Human Molecular Genetics, 2004, Vol. 13, No. 10 DOI: 10.1093/hmg/ddh117 Advance Access published on March 17, 2004
1025–1040
QRX, a novel homeobox gene, modulates photoreceptor gene expression Qing-liang Wang1,2, Shiming Chen5, Noriko Esumi2, Prabodh K. Swain6,{, Heidi S. Haines8, Guanghua Peng5, B. Michele Melia2, Iain McIntosh1, John R. Heckenlively9, Samuel G. Jacobson10, Edwin M. Stone8, Anand Swaroop6,7 and Donald J. Zack1,2,3,4,* 1
Received January 8, 2004; Revised and Accepted March 10, 2004
A novel paired-like homeobox gene, designated as Qrx, was identified by a yeast one-hybrid screen using the bovine Rhodopsin promoter Ret-1 DNA regulatory element as bait. Qrx is preferentially expressed in both the outer and inner nuclear layers of the retina. Its homeodomain is nearly identical to that of Rx/Rax, a transcription factor that is essential for eye development, but it shares only limited homology elsewhere. Although Qrx and Rx/Rax show similar DNA binding properties in vitro, the two proteins demonstrate distinct target selectivity and functional behavior in promoter activity assays. QRX synergistically increases the transactivating function of the photoreceptor transcription factors Crx and NRL and it physically interacts with CRX. Qrx is present in the bovine and human genomes, but appears to be absent from the mouse genome. Nonetheless, a 5.8 kb upstream region of human QRX is capable of directing expression in presumptive photoreceptor precursor cells in transgenic mice. These results indicate that Qrx may be involved in modulating photoreceptor gene expression. In addition, the finding of rare heterozygous QRX sequence changes in three individuals with retinal degeneration raises the possibility that QRX may be involved in disease pathogenesis.
INTRODUCTION The neuroretina has been studied as a model of neuronal function, development and degeneration (1). Photoreceptor cells, consisting of rods and cones, carry out phototransduction and pass their signals onto bipolar, retinal ganglion cells, and eventually to higher centers within the brain to generate vision. An interesting challenge in developmental neurobiology is deciphering the mechanisms that orchestrate how the many specialized retinal cell types develop and differentiate from a limited number of common progenitor cells (2–4). Retinal development is a function of differential gene regulation, defined by precise temporal and spatial patterns of
gene expression. Significant progress has been made in deciphering some of the complex regulatory network that functions to define the profile of gene expression in particular retinal cells. Among the cis-acting DNA elements involved are Ret-1/PCE-1, Ret-4, BAT-1 and NRE (5–10); the trans-acting transcription factors include CRX, NRL, RX/RAX, PAX6 and CHX10 (8,9,11–17). Rx/Rax, which belongs to the Q50 type paired-like class of homeodomain proteins, is one of the earliest expressed transcription factors in the developing eye, appears to be functionally conserved and is essential for normal eye development within vertebrates (15,18–22). Rx/Rax has also been implicated in determining the fate of Mu¨ller cells (19) and regulating photoreceptor-specific gene expression (23).
*To whom correspondence should be addressed at: Johns Hopkins University School of Medicine, Maumenee Bldg 809, 600 N. Wolfe St, Baltimore, MD 21287, USA. Tel: þ1 4105025230; Fax: þ1 4105025382; Email:
[email protected] { Present address: National Brain Research Center, Gurgaon, India, 122001.
Human Molecular Genetics, Vol. 13, No. 10 # Oxford University Press 2004; all rights reserved
Downloaded from http://hmg.oxfordjournals.org/ by guest on June 7, 2013
Program in Human Genetics and Molecular Biology, McKusick–Nathans Institute of Genetic Medicine, 2Department of Ophthalmology, 3Department of Molecular Biology and Genetics, 4Department of Neuroscience, Johns Hopkins University School of Medicine, Baltimore, MD, USA, 5Department of Ophthalmology and Visual Science, Washington University School of Medicine, St Louis, MO, USA, 6Department of Ophthalmology and Visual Sciences and 7Human Genetics, University of Michigan, Ann Arbor, MI, USA, 8Department of Ophthalmology, University of Iowa College of Medicine, Iowa City, IA, USA, 9Jules Stein Eye Institute, University of California–Los Angeles School of Medicine, Los Angeles, CA, USA and 10Department of Ophthalmology, University of Pennsylvania, Philadelphia, PA, USA
1026
Human Molecular Genetics, 2004, Vol. 13, No. 10
RESULTS Qrx encodes a novel paired-like homeodomain protein A bovine retinal cDNA-Gal4 activation domain fusion library was screened in a yeast one-hybrid assay using a bait that contains the bovine Rhodopsin promoter Ret-1/PCE-1 element (148 to 126 bp, referred to as Ret-1 below). In the primary screen of 2.07 million clones, 72 positives were identified by growth on histidine-deficient medium plates. After plasmid recovery, retransformation, and testing for both growth on histidine-deficient medium plates and positive activity in a bgalactosidase assay, 55 of these clones were confirmed as ‘true’ positives. Partial sequence analysis of these clones revealed that 42 clones matched Chx10 (36), five matched Vsx1 (a Chx10 homologue) (37–39), and eight represented novel sequences. Among the eight, two encoded the same 184 amino acid residue Q50-type paired-like homeodomain containing protein, which we designate as Qrx. In an independent screen using the BAT-1 element as bait (112 to 84 bp), three additional Qrx clones were identified. Clones encoding the human QRX orthologue were obtained by screening a lgt10 human retinal cDNA library with the bovine Qrx cDNA as probe. Human and bovine Qrx are 91% identical at the amino acid level, with 100% conservation in the homeodomain (Fig. 1A). The homeodomain begins at residue 27. A putative monopartite nuclear localization signal (NLS), PKKKHRR, that is homologous to the SV-40 T antigen NLS (CPKKKRKV) (40) is located four amino acids N-terminal to the beginning of the homeodomain. In addition, the Qrx homeodomain contains a sequence (QNRRAKWRRQE) that is homologous to a Crx sequence (KNRRAKCRQQR) that has been reported to be important for its nuclear localization (41).
BLAST analysis (42,43) indicated that the closest homologue of Qrx is Rx/Rax, to which it is 93% identical in the homeodomain. Outside the homeodomain, however, the only other region of significant homology is within the OAR domain, which is part of the C-terminal tail, in which a 20 amino acid stretch shows 60% identity. The OAR domain has been suggested to be involved in both transcriptional activation (44) and repression (45), but its role in Rx/Rax is unknown. Besides Qrx, 14 other Rx/Rax or Rx/Rax-related genes from nine different species have been identified to date. Amino acid sequence comparison and phylogenetic analysis indicate that human and bovine Qrx are the most distantly related members of the family and the most homologous member is the Drosophila Rx/Rax (Fig. 1D). Rx/Rax from human, mouse and rat seem to be more closely related to each other than to other members of the family. Rx1/Rax1 from zebrafish, astyanax and chicken are clustered together with the Rx2 from oryzias and zebrafish, while the chicken Rx/Rax2 (C-Rx2) is most homologous to the xenopus Rx and Rx2. Qrx appears to be deleted from the Mus musculus genome Multiple efforts were made to clone a murine (Mus musculus) Qrx ortholog, including screening of murine (M. musculus) retinal cDNA and genomic libraries with a human QRX cDNA probe, screening an arrayed murine genomic library with a human QRX cDNA and oligomer probes, and a reverse transcription PCR (RT–PCR) approach with several degenerate primers based on areas of conservation between the human and bovine sequences. Although several murine Rx/Rax clones and a number of other sequences were obtained, no clone representing a murine Qrx ortholog was identified, suggesting the possibility that there might not be a murine Qrx. To further address this issue, we pursued both experimental and bioinformatic comparative genomic approaches. Southern blot analysis was carried out with genomic DNA from eight species [human, bovine, M. musculus (C57BL6 and 129SvJ), Mus casteneus, Mus spretus, rat, zebrafish and Drosophila] using a human QRX coding region fragment as the probe. The human QRX probe detected the Qrx gene in both human and bovine as a single band (data not shown). In addition, several weak bands were identified in M. casteneus, M. spretus, zebrafish and Drosophila, but not in M. musculus or rat. Although the identities of these other bands are unknown, they could represent orthologs of QRX or some other homologous gene. However, since the murine Rx/Rax gene was not identified under the hybridization conditions employed, they are unlikely to represent Rx/Rax orthologs. Bioinformatic comparison of the human QRX locus on chromosome 19p13.3 (see mapping data below) to the conserved syntenic region in M. musculus (10C1, 79 kb from the centromere) provided additional evidence to support the absence of a Qrx gene in mouse. The murine region (derived from the publicly available NCBI Human–Mouse Homology Map and the Celera database, which was generated with data from the 129X1/SvJ, DBA/2J and A/J strains, with reported 6-fold coverage and >99% accuracy) shows significant overall conservation with the human region (Fig. 1E) (46–48).
Downloaded from http://hmg.oxfordjournals.org/ by guest on June 7, 2013
Mutations of a number of the transcription factors involved in the proliferation and differentiation of retinal cells can give rise to human retinal disease. As an example, Crx, the cone rod homeobox gene, is responsible for controlling the differentiation of photoreceptor cells by modulating the expression of photoreceptor genes, including Rhodopsin and Crx itself (12,13,24–26). Mutations of CRX have been found to cause cone-rod dystrophy (14,27), Leber congenital amaurosis (28,29) and retinitis pigmentosa (30). Another important factor in the regulation of Rhodopsin and in maintaining normal photoreceptor function is the neural retina leucine zipper protein, Nrl, which acts synergistically with Crx through a direct interaction with Crx’s homeodomain (12,31,32). Missense mutations of NRL can also cause autosomal dominant retinitis pigmentosa (33). Nrl is preferentially expressed in the rod photoreceptors (34), and its loss in mice results in the failure of rod photoreceptor development and an excess of functional S-cones (35). Here, we report the identification and characterization of a novel retinal homeobox gene, designated as Qrx (Q50-type retinal homeobox gene), which shares a nearly identical homeobox with Rx/Rax, plays a role in modulating the expression of Rhodopsin and possibly other photoreceptor genes, and may be associated with retinal degeneration.
Human Molecular Genetics, 2004, Vol. 13, No. 10
1027
The genes flanking QRX in human (the centromeric ZNFINGER RNA BP, and MATK and the telomeric MRP154, and APBA3) are preserved in mouse in terms of position and orientation. However, the interval between them is only 2 kb in mouse compared to 10 kb in human, and sequence analysis
does not reveal evidence of any homeobox-like sequence, nor even any substantial open reading frame, within the interval. In addition, BLAST analysis of human QRX against mouse genome databases did not reveal any significant matches except to the homeodomain itself.
Downloaded from http://hmg.oxfordjournals.org/ by guest on June 7, 2013
Figure 1. Qrx belongs to the paired-like class of homeobox genes. (A) Amino acid sequence alignment of bovine Qrx (B-Qrx), human QRX (H-QRX), and homeodomains of RX (RX-HD), PAX6 (PAX6-HD) and CRX (CRX-HD). Residues identical to human QRX are represented as dashed lines. The homeodomains are shaded in a box with the arrowhead depicting the position of the second intron and the asterisk marking the residues defining the paired-like class. The predicted sequence changes identified in the three human patients are shown at the corresponding positions with underlined letters above the wild-type human QRX amino acid sequence. (B) Human QRX gene structure. The QRX gene is composed of three exons. Its coding region is encoded by exons 2 and 3. HD represents the homeodomain. The numbers mark the nucleotide positions of intron 2 and the start and stop codons within the QRX cDNA. (C) Intron–exon boundary sequence of the human QRX gene. The 50 and 30 intron sequences match the canonical 50 -gt/30 -ag sequence as underlined. (D) Phylogenetic analysis of Rx/Rax and Rx/Rax-related genes. Amino acid sequences corresponding to human QRX (H-QRX-p, protein accession number AAP41547), bovine Qrx (B-Qrx-p, AAP41546), Drosophila Rx/Rax (D-Rx-p, NP_611526), zebra fish Rx3 (Z-Rx3-p, NP_571302), oryzias Rx3 (O-Rx3-p, CAC69975), mouse Rx/Rax (M-Rx-p, NP_038861), human RX/RAX (H-RX-p, NP_038463), rat Rx (R-Rx-p, NP_446130), xenopus Rx2 (X-Rx2-p, 042567), xenopus Rx (X-Rx-p, O42201), chicken Rx2/Rax2 (C-Rx2-p, Q9PVX0), zebra fish Rx1 (Z-Rx1-p, O42356), astyanax Rx1 (A-Rx1-p, Q9I9D5), chicken Rx1/RaxL (C-Rx1-p, Q9PVY0), oryzias Rx2 (O-Rx2-p, Q9I9A2) and zebra fish Rx2 (Z-Rx2-p, NP_571301) were used to generate the phylogenetic tree using the Geneworks program. The length of the horizontal lines represents the relative relatedness of the gene groups. (E) Schematic physical maps of the human QRX locus (19p13.3) (top panel) and conserved syntenic region in M. musculus (10C1.78–79) (bottom panel). The centromeric genes encoding ZN-FINGER RNA BP and MATK (megakyracytes-associated tyrosine kinase), as well as the telomeric genes encoding APBA3 and MRP154, are conserved in the same orientation in human and M. musculus. The physical distance between the flanking genes of QRX is 10 kb in human and 2 kb in mouse. This data was generated through the use of the Celera Discovery System and Celera’s associated databases.
1028
Human Molecular Genetics, 2004, Vol. 13, No. 10
Qrx is preferentially expressed in outer and inner nuclear layers of retina
A human QRX ‘promoter’ fragment is sufficient to drive GFP expression in presumptive photoreceptor precursors in transgenic mice Given the absence of Qrx in mice, we were interested in whether the human QRX promoter would still function in mice (i.e. whether the regulatory information was evolutionarily conserved). Transgenic mice were generated using a 8.1 kb genomic DNA fragment containing the human QRX locus (including 5.8 kb upstream sequence). To aid expression studies, an IRES-EGFP cassette was added to the downstream end of the genomic fragment. Analysis of postnatal day 1 retinas from four out of five analyzed independent transgenic lines revealed strong GFP expression in cells in the outer region of the neuroblast layer, presumptive developing photoreceptor cells of the forming outer nuclear layer (Fig. 2D and E). Also of potential interest is moderate GFP expression in isolated cells in the middle part of the neuroblast layer, perhaps migrating progenitor cells. These results suggest not only the strong conservation of the upstream regulatory complexes between human and mouse retina gene expression but also the likelihood of the involvement of QRX in the development and function of photoreceptors. Qrx and Rx/Rax bind to both the Ret-1 and BAT-1 elements Electrophoretic mobility shift assay (EMSA) and DNase I footprint analyses were carried out to characterize the DNA binding properties of QRX. Since the full-length QRX fusion protein was insoluble, recombinant QRX homeodomain GST fusion protein (QRX-HD-WT) containing six amino acids Nterminal to the homeodomain and five amino acids C-terminal to it, was expressed and purified from bacteria. As expected based on the initial one-hybrid results, QRX-HD-WT binds to both Ret-1 and BAT-1 oligos in EMSA (Fig. 3A). With larger amounts of QRX-HD-WT protein, faint additional bands of lower mobility were also evident (Fig. 3A, lanes 3 and 4),
Figure 2. Expression analysis of Qrx. (A) Northern blot analysis of bovine tissues using bovine Qrx as probe. The top panel shows the image of a northern blot with 10 mg of total RNA per lane isolated from bovine retinal pigment epithelium (P), retina (R), brain (B), heart (H), liver (L), muscle (M) and testis (T) hybridized with a 32P-labeled bovine Qrx cDNA probe. The bottom panel shows the same membrane stained with ethidium bromide under UV light before hybridization. (B) In situ hybridization of bovine Qrx in bovine retina. Expression of Qrx is seen in outer nuclear layer (ONL) and inner nuclear layer (INL), with 100 ng of antisense Qrx probe. The bar in the images represents 25 mm. (C) Parallel experiments as in (B) except that the same amount of sense probe was used. The bar in the images represents 25 mm. (D) DAPI staining of the retina from a human QRX-IRES-EGFP transgenic mouse at P1 stage. RPE (retinal pigment epithelium), NBL (neuroblast layer), GCL (ganglion cell layer). (E) Strong GFP expression in the developing neuroblast layer of retina of the same mouse as in (D).
possibly due to the binding of QRX-HD-WT multimers to the DNA target. The binding specificity of the fusion protein to Ret1 and BAT-1 sites was assessed by cold competitor assay. When wild-type cold DNA oligomer was added to either Ret-1 (Fig. 3A, lanes 5 and 6) or BAT-1 binding reactions (Fig. 3A, lanes 17 and 18), the binding of QRX-HD-WT to the labeled probes was competed away by the cold competitor. When the putative binding site in Ret-1 was mutated (MUT for Ret-1), the oligomer could no longer compete effectively with the probe (Fig. 3A, lanes 7 and 8). In the case of the BAT-1 site, when only one of the two putative QRX-HD-WT binding sites within the element was mutated, the cold competitor could still compete away the binding (Fig. 3A, MUT-1 and MUT-2, lanes 19–22). When both sites in the BAT-1 element were mutated (MUT-3), the mutant competitor showed minimal, if any, ability to compete with the labeled wild-type probe (Fig. 3A, lanes 23 and 24). A non-target sequence (NTS) cold competitor derived from another region of the Rhodopsin promoter showed no competition for the labeled Ret-1 probe (Fig. 3A, lanes 9 and 10). In addition to the EMSA analysis with the purified QRX homeodomain GST fusion protein (QRX-HD-WT), we also examined the interaction of native Qrx from bovine retinal nuclear extract with both Ret-1 and BAT-1 using an affinity purified rabbit polyclonal anti-Qrx antibody (6760) in the supershift assay. It appears that the anti-Qrx antibody largely eliminated the interaction of retina-enriched factors, presumably including both Qrx and Qrx-Crx complex in the retinal nuclear extract, with both Ret-1 and BAT-1 (Fig. 3B, lanes 3, 4 and 11, 12),
Downloaded from http://hmg.oxfordjournals.org/ by guest on June 7, 2013
The expression pattern of Qrx in adult tissues was examined by northern blotting (Fig. 2A) and RT–PCR (data not shown) analyses. Among the tissues tested with northern blotting [retina, retinal pigment epithelium (RPE), brain, heart, liver, muscle and testis], Qrx is only expressed in the retina, with a single abundant transcript of 2 kb. However, by RT–PCR analysis, Qrx expression was also detected weakly in the brain, testis and spleen. Based on RT–PCR analysis of the earliest bovine embryonic retinal tissue available, Qrx expression during development is evident as early as 4.5 months gestation (data not shown). For comparison, bovine Rhodopsin is minimally detectable at 5.2 months gestation and is highly expressed by 6.3–7.4 months gestation (7,49). By in situ hybridization analysis, Qrx is expressed in cells of the outer nuclear (ONL) and inner nuclear layers (INL), with stronger staining in the ONL (Fig. 2B and C). The expression of QRX in adult human retina has a similar pattern, which overlaps largely with that of RX/RAX (data not shown).
Human Molecular Genetics, 2004, Vol. 13, No. 10
1029
whereas it did not alter the non-specific interaction present in the 293 cell nuclear extract (Fig. 3B, lanes 7, 8, and 15, 16). In addition, the pre-immune serum (Fig. 3B, lanes 5 and 13) from the corresponding rabbit did not show any blockage of the binding activity. This is not surprising because the amino acids used as the peptide antigen (codon 88 to 107) to generate the anti-Qrx antibody (6760) are located in close vicinity to the C-terminus of the Qrx homeodomain (codon 27–86) and are likely to be required for the DNA binding activity of Qrx. To further explore the DNA binding properties of QRX in the context of the Rhodopsin promoter, and to compare it with other factors implicated in Rhodopsin regulation, DNase I footprinting analysis was performed with a probe spanning the Rhodopsin promoter from 225 to þ70 bp. Consistent with the EMSA results, the Ret-1 and BAT-1 sites were efficiently
protected by QRX-HD-WT (Fig. 4, lanes 1–5 and 14–18). As might be predicted from the high degree of homology between the QRX and Rx/Rax homeodomains, the Rx/Rax homeodomain pattern of protection is essentially identical to that of QRX (Fig. 4, lanes 6–9 and 19–22). However, the binding affinity of Rx/Rax may be somewhat lower in that 4fold more protein was required to achieve a similar degree of protection. Of potential significance, the protection pattern of QRX and Rx/Rax is similar to, but distinct from, that of Crx (12) (Fig. 4, lanes 10–13 and 23–26). The Ret-1 and BAT-1 protected sequences are overlapping but not identical, and Crx also protects the Ret-4 site. In addition, the patterns of hypersensitive sites are also distinct, presumably reflecting different DNA-protein conformations (Fig. 4, see arrows and arrowheads).
Downloaded from http://hmg.oxfordjournals.org/ by guest on June 7, 2013
Figure 3. QRX binds specifically to the Ret-1 and BAT-1 elements in EMSA. (A) EMSA with Ret-1 or BAT-1 probe in the absence (lanes 1–4 and lanes 11–16) or presence of cold oligo DNA competitors (lanes 5–10 and lanes 17–24). Lanes 1 and 11 do not contain protein. Lanes 2–4 contain increasing amounts of QRX-HDWT (1, 2 and 4 ng). Lanes 12–16 contain increasing amounts of QRX-HD-WT (1, 2, 4, 8 and 16 ng). Lanes 5–10 contain a constant amount of QRX-HD-WT (2 ng). Lanes 5, 7 and 9, contain a 200-fold excess of cold competitor Ret-1WT, Ret-1MUT, NTS (non-target sequence), respectively. Lanes 6, 8 and 10, a 10 000fold excess of cold competitor Ret-1WT (WT), Ret-1MUT (MUT), NTS (non-target sequence), respectively. Lanes 17–24 contain a constant amount of QRX-HDWT (4 ng). Lanes 17, 19, 21 and 23 contain a 200-fold excess of cold competitor BAT-1WT (WT), BAT-1MUT1 (MUT1), BAT-1MUT2 (MUT2) or BAT-1MUT3 (MUT3), respectively. Lanes 18, 20, 22 and 24, contain 10 000-fold excess of cold competitor BAT-1WT (WT), BAT-1MUT1 (MUT1), BAT-1MUT2 (MUT2) or BAT-1MUT3 (MUT3), respectively. (B) EMSA with Ret-1 or BAT-1 probe using nuclear extracts from either bovine retina or 293 cells in the presence (lanes 3, 4, 7, 8, 11, 12, 15, 16) or absence (lanes 2, 6, 10, 14) of the anti-Qrx antibody (6760). In lanes 5 and 13, the same volume (as lanes 3 and 11) of preimmune serum (PI) was used rather than the 6760 antibody.
1030
Human Molecular Genetics, 2004, Vol. 13, No. 10
QRX specifically transactivates the Ret-1 element in the presence of Crx and NRL To assess the transcriptional activity of QRX, we utilized a transient transfection assay with HEK293 cells. Using the bovine Rhodopsin 130 to þ70 and 225 to þ70 bp luciferase reporter that we employed previously to assess the transactivating activity of Crx and Nrl (8,9), we did not observe significant transactivating activity with QRX, either alone or in combination with Crx and/or NRL (data not shown). To develop a potentially more sensitive assay, and one that would be specific for individual DNA binding sites, thereby eliminating potential neighboring negative regulatory elements, luciferase reporter constructs were generated in which multimers of the Ret-1, BAT-1, Ret-4 or NRE/Ret-4 element were placed upstream of a minimal CMV promoter. Co-transfection of a QRX expression plasmid with each of these promoter elements again revealed minimal transactivating activity. This result was distinct from that with Rx/Rax, which could activate the BAT-1 construct (Fig. 5B) and has been reported to activate a Ret-1/PCE-1 site from the Arrestin promoter (23). The multimer reporters demonstrated a number of specific patterns in terms of their abilities to be activated by the various transcription factors that were tested. When used alone, Crx activated the Ret-1 and Ret-4 constructs (Fig. 5A and D), and
NRL selectively stimulated the NRE/Ret-4 containing reporter (Fig. 5C). However, when QRX was co-transfected with Crx and NRL, it further stimulated the expression of the Ret-1 construct by more than 10-fold over the effect of Crx and NRL combined in a dose dependent manner (Fig. 5E). The level of activation achieved with QRX was significantly higher (5-fold more) than what could be achieved simply by increasing the amount of Crx or NRL (data not shown). The transactivating effect peaks at 0.3 mg of QRX expression plasmid (Fig. 5E), which is relatively low for this type of assay. Furthermore, as little as 0.1 mg of QRX expression plasmid can achieve more than 5-fold activation. This transactivation activity of QRX is binding site specific, even though as noted above QRX binds to both the Ret-1 and BAT-1 sites, its ability to stimulate the activity of Crx and NRL is only observed with Ret-1 (BAT-1 data not shown). Rx/Rax also stimulates the combined activity of Crx and NRL, but only to about 50% the level of QRX (Fig. 5F). In order to dissect the relative contributions of Crx and NRL interactions in the observed QRX stimulation, QRX was tested with either of the two factors individually. QRX stimulates Crx alone significantly, to 50% the activity seen with Crx and NRL, whereas QRX stimulation of NRL alone is only minimal (Fig. 5F), suggesting that Crx is the key contributor in this three factor interaction.
Downloaded from http://hmg.oxfordjournals.org/ by guest on June 7, 2013
Figure 4. QRX homeodomain (QRX-HD-WT), as well as RX homeodomain (RX-HD) binds specifically to the Ret-1 and BAT-1 elements within the Rhodopsin promoter in DNase I footprint analysis. QRX-HD and RX-HD show identical pattern of binding to their target elements in the Rhodopsin promoter (Ret-1 and BAT1). Their patterns of binding are distinct from that of the CRX homeodomain (CRX-HD), as indicated by the arrows (for top strand) and arrowheads (for bottom strand). In addition, unlike CRX-HD, QRX-HD and RX-HD do not bind to the Ret-4 element. The target elements are bracketed and labeled accordingly. 0.25 mg of each protein was used in lanes 2, 6, 10, 15, 19 and 23; 1 mg in lanes 3, 7, 11, 16, 20 and 24; 4 mg in lanes 4, 8, 12, 17, 21 and 25; 16 mg in lanes 5, 9, 13, 18, 22 and 26.
Human Molecular Genetics, 2004, Vol. 13, No. 10
1031
Downloaded from http://hmg.oxfordjournals.org/ by guest on June 7, 2013
Figure 5. QRX specifically transactivates the Ret-1 element synergistically with Crx and NRL, whereas other factors show distinct target selectivity and QRX variants differentially affect its transcriptional activity. The boxes in the plots represent the experimental data points falling between the 25th and 75th percentile of each group. The bars on either side of the boxes depict the range of the data. The medians of each group are represented by a triangle. *P-value 0.0001 and **0.0001 P-value 0.0025 as compared to the reference group (the reference group is the vector group in A–E, the QRX group in F and the QRX-WT group at the corresponding DNA concentrations in G). (A) Effect of individual factors (1 mg/plate) on Ret-1CMV. (B) Effect of individual factors (1 mg/plate) on BAT-1 CMV. (C) Effect of individual factors (1 mg/plate) on NRE þ Ret-4. (D) Effect of individual factors (1 mg/plate) on Ret-4. (E) The effect of QRX on Ret1CMV in the presence of Crx and NRL at increasing DNA concentrations. (F) Effect of Rx on Ret-1CMV in the presence of Crx and NRL and QRX’s effect on Ret-1CMV in the presence of either Crx or NRL (shown in % activity of QRX with Crx and NRL). (G) Effect of the mutations of QRX at different DNA concentrations on the synergistic transactivation activity of QRX with Crx and NRL on Ret-1CMV.
QRX physically interacts with CRX, but not NRL The finding that Crx is essential for most of the transactivating function of QRX suggested that the QRX protein might physically interact with Crx, and that this interaction might be responsible for the observed synergy between the factors. To test this hypothesis, we performed co-immunoprecipitation and GST pull-down analyses. An anti-CRX antibody
precipitated radioactively labeled in vitro transcribed and translated QRX protein only in the presence of CRX (Fig. 6A, lane 1). The same antibody did not yield precipitate when used with the empty QRX expression vector (data not shown), nor did it precipitate QRX in the absence of CRX (Fig. 6A, lane 5). In addition, using recombinant proteins expressed in HEK 293 cells, QRX (QRX-Xp) was co-immunoprecipitated with Flag tagged CRX (CRX-Flag) (Fig. 6B, lane 1).
1032
Human Molecular Genetics, 2004, Vol. 13, No. 10
Downloaded from http://hmg.oxfordjournals.org/ by guest on June 7, 2013
Figure 5. Continued.
Consistent with these results, we found in a yeast two-hybrid assay that the Crx homeodomain interacts with Qrx (data not shown). In contrast, QRX failed to interact with NRL in a GST pull-down assay (Fig. 6C, lane 4). This is in contrast to the finding that both Rx and Crx interact with NRL (Fig. 6C, lanes 8 and 12), and that deletion of the transactivation domain containing amino-terminus of NRL (GdN), which has been shown to be involved in its interaction with Crx (32), reduced its interaction with both Rx and Crx (Fig. 6C, lanes 7 and 11).
Mapping of QRX to human chromosome 19p13.3, and identification of QRX sequence variants in cone-rod dystrophy and age-related macular degeneration patients Human QRX was mapped to chromosome 19p13.3 by radiation hybrid mapping, in which it was closely linked to the marker WI 6480 with a lod score of >3. The result was confirmed by somatic cell hybrid mapping to chromosome 19 and subsequently also by data from the human genome project. The cosmid F16403, which spans the entire QRX cDNA fragment, also maps to the same
Human Molecular Genetics, 2004, Vol. 13, No. 10
locus. By comparing the sequence of the human QRX cDNA to the cosmid F16403, the genomic structure of QRX was determined. The QRX cDNA is composed of three exons, with the coding region encoded by exons 2 and 3 (Fig. 1B). All the intron–exon boundary sequences match the canonical GT-AG consensus sequence (Fig. 1C).
To test the hypothesis that defects of QRX and its protein product could result in retinal disease, we screened a cohort of patients with retinopathies for QRX mutations: 322 cone-rod dystrophy (CORD), 107 Leber congenital amaurosis (LCA), 92 age-related macular degeneration (AMD), 14 autosomal recessive retinitis pigmentosa (ARRP) and 14 autosomal dominant retinitis pigmentosa (ADRP) patients, as well as a pool of 94 normal individuals. Three different heterozygous sequence variations were identified in patients with retinal disease that would be expected to alter the QRX protein (Table 1). In contrast, no amino acid altering sequence variations were observed in the pool of 94 normal control individuals. A 67-year-old patient with the clinical diagnosis of AMD was found to harbor a heterozygous G to A change at codon 87, which would be expected to result in a change from arginine to glutamine (CGG to CAG in Fig. 7A). This replacement of a bulky, positively charged amino acid by a neutral one occurs at the c-terminal border of the homeodomain (Fig. 1A), and suggested that it might affect the protein’s DNA binding activity. When this possibility was directly tested using a GST tagged recombinant QRX-HD-R87Q mutant protein, we found the R87Q alteration to cause an apparent increase in binding affinity for the Ret-1 and BAT-1 elements (Fig. 7C and D). This increase was not due to different amounts of the mutant or wild-type proteins used in the assay (Fig. 7B). A 78-year-old patient with macular degeneration and electroretinographic evidence of peripheral retinal involvement (which we interpret as CORD) was found to harbor a missense change in codon 137 (GGC to CGC). This change would be expected to result in the substitution of arginine for glycine (Table 1). Unfortunately, no family members of either of these two patients were available for interview, clinical examination or DNA analysis. A third heterozygous QRX sequence variation, a 6 bp (CCCGGG) insertion between codons 137 and 141, was identified in an unrelated CORD patient. This change would be expected to result in an in-frame insertion of a proline and a glycine (Fig. 1A). Since this region of the gene contains multiple repeats of the CCCGGG sequence, insertion or duplication of the CCCGGG sequence at multiple positions between codons 137 and 141 could result in the same final sequence. For simplicity, we designate this mutation as 140P_141Gdup (50). The same sequence change was also identified in the proband’s mother and two siblings. Although these relatives are said to have ‘normal’ vision, they all live in Pakistan and are not available for ophthalmologic exam and testing. Hence, the possibility that they have mild CORD cannot be excluded. If the relatives in fact do not have CORD, then this would suggest either incomplete penetrance, perhaps due to modifier or environmental influence, or that the 140P_141Gdup sequence change is not disease-associated. The R87Q, G137R and 140P_141Gdup sequence variants were expressed and tested in vitro to assess their effect on the interaction between QRX and CRX, and their functional consequence in the context of other transcription partners in vivo. In co-immunoprecipitation analyses, the G137R and 140P_141Gdup (140PGIns) proteins demonstrated decreased but not absent interaction with CRX, while the R87Q protein did not show decreased interaction (Fig. 6A). In transient transfection assays, all three sequence variants demonstrated subtle but statistically significant (P < 0.0001) and reproducible
Downloaded from http://hmg.oxfordjournals.org/ by guest on June 7, 2013
Figure 6. QRX physically interacts with CRX, which also interacts with NRL, and QRX variants demonstrate reduced interaction between QRX and CRX (A) In vitro transcribed and translated 35S-labeled QRX (WT and mutants) is co-immunoprecipitated with CRX using an anti-CRX antibody (lanes 1–5). Both the G137R (lane 3) and the 140P_141Gdup mutation (140PGIns in lane 4) resulted in reduced amount of QRX being precipitated suggesting decreased interaction between CRX and these mutant QRX proteins. However, the R87Q mutation resulted in little, if any, change compared to the wild-type (lane 2). In the absence of CRX, QRX is not precipitated (lane 5). Lanes 6–9 demonstrate that an approximately equal amount of each input protein was used. (B) Recombinant QRX (QRX-Xp) synthesized in HEK293 cells is co-precipitated with CRX-Flag. Immunoprecipitation using anti-Flag-M2 resin containing either CRX-Flag, or its vector control, with QRX-Xp was performed as described in experimental procedures. Lanes 1, 2, 4 and 5 are elutes (20 ml) from the resin with the indicated protein extracts; lanes 3, 6 and 7 are the inputs (10 ml) as indicated. Lanes 1, 2, 3 were probed with anti-Xpress (Invitrogen, 1 : 1000) and goat anti-mouse IgG-HRP and lanes 5, 6, 7 with anti-CRX P261 antibody (1 : 250) and anti-rabbit IgG-HRP. Arrows with labeling indicate the running positions for QRX-Xp and CRX-flag. (C) GST pull-down assays with in vitro transcribed and translated 35S-labeled QRX (QRX*), Rx/Rax (Rx*) or Crx (Crx*). QRX* did not show evidence of binding to columns with GST-NRL (GNRL, lane 4), GST-NRL with transactivation domain deleted (GdN, lane 3) nor with GST alone (lane 2). In contrast, both Rx and Crx interact with GNRL (lanes 8 and 12) and GdN (lanes 7 and 11) but not with GST alone (lanes 6 and 10).
1033
1034
Human Molecular Genetics, 2004, Vol. 13, No. 10
Table 1. Summary of QRX sequence variants found in patients with retinal disease DNA change
Protein change
Disease
Genetic class
CGG > CAG
R87Q
Heterozygous
GGC > CGC CCCGGG insertion
G137R 140P_141Gdup
Age-related macular degeneration Cone-rod dystrophy Cone-rod dystrophy
Heterozygous Heterozygous
difference from the QRX-WT in their ability to transactivate the Ret-1 element in the presence of Crx and NRL (Fig. 5G). At the DNA concentration that showed peak QRX-WT activity (0.3 mg/plate), R87Q and 140P_141Gdup resulted in increased transactivation activity of 30 and 50%, respectively, while G137R demonstrated 40% reduced transactivating activity at a concentration of 1 mg/plate.
DISCUSSION The transcriptional network implicated in the regulation of photoreceptor gene expression is becoming increasingly complex, and now includes Crx (12–14,51), Nrl (8,9, 31,32,34,52), Erx (53), Rx (15,16,19,21,23), Mash1 (17,54), Mok2 (55), nuclear receptor NR2E3 (35,56), thyroid hormone receptor beta 2 (57), the Sp family (58), and, at least in Drosophila, Pax6 (59–61). The combinatorial interplay of these factors is important not only for orchestrating normal
development and function, but also in mediating retinal pathology, as shown by reports that mutations in many of their genes have been associated with photoreceptor disease (14,27,33). Here, we have presented the characterization of a new transcription factor, Qrx, which appears to be involved in modulating the expression of photoreceptor specific genes and may play a role in retinal degeneration. Based on its structure, Qrx is a member of the Q50-type paired-like homeobox gene family, and its closest homolog is Rx/Rax. Rx/Rax and its related genes have been identified from nine different species, and some species express as many as three distinct family members (15,21,62). Unfortunately, however, relatively little functional information is available for these Rx/Rax-related genes, and whether Qrx represents a true ortholog of any of them is unclear. One finding, however, of potential relevance to Qrx is that expression of a putative dominant negative allele of a chicken Rx/Rax-related gene, cRaxL (C-Rax1), causes a decrease in expression of early photoreceptor markers (63). An interesting, and puzzling, aspect of the evolution of Qrx is that it appears to be absent from the M. musculus genome. Qrx seems to have been deleted from the M. musculus genome during evolution, rather than having evolved more recently than the branch point of M. musculus. Although it had been thought that the order Rodentia, to which M. musculus belongs, is more distantly related to the primate lineage than the order Certartiodactyla, to which bovine belongs (64), more recent molecular phylogenetic analysis indicates that M. musculus is a closer relative to human than is bovine (65,66). If so, a mechanism explaining the absence of a murine Qrx ortholog as being due to more recent evolution would necessitate an independent origin for
Downloaded from http://hmg.oxfordjournals.org/ by guest on June 7, 2013
Figure 7. The R87Q sequence change results in an apparent increase of the binding affinity of QRX to the Ret-1 and BAT-1 elements. (A) Heterozygous DNA sequence change (CGG to CAG) resulting in R87Q mutation. (B) Coomassie blue staining of QRX-HD-WT and QRX-HD-R87Q showing equivalent protein quantities (in nanograms). (C) EMSA comparing QRX-HD-WT and QRX-HD-R87Q DNA binding to the Ret-1* probe. The amount of protein (in nanograms) used in each reaction is indicated at the top of each lane. The quantity of probe shifted with QRX-HD-R87Q, compared to QRX-HD-WT, was increased 12.8-, 1.8- and 2.3-fold with 4, 8 and 16 ng of protein, respectively. (D) EMSA comparing QRX-HD-WT and QRX-HD-R87Q DNA binding to the BAT-1* probe. The amount of protein (in nanograms) used in each reaction is indicated at the top of each lane. The quantity of probe shifted with QRX-HD-R87Q, compared to QRX-HD-WT, was increased 4.6-, 4.8- and 2.3-fold with 4, 8 and 16 ng of protein, respectively.
Human Molecular Genetics, 2004, Vol. 13, No. 10
results, as well as previous studies, is shown in Figure 8. An unusual aspect of the model is that NRL is suggested to function by two mechanisms. With its already identified binding site, NRE, Nrl is shown to function in the standard manner as a DNA binding transcription factor (9). However, we suggest that with the Ret-1 site, NRL functions more as a coactivator or modulator, acting through protein–protein interaction with Crx without its direct DNA binding. This hypothesis is based on the finding that the addition of NRL with the Ret-1 reporter approximately doubles the activity observed with Crx and QRX alone (Fig. 5F), yet NRL does not demonstrate detectable binding to the Ret-1 sequence. NRL is shown as interacting directly with Crx, but not with QRX, based on the co-immunoprecipitation and GST pull-down assays shown in Figure 6. A second potential co-activator shown in the model is p300/CBP, as has been reported previously that it can both physically interact with Crx and enhance its transactivating function (68). An additional level of complexity, not indicated in Figure 8, is that negative acting transcription factors may bind the Rhodopsin promoter and repress expression in nonphotoreceptor retinal cells or modulate Rhodopsin promoter activity even in rods (51,52). Since Qrx does not exist in mouse genome, presumably its activity in regulating Rhodopsin expression could be compensated by other factors, including Rx, in mouse. The identification of QRX sequence changes in patients with retinal diseases is also consistent with a role for QRX in photoreceptor function and/or survival. Although the absence of existing or available relatives for the R87Q and G137R probands, as well as the presence of 140P_141Gdup in the proband’s presumably normal relatives, make it difficult to conclude that any of the identified sequence changes truly represent disease-causing mutations, the nature and positions of the sequence changes and the biochemical studies with the mutated proteins, are all supportive of this possible association. Each of the sequence changes led to an alteration that could be detected in our functional assays. The G137R mutation resulted in decreased interaction with CRX and decreased transactivation activity, the 140P_141Gdup led to decreased interaction with CRX but increased transactivation, and the R87Q mutation increased both transactivation and DNA binding activity. The possibility that increased transcription factor activity can cause eye disease is not novel, having been reported previously with both NRL (33) and PAX6 (69). Although this suggestion that different, and sometimes opposite, types of transcription factor alterations can result in similar ocular phenotypes may at first seem paradoxical, it probably reflects the growing consensus that the retinal gene expression network is so tightly regulated that a variety of different perturbations can all result in retinal degeneration.
MATERIALS AND METHODS Yeast one hybrid assay Library screening was performed essentially as described in Chen et al. (12) except the target elements used here were Ret-1 (five copies) or BAT-1 (three copies).
Downloaded from http://hmg.oxfordjournals.org/ by guest on June 7, 2013
the gene in human and bovine ancestors, and this seems unlikely given the high degree of nucleic acid sequence homology between the bovine and human orthologs. A more likely mechanism involves a small deletion mediated by recombination between the flanking repetitive elements. Of potential significance, multiple genes immediately flanking Qrx are conserved between mouse and human, but there are local disruptions of the nearby synteny (Fig. 1E). Perhaps related to these findings, compared to the human genome, the mouse genome is evolutionarily more volatile (67), giving mice the ability to rapidly adapt to changing environments. It is therefore interesting that, as demonstrated by the transgenic experiments, the regulatory information from the human QRX promoter region is still largely recognized by the mouse. As a practical side benefit, a QRX promoter/GFP transgene may turn out to be a useful early marker of developing photoreceptor cells. The absence of murine orthologs for human genes, and vice versa, is not rare. It has been estimated to occur for as many as 3–5% of genes in both genomes (48). Furthermore, homeobox genes appear to have a higher than average frequency of such variation. In human, of the 302 homeobox loci we identified from a NCBI Locuslink search, 139 harbor already characterized homeobox genes. The other 163 represent novel homeobox genes predicted from their nucleotide sequences or expressed sequence tag (EST) clusters. Within the group of previously characterized 139 homeobox genes, nine (6.5%) are absent from the mouse genome based on the search of gene symbols, NCBI HomoloGene data and nucleotide sequence BLAST of both the NCBI and the Celera databases. Analysis of the known murine homeobox genes revealed a similar frequency for which human orthologs could not be identified. These ‘missing’ homeobox genes are likely to be important contributors in the processes that account for the biological, morphological, and functional difference between mouse and man. In the eye, for example, variation in properties such as the number of types of cones, the relative frequency of cones and rods and the presence or absence of a macular region could partly be determined by differences in the pattern of homeodomain proteins expressed in the retina. The absence of a murine Qrx has made functional and developmental analysis more difficult. Nonetheless, several lines of evidence do provide insights into the role of Qrx within the retina. Although results of transient transfection experiments must always be interpreted with caution in terms of their relevance to the in vivo situation, the data presented suggest that Qrx may play a role in modulating the expression of Rhodopsin and perhaps other photoreceptor genes. While QRX did not demonstrate significant transactivating activity on its own, when used in combination with Crx and NRL, it stimulated reporter gene expression more than 10-fold. This effect was observed only with the Ret-1 site and not with the Ret-4 or BAT-1 sites. In the same assay with the same site, Rx/ Rax showed less than half the activity demonstrated by QRX, even though in the DNA binding assays it demonstrated essentially identical sequence specificity. Rx/Rax was the most potent activator of the BAT-1 reporter construct. These findings are noteworthy since although homeodomain proteins demonstrate in vivo specificity, they often show significant promiscuity in in vitro binding studies and cell transfection assays. A schematic model of Rhodopsin regulation that integrates these
1035
1036
Human Molecular Genetics, 2004, Vol. 13, No. 10
Figure 8. Model of combinatorial network regulating Rhodopsin expression. The primary target of QRX within the bovine Rhodopsin promoter appears to be the Ret-1 element. Upon binding to Ret-1 and CRX, which is associated with NRL through protein–protein interaction, QRX forms a heterotrimeric complex with CRX and NRL and together they are able to recruit other co-activators and/or components of the basal transcription machinery (PolII complex) to activate transcription. RX, bound to the BAT-1 element, may be forming a second complex with NRL and CRX, bound to the adjacent NRE and Ret-4 elements, respectively. However, whether these two complexes function coordinately is unclear. For simplicity, putative repressors, are not included in the diagram.
Radiation hybrid mapping
Northern blot analysis Total RNA was isolated from different fresh bovine tissues (RPE, retina, brain, heart, liver, muscle and testis) using the TRIZOL reagent (Invitrogen). Approximately 10 mg of total RNA from each tissue was used. After UV crosslinking, the membrane was hybridized to a probe generated from the full-length bovine Qrx cDNA followed by washing and overnight autoradiography. In situ hybridization In situ hybridization experiments were performed essentially as previously described (70), except that DIG-labeled bovine Qrx riboprobes were generated by in vitro transcription of either T3 (antisense probe) or T7 (sense probe) promoter driven plasmid vector containing partial bovine Qrx cDNA (191 basepairs), which has been digested with either EcoRI (sense probe) or BamHI (antisense probe). Eight mm thick cryosections of bovine retina, fixed in fresh 4% paraformaldehyde, were used in the hybridization with buffer containing 100 ng/ml of either antisense or sense DIG-labeled riboprobe. Generation and analysis of the QRX-IRES-EGFP transgenic mice A polylinker containing EcoRI–Af lII–SmaI–AgeI–BamHI sites was inserted into the EcoRI and BamHI sites of the pIRES2EGFP vector (BD Biosciences #632306). The 8.1 kb human QRX fragment generated by digesting the cosmid F16403 with BsaAI and Age1 enzymes was directionally cloned into the SmaI and AgeI sites of the above modified pIRES2-EGFP vector. The resulting construct was further digested with Af lII to release the 9.6 kb QRX-IRES-EGFP fragment for microinjection into the mouse embryos of SJL/J mice at the Johns
Expression and purification of GST-tagged QRX-HD-WT, QRX-HD-R87Q, Rx-HD, Crx-HD Bovine Qrx cDNA fragment corresponding to codons 20–91 or mouse Rx/Rax cDNA fragment corresponding to codons 130– 200 was cloned in frame into the BamHI (50 ) and EcoRI (30 ) sites of the pGEX-4T-2 vector (Pharmacia). Since the amino acid sequence within this region is 100% conserved between human and bovine Qrx and also between human and mouse Rx/Rax, we designate the constructs as QRX-HD-WT and RXHD, respectively. The QRX-HD-WT was also used as the template for the generation of QRX-HD-R87Q mutant construct using QuikChange site-directed mutagenesis kit (Strategene). CRX-HD was the same as previously described (12). All the expression constructs were transformed and expressed in bacteria strain BL21 (Pharmacia). The GST fusion proteins were purified following protocol provided by Pharmacia. The concentrations and quality of purified proteins were determined by both SDS–PAGE/Coomassie blue staining and spectrophotometry.
Generation of polyclonal anti-Qrx antibody (6760) Peptide antigens corresponding to the region of codon 88–107 of both human (LESGSGAVAAPRLPEAPALP) and bovine (LESGSGAVAAPRLPEVPALP) Qrx were synthesized according to standard procedure at Genemed Synthesis Inc. The antigens were injected as a mixture into two rabbits according to standard immunization protocol and antisera were obtained and affinity-purified using the peptide antigens.
Downloaded from http://hmg.oxfordjournals.org/ by guest on June 7, 2013
The MIT Genebridge radiation hybrid panel was analyzed using primers corresponding to human QRX cDNA sequences (forward: 50 -CAGTCTGGGCCAGCGCCACCTC-30 , reverse: 50 -GGCTGGAGTGCAGCAGTGTG-30 ) that amplify the 494 bp PCR products. Analysis of the resulting PCR pattern was submitted to the database (http://www-genome.wi.mit.edu/ cgi-bin/contig/rhmapper), which indicated close linkage to the marker WI6480 with a LOD score of more than 3.
Hopkins Transgenic Core Facility. The resulting transgene positive mice were genotyped by PCR analysis and further bred for multiple generations to eliminate the rd1 locus from the background. Out of a total of 12 transgenic founders, nine independent lines were established. GFP expression in the perinatal retinas of five of these nine lines have been analyzed. Ten micrometer cryosections of the eyecups from transgene positive mice were generated for standard histological processing and viewed with a Nikon Eclipse E1000 microscope using DAPI (excitation: 330–380 nm; dichroic mirror: 400 nm; barrier: 435–485 nm) or FITC (excitation: 465–495 nm; dichroic mirror: 505 nm; barrier: 515–555 nm) filter sets.
Human Molecular Genetics, 2004, Vol. 13, No. 10
EMSA
DNase I footprinting DNase I footprintings were performed essentially the same as previously described (71), except that up to 16 ng of purified GST-tagged QRX-HD-WT, RX-HD, CRX-HD proteins were used to mix with either top (forward) or bottom (reverse) strand labeled DNA probes corresponding to bovine Rhodopsin promoter 225 to þ70 bp. Human subjects Informed consent was obtained from participants after explanation of the studies. The research procedures were in accordance with institutional guidelines and the Declaration of Helsinki. Clinical ocular examinations and visual function testing were performed in the patients with retinal degenerations that had candidate gene screening.
forward: 50 -CTGGAGTCAGGCTCGGGT-30 , exon3B/reverse: 50 -CGGTTGCTCCATGACCTC-30 ) were used to amplify all but 90 base pairs of the coding sequence [the sequence encoding codons 1–30 is so GC rich (72%) that we were not able to amplify it despite using four different oligonucleotide primer pairs and numerous different reaction conditions]. The rest of exons 1 and 2 and the first portion of exon 3 (3A) were amplified with the Qiagen Taq DNA Polymerase Kit containing Coenzyme Q. The second portion of exon 3 (3B) was amplified with MasterAmp 2X PCR PreMix L from Epicentre. PCR products were denatured for 3 min at 94 C and electrophoresed on 6% polyacrylamide/5% glycerol gels at 25 W for 3 h at room temperature. Following electrophoresis, gels were stained with silver nitrate (72). PCR products exhibiting an electrophoretic shift in SSCP analysis were sequenced. Mutations were identified by the approximately equal peak intensity of two fluorescent dyes at the mutant base. The 140P_141Gdup sequence was further confirmed by cloning the mutant PCR products into the pGEM T-EASY vector (Promega) and sequencing multiple clones to identify both the mutant and normal allele sequences. All sequencing was bi-directional. Generation of target element containing reporter constructs, QRX wild-type and mutant expression constructs and transient transfection assays Ret-1 (10 copies), BAT-1 (three copies), Ret-4 (four copies) and Ret-4 þ NRE (four copies) were individually and directionally cloned into the XhoI site of the modified PGL2-promoter vector (Clontech), in which the SV40 promoter has been substituted by a minimal CMV promoter. QRX wild-type cDNA was inframe cloned into the BamHI (50 ) and EcoRI site of the pcDNA3.1HisB vector (Invitrogen). The mutant clones (R87Q, G137R and 140P_141Gdup) were generated using QuikChange site-directed mutagenesis kit (Strategene) using the wild-type QRX construct as the template in the initial PCR reactions. The resulting mutant constructs were sequenced and correct inserts were cut out and pasted into the pcDNA3.1HisB vector to eliminate any potential secondary mutations. The DNA of the resulting subclones were purified twice through CsCl gradient. The Crx and NRL expression constructs, the transfection of 293 cells and luciferase and b-gal assays were essentially as previously described (12). In vitro translation and GST pull-down assay
Screening of QRX mutations using single-strand conformational polymorphism (SSCP) and automated sequencing Samples from patients and normal controls were screened in an identical manner for mutations in the 50 UTR and coding sequences of the QRX gene using single-strand conformational polymorphism (SSCP) and sequencing analyses. Four pairs of oligonucleotide primers (exon1/forward: 50 -ACCTTGAGC TAATCCTCCCC-30 , exon1/reverse: 50 -CACCCTGAGGGG ATCTCTG-30 ; exon2B/forward: 50 -CCAAGAAGAAGCACCG GA-30 , exon2B/reverse: 50 -AAAGATGTGTGTGTTGGGG G-30 ; exon3A/forward: 50 -TCAGACGGGAGTCAGAACCC-30 , exon3A/reverse: 50 -TCCAGGGGCAGCGACATG-30 ; exon3B/
QRX, Rx/Rax and Crx were translated and radio-labeled with 35 S-methionine using coupled in vitro transcription/translation system (Promega) and Glutathione Sepharose bound GST, GSTDNRL and GST-NRL were prepared as described earlier (32). Equal amounts (30%) of each in vitro translated protein (QRX, Rx/Rax and Crx) were incubated separately with GST, GST-DNRL or GST-NRL (100 mg of protein). After the final wash, the Glutathione Sepharose bound protein was solubilized in 2 SDS sample buffer by heating for 5 min at 100 C. Proteins labeled with 35 S-methionine were analyzed by SDS–PAGE followed by autoradiography. Along with the pull-down proteins, 10% of the total protein used for binding was analyzed as input to check the translation efficiency of the individual protein.
Downloaded from http://hmg.oxfordjournals.org/ by guest on June 7, 2013
For the study of purified QRX-HD-WT GST fusion protein, fill-in labeled radioactive (32P) double stranded short DNA probes (10 000 cpm/ml) corresponding to the Ret-1WT (50 -GGCCCACC TGGAAGCCAATTAAGCC-30 ) and BAT-1WT (50 -GCAGC AGTGAGGATTAATATGATTAATAACGCC-30 ) elements were used to mix with purified GST-tagged QRX-HD-WT protein in the presence or absence of appropriate amount of non-radioactive competitors. The DNA sequence for the mutant non-radioactive competitors are as following: Ret-1MUT (MUT: 50 -GGCCCACC TGGAAGCACCGGCAGCC-30 ), NTS (non-target sequence: 50 -GGCCTCTGCTCTTTCCCAGGGTCCCC-30 ), BAT-1MUT1 (MUT1: 50 -GCAGCAGTGAGGATGCCGATGATTAATAAC GCC-30 ), BAT-1MUT2 (MUT2: 50 -GCAGCAGTGAGGATTAA TATGCGGCATAACGCC-30 ), BAT-1MUT3 (MUT3: 50 -GCAGC AGTGAGGATGCCGATGCGGCATAACGCC-30 ). Assays were carried out as previously described (71). The analysis of native Qrx protein in nuclear extract from bovine retina and the control 293 cell nuclear extract were carried out essentially the same as above except that 4 mg of nuclear extract, either from retina or 293 cells, and 0.1 mg of Poly(dI-dC) were used in each reaction. Nuclear extracts were added 20 min before the addition of 1.5 ml (Fig. 3B, lanes 3, 7, 11, 15) or 4.5 ml (Fig. 3B, lanes 4, 8, 12, 16) of the anti-Qrx antibody (6760). As controls, 1 ml of unpurified preimmune serum (PI) was added instead in lanes 5 and 13 of Figure 3B.
1037
1038
Human Molecular Genetics, 2004, Vol. 13, No. 10
In vitro and in vivo expression and co-immunoprecipitation analysis of CRX and wild-type and mutant QRX proteins
Statistical analysis of transient transfection data Relative luciferase activities from each experiment were analyzed and compared among transcription factors using a linear mixed model. This is analogous to analysis of variance (ANOVA), except that rather than averaging correlated observations prior to analysis as required by ANOVA, all data points are used in the analysis and the correlation in observations arising from using the same solution mix to transfect two tissue culture plates was modeled directly, using the assumption that the two tissue culture plates were exchangeable experimental units. No formal adjustment for multiple comparisons of transcription factors was made; however, only P-values 0.0025 were considered statistically significant. All analyses were performed using SAS statistical software (SAS Inc.).
ACKNOWLEDGEMENTS The authors wish to thank Drs Peter H. Mathers for providing the human Rx/Rax cDNA clone and for useful discussion and suggestions on the manuscript; Jeremy Nathans for the human retinal cDNA lgt10 phage library; Ching-Hwa Sung for the bovine retinal cDNA-Gal4 activation domain fusion library; Debora B. Farber and Mark Verardo for the Y79 cells; Adrian M. Timmers for the fetal bovine retinal RNA samples for initial
REFERENCES 1. Dowling, J. (1987) The Retina: An Approachable Part of the Brain. Belknap Press of Harvard University, Cambridge, MA. 2. Turner, D.L. and Cepko, C.L. (1987) A common progenitor for neurons and glia persists in rat retina late in development. Nature, 328, 131–136. 3. Adler, R. (2000) A model of retinal cell differentiation in the chick embryo. Prog. Retin. Eye Res., 19, 529–557. 4. Livesey, F.J. and Cepko, C.L. (2001) Vertebrate neural cell-fate determination: lessons from the retina. Nat. Rev. Neurosci., 2, 109–118. 5. Kikuchi, T., Raju, K., Breitman, M.L. and Shinohara, T. (1993) The proximal promoter of the mouse arrestin gene directs gene expression in photoreceptor cells and contains an evolutionarily conserved retinal factor-binding site. Mol. Cell. Biol., 13, 4400–4408. 6. Chen, S. and Zack, D.J. (1996) Ret 4, a positive acting rhodopsin regulatory element identified using a bovine retina in vitro transcription system. J. Biol. Chem., 271, 28549–28557. 7. DesJardin, L.E. and Hauswirth, W.W. (1996) Developmentally important DNA elements within the bovine opsin upstream region. Invest. Ophthalmol. Vis. Sci., 37, 154–165. 8. Kumar, R., Chen, S., Scheurer, D., Wang, Q.L., Duh, E., Sung, C.H., Rehemtulla, A., Swaroop, A., Adler, R. and Zack, D.J. (1996) The bZIP transcription factor Nrl stimulates rhodopsin promoter activity in primary retinal cell cultures. J. Biol. Chem., 271, 29612–29618. 9. Rehemtulla, A., Warwar, R., Kumar, R., Ji, X., Zack, D.J. and Swaroop, A. (1996) The basic motif-leucine zipper transcription factor Nrl can positively regulate rhodopsin gene expression. Proc. Natl Acad. Sci. USA, 93, 191–195. 10. Yu, X., Leconte, L., Martinez, J.A. and Barnstable, C.J. (1996) Ret 1, a cis-acting element of the rat opsin promoter, can direct gene expression in rod photoreceptors. J. Neurochem., 67, 2494–2504. 11. Freund, C., Horsford, D.J. and McInnes, R.R. (1996) Transcription factor genes and the developing eye: a genetic perspective. Hum. Mol. Genet., 5, 1471–1488. 12. Chen, S., Wang, Q.L., Nie, Z., Sun, H., Lennon, G., Copeland, N.G., Gilbert, D.J., Jenkins, N.A. and Zack, D.J. (1997) Crx, a novel otx-like paired-homeodomain protein, binds to and transactivates photoreceptor cell-specific genes. Neuron, 19, 1017–1030. 13. Furukawa, T., Morrow, E.M. and Cepko, C.L. (1997) Crx, a novel otx-like homeobox gene, shows photoreceptor-specific expression and regulates photoreceptor differentiation. Cell, 91, 531–541. 14. Freund, C.L., Gregory-Evans, C.Y., Furukawa, T., Papaioannou, M., Looser, J., Ploder, L., Bellingham, J., Ng, D., Herbrick, J.A., Duncan, A. et al. (1997) Cone-rod dystrophy due to mutations in a novel photoreceptor-specific homeobox gene (CRX) essential for maintenance of the photoreceptor. Cell, 91, 543–553. 15. Mathers, P.H., Grinberg, A., Mahon, K.A. and Jamrich, M. (1997) The Rx homeobox gene is essential for vertebrate eye development. Nature, 387, 603–607. 16. Furukawa, T., Kozak, C.A. and Cepko, C.L. (1997) Rax, a novel paired-type homeobox gene, shows expression in the anterior neural fold and developing retina. Proc. Natl Acad. Sci. USA, 94, 3088–3093. 17. Hatakeyama, J., Tomita, K., Inoue, T. and Kageyama, R. (2001) Roles of homeobox and bHLH genes in specification of a retinal cell type. Development, 128, 1313–1322. 18. Eggert, T., Hauck, B., Hildebrandt, N., Gehring, W.J. and Walldorf, U. (1998) Isolation of a Drosophila homolog of the vertebrate homeobox gene Rx and its possible role in brain and eye development. Proc. Natl Acad. Sci. USA, 95, 2343–2348.
Downloaded from http://hmg.oxfordjournals.org/ by guest on June 7, 2013
Co-immunoprecipitation experiments using in vitro translated wild-type and mutant QRX (35S-labeled) and Crx were carried out using anti-Crx antibody P261, essentially as described (73) with minor modifications. The coupling time with Protein-A beads was reduced to 3 h instead of overnight, and a cocktail of protease inhibitors (Roche Molecular Biochemicals) was added to each reaction. In vivo expression of recombinant QRX-Xp (Xpress tagged hQRX) and CRX-Flag (Flag-tagged hCRX) were carried out by individually transfecting HEK293 cells cultured on 100 mm plates with 5 mg of hCRX-Flag and hQRX-pcDNA3.1/HisB, respectively. Pull-down assays were performed with the above 293 cell extracts after the cells were washed twice with PBS (pH 7.4) and lysed 48 h post transfection in 1 ml of lysis buffer containing 50 mM Tris–Cl (pH 7.4), 150 mM NaCl, 1 mM EDTA, 0.1% NP40 and 10 ml of a protease inhibitor cocktail (Sigma) at 4 C for 20 min. Soluble cell extracts (950 ml) containing CRX-Flag, or empty vector, were incubated with 40 ml of anti-Flag M2 affinity resin equilibrated with TBS (50 mM Tris–Cl, pH 7.4, 150 mM NaCl) at 4 C for 3 h with gentle agitation. The bound resin was spun down and washed three times with 0.5 ml of TBS followed by incubation with cell extracts containing QRX-Xp for an additional 3 h at 4 C. After 3 washes with 1 ml TBS, proteins bound to the resin were eluted with 100 ml of 3 Flag peptide (150 ng/ml in TBS) and analyzed using SDS–PAGE and western blots with primary antibodies specific to CRX (73) or Xpress (the expression tag for hQRX), and secondary goat anti-mouse IgG-HRP (for antiXpress) or anti-rabbit IgG-HRP (for anti-CRX), respectively. ECL plus western blotting detection reagents and Hyperfilm for ECL were used for detection.
RT–PCR analysis and Chen-wei Wang for technical assistance. The research is supported in part by grants from the National Eye Institute, Steinbach Foundation, Foundation Fighting Blindness and Macula Vision Foundation, and generous gifts from Mr and Mrs Stevie and Marshall Wishnack and Mr and Mrs Robert and Clarice Smith. D.J.Z. is the Guerrieri Professor of Genetic Engineering and Molecular Ophthalmology, and the recipient of a Research to Prevent Blindness Senior Investigator Award.
Human Molecular Genetics, 2004, Vol. 13, No. 10
38. Passini, M.A., Levine, E.M., Canger, A.K., Raymond, P.A. and Schechter, N. (1997) Vsx-1 and Vsx-2: differential expression of two paired-like homeobox genes during zebrafish and goldfish retinogenesis. J. Comp. Neurol., 388, 495–505. 39. Levine, E.M., Passini, M., Hitchcock, P.F., Glasgow, E. and Schechter, N. (1997) Vsx-1 and Vsx-2: two Chx10-like homeobox genes expressed in overlapping domains in the adult goldfish retina. J. Comp. Neurol., 387, 439–448. 40. Kalderon, D., Roberts, B.L., Richardson, W.D. and Smith, A.E. (1984) A short amino acid sequence able to specify nuclear location. Cell, 39, 499–509. 41. Fei, Y. and Hughes, T.E. (2000) Nuclear trafficking of photoreceptor protein crx: the targeting sequence and pathologic implications. Invest. Ophthalmol. Vis. Sci., 41, 2849–2856. 42. Altschul, S.F., Gish, W., Miller, W., Myers, E.W. and Lipman, D.J. (1990) Basic local alignment search tool. J. Mol. Biol., 215, 403–410. 43. Zhang, Z., Schaffer, A.A., Miller, W., Madden, T.L., Lipman, D.J., Koonin, E.V. and Altschul, S.F. (1998) Protein sequence similarity searches using patterns as seeds. Nucl. Acids Res., 26, 3986–3990. 44. Simeone, A., D’Apice, M.R., Nigro, V., Casanova, J., Graziani, F., Acampora, D. and Avantaggiato, V. (1994) Orthopedia, a novel homeobox-containing gene expressed in the developing CNS of both mouse and Drosophila. Neuron, 13, 83–101. 45. Norris, R.A. and Kern, M.J. (2001) Identification of domains mediating transcription activation, repression, and inhibition in the paired-related homeobox protein, Prx2 (S8). DNA Cell Biol., 20, 89–99. 46. Venter, J.C., Adams, M.D., Myers, E.W., Li, P.W., Mural, R.J., Sutton, G.G., Smith, H.O., Yandell, M., Evans, C.A., Holt, R.A. et al. (2001) The sequence of the human genome. Science, 291, 1304–1351. 47. Lander, E.S., Linton, L.M., Birren, B., Nusbaum, C., Zody, M.C., Baldwin, J., Devon, K., Dewar, K., Doyle, M., FitzHugh, W. et al. (2001) Initial sequencing and analysis of the human genome. Nature, 409, 860–921. 48. Mural, R.J., Adams, M.D., Myers, E.W., Smith, H.O., Miklos, G.L., Wides, R., Halpern, A., Li, P.W., Sutton, G.G., Nadeau, J. et al. (2002) A comparison of whole-genome shotgun-derived mouse chromosome 16 and the human genome. Science, 296, 1661–1671. 49. DesJardin, L.E., Timmers, A.M. and Hauswirth, W.W. (1993) Transcription of photoreceptor genes during fetal retinal development. Evidence for positive and negative regulation. J. Biol. Chem., 268, 6953–6960. 50. den Dunnen, J.T. and Antonarakis, S.E. (2001) Nomenclature for the description of human sequence variations. Hum. Genet., 109, 121–124. 51. Wang, X., Xu, S., Rivolta, C., Li, L.Y., Peng, G.H., Swain, P.K., Sung, C.H., Swaroop, A., Berson, E.L., Dryja, T.P. et al. (2002) Barrier to autointegration factor interacts with the cone-rod homeobox and represses its transactivation function. J. Biol. Chem., 277, 43288–43300. 52. Mitton, K.P., Swain, P.K., Khanna, H., Dowd, M., Apel, I.J. and Swaroop, A. (2003) Interaction of retinal bZIP transcription factor NRL with Flt3-interacting zinc-finger protein Fiz1: possible role of Fiz1 as a transcriptional repressor. Hum. Mol. Genet., 12, 365–373. 53. Martinez, J.A. and Barnstable, C.J. (1998) Erx, a novel retina-specific homeodomain transcription factor, can interact with Ret 1/PCEI sites. Biochem. Biophys. Res. Commun., 250, 175–180. 54. Tomita, K., Nakanishi, S., Guillemot, F. and Kageyama, R. (1996) Mash1 promotes neuronal differentiation in the retina. Genes Cells, 1, 765–774. 55. Arranz, V., Dreuillet, C., Crisanti, P., Tillit, J., Kress, M. and Ernoult-Lange,M. (2001) The zinc finger transcription factor, MOK2, negatively modulates expression of the interphotoreceptor retinoid-binding protein gene, IRBP. J. Biol. Chem., 276, 11963–11969. 56. Milam, A.H., Rose, L., Cideciyan, A.V., Barakat, M.R., Tang, W.X., Gupta, N., Aleman, T.S., Wright, A.F., Stone, E.M., Sheffield, V.C. et al. (2002) The nuclear receptor NR2E3 plays a role in human retinal photoreceptor differentiation and degeneration. Proc. Natl Acad. Sci. USA, 99, 473–478. 57. Ng, L., Hurley, J.B., Dierks, B., Srinivas, M., Salto, C., Vennstrom, B., Reh, T.A. and Forrest, D. (2001) A thyroid hormone receptor that is required for the development of green cone photoreceptors. Nat. Genet., 27, 94–98. 58. Lerner, L.E., Gribanova, Y.E., Ji, M., Knox, B.E. and Farber, D.B. (2001) Nrl and Sp nuclear proteins mediate transcription of rod-specific cGMP- phosphodiesterase beta-subunit gene: involvement of multiple response elements. J. Biol. Chem., 276, 34999–35007.
Downloaded from http://hmg.oxfordjournals.org/ by guest on June 7, 2013
19. Furukawa, T., Mukherjee, S., Bao, Z.Z., Morrow, E.M. and Cepko, C.L. (2000) Rax, Hes1, and notch1 promote the formation of Muller glia by postnatal retinal progenitor cells. Neuron, 26, 383–394. 20. Mathers, P.H. and Jamrich, M. (2000) Regulation of eye formation by the Rx and pax6 homeobox genes. Cell Mol. Life Sci., 57, 186–194. 21. Chuang, J.C. and Raymond, P.A. (2001) Zebrafish genes rx1 and rx2 help define the region of forebrain that gives rise to retina. Dev. Biol., 231, 13–30. 22. Tucker, P., Laemle, L., Munson, A., Kanekar, S., Oliver, E.R., Brown, N., Schlecht, H., Vetter, M. and Glaser, T. (2001) The eyeless mouse mutation (ey1) removes an alternative start codon from the Rx/rax homeobox gene. Genesis, 31, 43–53. 23. Kimura, A., Singh, D., Wawrousek, E.F., Kikuchi, M., Nakamura, M. and Shinohara, T. (2000) Both PCE-1/RX and OTX/CRX interactions are necessary for photoreceptor-specific gene expression. J. Biol. Chem., 275, 1152–1160. 24. Furukawa, T., Morrow, E.M., Li, T., Davis, F.C. and Cepko, C.L. (1999) Retinopathy and attenuated circadian entrainment in Crx-deficient mice. Nat. Genet., 23, 466–470. 25. Livesey, F.J., Furukawa, T., Steffen, M.A., Church, G.M. and Cepko, C.L. (2000) Microarray analysis of the transcriptional network controlled by the photoreceptor homeobox gene Crx. Curr. Biol., 10, 301–310. 26. Furukawa, A., Koike, C., Lippincott, P., Cepko, C.L. and Furukawa, T. (2002) The mouse Crx 50 -upstream transgene sequence directs cell-specific and developmentally regulated expression in retinal photoreceptor cells. J. Neurosci., 22, 1640–1647. 27. Swain, P.K., Chen, S., Wang, Q.L., Affatigato, L.M., Coats, C.L., Brady, K.D., Fishman, G.A., Jacobson, S.G., Swaroop, A., Stone, E. et al. (1997) Mutations in the cone-rod homeobox gene are associated with the cone-rod dystrophy photoreceptor degeneration. Neuron, 19, 1329–1336. 28. Freund, C.L., Wang, Q.L., Chen, S., Muskat, B.L., Wiles, C.D., Sheffield, V.C., Jacobson, S.G., McInnes, R.R., Zack, D.J. and Stone, E.M. (1998) De novo mutations in the CRX homeobox gene associated with Leber congenital amaurosis. Nat. Genet., 18, 311–312. 29. Swaroop, A., Wang, Q.L., Wu, W., Cook, J., Coats, C., Xu, S., Chen, S., Zack, D.J. and Sieving, P.A. (1999) Leber congenital amaurosis caused by a homozygous mutation (R90W) in the homeodomain of the retinal transcription factor CRX: direct evidence for the involvement of CRX in the development of photoreceptor function. Hum. Mol. Genet., 8, 299–305. 30. Sohocki, M.M., Sullivan, L.S., Mintz-Hittner, H.A., Birch, D., Heckenlively, J.R., Freund, C.L., McInnes, R.R. and Daiger, S.P. (1998) A range of clinical phenotypes associated with mutations in CRX, a photoreceptor transcription-factor gene. Am. J. Hum. Genet., 63, 1307–1315. 31. Swaroop, A., Xu, J.Z., Pawar, H., Jackson, A., Skolnick, C. and Agarwal, N. (1992) A conserved retina-specific gene encodes a basic motif/leucine zipper domain. Proc. Natl Acad. Sci. USA, 89, 266–270. 32. Mitton, K.P., Swain, P.K., Chen, S., Xu, S., Zack, D.J. and Swaroop, A. (2000) The leucine zipper of NRL interacts with the CRX homeodomain. A possible mechanism of transcriptional synergy in rhodopsin regulation. J. Biol. Chem., 275, 29794–29799. 33. Bessant, D.A., Payne, A.M., Mitton, K.P., Wang, Q.L., Swain, P.K., Plant, C., Bird, A.C., Zack, D.J., Swaroop, A. and Bhattacharya, S.S. (1999) A mutation in NRL is associated with autosomal dominant retinitis pigmentosa. Nat. Genet., 21, 355–356. 34. Swain, P.K., Hicks, D., Mears, A.J., Apel, I.J., Smith, J.E., John, S.K., Hendrickson, A., Milam, A.H. and Swaroop, A. (2001) Multiple phosphorylated isoforms of NRL are expressed in rod photoreceptors. J. Biol. Chem., 276, 36824–36830. 35. Mears, A.J., Kondo, M., Swain, P.K., Takada, Y., Bush, R.A., Saunders, T.L., Sieving, P.A. and Swaroop, A. (2001) Nrl is required for rod photoreceptor development. Nat. Genet., 29, 447–452. 36. Liu, I.S., Chen, J.D., Ploder, L., Vidgen, D., van der Kooy, D., Kalnins, V.I. and McInnes, R.R. (1994) Developmental expression of a novel murine homeobox gene (Chx10): evidence for roles in determination of the neuroretina and inner nuclear layer. Neuron, 13, 377–393. 37. Hayashi, T., Huang, J. and Deeb, S.S. (2000) RINX(VSX1), a novel homeobox gene expressed in the inner nuclear layer of the adult retina. Genomics, 67, 128–139.
1039
1040
Human Molecular Genetics, 2004, Vol. 13, No. 10
67. O’Brien, S.J., Menotti-Raymond, M., Murphy, W.J., Nash, W.G., Wienberg, J., Stanyon, R., Copeland, N.G., Jenkins, N.A., Womack, J.E. and Marshall Graves, J.A. (1999) The promise of comparative genomics in mammals. Science, 286, 458–462, 479–481. 68. Yanagi, Y., Masuhiro, Y., Mori, M., Yanagisawa, J. and Kato, S. (2000) p300/CBP acts as a coactivator of the cone-rod homeobox transcription factor. Biochem. Biophys. Res. Commun., 269, 410–414. 69. Schedl, A., Ross, A., Lee, M., Engelkamp, D., Rashbass, P., van Heyningen, V. and Hastie, N.D. (1996) Influence of PAX6 gene dosage on development: overexpression causes severe eye abnormalities. Cell, 86, 71–82. 70. Schaeren-Wiemers, N. and Gerfin-Moser, A. (1993) A single protocol to detect transcripts of various types and expression levels in neural tissue and cultured cells: in situ hybridization using digoxigenin-labelled cRNA probes. Histochemistry, 100, 431–440. 71. Chen, S. and Zack, D.J. (2000) Cloning and characterization of retinal transcription factors, using target site-based methodology. Methods Enzymol., 316, 590–610. 72. Bassam, B.J., Caetano-Anolles, G. and Gresshoff, P.M. (1991) Fast and sensitive silver staining of DNA in polyacrylamide gels. Anal. Biochem., 196, 80–83. 73. La Spada, A.R., Fu, Y.H., Sopher, B.L., Libby, R.T., Wang, X., Li, L.Y., Einum, D.D., Huang, J., Possin, D.E., Smith, A.C. et al. (2001) Polyglutamine-expanded ataxin-7 antagonizes CRX function and induces cone-rod dystrophy in a mouse model of SCA7. Neuron, 31, 913–927
Downloaded from http://hmg.oxfordjournals.org/ by guest on June 7, 2013
59. Halder, G., Callaerts, P. and Gehring, W.J. (1995) Induction of ectopic eyes by targeted expression of the eyeless gene in Drosophila. Science, 267, 1788–1792. 60. Sheng, G., Thouvenot, E., Schmucker, D., Wilson, D.S. and Desplan, C. (1997) Direct regulation of rhodopsin 1 by Pax-6/eyeless in Drosophila: evidence for a conserved function in photoreceptors. Genes Dev., 11, 1122–1131. 61. Punzo, C., Seimiya, M., Flister, S., Gehring, W.J. and Plaza, S. (2002) Differential interactions of eyeless and twin of eyeless with the sine oculis enhancer. Development, 129, 625–634. 62. Chuang, J.C., Mathers, P.H. and Raymond, P.A. (1999) Expression of three Rx homeobox genes in embryonic and adult zebrafish. Mech. Dev., 84, 195–198. 63. Chen, C.M. and Cepko, C.L. (2002) The chicken RaxL gene plays a role in the initiation of photoreceptor differentiation. Development, 129, 5363–5375. 64. Easteal, S. (1988) Rate constancy of globin gene evolution in placental mammals. Proc. Natl Acad. Sci. USA, 85, 7622–7626. 65. Murphy, W.J., Eizirik, E., Johnson, W.E., Zhang, Y.P., Ryder, O.A. and O’Brien, S.J. (2001) Molecular phylogenetics and the origins of placental mammals. Nature, 409, 614–618. 66. Murphy, W.J., Eizirik, E., O’Brien, S.J., Madsen, O., Scally, M., Douady, C.J., Teeling, E., Ryder, O.A., Stanhope, M.J., de Jong, W.W. et al. (2001) Resolution of the early placental mammal radiation using Bayesian phylogenetics. Science, 294, 2348–2351.
