In vertebrates and in some higher invertebrates, the establishment of left-right patterning is directed by the Nodal signaling cascade, which involves the Transforming Growth Factor β-like molecule Nodal, its antagonists Cerberus/Dan and Lefty, and the transcription factor Pitx2. During early development, Nodal signaling directly activates the expression of Nodal itself, Lefty2 and Pitx2 in the left lateral plate mesoderm (LPM). This process is mediated by the transcription factor FoxH1, which recognizes conserved sequence motifs in the asymmetric enhancer (or ASE) of those genes. Therefore, Nodal signaling is amplified by self-induction, but is also strictly limited in space and time due to the feedback inhibition by Lefty.

In zebrafish, mouse and human, 2 Lefty genes have arisen by independent duplications. In the mouse embryo, Lefty1 is expressed in the midline (floor plate and notochord), where it prevents Nodal signaling from spreading to the right side, whereas Lefty2 is expressed in the left LPM, where it leads to the downregulation of Nodal signaling. In the chick, however, a single Lefty gene has been identified, Lefty2, which is expressed in both the midline and the left LPM. Although the role of Lefty2 as an inhibitor of Nodal signaling appears to be conserved in the chick embryo, it is currently unclear whether the expression of chick Lefty2 is also regulated by a Nodal-responsive enhancer. In the present study, we addressed this question by investigating the presence of an ASE enhancer in the cis-regulatory region of chick Lefty2 gene.

To identify the cis-regulatory region of chick Lefty2 gene responsible for driving asymmetric expression in the left LPM.

The asymmetric expression of mouse Lefty2 and human LEFTY1 and LEFTY2 genes is regulated by ASE enhancers that are found in their upstream regions and contain two FoxH1 binding sites each. We therefore investigated the presence of such binding elements in the upstream region of chick Lefty2 gene. A 3.0 kilobases (kb) genomic sequence upstream of the coding region was analyzed using TRANSFAC Patch 1.0 and MacInspector Release professional 8.4.1. This motif search identified 3 potential FoxH1 binding elements (AATC/ACACAT) closely located at -1709 to -1436 base pairs (bp) upstream of the chick Lefty2 initiation codon (Fig. 1A).

To determine if this region could be the ASE enhancer of chick Lefty2, we assessed its ability to drive transcription specifically in the left LPM, where chick Lefty2 is asymmetrically expressed (Fig. 1B). For this, the 274 bp DNA fragment was subcloned into an enhancer-less vector containing the human β-globin minimal promoter upstream of the eGFP coding sequence (i.e., ASE-eGFP; Fig. 1A), and introduced into chick embryos by electroporation in New culture. Together with the ASE-eGFP construct, embryos were co-transfected with the ubiquitous reporter pCAGGS-RFP to control for targeted area and electroporation efficiency. Our results showed that eGFP expression is specifically restricted to the posterior LPM on the left side (Fig. 1C; 17/18 embryos), mirroring the asymmetric expression pattern of chick Lefty2 (Fig. 1B). eGFP fluorescence starts to be detected approximately 2–3 h after the initial detection of Lefty2 transcripts in this region (e.g., stages HH8+ vs. HH8), which corresponds to the time required for eGFP gene transcription and protein synthesis. ASE-eGFP expression remains in the posterior region of the left LPM until the last stage analyzed (HH11; data not shown), in a similar pattern to Lefty2 asymmetric expression. However, eGFP expression was not clearly detected in the notochord domain of Lefty2 expression at any of the developmental stages tested (HH7-11). These observations suggest that chick Lefty2 ASE regulatory region is indeed a typical ASE enhancer, being able to specifically drive expression in the left LPM. Moreover, the presence of 3 FoxH1 binding elements indicates that chick Lefty2 expression may be directly induced by Nodal signaling in the left LPM, as previously shown for mouse Lefty2. In addition, our results indicate that chick Lefty2 expression in the midline is not induced by the ASE enhancer. Similarly to the transcriptional regulation of mouse Lefty1, midline expression may be under the control of regulatory sequences that do not contain FoxH1 motifs.

Unlike mouse Lefty2, the expression of chick Lefty2 is not detected in the left LPM at early stages. Based on their similar expression patterns and functions, it was proposed that the role of mouse Lefty2 has been taken by the Nodal antagonist Cerberus (Cer1) in chick left-right patterning. Indeed, chick Cer1 is expressed in the left LPM at early stages (HH7-9), and its transcription is also regulated by a Nodal-responsive enhancer containing two essential FoxH1 binding elements. Taken together, our results suggest that chick Lefty2 may replace Cer1 in the left LPM at later developmental stages (HH8-11) as a feedback inhibitor of Nodal signaling.

We have identified the asymmetric enhancer (ASE) of chick Lefty2 gene. This ASE sequence contains three conserved Nodal-responsive elements and is able to drive transcription specifically in the left lateral plate mesoderm, thus reproducing the asymmetric pattern of chick Lefty2 expression.

Although the identification of FoxH1 binding sites in the ASE enhancer suggests that chick Lefty2 transcription is regulated by Nodal signaling, further experimental evidence is required. Namely, functional studies are needed to assess if Nodal misexpression on the right side ectopically induces chick Lefty2 expression, whereas Nodal inhibition on the left side represses it. Moreover, as in the study of chick Cer1 transcriptional regulation, a mutagenesis analysis of the ASE enhancer would reveal whether the FoxH1 sites are indeed essential for the induction of left LPM expression. Taken together, these experiments would bring support to the hypothesis that, as in other chordates, Lefty2 is a direct target of Nodal signaling in the chick embryo.

Despite having similar asymmetric enhancers, the patterns of chick Lefty2 and Cer1 in the left LPM are slightly different, with Lefty2 expression starting at later developmental stages. Therefore, the question arises as to how Nodal signaling is able to differently regulate the transcription of Cer1 and Lefty2. One possible explanation may rely on the number of FoxH1 binding sites in their ASE enhancers: 2 in Cer1 and 3 in Lefty2 regulatory region (this study). In fact, the number of transcription factor binding sites in homotypic clusters is known to influence gene regulation. In particular, if transcription is activated only when all binding sites are occupied, having a higher number of sites would generate a time delay in gene expression. A simple way to address this possibility would be to evaluate the effect of mutating each of the FoxH1 binding elements in the chick Lefty2 enhancer in ASE-eGFP reporter assays.

Sequence analysis

The 3.0-kb genomic sequence upstream of the chick Lefty2 coding region was analyzed in silico using the software programs TRANSFAC Patch 1.0 (www.gene-regulation.com) and MacInspector Release professional 8.4.1 (www.genomatix.de) to identify potential transcription factor binding sites. The 274 bp region located between nucleotides -1709 and -1436 bp upstream the ATG (Chr3:17278480-17278753), which contains three FoxH1 binding elements, was termed the asymmetric enhancer (ASE) of chick Lefty2 gene.

DNA constructs

Lefty2 ASE enhancer was amplified by PCR from chick genomic DNA (forward primer: 5'-CTGGAGCTCACACCCTGAATGCACCATGG-3'; reverse primer: 5'-GACACTAGTACCAGGATGAAATCTCTCCC-3') and sub-cloned into the SacI/SpeI restriction sites of the p1229-eGFP enhancer-less vector, which contains the human β-globin minimal promoter and the eGFP coding sequence. The pCAGGS-RFP vector was used as a positive control for the efficiency and extent of transfection. This vector carries the coding sequence of monomeric RFP driven by the ubiquitous CAGGS promoter.

Chicken embryos

Fertilized chicken eggs were incubated for the appropriate period in a humidified incubator at 37.5°C. Embryos were staged according to Hamburger and Hamilton (HH).

Whole-mount in situ hybridization

Chick embryos were collected at stages HH7-12, fixed overnight in 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS) plus 0.1% Tween® 20 (PBT) at 4°C, dehydrated through a series of methanol solutions (25%, 50%, 75% and 100% methanol/PBT), and stored at -20°C until processed for hybridization. For this, embryos were rehydrated in PBT and bleached in 6% hydrogen peroxide in PBT for 1 h. Embryos were then rinsed 3 times in PBT, digested with proteinase K (10 μg/ml in PBT) for 5 min at room temperature, washed once in 2 mg/ml glycine/PBT and twice in PBT, and post-fixed in 4%PFA/0.2% glutaraldehyde/PBT for 20 min at room temperature. After fixation, embryos were rinsed twice in PBT and pre-hybridized for 2 h at 70°C in pre-hybridization solution (50% formamide, 5× standard saline citrate (SSC), pH 5, 0.1% Tween® 20, 50 μg/ml heparin), and incubated overnight at 70ºC in hybridization solution (pre-hybridization solution with 50 μg/ml torula RNA and 50 μg/ml salmon sperm DNA) containing 500 ng/ml of denatured riboprobe. The chick Lefty2 antisense riboprobe was generated by in vitro transcription in the presence of Digoxigenin-UTP (Roche), using the complete cDNA as template.

On the second day, embryos were washed twice in 50% formamide/4× SSC/1% SDS at 70°C and twice in 50% formamide/2× SSC at 65°C for 30 min each. Embryos were then rinsed 3 times in MABT (100 mM maleic acid, 150 mM NaCl, pH 7.5, 0.1% Tween®), blocked in 10% goat serum in MABT for 2 h at room temperature, and incubated in 1% goat serum in MABT with 1:5000 alkaline phosphatase-coupled anti-Digoxigenin antibody (Roche) overnight at 4°C. On the third day, embryos were washed 5 times in MABT for 45 min each, rinsed twice in NTMT (100 mM NaCl, 100 mM TrisHCl, pH 9.5, 50 mM MgCl2, 0.1% Tween®) for 10 min each, and incubated in BM Purple solution (Roche) in the dark for up to 12 h. Stained embryos were rinsed in PBT, fixed overnight in 4% PFA/PBT and stored in PBS at 4°C until imaging.

Embryo electroporation

HH3-4 chick embryos were processed for ex ovo culture and placed inside a silicone pool containing a 2 mm square cathode (CY700-1Y electrode; Nepa Gene). After being covered with warmed Hank’s buffer (Gibco-BRL), each embryo was injected with DNA solution (3 mg/ml ASE-eGFP; 0.5 mg/ml pCAGGS-RFP; 0.1% Fast Green; Sigma) using a IM-300 microinjector (Narishige), and electroporated with 5 pulses (10 V for 50 ms at 350 ms intervals) using a 2 mm square anode (CY700-2 electrode; Nepa Gene) and a square wave electroporator (TSS20 Ovodyne; Intracel). Embryos were then transferred to 35 mm Petri dishes filled with 1.5 ml of albumen, incubated at 37.5ºC and imaged at stages HH7-11.

Image acquisition

Whole chick embryos were observed under a Zeiss SteREO Lumar.V12 fluorescence stereomicroscope and photographed with a Zeiss MRc.Rev3 color camera and ZEN 2 Pro software (Carl Zeiss). Bright-field and fluorescence images were assembled using Adobe Photoshop (Adobe Systems).

This work was supported by Fundação para a Ciência e a Tecnologia (POCTI/BME/46257/2002 and SFRH/BPD/102261/2014).

We are grateful to Sofia Andrade and Telmo Pereira for technical assistance, Juan Carlos Izpisúa Belmonte for the chick Lefty2 riboprobe vector, and Domingos Henrique for the pCAGGS-RFP construct.

Not Applicable.