SEARCH

SEARCH BY CITATION

Keywords:

  • T-box transcription factor;
  • cleft palate;
  • nasal septum;
  • ankyloglossia

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

TBX22 belongs to the T-box family of transcription factors and was originally found in an in silico approach designed to identify new genes on the human Xq12-q21 region. Mutations in TBX22 have been reported in families with X-linked cleft palate and ankyloglossia (CPX), but the underlying pathogenetic mechanism remained unknown. We have identified mouse Tbx22 and analyzed its expression during embryogenesis by reverse transcriptase-polymerase chain reaction and in situ hybridization. In mouse embryos, it is expressed in distinct areas of the head, namely the mesenchyme of the inferior nasal septum, the posterior palatal shelf before fusion, the attachment of the tongue, and mesenchymal cells surrounding the eye anlage. The localization in the tongue frenulum perfectly correlates with the ankyloglossia phenotype in CPX. Furthermore, we identified positionally conserved binding sites for transcription factors, two of which have been implicated previously in palatogenesis (MSX1, PRX2). Developmental Dynamics 226:579–586, 2003. © 2003 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

The vertebrate face develops from four pairs of primordia, the mandibular, maxillary, median, and lateral nasal processes, which are formed by mesenchymal cells derived from the cranial neural crest (Kaufman, 1994; McGonnell et al., 1998). In the course of development, the surfaces of these processes fuse and permanent connections by means of mesenchymal bridges are established, thereby closing previous clefts. Disturbance of these steps results in persisting clefts (Moore and Persaud, 1998).

Clefts of the lip and the palate are the most frequent congenital orofacial abnormalities. Most of the cases are sporadic and thought to be due to environmental and genetic susceptibility factors (Schutte and Murray, 1999). The palate is derived from two structures: an anterior portion, the primary palate, originating from the so-called intermaxillary segment, and a posterior portion, the secondary palate (Kaufman, 1994; Kerrigan et al., 2000; Yoon et al., 2000). The latter is formed by the fusion of two lateral palatal processes, which are outgrowths of the maxillary processes. At the same time, midline fusion proceeds with the free margin of the nasal septum, which grows down from the fused medial nasal processes, as well as anterior fusion with the primary palate.

Despite the considerable importance of cleft palate malformations, few genes controlling palate development have been identified or proposed. Mutations in TBX22 have been found to cause a nonsyndromic form of cleft palate (CPX) in humans that is associated with ankyloglossia (Braybrook et al., 2001). Affected individuals have cleft palates due to failure of palatal process fusion. TBX22 belongs to the T-box family of transcription factors, which is named after a DNA-binding domain of approximately 200 amino acids, the so-called T-box (for review, see Papaioannou and Silver, 1998, or Tada and Smith, 2001).

Here, we report the identification of mouse Tbx22, and describe its genomic organization, conserved putative transcription factor binding sites, and its expression in distinct areas of the embryonic head involved in secondary palate and tongue development.

RESULTS AND DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

Database searches with a human TBX22 mRNA resulted in the identification of mouse genomic sequences, which were assembled into contigs, and used for the in silico prediction of a mouse Tbx22 mRNA. With this sequence, a mouse expressed sequence tag (EST) database was screened and matching ESTs were assembled into mouse Tbx22 cDNA. Mouse Tbx22 (MM_TBX22, AF515700) spans 8.6 kb from the beginning of the first exon to the first identified polyadenylation signal on genomic sequences of the mouse XD region, known to be syntenic with human Xq21. Eight exons were identified by flanking consensus splice signals and comparison with the human cDNA (Fig. 1). The open reading frame consists of 1554 nucleotides (Fig. 2). It encodes a 517 amino acid polypeptide with 72% identity and 76% similarity to the human TBX22 protein. The recently published chicken Tbx22 protein shows 64% identity and 70% similarity to our amino acid sequence (Haenig et al., 2002). The T-box domain spans from exon 2 to 6 and on the amino acid level is 88% identical and 92% similar to the human one. Originally, the human gene was reported to consist of 7 exons coding for a truncated T-box domain, but an additional 5′-exon was reported later (Laugier-Anfossi and Villard, 2000; Braybrook et al., 2001). Our reported mouse cDNA and the published chicken cDNA is in agreement with this finding. On a Northern blot, we detected a transcript of approximately 2.4 kb in E14 mouse snout tissue (Fig. 3).

thumbnail image

Figure 1. Genomic structure of mouse and human TBX22. Sizes of introns and exons are indicated, and the T-box is depicted by shaded boxes. The exact size of mouse 3′-untranslated region remains to be determined.

Download figure to PowerPoint

thumbnail image

Figure 2. Alignment of human and mouse TBX22 coding and amino acid sequence. Vertical bars indicate the positions of exon–intron boundaries. The T-box is depicted by a horizontal line. Mouse Tbx22 cDNA (MM_TBX22, AF515700).

Download figure to PowerPoint

thumbnail image

Figure 3. Northern blot showing a mouse Tbx22 transcript of approximately 2.4 kb in embryonic day 14 snout tissue.

Download figure to PowerPoint

Comparative analysis of human, mouse, and rat TBX22 sequences revealed a stretch of 1.1 kb upstream of exon 1 sharing a high similarity (HS_TBX22_Promoter_Region, AF515702; MM_TBX22, AF515700; RN_TBX22_ Promoter_Region, AF515701). We screened this region of all three species for possible transcription factor binding sites by using the TransFac analysis tool and further refined the output by alignment to a consensus sequence. Five blocks containing one or two putative transcription factor binding sites were identified (Fig. 4). Blocks 1 and 3 harbour an S8/PRX2 and an MSX1 site overlapping each other. Block 2 contains an overlapping NKX2.5 and an MEF2 site. Block 4 consists of a single NKX2.5/HSX site. Block 5 contains two NFY sites close to each other in tail-to-tail orientation.

thumbnail image

Figure 4. Putative regulatory elements with positionally conserved binding sites in the mammalian along the consensus sequence of the 5′ upstream region of the TBX22 gene. Arrows indicate the direction of promoter activity, numbers denote the distance from the putative start ATG. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com]

Download figure to PowerPoint

The expression of Tbx22 was examined in mouse snout tissue at different developmental stages by reverse transcriptase-polymerase chain reaction (RT-PCR; Fig. 5). A PCR product of the expected size (196 bp) was obtained in embryonic day (E) 9–E11 complete embryos and in snout samples from E12–E16 embryos and confirmed by sequencing. PCR reactions with cDNA from adult snout yielded a very faint product; the use of internal primers in a subsequent nested PCR resulted in a distinct product of the expected size (148 bp; Fig. 5), indicating that Tbx22 transcripts are also present in cells of the adult snout.

thumbnail image

Figure 5. Expression of mouse Tbx22 determined by reverse transcriptase-polymerase chain reaction (RT-PCR). RNA was extracted from complete embryonic day (E) 9–E11 embryos, frontal parts of the head (E12), E14 and E16 snouts, and adult (A) snouts, including the eye portion (A1) or without the eye portion (A2). Primers tbex for/rev were used in reactions E9, E10, E11, E12, E14, E16, A1, A2, and a negative control reaction with H2O instead of template cDNA. Aliquots of products A1 and A2 were used in a second PCR with nested primers tbin for/rev (A1N and A2N). C1, control PCR with A1 cDNA and β-actin primers; C2, control PCR with A2 cDNA and β-actin primers. +, reverse transcriptase was added; −, negative control, no reverse transcriptase added.

Download figure to PowerPoint

In situ hybridizations on frontal sections of E12, E14, and E16 embryos revealed the expression of mouse Tbx22 in distinct areas of the embryonic head, namely the nasal septum, posterior palatal shelves before fusion, tongue, and the mesenchyme surrounding the eye anlage (Figs. 6–8); no additional sites of expression were seen on sagittal sections of an entire E14 embryo (not shown).

thumbnail image

Figure 6. Expression of mouse Tbx22 in the tissue bridge between the tongue and the floor of the mouth detected by in situ hybridizations. Frontal sections of embryonic day 14 embryos are arranged in an anterior (A) to posterior (K) order. Brightfield images are shown in the left column (A,C,E,G,J), darkfield images in the right column (B,D,F,H,K). White arrows in B,D,F, and H depict the areas of hybridization signals. Black arrows in J depict the genioglossus muscles. Asterisks indicate Meckel's cartilage. Light refringence by structures of the Meckel's cartilage is seen in D. No signals above the background were seen with sense probes (not shown). t, tongue; f, frenulum.

Download figure to PowerPoint

thumbnail image

Figure 7. Expression of mouse Tbx22 in the nasal septum (A–J,B–K) and in the tissue surrounding the eye anlage (L,M) detected by in situ hybridizations. Frontal sections of embryonic day 14 embryos, showing the nasal septum and palatal shelves arranged in an anterior (A) to posterior (K) order. Brightfield images are shown in the left column (A,C,E,G,J,L), darkfield images in the right column (B,D,F,H,K,M). In the inset of A, sloughing of suprabasal palatal shelf epithelial cells in close proximity to an outgrowth of the nasal septum can be seen. Black arrows in G and J depict the nasopharynx; white arrows in B,D,F,H, and M depict the areas of hybridization signals. The bright rim seen in M is due to refringence of light by the pigment layer of the eye. No signals above the background were seen with sense probes (not shown). ns, nasal septum; ps, palatal shelf; f, fused nasal septum and palatal shelves; l, lens of the developing eye.

Download figure to PowerPoint

thumbnail image

Figure 8. Frontal sections showing the expression of mouse Tbx22 in the posterior palatal shelves of an embryonic day (E) 12 embryo (A,B) and in the nasal septum of an E12 (C,D) and E16 (E,F) embryo detected by in situ hybridizations. Brightfield images are shown in the left column (A,C,E), darkfield images in the right column (B,D,F). White arrows depict the areas of hybridization signals. No signals above the background were seen with sense probes (not shown).

Download figure to PowerPoint

Expression of Tbx22 was detected in the tissue bridge between the tongue and the floor of the mouth (Fig. 6). On anterior sections, transcripts were found in mesenchymal cells at the base of the tongue, the frenulum, and in the adjacent superior layer of the floor of the mouth (Fig. 6B,D,F,H). No Tbx22 expression was seen on posterior sections where the broad base of the tongue is attached to the floor of the mouth (Fig. 6K). The tongue is formed from tissue swellings at the floor of the mouth, and during development, its basal attachment is anteriorly reduced to a thin tissue bridge, the frenulum. Disturbances of this process can result in an anteriorly extended and/or shortened frenulum, a condition called ankyloglossia (tongue-tie; Moore and Persaud, 1998). The observed Tbx22 expression, therefore, perfectly correlates with the ankyloglossia phenotype, which has been reported as an associated or sole feature in CPX families. We also found Tbx22 expression in mesenchymal cells surrounding the eye anlage (Fig. 7M).

In anterior snout sections, where the palatal shelves are not yet fused, Tbx22 transcripts were located in mesenchymal cells underneath the epithelium which lines the surface of the inferior nasal septum (Fig. 7B). During palate development, the inferior nasal septum fuses with the right and left palatal shelf to form the secondary palate, and we found Tbx22 to be expressed in outgrowths of the nasal septum, which expand in the direction of these shelves. In the inset of Figure 7A, sloughing of suprabasal palatal shelf epithelial cells in close proximity to such an outgrowth can be seen, a mechanism that takes place immediately before contact and as the first step of mesenchymal bridging of these structures. In a more posterior section, where the palatal shelves have further approached, Tbx22 expression was still confined to the inferior nasal septum (Fig. 7D). After fusion of the nasal septum with the palatal shelves, mesenchymal cells that seem to have migrated from the nasal septum and merged with the tissue of the palatal shelves showed a faint Tbx22 expression (Fig. 7F). In more posterior sections, the nasal septum is not fused to the secondary palate; instead, a fissure between these structures, the ventral extremity of the nasopharynx, was seen (Fig. 7G). On this section, a very faint seam of Tbx22 expression underlying the epithelial lining of the nasal septum was noted. More posteriorly, Tbx22 was not expressed in the nasal septum any more (Fig. 7K). The expression of Tbx22 in the nasal septum was also detected in E12 and E16 embryos (Fig. 8D,F). In another issue of this journal, the expression of mouse Tbx22 is demonstrated in the nasal septum on sagittal sections and in palatal shelves on very posterior frontal embryonic sections of different developmental stages where no relation to the nasal septum can be seen (Bush et al., 2002). In a very posterior section of an E12 embryo, we could also detect Tbx22 expression in posterior palatal shelves before fusion (Fig. 8B). That we did not detect Tbx22 in the palatal shelves of anterior sections before fusion with the nasal septum could either be explained by a very weak or restricted three-dimensional localization in this structure. However, taking our study and the study by Bush et al. together, another explanation would be that Tbx22 is expressed differently in anterior and posterior structures of the developing secondary palate or only at early stages of palatal shelf development before fusion. Of interest, only the anterior parts of the palatal shelves fuse with the nasal septum, whereas no connection is made with the posterior palatal shelves, leaving an opening, the nasopharynx. A recent study showed that the expression of Msx-1, another gene involved in secondary palate development, is restricted to the anterior region of the palatal shelves and that different genetic pathways act in the anterior and posterior palatal shelf mesenchyme (Zhang et al., 2002).

Of the transcription factors with a positionally conserved putative binding site within the 5′-region of TBX22, MSX1, which can act as a repressor, and S8/PRX2 represent strong candidates for upstream-acting factors as they have been previously implicated in palatogenesis (Satokata and Maas, 1994; Karg et al., 1997; Lu et al., 1999; van den Boogaard et al., 2000; Chesterman and Kern, 2002). Elucidating the spatiotemporal interaction of TBX22 with the putative upstream genes identified in our study should shed light on the complex molecular process underlying normal and disturbed palatogenesis.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

In Silico Identification of Mouse Tbx22

The nucleotide sequence of the human TBX22 mRNA (AF251684) was used to identify corresponding mouse genomic shotgun reads by screening the ENSEMBL trace archive (http://trace.ensembl.org) and the NCBI trace database (http://www.ncbi.nlm.nih.gov/Traces/trace.cgi). The Tbx22-positive reads were assembled into sequence contigs by using the GAP4 program package (Bonfield et al., 1995). In an iterative approach, the resulting contigs were used to screen the trace archives again, until the murine Tbx22 locus was assembled to completion. The resulting mouse genomic sequence was subjected to gene prediction by using the GenScan program (http://genome.dkfz-heidelberg.de/cgi-bin/GENSCAN/genscan.cgi; Burge and Karlin, 1997), and the predicted mRNA was used to screen the mouse EST-database (http://www.ncbi.nlm.nih.gov/blast/Blast.cgi) by using the BLASTN algorithm (Altschul et al., 1990). Matching ESTs were assembled to a mouse Tbx22 cDNA (MM_TBX22, AF515700).

Promoter Analysis of Mouse, Rat, and Human TBX22

The 5′-upstream regions of human, mouse, and rat TBX22 were comparatively analyzed. Sequences of rat and mouse promoters were obtained in an in silico approach as described for the mouse Tbx22 gene. A block of approximately 1,100 bp upstream of the putative start ATG with an average identity higher than 73% (mouse vs. human; rat vs. mouse: 88%) was thereby identified (MM_TBX22, AF515700; HS_TBX22_Promoter_Region, AF515702; RN_TBX22_Promoter_Region, AF515701). No clear transcription start signal could be found, though (PromoterScan II, http://www.molbiol.ox.ac.uk/ratio3.htm). The 1.1-kb sequence stretches of the three species were analyzed by using the TransFac analysis tool with a wider range of matrices available in a commercial trial version (Quandt et al., 1995). Limits of the core and matrix similarity were raised to 1.00 and 0.90, respectively, and resulting hits were aligned to a consensus sequence of the three promoter regions. Putative transcription factor binding sites within this sequence block of the three species, which were located at the exact same position with reference to the consensus sequence, were considered conserved ones.

Tissue Isolation

Tissues were dissected from embryos and adult mice of a mixed C57BL/6/C3H background. The day of the vaginal plug was considered as embryonic day 1 (E1). For RNA isolation, complete embryos (E9, E10, E11), frontal parts of the head (E12), or snouts (E14, E16, and adult mice) were dissected in cold phosphate-buffered saline (PBS) and frozen in liquid nitrogen. For in situ hybridization (ISH), heads and complete embryos (E12, E14, and E16) were isolated and fixed in Serra's fixative (60% ethanol, 30% formalin, 10% acetic acid) overnight at 4°C. The tissues were then dehydrated in 100% isopropanol, embedded in paraffin, sectioned (7 μm), and mounted on SuperFrost Plus slides (Menzel-Gläser).

RNA Isolation and RT-PCR

Total RNA was isolated from mouse tissues by using Trizol reagent (Invitrogen), according to the manufacturer's protocol. cDNA was synthesized from 5 μg of total RNA in a 10 μl oligo dT-primed reverse transcription reaction. A total of 1 μl of the reaction volume was used in subsequent 50-μl PCR amplifications with 10 pmol of primers tbex for (cccagcaaaagaaaagctca)/tbex rev (gactttcctggctgttgctc), tbin for (agagaggagatgcagcctga)/tbin rev (cgtcagagcaggagaaaaca), or β-actin for (tgaaccctaaggccaacc gtg)/β-actin rev (gctcatagctcttctccaggg). Amplifications consisted of 35 cycles (94°C for 45 sec, 53°C for 1 min, and 72°C for 1 min).

Northern Blot Analysis

A total of 1 μg poly A+ RNA was isolated from total RNA of E14 mouse snout tissue by using the RNeasy kit (Qiagen), separated by electrophoresis on a 1% agarose gel and transferred to a Hybond N+ membrane (Amersham Pharmacia Biotech). A 196-bp PCR fragment amplified with primers tbex for/tbex rev from E14 mouse snout cDNA was 32P-labelled and hybridized in QuiKHyb solution (Stratagene) for 2 hr at 65°C. Washing was performed in 2× standard saline citrate/0.1% sodium dodecyl sulfate at 65°C for 10 min, and signals were detected with a PhosphorImager (Amersham Biosciences).

In Situ Hybridization

The antisense and sense mouse Tbx22 riboprobes were synthesized from a linearized pGEM-T Easy vector (Promega) containing a 196-bp PCR fragment amplified with primers tbex for/rev, by using T7 or Sp6 polymerases and [35S]UTP (Amersham Pharmacia Biotech). Our protocol was based essentially on that of Adam et al. (1996) with some modifications (Krause et al., 1999). Slides were exposed with K.5 photographic emulsion (Ilford) for 3 weeks before development. Staining was performed by using Mayer's hemalum (Merck). The slides were observed under brightfield and darkfield illumination by using a Zeiss Axiophot microscope, and image collection was performed by using a Sony 3 CCD camera.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

We thank Bettina Lipkowitz and Ralph Schulz for excellent technical assistance.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES