinct but overlapping spatial-temporal expression patterns in different tissues during mouse embryonic and postnatal development. Analyses of individual NFI gene knockout mice demonstrate the distinct and vital role of these proteins in normal development (das Neves et al., 2001; Grunder et al., 2002; Steele-Perkins et al., 2003; Driller et al., 2007).

NFI-C gene has been shown to be essential for normal tooth root formation, since Nfic^−/− mice lack molar roots (Steele-Perkins et al., 2003). A similar phenotype (missing roots) is observed in patients with a condition called Radicular Dentin Dysplasia (RDD; OMIM 125400, also known as Rootless Teeth or Dentin Dysplasia Type I). This rare inherited autosomal dominant disorder is characterized by incomplete to absent root formation resulting in premature exfoliation of primary and permanent dentitions (Bixler, 1976; Wesley et al., 1976; Witkop, 1975).

NFIC, located on human chromosome 19p13.3, is encoded by 12 exons that are alternatively spliced producing three protein products: NFI-C1, NFI-C2, and NFI-C4 (Qian et al., 1995; Lamani et al., 2009). The mouse NFI-C homology is mapped to chromosome 10 and similar C-terminus variants to the human isoforms can be found in Ensembl (Flicek et al., 2011). We have previously shown the NFI-C2 transcript to be the most abundant with high expression in both dental and non-dental tissues, while NFI-C1 mRNA is present at very low quantities and NFI-C4 expression was only moderate (Lamani et al., 2009). However, how this translates to the NFI-C protein levels during the various stages of tooth formation is yet to be determined.

There is conflicting evidence regarding NFIC expression in specific epithelial and mesenchymal cell populations at significant stages for tooth root formation (Steele-Perkins et al., 2003; Chen et al., 2014). Defining spatial and temporal distribution of NFI-C in developing tooth tissues is critical for understanding its role in root formation and identification of its molecular networks involved in radicular dentin and periodontal ligament formation. We have generated and characterized an antibody recognizing specifically the NFI-C2 isoform, the major isoform of NFI-C. In this paper, we investigate the protein expression profile of NFI-C2 during mouse tooth formation, cytodifferentiation and development of the tooth crown and root. Furthermore, we map the cellular localization of the NFI-C proteins in primary human dental derived cells.

Material and Methods

Tissue Samples and Primary Dental Cell Cultures

All studies were carried out under the approval of the animal ethical committee for the University of Alabama at Birmingham and the University of Texas Health Science Center at San Antonio. Mouse maxillary and mandibular tissues embryonic day (E) 13.5, 16 and 18 and postnatal day (PN) 1, 6, 13 and 30 were fixed and demineralized (if older than 6 days PN) in 10% EDTA. These tissues were processed and embedded in paraffin, sectioned (4–6 μm) and mounted on silane-treated slides for immunohistochemistry and in situ hybridization using standard methods (Chen et al., 2002; Gluhak-Heinrich et al., 2010). Tissue samples for in situ hybridization were processed under RNase free conditions with DEPC water in all reagents.

Human primary dental cell populations were established from extracted teeth (obtained with informed consents) using previously published techniques (MacDougall et al., 1996; Chen et al., 2005). Primary cell populations of enamel organ epithelium (EOE), dental pulp (DP), dental follicle (DF) and periodontal ligament (PDL) were grown from these cultures as previously described (MacDougall et al., 1996; Chen et al., 2005, Gay et al., 2007; Lamani et al., 2009; Borovjagin et al., 2011).

NFI-C2 Peptide Antibody Production

Polyclonal anti-NFI-C2 serum was generated by injecting rabbits with the synthesized peptide LRPTRPLQTVPLWD representing the last 14 common amino acids (AA) between mouse and human NFI-C2 (417–430AA human accession EAW69325.1 and 426–439 mouse accession NP_032714.1), coupled to keyhole limpet hemocyanin (KLH) via a cysteine added to the N-terminus (Alpha Diagnostic International, San Antonio, TX). BLASTP program was used to search the protein databases for potential identity with other mouse or human proteins. Affinity-purified antibody was generated by absorbing sera to the specific NFI-C2 immunogen coupled to cyanogen bromide-activated agarose matrix (Alpha Diagnostic International, San Antonio, TX).

Fluorophore-linked Immunobsorbent Assay (FLISA)

Antibody specificity was analyzed by peptide neutralization and fluorophore-linked immunobsorbent assay (FLIZA). NFI-C2 antibody dilutions (1:5K and 1:10K, 1ml each) were incubated overnight at 4 °C with 1μg, 10μg, 20μg or 50μg of the NFI-C2 peptide (Alpha Diagnostic International, San Antonio, TX), centrifuged to pellet any immune complexes, and the supernatant was used for FLISA. Flat-bottom 96-well PVC microtiter plates were coated with 1μg of the NFI-C2 peptide or with a non-specific protein (Sigma-Aldrich, St. Louis, MO) as a negative control, and probed with various NFI-C2 antibody dilutions (1:100–1:100K) or peptide neutralized dilutions (1:5K and 1:10K). Anti-rabbit IRDye 800 (1:20K) was used to visualize the signal on an Odyssey Infrared Imaging System (LI-COR Biosciences, Lincoln, NE).

Western Blot Analysis

For western blot analysis, a confluent dental pulp cell culture (100mm plate) was lysed in RIPA lysis buffer containing protease inhibitors (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) and the total protein concentration determined using a Pierce BCA Protein Assay kit (Pierce Biotechnology, Rockford, IL). The protein lysate (20 μg) was run on a 10% SDS-polyacrylamide gel and transferred onto a PVDF membrane. The membrane was probed with either the NFI-C2 (1:500) or mouse monoclonal NFI-C4 (1:500, Abnova, Taiwan) antibody, followed by anti-rabbit IRDye 800 (1:20K) or anti-mouse IRDye 680 (1:20K), respectively. The signal was visualized using the Odyssey Infrared Imaging System.

Immunohistochemistry and Immunocytochemistry

Paraffin sections (10μm) of mouse tissues containing upper and lower jaws or just mandibles (E13.5, 16 and 18, and PN1, 6, 13 and 30) were deparaffinized, and rehydrated using standard methods (Chen et al., 2002). All tissues were permeabilized with 0.25% Triton X-100 and treated with Proteinase K (20μg/mL) for antigen retrieval. Peroxidase anti-peroxidase immunohistochemistry was performed with NFI-C2 (1:200) or NFI-C4 (1:100) antibody, followed by visualization using the SuperPicTure polymer detection kit (Invitrogen, Carlsbad, CA) and photographed under bright-field illumination (Nikon Instruments Inc., Melville, NY). Normal rabbit serum or no primary antibody was used as negative controls.

For immunocytochemistry, human primary dental cells were plated on 4-chamber glass slides (Lab Tek II, Nalge Nunc Int.) and allowed to grow for 3 days and fixed with 4% formaldehyde for 10 min and permeabilize with 0.025% Triton X-100 for another 10 min. Cells were stained with either the NFI-C2 (1:200) or NFI-C4 (1:100) antibody. An anti-rabbit or anti-mouse Alexa Fluor 488 (1:1000, Invitrogen, Carlsbad, CA), respectively, was used as the florescent probe to visualize the signal on a Nikon 90i microscope (Nikon Instruments Inc., Melville, NY). Dental pulp cells were counterstained with Golgin-97 antibody (1:500, Invitrogen, Carlsbad, CA) and labeled with anti-mouse Alexa Fluor 594 (1:1000, Invitrogen, Carlsbad, CA). Cells were incubated overnight at 4°C with the primary antibodies and for 1hr at room temperature with the fluorescent secondary antibodies.

In situ hybridization

Isoform specific PCR constructs for NFI-C2 and NFI-C4 were amplified from specific spliced junctions and a construct common to all transcripts was amplified from exon 6 to 8 region of NFI-C (Figure 1). All RNA probes were transcribed in vitro in the presence of Digoxigenin for production of antisense and sense probes with T3, T7 or SP6 polymerases, where appropriate. Sections were deparaffinized, treated with Proteinase K and hybridized with 1 μg/ml RNA probe at 58 °C overnight. Sections were then treated with RNase, washed with 5XSSC and with 50% formamide in 2XSSC. Alkaline phosphatase substrate (NTB/BCIP, Roche Diagnostic Corp, Dallas, TX) was used for detection of hybridization signal. The duration of hybridization signal development was from 1 h to overnight at 30 °C, depending on the probe and abundance of the transcript in the tissue. The sections were counterstained with methyl green.

Figure 1.

Stragety for the construction of cDNA templates for NFI-C mRNA probes. (A) Diagram of the NFI-C gene and its alternatively spliced transcripts (primers represented by arrows and untranslated region shown in red). (B) Primers designed for generating specific NFI-C probes.

Results

Production and Validation of NFI-C2 Antibody

A polyclonal antibody was generated in rabbits against a common COOH-terminal peptide of mouse and human NFI-C2 protein. The uniqueness of the peptide to NFI-C2 was confirmed using BLASTP program. The antibody was affinity purified and specificity confirmed by FLISA when probed against antigenic and control peptides (Figure 2). The specificity was further confirmed by peptide neutralization studies. Our data shows a decrease in NFI-C2 signal absorbance when using the neutralized affinity purified antibody (Table 1). This decrease ranges from 2.8 to 4.2 folds compared to the low (1:5K) or high (1:10K) antibody dilution and supports the data for the specificity of NFI-C2 antibody generated.

Figure 2.

Characterization of the NFI-C2 antibody by FLISA. The plate was coated with the NFI-C2 antigenic peptide (top row) or a non-specific protein (bottom row). Primary antibody dilutions as shown (1:100 – 1:100K) with the secondary antibody used alone as the control.

Table 1.

Peptide competition for NFI-C2 antibody

Antibody Dilutions	Absorbance at 800 nm
	Affinity pure NFI-C2	Neutralized Affinity pure NFI-C2
1:5K	15.87	5.48
1:10K	9.07	2.14

The NFI-C2 peptide antibody produced was compared with a commercially available mouse monoclonal antibody generated against the full-length human recombinant NFI-C4 protein. Both NFI-C proteins are expected to be approximately 48kDa. Western blot analysis of proteins extracted from human dental cells revealed a single NFI-C2 protein band running at about 52kDa (Figure 3A). The NFI-C4 protein recognized by the commercially available NFI-C4 antibody (Figure 3A) ran at a slightly low molecular weight (50kDa). Finally, the specificity was tested by immunohistochemistry using mouse 1PN tissues containing skeletal muscle, a known high NFI-C expressing tissue as previously demonstrated and documented by GeneCard (Chaudhry et al., 1997; Safran et al., 2003; Shmueli et al., 2003; Yanai et al., 2005;). The polyclonal NFI-C2 antibody detected a strong protein staining in the skeletal muscle (Figure 3B) in contrast to the monoclonal NFI-C4 antibody that showed no signal above the negative control (Figure 3C and D) in any tissue tested; this included testing on cryosections and using various antigen retrieval protocols. To ensure NFI-C transcripts were expressed in developing roots we performed in situ hybridization of 10PN mouse tissues with a probe common to both transcripts (NFI-C) and two probes that were specific to the C-2 and C-4 isoforms (Figure 4). We confirmed expression of both NFI-C2 and C4 transcripts and found it to be comparable to the staining with the common NFI-C probe (Figure 4). However, since we were unable to utilize the NFI-C4 antibody, in our further studies aimed to establish the temporal-spatial expression pattern of NFI-C protein during tooth formation we used only the NFI-C2 antibody.

Figure 3.

Specificity of the NFI-C2 antibody by Western blot analysis and immunohistochemistry. Western blot showing single band (~55 kDa) in a human dental pulp total protein lysate (A). A commercial antibody to NFI-C4 serves as a positive control. Immunohistochemistry of 1PN mouse tissue (B) showing staining of the positive control tissue skeletal muscle (SKM). Bar=100μm

Figure 4.

Expression of NFI-C gene during tooth root formation. In situ hybridization with the NFI-C (A, D, G), NFI-C2 (B, E, H), and NFI-C4 (C, F, I) mRNA probes in 10PN mouse mandibular molars. Dentin (D), odontoblasts (Od), dental pulp mesenchyme (DPM).

NFI-C2 Expression Pattern during Tooth Formation

To determine the temporal-spatial pattern of NFI-C2 during odontogenesis, mouse dental tissues at critical stages of tooth formation were analyzed by immunohistochemistry. At E13.5, NFI-C2 staining is seen in the mandibular and maxillary first molars (M1) at the late bud stage (Figure 5 A and B). The second maxillary molar (arrowhead) at the initiation stage is also stained. In all molars NFI-C2 is localized to both the tooth epithelial and condensing ectomesenchymal tissues. However, the relative expression level in the dental ectomesenchyme is more intense. Strong staining is also seen in the condensing mesenchyme that will give raise to mandibular bone (asterisk), in tissue surrounding the Meckel’s cartilage (MC) and within the tongue (T). No staining is observed in the control E13.5 tissue section (Figure 5C). At the late cap stage (E16) and bell (E18) NFI-C2 staining is much more intense and restricted (Figure 5D-I). Strong staining is seen in the dental pulp mesenchyme (DPM) at these developmental stages. At E16 a gradient of NFI-C2 expression in the DPM is observed with more intense signal in the DPM proximal to the overlaying inner enamel epithelium (IEE) (Figure 5D–F). This gradient of NFI-C2 expression in DPM becomes more focused at E18, when strong NFI-C2 expression becomes restricted to the interface with IEE while the remaining DPM demonstrates weaker staining (Figure 5H). The region of strong NFI-C2 expression corresponds to the layer of cells that will form the specialized dentin producing odontoblasts. Staining within the epithelial IEE, outer enamel epithelium (OEE) and stellate reticulum (SR) remained low (Figure 5F and I) with increased staining seen in the surrounding dental follicle (Figure 5I, black arrow). At these stages intense NFI-C2 staining is seen in the tongue (Figure 5D), an oral tissue known to express NFI-C (Su et al., 2004). Interesting, light staining is also seen within the SR associated with an enamel cord (Figure 5G, open arrowhead).

Figure 5.

NFI-C2 expression in the early stages of mouse developing teeth. Molar teeth at E13.5 (A & B), E16 (D, E & F), E18 (G, H & I) and 1 day PN (J, K & L) showing localization of the protein within both mesenchymal and epithelial components. No primary antibody control of E13.5 (C). Abreviations and symbols: first molar (M1), tongue (T), Meckel’s cartilage (MC), condensing bone mesenchyme (asterisk), second molar (arrowhead), condensing mesenchyme (CM), dental epithelium (DE), inner enamel epithelium (IEE), outer enamel epithelium (OEE), stellate reticulum (SR), dental pulp mesenchyme (DPM), enamel cord (open arrowhead), dental follicle (arrow), stratum intermedium (SI), ameloblasts (Am) and odontoblasts (OD). Bar=100μm

At the late bell/crown stage of development (PN1), gradual focusing of the NFI-C2 localization is seen associated with the progression of dentinogenesis associated with the cytodifferentation of odontoblasts (OD). The NFI-C2 staining is more intense in the single layer of OD that lines the dental pulp (Figure 5L). While the expression of NFI-C is maintained in DPM, its intensity is decreased as compared to early stages of tooth formation (Figure 5J–L compared to D–I). The staining detected in the IEE intensifies with the cytodifferentiation of ameloblasts (Am) and stratrum intermedium (SI) associated with the process of amelogenesis (Figure 5L).

At initiation of the HERS (PN6), intense staining for NFI-C2 is evident in both epithelial and mesenchymal components (Figure 6A–C). A positive signal is detected in the secretory ameloblasts, and NFI-C2 continues to be sustained within the odontoblast layer. Furthermore, high NFI-C2 levels are seen in the labial enamel forming ameloblasts of the forming incisor (Figure 6A). At this stage NFI-C2 signal is intense within the condensing dental pulp particularly at the leading front of the developing root an area rich in dental stem cells (Figure 6C, arrow). As the molar root extends into the underlying mesenchyme (PN13), NFI-C2 is expressed in seen in all the radicular associated tissues including OD, dental pulp (DP) as well as dental follicle (DF) (Figure 6D–F). NFI-C2 is also now detected in the newly forming root supporting structures of the periodontal ligament (PDL) (Figure 6F). This pattern continues to the completion of the root by PN30 (Figure 6G–I). Upon completion of crown and root formation, NFI-C2 expression remains robust in the dental pulp with intensity highest in the OD lining the pulp chamber. An intense signal is also maintained at the root apex zone enriched for dental stem cells. NFI-C2 moderate staining is also found in PDL fibroblasts and in acellular and cellular cementoblasts at the root apex (Figure 6I, arrowhead). Positive staining is also seen in surrounding supportive tissues including osteoblasts lining the alveolar bone (Av B) as well as in osteocytes (OC) (Figure 6H, arrow).

Figure 6.

NFI-C2 localization during root formation at initiation of the HERS seen at 6 PN (A, B & C) and at 13 PN (D, E & F) and in the fully developed roots at 30 PN (G, H & I). Abbreviations and symbols: molar (M), incisor (I), ameloblasts (Am), odontoblasts (OD), dental pulp (DP), HERS (arrow), dental follicle (DF) and periodontal ligament (PDL), demineralized enamel (asterisk), dentin (D), alveolar bone (Av B), cementoblasts (arrowhead), and osteocytes (OC). Arrow points to osteocytes within the alveolar bone. Bar=100μm

In rodents, unlike humans, the incisors are continually erupting and have an enamel matrix only on the labial side (Thesleff & Tummers, 2009). The incisors have highly proliferative apical regions, which are known as the epithelium stem cell niches. Our study shows that these regions express NFI-C2 (Figure 7A–C). Furthermore, the staining is more pronounced within the labial ameloblasts, which make enamel, when compared with the lingual ameloblasts that do not produce enamel (Figure 7A and B). The NFI-C2 signal was also seen at high levels in the OD and within the DPM of developing mouse incisor (PN1).

Figure 7.

NFI-C2 localization in the incisor epithelium stem cell niches at 1PN seen at low (A), medium (B) and higher (C) magnification. Abbreviations: molar (M), incisor (I), labial (La), lingual (Li), ameloblasts (Am), odontoblasts (OD), dental pulp mesenchyme (DPM), labial epithelium stem cell niche (asterisk) and lingual epithelium stem cell niche (arrowhead). Bar=100μm

Cellular Localization of NFI-C Proteins

We examined four types of primary human dental cells (EOE, DP, DF and PDL) by immunocytochemistry using the NFI-C2 and NFI-C4 antibodies. Figure 8 demonstrates that all dental cell types tested express both proteins. However, each isoform exhibits individual pattern of nuclear and cytosolic localization. The NFI-C2 antibody generated a broader staining profile with protein detection in the nucleus as well as within the cytoplasm (Figure 8A, C, E and G). However, the NFI-C2 staining intensity varied between the cell types. The EOE cells display stronger signal in the cytoplasm compared to the nuclear localization (Figure 8A). In mesenchymal-derived DP and DF cells, on the other hand, the difference in the protein expression between these two cellular compartments is less pronounced (Figure 8C and E). These cells express lower levels of cytosolic protein; however this is comparable to the lower nuclear protein levels. Examination of the PDL cells reveals a mixture of cells, some with NFI-C2 expression similar to the DP/DF cells (data not shown) and others similar to the pattern observed in EOE cells with more intense cytosolic than nuclear expression (Figure 8G). In contrast, in all cell types tested the NFI-C4 antibody only detected protein in the nucleus with negative staining of the nucleolus (Figure 8B, D, F and H). Furthermore, the NFI-C4 expression levels appeared similar in all dental cell type tested.

Figure 8.

NFI-C2 and NFI-C4 intracellular localization within human primary dental cells. Enamel organ epithelial (A & B), dental pulp (C & D), dental follicle (E & F) and periodontal ligament (G & H) cells show identical patterns of staining. Note: NFI-C4 is detected only in the nucleus in contrast to NFI-C2 seen also in the cytoplasm and the intensity of the cytoplasmic staining varies in the cell types tested. Bar=10μm

Based on the condensed unique NFI-C2 staining pattern in the cytoplasm, we investigated the possibility of this protein localizing within the Golgi apparatus (Figure 9). Using an antibody against golgin-97 protein, a marker for the Golgi complex (Figure 9B), we confirmed NFI-C2 presence in this organelle (Figure 9A) by its colocalization with the golgin-97 protein (Figure 9C).

Figure 9.

Immunocytochemistry showing localization of NFI-C2 within the Golgi apparatus in human dental pulp primary cells. Panel A is NFI-C2 staining using Alex Fluor 488 (green) and panel B shows staining with anti-Golgin-97 antibody using Alex Fluor594 (red). Panel C is the merged image showing co-localization (yellow) of the two proteins. Bar=10μm

Discussion

We have previously shown that C2 and C4 are the major NFI-C transcripts expressed in human dental and non-dental tissues (Lamani et al., 2009). Analysis of the genomic organization of mouse Nfi-c also revealed two reported variants, similar to the human orthologs (Flicek et al., 2011). From our earlier studies, we identified C2 as the major NFI-C protein in human dental cells. Furthermore, this isoform has been reported to be the major NFI-C protein in mouse mammary glands (Kannius-Janson et al., 2002).

To address the question of where this protein is localized during odontogenesis, we generated a polyclonal peptide antibody that recognizes the unique COOH terminus of the NFI-C isoform. We characterized this antibody by FLISA, Western blot analysis and immunostaining. Antibody specificity was also established by peptide neutralization studies. This comprehensive approach demonstrated specificity of the generated antibody to mouse and human NFI-C2 isoform with no detectable cross-reactivity with other NFI-C family members. In order to compare this isoform expression to the NFI-C4 protein, we used a commercially available monoclonal antibody raised against the full-length human recombinant protein. However, while NFI-C4 antibody was able to detect denatured protein in Western blots and human protein in immunocytochemistry, we were not able to obtain any detectable positive reaction in fixed mouse tissues. This was even the case using various antigen-unmasking protocols (data not shown). Hence, we continued our immunohistochemistry studies with the NFI-C2 antibody only.

Prior studies have analyzed mRNA expression of the four NFI family members in whole embryos (Chaudhry et al., 1997) as well as at limited stages of the developing tooth (Steele-Perkins et al., 2003; Chen at al., 2014). Chen at al. also looked at NFIC protein expression during late molar development in rats. However, their data contradict earlier studies and is not clear regarding NFIC expression during epithelial-mesenchymal interactions during early root formation (Chen at al., 2014). In our studies, using protein detection approach we confirmed the localization of NFI-C2 in the skeletal muscle and tongue. In addition, our study provides for the first time the detailed characterization of the NFI-C2 expression pattern in association with specific major events of tooth development from bud stage to root completion in the adult dentition. Although, Chaudhry et al. studied mouse embryos from E8.5 to E16.5, they reported detecting NFI-C in the tooth primordial at E15.5–16.5 corresponding to the late cap stage of tooth formation (Chaudhry et al., 1997). At this stage, localization was at high levels in the tooth DPM and very consistent with our data at the protein level. We began our study at the bud stage of odontogenesis, examining mouse tooth organs at the critical stages of tooth formation from early morphogenesis through maturation (completion of the crown and root).

In this study, we showed that NFI-C2 was expressed earlier in tooth development, as NFI-C2 was detectable in the condensing dental mesenchyme at bud stage. Although Chaudhry and colleagues did not comment on the NFI-C signal in the developing tooth prior to E15.5, from their data it is evident that this mRNA transcript is detectable at E14.5 within the condensing tooth ectomesenchyme of the mandibular molar visible in the sagittal whole embryo section. This mRNA expression correlates with the protein localization at the cap and early bell stages (E16 and E18, respectively). During these stages NFI-C2 expression was seen mainly in the DPM with stronger staining at the more differentiated interface in contact with the overlaying IEE. Again this correlates well with the previous mRNA in situ hybridization seen at slightly later stages E15.5–16.5 (Chaudhry et al., 1997; Steele-Perkins et al., 2003). All data implies that NFI-C2 is important for the early epithelial-mesenchymal interactions when signals from the IEE triggers odontoblast cytodifferentiation that in turn leads to differentiation of the enamel producing ameloblasts. Similar NFI-C mRNA expression in the dental pulp and odontoblasts in newborn mice by in situ hybridization again supports this role during dentinogenesis (Huang et al., 2010). However, the focus of this previous study was limited to the pattern of NFI-C expression in dental ectomesenchymal-derived tissues with no expression described in the epithelial-derived tissues. As the developing tooth progresses into late bell stage (PN1), we show increased staining becomes focused within the differentiating matrix producing cells: the enamel secreting ameloblasts and dentin forming odontoblasts. This is however in contrast to the results shown by Chen et al. (2014) during rat tooth development where no detectable NFIC staining was evident until after PN4 well after odontoblast and ameloblast formation. In the rat studies, a different mouse monoclonal antibody directed against expressed human NFIC was used. Our data indicates the NFI-C2 isoform is strongly associated with the maturation of ameloblasts and odontoblast cells and may be involved in regulation of both dentin and enamel extracellular matrix production.

As odontogenesis progresses to the early stage of root development (PN6/PN9), we found NFI-C2 continues to be localized in the tooth matrix producing cells, but with high expression also seen within the HERS, supporting ours and earlier in situ hybridization data (Steele-Perkins et al., 2003). Interestingly, in rat teeth no NFIC was detectable in the HERS (Chen et al., 2014). This may be due to the specificity of the monoclonal antibody used, cross reaction between species, exposed epitopes in tissues or species differences between rat and mouse expression pattern.

The HERS represents a epithelial cell bilayer formed from the reduced enamel epithelium through a fusion of the IEE and OEE, which is reported to be devoid of SR and stratum intermedium (SI) (Ten Cate, 1996). The HERS migrates apically into the underlying DP and is believed to induce and regulate root formation as well as the future periodontium (Zeichner-David et al., 2003). As the root formation continues (PN13) or is completed (PN30), we showed that the supportive tissues of the tooth (PDL, DF and alveolar bone) are also positive for NFI-C2. This suggests that NFI-C2 is important not only for initiation but also in maintaining and supporting the developed root structures. This is the first time that NFI-C has been examined in adult mouse dental tissues.

The continuously erupting mouse incisors have a lingual/labial asymmetry, also referred to as root and crown analogue, respectively. Only the labial aspects of the developing incisor produce an enamel matrix that is comparable to crown enamel formation (Beertsen & Niehof, 1986). In contrast, the lingual portion of the incisor has IEE that does not give rise to fully differentiation ameloblasts capable of enamel matrix production and resembles the molar root. Furthermore, incisors do not develop proper anchoring roots as they are subjected to constant attrition through food grinding, leading to stimulation of continuous growth. This growth is carried out by the epithelium stem cell (ESC) niches which are highly proliferative apical labial and lingual regions and as the name implies rich in stem cells (Thesleff & Tummers, 2009; Chavez et al., 2013). Other studies have shown that Nfi-c null mice develop abnormal incisors that lack growth and are highly disorganized (Steele-Perkins et al., 2003). However, there have been no other reports concerning NFI-C localization in the mouse continuously growing incisor. In our study we found that NFI-C2 was expressed in epithelial and mesenchymal tissues of developing mouse incisors including the ESC niches. Furthermore, its expression was more prominent in the labial ESC niche associated with ameloblast differentiation and enamel formation. The incisor phenotype of the Nfi-c null mouse combined with our findings of high NFI-C2 expression in the ESC niches suggests that this protein maybe required for normal and continuous incisor growth.

Based on our findings and the dental phenotype of the Nfi-c knockout mouse lacking molar roots (Steele-Perkins et al., 2003), it is reasonable to assume that NFI-C2, although part of the signaling network associated with the processes of early tooth development, is a transcription factor critical for the later events of odontogenesis associated with root development. This suggests that even though some signaling pathways for crown and root development may be similar, there are specific signaling events regulated by NFI-C2 that are unique to root formation. Deleting NFI-C may disrupt the nexus of critical signaling pathway leading to the absent root phenotype (as seen in Nfi-c null mouse). There is now emerging information related to TGF-β/BMP and SHH signaling in relation to NFI-C regulation during root development (Huang et al., 2010). These studies indicate that NFI-C is key to both TGF-β/BMP and SHH signaling pathways in the developing root epithelium. However, further research is needed to understand the role of NFI-C2 in complex molecular networks regulating tooth root formation and identification of root specific downstream targets of NFI-C2.

Finally, in this study we investigated the localization and differential expression of the two major NFI-C isoforms in primary human dental cells in vitro. As expected with all transcription factors, our data shows both NFI-C2 and -C4 are localized in the nucleus. Staining was seen within the entire nuclear structure with the exception of the nucleolus, dedicated to synthesis of rRNA and assembly of ribosomes. Interestingly, only the NFI-C2 isoform was detected in the cytoplasm associated with the Golgi complex. The stronger signal in the Golgi complex of the EOE and some of the PDL cells when compared with that of DP or DF cells suggests that this protein may have different functions in the various dental cells. Earlier studies of NFI-C2 protein in mammary glands have shown that this protein can be highly glycosylated and that the degree of glycosylation affects its function in gland development (Kane et al., 2002). Thus, the localization of NFI-C2 in Golgi apparatus of EOE and PDL may indicate that NFI-C2 undergoes post-translational modifications in these cells. These data implies that the glycosylation events are linked more to the functions of the protein in epithelial rather than mesenchymal tissues. Furthermore, our data suggests that cellular localization of individual NFI-C proteins is important in formation of the tooth root structure and its supporting tissues. Abnormal levels or distribution of these proteins may have a significant role in the root phenotype associated with RDD. Recent studies suggest that cytosolic NFI-C is important in dephosphorylation of p-Smad2/3 during late odontoblast differentiation and mineralization (Lee et. al., 2011). However, the biological significance of the NFI-C proteins is yet to be established and further studies are needed to determine their precise role in dental tissues during root formation.

In conclusion, we showed that during the early events of odontogenesis, NFI-C2 is associated with the condensing ectomesenchyme. The expression within the dental mesenchyme and the subsequently derived tissues of the odontoblasts, pulp, dental follicle and PDL persists throughout the tooth’s viable life span. However, as the tooth continues to develop, NFI-C2 is upregulated in epithelial derived tissues of ameloblasts and HERS and persists in cell populations within the PDL. These data establishes for the first time the cellular localization and the temporal-spatial expression pattern of NFI-C2 during later stages of tooth formation.