- Email: [email protected]
Human ARX gene: genomic characterization and expression
Molecular Genetics and Metabolism 77 (2002) 179–188 www.academicpress.com
Human ARX gene: genomic characterization and expression R. Ohira,a Y.-H. Zhang,b W. Guo,b K. Dipple,b S.L. Shih,b J. Doerr,c B.-L. Huang,b L.J. Fu,d A. Abu-Khalil,d D. Geschwind,d and E.R.B. McCabea,b,* a
Department of Human Genetics, David Geffen School of Medicine at UCLA, 10833 Le Conte Avenue, Los Angeles, CA 90095-1752, USA b Department of Pediatrics, David Geffen School of Medicine at UCLA, Los Angeles, CA 90095-1752, USA c Departments of Microbiology and Molecular Genetics, David Geffen School of Medicine at UCLA, Los Angeles, CA 90095-1752, USA d Department of Neurology and Program in Neurogenetics, David Geffen School of Medicine at UCLA, Los Angeles, CA 90095-1752, USA Received 28 June 2002; accepted 3 July 2002
Abstract Arx is a homeobox-containing gene with a high degree of sequence similarity between mouse and zebrafish. Arx is expressed in the forebrain and floor plate of the developing central nervous systems of these vertebrates and in the presumptive cortex of fetal mice. Our goal was to identify genes in Xp22.1–p21.3 involved in human neuronal development. Our in silico search for candidate genes noted that annotation of a human Xp22 PAC (RPCI1-258N20) sequence (GenBank Accession No. AC002504) identified putative exons consistent with an Arx homologue in Xp22. Northern blot analysis showed that a 3.3 kb human ARX transcript was expressed at high levels in fetal brain. A 5.9 kb transcript was expressed in adult heart, skeletal muscle, and liver with very faint expression in other adult tissues, including brain. In situ hybridization of ARX in human fetal brain sections at various developmental stages showed the highest expression in neuronal precursors in the germinal matrix of the ganglionic eminence and in the ventricular zone of the telencephalon. Expression was also observed in the hippocampus, cingulate, subventricular zone, cortical plate, caudate nucleus, and putamen. The expression pattern suggests that ARX is involved in the differentiation and maintenance of specific neuronal cell types in the human central nervous system. We also mapped the murine Arx gene to the mouse genome using a mouse/hamster radiation hybrid panel and showed that Arx and ARX are orthologues. Therefore, investigations in model vertebrates may provide insight into the role of ARX in development. The recent identification of ARX mutations in patients with various forms of menta1 retardation make such studies in model organisms even more compelling. Ó 2002 Elsevier Science (USA). All rights reserved. Keywords: Aristaless-Related Homeobox (ARX) gene, human; CNS developmental gene candidate, ARX; Developmental expression, human, ARX; Fetal CNS development, ARX; Homeobox gene; In situ hybridization; Transcription factor, ARX
1. Introduction Vertebrate brain development is a complex process that entails various steps, including cell identity determination, cell migration, and regional differentiation. Transcription factors are expressed in a space- and timedependent manner creating distinct regions, which eventually develop into various parts of the brain including forebrain, midbrain, hindbrain, cerebellum, and spinal cord [1]. Regionalization involves the interplay of genetic and environmental factors and a number of transcription factors have been implicated in prosencephalic regional*
Corresponding author. Fax: 1-310-206-4584. E-mail address: [email protected] (E.R.B. McCabe).
ization [2]. One way to study regional specification in the brain has been to observe regional and temporal expression of transcription factors and signaling molecules during development. Identifying the organizing center of a particular region of the brain is important in establishing a basis for understanding neuronal development. The midbrain organizer, the isthmus, has been identified and characterized as the spinal cord organizing center, the floor plate [3]. The forebrain organizer region remains to be identified, but it has been argued that once discovered, it will elucidate the roles of each of the key factors involved in forebrain development [4]. Aristaless Related Homeobox (ARX) is a member of the group II aristaless-related protein family, members of which are expressed primarily in the central and/or
1096-7192/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved. PII: S 1 0 9 6 - 7 1 9 2 ( 0 2 ) 0 0 1 2 6 - 9
180
R. Ohira et al. / Molecular Genetics and Metabolism 77 (2002) 179–188
peripheral nervous system [5]. It contains two conserved domains: an aristaless domain also known as a C-peptide and a homeobox domain belonging to the prd-like class. ARX has been previously shown to have a high degree of sequence similarity between mouse and zebrafish [4]. Arx is expressed in the forebrain and floor plate of the developing central nervous system (CNS) in each of these vertebrates and in the presumptive cortex and hippocampus of fetal mice [4,6]. The genomic localization of Arx remained undefined when we initiated these investigations, though subsequent reports placed it in Xp22 [7,8]. Mutations in ARX have been reported recently among patients with various forms of mental retardation [6,9]. Our goal was to identify and characterize genes in Xp22.1–p21.3 that might be involved in human CNS development and deleted in patients with contiguous gene syndromes involving Xp22.1–p21.3 [10]. An in silico search revealed that putative exons consistent with an Arx homologue had been identified in Xp22, within the chromosomal region of interest. Northern blot analysis and in situ hybridization were performed to characterize the expression patterns of human ARX. The murine Arx gene was mapped to the mouse X chromosome in a region syntenic with Xp22.
2. Materials and methods 2.1. In silico identification of ARX An in silico search for candidate CNS developmental genes was conducted to identify genes within the desired, defined region of the X chromosome, Xp22.1–p21.3. Exons putatively defined as ARX (GenBank Accession No. AC002504) had been identified by the Baylor Genome Center (http://www.hgsc.bcm.tmc.edu/). 2.2. Sequencing of ARX cDNA Primer pairs, A1 50 gcaagttcagcctggttaag30 (50 end of vector), A2 50 tacgcgcatacctggtgaagacgtccgggtag30 (nucleotides 1052-IVS2+11) and B1 50 ttatgagtatttcttccag gg30 (30 end of vector), B2 50 aagaagatgaggacgaggaaga Table 1 Primers used for sequencing of ARX Primer 1 2 3 4 5 6
nt 0
0
5 CCA ACA CAC ACC CAT CCA TC3 50 CTT GTT ACC GCT TGT CCT GAG30 50 GGA CGA AGA AGA TGA GGA CGA GGA AGA GGA AC30 50 TGA AAG CTG GGT GTC GGA ACA C30 50 TTC CCG AGC CTA CCT CCG CCT C30 50 GCT CCT GAG ACA GCC CAC AC30
)81 to )62 )222 to )202 695 to 726 1403 to 1428 1320 to 1339 1496 to 1515
ggaac30 (nucleotides 700–726), were specifically designed to obtain overlapping PCR fragments from a human fetal brain cDNA library (Human Brain, 50 -stretch plus cDNA library, Cat. No. HL3008a; Clontech Laboratories, Palo Alto, CA) that covered the putative coding sequence. The amplified cDNA fragments were used as probes to screen the human fetal brain library and the cDNA clone was sequenced by automated cycle sequencing (ABI377, Perkin–Elmer, Foster City, CA) using the primers in Table 1. 2.3. Characterization of ARX expression Northern blot analysis was performed using adult and fetal human multiple tissue northern (MTN) blots (Clontech) containing poly (A)þ RNA. The probe used on both blots was a 673 bp fragment from exon 2 (nucleotides 400–1073). The fragment was first amplified by PCR (primer A2 (see above) and 50 aagaggccaaggcgtcga agtc30 (IVS1–150 to IVS1–129)) from genomic DNA and digested with EcoRI and NotI (to obtain an optimal length), then probes were labeled with [a-32 P]dCTP using the Prime-It II Random Primer Labeling Kit and purified using NUC-TRAP (Stratagene, La Jolla, CA). The blots were first incubated in ExpressHyb Solution (Clontech) at 68 °C for 30 min and then in the same solution containing the labeled ARX probe (107 cpm) at 68 °C for 1 h. After incubation with the probe, the blots were washed twice in Solution 1 (2 SSC and 0.05% SDS) at 68 °C for 10 min and twice in Solution 2 (0.1 SSC and 0.1% SDS) at 70 °C for 5 min. Autoradiographs were obtained after exposing the X-OMAT scientific imaging film (Kodak, Rochester, NY) for 1–2 days at )80 °C. Human fetal tissue (15–22 gestational weeks) obtained from the Brain and Tissue Bank for Developmental Disorders at the University of Maryland, Baltimore, Maryland, was cryoprotected and then sectioned into 20 lm coronal sections that were thawmounted onto glass slides. The sections were fixed in 4% paraformaldehyde for 20 min followed by a 10-min wash in 0.1 M phosphate buffer (PB) (0.0774 M Na2 PO4 , 0.0226 M NaH2 PO4 , pH 7.4). The ARX insert was obtained by similar methods as the Northern probe and the EcoRI and NotI fragment was subcloned into pBluescript. Anti-sense and sense RNA riboprobes were transcribed in vitro from a linearized plasmid containing the ARX insert (673 bp region of exon 2) using an RNA Transcription Kit (Stratagene) and [35 S]UTP (Amersham, Arlington Heights, IL). The tissue sections were hybridized with either the sense or anti-sense radioactive riboprobe (107 cpm) and were incubated overnight at 60 °C. Following the hybridizations, the slides were treated with RNase A to remove unhybridized probe and then washed in descending concentrations of SSC in the presence of sodium thiosulfate. Autoradiographic
R. Ohira et al. / Molecular Genetics and Metabolism 77 (2002) 179–188
film (Amersham) was placed on the slides and allowed to develop for five days. The slides were dipped in NTB2 emulsion (Kodak) and stored in light tight boxes for 4–6 weeks. The slides were then developed and counterstained with cresyl violet and were examined by both bright and dark field microscopy. 2.4. Murine Arx mapping A mouse/hamster radiation hybrid panel (Research Genetics, Huntsville, AL) was used to map the murine Arx. The mouse Arx (GenBank Accession No. NM_007492) sequence was compared to hamster Arx (GenBank Accession No. AY038070) and human ARX (GenBank Accession No. AY038071) to find a unique region in the mouse sequence. The human ARX sequence was used as an additional species comparison to verify the unique murine sequence. Primers 50 gcgcgtcaa gcatagccgcg30 (nucleotides 1678–1697) and 50 gtggactttga gttggagtg30 (nucleotides 1936–1955)) were designed to this unique region and PCR was performed. The results were scored using the Mouse Genome Database (The Jackson Laboratory, Bar Harbor, ME) and the LOD Score Method (www.informatics.jax.org).
181
Table 2 Intron–exon boundaries for human ARX Intron
Exon
Intron
50 gccccagcc ttgcccgcag tttcttatag tgccttgcag ctccctgcag
ATGAG 1 GCAAG GCTCC 2 ACCAG GGAGG 3 TCCAG GTCTG 4 GGCAG GCTCT 5 TGTGC
gtaaggatgc gtatgcgcgt gtgagctgca gtaacgcgca taaaggctgc30
Sequencing of the human fetal brain ARX cDNA confirmed the putative exons identified in the genomic DNA (Table 2, Fig. 1). The ARX cDNA consists of five exons spanning a region of 1689 bp and coding for a deduced 562 amino acid protein. Table 2 shows the intron–exon boundaries for each of the five exons. The intron–exon boundaries were identified by comparison of the genomic and cDNA sequences. Comparison of the human ARX and murine Arx genes revealed 89% identity at the nucleotide level and 95% identity at the amino acid level. Fig. 2 shows the nucleotide and amino acid sequence alignment of both ARX and Arx.
(Figs. 3 and 4). Hybridization of the ARX exon 2 probe with the fetal MTN blot (Fig. 3) showed a 3.3 kb transcript at very high levels in fetal brain compared with fetal kidney, liver, and lung, where hybridization was minimal or absent. Hybridization with RNA from adult tissues showed a signal for a 5.9 kb transcript in heart, skeletal muscle, and liver (Fig. 4). A faint signal was seen in the other adult tissues (brain, thymus, spleen, kidney, placenta, lung, and blood leukocytes). Both MTN blots were probed for b-actin as a control and all lanes were loaded equally (data not shown). Human fetal brain sections at three different gestational weeks (gw) of development were examined for ARX expression. In situ hybridization results showed a pattern consistent with expression in neuronal precursor cells of the germinal zones. At 15 gw (Figs. 5A and 6A–D), intense signal was seen in the germinal matrix (gm) of the ganglionic eminence, and ventricular zone (vz) of the telencephalon, with less signal in the subventricular zone (svz), cortical plate (cp), caudate nucleus (cn), and putamen (p). By 19 gw (Fig. 5B), intense signal was still observed in the gm and vz, with lower signal in the svz, cp, cn, p, globus pallidus (gp), and substantia nigra (sn). At 22 gw (Figs. 5C and 6E– H), the ganglionic eminence (still showing high levels of expression) was observed to have expanded within the developing brain. Strong signals were still seen in the vz as well. Less intense signal was seen in the svz, cp, the cingulate (ci), and the hippocampus (h), particularly in the dentate gyrus (dg) and entorhinal cortex (E).
3.2. Characterization of ARX expression
3.3. Murine Arx mapping
Northern blot analysis showed that ARX had much higher expression in human fetal brain than adult brain
The radiation hybrid panel allowed mapping of the murine Arx gene to the mouse genome. Arx
3. Results 3.1. Sequencing of ARX cDNA
Fig. 1. Genomic structure of the human ARX gene. The numbered boxes indicate exons. The lines indicate introns. The numbers above the exons and below the introns indicate the sizes in basepairs (bp) (GenBank Accession No. AC002504). The coding region for ARX is 1689 bp (562 amino acids in deduced sequence).
182
R. Ohira et al. / Molecular Genetics and Metabolism 77 (2002) 179–188
Fig. 2. Nucleotide and amino acid comparison of human ARX (above) and murine Arx (below). Asterisks indicate regions of identity and hyphens indicate gaps in the gene sequence.
mapped to the murine X chromosome in the region containing Dmd, Gyk, and Zfx. The data showed the highest anchor LOD (14.2) to DXMit6, then LOD (12.6) to DXMit63 and finally LOD (11.7) to DXMit112. The Arx locus of interest was determined to be between DXMit112, 78.9 cR proximal and DXMit6.7, 7.04 cR distal, in the region syntenic with Xp22–p21.
4. Discussion An in silico search for candidate CNS developmental genes revealed an annotation of a human Xp22 PAC (RPCI1-258N20) sequence deposited by the Baylor Human Genome Center (GenBank Accession No. AC002504), which identified putative exons, suggesting an Arx homologue in Xp22, within the chromosomal
R. Ohira et al. / Molecular Genetics and Metabolism 77 (2002) 179–188
183
Fig. 2. (continued)
region of interest. The human ARX appeared to have a much higher expression in fetal brain than adult brain and increased expression in adult heart, skeletal muscle, and liver. There was very high expression of the 3.3 kb transcript in fetal brain. In situ hybridizations of human
fetal brain (15–22 gw) sections showed a pattern consistent with expression of ARX in neuronal precursor cells. An intense signal was seen in the germinal matrix and ventricular zone at all stages of development, while weaker signals were seen in the subventricular zone,
184
R. Ohira et al. / Molecular Genetics and Metabolism 77 (2002) 179–188
Fig. 3. Human fetal MTN blot probed for ARX (region of exon 2). A 3.3 kb transcript was seen in human fetal brain.
Fig. 4. Human adult MTN blot probed for ARX (fragment from exon 2). A 5.9 kb fragment is seen in the heart, skeletal muscle, and liver, but at this exposure minimal if any signal is observed in brain.
cortical plate, caudate nucleus, putamen, globus pallidus, substantia nigra (19 gw), cingulate, and hippocampus (22 gw). The murine Arx gene was mapped to the mouse genome to determine whether Arx and ARX are orthologues, or paralogues and simply members of the same gene family. Arx mapped to the murine X chromosome in the region containing Dmd, Gyk, and Zfx, Arx therefore is syntenic with the ARX locus, and Arx and ARX are orthologues. The expression patterns of Arx in mouse and ARX in human brain sections are remarkably similar [4,6]. The expression in the hippocampus is particularly interesting with the recent reports of ARX mutations in patients with mental retardation [6,9]. Although expression in the fetal MTN blots is similar in our hands and those of Bienvenu [6], all three groups differ in the expression patterns observed in the adult Northern blots [6,9]. These differences may be due to the fact that each group used a different probe, and we have observed differences in hybridization to MTN blots depending on the probe that we used (unpublished observations). ARX is a member of the group-II aristaless rented homeobox proteins involved in CNS development, containing two highly conserved domains, the C-peptide, and the prd-like class homeobox domain [5]. The
exact function of ARX is unknown, but the C-peptide domain is highly conserved with a similar sequence and position to the Otp (orthopedia) homeoprotein, a member of this protein family thought to be important in regionalization of the ventral diencephalon [11]. Deletion of the C-peptide in Otp reduces transactivation activity [11] and this domain may serve a similar function in Arx and ARX [4]. ARX is one of many homeodomain genes expressed during development. The common link between these genes is the highly conserved gene sequence called the homeobox, which encodes a protein domain of approximately 60 amino acids [12]. The homeodomain proteins act as nuclear transcriptional regulators by activating or repressing eukaryotic transcription. Many of these transcription factors contain various domains with functions such as DNA binding, dimerization, and transcriptional activation/repression [13]. There are several classes of homeobox genes with specific roles in CNS development: Dlx, Emx, and Otx. The Dlx genes are primarily involved in basal forebrain ganglia differentiation and are required for craniofacial development [14–21]. The Emx genes have been shown to play a critical role in establishing regions of forebrain and midbrain [22]. Recent data suggest a role for Emx proteins in neuroblast proliferation [22,23], and migration [24,25], as well as cognitive processes [23,26,27]. Mutations in Emx2 have been found in individuals with schizencephaly, a rare cell migration disorder, which causes full thickness clefting of the cerebral hemispheres [28,29]. Otx genes most likely play a crucial role in establishing the identity of various regions of the developing brain [30,31]. Specifically, Otx may be involved in fasciculating and guiding nerve fibers [32]. Arx was previously shown to be expressed in the forebrain and floor plate of the developing CNS of mouse and zebrafish and in the presumptive cortex and hippocampus of fetal mice [4,6]. In the forebrain, Arx is expressed in the intermediate and mantle zones, but in the presumptive cerebral cortex, expression is seen in the ventricular zone suggesting different functions of Arx in the two regions. Within the zona limitans intrathalamica (ZLI) of the mouse, a region within the diencephalon, Arx is expressed primarily in the ventral thalamus, ganglionic eminence, and floor plate at E10-11 d.p.c., similar to expression patterns for Dlx1 [33]. Comparisons of Arx and Brx1 expression show that the dorsalmost expression of Arx overlaps with Brx1. However, Arx expression appears to have a broad path containing the spindle-shaped cells while the Brx1 expression shows a narrow path on the dorsal side of the Arx path. Examination of three coronal levels of the ventral lateral geniculate nucleus indicate Arx expression only in level 2. However, expression of Arx and Dlx1 is seen in the external medullary lamina of levels 1 and 2 [33]. The expression patterns of Arx suggest a role in the forma-
R. Ohira et al. / Molecular Genetics and Metabolism 77 (2002) 179–188
185
Fig. 5. In situ hybridization of coronal sections from human fetal brain at 15 gestational weeks (gw) (A), 19 gw (B), and 22 gw (C) probed with an ARX probe. Strong signals are seen in the ventricular zone (vz) of the telencephalon and the germinal matrix (gm) of the ganglionic eminence. Less intense signals are seen in the cortical plate (cp), caudate nucleus (cn), putamen (p), globus pallidus (gp), subventricular zone (svz), substantia nigra (sn), cingulate (ci), and the hippocampus (h). No signal is seen in the internal capsule (ic).
tion of the reticular thalamic nucleus as well as the ventral lateral geniculate nucleus. Both human ARX and murine Arx were shown to be expressed in the hippocampus and human ARX was shown to be expressed in the cingulate, which suggest a role in cognitive processes and memory. Two recent reports have mapped Arx to the region syntenic to the human ARX [7,8]. Our results confirm that Arx maps to the X chromosome near Dmd and Zfx, which makes it syntenic to the human ARX locus. Therefore, Arx and ARX are in fact orthologues. Sequence comparison of the two genes showed 89% identity at the nucleotide level and 95% identity at the amino acid level. The close similarity of these two genes allows for their investigation in model vertebrates, which may provide valuable information regarding ARX expression and the implications of the disruption of its expression.
Of particular interest in studies of ARX orthologues in model organisms will be investigations to identify potential modifier genes. The phenotypes of patients with ARX mutations are highly variable, including mental retardation (syndromic and nonspecific), epilepsy (infantile spasms and myoclonic seizures), and dystonia [6,9]. Among the 19 families with ARX mutations, 14 had alanine-repeat expansions and 11 of the 14 had the identical polyalanine expansion (EX2_428_451dup) from 12 to 20 alanines in length. The other mutations included four missense mutations (L33P, Q163R, G286S, and P353L), and a frameshift mutation (IVS4_EX5del1517). It is interesting to note that 11 of the 19 (58%) of the families identified with ARX mutations had the identical polyalanine expansion. Among individuals with the identical ARX genotype (e.g., EX2_428_451dup), there was considerable phenotypic variability. The variability in phenotype shows
186
R. Ohira et al. / Molecular Genetics and Metabolism 77 (2002) 179–188
Fig. 6. High magnification of coronal sections from human fetal brain at 15 gw (A)–(D) and 22 gw (E)–(H) in bright (A), (C), (E), and (G) and darkfield (B), (D), (F), and (H) views. The basal ganglia of a 15 gw human fetal brain section shows intense staining in the germinal matrix (gm), and ventricular zone (vz), lighter staining in the cortical plate (cp), caudate nucleus (cn), and putamen (p), and no staining in the internal capsule (ic) (A)– (D). Staining of the vz and gm of a 22 gw brain section are observed at high magnification (E) and (F). The high magnification of the 22 gw hippocampus is shown (G) and (H). Less intense signal is seen in the dentate gyrus (dg) and the entorhinal cortex (E), (G), and (H). Bar equals 150 lm.
the complexity of this ‘‘single gene disorder,’’ a feature common to many ‘‘simple’’ Mendelian genetic diseases [34–37]. The complexity of proteomic networks and the influences of modifier genes on these networks are sure to contribute to the lack of genotype–phenotype corre-
lation. Investigations in model organisms and patients with ARX mutations may help us to identify and characterize genes modifying phenotypic expression following disruption of ARX, a gene critical to central nervous system development.
R. Ohira et al. / Molecular Genetics and Metabolism 77 (2002) 179–188
Acknowledgments These investigations were supported in part by grants from the NIH (RO1 HD22563, P3O HD34610, and MH60233; DHG). We thank Hooman Allayee from the Department of Human Genetics at the David Geffin School of Medicine at UCLA for assisting us with murine mapping.
[15]
References
[18]
[1] M. Hatten, Central nervous system neuronal migralion, Annu. Rev. Neurosci. 22 (1999) 511–539. [2] K. Shimamura, J. Rubenstein, Inductive interactions direct early regionalization of the mouse forebrain, Development 124 (1997) 2709–2718. [3] S.L. Ang, The brain organization, Nature 380 (1996) 25–27. [4] H. Miura, M. Yanazawa, K. Kato, K. Kitamura, Expression of a novel aristaless related homeobox gene ÔArxÕ in the vertebrate telencephalon, diencephalon and floor plate, Mech. Dev. 65 (1997) 99–109. [5] F. Meijlink, A. Veverdam, A. Brouwer, T. Oosterveen, D. Berge, Vertebrate aristaless-related genes, Int. J. Dev. Biol. 43 (1999) 651–663. [6] T. Bienvenu, K. Poireier, G. Friocourt, N. Bahi, D. Beaumont, F. Fauchereau, L.B. Jeema, R. Zemni, M. Vinet, F. Francis, P. Couvert, M. Gomot, C. Moraine, H. van Bokhoven, V. Kalscheurer, S. Frints, J. Gecz, K. Ohzaki, H. Chaabouni, J. Fryns, V. Desportes, C. Beldjord, J. Chelly, ARX, a novel Prd-classhomeobox gene highly expressed in the telencephalon, is mutated in X-linked mental retardation, Hum. Mol. Genet. 11 (2002) 981– 991. [7] H. Blair, V. Reed, E. Gormally, J. Wilson, J. Novak, R. McInnes, S. Phillips, B. Taylor, Y. Boyd, Positioning of five genes (CASK, ARX, SAT, IMAGE cDNAs 248928 and 253949) from the human X chromosome short arm with respect to evolutionary breakpoints on the mouse X chromosome, Mamm. Genome 11 (2000) 710–712. [8] D. Cunningham, Q. Xiao, A. Chatterjee, K. Sulik, D. Juriloff, F. Elder, W. Harrison, G. Schuster, P.A. Overbeek, G.E. Herman, exma: an X-linked insertional mutation that disrupts forebrain and eye development, Mamm. Genome 13 (2002) 179–185. [9] P. Stromme, M.E. Mangelsdorf, M.A. Shaw, K.M. Lower, S.M.E. Lewis, H. Bruyere, V. Lutcherath, A.K. Gedeon, R.H. Wallace, I.E. Scheffer, G. Turner, M. Partington, S.G.M. Frints, J.P. Fryns, G.R. Sutherland, J.C. Mulley, J. Gecz, Mutations in the human ortholog of Aristaless cause X-linked mental retardation and epilepsy, Nat. Genet. 30 (2002) 441–445. [10] E.R.B. McCabe, Disorders of glycerol metabolism, in: C.R. Scriver, A.L. Beaudet, W.S. Sly, D. Valle (Eds.), The Metabolic & Molecular Bases of Inherited Diseases, eighth ed., McGraw-Hill, New York, 2001, pp. 2217–2237. [11] A. Simeone, M.R. Dapice, V. Nigro, J. Casanova, F. Graziani, D. Acampora, V. Avantaggiato, Orthopedia, a novel homeoboxcontaining gene expressed in the developing CNS of both mouse and Drosophila, Neuron 13 (1994) 83–101. [12] W.J. Gehring, M. Affolter, T. Burglin, Homeodomain proteins, Annu. Rev. Biochem. 63 (1994) 487–526. [13] J. Vollmer, R. Clerc, Homeobox genes in the developing mouse brain, J. Neurosci. 71 (1998) 1–19. [14] S.A. Anderson, M. Qiu, A. Bulfone, D.D. Eisenstat, J. Meneses, R. Pederson, J.L.R. Rubenstein, Mutations of the homeobox genes Dlx-1 and Dlx-2 disrupt the striatal subventricular zone and
[16]
[17]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
187
differentiation of late born striatal neurons, Neuron 19 (1997) 27– 37. S.A. Anderson, D.D. Eisenstat, L. Shi, J.L.R. Rubenstein, Interneuron migration from basal forebrain to neocortex; dependence on Dlx genes, Science 278 (1997) 474–476. A. Bulfone, H.J. Kim, L. Puelles, M.H. Porteus, J.F. Grippo, J.L.R. Rubenstein, The mouse Dlx-2 (Tes-1) gene is expressed in spatially restricted domains of the forebrain, face and limbs in midgestation mouse embryos, Mech. Dev. 40 (1993) 129–140. M.J. Depew, J.K. Liu, J.E. Long, R. Presley, J.J. Meneses, R.A. Pedersen, J.L.R. Rubenstein, Dlx5 regulates regional development of the branchial arches and sensory capsules, Development 126 (1999) 3831–3846. D.D. Eisenstat, J.K. Liu, M. Mione, W. Zhong, G. Yu, S.A. Anderson, I. Ghattas, L. Puelles, J.L.R. Rubenstein, Dlx-1, Dlx-2, and Dlx-5 expression define distinct stages of basal forebrain differentiation, J. Comp. Neurol. 414 (1999) 217–237. M. Qui, A. Bulfone, S. Martinez, J.J. Meneses, K. Shimamura, R.A. Pedersen, J.L.R. Rubenstein, Null mutations of Dlx-2 results in abnormal morphogenesis of proximal first and second branchial arch derivatives and abnormal differentiation in the forebrain, Genes Dev. 9 (1995) 2523–2538. M. Qui, A. Bulfone, I. Ghattas, J.J. Meneses, L. Christensen, P.T. Sharpe, R. Presley, R.A. Pedersen, J.L.R. Rubenstein, Role of the Dlx homeobox genes in proximodistal patterning of the branchial arches: mutations of Dlx-1, Dlx-2, and Dlx-1 and Dlx-2 alter morphogenesis of proximal skeletal and soft tissue structures derived from the first and second arches, Dev. Biol. 185 (1997) 165–184. A. Simeone, D. Acampora, M. Pannese, M. DÕEsposito, A. Stornaiuolo, M. Gulisano, A. Mallamaci, K. Katsury, T. Druck, K. Huebner, E. Boncinelli, Cloning and characterization of two members of the vertebrate Dlx gene family, Proc. Natl. Acad. Sci. USA 91 (1994) 2250–2254. A. Simeone, M. Gulisano, D. Acampora, A. Stornaiuolo, M. Rambaldi, E. Boncinelli, Two vertebrate homeobox genes related to the Drosophila empty spiracles gene are expressed in the embryonic cerebral cortex, EMBO J. 11 (1992) 2541–2550. M. Gulisano, V. Broccoli, C. Pardini, E. Boncinelli, Emxl and Emx2 show different patterns of expression during proliferation and differentiation of the developing cerebral cortex, Eur. J. Neurosci. 8 (1996) 1037–1050. A. Mallamaci, S. Mercurio, L. Muzio, C. Cecchi, C.L. Pardini, P. Gruss, E. Boncinelli, The lack of Emx2 causes absence of Cajal– Retzius cells and reeler-like defects of neuronal migration in the developing cerebral cortex, J. Neurosci. 20 (2000) 1109–1118. M. Ogawa, T. Miyata, K. Nakajima, K. Yagyu, M. Seike, K. Ikenaka, H. Yamamoto, K. Mikoshimba, The reeler geneassociated antigen on Cajal–Retzius neurons is a crucial molecule for laminar organization of cortical neurons, Neuron 14 (1995) 899–912. M. Pellegrini, A. Mansouri, A. Simeone, E. Boncinelli, P. Gruss, Dentate gyrus formation requires Emx2, Development 122 (1996) 3893–3898. M. Yoshida, Y. Suda, I. Matsuo, N. Miyamoto, N. Takeda, S. Kuratani, S. Aizawa, Emxl and Emx2 functions in development of dorsal telencephalon, Development 124 (1997) 101–111. A. Faiella, S. Brunelli, T. Granata, L. DÕIncerti, R. Cardini, C. Lenti, G. Battaglia, E. Boncinelli, A number of schizencephalic patients including 2 brothers are heterozygous for germline mutations in the homeobox gene EMX2, Eur. J. Hum. Genet. 5 (1997) 186–190. T. Granata, G. Battaglia, L. DÕIncerti, S. Franceschetti, R. Spreafico, M. Savoiardo, G. Avanzini, Schizencephaly: clinical findings, in: R. Guerrini, F. Andermann, R. Canapicchi, J. Roger, B.J. Zifkin, P. Pfanner (Eds.), Dysplasias of Cerebral Cortex and Epilepsy, Lippincott-Raven, Philadelphia, 1996.
188
R. Ohira et al. / Molecular Genetics and Metabolism 77 (2002) 179–188
[30] E. Boncinelli, M. Gulisano, V. Broccoli, Emx and Otx genes in the developing mouse brain, J. Neurobiol. 24 (1993) 1356–1366. [31] A. Simeone, D. Acampora, A. Mallamaci, A. Stornaiuolo, M.R. DÕApice, V. Nigro, E. Boncinelli, A vertebrate gene related to orthodenticle contains a homeodomain of the bicoid class and demarcates anterior neuroectoderm in the gastrulating mouse embryo, EMBO J. 12 (1993) 2735–2747. [32] K.T. Nguyen, Ba-Charvet, Y. Von Boxberg, S. Guazzi, E. Boncinelli, P. Godement, A potential role for the OTX2 homeoprotein in creating early ÔhighwaysÕ for axon extension in the rostral brain, Development 125 (1998) 4273–4282. [33] K. Kitamura, H. Miura, M. Yanazawa, T. Miyashita, K. Kato, Expression patterns of Brx1 (Rieg gene), Sonic hedgehog, Nkx2.2, Dlx1 and Arx during zona limitans intrathalamica and embryonic
[34] [35]
[36]
[37]
ventral lateral geniculate nuclear formation, Mech. Dev. 67 (1997) 83–96. C.R. Scriver, P.J. Waters, Monogenic traits are not simple: lessons from phenylketonuria, Trends Genet. 15 (1999) 267–272. K.M. Dipple, E.R.B. McCabe, Phenotypes of patients with ‘‘simple’’ Mendelian disorders are complex traits: thresholds, modifiers and systems dynamics, Am. J. Hum. Genet. 66 (2000) 1729–1735. K.M. Dipple, E.R.B. McCabe, Modifier genes convert ‘‘simple’’ Mendelian disorders to complex traits, Mol. Genet. Metab. 71 (2000) 43–50. K.M. Dipple, J.K. Phelan, E.R.B. McCabe, Consequences of complexity within biological networks: robustness and health, or vulnerability and disease, Mol. Genet. Metab. 74 (2001) 45–50.
Human ARX gene: genomic characterization and expression R. Ohira,a Y.-H. Zhang,b W. Guo,b K. Dipple,b S.L. Shih,b J. Doerr,c B.-L. Huang,b L.J. Fu,d A. Abu-Khalil,d D. Geschwind,d and E.R.B. McCabea,b,* a
Department of Human Genetics, David Geffen School of Medicine at UCLA, 10833 Le Conte Avenue, Los Angeles, CA 90095-1752, USA b Department of Pediatrics, David Geffen School of Medicine at UCLA, Los Angeles, CA 90095-1752, USA c Departments of Microbiology and Molecular Genetics, David Geffen School of Medicine at UCLA, Los Angeles, CA 90095-1752, USA d Department of Neurology and Program in Neurogenetics, David Geffen School of Medicine at UCLA, Los Angeles, CA 90095-1752, USA Received 28 June 2002; accepted 3 July 2002
Abstract Arx is a homeobox-containing gene with a high degree of sequence similarity between mouse and zebrafish. Arx is expressed in the forebrain and floor plate of the developing central nervous systems of these vertebrates and in the presumptive cortex of fetal mice. Our goal was to identify genes in Xp22.1–p21.3 involved in human neuronal development. Our in silico search for candidate genes noted that annotation of a human Xp22 PAC (RPCI1-258N20) sequence (GenBank Accession No. AC002504) identified putative exons consistent with an Arx homologue in Xp22. Northern blot analysis showed that a 3.3 kb human ARX transcript was expressed at high levels in fetal brain. A 5.9 kb transcript was expressed in adult heart, skeletal muscle, and liver with very faint expression in other adult tissues, including brain. In situ hybridization of ARX in human fetal brain sections at various developmental stages showed the highest expression in neuronal precursors in the germinal matrix of the ganglionic eminence and in the ventricular zone of the telencephalon. Expression was also observed in the hippocampus, cingulate, subventricular zone, cortical plate, caudate nucleus, and putamen. The expression pattern suggests that ARX is involved in the differentiation and maintenance of specific neuronal cell types in the human central nervous system. We also mapped the murine Arx gene to the mouse genome using a mouse/hamster radiation hybrid panel and showed that Arx and ARX are orthologues. Therefore, investigations in model vertebrates may provide insight into the role of ARX in development. The recent identification of ARX mutations in patients with various forms of menta1 retardation make such studies in model organisms even more compelling. Ó 2002 Elsevier Science (USA). All rights reserved. Keywords: Aristaless-Related Homeobox (ARX) gene, human; CNS developmental gene candidate, ARX; Developmental expression, human, ARX; Fetal CNS development, ARX; Homeobox gene; In situ hybridization; Transcription factor, ARX
1. Introduction Vertebrate brain development is a complex process that entails various steps, including cell identity determination, cell migration, and regional differentiation. Transcription factors are expressed in a space- and timedependent manner creating distinct regions, which eventually develop into various parts of the brain including forebrain, midbrain, hindbrain, cerebellum, and spinal cord [1]. Regionalization involves the interplay of genetic and environmental factors and a number of transcription factors have been implicated in prosencephalic regional*
Corresponding author. Fax: 1-310-206-4584. E-mail address: [email protected] (E.R.B. McCabe).
ization [2]. One way to study regional specification in the brain has been to observe regional and temporal expression of transcription factors and signaling molecules during development. Identifying the organizing center of a particular region of the brain is important in establishing a basis for understanding neuronal development. The midbrain organizer, the isthmus, has been identified and characterized as the spinal cord organizing center, the floor plate [3]. The forebrain organizer region remains to be identified, but it has been argued that once discovered, it will elucidate the roles of each of the key factors involved in forebrain development [4]. Aristaless Related Homeobox (ARX) is a member of the group II aristaless-related protein family, members of which are expressed primarily in the central and/or
1096-7192/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved. PII: S 1 0 9 6 - 7 1 9 2 ( 0 2 ) 0 0 1 2 6 - 9
180
R. Ohira et al. / Molecular Genetics and Metabolism 77 (2002) 179–188
peripheral nervous system [5]. It contains two conserved domains: an aristaless domain also known as a C-peptide and a homeobox domain belonging to the prd-like class. ARX has been previously shown to have a high degree of sequence similarity between mouse and zebrafish [4]. Arx is expressed in the forebrain and floor plate of the developing central nervous system (CNS) in each of these vertebrates and in the presumptive cortex and hippocampus of fetal mice [4,6]. The genomic localization of Arx remained undefined when we initiated these investigations, though subsequent reports placed it in Xp22 [7,8]. Mutations in ARX have been reported recently among patients with various forms of mental retardation [6,9]. Our goal was to identify and characterize genes in Xp22.1–p21.3 that might be involved in human CNS development and deleted in patients with contiguous gene syndromes involving Xp22.1–p21.3 [10]. An in silico search revealed that putative exons consistent with an Arx homologue had been identified in Xp22, within the chromosomal region of interest. Northern blot analysis and in situ hybridization were performed to characterize the expression patterns of human ARX. The murine Arx gene was mapped to the mouse X chromosome in a region syntenic with Xp22.
2. Materials and methods 2.1. In silico identification of ARX An in silico search for candidate CNS developmental genes was conducted to identify genes within the desired, defined region of the X chromosome, Xp22.1–p21.3. Exons putatively defined as ARX (GenBank Accession No. AC002504) had been identified by the Baylor Genome Center (http://www.hgsc.bcm.tmc.edu/). 2.2. Sequencing of ARX cDNA Primer pairs, A1 50 gcaagttcagcctggttaag30 (50 end of vector), A2 50 tacgcgcatacctggtgaagacgtccgggtag30 (nucleotides 1052-IVS2+11) and B1 50 ttatgagtatttcttccag gg30 (30 end of vector), B2 50 aagaagatgaggacgaggaaga Table 1 Primers used for sequencing of ARX Primer 1 2 3 4 5 6
nt 0
0
5 CCA ACA CAC ACC CAT CCA TC3 50 CTT GTT ACC GCT TGT CCT GAG30 50 GGA CGA AGA AGA TGA GGA CGA GGA AGA GGA AC30 50 TGA AAG CTG GGT GTC GGA ACA C30 50 TTC CCG AGC CTA CCT CCG CCT C30 50 GCT CCT GAG ACA GCC CAC AC30
)81 to )62 )222 to )202 695 to 726 1403 to 1428 1320 to 1339 1496 to 1515
ggaac30 (nucleotides 700–726), were specifically designed to obtain overlapping PCR fragments from a human fetal brain cDNA library (Human Brain, 50 -stretch plus cDNA library, Cat. No. HL3008a; Clontech Laboratories, Palo Alto, CA) that covered the putative coding sequence. The amplified cDNA fragments were used as probes to screen the human fetal brain library and the cDNA clone was sequenced by automated cycle sequencing (ABI377, Perkin–Elmer, Foster City, CA) using the primers in Table 1. 2.3. Characterization of ARX expression Northern blot analysis was performed using adult and fetal human multiple tissue northern (MTN) blots (Clontech) containing poly (A)þ RNA. The probe used on both blots was a 673 bp fragment from exon 2 (nucleotides 400–1073). The fragment was first amplified by PCR (primer A2 (see above) and 50 aagaggccaaggcgtcga agtc30 (IVS1–150 to IVS1–129)) from genomic DNA and digested with EcoRI and NotI (to obtain an optimal length), then probes were labeled with [a-32 P]dCTP using the Prime-It II Random Primer Labeling Kit and purified using NUC-TRAP (Stratagene, La Jolla, CA). The blots were first incubated in ExpressHyb Solution (Clontech) at 68 °C for 30 min and then in the same solution containing the labeled ARX probe (107 cpm) at 68 °C for 1 h. After incubation with the probe, the blots were washed twice in Solution 1 (2 SSC and 0.05% SDS) at 68 °C for 10 min and twice in Solution 2 (0.1 SSC and 0.1% SDS) at 70 °C for 5 min. Autoradiographs were obtained after exposing the X-OMAT scientific imaging film (Kodak, Rochester, NY) for 1–2 days at )80 °C. Human fetal tissue (15–22 gestational weeks) obtained from the Brain and Tissue Bank for Developmental Disorders at the University of Maryland, Baltimore, Maryland, was cryoprotected and then sectioned into 20 lm coronal sections that were thawmounted onto glass slides. The sections were fixed in 4% paraformaldehyde for 20 min followed by a 10-min wash in 0.1 M phosphate buffer (PB) (0.0774 M Na2 PO4 , 0.0226 M NaH2 PO4 , pH 7.4). The ARX insert was obtained by similar methods as the Northern probe and the EcoRI and NotI fragment was subcloned into pBluescript. Anti-sense and sense RNA riboprobes were transcribed in vitro from a linearized plasmid containing the ARX insert (673 bp region of exon 2) using an RNA Transcription Kit (Stratagene) and [35 S]UTP (Amersham, Arlington Heights, IL). The tissue sections were hybridized with either the sense or anti-sense radioactive riboprobe (107 cpm) and were incubated overnight at 60 °C. Following the hybridizations, the slides were treated with RNase A to remove unhybridized probe and then washed in descending concentrations of SSC in the presence of sodium thiosulfate. Autoradiographic
R. Ohira et al. / Molecular Genetics and Metabolism 77 (2002) 179–188
film (Amersham) was placed on the slides and allowed to develop for five days. The slides were dipped in NTB2 emulsion (Kodak) and stored in light tight boxes for 4–6 weeks. The slides were then developed and counterstained with cresyl violet and were examined by both bright and dark field microscopy. 2.4. Murine Arx mapping A mouse/hamster radiation hybrid panel (Research Genetics, Huntsville, AL) was used to map the murine Arx. The mouse Arx (GenBank Accession No. NM_007492) sequence was compared to hamster Arx (GenBank Accession No. AY038070) and human ARX (GenBank Accession No. AY038071) to find a unique region in the mouse sequence. The human ARX sequence was used as an additional species comparison to verify the unique murine sequence. Primers 50 gcgcgtcaa gcatagccgcg30 (nucleotides 1678–1697) and 50 gtggactttga gttggagtg30 (nucleotides 1936–1955)) were designed to this unique region and PCR was performed. The results were scored using the Mouse Genome Database (The Jackson Laboratory, Bar Harbor, ME) and the LOD Score Method (www.informatics.jax.org).
181
Table 2 Intron–exon boundaries for human ARX Intron
Exon
Intron
50 gccccagcc ttgcccgcag tttcttatag tgccttgcag ctccctgcag
ATGAG 1 GCAAG GCTCC 2 ACCAG GGAGG 3 TCCAG GTCTG 4 GGCAG GCTCT 5 TGTGC
gtaaggatgc gtatgcgcgt gtgagctgca gtaacgcgca taaaggctgc30
Sequencing of the human fetal brain ARX cDNA confirmed the putative exons identified in the genomic DNA (Table 2, Fig. 1). The ARX cDNA consists of five exons spanning a region of 1689 bp and coding for a deduced 562 amino acid protein. Table 2 shows the intron–exon boundaries for each of the five exons. The intron–exon boundaries were identified by comparison of the genomic and cDNA sequences. Comparison of the human ARX and murine Arx genes revealed 89% identity at the nucleotide level and 95% identity at the amino acid level. Fig. 2 shows the nucleotide and amino acid sequence alignment of both ARX and Arx.
(Figs. 3 and 4). Hybridization of the ARX exon 2 probe with the fetal MTN blot (Fig. 3) showed a 3.3 kb transcript at very high levels in fetal brain compared with fetal kidney, liver, and lung, where hybridization was minimal or absent. Hybridization with RNA from adult tissues showed a signal for a 5.9 kb transcript in heart, skeletal muscle, and liver (Fig. 4). A faint signal was seen in the other adult tissues (brain, thymus, spleen, kidney, placenta, lung, and blood leukocytes). Both MTN blots were probed for b-actin as a control and all lanes were loaded equally (data not shown). Human fetal brain sections at three different gestational weeks (gw) of development were examined for ARX expression. In situ hybridization results showed a pattern consistent with expression in neuronal precursor cells of the germinal zones. At 15 gw (Figs. 5A and 6A–D), intense signal was seen in the germinal matrix (gm) of the ganglionic eminence, and ventricular zone (vz) of the telencephalon, with less signal in the subventricular zone (svz), cortical plate (cp), caudate nucleus (cn), and putamen (p). By 19 gw (Fig. 5B), intense signal was still observed in the gm and vz, with lower signal in the svz, cp, cn, p, globus pallidus (gp), and substantia nigra (sn). At 22 gw (Figs. 5C and 6E– H), the ganglionic eminence (still showing high levels of expression) was observed to have expanded within the developing brain. Strong signals were still seen in the vz as well. Less intense signal was seen in the svz, cp, the cingulate (ci), and the hippocampus (h), particularly in the dentate gyrus (dg) and entorhinal cortex (E).
3.2. Characterization of ARX expression
3.3. Murine Arx mapping
Northern blot analysis showed that ARX had much higher expression in human fetal brain than adult brain
The radiation hybrid panel allowed mapping of the murine Arx gene to the mouse genome. Arx
3. Results 3.1. Sequencing of ARX cDNA
Fig. 1. Genomic structure of the human ARX gene. The numbered boxes indicate exons. The lines indicate introns. The numbers above the exons and below the introns indicate the sizes in basepairs (bp) (GenBank Accession No. AC002504). The coding region for ARX is 1689 bp (562 amino acids in deduced sequence).
182
R. Ohira et al. / Molecular Genetics and Metabolism 77 (2002) 179–188
Fig. 2. Nucleotide and amino acid comparison of human ARX (above) and murine Arx (below). Asterisks indicate regions of identity and hyphens indicate gaps in the gene sequence.
mapped to the murine X chromosome in the region containing Dmd, Gyk, and Zfx. The data showed the highest anchor LOD (14.2) to DXMit6, then LOD (12.6) to DXMit63 and finally LOD (11.7) to DXMit112. The Arx locus of interest was determined to be between DXMit112, 78.9 cR proximal and DXMit6.7, 7.04 cR distal, in the region syntenic with Xp22–p21.
4. Discussion An in silico search for candidate CNS developmental genes revealed an annotation of a human Xp22 PAC (RPCI1-258N20) sequence deposited by the Baylor Human Genome Center (GenBank Accession No. AC002504), which identified putative exons, suggesting an Arx homologue in Xp22, within the chromosomal
R. Ohira et al. / Molecular Genetics and Metabolism 77 (2002) 179–188
183
Fig. 2. (continued)
region of interest. The human ARX appeared to have a much higher expression in fetal brain than adult brain and increased expression in adult heart, skeletal muscle, and liver. There was very high expression of the 3.3 kb transcript in fetal brain. In situ hybridizations of human
fetal brain (15–22 gw) sections showed a pattern consistent with expression of ARX in neuronal precursor cells. An intense signal was seen in the germinal matrix and ventricular zone at all stages of development, while weaker signals were seen in the subventricular zone,
184
R. Ohira et al. / Molecular Genetics and Metabolism 77 (2002) 179–188
Fig. 3. Human fetal MTN blot probed for ARX (region of exon 2). A 3.3 kb transcript was seen in human fetal brain.
Fig. 4. Human adult MTN blot probed for ARX (fragment from exon 2). A 5.9 kb fragment is seen in the heart, skeletal muscle, and liver, but at this exposure minimal if any signal is observed in brain.
cortical plate, caudate nucleus, putamen, globus pallidus, substantia nigra (19 gw), cingulate, and hippocampus (22 gw). The murine Arx gene was mapped to the mouse genome to determine whether Arx and ARX are orthologues, or paralogues and simply members of the same gene family. Arx mapped to the murine X chromosome in the region containing Dmd, Gyk, and Zfx, Arx therefore is syntenic with the ARX locus, and Arx and ARX are orthologues. The expression patterns of Arx in mouse and ARX in human brain sections are remarkably similar [4,6]. The expression in the hippocampus is particularly interesting with the recent reports of ARX mutations in patients with mental retardation [6,9]. Although expression in the fetal MTN blots is similar in our hands and those of Bienvenu [6], all three groups differ in the expression patterns observed in the adult Northern blots [6,9]. These differences may be due to the fact that each group used a different probe, and we have observed differences in hybridization to MTN blots depending on the probe that we used (unpublished observations). ARX is a member of the group-II aristaless rented homeobox proteins involved in CNS development, containing two highly conserved domains, the C-peptide, and the prd-like class homeobox domain [5]. The
exact function of ARX is unknown, but the C-peptide domain is highly conserved with a similar sequence and position to the Otp (orthopedia) homeoprotein, a member of this protein family thought to be important in regionalization of the ventral diencephalon [11]. Deletion of the C-peptide in Otp reduces transactivation activity [11] and this domain may serve a similar function in Arx and ARX [4]. ARX is one of many homeodomain genes expressed during development. The common link between these genes is the highly conserved gene sequence called the homeobox, which encodes a protein domain of approximately 60 amino acids [12]. The homeodomain proteins act as nuclear transcriptional regulators by activating or repressing eukaryotic transcription. Many of these transcription factors contain various domains with functions such as DNA binding, dimerization, and transcriptional activation/repression [13]. There are several classes of homeobox genes with specific roles in CNS development: Dlx, Emx, and Otx. The Dlx genes are primarily involved in basal forebrain ganglia differentiation and are required for craniofacial development [14–21]. The Emx genes have been shown to play a critical role in establishing regions of forebrain and midbrain [22]. Recent data suggest a role for Emx proteins in neuroblast proliferation [22,23], and migration [24,25], as well as cognitive processes [23,26,27]. Mutations in Emx2 have been found in individuals with schizencephaly, a rare cell migration disorder, which causes full thickness clefting of the cerebral hemispheres [28,29]. Otx genes most likely play a crucial role in establishing the identity of various regions of the developing brain [30,31]. Specifically, Otx may be involved in fasciculating and guiding nerve fibers [32]. Arx was previously shown to be expressed in the forebrain and floor plate of the developing CNS of mouse and zebrafish and in the presumptive cortex and hippocampus of fetal mice [4,6]. In the forebrain, Arx is expressed in the intermediate and mantle zones, but in the presumptive cerebral cortex, expression is seen in the ventricular zone suggesting different functions of Arx in the two regions. Within the zona limitans intrathalamica (ZLI) of the mouse, a region within the diencephalon, Arx is expressed primarily in the ventral thalamus, ganglionic eminence, and floor plate at E10-11 d.p.c., similar to expression patterns for Dlx1 [33]. Comparisons of Arx and Brx1 expression show that the dorsalmost expression of Arx overlaps with Brx1. However, Arx expression appears to have a broad path containing the spindle-shaped cells while the Brx1 expression shows a narrow path on the dorsal side of the Arx path. Examination of three coronal levels of the ventral lateral geniculate nucleus indicate Arx expression only in level 2. However, expression of Arx and Dlx1 is seen in the external medullary lamina of levels 1 and 2 [33]. The expression patterns of Arx suggest a role in the forma-
R. Ohira et al. / Molecular Genetics and Metabolism 77 (2002) 179–188
185
Fig. 5. In situ hybridization of coronal sections from human fetal brain at 15 gestational weeks (gw) (A), 19 gw (B), and 22 gw (C) probed with an ARX probe. Strong signals are seen in the ventricular zone (vz) of the telencephalon and the germinal matrix (gm) of the ganglionic eminence. Less intense signals are seen in the cortical plate (cp), caudate nucleus (cn), putamen (p), globus pallidus (gp), subventricular zone (svz), substantia nigra (sn), cingulate (ci), and the hippocampus (h). No signal is seen in the internal capsule (ic).
tion of the reticular thalamic nucleus as well as the ventral lateral geniculate nucleus. Both human ARX and murine Arx were shown to be expressed in the hippocampus and human ARX was shown to be expressed in the cingulate, which suggest a role in cognitive processes and memory. Two recent reports have mapped Arx to the region syntenic to the human ARX [7,8]. Our results confirm that Arx maps to the X chromosome near Dmd and Zfx, which makes it syntenic to the human ARX locus. Therefore, Arx and ARX are in fact orthologues. Sequence comparison of the two genes showed 89% identity at the nucleotide level and 95% identity at the amino acid level. The close similarity of these two genes allows for their investigation in model vertebrates, which may provide valuable information regarding ARX expression and the implications of the disruption of its expression.
Of particular interest in studies of ARX orthologues in model organisms will be investigations to identify potential modifier genes. The phenotypes of patients with ARX mutations are highly variable, including mental retardation (syndromic and nonspecific), epilepsy (infantile spasms and myoclonic seizures), and dystonia [6,9]. Among the 19 families with ARX mutations, 14 had alanine-repeat expansions and 11 of the 14 had the identical polyalanine expansion (EX2_428_451dup) from 12 to 20 alanines in length. The other mutations included four missense mutations (L33P, Q163R, G286S, and P353L), and a frameshift mutation (IVS4_EX5del1517). It is interesting to note that 11 of the 19 (58%) of the families identified with ARX mutations had the identical polyalanine expansion. Among individuals with the identical ARX genotype (e.g., EX2_428_451dup), there was considerable phenotypic variability. The variability in phenotype shows
186
R. Ohira et al. / Molecular Genetics and Metabolism 77 (2002) 179–188
Fig. 6. High magnification of coronal sections from human fetal brain at 15 gw (A)–(D) and 22 gw (E)–(H) in bright (A), (C), (E), and (G) and darkfield (B), (D), (F), and (H) views. The basal ganglia of a 15 gw human fetal brain section shows intense staining in the germinal matrix (gm), and ventricular zone (vz), lighter staining in the cortical plate (cp), caudate nucleus (cn), and putamen (p), and no staining in the internal capsule (ic) (A)– (D). Staining of the vz and gm of a 22 gw brain section are observed at high magnification (E) and (F). The high magnification of the 22 gw hippocampus is shown (G) and (H). Less intense signal is seen in the dentate gyrus (dg) and the entorhinal cortex (E), (G), and (H). Bar equals 150 lm.
the complexity of this ‘‘single gene disorder,’’ a feature common to many ‘‘simple’’ Mendelian genetic diseases [34–37]. The complexity of proteomic networks and the influences of modifier genes on these networks are sure to contribute to the lack of genotype–phenotype corre-
lation. Investigations in model organisms and patients with ARX mutations may help us to identify and characterize genes modifying phenotypic expression following disruption of ARX, a gene critical to central nervous system development.
R. Ohira et al. / Molecular Genetics and Metabolism 77 (2002) 179–188
Acknowledgments These investigations were supported in part by grants from the NIH (RO1 HD22563, P3O HD34610, and MH60233; DHG). We thank Hooman Allayee from the Department of Human Genetics at the David Geffin School of Medicine at UCLA for assisting us with murine mapping.
[15]
References
[18]
[1] M. Hatten, Central nervous system neuronal migralion, Annu. Rev. Neurosci. 22 (1999) 511–539. [2] K. Shimamura, J. Rubenstein, Inductive interactions direct early regionalization of the mouse forebrain, Development 124 (1997) 2709–2718. [3] S.L. Ang, The brain organization, Nature 380 (1996) 25–27. [4] H. Miura, M. Yanazawa, K. Kato, K. Kitamura, Expression of a novel aristaless related homeobox gene ÔArxÕ in the vertebrate telencephalon, diencephalon and floor plate, Mech. Dev. 65 (1997) 99–109. [5] F. Meijlink, A. Veverdam, A. Brouwer, T. Oosterveen, D. Berge, Vertebrate aristaless-related genes, Int. J. Dev. Biol. 43 (1999) 651–663. [6] T. Bienvenu, K. Poireier, G. Friocourt, N. Bahi, D. Beaumont, F. Fauchereau, L.B. Jeema, R. Zemni, M. Vinet, F. Francis, P. Couvert, M. Gomot, C. Moraine, H. van Bokhoven, V. Kalscheurer, S. Frints, J. Gecz, K. Ohzaki, H. Chaabouni, J. Fryns, V. Desportes, C. Beldjord, J. Chelly, ARX, a novel Prd-classhomeobox gene highly expressed in the telencephalon, is mutated in X-linked mental retardation, Hum. Mol. Genet. 11 (2002) 981– 991. [7] H. Blair, V. Reed, E. Gormally, J. Wilson, J. Novak, R. McInnes, S. Phillips, B. Taylor, Y. Boyd, Positioning of five genes (CASK, ARX, SAT, IMAGE cDNAs 248928 and 253949) from the human X chromosome short arm with respect to evolutionary breakpoints on the mouse X chromosome, Mamm. Genome 11 (2000) 710–712. [8] D. Cunningham, Q. Xiao, A. Chatterjee, K. Sulik, D. Juriloff, F. Elder, W. Harrison, G. Schuster, P.A. Overbeek, G.E. Herman, exma: an X-linked insertional mutation that disrupts forebrain and eye development, Mamm. Genome 13 (2002) 179–185. [9] P. Stromme, M.E. Mangelsdorf, M.A. Shaw, K.M. Lower, S.M.E. Lewis, H. Bruyere, V. Lutcherath, A.K. Gedeon, R.H. Wallace, I.E. Scheffer, G. Turner, M. Partington, S.G.M. Frints, J.P. Fryns, G.R. Sutherland, J.C. Mulley, J. Gecz, Mutations in the human ortholog of Aristaless cause X-linked mental retardation and epilepsy, Nat. Genet. 30 (2002) 441–445. [10] E.R.B. McCabe, Disorders of glycerol metabolism, in: C.R. Scriver, A.L. Beaudet, W.S. Sly, D. Valle (Eds.), The Metabolic & Molecular Bases of Inherited Diseases, eighth ed., McGraw-Hill, New York, 2001, pp. 2217–2237. [11] A. Simeone, M.R. Dapice, V. Nigro, J. Casanova, F. Graziani, D. Acampora, V. Avantaggiato, Orthopedia, a novel homeoboxcontaining gene expressed in the developing CNS of both mouse and Drosophila, Neuron 13 (1994) 83–101. [12] W.J. Gehring, M. Affolter, T. Burglin, Homeodomain proteins, Annu. Rev. Biochem. 63 (1994) 487–526. [13] J. Vollmer, R. Clerc, Homeobox genes in the developing mouse brain, J. Neurosci. 71 (1998) 1–19. [14] S.A. Anderson, M. Qiu, A. Bulfone, D.D. Eisenstat, J. Meneses, R. Pederson, J.L.R. Rubenstein, Mutations of the homeobox genes Dlx-1 and Dlx-2 disrupt the striatal subventricular zone and
[16]
[17]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
187
differentiation of late born striatal neurons, Neuron 19 (1997) 27– 37. S.A. Anderson, D.D. Eisenstat, L. Shi, J.L.R. Rubenstein, Interneuron migration from basal forebrain to neocortex; dependence on Dlx genes, Science 278 (1997) 474–476. A. Bulfone, H.J. Kim, L. Puelles, M.H. Porteus, J.F. Grippo, J.L.R. Rubenstein, The mouse Dlx-2 (Tes-1) gene is expressed in spatially restricted domains of the forebrain, face and limbs in midgestation mouse embryos, Mech. Dev. 40 (1993) 129–140. M.J. Depew, J.K. Liu, J.E. Long, R. Presley, J.J. Meneses, R.A. Pedersen, J.L.R. Rubenstein, Dlx5 regulates regional development of the branchial arches and sensory capsules, Development 126 (1999) 3831–3846. D.D. Eisenstat, J.K. Liu, M. Mione, W. Zhong, G. Yu, S.A. Anderson, I. Ghattas, L. Puelles, J.L.R. Rubenstein, Dlx-1, Dlx-2, and Dlx-5 expression define distinct stages of basal forebrain differentiation, J. Comp. Neurol. 414 (1999) 217–237. M. Qui, A. Bulfone, S. Martinez, J.J. Meneses, K. Shimamura, R.A. Pedersen, J.L.R. Rubenstein, Null mutations of Dlx-2 results in abnormal morphogenesis of proximal first and second branchial arch derivatives and abnormal differentiation in the forebrain, Genes Dev. 9 (1995) 2523–2538. M. Qui, A. Bulfone, I. Ghattas, J.J. Meneses, L. Christensen, P.T. Sharpe, R. Presley, R.A. Pedersen, J.L.R. Rubenstein, Role of the Dlx homeobox genes in proximodistal patterning of the branchial arches: mutations of Dlx-1, Dlx-2, and Dlx-1 and Dlx-2 alter morphogenesis of proximal skeletal and soft tissue structures derived from the first and second arches, Dev. Biol. 185 (1997) 165–184. A. Simeone, D. Acampora, M. Pannese, M. DÕEsposito, A. Stornaiuolo, M. Gulisano, A. Mallamaci, K. Katsury, T. Druck, K. Huebner, E. Boncinelli, Cloning and characterization of two members of the vertebrate Dlx gene family, Proc. Natl. Acad. Sci. USA 91 (1994) 2250–2254. A. Simeone, M. Gulisano, D. Acampora, A. Stornaiuolo, M. Rambaldi, E. Boncinelli, Two vertebrate homeobox genes related to the Drosophila empty spiracles gene are expressed in the embryonic cerebral cortex, EMBO J. 11 (1992) 2541–2550. M. Gulisano, V. Broccoli, C. Pardini, E. Boncinelli, Emxl and Emx2 show different patterns of expression during proliferation and differentiation of the developing cerebral cortex, Eur. J. Neurosci. 8 (1996) 1037–1050. A. Mallamaci, S. Mercurio, L. Muzio, C. Cecchi, C.L. Pardini, P. Gruss, E. Boncinelli, The lack of Emx2 causes absence of Cajal– Retzius cells and reeler-like defects of neuronal migration in the developing cerebral cortex, J. Neurosci. 20 (2000) 1109–1118. M. Ogawa, T. Miyata, K. Nakajima, K. Yagyu, M. Seike, K. Ikenaka, H. Yamamoto, K. Mikoshimba, The reeler geneassociated antigen on Cajal–Retzius neurons is a crucial molecule for laminar organization of cortical neurons, Neuron 14 (1995) 899–912. M. Pellegrini, A. Mansouri, A. Simeone, E. Boncinelli, P. Gruss, Dentate gyrus formation requires Emx2, Development 122 (1996) 3893–3898. M. Yoshida, Y. Suda, I. Matsuo, N. Miyamoto, N. Takeda, S. Kuratani, S. Aizawa, Emxl and Emx2 functions in development of dorsal telencephalon, Development 124 (1997) 101–111. A. Faiella, S. Brunelli, T. Granata, L. DÕIncerti, R. Cardini, C. Lenti, G. Battaglia, E. Boncinelli, A number of schizencephalic patients including 2 brothers are heterozygous for germline mutations in the homeobox gene EMX2, Eur. J. Hum. Genet. 5 (1997) 186–190. T. Granata, G. Battaglia, L. DÕIncerti, S. Franceschetti, R. Spreafico, M. Savoiardo, G. Avanzini, Schizencephaly: clinical findings, in: R. Guerrini, F. Andermann, R. Canapicchi, J. Roger, B.J. Zifkin, P. Pfanner (Eds.), Dysplasias of Cerebral Cortex and Epilepsy, Lippincott-Raven, Philadelphia, 1996.
188
R. Ohira et al. / Molecular Genetics and Metabolism 77 (2002) 179–188
[30] E. Boncinelli, M. Gulisano, V. Broccoli, Emx and Otx genes in the developing mouse brain, J. Neurobiol. 24 (1993) 1356–1366. [31] A. Simeone, D. Acampora, A. Mallamaci, A. Stornaiuolo, M.R. DÕApice, V. Nigro, E. Boncinelli, A vertebrate gene related to orthodenticle contains a homeodomain of the bicoid class and demarcates anterior neuroectoderm in the gastrulating mouse embryo, EMBO J. 12 (1993) 2735–2747. [32] K.T. Nguyen, Ba-Charvet, Y. Von Boxberg, S. Guazzi, E. Boncinelli, P. Godement, A potential role for the OTX2 homeoprotein in creating early ÔhighwaysÕ for axon extension in the rostral brain, Development 125 (1998) 4273–4282. [33] K. Kitamura, H. Miura, M. Yanazawa, T. Miyashita, K. Kato, Expression patterns of Brx1 (Rieg gene), Sonic hedgehog, Nkx2.2, Dlx1 and Arx during zona limitans intrathalamica and embryonic
[34] [35]
[36]
[37]
ventral lateral geniculate nuclear formation, Mech. Dev. 67 (1997) 83–96. C.R. Scriver, P.J. Waters, Monogenic traits are not simple: lessons from phenylketonuria, Trends Genet. 15 (1999) 267–272. K.M. Dipple, E.R.B. McCabe, Phenotypes of patients with ‘‘simple’’ Mendelian disorders are complex traits: thresholds, modifiers and systems dynamics, Am. J. Hum. Genet. 66 (2000) 1729–1735. K.M. Dipple, E.R.B. McCabe, Modifier genes convert ‘‘simple’’ Mendelian disorders to complex traits, Mol. Genet. Metab. 71 (2000) 43–50. K.M. Dipple, J.K. Phelan, E.R.B. McCabe, Consequences of complexity within biological networks: robustness and health, or vulnerability and disease, Mol. Genet. Metab. 74 (2001) 45–50.