• chick embryo;
  • heart;
  • subtracted library;
  • microarray


  1. Top of page
  2. Abstract
  7. Acknowledgements

A subtracted library was constructed of genes expressed specifically in the chick precardiac mesoendoderm. The subtracted library was obtained by hybridization of nucleic acids derived from a starting tester library of stage 4–7 chick precardiac mesoendoderm and a starting driver library of stage 2 area pellucida. Approximately 11,000 clones from the resulting subtracted library were printed onto a microarray. Screening of the microarray with probes derived from cardiac and noncardiac tissues, followed by in situ hybridization during chick embryo development, has identified multiple cardiac-specific genes, including several that have not been characterized previously. The microarray will be useful for future attempts to identify additional novel cardiac-specific genes, as well as to characterize patterns of gene expression during heart differentiation. Developmental Dynamics 230:290–298, 2004. © 2004 Wiley-Liss, Inc.


  1. Top of page
  2. Abstract
  7. Acknowledgements

One of the central goals of cardiac developmental biology is to identify and characterize genes that regulate cardiac development. Several transcription factors have been identified that are expressed during early cardiogenesis, including members of the Gata (Arceci et al., 1993; Laverriere et al., 1994; Jiang and Evans, 1996), Nkx (Lints et al., 1993; Tonissen et al., 1994; Schultheiss et al., 1995; Lee et al., 1996), Tbx (Chapman et al., 1996), and Hand (Srivastava et al., 1997) families, among others. Mutations in many of these genes have been reported to interfere with normal cardiac development (Lints et al., 1993; Basson et al., 1997; Kuo, 1997; Molkentin et al., 1997; Srivastava et al., 1997; Reiter et al., 1999; Bruneau et al., 2001). Several classes of signaling molecules have also been found to play important roles in the regulation of cardiac induction, including members of the Bmp (Zhang and Bradley, 1996; Schultheiss et al., 1997; Andree et al., 1998), Fgf (Reifers et al., 2000; Alsan and Schultheiss, 2002), Wnt (Pandur et al., 2002), and Wnt-inhibitor (Marvin et al., 2001; Schneider and Mercola, 2001; Tzahor and Lassar, 2001) families.

Most of these cardiac regulatory molecules have been identified through candidate gene approaches, which raises the question of whether additional regulatory molecules could be found using less targeted methods. Genetic screens, conducted mostly in zebrafish, have indeed identified novel cardiac regulatory molecules, such as UDP-glucose dehydrogenase (Walsh and Stainier, 2001) and a sphingosine-1-phosphate receptor (Kupperman et al., 2000), but such genetic screens are laborious to perform and will not detect genes that are required for embryonic events before heart formation, or that are important but not required for cardiac development. Recently, informatics-based approaches have been used to identify genes, such as myocardin (Wang et al., 2001) and hop (Shin et al., 2002), that were expressed specifically in libraries generated from cardiac tissue. However, the informatics approach is dependent on the existence of the appropriate cDNA libraries, and will not identify genes that are not unique to the heart but that are important, nevertheless, for heart development.

As a complement to these approaches, we have undertaken a screen to identify genes that are expressed specifically during early cardiac development, using the avian embryo as an experimental system. In the chicken embryo, cardiac progenitors gastrulate at Hamburger and Hamilton (HH) stage 3 and migrate into the anterior lateral plate mesoderm between HH stages 4–7 (Rosenquist and DeHaan, 1966; Garcia-Martinez and Schoenwolf, 1993). Between HH stages 8 and 10, the bilateral cardiac progenitor fields begin to express markers of terminal differentiation, such as myofibrillar genes (Han et al., 1992), and to converge in the anterior ventral midline to form the primitive heart tube, which begins beating at HH stage 10–11. We were particularly interested in identifying genes expressed at stages 4–7, because during this time, the cardiac progenitors become patterned to undergo cardiac differentiation and begin to express known cardiac regulatory genes (Schultheiss et al., 1995, 1997). Thus, genes expressed specifically at this time are good candidates to play roles in regulating cardiac development.

The avian embryo possesses several advantages for this type of study, including the possibility of isolating relatively large amounts of cardiac tissue throughout heart development, and the wealth of knowledge of avian heart development accumulated over the course of many years of experimental embryology. The main limitation of the avian embryo as a model system has been the relative paucity of genetic tools, when compared with other vertebrate animals such as the mouse or the zebrafish. The situation is beginning to change, with the growing availability of chicken expressed sequence tags (ESTs; Boardman et al., 2002) and with the ongoing chicken genome sequencing project, which is due to be completed shortly (Burt and Pourquie, 2003).

We have constructed a subtracted cDNA library of genes expressed specifically in the chicken precardiac mesoendoderm and have constructed a microarray of approximately 11,000 clones from the subtracted library. The current report describes the development of these resources and presents an initial analysis of genes identified by screening of the microarray.


  1. Top of page
  2. Abstract
  7. Acknowledgements

Creation of a Subtracted Library Enriched for Genes Expressed in the Heart Primordia

The majority of the mRNA molecules expressed in a given cell derive from a relatively small number of highly expressed genes, many of which are not expressed in a tissue-specific manner (Davidson, 1986). Therefore, an unmodified library constructed from cDNA from any tissue would primarily contain many copies of relatively few, commonly expressed genes. To obtain a collection of cDNAs enriched for genes that are expressed specifically in the heart primordia, we used subtractive hybridization to reduce the abundance of non–tissue-specific genes. In subtractive hybridization, nucleic acids from two populations, the tester (the tissue of interest) and the driver (the reference tissue), are hybridized together to remove common molecules from the tester population. HH stage 4 is the earliest stage in the chick embryo in which the precardiac mesoderm will autonomously differentiate into beating heart tissue if placed in culture (Biehl et al., 1985; Gonzalez-Sanchez and Bader, 1990). By stage 8, terminal differentiation genes, such as those coding for sarcomeric proteins, begin to be expressed at high levels (Han et al., 1992). As we are interested in genes involved in the regulation of early heart formation, the tester was derived from stage 4–7 precardiac mesoderm (Fig. 1). Because it is technically difficult to cleanly separate endoderm from mesoderm at these stages, endoderm from the precardiac region was also included in the tester. This precardiac endoderm is a known source of heart- inducing factors (Schultheiss et al., 1995); thus, the resulting library could also be used to analyze gene expression in this cardiac-inducing endoderm. The driver was derived from stage 2 whole embryos (Fig. 1). The choice of driver is an important issue for a successful subtractive hybridization. To minimize the risk of losing unknown cardiac genes that might also be expressed in other embryonic mesodermal tissues, pregastrula embryos, which lack any mesoderm cells (Schoenwolf et al., 1992), were chosen as the driver. This tissue has the advantage of not containing any mesodermal cells, although subtraction between mesodermal tester and nonmesodermal driver would also result in the isolation of mesoderm-specific genes that are not necessarily cardiac related. Cardiac and noncardiac clones in the subtracted library can be distinguished by differential screening of the microarrayed subtracted library with cardiac and noncardiac probes (see below). The use of a pregastrula driver also presents the opportunity to use this resource to identify genes involved in mesoderm formation in general (see Discussion section).

thumbnail image

Figure 1. Subtraction scheme. The tester was taken from stage 6 anterior–lateral mesoendoderm (red box), and the driver was taken from stage 2 area pellucida (blue circle). Directional lambda phage libraries were prepared from tester and driver mRNA. By using appropriate helper phage, single-stranded tester phagemid DNA (ssDNA; Ss Phagemid) and double-stranded driver DNA plasmid (dsDNA) were produced, and biotin-labeled RNA was produced from the driver plasmid. Tester single-stranded DNA (ssDNA) was hybridized with driver biotinylated RNA, and hybridized nucleic acids were removed with AvidinD agarose. The remaining tester ssDNA was converted to dsDNA plasmids to generate a subtracted tester library. See Experimental Procedures section for details.

Download figure to PowerPoint

The subtractive hybridization scheme is outlined in Figure 1 (see Experimental Procedures section for a detailed discussion). Briefly, directional lambda phage libraries were constructed from tester and driver mRNA. The tester library was converted to single-stranded DNA (ssDNA) through the use of helper phage. The driver library was converted to double-stranded plasmids, which were used as templates for RNA transcription in an in vitro transcription reaction. Biotinylated UTP was incorporated into the synthesized RNA. The biotinylated driver RNA was hybridized with the tester ssDNA, and streptavidin was used to remove tester:driver hybrids as well as unhybridized driver. The remaining unhybridized tester DNA was converted to double-stranded plasmid (dsDNA), resulting in a subtracted tester cDNA plasmid library.

To determine the efficiency of the subtraction, the relative abundance of several genes was compared in the subtracted tester and the starting tester cDNA using polymerase chain reaction (PCR). After subtraction, the ubiquitously expressed genes glyceraldehyde-3-phosphate dehydrogenase and EF-1α were reduced by 13–15 PCR cycles, while the abundance of known cardiac genes, including Nkx-2.5, Tbx-5, and Mef-2c, were reduced by only 2 cycles (data not shown). Therefore, the relative abundance of the cardiac genes had increased by approximately 11–13 cycles, or over 1,000-fold. The twofold loss of cardiac-specific genes is most likely due to nonspecific general loss of DNA during the subtraction procedure.

Production of a Microarray of the Subtracted Heart Library

Microarray technology provides several advantages for analyzing large populations of genes (Schena et al., 1995; Lockhart et al., 1996). It allows one to examine the expression patterns of many genes simultaneously. In addition, one can compare the gene expression profiles of multiple different tissues or of a single tissue under different experimental conditions. The original tester lambda library contained approximately 1 × 106 clones before subtraction. A 1,000-fold reduction in the abundance of the most commonly expressed genes (see above) suggested that the subtracted library contained on the order of one to several thousand independent clones. A cDNA microarray of 10,000 clones would be expected to contain at least one copy of each clone in the subtracted library. To construct such an array, PCR was used to amplify the cDNA inserts from 11,000 individually picked clones and the amplified DNA was spotted onto glass slides, as described in Experimental Procedures. The microarray will be made available to qualified researchers upon request.

Screening the Microarray

Traditionally, screening of a cDNA microarray is a differential screening, in which two fluorescently labeled probes (typically labeled with Cy3 and Cy5) are generated from two cDNA populations (Schena et al., 1995; Lockhart et al., 1996). The ratio of signals at each spot on the microarray is taken as a measure of the relative expression level of that gene in the two populations. For initial analysis of the microarray, we conducted two screens, based on known features of cardiac development and cardiac gene expression. One screen (the lateral–medial or LM screen) used probes generated from stage 6 precardiac anterior–lateral (AL) mesoendoderm and stage 6 anterior–medial (AM) mesoendoderm (Fig. 2). The latter tissue, which lies adjacent to the precardiac mesoderm, was chosen because our previous studies have shown that it is competent to give rise to cardiac tissue (when exposed to Bmp-2/4 signaling), although it does not do so under normal embryonic conditions (Schultheiss and Lassar, 1997; Andree et al., 1998). Therefore, any differences in gene expression between the two tissues would be reasonable candidates for being involved in cardiac differentiation. The second screen (the anterior–posterior or AP screen) used probes derived from stage 6 precardiac AL mesoendoderm and stage 6 posterior primitive streak (PPS; Fig. 2). The PPS was used because our previous studies have shown that it can be converted to cardiac tissue by a Bmp-independent mechanism (Schultheiss et al., 1997; Marvin et al., 2001). Thus, comparing gene expression in AL vs. PPS may identify a different class of cardiac genes than those identified in the LM screen.

thumbnail image

Figure 2. Origin of tissues used to generate microarray probes. For the lateral–medial screen, tissues were harvested from stage 6 anterior–lateral (AL) and anterior–medial (AM) mesoendoderm. For the anterior–posterior screen, tissues were harvested from stage 6 AL and posterior primitive streak (PPS). hn, Hensen's node.

Download figure to PowerPoint

To generate sufficient cDNA to screen the microarray, we used the method of mRNA amplification by in vitro transcription of cDNA (Baugh et al., 2001), which has been found to produce linear amplification of small amounts of starting material (see Experimental Procedures section). In our experience, two to four pieces of stage 6 AL or AM embryonic tissue, containing several thousand cells, could produce 5–10 μg of antisense RNA (aRNA) after one round of amplification, which is sufficient to generate several microarray probes.

The results of the initial LM and AP hybridization are given in Table 1. More than 90% of the spots in the array gave roughly equivalent signals (less than a twofold difference in signal) when screened with the two probes in each hybridization. In the LM screen, 93 spots (1.0%) had L:M ratios of greater than 3, and 78 (0.8%) had ratios between 2 and 3, whereas in the AP screen, 234 spots (2.3%) had ratios greater than 3 and 455 (4.5%) had ratios between 2 and 3. For initial analysis, all LM clones with ratios of greater than 3 were sequenced, as well as approximately 50 nonoverlapping AP clones. Each selected clone had a ratio of >3 on at least three hybridizations, including at least one in which the fluorescent tags were reversed. Meaningful sequence was obtained for 123 clones.

Table 1. Statistics From the LM and AP Screensa
  • a

    Results shown are from a single, representative example of each hybridization. The table shows the number and percentage of microarray spots in which the ratios of LM (or AP) signals are within the indicated ranges. Spots of poor quality were flagged by the Genepix Analysis software and eliminated from the analysis. LM, lateral–medial; AP, anterior–posterior.

Total9769 10123 

Table 2 summarizes the classes of genes identified in the initial LM and AP screens. Genes known to be expressed in the avian precardiac mesoendoderm were identified, including Tbx-20 (Iio et al., 2001), Gata-5 (Laverriere et al., 1994), and Bmp-2 (Schultheiss et al., 1997; Fig. 3G,J and data not shown). Several genes that had been reported previously to be expressed in the precardiac mesoderm of other vertebrates but not in chick were also identified, including the transcription factor Id2 (Jen et al., 1996), SH3BGR (Fig. 3F), which codes for an SH3-domain binding glutamic acid-rich protein of unknown function (Scartezzini et al., 1997; Egeo et al., 2000), and qik (Fig. 3I), a putative serine–threonine protein kinase that is the chick homologue of the murine msk gene (Ruiz et al., 1994), also with unknown function. As expected, Bmp-2, which is expressed in both the AL endoderm and PPS (Schultheiss and Lassar, 1997), was identified on only the LM screen. Although the screen was designed to exclude myofibrillar genes, which are largely expressed after stage 7, the two myofibrillar genes that were detected, cardiac troponin T and cardiac α-actin, have been reported to be expressed earlier in development (Hayward and Schwartz, 1986; Antin et al., 2002). It was of interest to note that clones that were represented by several spots on the array gave a very similar ratio at each spot, confirming the reproducibility of the approach.

Table 2. Clone Identified in the Initial Microarray Screensa
CategoryGeneNumber of occurrencesChick gene?Chick GenBank IDClosest geneSpecies
  • a

    All clones from the LM and approximately 50 nonoverlapping clones from the AP screen which had LM or AP ratios of >3 upon at least two hybridizations and also upon reversal of the fluorescent tags were characterized. If a clone matches a characterized chicken gene in GenBank, the GenBank accession number is listed. If there is no match with a previously characterized chick gene, the the GenBank accession number of the chicken EST is given, along with the GenBank Accession number and species of the closest characterized homologue. The category “unknown function” contains clones which matched a previously identified gene of unknown function, while the category “no match in database” contains those clones that did not match any published chick or nonchick sequence (including ESTs). LM, lateral–medial; AP, anterior–posterior; EST, expressed sequence tag.

CytoskeletalAlpha actin1YesM13756  
 Cardiac Troponin T1YesM10013  
 Fibrillin 21NoBU405593NM_001999Human
 Keratin 182NoBU454539X12881Human
Extracellular Matrix/Adhesion/ Cell SurfaceCadherin 111YesAF055342  
Metabolic1,2 Alpha mannosidase5NoBU141691AAF97058Human
 Apolipoprotein A110YesM18746  
 Biliverdin IX alpha reductase2NoCD760994X93086Human
 Gal beta 1,3 GalNAc alpha2,3 sialyltransferase4YesQ11200  
 GalNAC-4 Sulfotransferase6NoBU212596AF332472Human
 Transthyretin precursor1YesX60471  
 Ubiquitin-conjugating enzyme E2D 31NoAJ451547BC023266Mouse
 Bone Morphogenetic Protein 21YesAY237249  
 Retinoic Acid Receptor beta4YesS63196  
 Smad 62YesAF165889  
Transcription factorsGata-54YesU11888  
Other regulatoryBridging integrator 1 (BIN1)1NoAJ450340NM_139345Human
 MN1 tumor supressor1NoBU363646X82209Human
 Qin-induced kinase1YesAF219232  
 Rab6 GTPase activating protein3NoBU325580BC031714Human
No match 22    
thumbnail image

Figure 3. Selected genes identified in the microarray screens. Whole-mount in situ hybridization was performed on stage 6–7 embryos. A–L: 1,2-α-Mannosidase (A), Meis1 (B), Thrombomucin (C), Gal-beta-1,3-GalNAc-alpha-2,3-sialyltransferase (D), Msx-1 (E), SH3BGR (F), Tbx-20 (G), MN1 (H), qik (I), Gata-5 (J), Otokeratin (K), Bambi (L). All embryos are oriented with the anterior end at the top. hfr, heart forming region.

Download figure to PowerPoint

In addition to previously characterized cardiac genes, the screens also identified other genes that had not been reported previously to be expressed in the heart, or whose embryonic expression pattern had not been described previously. In situ hybridization was performed to examine the expression pattern of these previously uncharacterized genes during avian embryogenesis, a sample of which is shown in Figure 3. Several of these genes had expression patterns that were largely confined to the heart-forming region, including thrombomucin (McNagny et al., 1997; Fig. 3C), and 1,2-α-mannoisidase (Hamagashira et al., 1996) (Fig. 3A), while others were expressed in a more complex pattern, which included the precardiac mesoderm but also extended into other tissues, including otokeratin (Heller et al., 1998; Fig. 3K), Gal-beta-1,3- GalNAc-alpha-2,3-sialyltransferase (Fig. 3D), Meis1 (Moskow et al., 1995; Fig. 3B), Msx1 (Hill et al., 1989; Fig. 3E), Bambi (Onichtchouk et al., 1999; Fig. 3L), and the oncogene MN1 (Lekanne Deprez et al., 1995; Fig. 3H). The expression patterns of 1,2-α-Mannosidase (Fig. 3A) and Gal- beta-1,3-GalNAc-alpha-2,3-sialyltransferase (Fig. 3D) demonstrate that genes that are involved in basic metabolic pathways of the cell can be expressed in highly tissue-specific patterns and, therefore, may be involved in tissue-specific processes.

The screen also detected 22 highly differentially expressed clones that did not match any sequences in the published databases. These likely represent 3′-untranslated sequences, and their identification awaits isolation of longer clones (see Discussion section).


  1. Top of page
  2. Abstract
  7. Acknowledgements

We have constructed a subtracted library of cDNAs that is enriched for genes expressed in the chicken precardiac mesoendoderm. The subtraction procedure reduced the relative abundance of commonly expressed genes by at least 1,000-fold, which permitted the entire set of genes contained in the subtracted library to be spotted onto a single microarray of approximately 11,000 clones. By screening the microarray with probes derived from cardiac and noncardiac regions of the embryo, we have identified several previously uncharacterized genes that are expressed during early heart formation. These genes fall into several functional categories, including putative regulatory molecules, cytoskeletal elements, metabolic enzymes, membrane proteins, and components of the extracellular matrix. These genes are the focus of ongoing studies aimed at determining their function during heart differentiation and development.

It was possible to collect sufficient tester and driver tissue polyA-RNA to make the initial lambda libraries without the need for PCR amplification and the potential bias that such amplification can produce. Approximately 250 tester embryos and 40 driver embryos were sufficient to build the initial lambda phage libraries. It was possible to collect this amount of tissue in less than 2 weeks. This finding points out one of the advantages of the avian embryo for these types of studies, in that it is possible to collect tissue from virtually any region of the embryo in sufficient quantities to generate these types of tissue-specific libraries.

The microarray resource should be useful for many types of studies of gene expression during avian heart formation. For example, previous studies from our laboratory have established that AM mesoderm will initiate expression of cardiac genes upon exposure to Bmp-2 signaling (Schultheiss et al., 1997). In work in progress, we are constructing probes from AM tissue with or without Bmp-2 treatment to identify genes that are regulated by Bmp signaling during heart formation. In addition to precardiac mesoderm, the subtracted library also contains genes expressed in the endoderm of the precardiac region. Therefore, the microarray should be useful for characterizing this endoderm, which possesses cardiac-inducing properties (Nascone and Mercola, 1995; Schultheiss et al., 1995) and which contains cells that will give rise to liver tissue (Le Douarin, 1975). Because the driver used to construct the subtracted library was derived from stage 2 embryos, which possess no mesoderm or definitive endoderm, the microarray should also be useful to identify genes that are expressed specifically in the mesoderm and definitive endoderm, when screened with appropriate probes.

One of the main parameters that governs the usefulness of a microarray is the issue of sensitivity. If a given cDNA is highly differentially expressed between two tissues but is expressed at very low levels, then a probe may not contain enough molecules of the cDNA of interest to give a detectable signal upon hybridization. For example, although Nkx-2.5 was present in the microarray (and detectable with an Nkx-2.5 probe), it was not detected by either the LM or AP probes. We have found that many clones that produce a barely detectable yet differentially expressed signal upon array hybridization (e.g., Tbx-20) are expressed in highly tissue-specific patterns as revealed by in situ hybridization. This finding suggests that there are likely to be additional differentially expressed genes that are expressed at levels below the detectability of our screening methods. Increasing the total amount of probe should increase the sensitivity of the hybridization and, therefore, the detectability of rare transcripts to some degree. A further increase in sensitivity could be obtained by screening the microarray with subtracted probes generated using the same techniques that were used to produce the subtracted library.

Eighteen percent of the differentially expressed clones identified in the LM and AP screens did not match any sequences in the public databases (Table 2). This finding occurred most likely because these clones code for the 3′ untranslated regions of cDNAs that have not previously been sequenced. We have confirmed this by using several such clones as probes to screen a chick cDNA library. In each case, the clone in the microarray was found to be part of the 3′ untranslated region of a gene for which there were homologues in the public databases. As chicken genome sequence becomes increasingly available (Burt and Pourquie, 2003), it will become much easier to identify the genes from which these 3′ untranslated sequences are derived.


  1. Top of page
  2. Abstract
  7. Acknowledgements

Library Construction and Subtraction Hybridization

Dissected tissue, approximately 500 pieces of HH stage 4–7 AL mesoendoderm (tester; from 250 embryos) and 40 pieces of whole embryo HH stage 2 area pellucida (driver), was collected in Trizol (Sigma) and processed for total RNA purification. Two rounds of mRNA selection were performed by using Oligotex mRNA kit (Qiagen) before cDNA synthesis and construction of directional tester and driver λ-ZapII libraries (Stratagene). The phage libraries were titered and amplified according to standard protocols. The ssDNA was prepared from the tester library by mass conversion using the VCM13 helper phage (Schweinfest et al., 1995). From the driver library, dsDNA plasmids were prepared by using the ExAssist helper phage and the plate lysate method (Sambrook et al., 1989), followed by linearization with XhoI. RNA from this library was produced by in vitro transcription with T3 RNA polymerase (Promega) in the presence of UTP-biotin (Roche; ratio UTP-biotin/UTP : 1/3).

The method for subtraction was derived and modified from (Schweinfest et al., 1990; Klar et al., 1992; Lemaire et al., 1993). Hybridization between ssDNA from the tester library and biotinylated RNA from driver library was performed as follows: In a siliconized PCR tube (0.5 ml), 0.5 μg of ssDNA and 20 μg of RNA were added followed by ethanol precipitation. The nucleic acid pellet was dissolved in 10 μl of RNAase-free hybridization buffer (0.5 M NaCl, 50 mM HEPES pH 7.6, 0.2% sodium dodecyl sulfate [SDS], 2 mM ethylenediaminetetraacetic acid [EDTA], 50% formamide). A total of 5 μg of blocking oligo (GATCCACTAGTTCTAGAGCGGCCGCCACC- GCGGTGGAGCT) was included in the hybridization mixture to block the region of the Bluescript multiple cloning site that was present in both tester and driver (which, thus, could produce nonspecific hybridization). By using a PCR machine with a heated lid, the reaction mix was first heated to 70°C for 5 min, followed by incubation at 45°C for 29 hr. To stop the reaction, 400 μl of dilution buffer (0.5 M NaCl, 50 mM HEPES, 2 mM EDTA) was added. Biotinylated nucleotides (driver:tester hybrids and unhybridized tester) were removed with two extractions using AvidinD Agarose resin (Vector Labs). The remaining ssDNA from the tester library was ethanol precipitated, dissolved in 20 μl of 5 mM Tris pH 7.5, 0.1 mM EDTA, and converted into a plasmid library by using the Klenow fragment of DNA Polymerase and T3 primer. The plasmid DNA library was transformed into Escherichia coli and amplified by using standard methods (Stratagene).

Microarray Construction and Array Printing

Approximately 11,000 colonies were manually picked and grown in LB in 96-well plates. PCR amplification of inserts was performed by using T3 and T7 primers. PCR products were ethanol precipitated and redissolved in 3× standard saline citrate (SSC) before printing on L-lysine coated glass slides. The array was printed by using a GeneMachines OmniGrid Arrayer and 32 pins.

Fluorescent Probe Preparation and Array Hybridization

Tissue was dissected from chick embryos as diagrammed in Figure 2 (two pieces AL, four pieces AM, or two pieces PS), dissolved in Trizol directly, and processed for aRNA as described (Baugh et al., 2001). In our hands, one round of RNA amplification yields enough aRNA (4–10 μg) to produced cDNA probes for two to five hybridizations (0.7–2 μg of aRNA were used to produce each probe). cDNA probes were produced by reverse transcription of aRNA, incorporating aminoallyl-dUTP (aa-dUTP; Sigma) into the reaction buffer. Cy3 and Cy5 were linked to the aa-dUTP residues of the cDNA by using Fluorolink Cy3 and Cy5 Dye Packs (Amersham). Labeled probe was hybridized to arrays for 12 hr at 65°C and then washed in 0.5× SSC, 0.25% SDS, followed by a 0.05× SSC wash. Microarrays were scanned by using a GenePix 4000B scanner and analyzed by using GenePix 4.0 software.

Sequencing and In Situ Hybridization

Clones picked for sequencing were recovered from stored 96-well plates (stored at −80°C) and sequenced using T3 primers. Whole-mount in situ hybridization was performed as previously described (Schultheiss et al., 1995). Probes were generated by cutting plasmids with EcoRI and transcribing with T7 RNA polymerase.


  1. Top of page
  2. Abstract
  7. Acknowledgements

The authors thank Claire Bailey and Tyler Aldridge of the Harvard Bauer Center for Genomics Research for advice and assistance in construction of the microarray, and Cliff Tabin and Connie Cepko for making available resources and equipment for microarray construction. T.M.S. was funded by a Scientist Development Grant from the American Heart Association.


  1. Top of page
  2. Abstract
  7. Acknowledgements
  • Alsan BH, Schultheiss TM. 2002. Regulation of avian cardiogenesis by Fgf8 signaling. Development 129: 19351943.
  • Andree B, Duprez D, Vorbusch B, Arnold HH, Brand T. 1998. BMP-2 induces ectopic expression of cardiac lineage markers and interferes with somite formation in chicken embryos. Mech Dev 70: 119131.
  • Antin PB, Bates MA, Zhang W, Garriock RJ, Yatskievych TA. 2002. Precocious expression of cardiac troponin T in early chick embryos is independent of bone morphogenetic protein signaling. Dev Dyn 225: 135141.
  • Arceci RJ, King AAJ, Simon C, Orkin SH, Wilson DB. 1993. Mouse GATA-4: a retinoic acid-inducible GATA-binding transcription factor expressed in endodermally derived tissues and heart. Mol Cell Biol 13: 22352246.
  • Basson CT, Bachinsky DR, Lin RC, Levi T, Elkins JA, Soults J, Grayzel D, Kroumpouzou E, Traill TA, Leblanc-Straceski J, Renault B, Kucherlapati R, Seidman JG, Seidman CE. 1997. Mutations in human TBX5 [corrected] cause limb and cardiac malformation in Holt-Oram syndrome. Nat Genet 15: 3035.
  • Baugh LR, Hill AA, Brown EL, Hunter CP. 2001. Quantitative analysis of mRNA amplification by in vitro transcription. Nucleic Acids Res 29: E29.
  • Biehl J, Holtzer S, Bennett G, Sun T, Holtzer H. 1985. Cultured chick blastodisc cells diverge into lineages with different IF isoforms. Ann N Y Acad Sci 455: 158166.
  • Boardman PE, Sanz-Ezquerro J, Overton IM, Burt DW, Bosch E, Fong WT, Tickle C, Brown WR, Wilson SA, Hubbard SJ. 2002. A comprehensive collection of chicken cDNAs. Curr Biol 12: 19651969.
  • Bruneau BG, Nemer G, Schmitt JP, Charron F, Robitaille L, Caron S, Conner DA, Gessler M, Nemer M, Seidman CE, Seidman JG. 2001. A murine model of Holt-Oram syndrome defines roles of the T-box transcription factor Tbx5 in cardiogenesis and disease. Cell 106: 709721.
  • Burt D, Pourquie O. 2003. Genetics. Chicken genome—science nuggets to come soon. Science 300: 1669.
  • Chapman DL, Garvey N, Hancock S, Alexiou M, Agulnik SI, Gibson-Brown JJ, Cebra-Thomas J, Bollag RJ, Silver LM, Papaioannou VE. 1996. Expression of the T-box family genes, Tbx1-Tbx5, during early mouse development. Dev Dyn 206: 379390.
  • Davidson E. 1986. Gene activity in early development. Orlando: Academic Press.
  • Egeo A, Di Lisi R, Sandri C, Mazzocco M, Lapide M, Schiaffino S, Scartezzini P. 2000. Developmental expression of the SH3BGR gene, mapping to the Down syndrome heart critical region. Mech Dev 90: 313316.
  • Garcia-Martinez V, Schoenwolf GC. 1993. Primitive streak origin of the cardiovascular system in avian embryos. Dev Biol 159: 706719.
  • Gonzalez-Sanchez A, Bader D. 1990. In vitro analysis of cardiac progenitor cell differentiation. Dev Biol 139: 197209.
  • Hamagashira N, Oku H, Mega T, Hase S. 1996. Purification and characterization of hen oviduct alpha 1,2-mannosidase. J Biochem (Tokyo) 119: 9981003.
  • Han Y, Dennis JE, Cohen-Gould L, Bader DM, Fischman DA. 1992. Expression of sarcomeric myosin in the presumptive myocardium of chicken embryos occurs within six hours of myocyte commitment. Dev Dyn 193: 257265.
  • Hayward LJ, Schwartz RJ. 1986. Sequential expression of chicken actin genes during myogenesis. J Cell Biol 102: 14851493.
  • Heller S, Sheane CA, Javed Z, Hudspeth AJ. 1998. Molecular markers for cell types of the inner ear and candidate genes for hearing disorders. Proc Natl Acad Sci U S A 95: 1140011405.
  • Hill RE, Jones PF, Rees AR, Sime CM, Justice MJ, Copeland NG, Jenkins NA, Graham E, Davidson DR. 1989. A new family of mouse homeo box-containing genes: molecular structure, chromosomal location, and developmental expression of Hox-7.1. Genes Dev 3: 2637.
  • Iio A, Koide M, Hidaka K, Morisaki T. 2001. Expression pattern of novel chick T-box gene, Tbx20. Dev Genes Evol 211: 559562.
  • Jen Y, Manova K, Benezra R. 1996. Expression patterns of Id1, Id2, and Id3 are highly related but distinct from that of Id4 during mouse embryogenesis. Dev Dyn 207: 235252.
  • Jiang Y, Evans T. 1996. The Xenopus GATA-4/5/6 genes are associated with cardiac specification and can regulate cardiac-specific transcription during embryogenesis. Dev Biol 174: 258270.
  • Klar A, Baldassare M, Jessell TM. 1992. F-Spondin: a gene expressed at high levels in the floor plate encodes a secreted protein that promotes neural cell adhesion and neurite extension. Cell 69: 95110.
  • Kuo CT. 1997. GATA4 transcription factor is required for ventral morphogenesis and heart tube formation. Genes Dev 118: 10481060.
  • Kupperman E, An S, Osborne N, Waldron S, Stainier DY. 2000. A sphingosine-1-phosphate receptor regulates cell migration during vertebrate heart development. Nature 406: 192195.
  • Laverriere AC, MacNeill C, Mueller C, Poelmann RE, Burch JBE, Evans T. 1994. GATA-4/5/6, a subfamily of three transcription factors transcribed in developing heart and gut. J Biol Chem 269: 2317723184.
  • Le Douarin N. 1975. An experimental analysis of liver development. Medi Biol 23: 427455.
  • Lee K-H, Xu Q, Breitbart RE. 1996. A new tinman-related gene, nkx-2.7, anticipates the expression of nkx-2.5 and nkx-2.3 in zebrafish heart and pharyngeal endoderm. Dev Biol 180: 722731.
  • Lekanne Deprez RH, Riegman PH, Groen NA, Warringa UL, van Biezen NA, Molijn AC, Bootsma D, de Jong PJ, Menon AG, Kley NA, et al. 1995. Cloning and characterization of MN1, a gene from chromosome 22q11, which is disrupted by a balanced translocation in a meningioma. Oncogene 10: 15211528.
  • Lemaire P, Garrett N, Kato K, Gurdon JB. 1993. Construction of subtracted cDNA libraries enriched for cDNAs for genes expressed in the mesoderm of early Xenopus gastrulae. C R Acad Sci III 316: 931944.
  • Lints TJ, Parsons LM, Hartley L, Lyons I, Harvey RP. 1993. Nkx-2.5: a novel murine homeobox gene expressed in early heart progenitor cells and their myogenic descendants. Development 119: 419431.
  • Lockhart DJ, Dong H, Byrne MC, Follettie MT, Gallo MV, Chee MS, Mittmann M, Wang C, Kobayashi M, Horton H, Brown EL. 1996. Expression monitoring by hybridization to high-density oligonucleotide arrays. Nat Biotechnol 14: 16751680.
  • Marvin MJ, Di Rocco G, Gardiner A, Bush SM, Lassar AB. 2001. Inhibition of Wnt activity induces heart formation from posterior mesoderm. Genes Dev 15: 316327.
  • McNagny KM, Pettersson I, Rossi F, Flamme I, Shevchenko A, Mann M, Graf T. 1997. Thrombomucin, a novel cell surface protein that defines thrombocytes and multipotent hematopoietic progenitors. J Cell Biol 138: 13951407.
  • Molkentin JD, Lin Q, Duncan SA, Olson EN. 1997. Requirement of the transcription factor GATA4 for heart tube formation and ventral morphogenesis. Genes Dev 118: 10611072.
  • Moskow JJ, Bullrich F, Huebner K, Daar IO, Buchberg AM. 1995. Meis1, a PBX1-related homeobox gene involved in myeloid leukemia in BXH-2 mice. Mol Cell Biol 15: 54345443.
  • Nascone N, Mercola M. 1995. An inductive role for the endoderm in Xenopus cardiogenesis. Development 121: 515523.
  • Onichtchouk D, Chen YG, Dosch R, Gawantka V, Delius H, Massague J, Niehrs C. 1999. Silencing of TGF-beta signalling by the pseudoreceptor BAMBI. Nature 401: 480485.
  • Pandur P, Lasche M, Eisenberg LM, Kuhl M. 2002. Wnt-11 activation of a non-canonical Wnt signalling pathway is required for cardiogenesis. Nature 418: 636641.
  • Reifers F, Walsh EC, Leger S, Stainier DY, Brand M. 2000. Induction and differentiation of the zebrafish heart requires fibroblast growth factor 8 (fgf8/acerebellar). Development 127: 225235.
  • Reiter JF, Alexander J, Rodaway A, Yelon D, Patient R, Holder N, Stainier DY. 1999. Gata5 is required for the development of the heart and endoderm in zebrafish. Genes Dev 13: 29832995.
  • Rosenquist GC, DeHaan RL. 1966. Migration of precardiac cells in the chick embryo: a radioautographic study. Carnegie Inst Washington Contrib Embryol 38: 111121.
  • Ruiz JC, Conlon FL, Robertson EJ. 1994. Identification of novel protein kinases expressed in the myocardium of the developing mouse heart. Mech Dev 48: 153164.
  • Sambrook J, Fritsch EF, Maniatis T. 1989. Molecular cloning. New York: Cold Spring Harbor Laboratory Press.
  • Scartezzini P, Egeo A, Colella S, Fumagalli P, Arrigo P, Nizetic D, Taramelli R, Rasore-Quartino A. 1997. Cloning a new human gene from chromosome 21q22.3 encoding a glutamic acid-rich protein expressed in heart and skeletal muscle. Hum Genet 99: 387392.
  • Schena M, Shalon D, Davis RW, Brown PO. 1995. Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 270: 467470.
  • Schneider VA, Mercola M. 2001. Wnt antagonism initiates cardiogenesis in Xenopus laevis. Genes Dev 15: 304315.
  • Schoenwolf GC, Garcia-Martinez V, Dias MS. 1992. Mesoderm movement and fate during avian gastrulation and neurulation. Dev Dyn 193: 235248.
  • Schultheiss TM, Lassar AB. 1997. Induction of chick cardiac myogenesis by bone morphogenetic proteins. Cold Spring Harb Symp Quant Biol 62: 413419.
  • Schultheiss TM, Xydas S, Lassar AB. 1995. Induction of avian cardiac myogenesis by anterior endoderm. Development 121: 42034214.
  • Schultheiss TM, Burch JBE, Lassar AB. 1997. A role for bone morphogenetic proteins in the induction of cardiac myogenesis. Genes Dev 11: 451462.
  • Schweinfest CW, Henderson KW, Gu JR, Kottaridis SD, Besbeas S, Panotopoulou E, Papas TS. 1990. Subtraction hybridization cDNA libraries from colon carcinoma and hepatic cancer. Genet Anal Tech Appl 7: 6470.
  • Schweinfest C, Nelson P, Graber M, Demopoulos R, Papas T. 1995. Subtraction hybridization cDNA libraries. In: TymmsM, editor. In vitro transcription and translation protocols. Totowa, NJ: Humana Press, Inc. p 1330.
  • Shin CH, Liu ZP, Passier R, Zhang CL, Wang DZ, Harris TM, Yamagishi H, Richardson JA, Childs G, Olson EN. 2002. Modulation of cardiac growth and development by HOP, an unusual homeodomain protein. Cell 110: 725735.
  • Srivastava D, Thomas T, Lin Q, Kirby ML, Brown D, Olson EN. 1997. Regulation of cardiac mesodermal and neural crest development by the bHLH transcription factor, dHAND. Nat Genet 16: 154160.
  • Tonissen KF, Drysdale TA, Lints TJ, Harvey RP, Krieg PA. 1994. XNkx-2.5, a Xenopus Gene Related to Nkx-2.5 and tinman: evidence for a conserved role in cardiac development. Dev Biol 162: 325328.
  • Tzahor E, Lassar AB. 2001. Wnt signals from the neural tube block ectopic cardiogenesis. Genes Dev 15: 255260.
  • Walsh EC, Stainier DY. 2001. UDP-glucose dehydrogenase required for cardiac valve formation in zebrafish. Science 293: 16701673.
  • Wang D, Chang PS, Wang Z, Sutherland L, Richardson JA, Small E, Krieg PA, Olson EN. 2001. Activation of cardiac gene expression by myocardin, a transcriptional cofactor for serum response factor. Cell 105: 851862.
  • Zhang H, Bradley A. 1996. Mice deficient for BMP2 are nonviable and have defects in amnion/chorion and cardiac development. Development 122: 29772986.