We have isolated, sequenced and examined the expression pattern of two tandemly arranged homeobox- containing genes from the chicken. The predicted amino acid sequences of the homeodomain and the adjacent carboxyterminal portion of the protein of the first gene is virtually identical (99%) to that of murine homeobox 2.1 and hence we refer to it as Ghox 2.1 (Gallus Aomeobox). The closest mouse homologue of the second homeodomain is Hox 2.2 (95% identical within the homeobox), and hence referred to as Ghox 2.2. Northern analysis of embryonic RNA reveals major transcripts of 2 kb for Ghox 2.1 and 1·7 kb for Ghox 2.2. To investigate the transcript pattern, embryos of various stages were dissected into heads, trunks and limb buds and the RNA was analysed by Northern blotting and RNase protection assays. Ghox 2.1 transcripts are present in all three regions. Ghox 2.2 RNA is found in trunks and limb buds, but it is strikingly absent from the developing head. In situ hybridization with 35S-labelled antisense riboprobes derived from Ghox 2.1 demonstrates that this gene is expressed at high levels in spinal chord, myelencephalon and mesonephros. Dorsal root ganglia and the lung rudiment also contain Ghox 2.1 message, but in somewhat lower amounts. Mid- and forebrain, the heart, presomitic mesenchyme and notochord do not contain detectable levels of Ghox 2.1 mRNA. Of particular interest is the expression of Ghox 2.1 in a well-defined patch of mesenchymal tissue situated in an anterioproximal region of the limb bud.

There is much current interest in the function of homeobox-containing genes in the embryogenesis of invertebrates and vertebrates (Gehring, 1987; Feinberg et al. 1987). It has been suggested that homeobox-containing genes have a key role in orchestrating development (Gehring, 1985). To date, genes carrying the 180 bp long homeobox motif have been found in all metazoan genomes appropriately examined (McGinnis et al. 1984a), including those of insects (McGinnis et al. 1984b; Scott & Weiner, 1984; Beeman, 1987), sea urchins (Dolecki et al. 1986), nematodes (Way & Chalfie, 1988), and vertebrates (e.g. Boncinelli et al. 1985; Carrasco et al. 1984; McGinnis et al. 1984c; Falzon & Chung, 1988). Such a wide distribution of homeo-boxes and their extraordinary degree of sequence conservation suggests a fundamental biological function; yet this unversality also makes it difficult to elucidate their specific role in the development of species that have diverged as much as sea urchins and man.

To gain understanding of the role of homeobox-containing genes in vertebrate embryogenesis, several avenues of research are currently being pursued. First, spatiotemporal expression patterns are being studied (e.g. Breier et al. 1988; Carrasco & Malacinski, 1987; Davis et al. 1988; Holland & Hogan, 1988a; Utset et al. 1987). Second, attempts are made to experimentally alter the expression pattern and study the consequences on the body plan (Harvey & Melton, 1988) or to examine the expresssion pattern as a consequence of experimental embryonic induction (Sharpe et al. 1987). A third strategy involves intra- and interspecies comparisons of gene structures and of their spatiotemporal expression pattern (Hart et al. 1987; Holland & Hogan, 1988b; Gaunt, 1988; Oliver et al. 1988).

In many ways the chick embryo is a favourable experimental system to study the mechanisms of vertebrate development using any of the above approaches. The chief attraction lies in its accessibility for experimental manipulations, even at the earliest stages. These manipulations range from organizer grafting and making chimeras to gene transfer experiments (Waddington, 1934; Saunders & Gasseling, 1968; Hara, 1978; Le Douarin, 1973; Gray et al. 1988). Accordingly, a substantial body of data on the embryology of the chick has been collected (Romanoff, 1960; Patten, 1951; Lillie, 1952). Furthermore, birds provide a link between amphibians and mammals and thus contribute to interspecies comparisons of homeobox-containing genes across an evolutionary distance of several hundred million years (Feduccia, 1980). Very likely, these strengths will help to elucidate the role of homeobox-containing genes. As a first step in this direction, we have cloned several homeobox-containing genes from the chicken. Here we report the molecular characterization of two chick homeoboxes and describe the embryonic expression pattern of one of them.

Isolation of homeobox-containing chicken genomic DNAs

Chick homeobox-containing DNA fragments were isolated by screening a genomic chick library kindly provided by Dr D. Engel (Northwestern University). The library had been prepared by partial digestion of erythrocyte DNA with Mbol followed by cloning of the fragments into the Sall site of EMBL 3 bacteriophage. The phage was plated onto 24×24cm dishes, double replica filters were prepared (Genescreen Plus, Dupont) which were autoclaved at 100 °C for 2 min and prewashed for 1 h in 50 mm-Tris-HCl (pH = 8), Iw-NaCl, Imm-EDTA, 0·1% SDS. Prehybridization was performed at 37°C for 1 h in 10mm-Tris-HCl (pH = 7·6), 40% formamide, 4 × SSC (1 × SSC = 150 mm-sodium chloride, 15 mm-sodium citrate, pH = 7), 0·8×Denhardt’s solution, 1% SDS and 20 μg ml−1 sonicated salmon sperm DNA. Subsequently 107ctsmin−1 of nick-translated 32P-labelled probe (Rigby et al. 1977) were added, and the membranes were hybridized overnight at 37 °C. This was followed by washing twice for 15 min at room temperature in 2 × SSC, 0·1% SDS, and two 15 min washes in 0-2 x SSC, 0-1% SDS at 48°C. Plaque purification was performed using Colony/Plaque Screen nylon membranes (Dupont) using essentially the same procedures.

DNA sequencing

Positive clones were subjected to restriction analysis and appropriate fragments were subcloned into bacteriophage M 13 or plasmids pGem 3 or pGem 3z (Promega) and sequenced as single-stranded or double-stranded DNA by the dideoxynucleotide method (Sanger et al. 1977; Toneguzzo et al. 1988). Every sequence was determined several times. Constructs used for Northern blots, RNase protection assays and in situ hybridization were sequenced prior to use.

RNA isolation and Northern blotting

Tissues were obtained from chick embryos incubated at 37-5°C to the desired stage (Hamburger & Hamilton, 1951). RNA was prepared using the guanidine isothiocyanate method described in Davis et al. (1986). Tissue was rinsed in ice-cold phosphate-buffered saline (PBS) and collected into tubes embedded in dry ice. 5M-guanidine isothiocyanate was added to the frozen samples and the tissue homogenized immediately with a Polytron homogenizer. The RNA was pelleted through a CsCl cushion, resuspended in diethylpyro-carbonate-treated water, phenol-chloroform extracted and ethanol precipitated.

Poly (A)+ RNA was selected by passing total RNA over a oligo-dT cellulose column (Collaborative Research) using a procedure based on Davis et al. (1986).

For Northern blotting, 1% formaldehyde-agarose gels were prepared as described by Davis et al. (1986). 2 or 4 μg of poly (A)+ RNA were loaded in each well and the gel was run at 150 volts in MOPS buffer (pH = 6) for 3h. The gel was treated with 50 mm-NaOH for 30 min, rinsed in water, neutralized in 0·1 m-Tris-HCl (pH = 8) (twice, for 15 min), and soaked in 20 × SSC for 30min. Overnight blotting with 20 × SSC onto a Gene Screen membrane was followed by u.v. photocrosslinking (Church & Gilbert, 1984). Filters were prehybridized and hybridized at 42 °C as described for the library screen, except that the hybridization buffer also contained 10% dextran sulphate plus 0·1 μg ml−1 synthetic poly A RNA and the probes were synthesized by the random primer procedure (Feinberg & Vogelstein, 1983; 1984). After hybridization filters were washed twice for 15 min in 2 × SSC, 0·1% SDS at room temperature, followed by a 30 min wash in 0·2 × SSC, 0·1% SDS at 60°C.

RNase protection assays

The synthesis of 32P-riboprobes was based on a protocol described by Krieg & Melton (1988). Vectors containing the probes were linearized with EcoRI (for use with SP6 promoter) or Hmdlll (for T7 promoter). The riboprobe synthesis reaction mixture (20 μl) was composed as follows: 0·5 mm-rATP, rCTP, rGTP, 25 μM-rUTP, 40 mm-Tris-HCl (pH = 7·5), 6mm-MgCl2, 2mm-spermidine, lOmm-DTT, 50 units of RNase inhibitor (Boehringer), 6·25 qM-32P-UTP (800 Ci mmol−1) and 1 μg template DNA. The reaction (carried out at 37 °C) was initiated by adding 30 units of polymerase. At 45min, 15 additional units of enzyme were provided. At 80 min, the sample was treated for 15 min with 3 units of DNase I (Promega Biotech), then phenol-chloroform extracted and ethanol precipitated. To recover full-length transcript, the RNA was fractionated on a 6% acrylamide gel and the full-length riboprobe was eluted into 0·1% SDS, Imm-EDTA, 500mm-ammonium acetate, 10mm-magnesium acetate. After phenol-chloroform extraction and ethanol precipitation, the probe was resuspended in water to give 105 cts min−1μl−1.

1 μl of probe was hybridized overnight at 50 °C to 50 μg total RNA (or tRNA as a control) in 50% formamide, 0·4m-NaCl, 40mm-Pipes (pH = 6·4) and Imm-EDTA in a total reaction volume of 80 μl. Subsequently, the residual single-stranded RNA was digested with RNase A (40 μg ml−1) and RNase T1 (2 μg ml−1) for 15 min at room temperature. The samples were then treated for 15 min (37 °C) with proteinase K (0·14 mg ml−1) and SDS (0·5%). Following phenol-chloroform extraction and ethanol precipitation, the protected RNA was analysed on a 6% sequencing gel run at 60 W for 1–2h. HinQH cut pBR322 that had been end-labelled with 35S-ATP was used as a size standard.

In situ hybridization

Riboprobe synthesis

35S-labelled riboprobes were synthesized at 37°C in 20 μl of the following reaction mixture: 0·3 mm each of rATP, rCT’P, rGTP, 3 μM-35S-UTP (1000Ci mmol−1), 40mm-Tris-HCl (pH = 7·5), 6mm-MgCl2, 2mm-spermidine, 10mm-DTT, 50 units of RNase inhibitor and 1 μg of template DNA. The reaction was initiated by addition of 30 units of polymerase and spiked with a further 30 units after 45 min of incubation. At 90min, the reaction was treated with DNase, phenolchloroform extracted and ethanol precipitated twice. The RNA was then resuspended, at a concentration of 1 ngμ−1, in 50% formamide containing 50mm-DTT and stored at –80°C. Typically the specific activity achieved was about 109disintsmin−1μg−1 RNA.

Fixation, embedding and sectioning

After dissection, stage-17 to -25 embryos were rinsed in PBS and placed immediately in ice-cold fixative (85% ethanol, 5% glacial acetic acid, 4% formaldehyde) for approximately 30 min. Subsequently, the embryos were dehydrated through ethanol, taken into xylene and then into Paraplast (Monoject). 5 μm sections were cut and placed on poly-L-lysine-coated slides. Once dried at 42°C overnight, the slides were either processed or stored at −20°C with desiccant for a period of one week to one month. The slides were prepared by first washing them in 1% Lubrol detergent (Linbro, Flow Laboratories), followed by rinsing in water and drying at 150°C for 1h. Slides were then placed into a solution of 100 μg ml−1 poly-L-lysine hydrobromide (Sigma) for 1h, whereupon they were dried and stored in a vacuum desiccator prior to use.

Section pretreatment, hybridization reaction and washing

For hybridization, sections were brought to room temperature. They were dewaxed in xylene and rehydrated through a decreasing ethanol series into water. Slides were then immersed in 0·2 N-HCI for 20 min at room temperature. The acid was rinsed off with PBS and the slides were transferred into 2 × SSC and incubated at 70°C for 20 min. After rinsing with saline, the tissue was postfixed in 4% paraformaldehyde in PBS for 10 min at room temperature. After an additional rinse in saline, the sections were acetylated twice for 5 min in freshly prepared 26mm-acetic anhydride in triethanolamine (0·1 M, pH = 8). The sections were then placed in 2 × SSC for 15 min and either used immediately for hybridization or they were dried and stored at −80°C with desiccant for next day use.

Sections were prehybridized for 1 h at room temperature in 75/d hybridization buffer (300mm-NaCl, 0·1 mm-Tris-HCl [pH = 6·8], 01M-NaH2PO4 [pH = 6·8], 50ITIm-EDTA, 0·2% w/v Ficoll 400, 0·2% polyvinylpyrrolidone, 50% deionized formamide, 10% dextran sulphate, 500 pg ml−1 yeast tRNA, 500 μg ml−1 synthetic poly A RNA, 50mm-DTT and 250mm-thio-ATP). 35S-riboprobe was added to hybridization buffer, heated to 90°C for 3 min and quenched on ice. 25 μl of the denatured antisense or sense probe were added to each slide and a siliconized cover slip was placed on top of the preparation. Final probe concentration was approximately 0·02 ng μl−1. Hybridization was carried out for 3h in a humidified chamber kept at 50°C. Prior to washing, the coverslips were flushed off and the slides immersed twice for 45 min in a 50°C solution of 2 × SSC, 50% formamide and 20 mm-2-mercaptoethanol (SFM). This was followed by washing once at 50°C in 4 × SSC, 10 mm-Tris-HCl (pH = 7·7), Imm-EDTA and 20mm-2-mercaptoethanol (STEM). Subsequently the sections were incubated in 20 μg ml−1 RNase A (dissolved in STEM) for 30 min at 37°C, followed by a wash in STEM only. Finally, slides were washed in SFM at 50°C and then in 2 × SSC at room temperature. Slides were dehydrated through ethanol containing 300 mm-ammonium acetate and dried.

Autoradiography

Slides were exposed to X-ray film (Cronex, Dupont) for 1-3 days to examine both the quality of the hybridization and to estimate the exposure time. Subsequently, slides were dipped in Kodak NTB-2 emulsion (1:1 diluted with water) and air dried at room temperature. They were placed into light-tight, desiccant-containing boxes and stored at 4°C. The slides were prewarmed to room temperature before developing in Kodak D 19 developer, rinsed in water, fixed and rinsed again for > 15 min. After air drying, the sections were first stained in 0·2% toluidine blue and then dehydrated through ethanol, transferred to xylene and enclosed in Permount (Fisher Scientific).

Isolation of chicken genomic clones that contain homeoboxes

In an initial effort to search for homeobox-containing genes in the chicken genome, we probed various restriction digests of genomic chicken DNA with a homeobox-containing 0-6 kbp BamHI/PvuII fragment excised from plasmid 903 G (McGinnis et al. 1984a), which is derived from the Antennapedia locus of Drosophila. Blots were washed at moderate stringency (38°C in 0·2 × SSC). Between seven and nine crosshybridizing bands were observed (data not shown), confirming earlier work by McGinnis et al. (1984a) that the chick genome contains Antennapedia-hkc homeoboxes, and indicating that it should be feasible to use this probe for isolating homeobox-containing clones from a chicken genomic library. Hence a genomic library of recombinant EMBL 3 lambda phage was screened with the BamUl/PvuII fragment under reduced stringency. About 106 phage plaques were included, corresponding approximately to ten chicken genome equivalents. A total of 20 significantly crosshybridizing clones were identified in the primary screening and 16 of them were subsequently plaque purified. Restriction digestions and Southern transfer showed that all of the 16 phages were recombinants and bore four distinct inserts (data not shown). Here we report a partial molecular characterization of one of the EMBL 3 clones (lambda 17) that has two homeobox-containing genes which are located at an interval of about 3 kbp (H. Busse and G. Eichele, unpublished observation).

Sequence analysis suggests homology with mouse homeboxes Hox 2.1 and Hox 2.2

The second and third line in Fig. 1A shows the DNA sequence of both avian homeoboxes including part of their 5’ and 3’ flanking DNA. Comparison with the Antennapedia sequence (top line of Fig. 1A) reveals that within the homeobox (bp 1 to 180) the two chicken sequences are 72% (Ghox 2.1) and 77% (Ghox 2.2) identical with the Ant homeobox. Outside the homeobox, as far as we have sequence available, we do not see significant similarities amongst the chicken genes or between chicken and Drosophila sequences. The top lines in Fig. 2A & B display the predicted amino acid sequences of the two avian homeodomains. We have compared the two sequences with the currently known mouse homeodomains. We found that the homeo-domain of Ghox 2.1 is identical to that of Hox 1.3 (Odenwald et al. 1987) and 2.1 (Krumlauf et al. 1987). However, in a detailed comparison of the carboxy- and amino-terminal sequences, it becomes clear that Ghox 2.1 is most likely the chick homologue of Hox 2.1 (see Fig. 2A). Ghox 2.1, Hox 2.1 and also human homeobox Hu 1 (Levine et al. 1984), share a virtually identical upstream sequence. This is in stark contrast to the upstream sequence of Hox 1.3 (Fig. 2A). Furthermore, Ghox 2.1, Hox 2.1 and Hu 1 are almost identical in their carboxy-terminal peptide, with a single conservative substitution at position 69 (Thr to Ser).

Fig. 1

(A) DNA sequence of the homeobox plus 5’ and 3’ flanking regions of Antp from D. melanogaster (Schneuwly et al. 1986) and chicken homeobox genes Ghox 2.1 and Ghox 2.2. The homeobox region and stop codons are underlined. Krumlauf et al. (1987) have identified the 5’ boundary of the Hox 2.1 homeobox-containing exon at 18 nucleotides 5’ to the beginning of the homeobox. A comparison of nucleotide sequences of Ghox 2.1 and Hox 2.1 reveals that both sequences are essentially identical back to this nucleotide 18 and then diverge completely, suggesting that the splice sites of the two genes are in the same positions. (B,C) Schematic representation of the chicken sequences shown in Fig. 1A to illustrate the DNA (A,B) and RNA (C,D,E) probes used in this study. The homeobox is the shaded area. Sense (+) and antisense (−) riboprobes are indicated by arrows and extend between the indicated restriction sites.

Fig. 1

(A) DNA sequence of the homeobox plus 5’ and 3’ flanking regions of Antp from D. melanogaster (Schneuwly et al. 1986) and chicken homeobox genes Ghox 2.1 and Ghox 2.2. The homeobox region and stop codons are underlined. Krumlauf et al. (1987) have identified the 5’ boundary of the Hox 2.1 homeobox-containing exon at 18 nucleotides 5’ to the beginning of the homeobox. A comparison of nucleotide sequences of Ghox 2.1 and Hox 2.1 reveals that both sequences are essentially identical back to this nucleotide 18 and then diverge completely, suggesting that the splice sites of the two genes are in the same positions. (B,C) Schematic representation of the chicken sequences shown in Fig. 1A to illustrate the DNA (A,B) and RNA (C,D,E) probes used in this study. The homeobox is the shaded area. Sense (+) and antisense (−) riboprobes are indicated by arrows and extend between the indicated restriction sites.

Fig. 2

Comparison of the predicted amino acid sequences of the two chicken genes suggesting that they are the avian homologues of Hox 2.1 and Hox 2.2 (mouse) and Hu 1 and Hu 2 (man), respectively. A match of sequence is indicated by a dash. The homeodomain is underlined. (A) Ghox 2.1 compared with murine sequences Hox 2.1 (Krumlauf et al. 1987) and Hox 1.3 (Odenwald et al. 1987) and the human homeobox Hu 1 (Levine et al. 1984). (B) Sequence comparison of Ghox 2.2 with Hox 2.2 (Hart et al. 1987) and Hox 1.2 (Colberg-Poley et al. 1985) from mouse and human homeobox Hu 2 (Levine et al. 1984).

Fig. 2

Comparison of the predicted amino acid sequences of the two chicken genes suggesting that they are the avian homologues of Hox 2.1 and Hox 2.2 (mouse) and Hu 1 and Hu 2 (man), respectively. A match of sequence is indicated by a dash. The homeodomain is underlined. (A) Ghox 2.1 compared with murine sequences Hox 2.1 (Krumlauf et al. 1987) and Hox 1.3 (Odenwald et al. 1987) and the human homeobox Hu 1 (Levine et al. 1984). (B) Sequence comparison of Ghox 2.2 with Hox 2.2 (Hart et al. 1987) and Hox 1.2 (Colberg-Poley et al. 1985) from mouse and human homeobox Hu 2 (Levine et al. 1984).

Assuming that the chromosomal organization of homeobox genes is the same in birds and mammals, we expect that the 5’ neighbour of Ghox 2.1 is the homologue of Hox 2.2 (Hart et al. 1987) and human homeobox Hu 2 (Levine et al. 1984). This suggestion is supported by a comparison of Ghox 2.2 with currently known mouse homeoboxes. Both Hox 1.2 and 2.2 are 95% identical to Ghox 2.2. However, on closer examination of the carboxyterminal peptides (Fig. 2B), it is very clear that Ghox 2.2 matches Hox 2.2 much better than Hox 1.2. Specifically, out of 19 amino acids at the carboxy terminus of Ghox 2.2, there is 84% identity with Hox 2.2, but only 47% with Hox 1.2 (Fig. 2B). As Hart et al. (1987) have noted, Hox 2.2 and the human homeobox Hu 2 are earmarked by a run of five glutamic acid residues at the carboxy terminus. These residues are also perfectly conserved in Ghox 2.2, but are not present in Hox 1.2. A difference between Hox 2.2, and Ghox 2.2 and Hu 2 is that the murine gene has an additional amino acid at position −1. Based on these pieces of evidence and the similarity of the spatial expression pattern of Ghox 2.1 and Hox 2.1 (see below), we conclude that the avian homeobox-containing genes described in this report are the chick homologues of Hox 2.1 and Hox 2.2, and of Hu 1 and Hu 2, respectively.

Ghox 2.1 and 2.2 are expressed during late somitogenesis and organogenesis

To obtain an overall picture of the temporal expression pattern of Ghox 2.1 and 2.2, we have used Northern blots and RNase protection assays. The period covered in this study ranges from Hamburger and Hamilton stage 16 to stage 34. This time period spans the later phase of early development (somitogenesis, neural tube differentiation), and the early phase of organogenesis that is followed by extensive growth (Hamburger & Hamilton, 1951; Lillie, 1952). Also note that these chick stages correspond approximately to mouse embryos of 8-5 to 16 days p.c. (Rugh, 1977). The Northern blots in Fig. 3A, (lanes 1–3) represent the hybridization pattern when poly (A)+ RNA isolated from chick embryos is probed with fragment A from Ghox 2.1 (see Fig. 1B). These Northern blots reproducibly display a broad range of transcripts between 1·6 and 2 kb, with the major transcript at 2 kb. In addition, there are minor bands at 3·5 and 9·4 kb and a very weak band at 4·4kb. Probe B, derived from Ghox 2.2 (see Fig. 1C) (lanes 4–6 of Fig. 3A) reveals a major transcript at 1·7 kb and minor ones at 4-4 and 8·8 kb.

Fig. 3

(A). Northern blot analysis of embryonic transcripts showing expression of Ghox 2.1 and 2.2. Lanes 1–3 each contain 2 μg of poly (A)+ RNA and were hybridized with probe A (Ghox 2.1). Lanes 4–6 each contain 4 μg of poly (A)+ RNA and were hybridized with probe B (Ghox 2.2). Chick ribosomal RNA (1·9 and 4·6 kb) was used as a size marker. Lanes 1,4 - RNA from trunks of stage-20 embryos. Lanes 2,5 - RNA from trunks of stage-22 embryos. Lanes 3,6 - RNA from trunks of stage-24 embryos. (B-E) RNase protection assays of total RNA from chick embryos hybridized with antisense probes of Ghox 2.1 and 2.2. The bands represent full-length protection. As a control we also hybridized with sense probes D(+) and E(+), and see no protection (data not shown). (B) RNA hybridized with probe D(−) of Ghox 2.1: Lane 1 - RNA from trunks of stage-30 embryos; Lane 2 - RNA from heads of stage-20 embryos; Lane 3 - RNA from heads of stage-22 embryos; Lane 4 - RNA from heads of stage-24 embryos; Lane 5 - tRNA control; Lane 6 - probe only. Note that Lane 2 has been underloaded. Other experiments show that there is no reduction in the level of expression at this stage. (C) RNA hybridized with probe E(−) of Ghox 2.2: Lane 1 - RNA from trunks of stage-30 embryos; Lane 2 - RNA from heads of stage-20 embryos; Lane 3 - RNA from heads of stage-22 embryos; Lane 4 - RNA from heads of stage-24 embryos; Lane 5 - tRNA control; Lane 6 - probe only. (D) RNA hybridized with probe D(−) of Ghox 2.1: Lane 1 - probe; Lane 2 - tRNA control; Lane 3 - RNA from stage-30 limb buds; Lane 4 - RNA from stage-28 hind limb buds; Lane 5 - RNA from stage-22 limb buds. (E) RNA hybridized with probe E(−) of Ghox 2.2: Lane 1 - probe; Lane 2 - tRNA control; Lane 3 - RNA from stage-30 limb buds; Lane 4 - RNA from stage-28 hind limb buds; Lane 5 - RNA from stage-20 limb buds.

Fig. 3

(A). Northern blot analysis of embryonic transcripts showing expression of Ghox 2.1 and 2.2. Lanes 1–3 each contain 2 μg of poly (A)+ RNA and were hybridized with probe A (Ghox 2.1). Lanes 4–6 each contain 4 μg of poly (A)+ RNA and were hybridized with probe B (Ghox 2.2). Chick ribosomal RNA (1·9 and 4·6 kb) was used as a size marker. Lanes 1,4 - RNA from trunks of stage-20 embryos. Lanes 2,5 - RNA from trunks of stage-22 embryos. Lanes 3,6 - RNA from trunks of stage-24 embryos. (B-E) RNase protection assays of total RNA from chick embryos hybridized with antisense probes of Ghox 2.1 and 2.2. The bands represent full-length protection. As a control we also hybridized with sense probes D(+) and E(+), and see no protection (data not shown). (B) RNA hybridized with probe D(−) of Ghox 2.1: Lane 1 - RNA from trunks of stage-30 embryos; Lane 2 - RNA from heads of stage-20 embryos; Lane 3 - RNA from heads of stage-22 embryos; Lane 4 - RNA from heads of stage-24 embryos; Lane 5 - tRNA control; Lane 6 - probe only. Note that Lane 2 has been underloaded. Other experiments show that there is no reduction in the level of expression at this stage. (C) RNA hybridized with probe E(−) of Ghox 2.2: Lane 1 - RNA from trunks of stage-30 embryos; Lane 2 - RNA from heads of stage-20 embryos; Lane 3 - RNA from heads of stage-22 embryos; Lane 4 - RNA from heads of stage-24 embryos; Lane 5 - tRNA control; Lane 6 - probe only. (D) RNA hybridized with probe D(−) of Ghox 2.1: Lane 1 - probe; Lane 2 - tRNA control; Lane 3 - RNA from stage-30 limb buds; Lane 4 - RNA from stage-28 hind limb buds; Lane 5 - RNA from stage-22 limb buds. (E) RNA hybridized with probe E(−) of Ghox 2.2: Lane 1 - probe; Lane 2 - tRNA control; Lane 3 - RNA from stage-30 limb buds; Lane 4 - RNA from stage-28 hind limb buds; Lane 5 - RNA from stage-20 limb buds.

For Hox 2.1, Krumlauf et al. (1987) have shown that some of the weaker bands seen in Northern blots of developing mouse embryos are caused by cross hybridization with mRNAs derived from closely related homeobox loci. It should be noted that neither probe used in Fig. 3A contains any homeobox sequence, but they derive mostly from the usually less conserved 3’ untranslated region of Ghox 2.1 and 2.2 (Fig. 1). Thus the simplest interpretation of the Northern blots is that the strong bands represent transcripts from the 2.1 and 2.2 gene, and that weak bands could be either RNA precursors, transcripts initiated at multiple start sites or alternatively spliced RNAs.

That Ghox 2.1 and 2.2 are both expressed in chick embryos is further substantiated by RNase protection experiments (Fig. 3B-E). To investigate whether expression of the genes varies within the embryo during development, RNA was extracted from embryos dissected into heads, trunks and limbs and hybridized with the appropriate antisense probe.

Ghox 2.1 is expressed at all the stages that we have studied (stages 16–34). The RNase protection data in Fig. 3 demonstrates Ghox 2.1 expression in the developing head (Fig. 3B) and in the limb buds (Fig. 3D). That Ghox 2.1 is also expressed in the trunk region of the embryo was already obvious from the Northern blot of Fig. 3A (lanes 1-3). Extensive analysis of Northern blot data using chick β-actin as a standard and of RNase protection assays show that the highest levels of Ghox 2.1 expression are found in the trunk, with less expression in the limb buds and least in the head. Moreover, we find no striking temporal change in the level of expression in each of these regions.

Ghox 2.2 is also expressed between stages 16–34, in the trunk, (Fig. 3A, lanes 4–6) and in the limbs, as demonstrated by the RNase protection data in Fig. 3E. Analysis of available data shows that there is a higher level of Ghox 2.2 expression in the trunk region but that there is no temporal alteration in the level of expression in these two regions. There is a striking absence of Ghox 2.2 expression from the developing head. Fig. 3C illustrates that there are no protected fragments of full length in head RNA (lanes 2 to 4). By contrast, the positive control RNA from stage-30 trunks is fully protected (lane 1). That Ghox 2.2 transcripts are absent in the early head indicates some degree of tissue- or position-specificity of the activity of this gene.

Spatial localization of Ghox 2.1 message by in situ hybridization

For in situ hybridization we have used 35S-labelled riboprobes C or D (see Fig. IB). Probe C is derived from 3’ flanking sequence whereas probe D also contains part of the homeobox. We found no systematic differences between the two probes, either in Northern blots or in in situ hybridization experiments and hence have used either of the two probes in the analyses summarized below.

Expression in the central nervous system

To illustrate the specificity of our probes, Fig. 4A shows a transverse section through the dorsal hindlimb region of a Hamburger-Hamilton stage-25 embryo. It can easily be seen that the spinal chord is the main structure that hybridizes with antisense probe. When hybridizing an adjacent section with sense RNA, the spinal chord tissue and its environment display a low level of silver grain (Fig. 4B). Hence the control probe does not display any specific binding.

Fig. 4

(A,B) Transverse section illustrating probe specificity and dorsoventral distribution of Ghox 2.1 RNA in the spinal chord region of a Hamburger-Hamilton stage-25 embryo. The section shown in A was hybridized with antisense riboprobe D(−) and photographed under darkfield illumination. Emulsions had been exposed for 14 days. Strong hybridization is seen in the ependymal zone, the inner mantle zone and roof- and floor plates. Moderate hybridization is detected in the dorsal root ganglia. Lateral motor horn tissue does not express Ghox 2.1 RNA. Laterodorsally there is a region of weak hybridization that corresponds to somitic tissue. However, the signal is too low to be certain that it is specific. Fig. 4B shows hybridization of a nearby section with sense riboprobe D(+), using the same time of exposure as for the antisense specimens and also photographed under dark-field illumination. Sense probe hybridization results in low level of hybridization. Areas with higher cell density, such as the spinal chord and dorsal root ganglia, pick up some sense probe. By contrast, grain density over e.g. the loose mesenchyme underneath the ectoderm is very low. (C) Parasagittal section through a stage-22 embryo at the level of the forelimb autoradiographed for 16 days. Note the strong hybridization of antisense probe D(−) to the spinal chord and a weaker signal over the dorsal root ganglia and sclerotome tissue. The sympathetic nerve chord extends ventrally (circles) but does not hybridize. Legend: c, central canal; dg, dorsal root ganglion; e, ependymal zone; f, floor plate; i, inner mantle zone; me, mesenchyme; o, outer mantle zone; r, roof plate; s, sclerotome; sc, spinal chord; v, ventral horn. Bar: 100 μm (A,B) and 200 μm (C). message. This distribution pattern is consistent with the Northern blot and RNase protection data that show the highest level of transcript in the trunk and a considerably lower level in the head.

Fig. 4

(A,B) Transverse section illustrating probe specificity and dorsoventral distribution of Ghox 2.1 RNA in the spinal chord region of a Hamburger-Hamilton stage-25 embryo. The section shown in A was hybridized with antisense riboprobe D(−) and photographed under darkfield illumination. Emulsions had been exposed for 14 days. Strong hybridization is seen in the ependymal zone, the inner mantle zone and roof- and floor plates. Moderate hybridization is detected in the dorsal root ganglia. Lateral motor horn tissue does not express Ghox 2.1 RNA. Laterodorsally there is a region of weak hybridization that corresponds to somitic tissue. However, the signal is too low to be certain that it is specific. Fig. 4B shows hybridization of a nearby section with sense riboprobe D(+), using the same time of exposure as for the antisense specimens and also photographed under dark-field illumination. Sense probe hybridization results in low level of hybridization. Areas with higher cell density, such as the spinal chord and dorsal root ganglia, pick up some sense probe. By contrast, grain density over e.g. the loose mesenchyme underneath the ectoderm is very low. (C) Parasagittal section through a stage-22 embryo at the level of the forelimb autoradiographed for 16 days. Note the strong hybridization of antisense probe D(−) to the spinal chord and a weaker signal over the dorsal root ganglia and sclerotome tissue. The sympathetic nerve chord extends ventrally (circles) but does not hybridize. Legend: c, central canal; dg, dorsal root ganglion; e, ependymal zone; f, floor plate; i, inner mantle zone; me, mesenchyme; o, outer mantle zone; r, roof plate; s, sclerotome; sc, spinal chord; v, ventral horn. Bar: 100 μm (A,B) and 200 μm (C). message. This distribution pattern is consistent with the Northern blot and RNase protection data that show the highest level of transcript in the trunk and a considerably lower level in the head.

An important point to be made from Fig. 4 is that the spinal chord in the 5-day chick embryo displays a nonuniform distribution of Ghox 2.1 transcripts. Particularly strong hybridization is seen in the ependymal zone and the inner mantle zone. Also, roof- and floorplate both hybridize with antisense probe. By contrast, there is little signal over the neuroblasts of the outer mantle zone and in the ventral horn (motor column). The reduced transcript level in the ventral horn reflects the notion that at this stage the motor portion of the cord is further developed than the sensory region (Hamburger, 1948).

Fig. 4A indicates moderate hybridization over the dorsal root ganglia. This hybridization is more visible in a parasagittal section of a younger embryo (stage 22, Fig. 4C). The grain density above the ganglia is significantly higher than the background levels seen, for example, in ventral mesenchymal tissue, but it is less pronounced than in the spinal chord. In the same section, we detect hybridization of antisense probe to cells in between the dorsal root ganglia that represent sclerotome tissue. It could be argued that the hybridization in the ganglia and the sclerotome result from the higher cell density in these tissues. This is probably not the case since the sense control slides exhibit a uniform low level of hybridization across the entire field (data not shown).

The uneven distribution of Ghox 2.1 transcripts in the spinal chord of a stage-25 embryo contrasts with that in a neural tube cross section of a stage-17 embryo. Fig. 5D and E show a phase-contrast and a dark-field micrograph of a transverse section through the tail bud region. At this stage and in this most caudal region, the neural tube is undifferentiated and is partially attached to the surface ectoderm. Ghox 2.1 transcripts are uniformly present throughout the entire neural tube cross-section. It therefore appears that maturation of the neural tube into spinal chord is paralleled by a spatial restriction of Ghox 2.1 expression.

Fig. 5

Collection of sagittal sections of a Hamburger-Hamilton stage 17 embryo hybridized with antisense riboprobe D (−). (A,B,C) Strong hybridization is seen in the neural tube and in the hindbrain. Note, sections shown in B and C are slightly more mesial than that depicted in A. Mesonephric duct and the nephrogenous tissue of the intermediate cell mass also express Ghox 2.1 RNA. The lung primordia have just emerged at this stage and as can be seen in C, display moderate hybridization with antisense probe. D is a phase-contrast and E a dark-field micrograph of a sagittal section through the tail bud region, lateral to the section shown in A. D and E illustrate the uniform distribution of Ghox 2.1 RNA in the early neural tube that is still attached to the ectoderm. Exposure time was 6 weeks. Legend: a, atrium; cl, coelom; d, dorsal aorta; ec, ectoderm; 1, lung bud; m, mesonephric duct; my, myelencephalon; n, nephrogenous tissue; no, notochord; nt, neural tube; p, presomitic mesoderm; ve, ventricle of the heart; v4, fourth ventricle. Bar: 500 μm (A,B,C); 50 μm (D,E).

Fig. 5

Collection of sagittal sections of a Hamburger-Hamilton stage 17 embryo hybridized with antisense riboprobe D (−). (A,B,C) Strong hybridization is seen in the neural tube and in the hindbrain. Note, sections shown in B and C are slightly more mesial than that depicted in A. Mesonephric duct and the nephrogenous tissue of the intermediate cell mass also express Ghox 2.1 RNA. The lung primordia have just emerged at this stage and as can be seen in C, display moderate hybridization with antisense probe. D is a phase-contrast and E a dark-field micrograph of a sagittal section through the tail bud region, lateral to the section shown in A. D and E illustrate the uniform distribution of Ghox 2.1 RNA in the early neural tube that is still attached to the ectoderm. Exposure time was 6 weeks. Legend: a, atrium; cl, coelom; d, dorsal aorta; ec, ectoderm; 1, lung bud; m, mesonephric duct; my, myelencephalon; n, nephrogenous tissue; no, notochord; nt, neural tube; p, presomitic mesoderm; ve, ventricle of the heart; v4, fourth ventricle. Bar: 500 μm (A,B,C); 50 μm (D,E).

Fig. 5A to C represent a composite of several sagittal sections of a stage-17 embryo and illustrates the overall distribution of Ghox 2.1 transcripts at that time. Most notably, the entire neural tube (Fig. 5A and B) and the myelencephalon (Fig. 5A and C) strongly hybridize with antisense probe. The anterior border of hybridization is at the level of the fourth ventricle. In the mid- and forebrain we were unable to detect Ghox 2.1

Expression in the mesonephros

The hybridization signal seen along the midline of the sagittal section in Fig. 5A and B coincides with the mesonephros. Mesonephros consists of the mesonephric duct (also known as Wolffian duct) and the mesonephric tubules that are induced by the duct in the intermediate cell mass (see Saxen, 1987). The nephrogenous tissue of the intermediate cell mass extends dorsally to the coelomic cavity (Fig. 5A & B) and hybridizes with Ghox 2.1 antisense RNA. A short stretch of duct is visible in the more anterior regions at the level of the presumptive wing bud (Fig. 5A). A parasagittal section (Fig. 6A & B) through the mesonephric system of a stage-22 embryo reveals an intricate array of mesonephric tubules, glomeruli and blood vessels. Strong hybridization is seen with RNA in the tubular epithelium. By contrast neither glomeruli nor blood vessels express Ghox 2.1. Inspection of the hybridization pattern in Fig. 6 shows that grain density above some of the tubules is strikingly unequal. One tubule marked by arrow 1 hybridizes only partially and others located near the embryo’s surface do not hybridize above background levels (e.g. arrow 2). This differential hybridization pattern may reflect the well-known segmented nature of the mesonephros (see Discussion).

Fig. 6

Expression of Ghox 2.1 in mesonephros at Hamburger-Hamilton stage 22. Bright-field (A) and darkfield (B) illumination micrograph of the same field demonstrates strong hybridization of antisense probe D(−) to the majority of mesonephric tubules. By contrast, Ghox 2.1 RNA is absent in blood vessels and glomeruli. Note the nonuniform hybridization (arrow 1) or lack of hybridization (arrow 2) over some of the tubuli depicted. The ‘upper’ part of the tubule marked by arrow 1 expresses little Ghox 2.1 RNA, whereas the ‘lower’ portion has high grain density, i.e. contains more Ghox 2.1 message. Exposure time was 16 days. Legend: bv, blood vessels; cl, coelom; g, glomeruli; t, tubuli. Bar: 200 μm.

Fig. 6

Expression of Ghox 2.1 in mesonephros at Hamburger-Hamilton stage 22. Bright-field (A) and darkfield (B) illumination micrograph of the same field demonstrates strong hybridization of antisense probe D(−) to the majority of mesonephric tubules. By contrast, Ghox 2.1 RNA is absent in blood vessels and glomeruli. Note the nonuniform hybridization (arrow 1) or lack of hybridization (arrow 2) over some of the tubuli depicted. The ‘upper’ part of the tubule marked by arrow 1 expresses little Ghox 2.1 RNA, whereas the ‘lower’ portion has high grain density, i.e. contains more Ghox 2.1 message. Exposure time was 16 days. Legend: bv, blood vessels; cl, coelom; g, glomeruli; t, tubuli. Bar: 200 μm.

Expression in the limb bud

The early limb bud, up to about stage 22, consists of apparently undifferentiated mesenchyme enclosed in an ectodermal hull. Initially, the only obviously differentiated cell type present in the mesenchyme are the endothelial cells forming the vasculature. As the limb grows distally, the mesenchyme terminally differentiates into a well-defined pattern of histologically distinct tissues such as cartilage, bones, nerves, dermis and muscles.

Northern blots and RNase protection assays showed that Ghox 2.1 and Ghox 2.2 are expressed in the developing limb (Fig. 3D & E). To investigate the spatial distribution of Ghox 2.1 transcript, we prepared sections perpendicular to the proximodistal limb axis. Serial reconstructions show a striking localization of Ghox 2.1 transcripts in the proximoanterior portion of the bud (Fig. 7B). This patch of hybridization is not seen on a nearby section which was probed with sense riboprobe (Fig. 7C). It is important to note that hybridization does not correspond to the cytodifferentiation pattern. As is just visible in the bright-field micrograph of Fig. 7A, cells are beginning to differentiate in the centre of the limb bud, in a zone that will later be the humerus. Clearly, the patch of hybridization does not relate to this chondrogenic region.

Fig. 7

Distribution of Ghox 2.1 in the developing forelimb bud of a stage-22 embryo. Dorsal is to the top of the sections, ventral to the bottom. (A) A transverse section through the proximal limb bud photographed under bright-field illumination indicates that at this stage the mesenchyme is essentially uniform, with only a hint of cellular condensation in the centre. The dark-field photograph shown in B reveals a patch of Ghox 2.1 RNA expression at the anterior limb bud margin (right) that is revealed upon hybridization with antisense probe D(−). The rest of the mesenchymal tissue shows a low level of signal that is slightly above background as defined by an adjacent sense section (Fig. 7C). Hence, we are not certain whether these grains reflect specific hybridization or background. (C) Nearby proximal limb bud section hybridized with sense riboprobe D(+). Exposure time: 16 days. Legend: me, mesenchyme; ec, ectoderm; bv, blood vessel. Bar: 200 μm.

Fig. 7

Distribution of Ghox 2.1 in the developing forelimb bud of a stage-22 embryo. Dorsal is to the top of the sections, ventral to the bottom. (A) A transverse section through the proximal limb bud photographed under bright-field illumination indicates that at this stage the mesenchyme is essentially uniform, with only a hint of cellular condensation in the centre. The dark-field photograph shown in B reveals a patch of Ghox 2.1 RNA expression at the anterior limb bud margin (right) that is revealed upon hybridization with antisense probe D(−). The rest of the mesenchymal tissue shows a low level of signal that is slightly above background as defined by an adjacent sense section (Fig. 7C). Hence, we are not certain whether these grains reflect specific hybridization or background. (C) Nearby proximal limb bud section hybridized with sense riboprobe D(+). Exposure time: 16 days. Legend: me, mesenchyme; ec, ectoderm; bv, blood vessel. Bar: 200 μm.

In the chick genome, we have identified two closely situated homeobox-containing genes that code for proteins whose homeodomains are homologous to mouse Hox 2.1 and 2.2 (Krumlauf et al. 1987; Hart et al. 1987). Because of this homology, we tentatively refer to the chicken homeoboxes described in this study as Ghox 2.1 (Gallus Aomeobox) and Ghox 2.2. Northern analyses and RNase protection assays of embryonic RNA collected during late somitogenesis and early organo-genesis show that both Ghox 2.1 and Ghox 2.2 are expressed, but that there is no marked temporal change of the RNA levels. In situ hybridization with 35S-labelled antisense riboprobe to Ghox 2.1 provides further details about the expression pattern. Accordingly, Ghox 2.1 RNA is present at high levels in the spinal chord, myelencephalon and mesonephros. Lower, but specific hybridization is seen in the dorsal root ganglia, the lung rudiment, and the early limb bud. Mid- and forebrain, the heart, presomitic mesenchyme and notochord do not exhibit detectable levels of Ghox 2.1 mRNA.

In what follows, we will discuss the salient features of Ghox 2.1 expression. Where appropriate, we will compare the pattern of Ghox 2.1 expression in the chick with that of Hox 2.1 in the mouse embryo, which has recently been described by Holland & Hogan (1988a).

Nervous system

Along the rostrocaudal axis of the CNS, Hox 2.1 and Ghox 2.1 RNA show a very similar spatial distribution. In both species, the expression extends anteriorly as far as the fourth ventricle, ending dorsally at the thin roof plate of the fourth ventricle (Fig. 5C). Ventrally, the expression domain extends slightly more anterior. In both chicken and mouse, the highest level of expression in the CNS is seen in the caudal part of the myelencephalon. Ghox 2.1 expression extends posteriorly, throughout the neural tube, and there is no clearly identifiable caudal expression boundary. This can readily be seen in Fig. 5D and E, which depict the most caudal portion of the neural tube.

Spinal chord cross-sections show that chick and mouse both exhibit higher levels of expression in the sensory domain than in the ventrally situated motor region. However, a comparison of our Fig. 5 with fig. 4 in Holland & Hogan (1988a) suggests also a slight discrepancy in the hybridization pattern of the spinal cord. In the mouse, Hox 2.1 is not expressed in the ependymal zone, but in the chick, Ghox 2.1 transcripts are very abundant in this region. Holland and Hogan have pointed out that Hox 2.1 expression along the anteroposterior axis of the CNS becomes more constricted as development proceeds. This ‘rule’ also applies to the dorso ventral axis, as is demonstrated in a comparison of Fig. 4A and 5D and E. Early neural tube uniformly expresses Ghox 2.1, whereas the more developed spinal chord lacks Ghox 2.1 RNA in the ventral regions. Chick and mouse both show hybridization to the neural-crest-derived dorsal root ganglia (Le Douarin, 1982). In both species, the level of hybridization in these ganglia is less than in the spinal chord, but is clearly above background level. Neither species expresses 2.1 in the cells of the sympathetic trunk. In conclusion, the expression pattern of Ghox 2.1 agrees with the overall picture described for Hox 2.1 in the mouse.

Our preliminary in situ data indicate that Ghox 2.2 is also expressed in the CNS. RNase protection experiments (Fig. 3B & C) clearly show that Ghox 2.1 but not Ghox 2.2 is expressed in the head. This implies that the expression boundary of Ghox 2.1 lies more anteriorly than that of Ghox 2.2. Such differential expression boundaries of neighbouring homeoboxes have been noted before in the mouse, and have been interpreted as evidence for position-specific rather than tissue-type-specific gene expression (Dressier & Gruss, 1988).

Mesodermally derived structures

A striking similarity between Ghox 2.1 and Hox 2.1 is the strong hybridization in the mesonephros, a property shared with a number of other homeoboxes such as Hox 1.3 (Dony & Gruss, 1987), Hox 1.5 (Gaunt, 1988), Hox 3.1 (Awgulewitsch et al. 1986) and Hox 6.1 (Sharpe et al. 1988). Despite the fact that all tubules consist of epithelial cells, it is clear that they do not all express Ghox 2.1 (Fig. 6). This probably mirrors the segmented nature of the mesonephric tubule into regions with different secretory and metabolic functions (see e.g. Ganong, 1983). That chick mesonephros is segmented has been shown histologically by Stampfli (1951) and by Croisille and colleagues with a kidney-specific antibody (Croisille et al. 1971; Croisille, 1971, 1976). Using this antibody, Croisille and collaborators have shown the existence of proximal, intermediate and distal tubular segments. Interestingly, this antibody reacts with an antigen whose expression exhibits a sharp segmental boundary (Croisille et al. 1971), a pattern also seen for Ghox 2.1 (Fig. 6A & B). Whether this implies a role of Ghox 2.1 in defining the extent of tubular segments remains to be seen.

Northern blots and RNase protection assays provide strong evidence that Ghox 2.1 and Ghox 2.2 are expressed throughout early chick limb development. In situ hybridization experiments reveal a patch of Ghox 2.1 RNA at the anterior margin of the limb, in a proximal region. To date, a number of areas have been defined in the limb bud mesenchyme which have distinct properties. Examples of functional importance are the posteriorly located zone of polarizing activity (Saunders & Gasseling, 1968) and the progress zone at the distal end of the bud (Summerbell et al. 1973). A recently discovered antibody reveals a yet unknown antigen confined to an anterior, ventral zone in the early limb bud (Ohsugi & Ide, 1986). Finally, limbs contain regions of programmed cell death, known as the posterior and anterior necrotic zones (Saunders et al. 1962; Hinchliffe & Ede, 1973). Anteriorly, cell death begins at stage 21 (3-75 days) at the junction of bud and body wall. During the subsequent stages the anterior necrotic zone broadens and extends more distally (Hinchliffe & Ede, 1973). Preliminary data indicate that the patch of Ghox 2.1 expression diminishes in size as the anterior necrotic zone expands. It should be stressed, however, that the patch of Ghox 2.1 is unlikely to act as a global signal for inducing cell death in the limb, because no similar zone of expression is seen along the posterior margin that also undergoes programmed cell death.

Northern analysis shows that human homeoboxes HHO.cl, HHO.c8 and HHO.C13 are expressed in limbs of 6- to 8-week-old embryos (Mavilio et al. 1986; Simione et al. 1987). This period corresponds to chick stages 27 to 30. Hox 2.1 transcripts, however, have not been detected in 12-5 p.c. mouse limb buds by RNase protection assays (Jackson et al. 1985). This stage of development corresponds approximately to Hamburger-Hamilton stage 29/30 of chick. Although Ghox 2.1 transcripts are still present in chick limb buds at this stage, limb patterning has long been completed. It will be interesting to see whether mouse limb buds equivalent to chick stage 21/22 express Hox 2.1, and whether expression is also more pronounced at the anterior margin.

The core of this study provides an analysis of the temporal and spatial expression pattern of the chicken homeobox-containing gene Ghox 2.1. A detailed analysis of Ghox 2.2 expression is underway in our laboratory. We are studying chicken homeobox-containing genes because avian embryos are readily accessible to a series of well-defined experimental manipulations that profoundly alter the body plan. Our goal is now to exploit this useful property to find out exactly how homeobox-containing genes are involved in vertebrate development.

We wish to thank Drs Rolf Zeller, Rick Woychik and Heinz Busse for their invaluable contributions throughout this study and Dr Philip Leder for his generous support at the start of this project. This study was supported by a grant from American Cancer Society (NP 630). Sarah Wedden was supported by a NATO postdoctoral research fellowship and Kevin Pang in part by a NIH predoctoral training grant.

Awgulewitsch
,
A.
,
Utset
,
M. F.
,
Hart
,
C. P.
,
McGinnis
,
W.
&
Ruddle
,
F. H.
(
1986
).
Spatial restriction of expression of a mouse homoeo box locus within the central nervous system
.
Nature, Lond
.
320
,
328
336
.
Beeman
,
R.
(
1987
).
A homeotic gene cluster in the red flour beetle
.
Nature, Lond
.
327
,
247
s249
.
Boncinelu
,
E.
,
Simeone
,
A.
,
La Volpe
,
A.
,
Faiella
,
A.
,
Fidanza
,
V.
,
Acampora
,
D.
&
Scott
,
L.
(
1985
).
Human cDNA clones containing homeobox sequences
.
Cold Spring Harb. Symp
.
50
,
301
306
.
Breier
,
G.
,
Dressler
,
G. R.
&
Gruss
,
P.
(
1988
).
Primary structure and developmental pattern of Hox 3.1, a member of the murine Hox 3 homeobox gene cluster
.
EM BO J
.
7
,
1329
1336
.
Carrasco
,
A. E.
&
Malacinski
,
G. M.
(
1987
).
Localization of Xenopus homeobox gene transcripts during embryogenesis and in the adult nervous system
.
Devi Biol
.
121
,
69
81
.
Carrasco
,
A. E.
,
McGinnis
,
W.
,
Gehring
,
W. J.
&
De Robertis
,
E. M.
(
1984
).
Cloning of an X. laevis gene expressed during early embryogenesis that codes for a peptide region homologous to Drosophila homeotic genes
.
Cell
37
,
409
414
.
Church
,
G. M.
&
Gilbert
,
W.
(
1984
).
Genomic sequencing
.
Proc, natn. Acad. Sci. U.S.A
.
81
,
1991
1995
.
Colberg-Poley
,
A. M.
,
Voss
,
S. D.
,
Chowdhury
,
K.
,
Stewart
,
C.
,
Wagner
,
E. F.
&
Gruss
,
P.
(
1985
).
Clustered horneo boxes are differentially expressed during munne development
.
Cell
43
,
39
45
.
Croisille
,
Y.
(
1971
).
Appearance and disappearance of organspecific components during kidney tubulogenesis in chick and quail embryos
.
In Protides of the Biological Fluids
(ed.
H.
Peeters
), pp.
79
85
.
Oxford
:
Pergamon Press
.
Croisille
,
Y.
(
1976
).
On some recent contributions to the study of kidney tubulogenesis in mammals and birds
.
In Tests of Teratogenicity In Vitro
(ed.
J. D.
Ebert
&
M.
Marois
), pp.
149
170
.
Amsterdam
:
North-HolLond
.
Croisille
,
Y.
,
Gumpel-Pinot
,
M.
&
Martin
,
C.
(
1971
).
Sur l’organogenese du mésonéphros chez les oiseaux. Etude immunohistologique du tubule urinaire chez l’embryon de poulet
.
C. r. hebd. Séanc. Acad. Sci. Paris Ser. D
272
,
629
631
.
Davis
,
C. A.
,
Noble-Topham
,
S. E.
,
Rossant
,
J.
&
Joyner
,
A. L.
(
1988
).
Expression of the horneo box-containing gene En-2 delineates a specific region of the developing mouse brain
.
Genes & Development
2
,
261
371
.
Davis
,
L. G.
,
Dibner
,
M. D.
&
Battey
,
J. F.
(
1986
).
Basic Methods in Molecular Biology
.
New York
:
Elsevier
.
Dolecki
,
G. J.
,
Wannakrairoj
,
S.
,
Lum
,
R.
,
Wang
,
G.
,
Riley
,
H. D.
,
Carlos
,
R.
,
Wang
,
A.
&
Humphreys
,
T.
(
1986
).
Stagespecific expression of a horneo box-containing gene in the nonsegmented sea urchin embryo
.
EMBO J
.
5
,
925
930
.
Dony
,
C.
&
Gruss
,
P.
(
1987
).
Specific expression of the Hox-1.3 homeobox gene in murine embryonic structures originating from or induced by the mesoderm
.
EMBO J
.
6
,
2965
2975
.
Dressler
,
G. R.
&
Gruss
,
P.
(
1988
).
Do multigene families regulate vertebrate development?
Trends in Genetics
4
,
214
219
.
Falzon
,
M.
&
Chung
,
S. Y.
(
1988
).
The expression of rat homeobox-containing genes is developmentally regulated and tissue specific
.
Development
103
,
601
610
.
Feduccia
,
A.
(
1980
).
The Age of Birds
.
Cambridge
:
Harvard University Press
.
Feinberg
,
A.
&
Vogelstein
,
B.
(
1983
).
A technique for radiolabelling DNA restriction endonuclease fragments to high specific activity
.
Analyt. Biochem
.
132
,
6
13
.
Feinberg
,
A.
&
Vogelstein
,
B.
(
1984
).
Addendum: A technique for radiolabelling DNA restriction endonuclease fragments to high specific activity
.
Analyt. Biochem
.
137
,
266
267
.
Feinberg
,
A. A.
,
Utset
,
M. E.
,
Bogarad
,
L. D.
,
Hart
,
C. P.
,
Awgulewitsch
,
A.
,
Ferguson-Smith
,
A.
,
Fainsod
,
A.
,
Rabin
,
M.
&
Ruddle
,
F. H.
(
1987
).
Horneo box genes in murine development
.
Curr. Topics in devl Biol
.
23
,
233
256
.
Ganong
,
W. F.
(
1983
).
Medical Physiology
.
Los Altos
:
Lange
.
Gaunt
,
S. J.
(
1988
).
Mouse homeobox gene transcripts occupy different but overlapping domains in embryonic germ layers and organs: a comparison of Hox-3.I and Hox-1.5
.
Development
103
,
135
144
.
Gehring
,
W. J.
(
1985
).
The horneo box: a key to the understanding of development?
Cell
40
,
3
5
.
Gehring
,
W. J.
(
1987
).
Horneo boxes in the study of development
.
Science
236
,
1245
1252
.
Gray
,
G. E.
,
Glover
,
J. C.
,
Majors
,
J.
&
Sanes
,
J. R.
(
1988
).
Rectal arrangement of clonally related cells in the chicken optic tectum: lineage analysis with a recombinant retrovirus
.
Proc, natn. Acad. Sci. U.S.A. (In press)
.
Hamburger
,
V.
(
1948
).
The mitotic patterns on the spinal chord of the chick embryo and their relation to histogenetic processes
.
J. comp. Neurol
.
88
,
221
283
.
Hamburger
,
V.
&
Hamilton
,
H. L.
(
1951
).
A series of normal stages in the development of the chick embryo
.
J. Morph
.
88
,
49
92
.
Hara
,
K.
(
1978
).
‘Spemann’s Organizer’ in Birds
.
In Organizer - A Milestone of a Half-Century from Spemann
(ed.
O.
Nakamura
&
S.
Toivonen
), pp.
221
265
.
Amsterdam
:
Elsevier/North-HolLond
.
Hart
,
C. P.
,
Fainsod
,
A.
&
Ruddle
,
F. H.
(
1987
).
Sequence analysis of the murine Hox-2.2, -2.3, and -2.4 horneo boxes: evolutionary and structural comparisons
.
Genomics
1
,
182
195
.
Harvey
,
R. P.
&
Melton
,
D. A.
(
1988
).
Microinjection of synthetic Xhox-1A homeobox mRNA disrupts somite formation in developing Xenopus embryos
.
Cell
53
,
687
697
.
Hinchliffe
,
J. R.
&
Ede
,
D. A.
(
1973
).
Cell death and the development of limb form and skeletal pattern in normal and wingless (tvs) chick embryos
.
J. Embryol. exp. Morph
.
30
,
753
772
.
Holland
,
P. W. H.
&
Hogan
,
B. L. M.
(
1988a
).
Spatially restricted pattern of expression of the homeobox-containing gene Hox 2.1 during mouse embryogenesis
.
Development
102
,
159
174
.
Holland
,
P. W. H.
&
Hogan
,
B. L. M.
(
1988b
).
Expression of horneo box genes during mouse development; a review
.
Genes & Development
2
,
773
782
.
Jackson
,
I. J.
,
Schofield
,
P.
&
Hogan
,
B.
(
1985
).
A mouse horneo box gene is expressed during embryogenesis and in adult kidney
.
Nature, Lond
.
317
,
745
748
.
Krieg
,
P.
&
Melton
,
D.
(
1988
).
In vitro RNA synthesis with SP6 polymerase
.
Meth. Enzymol. (In press)
.
Krumlauf
,
R.
,
Hollond
,
P. W. H.
,
McVey
,
J. H.
&
Hogan
,
B. L. M.
(
1987
).
Developmental and spatial patterns of expression of the mouse horneo box gene, Hox 2.1
.
Development
99
,
603
617
.
Le Douarin
,
N.
(
1973
).
A biological cell labelling technique and its use in experimental embryology
.
Devi Biol
.
30
,
217
222
.
Le Douarin
,
N.
(
1982
).
The Neural Crest
.
Cambridge
:
Cambridge University Press
.
Levine
,
M.
,
Rubin
,
G.
&
Than
,
R.
(
1984
).
Human DNA sequences homologous to a protein coding region conserved between homeotic genes of Drosophila
.
Cell
38
,
667
673
.
Lillie
,
F. R.
(
1952
).
Lillie’s Development of the Chick. Revised by Howard C. Hamilton
.
New York
:
Holt, Rinehart & Winston
.
Malvilio
,
F.
,
Simeone
,
A.
,
Giampaolo
,
A.
,
Failla
,
A.
,
Zappavigna
,
V.
,
Acampora
,
D.
,
Poina
,
G.
,
Russo
,
G.
,
Peschle
,
C.
&
Boncinelu
,
E.
(
1986
).
Differential and stage-related expression in embryonic tissues of a new human homeobox gene
.
Nature, Lond
.
324
,
664
668
.
McGinnis
,
W.
,
Garber
,
R. L.
,
Wirz
,
J.
,
Kuroiwa
,
A.
&
Gehring
,
W. J.
(
1984a
).
A homologous protein coding sequence in Drosophila homeotic genes and its conservation in other metazoans
.
Cell
37
,
403
408
.
McGinnis
,
W.
,
Hart
,
C. P.
,
Gehring
,
W. J.
&
Ruddle
,
F. H.
(
1984c
).
Molecular cloning and chromosome mapping of a mouse DNA sequence homologous to homeotic genes of Drosophila
.
Cell
38
,
675
680
.
McGinnis
,
W.
,
Levine
,
M. S.
,
Hafen
,
E.
,
Kuroiwa
,
A.
&
Gehring
,
W.
(
1984b
).
A conserved DNA sequence in homeotic genes of Drosophila Antennapedia and bithorax complexes
.
Nature, Lond
.
308
,
428
433
.
Odenwald
,
W. F.
,
Taylor
,
C. F.
,
Palmer-Hill
,
F. J.
,
Friedrich
,
V.
,
Tani
,
M.
&
Lazzarini
,
R. A.
(
1987
).
Expression of a horneo domain protein in noncontact-inhibited culture cells and postmitotic neurons
.
Genes & Development
1
,
482
496
.
Ohsugi
,
K.
&
Ide
,
H.
(
1986
).
Position specific binding of a monoclonal antibody in chick limb buds
.
Devi Biol
.
117
,
676
679
.
Oliver
,
G.
,
Wright
,
C. V. E.
,
Hardwicke
,
J.
&
De Robertis
,
E. M.
(
1988
).
Differential antero-posterior expression of two proteins encoded by a homeobox gene in Xenopus and mouse embryos
.
EMBO J
.
7
,
3199
3209
.
Patten
,
B. M.
(
1951
).
Early Embryology of the Chick
.
New York
:
McGraw-Hill Book Co. Inc
.
Rigby
,
P. W. J.
,
Dieckmann
,
M.
,
Rhodes
,
C.
&
Berg
,
P.
(
1977
).
Labelling deoxyribonucleic acid to high specific activity in vitro by nick-translation with DNA polymerase I
.
J. molec. Biol
.
113
,
237
251
.
Romanoff
,
A. I.
(
1960
).
The Avian Embryo
.
Structural and Functional Development
.
New York
:
Macmillan Co
.
Rugh
,
R.
(
1977
).
A Guide to Vertebrate Development
.
Minneapolis
:
Burgess Publishing Co
.
Sanger
,
F.
,
Nicklen
,
S.
&
Coulson
,
A. R.
(
1977
).
DNA sequencing with chain terminating inhibitors
.
Proc. natn. Acad. Sci. U.S.A
.
74
,
5463
5467
.
Saunders
,
J. W.
&
Gasseling
,
M. T.
(
1968
).
Ectodermal-mesenchymal interactions in the origin of wing symmetry
.
In Epithelial-mesenchymal Interactions
(ed.
R.
Fleischmajer
&
R. E.
Billingham
), pp.
78
97
.
Baltimore
:
Williams and Wilkins
.
Saunders
,
J. W.
,
Gasseling
,
M. T.
&
Saunders
,
L. C.
(
1962
).
Cellular death in morphogenesis of the avian wing
.
Devi Biol
.
5
,
147
178
.
Saxen
,
L.
(
1987
).
Organogenesis of the Kidney
.
Cambridge
:
Cambridge University Press
.
Schneuwly
,
S.
,
Kuroiwa
,
A.
,
Baumgartner
,
P.
&
Gehring
,
W. J.
(
1986
).
Structural organization and sequence of the homeotic gene Antennapedia of Drosophila melanogaster
.
EMBO J
.
5
,
733
739
.
Scott
,
M. P.
&
Weiner
,
A. J.
(
1984
).
Structural relationships among genes that control development: sequence homology between Antennapedia, Ultrabithorax and fuzi tarazu loci of Drosophila
.
Proc. natn. Acad. Sci. U.S.A
.
81
,
4115
4119
.
Sharpe
,
C. R.
,
Fritz
,
A.
,
De Robertis
,
E. M.
&
Gurdon
,
J. B.
(
1987
).
A homeobox-containing marker of posterior neural differentiation shows the importance of predetermination in neural induction
.
Cell
50
,
749
758
.
Sharpe
,
P. T.
,
Miller
,
J. R.
,
Evans
,
E. P.
,
Burtenshaw
,
M. D.
&
Gaunt
,
S. J.
(
1988
).
Isolation and expression of a new mouse homeobox gene
.
Development
102
,
397
407
.
Simeone
,
A.
,
Ma Vino
,
F.
,
Acampora
,
D.
,
Giampaola
,
A.
,
Faiella
,
A.
,
Zappavigna
,
V.
,
D’Esposita
,
M.
,
Russo
,
G.
,
Boncinelli
,
E.
&
Peschle
,
C.
(
1987
).
Two human horneo box genes, cl and c8: structure analysis and expression in embryonic development
.
Proc. natn. Acad. Sci. U.S.A
.
84
,
4914
4918
.
Stampfli
,
H. R.
(
1951
).
Histologische Studien am Wolffschen KOrper (Mesonephros) der VOgel und Uber seinen Umbau zum Nebenhoden und Nebenovar
.
Rev. Suisse Zool
.
57
,
237
316
.
Summerbell
,
D.
,
Lewis
,
J.
&
Wolpert
,
L.
(
1973
).
Positional information in chick limb morphogenesis
.
Nature, Lond
.
244
,
492
496
.
Toneguzzo
,
F.
,
Glynn
,
S.
,
Levi
,
E.
,
Mjolsness
,
S.
&
Hayday
,
A.
(
1988
).
Use of a chemically modified T7 DNA polymerase for manual and automated sequencing of supercoiled DNA
.
BioTechn’iques
6
,
460
469
.
Utset
,
M. F.
,
Awgulewitsch
,
A.
,
Ruddle
,
F. H.
&
Mcginnis
,
W.
(
1987
).
Region-specific expression of two mouse homeobox genes
.
Science
235
,
1379
1382
.
Waddington
,
C. H.
(
1934
).
Experiments on embryonic induction
.
J. exp Biol
.
11
,
211
227
.
Way
,
J. C.
&
Chálete
,
M.
(
1988
).
Mec-3, a homeobox-containing gene that specifies differentiation of the touch receptor neurones in C. elegans
.
Cell
54
,
5
16
.