|Home | About | Journals | Submit | Contact Us | Français|
Expression of the basic helix-loop-helix factor Hairy and Enhancer of Split-1 (Hes1) is required for normal development of a number of tissues during embryonic development. Depending on context, Hes1 may act as a Notch signalling effector which promotes the undifferentiated and proliferative state of progenitor cells, but increasing evidence also points to Notch independent regulation of Hes1 expression. Here we use high resolution confocal scanning of EGFP in a novel BAC transgenic mouse reporter line, Tg(Hes1-EGFP)1Hri, to analyse Hes1 expression from embryonic day 7.0 (e7.0). Our data recapitulates some previous observations on Hes1 expression and suggests new, hitherto unrecognised expression domains including expression in the definitive endoderm at early somite stages before gut tube closure and thus preceding organogenesis. This mouse line will be a valuable tool for studies addressing the role of Hes1 in a number of different research areas including organ specification, development and regeneration.
Hes1 is a basic helix-loop-helix (bHLH) transcriptional repressor which is required for normal development of several tissues and organs including the nervous system (Ishibashi et al., 1995), the eyes (Tomita et al., 1996), the pancreas (Jensen et al., 2000), the thymus (Tomita et al., 1999) and the lungs (Ito et al., 2000). Hes1 is also expressed in the developing kidneys (Chen and Al-Awqati, 2005; Piscione et al., 2004), the intestine (Jensen et al., 2000) and the stomach (Nyeng et al., 2007). Loss of Hes1 in the developing nervous system leads to up-regulation of the neural differentiation factor Mash1 and subsequently premature neural differentiation (Ishibashi et al., 1995). Also, the loss of Hes1 results in premature endocrine differentiation in the developing mouse gut tube as well as in the pancreas, resulting in pancreatic hypoplasia (Apelqvist et al., 1999; Esni et al., 2004; Fujikura et al., 2006; Jensen et al., 2000). Cell culture studies have shown that Hes1 is a downstream target of Notch signalling in some contexts (Jarriault et al., 1995; Jarriault et al., 1998; Ohtsuka et al., 1999), and loss-of-function studies often show similar phenotypes when comparing Hes1 and other Notch pathway null mutants which all result in premature differentiation (Apelqvist et al., 1999; Esni et al., 2004; Fujikura et al., 2006).
In the pancreas, Hes1 regulates endocrine differentiation (Jensen et al., 2000) by inhibiting the expression of the pro-endocrine differentiation factor Neurog3 (Apelqvist et al., 1999; Fukuda et al., 2006), and studies in cell culture suggest that exocrine differentiation may be regulated by direct interactions between Hes1 and Pancreatic transcription factor 1a (Ptf1a) (Esni et al., 2004; Ghosh and Leach, 2006). In the adult pancreas, Hes1 is restricted to the centroacinar cells but becomes activated by pancreatic injury, e.g. during chemically induced pancreatitis (Jensen et al., 2005), indicating a possible role in regeneration. Hes1 is also upregulated during induction of pancreatic cancer (Miyamoto et al., 2003; Pasca di Magliano et al., 2006).
Hence, a key function of Hes1 is to prevent differentiation and keep a pool of progenitor cells in a proliferative state to ensure appropriate growth of the developing tissue. However, more recent studies point to several Notch independent ways of activating Hes1 expression. During somitogenesis, segmentation is controlled by a molecular clock (Pourquie, 2003) where FGF has been shown to induce the oscillations of Hes1 expression (Nakayama et al., 2008), while in retina explants, Sonic hedgehog (Shh) has been demonstrated to regulate Hes1 activity in a Gli2 dependent manner (Ingram et al., 2008; Wall et al., 2009).
Previous investigations of Hes1 mutant mice revealed formation of ectopic pancreatic tissue in the common bile duct (Sumazaki et al., 2004), stomach, and duodenum, along with gall bladder agenesis (Fukuda et al., 2006). It has recently been proposed that Hes1 acts in conjunction with the SRY-box containing HMG transcription factor Sox17 to define the pancreato-biliary boundary in the ventral posterior foregut (Spence et al., 2009).
Here, we have used a publicly available BAC-clone from the GENSAT project (Gong et al., 2003), where the coding sequence of Enhanced Green Fluorescent Protein (EGFP) is inserted in the 5′UTR of the Hes1 gene to generate a transgenic mouse line, Tg(Hes1-EGFP)1Hri. We present the analysis of Hes1-EGFP expression in the developing mouse embryo from late streak stage around e7.0 with focus on the pancreas, the intestine, the liver, the kidneys, and the lungs. Our results confirm previous reports on Hes1 expression and reveal novel Hes1 expression domains.
To investigate the temporal and spatial patterns of Hes1 expression, we performed pronuclear injections to generate a transgenic mouse line expressing EGFP under control of Hes1 regulatory sequences. We used a BAC clone containing more than 224 kb of chromosome 16 including the Hes1 gene (Fig.1A). The insertion of the whole 224 kb fragment was verified by Southern blots with probes that recognise the ends (Fig. 1B, C). Immunoflourescent stainings for Hes1 and EGFP on adjacent sections confirm co-expression of EGFP with Hes1 in the inner neural layer and the outer pigment layer of the future retina in the e10.5 developing eye (Fig. 1D, E), in e10.5 developing pancreas epithelium (but not in the mesenchymal cells) (Fig. 1F, G), in roof plate (but not in the ependymal cells) in e14.5 neural tube (Fig. 1I, J), and in e14.5 oesophagus (Fig. 1J, K). We have used the Tg(Hes1-EGFP)1Hri mouse line to analyse for EGFP expression at different stages of development in selected tissues.
We first analysed e7.0 and e8.0-e8.5 stage embryos using whole mount immunohistochemistry. The e7.0 late streak stage embryos show uniform EGFP expression in the endodermal cell layer but is excluded from the mesoderm and ectoderm (Fig. 2A, B). A day later at e8.0-e8.5 just before embryo turning, we find a distinct area of EGFP expression in the posterior part, and another expression area in the anterior part of the embryo (Fig. 2C). In the tail region, we observe strong expression of EGFP in the neuroepithelium posterior to the forming somites (Fig. 2C, D). In addition, we find EGFP in the posterior definitive endoderm with an anterior border at the 4th somite pair (Fig. 2E-I), and in the presomitic paraxial mesoderm (Fig. 2C and J). The latter correlates well with previous data showing that Hes1 is a part of the segmental clock in the presomitic mesoderm (Jouve et al., 2000). Anteriorly, we find EGFP expression in the endoderm of the foregut and in the anterior intestinal portal (AIP) (Fig.2K-M).
Analysis of e9.0 embryos by whole mount immunohistochemistry shows continued expression of EGFP in the endodermal epithelium of the primitive gut tube. The dorsal Pdx1-positive pancreatic primordium shows strong EGFP expression and marks the anterior border of the posterior, dorsal endodermal expression domain at the level of the 4th somite pair (Fig. 3A). Only the dorsal part of the gut tube from the pancreas bud to the posterior part of the embryo is positive for EGFP, and at the most posterior end we also detect EGFP expression in the notochord (Fig. 3B). The ventral pancreatic progenitor cells marked by Pdx1 expression as well as the ventral part of the primitive gut tube posterior to the Pdx1 domain do not express EGFP at this stage (Fig. 3A). However, there is EGFP expression in the ventral foregut endoderm anterior to the ventral pancreas, but this does not appear to be in liver progenitor cells as there is no overlap between EGFP and the expression of Prox1 (Fig. 6A-C) (see section 2.3.3). At e9.0, we also observe Hes1-EGFP expression in the mesonephric ridge and the presomitic mesoderm (Fig. 3A). In the head region of the embryo, EGFP expression is detected in the otic pit epithelium (Fig. 3A) demonstrating Hes1 expression before the development of the primordial cochlea where expression of Hes1 mRNA has been reported previously (Murata et al., 2009). Additionally, the optic vesicles express EGFP (Fig. 3A), a prelude to the Hes1-EGFP expression previously described in the mouse e13.5 retina (Ohtsuka et al., 2006). We also find EGFP expression in the frontonasal process and in the pharyngeal region (Fig. 3A) corresponding to Hes1 mRNA expression described by Rochais et al. (Rochais et al., 2009). Moreover, there is EGFP expression in the neuroepithelium of forebrain, midbrain, and hindbrain (Fig. 3A).
Slightly later at e9.5, endodermal EGFP expression becomes confined to the dorsal pancreas anlage (Fig. 3C, D), whereas the more posterior dorsal gut tube epithelium as well as the ventral foregut epithelium both have ceased EGFP expression (Fig. 3C, D). EGFP expression in the mesonephric ridge becomes more pronounced (Fig. 3C).
At e10.5, the formation of many organs is more defined, and Hes1-EGFP expression has now become organ specific as seen by expression in the mesonephric tubules and in both pancreatic buds (Fig. 3E, F, F’). The EGFP expression is less intense in the dorsal pancreas bud compared to the mesonephros and even lower in the ventral pancreas bud (Fig. 3F’, inset). A strong EGFP signal remains in the tail region (Fig 3E), where continued Hes1 expression is required in the presomitic mesoderm for normal somite segmentation (Jouve et al., 2000), but also the notochord and the posterior hindgut epithelium show EGFP expression in the most posterior part of the tail bud (Fig. 3G). Additionally, we see EGFP expression in the epidermal epithelium of the budding forelimbs (data not shown) corresponding with previous observations for Hes1 mRNA (Rochais et al., 2009). In the head region of the e10.5 embryo, EGFP expression in the developing eye is now discernable to be in the inner neural layer of the optic cup (Fig. 3E). The neural EGFP expression is most pronounced in the telencephalic vesicle and the hindbrain, but also the neuroepithelium of the mesencephalon and the dorsal neural tube show significant EGFP expression (Fig 3E). Overall, our findings of neural EGFP expression in e10.5 embryos reflect previously described neural expression of Hes1 promoter activated EGFP (Ohtsuka et al., 2006).
Hes1 driven EGFP expression in the brain is well described by the GENSAT project (www.GENSAT.org) and in Ohtsuka et al. 2006 (Ohtsuka et al., 2006). We have therefore focused on the spatiotemporal expression pattern in different organs of the trunk, where the Hes1 expression pattern is poorly characterized.
At e10.5, the pancreatic structures are well defined and distinct from the stomach, bile duct, and duodenum. Here, clear EGFP expression is seen in the dorsal pancreatic bud epithelium (Fig. 3F, F’, 4A, D). The ventral pancreatic bud also displays EGFP expression although at a lower level (Fig. (Fig.3F3F inset, inset,4D).4D). Others have reported Hes1 immunoreactivity in the pancreatic mesenchyme at e11.5 (Seymour et al., 2007) and we see that as well when performing Hes1 immunostainings (Fig. 1F), but we do not detect Hes1-EGFP expression in the pancreatic mesenchyme neither at e10.5 nor at e12.5 (Fig. 4A, B, D, E). These data correlate well with previous in situ hybridisation data (Lammert et al., 2000) demonstrating Hes1 mRNA specifically in the e10.5 pancreas epithelium. However, we do see scattered EGFP expressing cells in the mesenchyme surrounding the duodenum (Fig. 4A, D, arrow heads).
At e12.5, EGFP is broadly expressed in the developing pancreas in both Pdx1 positive and Pdx1 negative cells (Fig. 4B, C, E) and the expression have reached similar levels in the dorsal and the ventral pancreas (not shown). This is in line with previous data showing co-expression of Hes1 and Pdx1 in mouse e13.5 pancreas (Esni et al., 2004). It has also previously been demonstrated that Hes1 expression is down regulated in glucagon positive cells in the e12.5 pancreas (Jensen et al., 2000), and we can confirm that many glucagon positive as well as insulin positive cells are negative for EGFP (Fig 4B, E, arrow heads), but we also see examples of cells double positive for EGFP and glucagon (Fig. 4E, arrow) which can be explained by the longer half life of EGFP compared to endogenous Hes1 mRNA and protein (Corish and Tyler-Smith, 1999; Hirata et al., 2002; Jouve et al., 2000). In the central epithelium, there is a quite uniform expression of EGFP whereas only a subset of cells in the tips of the forming branches are EGFP positive (Fig. 4C). We find that Carboxypeptidase A (Cpa1) is only expressed in tip cells with very low or no EGFP-expression (Fig. 4F). At e14.5, Hes1-EGFP expression is entirely restricted to the trunk epithelium and absent from the forming acini determined to become exocrine cells (Fig. 4G, K, O).
In the e17.5 pancreas, the EGFP expression is primarily observed in the central trunk epithelium, adjacent to insulin and glucagon expressing cells. EGFP is also seen in the exocrine tissue as scattered single cells or small clusters of 2-3 cells (Fig. 4H, L, P). These EGFP positive cell clusters all co-express the ductal marker Sox9 (Fig. 4P inset) (Seymour et al., 2007)
At birth (P0), we find most of the EGFP expression located in proximity to the islets of Langerhans and the pancreatic ducts, but never in the differentiated insulin or glucagon positive cells (Fig. 4I, M). EGFP positive single cells can be observed in the central part of many acini (Fig. 4I, M, U, arrow heads). Eight days after birth (P8), EGFP expression is lost in the ducts and around the islets (Fig. 4J, N) but the single, scattered EGFP positive cells are still seen at the base of the acini (Fig. 4J, V, arrow heads) and they co-express Sox9 (Fig. 4V inset) corresponding to the centroacinar cells in agreement with previous reports (Furuyama et al., 2010; Miyamoto et al., 2003). These cells can still be found in adult pancreas although they are very rare (Fig. 4X, arrow head and inset).
Immediately after closure of the gut endoderm at e9.0, we observe a Hes1-EGFP positive streak along the dorsal part of the prospective duodenum, posterior to the dorsal pancreas area (Fig. 3A). However, this expression disappears already at e9.5 (Fig. 3C, D), and at e10.5 there is only a few EGFP expressing cells left in the dorsal duodenal epithelium close to the pancreas (Fig. 5B), but many dispersed EGFP positive cells in the mesenchyme surrounding the developing duodenum and gut tube (Fig. 5A, B, arrow heads). At e12.5, there is a layer of EGFP expressing cells peripheral to the duodenal and midgut epithelium (Fig. 5C-E). These cells are located more peripherally than the smooth musculature marked by the expression of smooth muscle actin (Fig. 5C, E). Consistent with an absence of these cells from the e12.5 hindgut (data not shown), we observe Hes1 expression to be overlapping with beta-III-tubulin expression, which marks enteric neurons (Fig. 5D).
At e14.5, EGFP expression appears in some but not all cells in the duodenal epithelium, and in the periphery we find a rim of cells with a strong EGFP signal together with a weaker EGFP signal in cells closer to the epithelium (Fig. 5G). We find EGFP expression in the e17.5 epithelium (Fig. 5H) confirming previously published data showing Hes1 immunoreactivity in the intestinal villi (Jensen et al., 2000). However, we do not find EGFP expression restricted to the intervillus regions (future crypts) and within the villus mesenchyme as others have previously detected by in situ hybridisation at e18.5 (Schroder and Gossler, 2002). Also at birth, we observe EGFP expression in cells distributed throughout the villus epithelium (Fig. 5I), whereas it becomes more intense near the crypts at postnatal day 8 (Fig. 5J). In adult tissue, we find an irregular distribution of EGFP expression in the duodenum. Occasionally, there is a single crypt which shows profound EGFP expression (Fig. 5K), and in other regions we find areas with EGFP expression along the villi (Fig. 5L), but parts of the duodenal tissue are negative for EGFP.
To determine if the EGFP expression observed in e9.0 ventral foregut endoderm anterior to the ventral pancreas is in the developing liver bud, we have made co-immunostainings for the liver and pancreas progenitor cells marker Prox1 (Burke and Oliver, 2002). Although there is nice co-expression of EGFP and Prox1 in the developing dorsal pancreas, we do not detect any EGFP expression in the Prox1 positive cells in the ventral part of the gut tube (Fig. 6A-C). Also at e10.5, we do not detect any EGFP expression in the developing liver bud epithelium or hepatoblasts (Fig. 6D, E), but at e12.5 and e14.5 we find EGFP expression in the Sox9 positive common bile duct epithelium (Fig. 6F, G) and in Sox9 positive primitive ductal cells along the portal veins (Fig. 6H, I). At e17.5, almost all the biliary epithelium along the portal veins express EGFP (Fig. 6J), and in newborn liver, the Sox9 positive bile ducts co-express EGFP (Fig. 6K). At P8, we still find EGFP positive cells along the portal veins and some bile ducts appear to be asymmetrical with EGFP expression in half of the epithelium whereas other bile ducts are completely negative for EGFP (Fig. 6L, M). These results correlate well with recent lineage tracing studies on Hes1 expression (Kopinke et al., 2011) and previous work showing asymmetrical Hes1 expression at e15.5 and symmetrical Hes1 expression at e18.5 in the developing bile ducts (Antoniou et al., 2009). We do not find EGFP expression in the adult liver (Fig. 6N) or in the common bile duct at e17.5 and older (data not shown).
We detect an incipient EGFP signal in the mesonephric ridge of e9.0-e9.5 embryos (Fig. 3A, C) that develops into strong EGFP expression in the mesonephric tubules at e10.5 (Fig. 3E, F and and7A)7A) and e12.5 (Fig. 7B). At e14.5, Hes1-EGFP expression is observed in the tubules, the comma-shaped and the s-shaped bodies and the capsules of the maturing glomeruli (Fig. 7C). This confirms the in situ hybridisation results reported by others at e13.5 (Chen and Al-Awqati, 2005; Piscione et al., 2004), except that we do not detect any EGFP expression in the condensed mesenchyme. At e17.5 and P0 we find EGFP expression primarily in stage I-III nephrons in the cortical region and in the collecting ducts, but also among the undifferentiated mesenchymal cells in the medullary region (Fig. 7D, D’, E, E’).
A shift in Hes1 promoter activity can be observed between P0 and P8 where the EGFP expression is lost in the nephritic epithelia within the nephrogenic zone as the kidneys become close to fully developed. Only the EGFP expressing cells in the medullary interstitium remain (Fig 7F, F’). Kidney tissue from adult is completely devoid of EGFP expression (data not shown).
Hes1 mRNA expression has previously been detected by northern blot in foetal lungs from e12.5 to e18.5 (Ito et al., 2000). We find that the lung buds at e12.5 are mainly negative for EGFP expression except for a weak signal in the main bronchi epithelium (Fig. 8A). However at e14.5 and e17.5, we detect EGFP in the epithelial cell layer lining the bronchioles (Fig. 8B, C) correlating well with reported immunohistochemical detection of Hes1 protein in the bronchiolar epithelium at e16.5 (Collins et al., 2004). This expression pattern continues through birth (Fig. 8D) and into adulthood (Fig. 8E, F).
Hes1 is expressed in multiple tissues during development, including organs derived from all three germ layers such as the brain, kidneys and pancreas (Chen and Al-Awqati, 2005; Jensen et al., 2000; Ohtsuka et al., 2006). In our hands, Hes1 expression is difficult to detect reliably by in situ hybridisation and immunohistochemical techniques, and the reports on Hes1 expression are often sparse and incomplete. Here we report a detailed analysis of the Tg(Hes1-EGFP)1Hri mouse strain that allows for easy investigation of Hes1 expression. We assume that most if not all Hes1 regulatory sequences are included in the BAC as it contains approximately 178 kb genomic sequence upstream and 43 kb downstream of the Hes1 coding region. There are small discrepancies between EGFP expression and the immunohistochemical detection of Hes1 protein, and this can be due to the polyclonal antiserum not being absolutely specific to Hes1, or that it may take longer time for EGFP than Hes1 to reach detection levels, or the fact that EGFP has a longer half life than the tightly regulated Hes1 transcription factor, or that certain distant regulatory sequences is not included in the transgenic mouse. But in general, the results confirm previously described expression patterns of Hes1 mRNA and Hes1 protein, indicating that the transgene has integrated in a region accessible for transcriptional regulation throughout development. There is no evidence for insertional mutagenesis as the mice breed well and develop normally both as heterozygotes and homozygotes.
In several tissues, Hes1 mRNA and protein levels are tightly regulated and oscillates in a two hour cycle where Hes1 itself plays a role as a repressor of its own expression (Hirata et al., 2002; Kobayashi et al., 2009). This requires a very short half life of Hes1 compared to the relatively stable EGFP which has a half life of approximately 26 hours (Corish and Tyler-Smith, 1999). Therefore, the EGFP expression in this transgenic mouse line will reflect Hes1 expression sites but not the post-trancriptional regulation.
Our analyses describe previously uncharacterised Hes1 expression in the endoderm of late streak stage (e7.0) and in 5-6 somite (e8.0) embryos as well as in e9.0 dorsal gut tube epithelium. There is an anterior border of the caudal expression domain around the 4th somite pair that seems to stay fixed during growth of the embryo until e9.0 where it correlates with the anterior border of Pdx1 expression outlining the foregut/midgut boundary (Zorn and Wells, 2009). The ventral foregut also shows Hes1 expression bordering the liver diverticulum determined by Prox1 expression and thus not including the prospective liver and the ventral pancreas. This suggests that Hes1 may have a role in patterning of endodermal cells even before gut tube closure. The broad Hes1 expression in the dorsal endoderm then becomes restricted to the dorsal pancreas epithelium at e9.5. Sox17 has been shown to regulate ventral foregut patterning in conjunction with Hes1, where the timely appearance of Hes1 is essential in order to repress Sox17, thereby defining the pancreato-biliary border with a Sox17 positive biliary part and a Hes1 positive pancreatic part (Spence et al., 2009). This model would predict that Hes1 mutants should display pancreas to bile duct conversion, but actually the opposite seems to occur (Sumazaki et al., 2004). The entire pancreatic epithelium shows Hes1 expression at e10.5, but as the pancreas starts to grow by branching morphogenesis at e12.5, Hes1 expression becomes restricted to the trunk epithelium and disappears from the Cpa1 positive cells. At e14.5, the tip cells destined to become exocrine acini are completely devoid of Hes1 expression, which corresponds well with Hes1 keeping pancreatic cells in a progenitor state. After the burst of endocrine differentiation at the secondary transition and towards the end of gestation, Hes1 expression declines and in general we do not detect Hes1 expression in the differentiated insulin or glucagon positive cells. However, we do occasionally find a few cells that are double positive for EGFP and a hormone, which is likely due to the longer half life of EGFP relative to Hes1 protein and mRNA. As expected we detect Hes1 expression in the centroacinar cells (Miyamoto et al., 2003).
Looking at other selected organs during embryonic development, we detect Hes1 expression in neuronal class III beta-tubulin positive enteric neurons in the duodenal epithelium. Similarly, we observe an equivalent distribution of EGFP positive cells in the stomach mesenchyme at all stages examined (data not shown). In the duodenal/intestinal epithelium, the Hes1 gene becomes active between e12.5 and e14.5 and is later found to be largely restricted to the proliferating crypt cells in neonatal and young animals. During liver development, we find the earliest Hes1 activity at e12.5 in the common bile duct epithelium and at the first stages of intrahepatic bile duct differentiation. Here, Sox9 has shown to be critical for bile duct formation (Antoniou et al., 2009) and it also marks progenitor cells in adult liver (Furuyama et al., 2010). In agreement with the work by Antoniou et al. (Antoniou et al., 2009), we find that Hes1 activity in the biliary epithelium co-localise with Sox9 expression during first the asymmetrical phase and later during the symmetrical phase of tubulogenesis, whereas mature bile ducts do not show Hes1 expression. The kidneys show Hes1 gene activity from the very early time points in the mesonephric ridge at e9.0, then in the mesonephric tubules, and later widely distributed in the first developmental stages of the nephrons. This correlates well with the more detailed descriptions of Hes1 mRNA detection in the nephrons by Chen et al. and Piscione et al. (Chen and Al-Awqati, 2005; Piscione et al., 2004). However, there are some small discrepancies since we do not see Hes1 activity in the condensed mesenchyme in the outer cortex as it has been reported at e13.5 by both groups (Chen and Al-Awqati, 2005; Piscione et al., 2004). This may reflect the delay in detection of EGFP expression compared to endogenous Hes1. In addition, the postnatal Hes1 gene activity that we find in the medullary interstitium has not been described before. We expected to find Hes1 gene activity in the developing lung buds at e12.5 as has been reported by Hes1 northern blot analysis (Ito et al., 2000), but we could only detect a faint signal in the main bronchi and nothing in the segmental bronchi at that stage. Subsequently, we detect Hes1 expression in the bronchiolar epithelia from e14.5 and onwards.
The results in this study support the concept of Hes1 as a key component in progenitor cells as it is observed in the trunk epithelium in the developing pancreas. This agrees with the reports on pancreas hypoplasia in Hes1 mutant mice (Jensen et al., 2000). Hes1 expression in the developing common bile duct is in agreement with incorrect development of the common bile duct in Hes1 mutant mice (Sumazaki et al., 2004). Also in intrahepatic bile duct development, Hes1 seems to play a role, but it has not been established whether it is in a Notch dependent manner (Antoniou et al., 2009). Along the same lines in the kidneys, it appears that Hes1 primarily is active during development of the tubules, but it is not clear which role Hes1 plays here.
The Tg(Hes1-EGFP)1Hri mouse reported on here will be a useful tool to study the role of Hes1 in a number of organ systems also allowing for prospective cell sorting.
From the BACPAC Resources Center (http://bacpac.chori.org/) we obtained a BAC clone, GENSAT1-BX38, deposited by the GENSAT project (http://gensat.org/). The BAC contains 224,313 bp of chromosome 16, including 178 kb upstream of the transcriptional start site of the Hes1 gene and 43 kb downstream of the Hes1 gene. It also contains a 1.1 kb EGFP-PolyA fragment inserted into the 5′UTR of the Hes1 gene, 27 bp upstream from the Hes1 translation start codon. DNA from the BAC clone was prepared from E. coli cells using the NucleoBond® Xtra midi kit (MACHEREY-NAGEL), and purified according to (Nagy, 2003). The purified BAC DNA was resuspended in sterile filtered injection buffer (Moreira et al., 2004) to a final concentration of 1.5 ng/μl, kept on ice until injection, which was performed the same day. The DNA was injected into the male pronucleus of B6D2F2 zygotes which were then introduced into pseudopregnant NMRI female mice. Tail biopsies from 13 male offspring were positive for EGFP by PCR analysis, and all were mated with C57BL/6 female mice. Germ line transmission of EGFP expression in the embryos was analyzed by PCR (Primers: Hes1 5′-UTR: 5′-CGAGCGGTGCCGCGTGTCTCTTCCTCCC-3′; EGFP antisense: 5′-CGGCGAGCT GCACGCTGCCGTCCTC-3′) and visually by fluorescence microcopy. Only one male founder gave rise to EGFP positive offspring. C57BL/6 female mice were time mated with heterozygous mutant males and EGFP positive embryos were identified by PCR using the primer pair described above. All mouse work conforms to the European Convention for the Protection of Vertebrate Animals used for Experimental and other Scientific Purposes (ETS 123).
The two probes for verification of insertion of the BAC ends were amplified by PCR from purified BAC Hes1-EGFP DNA. Primers: Probe1: 5′-GGA TTG TTT GTC TAG GAT TTG AGG-3′ and 5′-GGA AAG GAC ACT CTG ACC TGT T-3′ (approx. 500 bp). Probe2: 5′-CTT TTT TGA TTT AGT GCA ATG CAC C-3′ and 5′-GGT AGT GCA AGA CAA ACA AAC AAG-3′ (approx. 500 bp). 3μg genomic DNA samples from a wild-type, a heterozygous, and a homozygous mouse were digested with BglII and EcoRV and subjected to standard Southern blot analysis with 100 ng radio labelled Probe1. Likewise 3μg genomic DNA samples were digested with NdeI and subjected to standard Southern blot analysis with 100 ng radio labelled Probe2.
Whole mount immunofluorescent stainings were performed according to (Ahnfelt-Ronne et al., 2007) using the antibodies described below.
Embryos were dissected in cold PBS and fixed O/N at 4°C in Lilly’s formalin buffer, pH 7.4 (Bie & Berntsen, Denmark). The embryos were cryo protected O/N at 4°C in 30% sucrose in PBS and then embedded in Tissue-Tek (Sakura, Denmark). The tissue was cryo sectioned at 10μm. Immunostainings were performed according to (Klinck et al., 2008) using the antibodies described below.
Primary antibodies: Rabbit anti-GFP “Living Colors” (Clontech) diluted 1:2000. Goat anti-GFP (Abcam) diluted 1:1000. For the whole mount GFP stainings a Tyramide Signal Amplification (TSA) step was applied (PerkinElmer,USA). Rabbit anti-Hes1 1:800 with TSA (kind gift from Dr. T. Sudo), Goat anti-Sox17 (R&D systems) diluted 1:1000. Guinea Pig anti-Insulin (Abcam) diluted 1:500. Goat anti-Pdx1 (kind gift from Chris Wright), diluted 1:10.000. Mouse anti-E-cadherin (BD Biosciences) diluted 1:1000. Rat anti-E-cadherin (R&D Systems) diluted 1:1000. Guinea pig anti-Glucagon (Millipore) diluted 1:10000. Goat anti-Carboxypeptidase A (R&D Systems) diluted 1:1000. Mouse anti smooth muscle actin (DAKO) diluted 1:500. Rabbit anti-beta-III-tubulin (TuJ1) (Bio Site) diluted 1:1000. Rabbit anti-Sox9 (Millipore) diluted 1:500. Goat anti-HNF4a (Santa Cruz) diluted 1:100. Secondary Antibodies: Whole IgG biotin anti-rabbit (cat. #711-065-152; for TSA amplification), Cy2-anti-goat (cat. #705-225-147), Cy3-anti-rabbit (cat. #711-165-152), Cy3-anti-rat (Cat. No. 712-165-153), Cy5-anti-mouse (cat. #715-175-151), all from Jackson ImmunoResearch Laboratories raised in donkey. Diluted 1:500.
Images were obtained using a Zeiss LSM510 META Axio Imager connected to a LSM 510 laser module with the following lasers: Ar laser (488 nm; 50% laser power), He/Ne laser (543 nm), and He/Ne laser (633 nm) (Carl Zeiss; Germany). Embryos stained as whole mount were cleared in BABB (benzyl alcohol:benzyl benzoate1:2) before scanning. Alignment of images to produce the composite e10.5 image was done manually with Canvas 9 (ACD systems). List of Objectives: Plan-Neofluar 10x/0.3, WD 5.5 mm; Achroplan 20x/0.5 W Ph2, WD 7.9 mm; Plan-Neofluar 25x/0.8, WD 0,21; EC Plan-Neofluar 40x/1,30 Oil DIC, WD 0.21 mm.
We thank Karsten Skole Marckstrøm, Lene Petersen, Anette Bjerregaard and Lisbeth Ahm Hansen for technical assistance, Chris Wright for the anti-Pdx1 antibody, and Dr. T. Sudo for the anti-Hes1 antibody. This work was made possible by support from the JDRF (10-2008-580 and 1-2009-308), the NIH (DK072495), and the EU 6th Framework Program.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.