Abstract

Cone–rod homeobox (CRX), a paired-like homeobox transcription factor, plays a major role in photoreceptor development and maintenance of the retina. Fifteen different mutations in the CRX gene have been identified as a cause of blinding retinal dystrophy. As a step towards characterizing the underlying pathophysiology of disease, temporal and spatial gene expression patterns during human and mouse eye development were investigated for CRX and for downstream retinally expressed genes, postulated to be transactivated by CRX. We found that human CRX was expressed at 10.5 weeks post-conception (p.c.). This was significantly later than observed in mouse development. Immunocytochemistry in human retina showed that CRX protein was not detected until >4 weeks later at 15 weeks p.c., implying that it would be unable to transactivate PDEB, IRBP and arrestin, which were all expressed before 15 weeks. These data therefore eliminate CRX as the major transcriptional activator of these three genes from a wide group of retinal genes that can be transactivated by CRX in vitro. Additionally, PDEB was expressed 2 weeks before CRX whereas murine Pdeb was expressed after Crx, highlighting a potential difference for the role of PDEB in human eye development. Previous data had shown CRX expression in the adult human retina to be photoreceptor-specific; however, we demonstrate that this gene is also expressed in the inner nuclear layer (INL) of the human and mouse retina by insitu hybridization and immunocytochemistry. INL localization of murine Crx was confirmed in rd/rd,cl mice, as in this mouse model the photoreceptors are absent. We have found important differences in the temporal expression of this gene in human and mouse retina, although spatial expression of the CRX gene appears to be conserved. In addition, downstream targets of CRXin vitro might not represent in vivo function during development. These data support concerns about the extent to which we can extrapolate from rodent models regarding embryonic development and disease pathophysiology.

Received April 11, 2001; Revised and Accepted June 1, 2001.

INTRODUCTION

The process of retinal development involves a hierarchy of transcription factors regulating increasingly complex programmes of gene expression and cell fate determination (1,2). One such transcription factor is the cone–rod homeobox (CRX) gene (35). CRX is a member of the OTX-like homeobox gene family, a group of transcription factors that encode paired-like (prd) homeodomain proteins. The OTX proteins are involved in the regulation of anterior head structure and sensory organ development (6). Localized to human chromosome 19q13.3, mutations in the CRX gene have been found to cause autosomal dominant cone-rod dystrophy (4,79) and autosomal recessive Leber congenital amaurosis (8,1012). Both these diseases are clinically and genetically heterogeneous. Cone-rod dystrophy is a severe form of retinal dystrophy, which often begins in the second decade of life and is untreatable and incurable (13). Leber congenital amaurosis is characterized by total blindness at birth or shortly thereafter (14), with no detectable electroretinogram (derived from photoreceptor cells).

The CRX protein binds to a conserved site or CRX-binding element (CBE 1), C/TTAATC/T, which is present in the upstream region of many photoreceptor-specific genes such as rhodopsin, arrestin and interphotoreceptor retinoid-binding protein (IRBP) (3,5). Recently, a second and weaker CBE-like motif has been identified, giving an 11-base motif in a head-to-tail arrangement with CBE1 (C/TTAATC/TG/AGA/CTT/C). This motif has been identified in the upstream regions of retinally expressed genes, which are down-regulated in the absence of Crx in mouse retina (15,16). In vitro, CRX is capable of transactivating retinal gene reporter-constructs carrying the C/TTAATC/T binding site (3,5); however, transactivation by CRX invivo remains to be confirmed.

The CRX transcription factor has been shown to be critical for the differentiation of photoreceptors during retinal development and also the maintenance of these structures in adult tissue (3). Over-expression of retroviral Crx in developing retinal cells results in a marked increase in rod photoreceptors and an almost complete absence of amacrine interneurons (3), highlighting the importance of this transcription factor in morphogenesis and the determination of cell function. Targeted disruption of murine Crx results in the failure of photoreceptor outer segment (OS) growth and thus the absence of phototransduction (16). Characterization of gene expression in the Crx null mouse has indicated that a number of photoreceptor genes are down-regulated in the absence of Crx (16), and the majority of these genes are crucial either to phototransduction or to the structural formation of the photoreceptor cells.

Studying the temporal and spatial expression of developmental genes can be used for dissecting biological development and disease pathophysiology (17). To date in adult tissues, CRX expression has been shown to be specific to the photoreceptor cell layer of the retina (35) and the pinealocytes of the brain (5). In the developing murine retina, Crx expression, as detected by in situ hybridization, is observed at embryonic day (E)12.5, coinciding with cone photoreceptor cell genesis (3). At postnatal day 6 (P6), peak expression is seen in correlation with the increase in photoreceptor cells expressing rhodopsin and other phototransduction genes. The expression pattern of CRX during human development has not been reported, and the data on expression of some of the retinal genes that CRX is believed to regulate is limited. In the present study, RT–PCR and in situ hybridization analysis were carried out to evaluate temporal and spatial gene expression during human retinal development, and immunocytochemistry was undertaken to confirm protein expression. Much of what is reported about this gene’s function is based on mouse studies. Comparative gene expression patterns in the retina were therefore investigated to highlight any potential human/mouse differences in the expression of this and related genes during development.

RESULTS

RT–PCR analysis of CRX expression in the human and mouse developing eye

To identify the point during human eye development at which CRX is expressed, RT–PCR analysis was carried out on total RNA extracted from human fetal and adult eyes. Initially, the sine oculis homeobox gene 3 (SIX3/Six3) was amplified from both human and mouse cDNA as a positive control. SIX3 is expressed in the developing human eye as early as 5–7 weeks gestation and expression is also maintained in mature ocular tissue (18). In the mouse, Six3 expression is seen in the optic vesicles at E9.5 and by E11.5 expression is observed in the neural retina (19). SIX3/Six3 was present in all ages tested of both human (Fig. 1A) and mouse cDNA (Fig. 2A), confirming RNA integrity of all samples. Exonic primers were designed to span intron 1 of the CRX gene, to avoid the possibility of amplifying from contaminating genomic DNA. Amplification of the CRX gene produced positive results from 10.5, 11.0 and 13.5 weeks post-conception (p.c.) and from the adult control (Fig. 1B). It was not possible to amplify CRX from either of the two younger fetal samples (8.5 and 9.5 weeks p.c.) and therefore the onset of CRX expression must occur between 9.5 and 10.5 weeks of development. Gene expression is presumably maintained from 10.5 weeks to adult, as shown with CRX expression studies in the mouse (3). An identical procedure was used to amplify Crx from embryonic mouse eyes. Expression was seen first at E10.5 and this expression was maintained in all cDNAs tested, both embryonic and postnatal (Fig. 2B).

Retinal-specific gene expression

A number of retinal-specific genes containing CRX binding sites in their promoter were also analysed by RT–PCR to determine their onset of expression relative to that of CRX. If CRX was expressed before these genes, it is possible that CRX could transactivate them in vivo. Arrestin was expressed at 13.5 weeks p.c. in the human eye (Fig. 1C) and at E14.5 in the mouse eye (Fig. 2C), although a low level of expression was observed. Human IRBP is expressed at the same time as CRX (10.5 weeks), therefore CRX could potentially transactivate IRBPin vivo (Fig. 1D). Irbp in mouse was detected at E10.5, the same time as Crx is first expressed (Fig. 2D), thus showing conserved temporal expression of IRBP/Irbp and CRX/Crx between man and mouse. A striking result was obtained for PDEB in the human retina, where itwas expressed from an early stage (8.5 weeks p.c.) of development (Fig. 1E) preceding CRX expression by 2 weeks, whereas expression of Pdeb in the murine retina begins at E11.5 (Fig. 2E), after Crx gene expression was first detected. Hence, comparative temporal expression of PDEB/Pdeb and CRX/Crx is not conserved between species.

Rhodopsin expression was absent from all human fetal cDNAs, but was observed in the adult retinal control cDNA (Fig. 1F). In the mouse, rhodopsin expression was only observed postnatally, from P5 onwards, corresponding with rod photoreceptor OS development (Fig. 2F). In other species such as rat (20) and in human fetal retina (21) opsins have been detected before birth by immunocytochemical studies. Other studies in rats report later onset of expression (22). These differences may be due in part to detecting cell-specific expression, rather than a more global RT–PCR approach. We have tested our 13.5 and 15 week retinal sections for rhodopsin expression by in situ hybridization and found no expression (data not shown). Retinoschisin (RS1) was not expressed in any of the human fetal ages tested and neither were the α- and γ-subunits of phosphodiesterase (PDEA/PDEG), nor green or blue cone opsin (Fig. 1G–K). Comparable results were observed with mouse samples, with each gene having a specific time-of-onset during postnatal retinal maturation (Fig. 2G–K). Expression of NRL was not seen at any of the developmental stages tested in human fetal retina (Fig. 1L) and in mouse it was first observed at P1 (Fig. 2L).

In situ analysis of CRX expression in human fetal and adult retina

The expression of CRX was further investigated in the developing human retina by in situ hybridization analysis. Retinal sections of 10 and 12 weeks of development were tested, since the RT–PCR data had shown CRX expression at these times. However, we were unable to detect expression of CRX, indicating that it is likely that the in situ analysis was not sensitive enough to detect CRX expression at low levels. At 13 weeks p.c., CRX expression was observed in the neural retina adjacent to the retinal pigment epithelium (RPE), where newly born photoreceptors are present (Fig. 3A). At 15 weeks p.c., CRX expression was much stronger and was localized specifically to the developing outer nuclear layer (ONL) of the retina (Fig. 3B), consistent with the in situ expression pattern in mouse embryonic retina (3). In adult tissue CRX is expressed in the ONL of the retina, which contains the nuclei of the rod and cone cells (Fig. 3C). However, expression was not limited to the ONL of the adult retina, as staining was also detected within the inner nuclear layer (INL) of the retina. Whereas the staining in the ONL was strong and diffuse, presumably in both the rod and cone nuclei, staining in the INL was weaker and punctate, suggesting that CRX expression in the INL may be localized to a particular subset of retinal cells.

CRX immunocytochemistry in human retina

We generated an N-terminal peptide antibody, which detected a specific 37 kDa band by western blotting of total mouse retinal extract (Fig. 4A). The specificity of detection was confirmed by western blotting with pre-immune serum (which did not detect this band) and pre-incubation of the antibody with the peptide used to generate it (this removed the ability of the antibody to detect the band; data not shown). Using this antibody we first saw immunochemical localization of CRX protein in the fetal retina at 15 weeks p.c., where strong staining was present in the photoreceptor cells adjacent to the RPE (Fig. 4B). Weak staining of the ganglion cell layer (GCL) was also noted at 15 weeks p.c. Similar GCL staining was present in controls without primary antibody, suggesting non-specific staining by the secondary antibody. No staining of the GCL was detected by in situ hybridization (Fig. 3B) in the same age embryos. Using the same conditions for immunohistochemistry, we did not detect protein at 13 weeks p.c. (Fig. 4C). Thus, CRX protein was only detected 2 weeks after we could detect CRX mRNA by in situ hybridization (at 13 weeks p.c.). This apparent interval between mRNA expression and protein localization may, however, reflect a difference in threshold of sensitivity of in situ hybridization compared with the immunocytochemistry technique.

Crx gene and protein expression in rodless coneless (rd/rd,cl) retina

To corroborate CRX expression in the INL of human retina, we took advantage of the rd/rd,cl mouse, where the ONL of the retina degenerates completely by P80 due to a lesion of rods (rd/rd) and a transgenic ablation (cl) of cone photoreceptors (23). Crx was amplified from rd/rd,cl retinal cDNA and also from wild-type mouse retinal cDNA isolated at P80. Crx RT–PCR products were observed in both samples, confirming that Crx expression is not confined to the ONL of the retina (Fig. 5A). It was not possible to amplify rhodopsin, green or blue cone opsin from the rd/rd,cl mice (Fig. 5B), inferring that the RT-PCR products were not due to any remaining rod and cone photoreceptor nuclei.

To confirm that Crx expression in rd/rd,cl mice was localized to the INL, in situ hybridization was performed on eyes taken from adult rd/rd,cl mice, which were then compared with wild-type eyes. The spatial expression of Crx in wild-type eyes reflected that of the human adult retina, as staining was observed in both the ONL and INL of the retina (Fig. 6A). It was evident that INL Crx staining was noticeably weaker than that of the ONL, consistent with the human data (compare Fig. 6A with Fig. 3C). Expression was also observed in the INL of the rd/rd,cl mice (Fig. 6B), confirming the RT–PCR data. By immunocytochemistry, Crx protein was localized to both the ONL and INL in wild-type adult mouse retina (Fig. 6C). In the rd/rd,cl mouse retina, Crx protein was localized to the INL (Fig. 6D). Interestingly, by both in situ hybridization and immunocytochemistry, the INL staining of the rd/rd,cl retina was stronger than previously observed in either the INL of the adult human sections or the wild-type mouse. This suggests that either the ONL acts as a competitor for the Crx probe during hybridization or, alternatively, that Crx expression is up-regulated in mice lacking rod and cone photoreceptors. To investigate this possibility, sections of rd/rd,cl eyes were taken before (P9) and after (P13) retinal degeneration had started. By immunocytochemistry we were unable to detect up-regulation of Crx protein, with similar levels of Crx detected at each time point examined (data not shown).

DISCUSSION

Between 8 and 20 weeks of fetal development, the human retina progresses from a basic two-layered structure comprising the inner and outer neuroblastic layers to a highly differentiated and multicellular structure (24,25). During this time the cells destined to become photoreceptors begin to develop in the outer neural layer of the retina. Cone nuclei are born earlier than those of rod photoreceptors, first appearing at ∼10 weeks of development during formation of the fovea (21). Rod nuclei, however, are born a little later than this and are subsequently not present until ∼12 weeks of development (24,25). Our own observations show that CRX is not expressed until 10.5 weeks in the human retina. Therefore, it is likely that the first cells to express CRX in the developing human retina are the newly born cone photoreceptor nuclei. Continued CRX expression in the outer retina coincides with cone and rod specification and structural maturation of the photoreceptors. It is important to note that the onset of expression occurs later in human development than it does in the mouse. In the mouse retina we detected Crx expression at E10.5. This equates to ∼30 days (∼4 weeks) in human development based on Carnegie comparisons, highlighting a difference in temporal CRX expression between species. At birth the photoreceptors are fully differentiated in the human retina. In mouse retina, however, rod maturation is not complete until ∼P16 (26). The role of CRX in the timing of differentiation is therefore different in mouse than human, suggesting that other factors/cues are involved.

The early expression of PDEB was unexpected and suggests that transactivation of PDEB occurs independently of CRX. Furthermore, presence of a CRX binding site in the promoter region of a gene, as is the case for PDEB, should not be regarded as evidence that CRX transactivates the gene in vivo, even though it is possible to drive expression of a PDEB reporter construct with CRX (5). Previous work on PDEB has indicated the presence of a conserved AP-1 element in both the human and the mouse proximal 5′ regions of the gene (27). Additionally, mutation of the CRX binding site in the PDEB promoter had no effect on transcription of a reporter gene, suggesting that CRX is not important in transactivation of PDEB (27). The reason for such an early expression of PDEB is unclear, although this observation suggests that the PDEB may have a role during the development of the human retina that is alternative to its role in phototransduction in mature retina. The early expression of Pdeb is not mirrored within mouse retinal development, where it occurs after Crx gene expression. This highlights a potential difference for the role of PDEB/Pdeb between these two species.

The expression pattern of NRL was included in this study, as previous work has suggested that CRX and NRL work synergistically in the activation of a number of retinal targets (5). Expression of NRL was not detected in any of the fetal tissues we tested. Therefore, if CRX is responsible for activation of retinal genes expressed at 13.5 weeks or earlier, such as IRBP and arrestin, then it must do so independently of NRL. Very recent data have shown that Nrl null mice completely lack rod photoreceptors and imply that Nrl plays an important role in directing gene expression in rod photoreceptors (28). Nrl expression has been detected slightly earlier than in our study at E18.5 in mice, in line with our view that Crx is acting independently of Nrl early in development. Another anomaly we found with Nrl expression was that in mouse, Nrl was expressed much later (P1) than Pdeb (E12). However, in rat retina it has been shown by northern blotting that Nrl expression precedes Pdeb by two postnatal days (22). Thus, even between species of the rodent lineage there seem to be differences in temporal gene expression.

Although, IRBP has a CRX binding site in its promoter, it seems unlikely that IRBP is itself regulated by CRXin vivo since it is not down-regulated in the absence of Crx in the null mouse (16). Furthermore, it has been shown recently that OTX2 binds specifically to the IRBP promoter in yeast one-hybrid studies, whereas CRX was not identified (29). Thus, IRBP is another example of a gene containing a CRX binding site in its promoter, which can be transactivated by CRX in vitro; however, other studies show that CRX is unlikely to be the main transcription factor involved in IRBP gene expression.

Whether CRX can function as a transcriptional activator in vivo relies on the protein being present, rather than just evidence that the gene is being expressed. We detected CRX expression by RT–PCR at 10.5 weeks p.c. and by in situ hybridization at 13 weeks. In contrast, significant levels of CRX protein were only detected at 15 weeks p.c., whereas no protein was detectable at 13 weeks p.c. using the same immunohistochemical conditions. Although limitations to the sensitivity of the immunostaining mean that we cannot rule out the presence of low levels of protein before 15 weeks, our data does point to the possibility of post-transcriptional control of CRX expression.

Transcript localization and translational regulation are two post-transcriptional mechanisms for the spatial and temporal regulation of protein production (30). There are a number of examples showing that even though a gene is transcribed (e.g. nanos and oskar mRNAs), and is localized in the right cells, it is translationally repressed by proteins binding to cis-acting elements in 3′-untranslated region (3′-UTR) sequences (31,32). Translation is activated at the appropriate time by proteins binding to 5′-UTR sequences (33,34). Thus, we propose that retinally-expressed genes transcribed before 15 weeks p.c. in human fetal retina occur independently of CRX. Thus, in addition to PDEB and IRBP, we would exclude arrestin from transactivation by CRX in vivo.

The in situ hybridization and immunocytochemistry data convincingly demonstrate that the CRX gene and protein are expressed in the INL of both human and mouse retina. This is consistent with previous expression data observed in mouse (35) and zebrafish retina (36). INL expression of CRX has been previously documented in the retina of P6 mice (3) and this was thought to correspond to developing photoreceptors trapped on the vitreal side of the inner plexiform layer (37). However, we have clearly shown INL layer CRX expression in human adult retina. This was verified in the rd/rd,cl mouse retina which has no photoreceptor cells present. It is clear that expression of retinal CRX is not exclusive to the photoreceptors of the ONL and that its role in development may not be limited to differentiation and regulation of the photoreceptors as previously thought. The exact role of CRX in the inner retina remains to be determined. However, in Crx null mice circadian entrainment is attenuated (16). This is the only retinal modification, other than complete lesion, that has been shown to impair circadian photo-entrainment in mammals. This suggests that Crx may have a role, directly or indirectly, in circadian rhythmicity, and perhaps this is mediated via cells of the inner retina. We are following a number of lines of investigation to test this hypothesis.

The reliance on extrapolation from rodent embryonic expression studies to humans has been justified by strong similarities in the organization, morphology and evolutionary conservation of many key genes. However, recent studies have shown that there are both temporal and spatial differences between species (38). This could explain why attempts to make mouse models by gene targeting often produce either no phenotype or phenotypes that do not resemble the human condition (39). The extent to which our findings represent true functional differences for CRX remains to be determined. The Crx null mouse shows attenuated circadian photo-entrainment, whereby it takes approximately twice as long to entrain to a new light stimulus compared with wild-type mice (16). No circadian rhythm disturbance, such as abnormal sleep pattern or symptoms akin to prolonged ‘jet-lag’, has been reported. Such disturbances, however, have not been thoroughly investigated in humans with the CRX mutation. If humans do not have these disturbances it would suggest, in the mouse, that Crx has other functions, or in humans, that there is a compensatory mechanism. Species differences in longevity or modifier genes could be involved in this human/mouse difference; however, this deserves further investigation. The current work emphasizes the value of comparative studies in validating animal models of human disease and the limited degree of extrapolation that can be drawn from in vitro experiments.

MATERIALS AND METHODS

Tissues utilized for studies

Human fetal tissue collected from social terminations of pregnancy was obtained from the MRC Tissue Bank, Hammersmith Hospital, London, with ethical approval. Staging of fetal embryos was by last menstrual period and crown–rump length. Specimens were transferred, <4 h post-operatively, to liquid N2 for RT–PCR, or to 4% (w/v) paraformaldehyde in phosphate-buffered saline (PBS) for fixation at 4°C for 18 h for in situ hybridization analysis. Mouse embryos were obtained from matings of C57BL/6 × CBA mice. The day on which the vaginal plug was detected was designated as E0.5. Tissue from rd/rd,cl mice was taken at P80 and tissue from age-matched wild-type eyes was taken when the ONL was absent by histological analysis.

Expression analysis using RT–PCR

Total RNA was extracted from human eyes using the RNAzol RNA extraction kit (Biogenesis), whilst TRIzol (Gibco BRL) was used to isolate total RNA from mouse eyes according to the manufacturer’s instructions. The SuperScript pre-amplification system (Gibco BRL) was used to generate cDNA for PCR from 1 µg of total RNA. Identical aliquots of cDNA were used for amplification using gene-specific primers. The CRX exon 1 forward primer (5′-CCCTGACTTGGGCCTCAGT-3′) and exon 2 reverse primer (5′-CCACCTCCTCACGGGCATA-3′) were designed to span intron 1 and amplify a product of 250 bp. The PCR profile typically incorporated an initial denaturing step at 94°C for 5 min followed by 30 cycles of denaturation at 94°C for 45 s, annealing at 55°C for 45 s and extension at 72°C for 45 s, with a final extension at 72°C for 5 min. The time of expression for each gene was tested on three independent tissue samples at each time point. Primers used to amplify other retinal genes were as follows:

Rhodopsin (436 bp), 5′-ACCAGATCACCTGCTCTAAC/5′-TCTTGGACACGGTAGCAGAGG;

PDEB (425 bp), 5′-AGGAGACCCTGAACATCTACC/5′-ATGAAGCCCACTTGCAGC;

PDEA (464 bp), 5′-CTCCATGGGTCCTCTATC/5′-CCGACTTGAAGCTTAGGG;

PDEG (594 bp), 5′-GCAGCAGGAGGGAGTC/5′-GACCAAGCCTCTCTGTGG;

arrestin (406 bp), 5′-AAAAAGTGCCACCAAACAGC/5′-GCGTCATTCTTGTCTCTCTTCC;

RS1 (434 bp), 5′-ACCAGATCACCTGCTCTAAC/5′-ACACTTGCTGACGCACTCC;

IRBP (1097 bp), 5′-CCTTTGCACACACCATGC/5′- CATGATATAGGTGAACTCC;

NRL (287 bp), 5′-GTGCCTCCTTCACCCACC/5′-CAGACATCGAGACCAGCG;

green opsin (231 bp), 5′-GCCCAGACGTGTTCAGCG/5′-GACCATCACCACCACCAT;

blue opsin (176 bp), 5′-GGTCACTGGCCTTCCTGG/5′-TGCAGGCCCTCAGGGATG;

SIX3 (242 bp), 5′-AGTCCACACACACTCCCAC/5′-CGTCATGCAGGTGGGGTCG.

Where primers designed against the human gene sequence were not compatible with mouse, specific primers to mouse homologous genes were used:

Pdeg (282 bp), 5′-GTCTCTGCCAGCCTCACC/5′-CTAAATGATGCCATACTGGG;

Pdea (717 bp), 5′-CTCCATGGGTCCTCCATC/5′-CTGGATGCAACAGGACTT;

Rs1 (434 bp), 5′-ATCAGATCACTTGCACCA/5′-ACACTTGCCGGCACACTC;

blue opsin (339 bp), 5′-CAGCCTTCATGGGATTTG/5′-GTGCATGCTTGGAGTTGA;

Irbp (427 bp), 5′-CCCTCCCCAGAAGTCTTT/5′-CAGCCTCTTCATGATGTA;

Nrl (287 bp), 5′-GTGCCTCCTTCACCCACC/5′-GCTGCCGGCAACTCGC;

arrestin (390 bp), 5′-AAAAAGTGCAGCCAAACAGC/5′-CATCTTTCTTCCCTTCTGTG.

In situ hybridization

Fresh human or murine eyes were fixed overnight in 4% paraformaldehyde in PBS and incubated overnight again in 20% sucrose in PBS at 4°C. The eyes were orientated and flash frozen in O.C.T. compound (BDH) using a chamber of dry ice and isopentane. In situ hybridization analysis was performed with digoxigenin-labelled riboprobes on 10 µm cryostat sagital sections through the central retina. The probes were generated from a pGEM-T plasmid (Promega) into which either a 445 bp fragment of the human 3′ CRX cDNA (forward primer, 5′-GGACTACAAGGATCAGAGTGCCTG-3′; reverse primer, 5′-GTTAAGTTATCAAGCCCCCCTACC-3′) or a 300 bp 3′ Crx mouse fragment (forward primer, 5′-CTCCTCCAGCTTAGATTC-3′; reverse primer, 5′-CCCGGAGTTCTAAGCCAA-3′) had been cloned. The plasmids were linearized with either NotI (for antisense probe) or NcoI (for sense probe). Digoxigenin-labelled riboprobes were generated using either SP6 RNA polymerase (antisense) or T7 RNA polymerase (sense) in the presence of digoxigenin RNA labelling mix (Roche) in accordance with the manufacturer’s instructions. Immediately before use, the probes were diluted to a concentration of 1 ng/µl in hybridization buffer (10 mM Tris–HCl pH 7.5, 200 mM NaCl, 5 mM NaH2PO4, 5 mM Na2HPO4, 5 mM EDTA, 1 mg/ml tRNA, 50% formamide, 10% dextran sulphate, 1× Denhardt’s). The probes were denatured for 5 min at 70°C and quenched on ice. Hybridization was performed overnight at 65°C in a humidified chamber of 50% formamide and 2× SSC. Immunodetection was carried out using a 1:1000 dilution of anti-digoxigenin alkaline phosphatase-conjugated Fab fragments (Roche). Hybrids were visualized with the BCIP/NBT substrate (Roche). Images were viewed using a Leitz Aristoplan microscope with digital image capture (Olympus DP10 camera). In situ experiments were repeated on three different eyes at each time point.

Generation of an N-terminal antibody to human CRX

A peptide was synthesized commercially using Fmoc chemistry (Sigma-Genosys) corresponding to the first 24 amino acids of CRX protein. The peptide was conjugated to keyhole limpet hemocyanin and then used to generate a rabbit polyclonal antibody by standard protocols. Serum samples were desalted prior to IgG fraction by DEAE affi-gel chromatography (Bio-Rad). Protein concentrations were estimated using a Bradford assay (Pierce).

Western blotting

Mouse retinal tissue was dissected in ice-cold PBS, added to solubilization buffer (150 mM NaCl, 1% Tergitol, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM Tris–HCl pH 8.0) containing protease inhibitor cocktail (Sigma) and kept on ice for 15 min. Tissue was sonicated, placed on ice for 10 min and then centrifuged at 10 000 g for 2 min. The supernatant was then mixed 1:1 with Laemmli sample buffer, denatured for 5 min at 100°C and the proteins were separated by 12% SDS–PAGE relative to low molecular weight standards (Biorad). Each sample loaded on a gel contained protein equivalent to one mouse retina in a 10 µl volume. Proteins were electrophoretically transferred onto nitrocellulose membrane (Schleicher and Schuell) using a Tris-glycine buffer system (48 mM Tris, 39 mM glycine, 0.037% SDS, 20% methanol). The membrane was pre-incubated in blocking buffer (5% non-fat milk, 0.1% Tween-20 in PBS) at 4°C for several hours followed by incubation with primary CRX antibody (diluted 1:250 in blocking buffer) for 1 h at room temperature. After three washes in PBS, 0.1% Tween-20, the membrane was incubated for 1 h at room temperature with a 1:3000 dilution of goat anti-rabbit conjugated-HRP in blocking buffer. After three final washes the peroxidase reaction was visualized by enhanced chemiluminescence (Amersham Pharmacia Biotech).

Immunocytochemistry

The sections for immunocytochemistry were taken from the same eyes that were used for in situ hybridization at each time point. To reduce non-specific labelling, 10 µm cryostat sections were incubated for 2 h at room temperature in blocking solution (PBS, 0.5% bovine serum albumin, 0.2% Triton X-100, 0.5% sodium azide) containing 2% normal donkey serum (Sigma). The sections were incubated with the primary CRX antibody in the same buffer (1:40) for 16 h at 4°C, rinsed in blocking solution and then incubated for 30 min at room temperature in secondary antibody (1:200) conjugated to fluorescein isothiocyanate (affinity-purified and species-absorbed donkey anti-rabbit IgG, Chemicon). Immunolabelled sections were mounted with Immunofluore (ICN) and were examined with a Leica laser scanning confocal microscope. Phase and fluorescence confocal images were exported to Photoshop 4.0 (Adobe) for annotation and dye-sublimation prints were generated.

ACKNOWLEDGEMENTS

The authors acknowledge the MRC Tissue Bank for provision of human fetal tissue. This work was supported by Fight for Sight prize studentships (L.C.B. and A.R.), British Retinitis Pigmentosa Society (GR518), Medical Research Council (J.C.S.) and the Child Health Research Appeal Trust (J.K.L.H.).

NOTE ADDED IN PROOF

A novel 12 bp deletion mutation in CRX has been identified by Tzekov, R.T. et al. (40).

+

To whom correspondence should be addressed. Tel: +44 20 7594 3007; Fax: +44 20 7594 3015; Email: c.gregory-evans@ic.ac.uk

Figure 1. RT–PCR analysis during early human eye development. Lanes 1–5 are cDNA samples from 8.5, 9.5, 10.5, 11.0, and 13.5 weeks p.c., respectively. Lane 6 is adult cDNA and lane 7 is a negative control. The retinally expressed genes tested by RT–PCR were (A) SIX3, (B) CRX, (C) arrestin, (D) IRBP, (E) PDEB, (F) rhodopsin, (G) RS1, (H) PDEA, (I) PDEG, (J) red/green cone opsin, (K) blue cone opsin and (L) NRL. The DNA marker for (A) and (H–L) is ϕX174 (Gibco) and for (B–G) is 100 bp ladder (Gibco).

Figure 1. RT–PCR analysis during early human eye development. Lanes 1–5 are cDNA samples from 8.5, 9.5, 10.5, 11.0, and 13.5 weeks p.c., respectively. Lane 6 is adult cDNA and lane 7 is a negative control. The retinally expressed genes tested by RT–PCR were (A) SIX3, (B) CRX, (C) arrestin, (D) IRBP, (E) PDEB, (F) rhodopsin, (G) RS1, (H) PDEA, (I) PDEG, (J) red/green cone opsin, (K) blue cone opsin and (L) NRL. The DNA marker for (A) and (H–L) is ϕX174 (Gibco) and for (B–G) is 100 bp ladder (Gibco).

Figure 2. RT–PCR analysis during mouse eye development. Lanes 1–6 are cDNA samples from E9.5, E10.5, E11.5, E12.5, E13.5 and E14.5, respectively. Lane 7 is adult cDNA and lane 8 is a negative control. Lanes 9–20 are cDNA samples from P1–P12, respectively. Lane 21 is adult cDNA and lane 22 is a negative control. The genes tested by RT–PCR were (A) Six3, (B) Crx, (C) arrestin, (D) Irbp, (E) Pdeb, (F) rhodopsin, (G) Rs1, (H) Pdea, (I) Pdeg, (J) green opsin, (K) blue opsin and (L) Nrl. The DNA marker used was ϕX174 (Gibco).

Figure 2. RT–PCR analysis during mouse eye development. Lanes 1–6 are cDNA samples from E9.5, E10.5, E11.5, E12.5, E13.5 and E14.5, respectively. Lane 7 is adult cDNA and lane 8 is a negative control. Lanes 9–20 are cDNA samples from P1–P12, respectively. Lane 21 is adult cDNA and lane 22 is a negative control. The genes tested by RT–PCR were (A) Six3, (B) Crx, (C) arrestin, (D) Irbp, (E) Pdeb, (F) rhodopsin, (G) Rs1, (H) Pdea, (I) Pdeg, (J) green opsin, (K) blue opsin and (L) Nrl. The DNA marker used was ϕX174 (Gibco).

Figure 3.In situ hybridization analysis of CRX expression in human fetal and adult retina. Left panels are haematoxylin and eosin (H&E) stained sections adjacent to corresponding in situ in right panels. (A) In human fetal retina at 13 weeks p.c., staining is observed in newly born photoreceptors (NBPR), lying just vitreal to the RPE. NBL, neuroblastic layer; PGCL, presumptive ganglion cell layer. (B) In human fetal retina at 15 weeks p.c., expression is seen throughout the developing ONL, and is absent from the developing INL. (C) In adult human retina, CRX is localized to both ONL and INL. Scale bars, 100 µm.

Figure 3.In situ hybridization analysis of CRX expression in human fetal and adult retina. Left panels are haematoxylin and eosin (H&E) stained sections adjacent to corresponding in situ in right panels. (A) In human fetal retina at 13 weeks p.c., staining is observed in newly born photoreceptors (NBPR), lying just vitreal to the RPE. NBL, neuroblastic layer; PGCL, presumptive ganglion cell layer. (B) In human fetal retina at 15 weeks p.c., expression is seen throughout the developing ONL, and is absent from the developing INL. (C) In adult human retina, CRX is localized to both ONL and INL. Scale bars, 100 µm.

Figure 4. Immunocytochemical localization of CRX in fetal human retina. (A) Western blot of N-terminal peptide antibody showing a single band at 37 kDa. (B) Left panel, H&E staining; right panel, antibody localization at 15 weeks p.c. Green fluorescence present in newly formed photoreceptors. (C) No antibody detected at 13 weeks p.c. Scale bars, 100 µm.

Figure 4. Immunocytochemical localization of CRX in fetal human retina. (A) Western blot of N-terminal peptide antibody showing a single band at 37 kDa. (B) Left panel, H&E staining; right panel, antibody localization at 15 weeks p.c. Green fluorescence present in newly formed photoreceptors. (C) No antibody detected at 13 weeks p.c. Scale bars, 100 µm.

Figure 5. RT–PCR analysis from wild-type and rd/rd,cl mouse retinal cDNA. Lane 1, rd/rd,cl cDNA; lane 3, wild-type cDNA; lanes 2 and 4 are negative controls. The genes amplified were (A) CRX, (B) rhodopsin, green opsin and blue opsin, respectively. The DNA marker used was ϕX174 (Gibco).

Figure 5. RT–PCR analysis from wild-type and rd/rd,cl mouse retinal cDNA. Lane 1, rd/rd,cl cDNA; lane 3, wild-type cDNA; lanes 2 and 4 are negative controls. The genes amplified were (A) CRX, (B) rhodopsin, green opsin and blue opsin, respectively. The DNA marker used was ϕX174 (Gibco).

Figure 6.In situ hybridization and immunocytochemical (ICC) analyses in wild-type and rd/rd,cl retinal sections. (A) Crx mRNA expression in wild-type retina present in ONL and more weakly in INL. Left panel, H&E; right panel, in situ. (B) Crx mRNA expression in rd/rd,cl retina present in INL. Left panel, H&E; right panel, in situ. (C) CRX antibody localization in wild-type retina. Left panel, phase contrast; right panel, ICC. (D) CRX antibody localization in rd/rd,cl retina in the INL. Left panel, phase contrast; right panel, ICC. (A and B), 10× magnification; (C and D), 20× magnification.

Figure 6.In situ hybridization and immunocytochemical (ICC) analyses in wild-type and rd/rd,cl retinal sections. (A) Crx mRNA expression in wild-type retina present in ONL and more weakly in INL. Left panel, H&E; right panel, in situ. (B) Crx mRNA expression in rd/rd,cl retina present in INL. Left panel, H&E; right panel, in situ. (C) CRX antibody localization in wild-type retina. Left panel, phase contrast; right panel, ICC. (D) CRX antibody localization in rd/rd,cl retina in the INL. Left panel, phase contrast; right panel, ICC. (A and B), 10× magnification; (C and D), 20× magnification.

References

1 Zipursky, S.L. and Rubin, G.M. (
1994
) Determination of neuronal cell fate: lessons from the R7 neuron of Drosophila.
Annu. Rev. Neurosci.
 ,
17
,
373
–397.
2 Cepko, C., Austin, C., Yang, X., Alexiades, M. and Ezzeddine, D. (
1996
) Cell fate determination in the vertebrate retina.
Proc. Natl Acad. Sci. USA
 ,
93
,
589
–595.
3 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.
4 Freund, C.L., Gregory-Evans, C.Y., Furukawa, T., Papaioannou, M., Looser, J., Ploder, L., Bellingham, J., Ng, D., Herbrick, J.-A.S., 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.
5 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.
6 Finkelstein, R. and Boncinelli, E. (
1994
) From fly head to mammalian forebrain: the story of otd and OTX.
Trends Genet.
 ,
10
,
310
–315.
7 Swain, P.K., Chen, S., Wang, Q.-L., Affatigato, L.M., Copats, 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 cone-rod dystrophy photoreceptor degeneration.
Neuron
 ,
19
,
1329
–1336.
8 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.
9 Sankila, E.M., Joensuu, T.H., Hamalainen, R.H., Raitanen, N., Valle, O., Ignatius, J. and Cormand, B. (
2000
) A CRX mutation in a Finnish family with dominant cone-rod retinal dystrophy.
Hum. Mutat.
 ,
16
,
94
.
10 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.
11 Swaroop, A., Wang, Q.-L., Wu, W., Cook, J., Coats, C., Xu, S., Zack, D.J. and Sieving, P.A. (
1999
) Leber congenital amaurosis caused by 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.
12 Silva, E., Yang, J.-M., Li, Y., Dharmaraj, S., Sundin, O.H. and Maumenee, I.H. (
2000
) A CRX null mutation is associated with both Leber congenital amaurosis and a normal ocular phenotype.
Invest. Ophthalmol. Vis. Sci.
 ,
41
,
2076
–2079.
13 Evans, K., Duvall-Young, J., Fitzke, F.W., Arden, G.B., Bhattacharya, S.S. and Bird, A.C. (
1995
) Chromosome 19q cone-rod retinal dystrophy: ocular phenotype.
Arch. Ophthalmol.
 ,
113
,
195
–201.
14 Schroeder, R., Mets, M.B. and Maumenee, I.H. (
1987
) Leber’s congenital amaurosis.
Arch. Ophthalmol.
 ,
105
,
356
–359.
15 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.
16 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.
17 Kumar, R. and Zack, D.J. (
1995
) Regulation of visual pigment gene expression. In Wiggs, J. (ed.), Molecular Genetics of Ocular Disease. Wiley-Liss Inc., New York, NY, pp.
139
–160.
18 Granadino, B., Gallardo, M.E., Lopez-Rios, J., Sanz, R., Ramos, C., Ayuso, C., Bovolenta, P. and Rodriguez de Cordoba, S. (
1999
) Genomic cloning, structure, expression pattern and chromosomal location of the human SIX3 gene.
Genomics
 ,
55
,
100
–105.
19 Oliver, G., Mailhos, A., Wehr, R., Copeland, N.G., Jenkins, N.A. and Gruss, P. (
1995
) Six3, a murine homologue of the sine oculis gene, demarcates the most anterior border of the developing neural plate and is expressed during eye development.
Development
 ,
121
,
4045
–4055.
20 Cepko, C.L. (
1996
) The patterning and onset of opsin expression in vertebrate retinae.
Curr. Opin. Neurobiol.
 ,
6
,
542
–546.
21 Xiao, M. and Hendrickson, A. (
2000
) Spatial and temporal expression of short, long/medium, or both opsins in human foetal cones.
J. Comp. Neurol.
 ,
425
,
545
–559.
22 He, L., Campbell, M.L., Srivastava, D., Blocker, Y.S., Harris, J.R., Swaroop, A. and Fox, D.A. (
1998
) Spatial and temporal expression of AP-1 responsive rod photoreceptor genes and bZIP transcription factors during development of the rat retina.
Mol. Vis.
 ,
4
,
32
.
23 Freedman, M.S., Lucas, R.J., Soni, B., von Schantz, M., David-Gray, Z. and Foster, R. (
1999
) Regulation of mammalian circadian behaviour by non-rod, non-cone ocular photoreceptors.
Science
 ,
284
,
502
–504.
24 Remington, L.A. and McGill, E.C. (
1997
) Ocular embryology. In McGill, E.C. (ed.), Clinical Anatomy of the Visual System. Butterworth-Heinemann, Oxford, UK, pp.
103
–121.
25 Jakobiec, F.A. and Ozanics, V. (
1982
) Prenatal development of the eye and its adenexa. In Jakobiec, F.A. and Ozanics, V. (eds), Ocular Anatomy, Embryology and Teratology. Harper and Row, London, UK, pp.
11
–96.
26 Rohrer, B., Korenbrot, J.I., LaVail, M.M., Reichardt, L.F. and Xu, B (
1999
) Role of neurotrophin receptor TrkB in the maturation of rod photoreceptors and establishment of synaptic transmission to the inner retina.
J. Neurosci.
 ,
19
,
8919
–8930.
27 Di polo, A., Lerner, L.E. and Farber, D.B. (
1996
) Transcriptional activation of the human rod cGMP-phosphodiesterase β-subunit gene is mediated by an upstream AP-1 element.
Nucleic Acids Res.
 ,
25
,
3863
–3867.
28 Mears, A.J., Kondo, M., Takada, Y., Swain, P.K., Nelson, B., Sieving, P.A. and Swaroop, A. (
2001
) NRL is a major determinant of rod photoreceptor cell fate: study of a knockout mouse.
Invest. Ophthalmol. Vis. Sci.
 ,
42
,
S527
.
29 Fong, S.L. and Fong, W.B. (
1999
) Elements regulating the transcription of human interstitial retinoid binding protein (IRBP) gene in cultured retinoblastoma cells.
Curr. Eye Res.
 ,
18
,
283
–291.
30 Lipshitz, H.D. and Smibert, C.A. (
2000
) Mechanisms of RNA localisation and translational regulation.
Curr. Opin. Genet. Dev.
 ,
10
,
476
–488.
31 Kim-Ha, J., Kerr, K. and Macdonald, P.M. (
1995
) Translational regulation of oskar mRNA by Bruno, an ovarian RNA-binding protein, is essential.
Cell
 ,
81
,
403
–412.
32 Smibert, C.A., Wilson, J.E., Kerr, K. and Macdonald, P.M. (
1996
) Smaug protein represses translation of unlocalised nanos mRNA in the Drosophila embryo.
Genes Dev.
 ,
10
,
2610
–2620.
33 Wilson, J.E., Connell, J.E. and Macdonald, P.M. (
1996
) aubergine enhances oskar translation in the Drosophila ovary.
Development
 ,
122
,
1631
–1639.
34 Crucs, C., Chatterjee, S. and Gavis, E.R. (
2000
) Overlapping but distinct RNA elements control repression and activation of nanos translation.
Mol. Cell
 ,
5
,
457
–467.
35 Chen, S., McMahan, B. and Xu, S. (
2000
) Localisation of the CRX protein in the mature and developing mouse retina.
Invest. Ophthalmol. Vis. Sci.
 ,
41
,
S392
.
36 Liu, Y., Shen, Y.-C., Rest, J.S., Redmond, P.A. and Zack, D.J. (
2001
) Isolation and characterization of a zebrafish homologue of the cone rod homeobox gene.
Invest. Ophthalmol. Vis. Sci.
 ,
42
,
481
–487.
37 Young, R.W. (
1985
) Cell death during differentiation of the retina in the mouse.
J. Comp. Neurol.
 ,
229
,
362
–373.
38 Fougerousse, F., Bullen, P., Herasse, M., Lindsay, S., Richard, I., Wilson, D., Suel, L., Durand, M., Robson, S., Abitbol, M. et al. (
2000
) Human-mouse differences in the embryonic expressions patterns of developmental control genes and disease genes.
Hum. Mol. Genet.
 ,
9
,
165
–173.
39 Wynshaw-Boris, A. (
1996
) Model mice and human disease.
Nat. Genet.
 ,
13
,
259
–260.
40 Tzekov, R.T., Liu, Y., Sohocki, M.M., Zack, D.J., Daiger, S.P., Heckenlively, J.R. and Birch, D.G. (
2001
) Autosomal dominant retinal degeneration and bone loss in patients with a 12-bp deletion in the CRX gene.
Invest. Ophthalmol. Vis. Sci.
 ,
42
,
1319
–1327.