The Iroquois Homeobox Gene Irx2 Is Not Essential for Normal ...

MOLECULAR AND CELLULAR BIOLOGY, Nov. 2003, p. 8216–8225 0270-7306/03/$08.00⫹0 DOI: 10.1128/MCB.23.22.8216–8225.2003 Copyright © 2003, American Society for Microbiology. All Rights Reserved.

Vol. 23, No. 22

The Iroquois Homeobox Gene Irx2 Is Not Essential for Normal Development of the Heart and Midbrain-Hindbrain Boundary in Mice Me´lanie Lebel,1,2 Pooja Agarwal,1,2,3 Chi Wa Cheng,1,4 M. Golam Kabir,5,6 Toby Y. Chan,5,6 Vijitha Thanabalasingham,1 Xiaoyun Zhang,1 Dana R. Cohen,1,2 Mansoor Husain,5,6 Shuk Han Cheng,4 Benoit G. Bruneau,1,2,3,5* and Chi-Chung Hui1,2* Program in Developmental Biology1 and Program in Cardiovascular Research,3 The Hospital for Sick Children, and Department of Molecular and Medical Genetics, University of Toronto,2 Toronto, Ontario M5G 1X8, and The Heart and Stroke/Richard Lewar Centre for Cardiovascular Research at the University of Toronto5 and Division of Cellular and Molecular Biology, Toronto General Hospital Research Institute, Toronto General Hospital, and Department of Medicine, University of Toronto,6 Toronto, Ontario M5G 2C4, Canada, and Department of Biology and Chemistry, City University of Hong Kong, Kowloon, Hong Kong, China4 Received 11 April 2003/Returned for modification 12 June 2003/Accepted 14 August 2003

The Iroquois homeobox (Irx) genes have been implicated in the specification and patterning of several organs in Drosophila and several vertebrate species. Misexpression studies of chick, Xenopus, and zebra fish embryos have demonstrated that Irx genes are involved in the specification of the midbrain-hindbrain boundary. All six murine Irx genes are expressed in the developing heart, suggesting that they might possess distinct functions during heart development, and a role for Irx4 in normal heart development has been recently demonstrated by gene-targeting experiments. Here we describe the generation and phenotypic analysis of an Irx2-deficient mouse strain. By targeted insertion of a lacZ reporter gene into the Irx2 locus, we show that lacZ expression reproduces most of the endogenous Irx2 expression pattern. Despite the dynamic expression of Irx2 in the developing heart, nervous system, and other organs, Irx2-deficient mice are viable, are fertile, and appear to be normal. Although chick Irx2 has been implicated in the development of the midbrain-hindbrain region, we show that Irx2-deficient mice develop a normal midbrain-hindbrain boundary. Furthermore, Irx2-deficient mice have normal cardiac morphology and function. Functional compensation by other Irx genes might account for the absence of a phenotype in Irx2-deficient mice. Further studies of mutant mice of other Irx genes as well as compound mutant mice will be necessary to uncover the functional roles of these evolutionarily conserved transcriptional regulators in development and disease. 16). After the neural plate is specified, the Irx genes are also involved in the anteroposterior and dorsoventral patterning of the neural tube. In chicken embryos, Irx3 is implicated in determining the positioning of zona limitans intrathalamica at the border between the diencephalon and telencephalon (21) as well as in the dorsoventral patterning of the spinal cord (5). Gene knockdown studies have shown that Irx genes are required for the formation of the midbrain-hindbrain boundary (MHB) in Xenopus (Xiro1) and zebra fish (ziro1 and ziro7) (14, 19). Furthermore, recent studies have illustrated a critical role for Irx2 in chick MHB formation; misexpression of Irx2 promotes cerebellum development in an Fgf8-dependent manner (T. Ogura, personal communications; reviewed in reference 15). The Irx genes have also been implicated in the patterning and specification of the developing heart. In chicken embryos, misexpression of Irx4 affects chamber-specific gene expression (2). All six Irx genes display highly specific expression patterns in the developing mouse heart (6, 7, 9, 24), and inactivation of Irx4 in mice results in aberrant ventricular gene expression and adult-onset cardiomyopathy (7). Despite numerous studies suggesting important function for Irx genes in vertebrate development (2, 3, 5, 13, 14, 16, 19–22), only a few genetic studies illustrate the physiological role of the Irx genes. In mice, targeted inactivation of Irx4 resulted in aberrant ventricular gene expression, including reduced ex-

The Drosophila Iroquois complex (Iro-C) was first identified as prepattern genes that control the patterning of notum bristles (see reference 15 for a review). Iro-C contains three highly related homeobox genes (araucan [ara], caupolican [caup], and mirror [mir]), which encode transcriptional regulators that control the expression of the proneural genes, such as achaete and scute. These Iroquois homeobox (Irx) genes also play regulatory roles in the development of other structures, including the head, eye, and wing veins. The structure and organization of vertebrate Irx genes are evolutionarily conserved (15). For example, in mouse and human, six Irx genes (Irx1 to Irx6) are found in two clusters; Iro-A contains Irx1, Irx2, and Irx4, and Iro-B contains Irx3, Irx5, and Irx6 (26). Similar to their Drosophila counterparts, the vertebrate Irx genes also possess regulatory functions in the specification and patterning of early embryos and several organs. Misexpression studies have revealed the roles of three Xenopus Irx genes (Xiro1, -2, and -3) in the regulation of vertebrate proneural genes and the specification of the neural plate (3,

* Corresponding authors. Mailing address: Program in Developmental Biology, The Hospital for Sick Children, 555 University Ave., Toronto, Ontario M5G 1X8, Canada. Phone: (416) 813-5681. Fax: (416) 597-9497. E-mail for Chi-Chung Hui: [email protected]. E-mail for Benoit G. Bruneau: [email protected]. 8216

VOL. 23, 2003

Irx2 GENE IS DISPENSABLE FOR MOUSE DEVELOPMENT

8217

FIG. 1. Disruption of Irx2 by gene targeting. (A) Targeting strategy, showing the partial restriction map of the wild-type (WT) locus, the targeting vector, and the targeted alleles before and after the excision of the neomycin (neo) cassette by Cre-loxP recombination. Homologous recombination replaces the SphI-EcoRI fragment, which contains parts of exon 1 and intron 1, of Irx2 with the lacZ reporter gene (IRES-NLSLacZ-PolyA) and the neo selection gene (PGK-neo). (B) Genotyping of progeny of heterozygous mutants by Southern blot analysis. The sizes of the BamHI fragments detected by the 5⬘ flanking probe in the wild-type and mutant alleles are 15 and 7.1 kb, respectively. (C) PCR amplificationgenerated wild-type (255-bp) and mutant (267-bp) bands. (D) RT-PCR analysis of the mutant transcript. The products of the F2-R2 reaction, which cover the region 3⬘ of the insertion, could be detected in all genotypes. The products of the F1-R1 reaction, which are disrupted by the insertion, could not be detected in homozygous Irx2 mutants. (E) Immunohistochemistry and in situ RNA hybridization of E10.5 transverse embryos sections, showing the specificity of the Irx2 antibody. Western blot analysis of protein extract from E10.5 wild-type (⫹/⫹) and Irx2LacZNeo/LacZNeo (⫺/⫺) embryos is shown. The Irx2 protein band (⬃50 kDa) is missing in the Irx2LacZNeo/LacZNeo embryo extract.

pression of the basic helix-loop-helix transcription factor Hand1/eHand and increased expression of Irx2. The up-regulation of Irx2 expression in Irx4-deficent embryos suggests that functional compensation may partly account for the mild heart phenotype (7). The Fused-toes (Ft) mouse mutation consists of a deletion of six genes, including the entire IroB cluster (27). Ft mutant embryos exhibit severe craniofacial, forebrain, and ventral neural tube defects; malformations of limb and heart; and

random left-right asymmetry (29). Since Irx3, Irx5, and Irx6 are expressed in many of the affected structures, at least some of the Ft mutant phenotypes are caused by the loss of Irx gene function. We present here a mutational analysis of Irx2 function in mice by generation of a loss-of-function allele of Irx2 by homologous recombination in embryonic stem (ES) cells. Although other studies have suggested a role for Irx2 in MHB and heart development, our results indicate that Irx2-deficient

8218

LEBEL ET AL.

MOL. CELL. BIOL.

FIG. 2. Expression of the lacZ reporter gene in heterozygous Irx2 mutant embryos. (A to C) Whole-mount X-Gal staining of E10.5 wild-type (WT) (A), Irx2LacZNeo/⫹ (B), and Irx2LacZ/⫹ (C) embryos. (D) Whole-mount RNA in situ hybridization of Irx2 in E10.5 embryos. (E to BB) X-Gal staining on sections of E15 heterozygous Irx2 mutant embryos. Transverse sections of spinal cord (E, F, and G), heart (H, I, and J), lung (K, L, and M), kidney (N, O, and P), pancreas (Q, R, and S), eye (T, U, and V), whiskers (W, X, and Y), and hair follicle (Z, AA, and BB) were stained, showing lacZ expression in Irx2LacZNeo/⫹ (E, H, K, N, Q, T, W, and Z) and Irx2LacZ/⫹ (F, I, L, O, R, U, X, and AA) embryos and Irx2 RNA expression in wild-type embryos (G, J, M, P, S, V, Y, and BB).

mice develop normally, are viable and fertile, and do not display any obvious defects in neural, heart, and limb development. MATERIALS AND METHODS Generation of Irx2 mutant mice. Genomic clones of Irx2 were isolated from a 129/Sv genomic library by using the Irx2 cDNA as a probe (10). A loss-offunction mutation in Irx2 was generated by replacing part of exon 1 and intron 1 with an internal ribosome entry site (IRES)-lacZ reporter gene and a PGK-neo

cassette by homologous recombination in R1 ES cells. ES cell lines with the desired mutation were identified by Southern analysis, and generation of the mutant mice (Irx2LacZNeo/⫹) was performed by standard procedures (23). To delete the PGK-neo cassette at the targeted locus, Irx2LacZNeo/⫹ mice were crossed with NLS-Cre transgenic mice, which show ubiquitous Cre recombinase expression (a gift from C. Lobe), and Irx2LacZ/⫹ mice were generated. Mice were maintained in a mixed 129/Sv and CD1 background. X-Gal staining. Embryos were fixed in 2.7% formaldehyde–0.02% NP-40 in 1⫻ phosphate-buffered saline. Embryonic day 10.5 (E10.5) embryos were fixed


VOL. 23, 2003

8219

TABLE 1. Hemadynamic analysis of Irx2 mutant micea Genotype (n)

LVSP (mm Hg)

LVEDP (mm Hg)

ABPS (mm Hg)

ABPD (mm Hg)

Peak positive dP/ dT (mm Hg/s)

Peak negative dP/dT (mm Hg/s)

HR (bpm)

Wild type (6) Irx2⫹/⫺ (12) Irx2⫺/⫺ (4)

93.51 ⫾ 5.70 96.66 ⫾ 5.49 98.27 ⫾ 4.21

⫺3.30 ⫾ 0.44 ⫺1.44 ⫾ 2.81 ⫺1.02 ⫾ 1.41

96.35 ⫾ 7.99 103.55 ⫾ 4.27 103.43 ⫾ 4.64

62.53 ⫾ 8.93 64.54 ⫾ 3.85 65.31 ⫾ 4.09

5,635 ⫾ 852 5,551 ⫾ 574 5,808 ⫾ 571

⫺4,683 ⫾ 414 ⫺4,619 ⫾ 592 ⫺4,884 ⫾ 428

190 ⫾ 12 216 ⫾ 48 229 ⫾ 18

a LVSP, left ventricle systolic pressure; LVEDP, left ventricle end-diastolic pressure; ABPS, aortic systolic pressure; ABPD, aortic diastolic pressure; dP/dT, first derivative of the change in left ventricular pressure/time; HR, heart rate; bpm ⫽ heart beats per minute. Analysis of variance shows no significant difference between groups. Values are means ⫾ standard errors.

for 1 h at 4°C, and E15 embryos were fixed for 16 h (overnight) at 4°C. Cryosections (14 ␮m) of E15 embryos and E10.5 whole embryos were subjected to X-Gal (5-bromo-4-chloro-3-indolyl-␤-D-galactopyranoside) staining at 37°C for 48 h, as described previously (11). RT-PCR. RNA was extracted from E11.5 embryos by using Trizol (Invitrogen), and reverse transcription-PCR (RT-PCR) was performed with the SuperScript one-step RT-PCR with platinum Taq kit (Stratagene) according to the manufacturer’s instructions. For real-time RT-PCR, RNA was extracted from E10.5 embryos by using Trizol reagent. One to five micrograms of RNA was reverse transcribed by using the SuperScript first-strand RT-PCR synthesis system (Invitrogen) with oligo(dT)12-18 primers. Quantitative real time PCR was performed with the ABI 7000 sequence detector (Applied Biosystems). Gene expression was quantified by using customized Assays-on-Demand (Applied Biosystems) for Irx3 (Mm00500463 ml), Irx4 (Mm00502170 ml), Irx5 (Mm00502107 ml), and Irx6 (Mm00517712 ml). Specific quantitative assays for Irx1 (forward primer 5⬘-TTA TCCCTATGGTCAGTTTCAATACG-3⬘ and reverse primer 5⬘-CGTTGAGCC AGGCTTTCAG-3⬘) and Irx2 (forward primer 5⬘-ACGCACACCACCGGAAT G-3⬘ and reverse primer 5⬘-ATGGATAGGCCGCACTGC-3⬘) were developed by using Primer Express (Applied Biosystems), following the recommended guidelines based on sequences from GenBank. Gene expression data was normalized by using TaqMan rodent GAPDH (glyceraldehyde-3-phosphate dehydrogenase) (Applied Biosystems). In situ hybridization. Tissues and embryos were fixed in 4% paraformaldehyde in phosphate-buffered saline for 16 h (overnight) at 4°C. Cryosections (14 ␮m) or whole-mount embryos were subjected to in situ hybridization with digoxigenindUTP-labeled riboprobes, as described previously (11). Plasmids for generating the riboprobes were Irx1-5 (6, 10), Fgf8 (28), Otx2 (1), and Pax2 (12). A cDNA probe for Irx6 (nucleotide residues 742 to 1727) was generated by RT-PCR based on published sequence (26). Antibody production. Irx2 antibodies used in the Western blot analysis were affinity-purified rabbit polyclonal antisera raised against the carboxyl terminus (amino acid residues 391 to 458) of Irx2 fused to glutathione S-transferase by standard procedures (17). Immunohistochemistry. Paraffin sections (7 ␮m) were used for immunohistochemistry. The slides were deparaffinized, rehydrated, boiled with EDTA for antigen retrieval, and blocked with goat serum. The sections were then incubated with Irx2 antibodies (1/50) overnight at 4°C and then with an anti-rabbit secondary antibody coupled to biotin. We used the standard Vectastain ABC-AP kit (Vector) and the red substrate kit (Vector) to visualize the signal. Skeletal staining. Alcian blue and alizarin red staining of bone and cartilage was performed on newborn skeletons as previously described (23). Cardiac physiology. Mice (8 to 10 weeks old) were anesthetized by intraperitoneal injection of ketamine (100 mg/kg) and xylazine (10 mg/kg). Echocardiography was performed with a Sonos 5500 (Hewlett-Packard) with a 12-MHz transducer as previously described (25). For in vivo hemodynamic analysis, the right carotid artery and jugular vein were catheterized with a 1.4-F high-fidelity micromanometer catheter (model SPR-671; Millar Instruments, Houston, Tex.)

to obtain heart rate, aortic pressure, left ventricular (LV) systolic pressure, LV end-diastolic pressure, right ventricular (RV) systolic pressure, RV end-diastolic pressure, and the peak positive and negative first derivatives of the LV and RV pressures (⫾dP/dt). Electrocardiography was performed as previously described (25). The QT interval corrected for heart rate (QTc) was defined as QT/ 冑共60,000/HR兲 ⫻ 10, where QT is the time from the beginning of ventricular depolarization to the beginning of ventricular repolarization in milliseconds and HR is the heart rate in beats per minute.

RESULTS Gene targeting of Irx2. To determine the role of Irx2 in mouse development, we inactivated Irx2 by homologous recombination in ES cells. The targeting vector was designed to insert an IRES-lacZ reporter gene and a neomycin resistance gene (neo) into the first exon of Irx2 (Fig. 1A). As a result of the insertion, several stop codons were generated after the initiation codon of Irx2. Thus, this targeted mutation was expected to generate a null allele. After electroporation and drug selection, 192 independent ES clones were analyzed by Southern blot hybridization with a flanking probe. Two targeted mutant clones were obtained, and one of them was used to generate chimeric mice. Germ line transmission of the chimeric mice resulted in the Irx2LacZNeo/⫹ mouse line, and the Irx2LacZ/⫹ mouse line was obtained by deletion of the neo cassette by Cre-loxP recombination (see Materials and Methods). Southern blot (Fig. 1B) and PCR (Fig. 1C) analyses were used to detect the wild-type and mutant alleles of Irx2. Both Irx2LacZNeo/⫹ and Irx2LacZ/⫹ mice appear to be normal and show normal fertility. RT-PCR was performed to analyze the nature of the mutant transcripts (Fig. 1D). The RT-PCR results revealed that while the IRES-lacZ cassette disrupted the transcript at the site of insertion, the Irx2 transcript 3⬘ to the insertion appeared to be normal. In addition, our data indicated that the level of Irx2 transcripts was reduced by the insertion of the IRES-lacZ cassette. As shown below, we also observed a significant reduction of Irx2 transcript in the mutants by real-time RT-PCR and whole-mount RNA in situ hybridization (see Fig. 5 and 6E and F). Furthermore, Western blotting analysis with an Irx2-specific antibody detected no Irx2

TABLE 2. Electrocardiographic analysis of Irx2 mutant micea Pdur (ms)

PR (ms)

QRS (ms)

QTmax (ms)

QT (ms)

QTc (ms)

QRSamp (mV)

Genotype (n)

HR (bpm)

Wild type (9) Irx2⫹/⫺ (3) Irx2⫺/⫺ (9)

293 ⫾ 31 20.221 ⫾ 46 34.11 ⫾ 2.92 12.78 ⫾ 1.99 13.78 ⫾ 2.06 76.11 ⫾ 3.59 52.04 ⫾ 2.52 0.83 ⫾ 0.06 0.15 ⫾ 0.02 0.94 ⫾ 0.12 256 ⫾ 116 24.0 ⫾ 0 34.67 ⫾ 1.66 13.33 ⫾ 1.20 13.33 ⫾ 1.20 83.50 ⫾ 9.38 59.08 ⫾ 1.67 1.12 ⫾ 0.26 0.15 ⫾ 0.02 0.85 ⫾ 0.13 240 ⫾ 81 20.22 ⫾ 1.48 34.66 ⫾ 1.08 10.11 ⫾ 0.69 10.44 ⫾ 1.95 72.00 ⫾ 1.67 52.48 ⫾ 2.79 1.48 ⫾ 0.30 0.19 ⫾ 0.02 1.26 ⫾ 0.11

Pamp (mV)

Tamp (mV)

a HR, heart rate; bpm, beats per minute; Pdur, duration of atrial depolarization; PR, time between the start of atrial depolarization and the start of ventricular deporization; amp, QRS, amplitude of segment. Analysis of variance shows no significant difference between groups. Values are means ⫾ standard errors.

8220

LEBEL ET AL.

FIG. 3. MHB integrity in Irx2LacZ/LacZ embryos. Whole-mount RNA in situ hybridization for markers of the MHB in E10 wild-type (WT) (A, C, and E) and Irx2LacZ/LacZ (B, D, and F) embryos is shown. Expression of Otx2 (A and B), Fgf8 (C and D), and Pax2 (E and F) appears to be normal in Irx2LacZ/LacZ embryos.

protein in E10.5 embryos, suggesting that our gene-targeting strategy had successfully generated an Irx2-deficient mouse strain (Fig. 1E). lacZ reporter gene expression reproduces endogenous Irx2 expression in embryos. X-Gal staining of heterozygous Irx2 mutant embryos showed that lacZ expression mimics the endogenous expression of Irx2 in the central nervous system, lungs, kidneys, and pancreas (Fig. 2). Whiskers, hair follicles, and skin also showed lacZ expression, but the level of expression appeared to be weaker than endogenous Irx2 RNA expression. However, lacZ expression could not be detected in the neural retina, which is known to express high levels of Irx2 (10, 18, 24). Some differences were observed between lacZ expression in Irx2LacZNeo/⫹ and Irx2LacZ/⫹ embryos. Compared with Irx2LacZ/⫹ embryos, Irx2LacZNeo/⫹ embryos showed a much higher level of lacZ expression in the heart at E9.5 and E10.5 (Fig. 2B, C, and D and data not shown). Interestingly, later in development (at E15), lacZ expression in the heart became comparable in the two lines (Fig. 2H and I). Weak lacZ expression was detected in the eye sclera of Irx2LacZ/⫹, but not Irx2LacZNeo/⫹, embryos (Fig. 2T, U, and V). Stronger lacZ expression was also detected in the hair follicles of Irx2LacZ/⫹ embryos (Fig. 2Z and AA). These results indicate that the expression of the lacZ reporter gene reproduces most of the endogenous Irx2 expression, although there are subtle differences in lacZ expression between the two mutant lines.

MOL. CELL. BIOL.

Homozygous Irx2 mutant mice are viable and fertile. To determine the mutant phenotype of mice lacking Irx2, heterozygous Irx2 mutant mice were intercrossed to generate homozygous mutant mice. Homozygous mutant mice were obtained at normal Mendelian ratios (26 ⫹/⫹:47 ⫹/⫺:19 ⫺/⫺ for the Irx2LacZNeo/⫹ line and 42 ⫹/⫹:70 ⫹/⫺:34 ⫺/⫺ for the Irx2LacZ/⫹ line), showing that the mutation is not lethal. Both Irx2LacZNeo/LacZNeo and Irx2LacZ/LacZ mice had no gross morphological defects and could reproduce normally. These results indicate that Irx2 is dispensable for embryonic development and is not required for most adult functions, including reproduction. Normal heart functions in homozygous Irx2 mutant mice. All six Irx genes display distinctive expression patterns in the developing mouse heart (6, 7, 9, 24). Irx2 is expressed in the ventricular septum of the heart and in the ventricular conduction system as early as E10.5 (7, 9). Histological staining of 6-month-old Irx2-deficient hearts did not reveal any aberrations in cardiac morphology, suggesting that Irx2 is not essential for heart development (data not shown). Physiological studies have previously revealed a cardiomyopathy phenotype in adult Irx4-deficient mice, although the animals have normal cardiac morphology during embryogenesis and in early postnatal life (7). To examine whether Irx2-deficient mice exhibit any subtle heart anomalies, cardiac function in Irx2LacZ/LacZ mice and their littermates was assessed by surface eight-lead and signal-averaged electrocardiography, two-dimensional echocardiography and Millar catheter-based invasive hemodynamic studies (25). Eight- to 10-week-old Irx2LacZ/LacZ mice showed no discernible cardiac phenotype compared to their heterozygous or wild-type littermates (Tables 1 and 2 and data not shown). Furthermore, Northern blot and real-time RTPCR analyses of markers for cardiac hypertrophy, including atrial natriuretic factor and ␤-myosin heavy chain, which were previously shown to be up-regulated in Irx4-deficient mice (7), showed that Irx2LacZ/LacZ mice displayed normal expression indistinguishable from that of wild-type mice (data not shown). Together, these results suggest that Irx2 is not required for normal cardiac development and function. The MHB appears to be normal in Irx2 mutants. In Xenopus embryos, Xiro1 is required for the formation of the MHB, which gives rise to the isthmic organizer. Misexpression studies indicate that Xiro1 activates the expression of Otx2, Gbx2, Fgf8, and En2 sequentially during the induction and the positioning of the isthmus organizer (14). Similarly, in zebra fish, ziro1 and ziro7 possess overlapping functions in the establishment of the isthmic organizer (19). Recent studies have revealed that chick Irx2 plays an important role in the specification of the MHB; in collaboration with the Fgf8 signal from the isthmic organizer, misexpression of Irx2 induces ectopic cerebellum development (T. Ogura, personal communication). In chicks and mice, Irx2 expression is very strong in the rhombomere 1 and 2 and weak in the rostral midbrain but is not detected in the caudal midbrain near the isthmus. To determine whether inactivation of Irx2 resulted in any defect in MHB patterning, we examined the expression of Otx2, Fgf8, and Pax2 in wild-type and Irx2LacZ/LacZ embryos. In mice, Otx2 and Gbx2 are responsible for the proper positioning of the isthmus, while Fgf8 and En2 play a later role in isthmus function (30). In E10 Irx2LacZ/LacZ embryos, Otx2 was normally expressed in the forebrain and the

VOL. 23, 2003


8221

FIG. 4. Normal digit development in Irx2-deficient mice. Whole-mount in situ hybridization of E13.5 limb buds is shown. (A to F) Expression of Irx1 (B, C, E, and F) and Irx2 (A and D) in wild-type (WT) forelimb buds (A and B), WT hindlimb buds (D and E), Irx2 mutant forelimb buds (C), and Irx2 mutant hindlimb buds (F). (G to J) Skeletal staining of WT and Irx2 mutant newborn limbs, using alcian blue and alizarin red, revealed no significant difference in cartilage and bone development.

midbrain, including the MHB (Fig. 3A and B). Fgf8 and Pax2 also showed normal MHB expression in Irx2LacZ/LacZ embryos (Fig. 3C, D, E, and F). Furthermore, whole-mount and histological analyses of adult brains did not reveal any obvious defects in the mutants (data not shown). These observations suggest that Irx2 is not required for the patterning of the midbrain-hindbrain region. Normal digit development in Irx2 mutant mice. As reported previously, Irx1 and Irx2 are expressed in the metatarsal cartilage and the cartilage of the phalanges around E13.5 (31) (Fig. 4A, B, D, and E). To examine digit development in Irx2deficient mice, we performed alcian blue and alizarin red staining of newborn skeleton (Fig. 4G to J). The stainings of the bones and cartilage were similar for wild-type and Irx2-deficient limbs, indicating that Irx2 is not required for the pattern-

ing and development of the digits. Furthermore, we did not observe any significant difference of Irx1 expression in E13.5 Irx2-deficient limbs (Fig. 4A, B, C, E, and F). Expression of Irx genes in Irx2-deficient embryos. The absence of a discernible phenotype in mice lacking Irx2 could be attributable to functional redundancy among the Irx genes. Functional redundancy of Iroquois genes has been revealed by previous studies on Iroquois mutations in Drosophila, zebra fish, and mouse. In Drosophila, mutations affecting only one Iroquois gene have very mild effects, whereas deletion of two or three Iroquois genes results in more pronounced phenotypes (8). In zebra fish embryos, the development of the MHB region is affected by knockdown of both ziro1 and ziro7, while the knockdown of either gene alone has no effect (19). In mice, the mild cardiac phenotype of Irx4 mutants might be due to partial

8222

LEBEL ET AL.

MOL. CELL. BIOL.

FIG. 5. Expression levels of other Irx genes show no major difference in E10.5 Irx2LacZ/LacZ embryos. Real-time RT-PCR was performed on RNA extracted from E10.5 wild-type (WT) and Irx2LacZ/LacZ embryos (n ⫽ 4 for each value). Results are shown as abundance relative to GADPH (internal control). Irx2 RNA is strongly reduced in Irx2LacZ/LacZ embryos, but the transcript levels of other Irx genes are comparable in WT and Irx2LacZ/LacZ embryos. Error bars indicate standard deviations.

compensation by up-regulation of Irx2 expression (7). To determine whether similar compensation is involved in Irx2 mutants, we examined the expression of other Irx genes in Irx2 mutant embryos. We first assessed the level of transcription of Irx genes in the Irx2-deficient embryos by performing real-time RT-PCR on E10.5 embryo RNA (Fig. 5). No major change of Irx1, -3, -4, -5, and -6 expression could be detected in Irx2deficient embryos. In contrast, the Irx2 transcript level was drastically reduced in Irx2-deficient embryos. To further examine whether the expression of other Irx genes is affected in Irx2-deficient embryos, whole-mount in situ hybridization was performed on wild-type and Irx2-deficient embryos. As shown in Fig. 6, no difference in the expression of Irx1 (Fig. 6A and B), Irx4 (Fig. 6I and J), Irx5 (Fig. 6G and H), and Irx6 (Fig. 6K and L) could be detected between E10 wild-type and homozygous Irx2 mutant embryos. However, ectopic Irx3 expression was detected in the MHBs of some homozygous Irx2 mutant embryos (four of seven of the embryos examined). While it remains unclear why only a fraction of homozygous Irx2 mutant embryos showed this defect, the ectopic expression of Irx3

appeared to have a very minor or no effect on MHB development (Fig. 3). DISCUSSION There is increasing evidence suggesting the importance of Irx genes in the specification and patterning of vertebrate embryos. To determine the role of Irx2 in mouse development, we have generated a mutant mouse strain by targeted disruption of the protein-coding region of Irx2 with an IRES-lacZ reporter gene. We show here that the expression of the lacZ reporter gene reproduces most of the endogenous expression of Irx2. In the homozygous Irx2 mutant embryos, Irx2 transcripts are significantly reduced and Irx2 protein could not be detected, confirming that the targeted mutation results in a loss-of-function allele. However, Irx2-deficient mice are viable and fertile and do not display any obvious defects, indicating that Irx2 is dispensable for development, fertility, and adult homeostasis. During mouse embryogenesis, Irx2 displays a highly dynamic

VOL. 23, 2003


8223

FIG. 6. Expression pattern of Irx genes in Irx2LacZ/LacZ embryos. Whole-mount RNA in situ hybridization of E10 wild type (A, C, E, G, I, and K) and Irx2LacZ/LacZ (B, D, F, H, J, and L) embryos is shown. (E and F) Irx2 RNA is significantly reduced in Irx2LacZ/LacZ embryos. (C and D) Ectopic expression of Irx3 (arrowhead) is detected in the MHBs of some Irx2LacZ/LacZ embryos. Expression of Irx1 (A and B), Irx4 (I and J), Irx5 (G and H), and Irx6 (K and L) is similar in wild-type and Irx2LacZ/LacZ embryos. The embryos in panels K and L were overstained to reveal the faint Irx6 expression in the heart and limb buds.

expression pattern in various tissues and organs, suggesting that it might possess multiple developmental functions (4, 8, 7, 9, 10, 18). Recent studies have strongly indicated a critical role for chick Irx2 in the development of the midbrain-hindbrain region (T. Ogura, personal communications). However, we show here that Irx2-deficient mice do not exhibit any defects in MHB patterning. During heart development, Irx2 is expressed in the cardiac interventricular septum (7, 9). It has been suggested that Irx2 might play a role in septum specification and, later during heart development, in the specification of components of the ventricular conduction system (9). Moreover, Irx2

expression is up-regulated in the hearts of Irx4-deficient mice, which show adult-onset cardiomyopathy (7). However, Irx2deficient hearts did not exhibit any morphological abnormalities or defects of cardiac function; it remains possible that in aged mice a subtle phenotype might become more apparent. Furthermore, we did not detect any obvious defects in the dorsoventral patterning of the hindbrain and spinal cord in the Irx2 mutant (data not shown). Similarly, limbs, lungs, and hair follicles, where Irx2 is also strongly expressed, appear to be normal in Irx2 mutants (Fig. 4 and data not shown). Together, our results suggest that Irx2 is not essential for the specification

8224

LEBEL ET AL.

MOL. CELL. BIOL.

and development of multiple cell types in many tissues and organs which show robust expression of Irx2 during development. A recent study has shown that Irx4 is involved in the regulation of neural retina expression of slit, which is implicated in axon guidance (20). In both embryonic and adult retinas, Irx2 is specifically expressed in the ganglion cell layer (10; unpublished data). We found that all major cell types develop normally in Irx2-deficient retinas (data not shown). It will be intriguing to determine whether subtle neuronal guidance defects can be found in the Irx2 mutant mice. The lack of a phenotype in Irx2-deficient mice is likely due to functional compensation by other Irx genes. Functional compensation between the Irx genes might involve increased or ectopic expression of other Irx genes. In Irx4-deficient mutants, Irx2 expression is up-regulated in the heart (7). However, we have been unable to detect increased or ectopic expression of Irx1 and Irx4, other members of the IroA gene cluster to which Irx2 belongs. Similarly, we did not find any misexpression of Irx5, which is structurally most related to Irx2, and Irx6. Interestingly, in some Irx2 mutant embryos, we found that Irx3 is ectopically expressed at the MHB, suggesting that there is a cross-regulation of Irx genes between the two Iro gene clusters. It remains to be determined whether this cross-regulation is direct or indirect and why only some of the Irx2 mutants show this misexpression of Irx3. Our results revealed that Irx2-deficient embryos show no major alteration in the expression of other Irx genes, suggesting that increased or ectopic expression of other Irx genes may not be involved in the functional compensation of Irx2-deficient mice. Since Irx1 shows an expression pattern almost identical to that of Irx2 and some of the other Irx genes also show overlapping expression with Irx2 during embryonic development, it is highly probable that these Irx genes might substitute for the function of Irx2 in Irx2deficient mice. In summary, we have shown here that Irx2 is not essential for normal development and adult homeostasis in mice. As suggested by genetic studies with Drosophila (8), the vertebrate Irx genes likely share extensive overlapping functions during development. Further studies of mutant mice of other Irx genes as well as compound mutant mice will be needed to dissect the functional roles of this class of evolutionarily conserved transcriptional regulators in development and disease. ACKNOWLEDGMENTS This research was funded by operating grants from the Canadian Institutes of Health Research (CIHR) to C.-C.H and B.G.B and by a CIHR interdisciplinary group grant to B.G.B. and M.H. The work described in this paper was also partially supported by a grant from the Research Grants Council of the Hong Kong SAR (project no. City U 1164/02 M). M.L. is the recipient of a doctoral award from the Fond de la Recherche en Sante´ du Que´bec, P.A. is the recipient of a National Science and Engineering Research Council of Canada Scholarship, T.Y.C was supported by studentships by the Canadian Hypertension Society and Heart and Stroke Foundation of Ontario, M.H. holds a Clinician Scientist Award from CIHR, B.G.B. holds a Canada Research Chair in Developmental Cardiology, and C.-C.H. is a Research Scientist of the National Cancer Institute of Canada. We thank C. Lobe for the gift of the NLS-Cre mice. REFERENCES 1. Ang, S. L., R. A. Conlon, O. Jin, and J. Rossant. 1994. Positive and negative signals from mesoderm regulate the expression of mouse Otx2 in ectoderm explants. Development 120:2979–2989.

2. Bao, Z. Z., B. G. Bruneau, J. G. Seidman, C. E. Seidman, and C. L. Cepko. 1999. Regulation of chamber-specific gene expression in the developing heart by Irx4. Science 283:1161–1164. 3. Bellefroid, E. J., A. Kobbe, P. Gruss, T. Pieler, J. B. Gurdon, and N. Papalopulu. 1998. Xiro3 encodes a Xenopus homolog of the Drosophila Iroquois genes and functions in neural specification. EMBO J. 17:191–203. 4. Bosse, A., A. Zu ¨lch, M. B. Becker, M. Torres, J. L. Go ´mez-Skarmeta, J. Modolell, and P. Gruss. 1997. Identification of the vertebrate Iroquois homeobox gene family with overlapping expression during early development of the nervous system. Mech. Dev. 69:169–181. 5. Briscoe, J., A. Pierani, T. M. Jessell, and J. Ericson. 2000. A homeodomain protein code specifies progenitor cell identity and neuronal fate in the ventral neural tube. Cell 101:435–445. 6. Bruneau, B. G., Z. Z. Bao, M. Tanaka, J. J. Schott, S. Izumo, C. L. Cepko, J. G. Seidman, and C. E. Seidman. 2000. Cardiac expression of the ventriclespecific homeobox gene Irx4 is modulated by Nkx2.5 and dHand. Dev. Biol. 217:266–277. 7. Bruneau, B. G., Z. Z. Bao, D. Fatkin, J. Xavier-Neto, D. Georgakopoulos, C. T. Maguire, C. I. Berul, D. A. Kass, M. L. Kuroski-de Bold, A. J. de Bold, D. A. Conner, N. Rosenthal, C. L. Cepko, C. E. Seidman, and J. G. Seidman. 2001. Cardiomyopathy in Irx4-deficient mice is preceded by abnormal ventricular gene expression. Mol. Cell. Biol. 21:1730–1736. 8. Cavodeassi, F., J. Modolell, and J. L. Go ´mez-Skarmeta. 2001. The Iroquois family of genes: from body building to neural patterning. Development 128:2847–2855. 9. Christoffels, V. M., A. G. Keijser, A. C. Houweling, D. E. W. Clout, and A. F. Moorman. 2000. Patterning the embryonic heart: identification of five mouse Iroquois homeobox genes in the developing heart. Dev. Biol. 224:263–274. 10. Cohen, D. R., C. W. Cheng, S. H. Cheng, and C.-C. Hui. 2000. Expression of two novel mouse Iroquois homeobox genes during neurogenesis. Mech. Dev. 91:317–321. 11. Ding, Q., J. Motoyama, S. Gasca, R. Mo, H. Sasaki, J. Rossant, and C.-C. Hui. 1998. Diminished Sonic hedgehog signaling and lack of floor plate differentiation in Gli2 mutant mice. Development 125:2533–2543. 12. Dressler, G. R., U. Deutsch, K. Chowdhury, H. O. Nornes, and P. Gruss. 1990. Pax2, a new murine paired-box-containing gene and its expression in the developing excretory system. Development 109:787–795. 13. Glavic, A., J. L. Go ´mez-Skarmeta, and R. Mayor. 2001. Xiro-1 controls mesoderm patterning by repressing bmp-4 expression in the Spemann organizer. Dev. Dyn. 222:368–376. 14. Glavic, A., J. L. Go ´mez-Skarmeta, and R. Mayor. 2002. The homeoprotein Xiro1 is required for midbrain-hindbrain boundary formation. Development 129:1609–1621. 15. Go ´mez-Skarmeta, J. L., and J. Modolell. 2002. Iroquois genes: genomic organization and function in vertebrate neural development. Curr. Opin. Genet. Dev. 12:403–408. 16. Go ´mez-Skarmeta, J. L., A. Glavic, E. de la Calle-Mustienes, J. Modolell, and R. Mayor. 1998. Xiro, a Xenopus homolog of the Drosophila Iroquois complex genes, controls development at the neural plate. EMBO J. 17:181–190. 17. Harlow, E., and D. Lane. 1988. Antibodies: a laboratory manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. 18. Houweling, A. C., R. Dildrop, T. Peters, J. Mummenhoff, A. F. Moorman, U. Ru ¨ther, and V. M. Christoffels. 2001. Gene and cluster-specific expression of the Iroquois family members during mouse development. Mech. Dev. 107: 169–174. 19. Itoh, M., T. Kudoh, M. Dedekian, C. H. Kim, and A. B. Chitnis. 2002. A role for iro1 and iro7 in the establishment of an anteroposterior compartment of the ectoderm adjacent to the midbrain-hindbrain boundary. Development 129:2317–2327. 20. Jin, Z., J. Zhang, A. Klar, A. Che´dotal, Y. Rao, C. L. Cepko, and Z. Z. Bao. 2003. Irx4-mediated regulation of Slit1 expression contributes to the definition of early axonal paths inside the retina. Development 130:1037–1048. 21. Kobayashi, D., M. Kobayashi, K. Matsumoto, T. Ogura, M. Nakafuku, and K. Shimamura. 2002. Early subdivisions in the neural plate define distinct competence for inductive signals. Development 129:83–93. 22. Kudoh, T., and I. B. Dawid. 2001. Role of the iroquois3 homeobox gene in organizer formation. Proc. Natl. Acad. Sci. USA 98:7852–7857. 23. Mo, R., A. M. Freer, D. L. Zinyk, M. A. Crackower, J. Michaud, H. H. Heng, K. W. Chik, X. M. Shi, L. C. Tsui, S. H. Cheng, A. L. Joyner, and C.-C. Hui. 1997. Specific and redundant functions of Gli2 and Gli3 zinc finger genes in skeletal patterning and development. Development 124:113–123. 24. Mummenhoff, J., A. C. Houweling, T. Peters., V. M. Christoffels, and U. Ru ¨ther. 2001. Expression of Irx6 during mouse embryogenesis. Mech. Dev. 103:193–195. 25. Mungrue, I. N., R. Gros, X. You, A. Pirani, A. Azad, T. Csont, R. Schulz, J. Butany, D. J. Stewart, and M. Husain. 2002. Cardiomyocyte overexpression of iNOS in mice results in peroxynitrite generation, heart block, and sudden death. J. Clin. Investig. 109:735–743. 26. Peters, T., R. Dildrop, K. Ausmeier, and U. Ru ¨ther. 2000. Organization of mouse Iroquois homeobox genes in two clusters suggests a conserved regulation and function in vertebrate development. Genome Res. 10:1453–1462.

VOL. 23, 2003


27. Peters, T., K. Ausmeier, R. Dildrop, and U. Ru ¨ther. 2002. The mouse Fused toes (Ft) mutation is the result of a 1.6-Mb deletion including the entire Iroquois B gene cluster. Mamm. Genome 13:186–188. 28. Tanaka, A., K. Miyamoto, N. Minamino, M. Takeda, B. Sato, H. Matsuo, and K. Matsumoto. 1992. Cloning and characterization of an androgeninduced growth factor essential for the androgen-dependent growth of mouse mammary carcinoma cells. Proc. Natl. Acad. Sci. USA 89:8928–8932.

8225

29. van der Hoeven, F., T. Schimmang, A. Volkmann, M. G. Mattei, B. Kyewski, and U. Ru ¨ther. 1994. Programmed cell death is affected in the novel mouse mutant Fused toes (Ft). Development 120:2601–2607. 30. Wurst, W., and L. Bally-Cuif. 2001. Neural plate patterning: upstream and downstream of the isthmic organizer. Nat. Rev. Neurosci. 2:99–108. 31. Zu ¨lch, A., M. B. Becker, and P. Gruss. 2001. Expression pattern of Irx1 and Irx2 during mouse digit development. Mech. Dev. 106:159–162.