RhoA is highly up-regulated in the process of early heart development of the chick and important for normal embryogenesis

Abstract

We have used molecular techniques, combined with classic embryological methods, to identify up-regulated genes associated with early heart development. One of the cDNAs identified and isolated by screening a chick lambda cDNA library was the small guanosine triphosphatase RhoA. RhoA has at least three different length mRNA species, each varying in the length of the 3′ untranslated region. In situ hybridisation and immunocytochemistry analysis of RhoA expression show marked up-regulation in the heart-forming region. In other systems, RhoA signalling has been shown to be important for both gene expression and morphology. To investigate the function of RhoA in early heart development, we used small interfering RNAs (siRNA) in early chick embryos. Disruption of RhoA expression by siRNA treatment resulted in lack of heart tube fusion and abnormal head development. These data indicate that RhoA is important for normal embryogenesis. Developmental Dynamics 227:35–47, 2003. © 2003 Wiley-Liss, Inc.

INTRODUCTION

The heart is the first organ to be formed in vertebrates. The primordium of this organ is derived from mesodermal cells, which are involved in a complex series of morphogenic interactions with cells from several embryonic origins (Olson and Srivastava, 1996; Harvey, 1996, 1999; Zaffran and Frasch, 2002). Soon after gastrulation, myocardial precursor cells migrate to the anterior lateral mesoderm and are committed to the cardiac lineage in response to signals thought to originate from the adjacent endoderm (Schultheiss et al., 1995). The cardiac progenitor cells are developed further into cardiac primordia, which then fuse to form the primitive cardiac tube (Olson and Srivastava, 1996). In higher vertebrates, the tubular heart undergoes looping and later septation, which results in the formation of an integrated four-chambered organ. After cardiac looping, cardiac myocytes express muscle genes, the specific set of which is characteristic for each chamber, thereby yielding chamber-specific properties (reviewed in Zaffran and Frasch, 2002).

The events of early cardiogenesis have been well defined in terms of classic embryology using techniques of microdissection and transplantation or explant culture (Sater and Jacobson, 1990), usually by using avian or amphibian embryos. These studies have resulted in the generally accepted view of induction of the lateral plate mesoderm by the pharyngeal endoderm (or its precursor, the dorsal blastopore lip). At a molecular level, the commitment of vertebrate cardiogenic mesoderm is being elucidated and many new studies are implicating various growth and transcription factors (Lough and Sugi, 2000; Abu-Issa et al., 2002; Chen et al., 2002; Lopez-Sanchez et al., 2002; Nemer and Nemer, 2002; Szeto et al., 2002).

With this in mind, we set out to develop a protocol for identifying genes whose activities are specifically up-regulated during cardiogenesis. The larger size, greater accessibility, and ease of manipulation of the early chick embryo, and well-defined stages by Hamburger Hamilton (HH; Hamburger and Hamilton, 1951) have provided a strong rationale for use of this model in this study. The protocol developed involved differential amplification and differential display (DD) of gene fragments (DD-PCR [PCR, polymerase chain reaction]), developmental dot blots to confirm up-regulation, and the generation of a full-length cDNA phage library. To isolate and characterise genes that are associated with early heart development, this library was screened with DD-PCR fragments from the differential display that had been shown to be up-regulated in anterior lateral plate mesoderm (LPM) and heart primordium (HP) (Kaarbø et al., 2000; Sneesby et al., 2001). One of the first cDNAs that was isolated by cDNA library screening by using the DD-probes and shown to be up-regulated, encodes the small guanosine triphosphatase (GTPase) RhoA. Four different transcripts of RhoA, varying in the length of the 3′ untranslated region (UTR), were identified. In situ hybridisation and immunocytochemistry analysis revealed marked up-regulation of RhoA in the heart primordium (HH stage 5 onwards).

Rho GTPases act as molecular switches controlling a variety of signal transduction pathways in eukaryotes. They use a simple biochemical strategy, cycling between two conformational states—active (bound to GTP) and inactive (bound to GDP), reviewed by Etienne-Manville and Hall (2002). Formerly principally known for their role in regulating the actin cytoskeleton, they are now known to also influence cell polarity, microtubule dynamics, membrane transport pathways, and transcription factor activity. It is thought that, by the switching on of a single GTPase, several distinct signalling pathways can be coordinately activated. We show here that disruption of RhoA expression in vivo by using small interfering RNA (siRNA) injected into anterior lateral plate mesoderm caused severe defects in cardiac tube fusion and head formation. These results indicate that RhoA expression is necessary for the development of normal heart and head morphology in the early chick embryo.

RESULTS

RhoA Is Highly Up-Regulated in Early Heart Development and Appears as Different Length Transcripts Varying in the 3′-UTR

DD amplicon 54 represents a gene up-regulated during early heart development.

To identify significantly up-regulated gene transcripts involved in the process of cardiogenesis, cDNA fragments were isolated by differential display and subtractive hybridisation. From the differential display, the DD54 cDNA fragment was not evident in posterior mesoderm (HH stage 4; PM), expressed moderately in anterior LPM at the level of Henson's node (HH stage 4), and markedly up-regulated in HP (HH stage 7–9), while at the same time appearing in the subtracted molecular libraries LPM-PM and HP-LPM and as well evident in the artificially induced posterior section (Psi; Fig. 1A).

Figure 1. Differential display identifies a gene (DD54) that is markedly up-regulated during heart development. A: Multiple DD–polymerase chain reaction products when generated from cDNA derived from various tissues in the developing chick embryo. A section has been magnified to indicate bands involved in the selection of fragment 54 (indicated by arrow). This fragment is present in LPM, HP, and PSi cap-cDNA samples, as well as in the subtracted cap-cDNA samples LPM-PM and HP-LPM. B: Developmental dot blots. Plasmid DNA with DD54 fragment was transferred to nylon membranes and hybridised to ³²P-labeled cDNA from tissues representing three stages of development in the early chick embryo. PM, posterior mesoderm; LPM-PM, cDNA derived from RNA remaining in LPM RNA after subtracted hybridisation with PM cDNA; LPM, lateral plate mesoderm; HP-LPM, cDNA derived from RNA remaining in HP RNA after subtracted hybridisation with LPM cDNA; HP, heart primordium; PSi, artificially induced posterior mesoderm.

Confirmation of developmental up-regulation of DD54 mRNA.

Northern hybridisation is normally used to confirm authenticity of cloned amplicons. As the amount of tissue required to obtain sufficient RNA for blots was necessarily limited, we devised a method whereby dot blots of selected DD-derived cDNA-containing plasmids, as well as one control plasmid containing actin cDNA, were probed by ³²P-labelled cDNA of the three pertinent tissue sources. By these analyses, DD54 mRNA expression was indicated as being up-regulated in LPM and HP compared with PM. Actin levels were also measured to indicate cDNA probe labelling efficiency in each case (Fig. 1B).

DD54 is identified as RhoA cDNA.

DD54 was used as a probe to screen a stage 6–9 heart primordium cDNA library. The sequence of the cDNA isolated encodes for chick RhoA. Four different mRNA species that differ in the length of the 3′-UTR were identified. These species have been designated RhoA_1,2,3, and ₄ and have the following total lengths (minus polyA tails): 1249, 1371, 1561, and 1726 bp. Each species appears to be generated by alternative putative polyadenylation signals. Figure 2 depicts the sequence of the longest RhoA transcript isolated (aligned with the human sequence), which includes 48 bp of 5′-UTR, 579 bp of open reading frame (ORF), and 1124 bp of 3′-UTR (Genebank accession no: AJ294705). The putative alternative polyadenylation signals are indicated in bold and underlined.

Figure 2. Clustal W multiple sequence alignment result of chick RhoA and human protooncogene RhoA. Stop codon and putative polyadenylation signals are highlighted and underlined. Chick sequence includes the identified RhoA cDNA species and is of the largest identified RhoA cDNA (RhoA₄). This sequence includes 48 bp of 5′ untranslated region (UTR), 582 bp of open reading frame, and 1096 bp of 3′-UTR sequence. The start codon (ATG), stop codon (TAA), and putative alternative polyadenylation signals are shown in bold type and underlined.

Moscow and coworkers have reported multiple length transcripts of RhoA mRNA in human breast cancer cell lines, and three polyadenylation signals were identified that terminate these RhoA transcripts (Moscow et al., 1994). Alignment of the human RhoA protooncogene cDNA and the Gallus gallus RhoA cDNA (Fig. 2) shows substantial sequence identity in the 3′-UTR. The three polyadenylation signals identified in human are the same as three of the putative polyadenylation signals in chick. As well, there appears to be highly conserved sequence preceding these three polyadenylation signals.

Differential expression of RhoA transcripts.

As described above, DD54 was designated as RhoA₁ and used to isolate three other putative RhoA transcripts from the chick cDNA library. To confirm expression of the different transcripts of RhoA in earlier stages of the chick embryo and investigate whether the expression of particular length transcripts had developmental significance, PCR of chick embryonic cDNA was used to amplify products representing the different length RhoA transcripts. RNA was extracted from tissue from the heart-forming region of lateral plate mesoderm (HH stages 4–5), heart primordium (HH stages 7–9), and from the noncardiogenic region, the posterior mesoderm, for both HH stages. All four transcripts were reproducibly detected in all four tissue sources, with the putative RhoA₁ at a very low level, often only evident as a diffuse band just beneath RhoA₂. Only the RhoA₂ and RhoA₃ products were able to be cloned; sequencing of these confirmed the expected sequence in each case. Both the no template control and no reverse transcription (RT) controls were negative, indicating no contamination with genomic DNA (results not shown). The different sizes of the transcripts amplified by PCR, approximately 670, 790, 980, and 1,140 bp for RhoA_1–4, respectively, are consistent with the different length transcripts identified from the cDNA library. It should be noted that the putative RhoA₃ cDNA might not be a true reverse transcript. As can be seen in Figure 2, the polyT primer used may have primed from the polyA sequence 3′ of the third putative polyadenylation signal sequence, thus producing a falsely truncated cDNA during library construction as well by the RT-PCR experiment. Comparison of the band intensity of the different PCR products suggests that RhoA₂ is the most highly expressed transcript in all tissues tested. The relative level of expression of the different transcripts (within a tissue location) is indicated as having changed between the heart-forming regions LPM (stage 4–5; Fig. 3, lane 2) and HP (stage 7–9; Fig. 3, lane 4), whereas there was no difference in proportional levels of amplicons in PM for both stages tested (Fig. 3, lanes 3 and 5). These findings suggest that RhoA₂ is up-regulated (compared with other length transcripts) coincident with early heart development.

Figure 3. Polymerase chain reaction (PCR) products of different RhoA transcripts amplified from chick embryonic cDNA. The 5′-primer used was designed to anneal to the beginning of the 3′ untranslated region and common for all the different RhoA transcripts. The 3′-primer used was designed to anneal to the different transcripts reversed transcribed using a T7-(dT)₂₄ primer. The products generated were resolved by electrophoresis on a 1% agarose gel. A total of 1 μg of 1 kb+ ladder was used as a standard; the mobility of the PCR products is indicated by the respective transcript name. Lane 1, 1 kb plus ladder; lane 2, LPM (HH 4–5); lane 3, PM (HH 4–5); lane 4, HP (HH 7–9); lane 5, PM (HH 7–9). LPM, lateral plate mesoderm; PM, posterior mesoderm; HP, heart primordium; HH, Hamburger and Hamilton.

RhoA expression confirmed by in situ hybridisation and immunocytochemistry.

The expression pattern of RhoA mRNA during early embryonic chick development was analysed by whole-mount in situ hybridisation. These analyses showed significant RhoA expression in the anterior lateral plate (HH stages 4–5) and heart primordium (HH stages 6–9; Fig. 4A-A,C,D,F), and no significant signal in these regions by using the DD54 antisense probe. Intense staining of RhoA mRNA was evident in stage 7 (one to two somites) embryos (Fig. 4A–C) in the region of the headfold and in the heart primordium, that is, in a region encompassing the border of the anterior intestinal portal and amniocardiac vesicle. Figure 4A–D depicts a magnification of a stage 8 (four to six somites) embryo in which strong staining for this message is evident in heart primordium region, that is, in a crescent encompassing vitelline vein, heart primordia, and the border of the anterior intestinal portal. As well, relatively intense staining was evident in the neural fold. A section of this embryo encompassing neural groove and heart primordial regions lateral to it (Fig. 4A–F) indicates that RhoA expression at this cellular level is widespread, being present in all three germ layers and with intense staining in the lateral plate mesoderm as it ingresses into the amniocardiac vesicle. In contrast, GAPDH expression appeared relatively ubiquitous in all embryos tested (data not shown). Later embryos (e.g., stage 14, 17 somites; results not shown) exhibited a more general staining pattern in keeping with the findings of Malosio et al. (1997), who described ubiquitous RhoA expression in 6-day chick embryos.

Figure 4. A: RhoA expression is up-regulated in early heart development. Whole-mount in situ hybridisation studies of chick embryos. Stage 4 (A), stage 7 (C), and stage 8 (D) embryos probed with DD54 antisense riboprobe. Stage 4–5 (B) and stage 8 (E) embryos probed with DD54 sense riboprobe. F: A section from embryo in D from the region depicted by the white line in D. Arrows in each case indicate the heart-forming region of one side. S, somite; P, proencephalon; HN, Hensen's node; PS, primitive streak; N, neural tube. B: RhoA protein expression in the early chick as detected by immunocytochemistry. Stage 5 (A) and stage 8 (C) embryo stained with RhoA antibody; stage 5 (B) and stage 8 (D) negative controls. The arrows indicate heart-forming regions: lateral plate in stage 5, heart primordium in stage 8.

A polyclonal RhoA antibody was used to detect RhoA protein abundance in early chick embryos. Immunocytochemistry was performed on whole-mount embryos at HH stages 5 and 8 (Fig. 4B). RhoA was detected in both the lateral plate (HH stage 5) and the heart primordium (HH stage 8). RhoA was also detected in stage 5 primitive streak and stage 8 neural fold. These results are consistent with the results of in situ analysis of RhoA mRNA expression.

RhoA plays an important role in embryogenesis: Treatment with RhoA siRNA induces cardia bifida and abnormal head development in cultured chick embryos.

To investigate the role of RhoA in early heart development, we injected chick embryos with double-stranded (ds) siRNA homologous to RhoA (siRNA-RhoA) to inhibit RhoA expression. The use of ds siRNA in this way blocks the activity of homologous cellular genes (Fire, 1999) and has been shown to interfere with the function of specific genes in both invertebrate and vertebrate species (Cogoni and Macino, 1997; Misquitta and Paterson, 1999; Wargelius et al., 1999; Clemens et al., 2000).

In vertebrates, myocardial precursor cells migrate from the primitive streak to the anterior lateral mesoderm to form paired cardiac fields. The bilateral heart primordia then migrate to the midline and fuse to form a single heart tube. When siRNA homologous to RhoA was injected into the anterior lateral plate (at the level of Hensen's node) of chick embryos (HH stage 4–6) cultured in vitro, the fusion of the heart primordia was inhibited (Fig. 5A). This result suggests that RhoA is involved in regulating the migration of cardiac precursors to the ventral midline. Figure 5C illustrates one of the most severe cases, in which no fusion of the heart tube occurred, head formation was abnormal, and somitogenesis was disrupted. The abnormal cardiac, head, and somite development caused by disrupted RhoA expression correlates with the expression pattern of RhoA (Fig. 4). Embryos with cardia bifida that were observed 48 hr after injection of siRNA-RhoA did not develop any further than stage 11. Nevertheless, affected embryos, even with cardia bifida, usually had beating hearts. The abnormal head development observed in siRNA-RhoA–treated embryos occurred either in conjunction with cardia bifida or in embryos with normal heart formation. The negative control used in these experiments was a scrambled siRNA template with the same nucleotide content as the siRNA-RhoA but designed not to target any known mRNA. Embryos injected with scrambled siRNA at the same concentration as siRNA-RhoA did not show abnormal development (Fig. 5B,D).

Figure 5. Disruption of RhoA expression by small interfering RNAs (siRNA) treatment induced cardia bifida and abnormal head development. Chick embryos treated with siRNA-RhoA (A,C) and control siRNA (B,D). The images are of embryos 24 hr (A,B) and 48 hr (C,D) after siRNA injection, respectively. The position of the heart is indicated by an arrow.

We used real-time quantitative PCR to measure the level of RhoA mRNA normalised to GAPDH in embryos treated with siRNA. The relative expression level of RhoA in siRNA-treated embryos compared with negative control embryos (embryos injected with scrambled siRNA) was reduced by 50% after 24 hr and 80% after 48 hr (Fig. 6). In stage 11 embryos (24 hr after injection), RhoA mRNA expression in siRNA-affected embryos was reduced by twofold and in stage 15 embryos (48 hr after injection) by fivefold compared with the negative control. As expected, the expression level of RhoA mRNA normalised to GAPDH in the negative control samples was higher in stage 15 embryos than in stage 11 embryos, consistent with the expression of RhoA mRNA increasing as the embryo developed.

Figure 6. Small interfering RNAs (siRNA) -RhoA treatment of early chick embryo disrupts RhoA expression. Real-time quantification polymerase chain reaction (PCR) analysis of RhoA in chick embryo mRNA after treatment with small interfering RNAs (siRNA) homologues to RhoA or randomised negative control. RNA was extracted from siRNA-treated embryos 24 hr and 48 hr after injection of Hamburger and Hamilton stage 4–6 embryos and subjected to reverse transcription and real-time PCR. The RhoA expression was normalised to GAPDH as described in the Experimental Procedures section and is shown for siRNA-RhoA–treated embryos compared with control embryos treated with scrambled siRNA template 24 and 48 hr after treatment. Values represent the ± standard error of mean of three independent experiments.

DISCUSSION

We have combined the merits of classic embryology and modern molecular biology to dissect embryonic events. This method used a modified differential display technique as well as other molecular innovations to detect and characterise potential novel gene transcripts associated with early cardiac tissue commitment. In the present study, we have identified RhoA as a gene whose expression is markedly up-regulated in the tissue primordia giving rise to chick heart development.

RhoA Is Highly Up-Regulated in Early Heart Development and Appears as Multiple Length Transcripts Varying in the 3′-UTR

Differential display, developmental dot blots, and in situ hybridisation experiments during early development clearly show that RhoA is highly up-regulated during cardiogenesis. Immunocytochemistry confirms this expression pattern at the protein level. RT-PCR amplification indicates that relative higher expression of the second shortest transcript may be associated with heart development. Four different transcripts of RhoA, varying in the length of the 3′-UTR were identified when screening the HP embryonic chick λ cDNA library with the DD54 probe. RT-PCR of chick embryonic RNA confirmed the expression of multiple RhoA transcripts in chick. Multiple length transcripts of RhoA have been reported in human breast cancer cell lines (Moscow et al., 1994) and sequence analyses of different Drosophila Rho1 cDNA RhoA homologues reveal at least six different classes of transcripts as a result of alternative splicing in both the 5′- and 3′-UTRs (Magie et al., 1999). As with chick and human RhoA, all the transcripts share an identical ORF. Alignment of chick and human RhoA cDNA (Fig. 2) shows the three polyadenylation signals identified for human RhoA are identical to three of the putative chick polyadenylation signals. Of interest, the regions 5′ of each of these polyadenylation signals are highly conserved (∼90%). The fourth polyadenylation signal identified in chick (RhoA₃) is not identified in human and, as discussed earlier and because it is not represented in human, may well not be a true transcript. The sequence similarity between human and chick in the 3′-UTR and the almost identical region before the polyadenylation signal suggests that the 3′-UTR with its multiple polyadenylation signals has functional significance. Alternative polyadenylation of the GAPDH gene (Mezquita et al., 1998) and Tlk (Touled-like kinase) gene (Shalom and Don, 1999) have been associated with testis differentiation in chick and mouse, respectively. Furthermore, Fraidenraich and coworkers have demonstrated that regulatory elements governing Fgf4 expression patterns in developing mouse embryos are located within the 3′-UTR of the gene (Fraidenraich et al., 1998). In embryonic development, many mRNAs encoding essential morphogenic proteins are selectively targeted to different subcellular domains and, therefore, selective protein location (St. Johnston, 1995). Ainger and coworkers have reported that a small element in the 3′-UTR of myelin basic protein mRNA, designated the RNA transport signal is necessary for cytoplasmic RNA transport in oligodendrocytes (Ainger et al., 1997). The proportionally greater expression of the second shortest RhoA (RhoA₂) transcript detected in developing heart tissue, compared with other embryonic tissues in the present study, suggests a possible developmental role for this 3′-UTR that is yet to be elucidated.

RhoA Is Important for Vertebrate Embryonic Organogenesis

Disruption of RhoA expression in vivo using small interfering RNA injected into anterior lateral plate mesoderm, confirmed by using real-time PCR, has indicated that RhoA expression is necessary for the development of normal heart and head morphology in the early chick embryo. The fusion of the heart primordia was inhibited in many cases. This result suggests that RhoA is involved in regulating the migration of cardiac precursors to the ventral midline.

RhoA belongs to the vertebrate Rho family of the Ras superfamily of small GTPases. RhoA, Rac, and Cdc42 are commonly known as key regulators of the actin cytoskeleton. Through their interaction with multiple target proteins, they also ensure coordinated control of other cellular activities such as adhesion and gene transcription (Hall, 1998). We have in this study shown that RhoA expression is important for early heart development and head formation. This finding is consistent with results from other studies of Rho family GTPases in Drosophila, Xenopus, and vertebrates, which suggest that these proteins are important for early embryogenesis (Magie et al., 1999; Wunnenberg-Stapleton et al., 1999; Wei et al., 2002). Magie and colleagues reported that loss of Rho1 function in Drosophila results in severe defects in morphogenetic processes, such as a serious defect in head involution and imperfect dorsal closure in embryos (Magie et al., 1999). In Xenopus, XRhoA has been suggested to be the first intracellular signalling molecule implicated in head formation (Wunnenberg-Stapleton et al., 1999). Wei and coworkers (2002) have suggested a critical role for RhoA family proteins in heart development through showing that inhibition of these proteins en masse, in studies using mice, resulted in disruption of cardiac morphogenesis, including cardiac looping and chamber maturation (Wei et al., 2002). The same group has also reported that Rho kinases, which are direct downstream effectors of RhoA, play an essential role in vertebrate embryonic organogenesis. Inhibition of Rho kinases in early chick embryos blocked migration and fusion of the bilateral heart primordia. Also, inhibition of RhoA kinases induced up-regulated expression of cardiac α-actin, SRF, and GATA4 mRNA levels in stage 8 embryos (Wei et al., 2001). It would be expected that down-regulation of RhoA would exhibit the same effects that result when Rho kinases are inhibited. Thus, these reports are consistent with our results and suggest that RhoA is a key regulator for vertebrate embryogenesis.

The mechanism by which RhoA regulates the diverse multiple cellular processes described above is not well understood. One possibility is that RhoA mediates distinct cellular functions through the control of transcriptional activation. This hypothesis is consistent with many studies in which RhoA has been implicated in the regulation of transcription factors (Hill et al., 1995; Chang et al., 1998; Wei et al., 1998; Kim and Cochran, 2000).

RhoA has been reported to regulate the transcription factors Ap-1, NFκB, GATA4, and SRF (Hill et al., 1995; Chang et al., 1998; Montaner et al., 1998; Wei et al., 1998; Charron et al., 2001). The most studied transcription target of RhoA is the c-fos promoter, which is activated by RhoA through the serum response element (SRE). Activated RhoA was shown to mediate a signal transduction pathway that led to activation of SRF (Hill et al., 1995). This finding was further supported by Wei and coworkers (1998), who showed that RhoA-dependent signalling by means of SRF is necessary in the activation of c-fos and skeletal α-actin promoters in muscle cells (Wei et al., 1998). In the same study, overexpression of the dominant negative mutant form of RhoA (N19-RhoA) inhibited morphologic differentiation in myoblasts. RhoA has also been shown to potentiate transcriptional activity of the cardiac-specific transcription factor GATA4, by stimulating the transcriptional activity of its activation domain (Charron et al., 2001). The identification of RhoA as a transcriptional regulator of GATA4 suggests a mechanism by which RhoA may be implicated in controlling heart gene expression.

These studies of RhoA together with our results strongly support the view that RhoA is involved in and important for early cardiogenesis. Chen and Schwartz (1996) demonstrated that cotransfection of the cardiac transcription factors Nkx2.5 and SRF in nonmyogenic fibroblasts resulted in activation of cardiac α-actin expression (Chen and Schwartz, 1996). RhoA has been shown to activate the skeletal α-actin promoter in myoblasts, through regulation by β1-integrin and PI 3 kinase (Wei et al., 2000). Hill and coworkers demonstrated that RhoA regulates transcriptional activity by SRF in 3T3 cells (Hill et al., 1995). In addition, it has been demonstrated that RhoA-dependent activation of SRF results in the expression of muscle-specific genes (Wei et al., 1998). As mentioned above, RhoA has also been found to potentiate transcriptional activity of the heart-specific transcription factor GATA4 (Charron et al., 2001). These findings indicate that there is a link between GATA4 and possibly other transcription factors, such as SRF and RhoA in heart, which results in transcription, expression, and activation of important genes associated with early heart development. Further studies will be required to investigate whether inhibition of RhoA will affect the level of transcription of genes shown to be important in early heart development.

In summary, a method of molecular dissection of the embryonic events has identified the marked up-regulation of a RhoA transcript during cardiogenesis. This finding has not been reported previously. These studies show that RhoA has at least three different length mRNA species that differ in the length of their 3′-UTRs. In situ hybridisation and immunocytochemical analysis confirmed marked up-regulation of RhoA in heart primordial regions (stages 6–8), and RT-PCR amplification indicates that relative higher expression of the second shortest transcript may be associated with heart development. As well, disruption of transcription in vivo by using small interfering RNA injected into lateral plate mesoderm indicates that RhoA expression is necessary for the development of normal heart and head formation in the early chick embryo.

EXPERIMENTAL PROCEDURES

Collection of Chick Embryos for Microdissection and Culturing

Fertile hen eggs were obtained from commercial sources and incubated at 38°C for 18–30 hr. Yolks were decanted into prewarmed 0.9% saline. Blastoderms were circumcised with scissors and gently floated free by pipetting and then spread dorsal side up on glass Petri dishes after which most of the saline was removed.

RNA Sampling of Embryos

Embryos were digested with 0.4% trypsin for 90 sec. Digestion was stopped by addition of 0.8% soybean trypsin inhibitor. The target regions of the embryo were then scraped with a fine glass needle so that the surface layer of cells (ectoderm) was drawn back. A micromanipulator equipped with a fine glass capillary attached to a suction tube was then brought into the target region, and mesoderm cells (<200) were gently drawn into the capillary.

To identify significantly up-regulated gene transcripts involved in the process of cardiogenesis, cells from a noncardiogenic region (HH stage 4–5 PM), the cardiogenic region (HH stage 4–5, anterior LPM) as defined by Rawles (1936), the HP of stage 6–9, and from in vitro induced posterior sections (PSi, see below) were sampled. Ten posterior sections were cocultured with anterior mesendoderm (both from stage 4). Six explants arbitrarily sampled at 5 and 24 hr in culture (thus, encompassing the process of induction) were pooled and provided the samples described as PSi.

To be able to establish a DD method suitable for the analysis of these sampled tissues, a modification of the method of Zimmerman and Schultz was introduced to yield sufficient starting material (Zimmermann and Schultz, 1994). A “capfinder” cDNA cloning kit (Clontech) was used to create full-length cDNA molecular libraries for the following tissue samples: PM, LPM, HP, and PSi. Efficacy of all these “cap” libraries was confirmed by the successful amplification of GAPDH cDNA. As well, cDNA for Nkx2.5 (an early marker of cardiogenesis) was successfully detected by PCR in the LPM and HP cap libraries. To allow the detection of rare transcripts, cap libraries for DD were also generated from subtracted RNA populations by using oligotex mRNA derived from PM and LPM (Kuribayashi-Ohta et al., 1993). This procedure resulted in creation of subtracted cap cDNA: LPM-PM and HP-LPM. The differential display products from these were compared with the display products obtained from the primary cap libraries described above.

Dissection and Culture of Embryo Explants

As a confirmation of correct tissue sampling, full cross-section explants of the cardiogenic regions (stage 4–5 and 6–9) and noncardiogenic regions (stage 4–5) were cultured on Wolff and Haffin's semisolid medium (Sarasa and Climent, 1991; seven parts v/v 1% bactoagar [Difco] in Gey's saline, three parts Tyrode's saline, and three parts horse serum). For coculture experiments, embryos were isolated and positioned ventral side up on a layer of 1% agar. Explants consisting of anterior mesendoderm (endoderm and mesoderm) from the central anterior region and posterior section (full cross-section at the posterior end of the primitive streak) were dissected away with fine glass needles and cultured side by side on Wolff and Haffen's medium. Control cultures containing either one or the other tissue were also cultivated.

All animal tissue samples were obtained in accordance with the current NH&MRC code of practice for the care and use of animals in research in Australia.

Differential Display

A total of 0.6 ng of cap cDNA was used as template in a total reaction mix of 20 μl containing 1 μM of 3′-primer (Table 1), 0.5 μM 5′-primer (Table 1), 2 μM dNTPs, 10 μCi [α³³P]ATP, 2.5 mM MgCl₂, 1× reaction buffer and 2.5 U of Taq polymerase. Samples were overlaid with mineral oil and subjected to 40 cycles of PCR according to the following protocol: 94°C for 30 sec, 42°C for 60 sec, 72°C for 30 sec, and a final 5 min extension at 72°C. PCR products were electrophoresed on a 6% denaturing polyacrylamide sequencing gel alongside a known sequencing reaction (to provide size markers), and run for approximately 3 hr at 50 W to display fragments ranging in size from 70 to 500 bp. The gel was then autoradiographed overnight, and the display of fragments thus obtained was examined for bands of interest.

Table 1. Primer Sequences Used in Polymerase Chain Reaction (PCR) Amplifications

PCR application	5′-primer	3′-primer
“cap” cDNA amplification	5′-TGCTGCGAGAAGACGACAGAAT-3′	5′-(T)30(A,G,C)(A,G,C,T)-3′
GAPDH amplification	5′-ACGCCATCACTATCTTCCAG-3′	5′-CAGCCTTCACTACCCTCTTG-3′
Nkx2.5 amplification	5′-CCTTCCCCGGCCCCTACTAC-3′	5′-CTGCTGCTTGAACCTTCTCT-3′
Differential Display	5′-CTTGATTGCC-3′	5′-(T)11(A,C,G)T-3′
	5′-CTGATCCATG-3′	5′-(T)11(A,C,G)G-3′
	5′-TACAACGAGG-3′	5′-(T)11GG-3′
	5′-TGGATTGGTC-3′	5′-(T)11CT-3′
	5′-CTTTCTACCC-3′
	5′-TTTTGGCTCC-3′
	5′-AGTGAGCTAG-3′
RhoA 3′UTR, Reverse transcription	5′-GGCCAGTGAATTGTAATACGACTCACTATAGGGAGGCGG-(Dt)24-3′
RhoA 3′UTR amplification	5′-CGGAATTCTCCGGGTGCCTTCTCTTATAAAG-3′	5′-GCCGGAATTCCTCACTATAGGGAGGCGG-3′
RhoA amplification, Real time PCR	5′-GTTGGCTTTGTGGGATAC-3′	5′-CAGAAATGCTTCACTTCCG-3′
GAPDH amplification, Real time PCR	5′-ACTTTGGCATTGTGGAGGGTC-3′	5′-TGGACGCTGGGATGATGTTC-3′

Amplification and Cloning of Bands of Interest

By using the autoradiograph as a template, the piece of gel containing the cDNA fragment of interest was cut out, purified, and reamplified. Reamplified products were ligated into a bidirectional TA cloning vector (Invitrogen) and used to transform bacteria. Recombinant plasmids were selected by restriction digest and purified by using a Plasmid Mini Kit (Qiagen), and inserts were sequenced using the ABI prism protocol and an ABI/Perkin-Elmer 373A DNA sequencer.

Sequence Analysis of Amplicons

Vector insert sequences, edited to remove identified polyA or polyT tails, were analysed for homology by using BLAST (Altschul et al., 1990) against the nonredundant DNA database (NCBI), were also translated into protein reading frames, and were analysed for homology against nonredundant protein databases.

Developmental Dot Blots

Plasmid DNA samples were denatured by heating at 95°C for 10 min followed by incubation at 4°C for 10 min. A total of 100 and 300 ng of DNA per dot well were transferred to GeneScreen Plus membranes (Du Pont) by using a Hybri-Dot manifold (BRL). Membranes and DNA were then crosslinked by exposure to ultraviolet light for 4 min. cDNA (PM, LPM, and HP) probes were labelled by random hexamer priming (Boehringer) by using [α³²P]ATP (Bresatec) to a specific activity of >10⁹ counts/min/μg of probe DNA. Probes were added to hybridization solution at a concentration of 0.5 and 10 ng/ml. Hybridisation was carried out in 5× SSPE (0.15 M sodium chloride, 0.01 M sodium dihydrogen orthophosphate, 1 mM EDTA, pH 7.7), 5 × Denhardt's solution, 0.5% sodium dodecyl sulphate (SDS), 100 μg/ml fragmented denatured herring sperm DNA at 65°C for 18 hr. Membranes were washed twice in 3× SSPE, 0.1% SDS for 10 min at room temperature, and twice in 1 × SSPE, 0.1% SDS for 30 min at 65°C. Washed membranes were wrapped in plastic film (Gladwrap) and exposed to Fuji X-ray film in an X-ray cassette equipped with an intensifying screen (Kodak) at −70°C overnight. Resultant dots were quantitated by using ultraviolet photometry (UVP) analysis software and normalised according to radiolabel activity. In these analyses, dot signal data for empty plasmids (without inserts) provided the background level. A plasmid containing smooth muscle full-length α-actin cDNA, pα (kindly provided by Dr. C.C. Kumas), which hybridises to α-, β-, and γ-actins (Murrell et al., 1994), was used as a positive control to test the integrity of the labelled cDNA probes.

Construction and Screening of cDNA Libraries

Cap cDNA from lateral plate mesoderm (stage 4) and from heart primordium (stages 6–9) was cloned into λZap express (Stratagene) by using the methods described in the Clontech Capfinder handbook. Plaque lifts were performed according to Stratagene's λZap handbook and hybridisation of probe DNA to phage DNA immobilised on Genescreen plus membranes was undertaken according to the methods described in the Genescreen handbook.

Fixing and Storage of Whole-Mount Chick Embryos for In Situ Hybridisation and Immunocytochemistry Analysis

Whole stage 4–9 chick embryos were fixed in 4% paraformaldehyde in phosphate buffered saline (PBS), pH 7.4, for 20 min at room temperature. After this step, they were washed twice for 5 min in PBS. Embryos were then stored at −20°C in 70% ethanol until use.

Whole-Mount In Situ Hybridisation of Early Chick Embryos

Fixed embryos were incubated in 100% ethanol for 30 min at room temperature. After this, they were washed twice for 5 min in PBS, after which they were incubated in 0.3% Triton X-100 in PBS for 15 min and then washed again twice in PBS. Permeabilisation was achieved by digestion for 10 min by 3 or 10 μg of proteinase K per ml in Tris/EDTA (10 mM Tris-CL, pH 8.0 and 1 mM EDTA). Post-fixation was for 5 min in 4% paraformaldehyde followed by two washes for 5 min in PBS. Embryos were prehybridised in 4 × SSC (0.03 M sodium chloride, 3 mM tri-sodium citrate), 50% deionised formamide at 37°C for 10 min. Hybridisation took place overnight at 42°C under the following conditions: 40% deionised formamide, 10% dextran sulphate, 1 × Denhardt's solution, 4 × SSC, 10 mM dithiothreitol, 1 mg/ml fragmented denatured herring sperm DNA, 1 mg/ml transfer RNA, 200 ng/ml of probe. Washes were as follows: 2 × SSC, 55°C twice for 15 min; 1 × SSC, 55°C twice for 15 min; and 0.1 × SSC, 37°C for 60 min. Detection was performed by using the Dig Detection Kit (Boehringer) with colour reaction taking place for up to 24 hr. Reactions were stopped and embryos were stored in H₂O until mounting in glycerol.

Whole-Mount Immunocytochemistry of Early Chick Embryos

RhoA was detected in whole-mount chick embryos by a polyclonal rabbit RhoA antibody (Santa Cruz). Controls included embryos incubated with no primary antibody, preimmune serum of the species used to raise the primary antibody, or no secondary antibody. Permeabilisation was achieved by digestion for 10 min by using 10 μg of proteinase K per ml in PBS. Post-fixation was for 20 min in 4% paraformaldehyde at 4°C followed by two washes for 5 min in PBS. Endogenous peroxidase was quenched by incubation in 3% H₂O₂ and 0.3% Triton X-100 in PBS for 5 min and the embryos were then washed again twice in PBS. Nonspecific binding sites were blocked with 10% goat serum in PBS for 1 hr. The embryos were then incubated overnight with RhoA antibody (Santa Cruz; 1/200 in PBS) in 1% goat serum. This procedure was followed by three washes of 10 min in PBS with 1% goat serum and 0.1% Tween 20. The embryos were then incubated at room temperature for 30 min in secondary goat anti-rabbit antibody horseradish peroxidase conjugate (1/200 in PBS) in 1% goat serum. After three 5-min washes in PBS, peroxidase activity was detected by incubation in 0.05% filtered diaminobenzidine (DAB), 0.004% H₂O₂ in PBS for 5 min. This was followed by three 5-min washes in PBS before mounting in glycerol.

RT-PCR Analysis of 3′-UTR Expression

Tissue from cardiogenic regions (LPM stage 4–5 and HP stage 6–9) and noncardiogenic regions (PM stage 4–5 and stage 6–9) were microdissected, and RNA from each tissue was extracted by using a method based on that reported by Chomczynski and Sacchi (1987). Genomic DNA was digested with 2 U of DNase per 10 μg of RNA at 37°C for 30–60 min.

cDNA templates were made by reverse transcription using a T7(dT₂₄) primer (Table 1) and 200 U of Superscript II (Invitrogen). A primer specific to the 3′-ORF of chick RhoA and a primer annealing to template reversed transcribed using the T7(dT₂₄) primer were used for PCR amplification (see Table 1). The cDNA synthesis was carried out in a 50-μl reaction volume consisting of 1 μl cDNA, 1× PCR buffer (Promega), 1.5 mM MgCl₂, 0.2 mM dNTPs, 0.2 μM of 5′-primer, 0.4 μM of 3′-primer, and 2.5 U of Taq polymerase. The Taq polymerase was added after the first denaturing step. Amplification was carried out in a thermal cycler (MJ Research) for 45 cycles, each cycle consisting of DNA denaturation for 30 sec at 94°C, annealing at 58°C for 60 sec and finally elongation for 60 sec at 72°C. Before the first cycle, the sample was heated and denatured for 3 min at 94°C. After the last cycle, the sample was kept at 72°C for a 6-min extension step. The PCR products were analysed by electrophoresis on a 1% agarose gel in TBE buffer, TA-cloned, and sequenced as described earlier.

Inhibition of RhoA by Using Small Interfering RNA and In Vitro Culture of Whole Embryos

Embryos between stages 4 and 5 were placed inside Whatman paper rings and laid ventral side up on albumin-agar plates as described by (Chapman et al., 2001). Small interfering RNAs (siRNA) homologous to the 3′-ORF of RhoA (siRNA-RhoA) and a scrambled negative control siRNA were synthesised by using the Silencer siRNA Construction Kit (Ambion). The sequences of the siRNA template were as follows: siRNA-RhoA, 5′-AAGAAAAAGTCCGGGTGCCTTCCTGTCTC-3′; Scrambled siRNA, 5′-AACTGCATTGAAAGGCAGTCGCCTGTCTC-3′ siRNAs targeted to RhoA and the negative control were diluted in Dulbecco's modified Eagle medium (Gibco) and injected at a concentration of 3,000 pmol in a 1-μl volume into the heart-forming region (LPM) of the chick embryo by using a micromanipulator (Narishige) and a glass needle. General toxicity effects were not apparent with siRNA quantities of 200–3,000 pmol. Embryos were incubated for 24 or 48 hr for phenotypic analysis and tissue sampling for real-time PCR.

Real-Time PCR Quantification of RhoA mRNA

Quantitative levels of RhoA mRNA from siRNA-treated chick embryos were assessed by real-time PCR (iCycler, Bio-Rad). RNA was extracted from siRNA-treated in vitro cultured chick embryos by using a modified method based on that by Chomczynski and Sacchi (1987) after DNAse treatment as described earlier. cDNA templates were prepared by reverse transcription using a anchored dT₃₀MN primer (Table 1) and 200 U of Superscript II (Invitrogen). Both primer pairs used for real-time PCR (see Table 1) were designed to span an intron to ensure that the products generated were from cDNA; the specificity of the primers was checked by BLASTn search. cDNA synthesis was carried out as described by Maxwell et al. (2002). Amplification of a single product of the correct size was also confirmed by agarose gel electrophoresis and verified by sequence analysis. Duplicate reactions for all samples were analysed and the RhoA mRNA level was quantitated by real-time PCR for three embryos per group.

Data were analysed by using iCycler Software (Bio-Rad), initially converting into threshold values (Ct), which refers to the cycle number during exponential amplification at which the PCR product crosses a set threshold. Standard curves for RhoA and for the housekeeping gene GAPDH were generated by using samples containing known amounts of linearised plasmid containing the ORF of the respective genes. The Ct values were then plotted against the log of the template amounts. The subsequent RhoA or GAPDH cDNA Ct values were within the range of Ct values used when generating the standard curves. Relative quantitation was obtained by plotting the average RhoA Ct value of the sample against the cycle thresholds of the respective standard curve to convert the Ct values into actual copy numbers of input cDNA. To adjust for variations in the amount of input cDNA, the average copy number for RhoA was normalised against the average copy number for GAPDH by dividing the target amount by the endogenous reference amount.

Acknowledgements

We thank Professor Roger Willis for his longstanding encouragement and support. Thanks also to Alyssa Scurrah for her assistance with aspects of the study.