Leber congenital amaurosis (LCA) is the most severe inherited retinal dystrophy resulting in markedly impaired vision or blindness at birth. LCA is characterized by an extinguished electroretinogram in infancy, which is thought to be indicative of an early and severe impairment of both the rod and cone photoreceptors in the human retina. Recently, the aryl hydrocarbon receptor-interacting protein-like 1 (AIPL1) gene was identified as the fourth causative gene of LCA. AIPL1 encodes a 384 amino acid protein of unknown function. We have generated a polyclonal antibody against a peptide from a unique region within the primate AIPL1 protein, which detects a protein of ∼43 kDa in human retinal extracts. A screen of human tissues and immortalized cell lines with this antibody reveals AIPL1 to be specific to human retina and cell lines of retinal origin (Y79 retinoblastoma cells). Within the retina, AIPL1 was detected only in the rod photoreceptor cells of the peripheral and central human retina. The AIPL1 staining pattern extended within the rod photoreceptor cells from the inner segments, through the rod nuclei to the rod photoreceptor synaptic spherules in the outer plexiform layer. AIPL1 was not detected in the cone photoreceptors of peripheral or central human retina. This study is the first to suggest that AIPL1 performs a function essential to the maintenance of rod photoreceptor function.
Received December 12, 2001; Revised and Accepted January 31, 2002.
Leber congenital amaurosis (LCA), first described by Theodore Leber in 1869 (1), is the most severe form of early-onset inherited retinal dystrophy responsible for congenital blindness or severely impaired vision. Clinically, patients affected by LCA present in infancy with pendular nystagmus, unusual roving eye movements and absent ocular pursuit upon ophthalmic examination. LCA patients may habitually rub their eyes with the fist or fingers (the oculodigital reflex) (2) and have a higher incidence of keratoconus than in the normal population. In early infancy, the optic discs and fundus are normal in appearance showing no abnormality upon fundoscopy. However, progressive abnormalities in the fundus appear with time, including attenuation of the retinal vasculature, optic nerve pallor, bone corpuscular pigmentation, atrophy of the retinal pigment epithelium (RPE) and occasionally irregular yellow pigmentation within the peripheral and mid-peripheral retina. In the absence of fundus abnormalities in early infancy, LCA is differentiated from other early-onset retinal dystrophies by electroretinographic (ERG) testing. The ERG recordings in LCA patients are markedly attenuated or absent. Additional systemic disorders, of which psychomotor retardation is the most frequent, have been reported in association with LCA (3–5).
LCA is a genetically heterogeneous disorder with an autosomal recessive mode of inheritance (6). To date, six causative genes have been identified in LCA, the first of which maps to chromosome 17p13.1 in humans (7,8) and translates to a photoreceptor-specific guanylate cyclase (retGC-1 or GUCY2D) (9). RetGC-1 is an essential component of the phototransduction cascade, and mutations in retGC-1 impair the recovery of the dark state after photo-excitation of the photoreceptor cells. The situation is equivalent to sustained photo-excitation of the photoreceptors and patients with LCA caused by mutations in retGC-1 present with severe hyperopia and photophobia (10). The second locus for LCA was assigned to chromosome 1p31 (11) and the affected gene encodes a microsomal membrane protein (RPE65) of the RPE (12). RPE65 is a crucial component of the visual cycle in that it facilitates the isomerization of all-trans-retinylester to 11-cis-retinol and hence the regeneration of the universal chromatophore 11-cis-retinal (13,14). Mutations in the RPE65 gene decrease the production of functional visual pigments leading to a situation equivalent to the sustained absence of photo-excitation of photoreceptor cells. Patients with LCA caused by mutations in RPE65 present with moderate or no hyperopia and sometimes low myopia upon ophthalmic examination (10). The third causative gene of LCA is a cone–rod homeobox (CRX) gene, which maps to chromosome 19q13.3 in humans (15,16). The CRX protein is a photoreceptor-specific homeodomain transcription factor that plays an essential role in the differentiation, development and maintenance of photoreceptor cells through the transactivation of several photoreceptor-specific gene promoters, including rhodopsin (16–18). Recently, two other LCA causative genes have been identified, CRB1 (19,20) and RPGRIP1 (21). CRB1 is a retina and CNS expressed protein that is homologous to the Drosophila crumbs protein that is important in cell polarity and cell–cell contact, although the role of CRB1 in the retina remains to be identified. RPGRIP1 was identified as an interacting partner of the X-linked retinitis pigmentosa protein RPGR and is thought to be localized to the photoreceptor connecting cilium where it may play a role in transport between the inner and outer segments. The diverse nature of these LCA causative genes illustrates the complexity of retinal cell biology and highlights that many, as yet unstudied, pathways are essential for the normal function of the retina.
The aryl hydrocarbon receptor interacting protein-like 1 (AIPL1) gene was the fourth gene found to cause LCA. AIPL1 maps to within 2.5 Mb distal to the retGC-1 locus on chromosome 17p13.1 in humans (22). The AIPL1 gene encompasses six exons encoding a protein of 384 amino acids in length. The protein sequence includes three consecutive 34 amino acid tetratricopeptide repeat (TPR) motifs, which are thought to mediate specific protein interactions. TPR motifs are ubiquitously conserved in structurally unrelated proteins that participate in diverse biological functions, including the co-ordination of multiprotein complex assembly and protein translocation (23). A 56 amino acid polyproline-rich sequence of high flexibility encompassing multiple O-glycosylation sites and putative phosphorylation sites is present at the C-terminus of the AIPL1 protein in humans. The function of AIPL1 in normal vision is unknown. However, AIPL1 is similar (49% identity) to the aryl hydrocarbon receptor-interacting protein (AIP) (24), also known as the aryl hydrocarbon receptor-activated protein (ARA9) (25) or the X-associated protein (XAP2) (26). The aryl hydrocarbon receptor is a cytosolic ligand-activated transcription factor that mediates adaptive and toxic responses to environmental pollutants such as dioxin by increasing the transcription of xenobiotic metabolizing enzymes. AIP facilitates the transactivation activity of the cognate transcription factor by regulating its nuclear translocation (24–31 and reviewed in 32). The inclusion of the TPR motifs in the AIPL1 protein and the similarity of AIPL1 to AIP collectively suggest that AIPL1 may be involved in retinal protein folding or cellular translocation.
Northern blot hybridization and in situ hybridization demonstrated the expression of AIPL1 in the pineal gland and retina (22). This study is the first description of the localization of AIPL1 protein in the adult human retina and photoreceptor cells. Our results are surprising in that AIPL1 is produced exclusively in the rod photoreceptor system of the adult human retina and is absent from cones, and suggest that AIPL1 performs an essential function in the maintenance of rod photoreceptors.
Characterization of AIPL1 expression
A rabbit polyclonal antiserum was raised against a keyhole limpet haemocyanin (KLH)-conjugated peptide AEPATEPPPSPGHSLQH comprising amino acid residues 368–384 of the predicted human AIPL1 protein sequence. The polyclonal antiserum, Ab-hAIPL1, was characterized against human tissue homogenates and AIPL1 recombinant proteins. A single protein of ∼43 kDa, corresponding to the predicted molecular weight of AIPL1 (43.865 kDa), was detected by the Ab-hAIPL1 antiserum in human retinal tissue extracts under reducing conditions (Fig. 1A). The 43 kDa protein in human retinal tissue was not detected by Ab-hAIPL1 in the presence of competing peptide (1 µg/ml) (Fig. 1A). The pre-immune serum showed no reactivity with the 43 kDa protein or any other protein species (Fig. 1A), indicating that the reactivity against the 43 kDa retinal protein was specific. The Ab-hAIPL1 antiserum was also able to specifically react against heterologously produced AIPL1 recombinant proteins (Fig. 1B). Single bands of ∼71 and 50 kDa were detected in immunoblots of purified glutathione S-transferase (GST)-tagged AIPL1 (GST-AIPL1) and polyhistidine-tagged AIPL1 (His-AIPL1), respectively, corresponding to the predicted molecular weight of each of these proteins. Ab-hAIPL1 showed no reactivity against any proteins of bacterial origin or GST on its own. Hence, the Ab-hAIPL1 antiserum was able to react with the AIPL1 epitope in both recombinant AIPL1 and AIPL1 in retinal extracts. The Ab-hAIPL1 antiserum was characterized against a panel of eight other human tissues including neocortex, cerebellum, spinal cord, kidney, lung, liver, testes and heart (Fig. 1C). The 43 kDa protein was only detected in human retina. This confirmed previous analyses using multi-tissue northern blot hybridization, which demonstrated mRNA expression in adult human retina (22). Expression of AIPL1 in the Y79 retinoblastoma cell line was investigated by immunoblotting with Ab-hAIPL1, as retinoblastoma cells share many photoreceptor characteristics (33) (Fig. 1D). Ab-hAIPL1 detected a specific band of ∼43 kDa in whole cell preparations of Y79 cells, demonstrating that Y79 retinoblastoma cells also express AIPL1. Ab-hAIPL1 failed to detect a specific product in five other cell lines of non-retinal origin, including two cell lines of human corneal origin (HCEF and EK1B) (a gift of Mr N.D.Ebenezer).
Immunocytochemistry of AIPL1 in Y79 cells
Undifferentiated Y79 retinoblastoma cells grown in suspension were attached to glass slides by cytospin centrifugation, and the subcellular distribution of AIPL1 was determined by fluorescence immunocytochemistry using Ab-hAIPL1 (Fig. 2). AIPL1 was detected both in the nucleus (n) and in the cytoplasm (c) of Y79 cells, although immunostaining with Ab-hAIPL1 was more intense in the cytoplasm and in some cells was perinuclear. The staining pattern throughout the Y79 cells was granular in appearance. No immunostaining was detected either with the pre-immune serum or in the presence of competing peptide (data not shown).
Immunohistochemistry of AIPL1 in human retina
In 10% neutral-buffered formalin-fixed, paraffin-embedded sections of adult human eye, specific immunolabelling with Ab-hAIPL1 was detected only in the retina, and was excluded from all other parts of the eye including the lens, cornea, sclera, choroid and optic nerve (data not shown). Within the retina, immunolabelling with Ab-hAIPL1 was detected in the photoreceptor inner segments, the outer nuclear layer (ONL) and the outer plexiform layer (OPL) extending to the rod spherules and cone pedicles in the OPL (Fig. 3). Immunolabelling with Ab-hAIPL1 was absent from the RPE, the inner nuclear layer (INL) and the inner plexiform layer (IPL) (Fig. 3). The connecting cilia of the photoreceptor inner segments appeared to be more intensely labelled with the Ab‐hAIPL1 antiserum. Towards the outer limit of the ONL, a clear absence of Ab-hAIPL1 immunolabelling was noticeable in the region of the cone photoreceptor cell bodies. Rod photoreceptor cell bodies of the ONL were intensely labelled by Ab-hAIPL1, and the distribution of Ab-hAIPL1 immunolabelling around the rod photoreceptor nuclei appeared to be asymmetric with more intense labelling apically. The immunolabelling of the OPL with the Ab-hAIPL1 antiserum was punctate in appearance, with intense labelling visible in the region of the photoreceptor spherules and pedicles. The absence of Ab-hAIPL1 immunostain in the cone photoreceptor cell bodies was confirmed at high titres (1:50) of Ab-hAIPL1 antiserum and long chromophore development times (data not shown). Furthermore, immunolabelling of the central retina (macula) with Ab-hAIPL1 demonstrated a similar distribution of AIPL1 and confirmed the apparent exclusion of AIPL1 labelling from the cone photoreceptor cell bodies (data not shown). No reaction product was observed when the primary antibody was omitted or when immunolabelling was performed in the presence of the peptide against which Ab-hAIPL1 was raised (data not shown).
Immunofluorescence confocal microscopy of AIPL1
Double immunofluorescent labelling was performed with the Ab-hAIPL1 antiserum and a panel of retinal markers in paraformaldehyde-fixed agarose-embedded human retina. In the peripheral human retina, AIPL1 (Fig. 4, AIPL1) was detected in the rod photoreceptor inner segments, with intense labelling in the region of the photoreceptor connecting cilia, but was absent from the rod photoreceptor outer segments. The rod photoreceptor cell bodies in the ONL were intensely labelled with Ab-hAIPL1. Though Ab-hAIPL1 labelling was not absent from the nuclei of the rod photoreceptors, labelling was distributed in the rod photoreceptor cell bodies in an asymmetric, perinuclear fashion. Immunolabelling with the Ab-hAIPL1 antiserum was absent from the cone photoreceptor cell bodies. Punctate labelling of the OPL could be detected with Ab-hAIPL1, which terminated with intense labelling of the rod photoreceptor synaptic spherules. The distribution of AIPL1 in the peripheral adult human retina by confocal microscopy thus confirmed that previously shown by immunohistochemistry (Fig. 3). Immunofluorescent labelling with the monoclonal antibody 1D4 detected rhodopsin in the rod photoreceptor outer segments (Fig. 4, 1D4). The signals for Ab-hAIPL1 and 1D4 did not co-localize (Fig. 4, MERGE), demonstrating the exclusion of AIPL1 from the rod photoreceptor outer segments. An enlargement of the merged image (Fig. 4, ZOOM) emphasizes the intense Ab-hAIPL1 immunosignal detected in the connecting cilia of the rod photoreceptor cells as well as the absence of Ab-hAIPL1 signal from the rod photoreceptor outer segments.
The lectin peanut agglutinin (PNA) was used as a marker for the extracellular matrix (ECM) sheaths of the cone photoreceptor outer segments and cone photoreceptor inner segments (Fig. 5A, PNA). PNA labelling did not co-localize with the Ab‐hAIPL1 immunosignal (Fig. 5A, MERGE), demonstrating the absence of AIPL1 from the cone inner and outer segments. The absence of Ab-hAIPL1 immunosignal from the cone photoreceptor outer and inner segments is highlighted in an enlargement (Fig. 5A, ZOOM) of the merged image.
The monoclonal antibody 7G6 was used as a marker for cone photoreceptors extending from the cone inner segments to the cone synaptic pedicles in the OPL (Fig. 5B, 7G6). 7G6 labelling did not co-localize with the Ab-hAIPL1 immunosignal (Fig. 5B, MERGE), demonstrating the absence of AIPL1 from the cone photoreceptor cells, including the cell bodies, outer and inner segments and the cone synaptic pedicles in the OPL. In an enlargement of the merged image (Fig. 5B, ZOOM), the absence of Ab-hAIPL1 immunosignal from the cone photoreceptor inner segments, cone photoreceptor cell bodies and cone photoreceptor spherules is clearer. Collectively, these results demonstrate the exclusion of AIPL1 from the cone photoreceptor cells in the peripheral retina, and confine the AIPL1 distribution to the rod photoreceptors extending from the connecting cilia of the rod photoreceptor inner segments to the synaptic terminals of the rod photoreceptor spherules in the OPL. In each experiment, the Ab-hAIPL1 immunosignal was absent from cone photoreceptor cells, even at high titres of the Ab‐hAIPL1 antiserum, suggesting the absence of low levels of AIPL1 expression in cone photoreceptors. Furthermore, no AIPL1 reaction product was observed when the Ab-hAIPL1 antibody was omitted or when immunolabelling was performed in the presence of competing peptide (data not shown). No cross-reaction was detected between the primary and secondary antibodies of each double labelling experiment and no immunosignals were detected in the presence of either primary or secondary antibodies on their own (data not shown).
Double immunofluorescent labelling was performed with the Ab-hAIPL1 antiserum and the 7G6 monoclonal antibody in the macula and fovea of the human retina. Within the rim of the foveal pit (Fig. 6), the Ab-hAIPL1 immunosignal was limited to the small population of rod photoreceptor cells in this region (white arrowheads) and did not co-localize with the 7G6 immunosignal observed for the population of densely packed cone photoreceptor cells. Within the foveal pit, which contains exclusively cone photoreceptor cells, the Ab-hAIPL1 immunosignal was absent (data not shown). These results demonstrate the exclusion of AIPL1 from the cone photoreceptor cells in the central human retina.
Within the adult human retina, our studies show that AIPL1 protein is produced exclusively in the rod photoreceptors and is not present in the cone photoreceptors. This result is intriguing in light of the fact that LCA is thought to be the consequence of either the impaired function, development or extremely early degeneration of both the rod and cone photoreceptor systems based on the absence of ERG recordings in LCA patients. The sustained expression of AIPL1 in adult photoreceptors suggests that AIPL1 performs an essential function in the maintenance of the rod photoreceptors in the adult human retina. It is possible that AIPL1 is expressed in both rod and cone photoreceptor precursors and other cell types during eye development. The early degeneration or dysfunction of both rod and cone photoreceptor progenitor cells in the absence of functional AIPL1 may explain why the cone cells fail to function in AIPL1-LCA. Alternatively, AIPL1 may be expressed exclusively in developing rod photoreceptor cells as well as in the adult retina. In this scenario, an extremely early degeneration of rod photoreceptor precursors due to the absence of functional AIPL1 could lead to the dysfunction or degeneration of cone photoreceptors. The importance of rod derived cone survival factors in retinal degeneration is well established (34), although the effects of such early rod cell loss upon cone cell survival are unknown. Another possibility is that the loss of AIPL1 function in rod precursors may result in an inhibition of the final maturation of cone precursors due to the absence of an inductive or permissive signal (34). The determination of the spatial and temporal expression of AIPL1 in the developing human retina, therefore, will be fundamental in clarifying the pathogenesis of LCA caused by mutations in AIPL1.
The proteins encoded by the different LCA causative genes each participate in distinct and essential physiopathological pathways in the retina, including phototransduction (retGC-1), retinoid metabolism (RPE65), photoreceptor development (CRX), cell polarity (CRB1) and intracellular transport (RPGRIP1). Attempts have been made to correlate the genetic heterogeneity of LCA with clinically heterogeneous features (10). For example, mutations in the retGC-1 gene are responsible for LCA characterized by congenital severe cone–rod dystrophy, whilst mutations in the RPE65 gene cause LCA characterized by congenital severe but progressive rod–cone dystrophy and CRX mutations are associated with LCA characterized by cone–rod dystrophy (10). The exclusive localization of AIPL1 to rod photoreceptors suggests that mutations in AIPL1 should result in LCA characterized by congenital severe rod–cone dystrophy. In addition to the heterogeneity in LCA, the LCA causative genes also display a degree of allelism with some mutations resulting in clinical physiopathological diagnoses distinct from LCA. Heterozygous mutations in exon 13 of the retGC-1 gene result in progressive autosomal dominant cone–rod dystrophy (CORD6) (35–37). Similarly, a four amino acid deletion mutation (P351Δ12) in the C-terminal polyproline-rich region of AIPL1 has been identified in two probands given the clinical diagnosis of autosomal dominant cone–rod dystrophy and juvenile RP, respectively (38), though all other AIPL1 mutations are associated with LCA. This diagnosis is surprising in the light of our findings, as we would predict that dominant mutations in AIPL1 should result in rod–cone dystrophy. Fundoscopic and electrodiagnostic analyses of dominant AIPL1 cases and heterozygous ‘carriers’ of AIPL1 mutations combined with cellular and molecular studies of AIPL1 function may clarify the role of AIPL1 in rod–cone and cone–rod dystrophies.
The epitope for our antiserum Ab-hAIPL1 resides in the 56 amino acid C-terminal polyproline-rich region of the predicted human AIPL1 sequence. A comparative analysis of the AIPL1 gene in different species demonstrated that the polyproline-rich region is conserved only in primates (human, chimpanzee, baboon, rhesus monkey and squirrel monkey) and is absent in the mouse, rat and cow (39). Ab-hAIPL1, directed against this C-terminal extension, does not cross-react with AIPL1 in these species (data not shown) confirming the apparent primate specificity of this region. The results of our study are based on the expression of this C-terminal domain and we cannot exclude the potential expression of an AIPL1 variant lacking the polyproline-rich region, and thus our antibody epitope, in cone photoreceptors. However, the polyproline-rich region is encoded by exon 6 of the AIPL1 gene, which also encodes the third TPR motif (22), and it is thus unlikely that alternative splicing produces an AIPL1 isoform lacking the polyproline-rich region. Furthermore, it is evident that the polyproline-rich region performs an important if not essential function in normal primate vision, arguing against the existence of an AIPL1 variant lacking the polyproline-rich region. Three mutations (R302L, A336Δ2 and P376S) have been identified in the AIPL1 polyproline-rich region, which cause the severe phenotype of LCA (22,38). In addition, an LCA-causing homozygous nonsense mutation at codon 278 of AIPL1 (W278X), if expressed, truncates the C-terminal 107 amino acids of AIPL1, including the entire polyproline-rich region (22,38) and once again emphasizing the importance of this region in normal primate vision. The observed rod cell exclusivity of AIPL1 expression in our study could be a consequence of cone-specific epitope masking. We feel that this is highly unlikely, as we have observed the same pattern of expression under a wide range of tissue fixations and antigen retrieval procedures. Such a cell-specific epitope masking would, however, still indicate fundamental differences in AIPL1 function between rods and cones.
There is a higher propensity for LCA patients to develop keratoconus, a chronic non-inflammatory corneal thinning disorder. There is evidence for a genetic role in the etiology of keratoconus (40,41) and it has been demonstrated that AIPL1 is a locus for the combined phenotype of LCA and keratoconus, which cosegregate in an autosomal recessive manner (42,43). However, in a study of 12 LCA patients from four consanguineous pedigrees, there was a striking phenotypic variability with respect to keratoconus even though the W278X mutation in AIPL1 was the underlying cause of LCA in each affected subject (43). Three of the affected subjects had definite keratoconus, and two were suspected of developing keratoconus based on mild cone formation in the cornea of at least one eye (43). Our results demonstrate that AIPL1 protein is expressed only in immortalized cells of retinal origin and in the retina of the adult human eye. AIPL1 is not expressed either in immortalized cells of corneal origin or in the cornea of the adult human eye. Therefore, there is no simple relationship between AIPL1 expression and the development of keratoconus. Mutations in RetGC1 have also been associated with LCA and keratoconus with no obvious mechanistic link between the photoreceptor protein and corneal disease (44). It is possible that the combined phenotype of LCA and anterior keratoconus may arise due to unknown developmental effects of AIPL1 functional ablation or combined genetic and environmental factors such as repeated eye rubbing (45).
The role of AIPL1 within the rod photoreceptors of the retina has yet to be established. However, it is interesting that within the highly polarized photoreceptor cell, AIPL1 is excluded from the rod photoreceptor outer segments and otherwise extends throughout the cell from the inner segments to the rod spherules. The similarity of AIPL1 to AIP as well as the inclusion of the TPR motifs in the AIPL1 sequence suggests that AIPL1 may participate in protein complex formation and/or translocation within the rod photoreceptor cells of the retina. Furthermore, proline-rich regions similar to those found in AIPL1 mediate protein interactions in which the participants are involved in cellular processes that require the rapid recruitment or interchange of proteins, including cytoskeletal rearrangements, signal transduction and transcriptional initiation (46). Hence, AIPL1 may well function in protein complex formation and translocation in the rod photoreceptors of the human retina, and it is important to identify the partner proteins that interact with AIPL1.
In conclusion, this study is the first to demonstrate that AIPL1, a product of a gene that causes Leber congenital amaurosis, is expressed exclusively in the rod photoreceptors of the adult human retina. This result will impact on future studies of AIPL1, including the development of appropriate animal models and potential therapies for LCA caused by mutations in AIPL1. Further evaluation of the expression of AIPL1 in the developing human retina and the function of AIPL1 within the rod photoreceptors will provide insight into the biology of normal vision and the pathology of LCA.
MATERIALS AND METHODS
Antibody, SDS–PAGE and immunoblotting
Rabbit polyclonal antiserum, Ab-hAIPL1, was raised against peptide AEPATEPPPSPGHSLQH (amino acid residues 368–384) conjugated to KLH (Genosys Biotechnologies, Cambridge, UK). Total protein extracts were prepared by homogenization of tissues or cells in sample buffer (0.0625 M Tris–HCl, 2.5% SDS, 5% 2-mercaptoethanol, 10% glycerol, 0.25% bromophenol blue). Total protein concentrations in tissue and cell extracts were assayed using the Bradford microassay (Bio-Rad Laboratories, Hercules, CA) to ensure even loading on gels. Protein samples were resolved by SDS–PAGE on 12% gels and electroblotted onto ECL nitrocellulose (Amersham, Little Chalfont, UK). The nitrocellulose membrane was incubated overnight in blocking solution [5% Marvel non-fat milk powder, 1× phosphate-buffered saline (PBS), 0.1% Tween-20], following which blots were hybridized with rabbit pre-immune serum or Ab-hAIPL1 (1:2000) for 1 h at room temperature. In competition assays, peptide (1 µg/ml) against which the antiserum was generated was incubated with the Ab-hAIPL1 antiserum (1:2000) for 30 min on ice prior to incubation with the nitrocellulose membrane for 1 h at room temperature. Subsequently, the membrane was incubated with mouse anti-rabbit secondary antibody (1:20000) conjugated to horseradish peroxidase (Pierce, Rockford, IL) for 1 h at room temperature. Bands were visualized using enhanced chemiluminescence (Amersham).
Retinoblastoma cell culture and immunocytochemistry
Y79 human retinoblastoma cells (47) (a gift of Dr D.Trump, University of Cambridge, UK) were maintained in suspension culture in RPMI 1640 medium (Gibco BRL, Paisley, UK) supplemented with 15% fetal bovine serum, 2 mM l-glutamine and 50 µg/ml gentamycin (Sigma, Poole, UK) with medium changes every 2–3 days. For western analysis of endogenous AIPL1, Y79 cells were washed three times with ice cold PBS before whole cell homogenization in sample buffer. For immunocytochemistry, Y79 cells were washed three times in ice cold PBS and fixed in 3.7% formaldehyde (TAAB Laboratories, Reading, UK) in PBS for 15 min at room temperature. The Y79 retinoblastoma cells were fixed to glass slides by cytospin centrifugation, permeabilized in 0.1% Triton X-100 (Sigma) in PBS for 5 min at room temperature and blocked with 3% bovine serum albumin (BSA) (Sigma) and 10% normal swine serum (NSS) (DAKO Ltd, Ely, UK) in PBS for 30 min at room temperature. Cells were subsequently incubated in rabbit pre‐immune serum or Ab-hAIPL1 antiserum (1:250) in blocking solution for 1 h at room temperature, followed by swine anti-rabbit secondary antibody conjugated to the TRITC fluorophore (DAKO) in blocking solution for a further 1 h at room temperature. 4′6-diamidino-2-phenylindole (DAPI) (Sigma) was included in the final PBS wash before the cells were mounted in fluorescent mounting medium containing 15 mM sodium azide (DAKO) and visualized with a Zeiss laser scanning confocal microscope.
Adult human retinae were fixed with 10% neutral-buffered formalin within 2 min of enucleation for at least 24 h. The retinae were dehydrated with increasing concentrations of industrial methylated spirits (IMS), equilibrated in xylene and embedded in paraffin wax. Retinal sections were cut with a microtome at a setting of 8 µm, floated out onto 20% methanol, expanded on water pre-heated to 40°C and mounted on glass slides treated with 2% aminopropyltriethoxysilane (APES) (Sigma) in acetone. The retinal sections were deparaffined successively with xylene, 100 and 95% IMS before proceeding with the immunohistochemistry. Endogenous peroxidase activity was blocked by incubating the retinal sections with 0.5% hydrogen peroxide (Sigma) in methanol for 30 min at room temperature. Antigen retrieval was accomplished by microwaving the retinal sections four times for 2.5 min each at 800 W in Tris-buffered saline (TBS) (100 mM Tris–HCl, 150 mM NaCl, pH 7.6) containing 5% urea. The retinal sections were blocked with 10% NSS (DAKO) and 2% BSA in TBS for 45 min at room temperature, incubated with Ab‐hAIPL1 antiserum (1:500 or 1:1000) in TBS containing 0.1% BSA overnight at 4°C, and finally with biotin-conjugated swine anti-rabbit IgG (1:300) (DAKO) in TBS containing 0.1% BSA for 45 min at room temperature. The immunoreaction was visualized with streptavidin biotinylated horseradish peroxidase complex (DAKO) and 3′,3′-diaminobenzidine (DAB) (Sigma), 0.03% hydrogen peroxide in TBS. The retinal sections were successively dehydrated with 70, 95 and 100% IMS, clarified with three successive changes of xylene and mounted in dibutyl phthalate xylene (DPX) mounting medium (Merck, Lutterworth, UK). To verify the specificity of the immunostaining, retinal sections were also stained with rabbit pre‐immune serum and Ab-hAIPL1 antiserum pre-absorbed with the peptide epitope (30 µg/µl). The retinal sections were visualized with an Olympus BX50 light microscope using bright field and differential interference contrast (Nomarski) optics. Photography was done using an integral Olympus DP10 digital camera.
Fluorescence scanning confocal microscopy
Adult human retinae were fixed with freshly prepared 4% paraformaldehyde in isotonic PBS pH 7.3 within 2 min of enucleation for at least 2 h. The retinae were washed extensively with chilled PBS, embedded in 5% low melting temperature agarose and cut with a vibratome at a setting of 100 µm. The retinal sections were blocked with 5% normal donkey serum (NDS) (Jackson ImmunoResearch Laboratories, West Grove, PA) and 0.5% BSA in PBS overnight at 4°C and double labelled with Ab-hAIPL1 antiserum (1:1000) in the presence of a second retinal marker. Rhodopsin, a marker for rod photoreceptor outer segments, was detected with the mouse monoclonal antibody 1D4 (National Cell Culture Centre, USA) (1:200). Cone photoreceptor outer and inner segment ECM oligosaccharides were detected with PNA (1:200) (Vector laboratories, Burlingame, CA). Cone photoreceptors were detected with the mouse monoclonal antibody 7G6 (1:100) (a generous gift of Dr P.MacLeish, Morehouse School of Medicine Neuroscience Institute, USA). Secondary antibody incubations were performed with CY3-conjugated donkey anti-rabbit (Molecular Probes, Eugene, OR), CY2-conjugated donkey anti-mouse (Molecular Probes) and CY2-conjugated streptavidin (Molecular Probes). All antibody incubations were performed in block at 4°C overnight. DAPI was included in the final PBS wash and the retinal sections were mounted in fluorescent mounting medium containing 15 mM sodium azide (DAKO) and visualized with a Zeiss laser scanning confocal microscope. Retinal sections were stained with rabbit pre-immune serum and Ab-hAIPL1 antiserum pre-adsorbed with the peptide epitope (30 µg/ml) to verify the specificity of the immunostain. Retinal sections were also incubated with either the primary or secondary antibodies on their own to confirm that the signal observed did not arise non-specifically from any of the antibodies per se. In addition, rabbit primary antibodies were incubated with mouse secondary antibodies and vice versa to verify that immunosignals did not arise from cross-reaction between the antibodies in the double label procedure.
We are grateful to Dr D.Trump for the provision of Y79 retinoblastoma cells and Mr N.D.Ebenezer for the provision of HCEF and EK1B cells. We would like to thank Dr P.MacLeish for provision of the 7G6 cone marker. We are very grateful to Dr V.Pearson for assistance in immunohistochemical technique, and would also like to thank Mr R.Alexander and Miss R.Hall. We are grateful to the Brain Bank, Department of Neuropathology, Institute of Psychiatry, for the provision of human tissue samples. We are grateful for financial support from the Wellcome Trust (J.v.d.S. is the recipient of a Wellcome Trust Travelling Research Fellowship).
To whom correspondence should be addressed. Tel: +44 207 6086944; Fax: +44 207 6086862; Email: email@example.com