MCS9.7 enhancer activity is highly, but not completely, associated with expression of Irf6 and p63



Background: DNA variation in Interferon Regulatory Factor 6 (IRF6) contributes risk for orofacial clefting, including a common DNA variant rs642961. This DNA variant is located in a multi-species conserved sequence that is 9.7 kb upstream from the IRF6 transcriptional start site (MCS9.7). The MCS9.7 element was shown to possess enhancer activity that mimicked the expression of endogenous Irf6 at embryonic day 11.5 in transient transgenic embryos, and also contains a p63 binding site that transactivates IRF6 expression. To analyze whether the MCS9.7 enhancer is sufficient to drive IRF6 expression, we generated stable transgenic murine lines that carry a MCS9.7-lacZ transgene. We hypothesized that MCS9.7 was sufficient to recapitulate the endogenous expression of Irf6 at other time-points during embryonic development. Results: We observed that MCS9.7 activity recapitulated endogenous Irf6 expression in most tissues, but not in the medial edge epithelium (MEE) at E14.5, when Irf6 expression was high during secondary palatal fusion. Also, while MCS9.7 activity and Irf6 expression were associated with p63 expression, we observed MCS9.7 activity and Irf6 expression in periderm, although p63 was absent. Conclusion: These data suggest that MCS9.7 enhancer activity is not sufficient to recapitulate IRF6 expression, and that p63 expression is not always necessary nor sufficient for transactivation of IRF6. Developmental Dynamics 241:340–349, 2012. © 2011 Wiley Periodicals, Inc.


Formation of the lip and palate in human and mouse is a complex series of cellular processes during embryonic development (Cox, 2004). Lip and palate are developed at different time points, but use common cellular processes that include proliferation, differentiation and migration. Because disruption in any of these developmental processes can lead to orofacial abnormalities, cleft lip and palate is one of the most common congenital birth defects. To understand the development of lip and palate, it is important to define the basic players and regulatory pathways that are involved. In the mouse, the first signs for lip development start at embryonic day (E) 9.5 with the formation of the frontonasal prominence, and the paired maxillary and mandibular processes that surround the oral cavity. Subsequently, the invagination of the nasal placodes split the lower portion of the frontonasal prominence into paired medial and lateral nasal processes. The upper lip becomes continuous by the merging of the facial processes at E11.5. For the development of secondary palate, the epithelial and the mesenchymal cell types play determinant roles for proper formation in embryos. More specifically, secondary palatal shelves go through at least three critical developmental processes; outgrowth of the palatal shelves from the maxillary prominence along the side of the tongue, elevation and apposition of the palatal shelves, whereby the two epithelial layers form a medial edge seam, and then fusion of the palatal shelves to yield a mesenchymal confluence that separates the nasal from oral cavity. During the apposition and fusion steps, the critical tissues are the periderm and the medial edge epithelium of secondary palatal shelves, where both of these cell types must cease proliferation and eventually disappear (Cox, 2004; Knight et al., 2006; Dudas et al., 2007).

Van der Woude syndrome (VWS), an autosomal dominant genetic disorder, is the most common syndromic form of orofacial clefting, and accounts for 2% of all cases of cleft lip and palate (Burdick et al., 1985). Likely etiologic mutations were identified in interferon regulatory factor 6 (IRF6) in 75% of probands with VWS (Kondo et al., 2002; de Lima et al., 2009). Morphological and histological analyses of mutant murine strains showed that Irf6 is required for the proper development of orofacial, limb and cutaneous tissues (Ingraham et al., 2006; Richardson et al., 2006, 2009a). To dissect the role of IRF6 and the potential effects of DNA variants, it is important to study its expression pattern during embryonic development and how that pattern is regulated. It was shown that Irf6 is expressed in the epithelium of the lateral and medial nasal prominences as well as in the epithelium of the maxillary and mandibular prominences. In addition, Irf6 was strongly expressed in the medial edge epithelium throughout palatal development (Knight et al., 2006). In addition, Irf6 plays a critical role in the formation and maintenance of the periderm during secondary palatal development (Richardson et al., 2009a). In stratified skin, Irf6 expression was detected mainly in the suprabasal layers, specifically in the spinous layer at E17.5 (Ingraham et al., 2006; Richardson et al., 2006, 2009a).

Studying the regulation of genes involved in developmental processes of the embryo will allow us to identify rare and common mutations in regulatory elements that may contribute risk for human genetic diseases. For example, a common DNA variant (rs642961) was observed to be significantly associated with the isolated cleft lip (Rahimov et al., 2008). Furthermore, this DNA variant was located in a multi-species conserved sequence (MCS) that was 9.78 kb upstream from the IRF6 transcription start site (MCS9.7). The MCS9.7 element was shown to possess enhancer activity that mimicked endogenous Irf6 expression at E11.5. More recently, the MCS9.7 enhancer element was shown to be bound by the trans regulator deltaN-p63 and to transactivate the expression of Irf6 in epithelial cells of secondary palatal shelves at E14.5 (Thomason et al., 2010). These data suggested that deltaN-p63 was a critical regulator of Irf6 expression through MCS9.7.

In this study, we generated stable transgenic murine lines that express a lacZ reporter gene driven by the MCS9.7 enhancer element. Murine embryos were harvested between E9.5 and E17.5 to determine lacZ expression during the development of the oral cavity and the face. Furthermore, we studied lacZ expression in the different cell layers of the epidermis, because of previously reported cutaneous expression of Irf6 and cutaneous defect in Irf6 knockout mice (Ingraham et al., 2006; Richardson et al., 2006). Our data suggest that the MCS9.7 enhancer element is sufficient to recapitulate the expression of endogenous Irf6 in most tissues and time points. An exception was the medial edge epithelium at E14.5, a critical time point when the palatal shelves are starting to fuse. The high sensitivity of the lacZ transgene also led to the identification of previously undetected sites of Irf6 expression during embryonic development.


Stable MCS9.7-lacZ Lines Recapitulated Bgal Activity of Transient Transgenic Murine Embryos

Previously, we observed that MCS9.7 appeared to recapitulate endogenous Irf6 expression at E11.5 using transient transgenic embryos (Rahimov et al., 2008). To test whether MCS9.7 recapitulates the endogenous expression of Irf6 at later time points and in other tissues, we generated stable murine transgenic lines using the same vector. We obtained four independent lines, but one could not be propagated and one did not show Bgal activity, although it was positive for the transgene by polymerase chain reaction (PCR). Using the two remaining lines, Bgal activity was first studied at E9.5 because it is the beginning of the development of the frontonasal prominences as well as the maxillary and mandibular processes. Macroscopically, Bgal staining was observed in the facial prominences, 3rd branchial arch, hindbrain, fore-, and hindlimb buds as well as in the somites (Fig. 1A). At E10.5, additional Bgal activity was observed in the branchial arches and the apical ectodermal ridge of the fore- and hindlimb buds (Fig. 1B). Bgal activity was maintained at E11.5, with additional intense staining in the dorsal surface of the fore- and hindlimb buds (Fig. 1C). At E13.5, additional Bgal activity was detected in facial structures, including follicles of the whisker pad (Fig. 1D). In the previous study, images were taken for nine transient transgenic embryos that showed a common pattern of Bgal activity ( Here, the pattern of Bgal activity in E11.5 embryos from both stable transgenic lines was mimicked by two of the nine transient transgenic embryos, with blue staining in the epithelium of orofacial tissues, limbs, eyes, hindbrain, and somites (Supp. Fig. S1, which is available online). The other seven transient embryos showed blue staining in all of these tissues, except the hindbrain. Thus, at this gross level and single time point, the pattern of MCS9.7 activity, observed in multiple transient and stable transgenic lines, appeared to recapitulate the previously observed pattern of Irf6 expression (Knight et al., 2006). Optical tomography (OPT) movies were taken to show the three-dimensional (3D) pattern of Bgal activity in the stable MCS9.7-lacZ embryos at E11.5 and E13.5 (Supp. Fig. S1; Supp. Movies).

Figure 1.

MCS9.7-lacZ enhancer activity during embryonic development. Bgal activity was visualized in whole-mount murine embryos between embryonic day (E) 9.5 and E13.5. A–D: Bgal activity was detected in the 1st (I), 2nd (II), and 3rd (III) branchial arches (A,B), maxillary (mx) and mandibular (m) processes (A,B), hindbrain (hb; A,B), dorsal surface of fore- (flb) and hindlimb buds (hlb; C,D), apical ectoderm ridge (aer; B), and the somites (s; A–D). Scale bar = 1.0 mm.

MCS9.7-lacZ Expression in Craniofacial Tissues

To evaluate MCS9.7 enhancer activity in the developing orofacial tissues, we determined the pattern of Bgal activity from E10.5 to E14.5 in whole-mount embryos. At E10.5, Bgal activity was observed in the lateral and medial nasal prominences, in the epithelium between the maxilla and mandibular prominences and between the branchial arches, as well as in the 3rd branchial arch and the hindbrain (Fig. 2A,E). At E11.5, the development of the upper and lower lip is completed. At this time point, a similar pattern of Bgal activity was observed, however intense staining was also observed in the nares, in the inter-maxillary segment, and in the medial area of the lateral nasal and maxillary processes (Fig. 2B,F). At E13.5, the facial processes had fused completely and Bgal activity was detected in the nasal pits, vibrissae, supra-orbital hair follicles, and in the ectoderm of the developing eyelid and outer ears (Fig. 2C,G). At E14.5, we did not perform a whole-mount staining for the entire embryo because the epidermis starts to stratify leading to reduced penetration of staining solution into the embryo. Therefore, only a head at E14.5 was used for staining. We observed similar Bgal activity compared with E13.5, with additional strong staining in the submandibular glands and outer ear (Fig. 2D,H). We did not observe obvious differences in Bgal activity between the two stable transgenic lines. These data are consistent with the hypothesis that lacZ expression was not altered by position effect, and that the observed Bgal staining reflected activity of the MCS9.7 enhancer.

Figure 2.

MCS9.7-lacZ activity in craniofacial tissues. Bgal activity was visualized in whole-mount transgenic murine embryos between embryonic day (E) 10.5 and E14.5. A–H: Frontal (A–D) and lateral (E–H) views of the same head. Bgal activity was observed in the medial nasal processes (mnp), maxillary processes (mx), mandibular processes (m), hindbrain (hb), nares (n), 2nd branchial arch (II), vibrissae (v), sub-mandibular gland (sg), and outer ear (e). Scale bar = 1.0 mm.

MCS9.7 Activity During Palatal Development

In the oral cavity, Irf6 was previously observed in the tooth germ, the palatal rugae and in all epithelium of oral and nasal cavities (Kondo et al., 2002; Knight et al., 2006). Importantly, Irf6 was observed in the medial edge epithelium (MEE) throughout palatal development, until the MEE disappears during palatal fusion. To test whether MCS9.7 is sufficient to drive this pattern of expression, we determined Bgal activity in E13.5 and E14.5 heads. In whole-mount embryos, we observed Bgal activity in the tooth germ and palatal rugae at both E13.5 (Fig. 3A) and E14.5 (Fig. 3B). Also, Bgal activity was observed in the MEE of the palatal shelves at E13.5. Unexpectedly, Bgal activity was not observed in the MEE of the palatal shelves at E14.5. To determine the cell type presenting the Bgal activity, we performed X-gal staining on coronal sections of E13.5 and E14.5 heads. At E13.5, Bgal activity was observed in all epithelia of the oral and nasal cavities (Fig. 3C,E). At E14.5, Bgal activity continued in the oral epithelium of the palate and the tongue, but was greatly reduced in the nasal epithelium and absent in the MEE (Fig. 3D,F).

Figure 3.

MCS9.7-lacZ enhancer activity in the oral cavity. A: Bgal activity in the medial edge and rugae of secondary palatal shelves (p) and in the incisor and molar tooth germs (tg) from whole-mount staining at embryonic day (E) 13.5. B: At E14.5, Bgal activity was similar except it was absent at the medial edge of the palatal shelves (arrows). C,E: Staining of coronal sections of E13.5 heads showed Bgal activity along oral, nasal, and tongue epithelium. D,F: However, Bgal activity was not observed in the medial edge epithelium of secondary palatal shelves just prior to palatal fusion despite the dark blue staining in oral epithelium and tooth germ. E and F images are close up for C and D. Scale bar = 0.75 mm (A,B), 0.5 mm (C,D), and 0.2 mm (E,F).

To confirm the change observed in MCS9.7 activity in the MEE at E14.5, immunofluorescent staining was performed on serial sections of transgenic murine embryo at E13.5 and E14.5. First, we performed immunostaining for keratin 14 (K14) at both of these time points (Fig. 4A–C). The K14 staining included the first and last slides in the series of sections to ensure that the MEE was present throughout (data not shown for last slide) the experiment. We also confirmed the presence of endogenous Irf6 protein in the MEE at both time points (Fig. 4D–F). However, while Bgal protein was detected in the oral epithelium and MEE at E13.5 (Fig. 4G) and in the oral epithelium at E14.5 (Fig. 4H,I), we did not detect the presence of Bgal protein in the MEE (Fig. 4H,I). In addition, the pattern of expression for p63 protein paralleled the expression of Bgal protein (Fig. 4J–L). The expression pattern of p63 is significant because it was shown previously to bind to the MCS9.7 enhancer element in primary human keratinocytes and to regulate its activity in vitro (Thomason et al., 2010). In sum, these results suggested that MCS9.7 is sufficient to recapitulate endogenous expression of Irf6 in all previously studied tissues during murine embryonic development, but not in the MEE at E14.5. Additionally, these results support a role for p63 in regulating MCS9.7 activity in vivo.

Figure 4.

Expression of Krt14, Irf6, p63, and Bgal proteins in embryonic murine secondary palate. Fluorescent immunostaining was performed on serial coronal sections of heads of MCS9.7-lacZ transgenic embryo at embryonic day (E) 13.5 and E14.5. A–C: Krt14 serves as a marker for epithelial cells, including the medial edge epithelium (MEE). D–F: Irf6 is expressed in MEE throughout palatal development. G–I: Bgal was observed in all epithelium, including MEE at E13.5 (G), but was not detected in the MEE at E14.5 (H, I). J–L: Staining for p63 was identical to Bgal. Secondary palatal shelf (p); nasal septum (ns). White arrows indicate the MEE. Scale bar = 100 μm (A, B) and 50 μm (C).

MCS9.7 Activity in the Periderm, Basal Epithelium, and Stratified Epidermis

To better characterize the dynamic changes in spatiotemporal activity of MCS9.7 in epithelia, we performed Bgal staining on additional tissues and time points. In coronal sections of E13.5 head, Bgal activity appeared to localize to the periderm of oral and nasal epithelia (Fig. 5A). However, at E14.5, we observed Bgal activity in some regions of the basal layer of the nasal and oral epithelia (Fig. 5B). A similar dynamic pattern of Bgal activity was observed in facial epidermis. For example, at E13.5, Bgal activity was restricted to the periderm (Fig. 5C), but at E14.5, Bgal activity could also be seen in regions of the basal layer of the epidermis (Fig. 5D). Finally, we studied Bgal activity in dorsal skin at E17.5, where endogenous Irf6 is expressed primarily in the spinous layer (Ingraham et al., 2006). We observed Bgal activity in the suprabasal layers of the epidermis (Fig. 5F). Similarly, in adult skin from ear and footpad, we observed Bgal activity in the granular and cornified layers of the epidermis with occasional spinous Bgal-positive keratinocytes (data not shown). We conclude that MCS9.7 enhancer activity is highly dynamic in the different epithelial layers and time points.

Figure 5.

Dynamic epithelium-specific enhancer activity of MCS9.7. A,B: In the oral epithelium, Bgal activity was observed primarily in the periderm at embryonic day (E) 13.5 (A) but was also observed in the basal layer at E14.5 (B). C,D: A similar change in expression was observed in the facial epidermis from E13.5, (C) to E14.5 (D). E,F: At E17.5, Bgal activity was not observed in transgene negative embryonic back skin (E), yet was detected in the suprabasal layers in transgene-positive animal (F). Maxilla (mx), mandible (m), basal layer (b), periderm (p), dermis (d), suprabasal layer (s). Scale bar = 200 μm (A-F).

To correlate this dynamic MCS9.7 activity with endogenous Irf6 and p63 expression, we performed dual immunostaining of coronal sections of E13.5 and E14.5 heads. As observed previously (Knight et al., 2006; Richardson et al., 2009a; Thomason et al., 2010), Irf6 and p63 protein were only detected in the epithelium of oral tissues at both time points (Fig. 6A,B). However, with higher magnification, Irf6 and p63 staining at E13.5 were distinct in most regions. Irf6 staining, like the Bgal activity, was generally localized to the periderm, whereas p63 was only detected in the basal cells (Fig. 6A1). Only one region of oral epithelium showed overlapping staining of p63 and Irf6 (Fig. 6A2). Conversely, at E14.5 most regions showed overlapping staining of Irf6 and p63 (Fig. 6B1). The distinct areas included the oral epithelium and also the MEE. As described above for the MEE, Irf6 remained strongly expressed, but p63 was greatly decreased or absent (Fig. 6B2). Overall, Irf6 expression and MCS9.7 activity are highly associated, except at the MEE at E14.5. Surprisingly, Irf6 expression and MCS9.7 activity in the periderm appear to be independent of p63 expression, and Irf6 expression in the MEE at E14.5 appears to be independent of MCS9.7 activity and of p63 expression.

Figure 6.

Dynamic expression of Irf6 and p63 in palatal development. A,B: At embryonic day (E) 13.5 (A), expression of Irf6 (red) and p63 (green) appeared primarily distinct, but at E14.5 (B) they appeared to overlap primarily. Higher magnification showed p63 only in the basal layer at E13.5 (A1, A2), with Irf6 primarily in the periderm (A1) with some localized expression in basal cells (A2). At E14.5, both p63 and Irf6 were observed in the basal layer of oral epithelium (B1), but while p63 was greatly decreased in the medial edge epithelium (MEE), Irf6 remained strong (B2). Maxilla (mx), mandible (m), basal layer (b), dermis (d), palate (p). Scale bar = 100 μm (A, B) and 50 μm (A1, A2, B1, B2). A and B were stretched horizontally, and A was compressed vertically to make the images fit the figure.

MCS9.7 Activity in Other Tissues

In this study, we observed Bgal activity in tissues where endogenous Irf6 expression had not been reported, including the hindbrain, the somites, nonepithelial tissues of the tongue, as well as the cartilage and muscle of fore- and hindlimbs (Supp. Fig. S2). To test for endogenous expression of Irf6 in these tissues, we performed immuno-staining with the Irf6 antibody on transverse sections of the hindbrain and sagittal sections on the forelimb (Supp. Fig. S3). In the brain, we observed Irf6 expression most prominently in the medullary raphe. This structure is notable because, like the palate, it is formed by the fusing of two tissues. In the limbs, we observed Irf6 expression in the cartilage primordium of the humerus and the metacarpel bones. This expression is notable because previous studies showed a delay in bone formation in embryos that lacked Irf6 (Ingraham et al., 2006; Richardson et al., 2006). Under higher magnification, we observed a striated pattern of staining in these tissues. To further validate endogenous Irf6 expression in these regions, we compared the immuno-staining results with available whole-mount in situ hybridization at EMAGE (Richardson et al., 2009b) and sagittal sections at Eurexpress (Diez-Roux et al., 2011). At E14.5, a similar pattern of expression was observed in medullary oblongata and primordium of cartilage in fore- and hindlimbs. In general, these expression patterns for limb and brain were consistent with the Bgal activity shown in transgenic lines, and may help to explain abnormalities in brain (Nopoulos et al., 2007a, b) and limb (Kantaputra et al., 2004) development in patients with Van der Woude syndrome.


In our previous study, we showed that MCS9.7 had enhancer activity at E11.5 in transient transgenic embryos (Rahimov et al., 2008). Given the high similarity between the enhancer activity of MCS9.7 and endogenous expression of Irf6 at that time point, we hypothesized that MCS9.7 would be sufficient to recapitulate endogenous expression of Irf6 during embryonic development. In fact, using the same MCS9.7-lacZ transgene in two transgenic murine lines, we observed Bgal activity in most, but not all, tissues and cell types expressing endogenous Irf6. As expected from previous studies (Kondo et al., 2002; Ingraham et al., 2006; Knight et al., 2006; Richardson et al, 2006; Richardson et al., 2009a), we observed Bgal activity along the edges of facial growth projections and branchial arches, the apical ridge of the limb buds, the palatal rugae, the medial edge of the secondary palatal shelves at E13.5, the tooth germs, the periderm, the oral and nasal epithelia, the hair follicles, and the epidermis of the skin. Thus, MCS9.7 is sufficient to recapitulate endogenous Irf6 expression in most tissues. However, four observations were unexpected; the dynamic activity of the MCS9.7 enhancer in oral epithelium and skin, the lack of MCS9.7 activity in the medial edge epithelium (MEE) at E14.5, the expression of Irf6 and activity of MCS9.7 in the periderm in the absence of p63, and the strong enhancer activity in tongue mesenchyme.

In this study, we observed an obvious and consistent transition in MCS9.7 activity from the periderm at E13.5 to regions of the basal epithelial layer at E14.5, and this transition was highly associated with similar changes in expression of endogenous Irf6. Previous studies of Irf6 did not describe this change in cell type expression (Kondo et al., 2002; Knight et al., 2006; Richardson et al., 2009a; Thomason et al., 2010). However, this transition is difficult to detect because the two cell types are adjacent, and the periderm is a very thin cell layer. In the previous studies, three used RNA in situ hybridization and only one used immunostaining to detect endogenous Irf6 expression. RNA in situ lacks sufficient resolution to discriminate between two, single cell layers. And, the difference with immunostaining at E13.5 and E14.5 only became apparent in the current study after performing the Bgal staining for the transgene.

Another unexpected result is that we did not observe MCS9.7 activity in the medial edge epithelium (MEE) at E14.5, although endogenous Irf6 protein was expressed in these same cells. We observed a similar result in cell culture, where undifferentiated primary keratinocytes expressed Irf6, but did not show MCS9.7 activity under these growth conditions (Biggs et al., 2011). Because we obtained the same results in two independent transgenic lines, the absence of MCS9.7 activity in these cells was likely not due to a genomic effect based on the site of insertion. We conclude that MCS9.7 is sufficient to drive Irf6 expression in many tissues and time points, but it is not sufficient for Irf6 expression in all tissues.

The lack of Bgal activity in the MEE at E14.5 coincided with a lack of expression of p63 in those cells. This observation is consistent with recent studies showing that p63 binds to MCS9.7 enhancer elements in primary human keratinocytes and that mutation of the two p63 binding sites abolishes the enhancer activity in cell culture (Thomason et al., 2010). Furthermore, expression of Irf6 in oral and nasal epithelia was remarkably reduced in p63 null mice at E13.5 (Moretti et al., 2010; Thomason et al., 2010). If p63 is not in the MEE at E14.5, and MCS9.7 is not active, then it remains to determine which trans and cis factors are responsible for the expression of Irf6 at this critical point in palatal development. Although no clear answer to this question is currently available, previous studies showed that expression of Irf6 in the MEE at E14.5 was dependent on Tgfb3 signaling (Knight et al., 2006; Xu et al., 2006). Thus, we speculate that expression of Irf6 in the MEE at E14.5 involves factors that transduce the Tgfb3 signal in the MEE. If true, then one role for Tgfb3 signaling during palatal development would be to remove p63 from the cells of the MEE, potentially by an Irf6-dependent targeting of p63 protein to the proteosome for degradation (Moretti et al., 2010). As such, we propose that Tgfb3 could serve to close the p63-Irf6 negative feedback loop in the epithelial cells of the MEE (Fig. 7).

Figure 7.

A proposed model for the regulation of Irf6 in medial edge epithelium (MEE). Solid lines are based on our studies and previously published data (Thomason et al., 2010; Moretti et al., 2010). Dashed line represents a hypothetical regulation.

While the expression of Irf6 in the MEE was not associated with p63 expression or with MCS9.7 activity, the expression of Irf6 in the periderm was associated with MCS9.7 activity, but not p63 expression. One possible explanation is that Irf6 expression and MCS9.7 activity were due to p63, but whereas p63 protein was degraded, Irf6 and Bgal proteins purdured. Another possible explanation is that Irf6 expression and MCS9.7 activity are independent of p63 expression in the periderm. Thus, some other trans factor may be activating the MCS9.7 enhancer in these cells. A potential candidate is AP-2α, because it is expressed in early ectodermal development (Byrne et al., 1994), and because the common SNP rs642961 that is located in MCS9.7 was shown to be associated with cleft lip and alter binding by AP-2α (Rahimov et al., 2008).

The final set of unexpected results was the presence of lacZ expression in several tissues where endogenous Irf6 had not been previously reported. The strong and highly localized expression in the tongue was most surprising. Previous in situ studies detected Irf6 expression in the thyroglossal duct at the base of the tongue (Kondo et al., 2002), but other in situ and immunostaining experiments did not demonstrate obvious expression of Irf6 mRNA and protein in the core region of the tongue (Knight et al., 2006; Richardson et al., 2009a; Thomason et al., 2010). The unexpected MCS9.7 activity and Irf6 expression in the hindbrain is also of interest. The expression of Irf6 in the medullary raphe may be analogous to other structures in the body such as the palate, the penis, and the scrotum, as tissues that form a raphe and express Irf6 (Kondo et al., 2002). This expression pattern suggests a common function for Irf6 to fuse structures during development. Also, the broader expression of Irf6 in regions of developing neuroepithelium may provide a molecular rationale for the previously observed association between patients with VWS and cognitive disorders, as well as abnormalities in brain structure (Nopoulos et al., 2007a, b). In the limb, we observed MCS9.7 activity in the humerus, and the radius primordia in fore- and hindlimb at E11.5 and E13.5. In our previous study of Irf6 null embryos, we observed abnormal bone and skeleton development (Ingraham et al., 2006). Our immuno-staining showed that Irf6 was expressed in the primordial bone during embryonic development. Thus Irf6 may have a direct role in bone development.

Finally, the results presented here will help to identify cis and trans factors that contribute risk for cleft lip and palate. Several genome wide association studies (GWAS) identified DNA variants in gene-poor areas that are significantly associated with complex diseases, including cleft lip and palate (Ghoussaini et al., 2008; Birnbaum et al., 2009; Tuupanen et al., 2009; Beaty et al., 2010). These observations may suggest that etiologic DNA variants reside in regulatory elements such as DNA enhancers. This hypothesis was already proven for IRF6, where rs642961, a common DNA variant in the MCS9.7 enhancer element, was highly associated with isolated cleft lip (Rahimov et al., 2008). Also, this variant does not account for all of the attributable risk for cleft lip and palate at the IRF6 locus. In addition, approximately 25% of families that have at least one individual with VWS lack an etiologic mutation in one of the exons of IRF6 (de Lima et al., 2009). Together, these results suggest the need for further mutation studies in putative regulatory elements. The ability of MCS9.7 activity to recapitulate endogenous IRF6 expression provides strong rationale for searching for disease-associated variants in this region. Finally, our data will allow us to correlate Bgal activity to expression of putative trans factors that may regulate IRF6 through MCS9.7. We can then use the MCS9.7-lacZ mice as an in vivo model to directly screen for genetic interactions with factors that regulate the expression of IRF6 in lip and palate.


Generation of Transgenic Mice and Genotyping

The 607-bp sequence element of MCS9.7 was cloned into the XbaI site upstream from the SV40 basal promoter, the lacZ reporter gene and the SV40 polyA site of the Hsp68 expression plasmid. This is the same vector that was used to generate transient transgenic embryos described previously (Rahimov et al., 2008). Transgenic mice were generated with this vector by standard procedures at the University of Iowa transgenic facility using fertilized eggs from C57BL/6 crosses. Transgenics were identified by PCR with MCS9.7 forward- (F: CTCCTCTGTTTGCTTAGCTTACCTC) and reverse-primer (R: CTGGTAAATGGTGAGTAGGAAGTTG) using genomic DNA extracted from tail tissue. Animal use procedures were approved by the All-University Committee on Animal Use and Care at Michigan State University, and followed National Institute of Health guidelines.

Bgal Activity Staining of Whole-Mount Embryos

Embryos from pregnant transgenic female mice were dissected and kept in 1.5-ml tubes containing cold phosphate buffered saline (PBS) until all embryos were processed. Staining for beta-galactosidase (Bgal) was done simultaneously for all embryos harvested in that day with slight modification of a published procedure (Parmar et al., 2001). Embryos were fixed in 4% paraformaldehyde for 15 min for embryos at E9.5 and E10.5, and 30 min for embryos from E11.5 to E14.5. Embryos were then washed 4 times for 5 min each with cold PBS buffer and 2 times for 5 min in Bgal washing buffer. One half milliliter of Xgal (20 ng/ml) in Bgal washing solution was added to each embryo, and vials were incubated for 15–45 min depending on the appearance of blue color. Embryos were then washed 3 times with cold PBS buffer for 5 min, followed by 15-min fixation in 4% paraformaldehyde (Parmar et al., 2001). For this assay, the Bgal staining buffer was always made fresh.

Bgal Activity Staining of Cryosections

For OCT tissue sectioning, the embryos were processed with modifications of the above procedure. The fixing step in 4% paraformaldehyde was extended for E13.5 and E14.5 embryonic heads up to 2 hr. After fixing, embryos were transferred to 5% sucrose for 2 hr, then to 30% sucrose and incubated overnight. Embryos that sank to the bottom of the tubes were embedded in OCT solution and stored at −80°C until use. Twelve-micron-thick sections were post-fixed for 1 min at 4°C using Bgal fixing solution. Slides were rinsed in PBS containing 2 mM MgCl2 at 4°C, followed by a 10-min wash in the same solution. Slides were washed in Bgal washing buffer for 10 min on ice, followed by staining for 6 hr or overnight at room temperature in the dark using Bgal staining buffer. The slides were washed twice in PBS containing 2 mM MgCl2 for 5 min each at room temperature, and then rinsed in distilled H2O. Sections were counterstained for 10 min in 1% Orange G solution dissolved in 2% phosphotungstic acid, followed by three washes in distilled H2O for 5 min each time, and dehydrated through ascending concentrations of methanol (50%, 70%, and 100% ethanol, 2 min each in solution). Sections were cleared and rinsed twice for 3 min in xylene and then mounted using Permount solution.

Fluorescent Immunostaining

Expression of Irf6, Bgal, p63, and K14 proteins was performed on paraffin and frozen sections from heads of transgenic embryos. Tissues were deparaffinized and rehydrated in a series of ethanol dilutions. Sections were blocked with 10% normal goat serum in 1% PBS-bovine serum albumin for 1hr, then incubated overnight at 4°C with the following primary antibodies: polyclonal rabbit anti-Irf6 (1:500, Irf6-SPEA), rabbit polyclonal Bgal (1:200, Santa Cruz, CA), mouse monoclonal p63 (1:150, 4A4 against residues 1-205 of ΔNp63, Santa Cruz, CA), rabbit polyclonal p63 (1:150, clone H-129 detects TAp63 and ΔNp63, Santa Cruz, CA), mouse monoclonal keratin 14 (1:100, Abcam, MA). After rinsing in PBS, sections were incubated with secondary antibodies conjugated to Alexa Fluorophore 488 or 555 (Molecular Probes, Invitrogen, CA). The nuclei were counterstained with DAPI (4′,6-diamidine-2-phenylidole-dihydrochloride) in PBS (1:1,000). The images were taken using a Nikon Eclipse 90i fluorescent microscope. The polyclonal rabbit anti-Irf6 antibody was produced by co-injecting two peptides as described previously (Bailey et al., 2005). However, two distinct antibodies were affinity purified with each of the two peptides. We called these Irf6-SPEA and Irf6-EDEL, which are the first four amino acids for each peptide. For all tissues and time points tested, pretreatment of the Irf6-SPEA antibody with the immunizing peptide abrogated staining (data not shown).

Optical Projection Tomography (OPT)

To visualize MCS9.7-lacZ expression pattern in 3D, whole murine embryos were mounted in 1% agarose, dehydrated in methanol, and then cleared overnight in BABB (1 part benzyl alcohol: 2 parts benzyl benzoate). The samples were then imaged using a Bioptonics OPT Scanner 3001 (Bioptonics, UK) using brightfield to detect the lacZ staining and for tissue autofluorescence (excitation 425 nm/ emission 475 nm) to capture the anatomy (Sharpe et al., 2002). The resulting images were reconstructed using Bioptonics proprietary software, automatically thresholded and merged to a single 3D image output using Bioptonics Viewer software.