Human Molecular Genetics Advance Access originally published online on April 6, 2006
Human Molecular Genetics 2006 15(10):1680-1689; doi:10.1093/hmg/ddl091
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Disease mechanisms in late-onset retinal macular degeneration associated with mutation in C1QTNF5
MRC Human Genetics Unit, Western General Hospital, Edinburgh EH4 2XU, UK
* To whom correspondence should be addressed. Tel: +44 1313322471; Fax: +44 1314678456; Email: xinhua.shu{at}hgu.mrc.ac.uk
Received January 31, 2006; Accepted March 29, 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSSUPPLEMENTARY MATERIALREFERENCES |
|---|
Late-onset retinal macular degeneration (L-ORMD) is an autosomal dominant condition resembling age-related macular degeneration (AMD) in which a key pathological feature is a thick extracellular sub-retinal pigment epithelial (RPE) deposit. L-ORMD is caused by mutation in the C1QTNF5 (CTRP5) short-chain collagen gene, but the disease mechanism is unknown. Here, we first show that wild-type C1QTNF5 is secreted, whereas mutant C1QTNF5 is misfolded and retained within the endoplasmic reticulum (ER). Secondly, the ER retained mutant protein has a shorter half-life than wild-type C1QTNF5 and is preferentially degraded by proteasomes. Thirdly, C1QTNF5 is shown to interact with the membrane-type frizzled related protein (MFRP), on the basis of yeast two-hybrid, protein pull-down and co-immunoprecipitation assays and RPE co-localization. These data suggest that L-ORMD is due to insufficient levels of secreted C1QTNF5, compromised RPE cell function resulting from ER retention of the mutant protein or both mechanisms.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSSUPPLEMENTARY MATERIALREFERENCES |
|---|
Age-related macular degeneration (AMD) is the commonest cause of blind registration in the developed world (1
). It is characterized by a late-onset degeneration of the retinal macula and represents the advanced stage of a more common disorder, age-related maculopathy. An important pathological feature of AMD is the accumulation of both focal (drusen) and diffuse extracellular (basal) deposits in the macula, between the retinal pigment epithelium (RPE) and the adjacent Bruch's membrane (2). These deposits lead to dysfunction and later death of RPE and associated photoreceptors. The nature of the proteins within the diffuse extracellular (basal) deposits have not been elucidated, but the focal deposits (drusen) include >100 proteins, together with lipids and glycosaminoglycans (3–5). Genetic factors are implicated in AMD on the basis of twin, family and association studies but it appears to be a genetically complex disorder (6).
An important means of elucidating disease mechanisms in a genetically complex disorder such as AMD, is to take advantage of information from a Mendelian disorder in which AMD is also present. Late-onset retinal macular degeneration (L-ORMD), originally named late-onset retinal degeneration (7–10), and also referred to as autosomal dominant haemorrhagic macular dystrophy (11) or late-onset macular degeneration (12), is a rare disorder with onset in the fifth to sixth decade with drusen-like deposits, followed by both macular degeneration and, in late stages, choroidal neovascularization and disciform scarring and severe peripheral chorioretinal atrophy (7–12). L-ORMD is caused by a single founder Ser163Arg mutation in the Complement 1q Tumour Necrosis Factor 5 gene (C1QTNF5, previously called CTRP5) (10,12). A primary feature of the disease is a thick (
50 mM) extracellular sub-RPE deposit, which is worst in the macula but extends into the peripheral retina (7,8). This deposit is indistinguishable at the microscopic level from basal deposits seen in AMD (8). This suggests that accumulation of basal deposits is a sufficient cause for progression to AMD, and L-ORMD has been described as an excellent model for the most severe ‘wet’ or neovascular form of AMD (11).
C1QTNF5 encodes an N-terminal signal peptide, a short collagen (Gly-X-Y) repeat and a C-terminal globular complement 1q (gC1q) domain, which is proposed to be necessary for trimerization of the collagen domain. It is widely expressed but shows highest expression in RPE and ciliary epithelium (10,12). The Arg163 mutant gC1q domain is unstable and forms high molecular weight aggregates in vitro (10). C1QTNF5 is expressed as a dicistronic transcript in RPE cells together with the membrane-type frizzled related protein (MFRP) gene (10,12), which is mutated in autosomal recessive nanophthalmos subjects (13) and in the mouse rd6 retinal degeneration (14). MFRP is an integral membrane protein of unknown function, which contains a cysteine-rich domain with close homology to frizzled, which can act as a Wnt or soluble Frizzled Related Protein (sFRP) receptor in intra-cellular signalling, and two cubilin (CUB) low-density lipoprotein receptor (LDLa) domains (13–15). Dicistronic transcripts commonly encode proteins involved in the same functional pathway (16).
In order to understand the pathogenic mechanisms associated with the C1QTNF5 mutation causing L-ORMD, we investigated the properties of C1QTNF5 with regard to secretion, cell adhesion, protein degradation and interactions. We found that the C1QTNF5 mutation abolished secretion and resulted in protein retention within the endoplasmic reticulum (ER) and preferential proteasomal degradation compared with wild-type protein. We also found that C1QTNF5 interacts with MFRP, indicating that C1QTNF5 and MFRP are involved in the same functional pathway.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSSUPPLEMENTARY MATERIALREFERENCES |
|---|
C1QTNF5 self-assembly is unaffected by mutation
Previous work showed that wild-type C1QTNF5 gC1q domain can form a major trimeric and weak hexameric species in vitro, while the mutant forms high molecular weight aggregates (10
). To explore the proposed oligomerization of C1QTNF5 further, we used glutathione S-transferase (GST) tagged wild-type gC1q domain to pull down hexa-histidine (His) tagged gC1q domain (Ser163, wild-type or Arg163, mutant) in vitro. The results showed that wild-type gC1q domain is involved in self-assembly for both wild-type and mutant, and that an interaction also occurred between mutant and wild-type gC1q domains (Fig. 1A and B). This was confirmed by yeast two-hybrid analysis, as a positive interaction was detected when either full length wild-type C1QTNF5 or its gC1q domain was paired with itself or with the mutant growing on triple selection plates lacking leucine (–Leu), tryptophan (–Trp) and histidine (–His), and metabolizing X-gal with ß-galactosidase by formation of blue colour (Fig. 1C and data not shown), further suggesting that self-interaction or cross-interaction of wild-type and mutant C1QTNF5 depends on the gC1q domain.
|
C1QTNF5 shares an aromatic motif with the fibrillar collagens and the C1q complement proteins (17
) (Fig. 1D), which is thought to be critical for the initiation of collagen alpha1(X) assembly (18). To explore the role of the aromatic motif in C1QTNF5 assembly, we constructed missense and deletion mutations in this domain (see Materials and Methods) and examined oligomerization of the resultant peptides. Deletion of the entire aromatic motif of C1QTNF5 gC1q domain resulted in complete loss of self-assembly, while mutation of the highly conserved aromatic amino acids corresponding at residues F143 (F143A) and G149 (G149A) resulted in loss of the trimeric form of the protein but retention of monomeric and dimeric forms (Fig. 1E). Mutation of the conserved F153 residue (F153A) resulted in total loss of self-assembly (Fig. 1E). The aromatic motif is therefore required for efficient multimerization of the C1QTNF5 gC1q domain.
C1QTNF5 mutation affects both secretion and cell adhesion
C1QTNF5 has a signal peptide and is proposed to be secreted into the extracellular matrix. To investigate whether the mutation in C1QTNF5 affects its secretion, stably transfected HEK-293 EBNA cell lines expressing similar level of either wild-type or mutant forms of C1QTNF5 were examined for their ability to secrete His-tagged protein (Fig. 2A). The results showed that while wild-type protein is secreted, the mutant is not secreted (Fig. 2A) and accumulates as a high molecular weight aggregates within the cell (Fig. 2B).
|
The effect of C1QTNF5 mutation on cell adhesion was also examined by plating HEK-293 EBNA cells on plates coated with laminin or fibronectin. The results showed that cells expressing mutant protein similar level to wild-type show a statistically significant reduction (P<0.00001) in cell adhesion to laminin-coated plates, compared with wild-type or control plates (Fig. 2C). No difference in cell adhesion was seen between wild-type and mutant cells plated on fibronectin-coated plates (data not shown).
Mutant C1QTNF5 is retained in the ER and preferentially degraded by the proteasome
Analysis of ARPE-19 cells transiently expressing either wild-type or mutant C-terminal GFP-tagged C1QTNF5 showed that cells expressing the mutant protein exhibit cytoplasmic aggregates which co-localize with the ER chaperones calreticulin (Fig. 3A) and protein disulphide isomerase (Fig. 3B), in contrast to wild-type (Fig.3A and B) or GFP control cells (data not shown). Although wild-type predominantly localizes in ER, some protein trafficks to Golgi apparatus and co-localizes with Golgi apparatus 58K protein. There is no co-localization of the aggregated mutant with Golgi apparatus 58K protein. (Supplementary Material, Fig. S1). Immunoblot analysis of non-denaturing gels for wild-type or mutant C1QTNF5 revealed that while wild-type migrated as a major trimeric and weak hexameric species, the mutant formed high-molecular weight aggregates which are retained in the stacking gel (data not shown).
|
Misfolded secretory proteins that are retained within the ER are often translocated from ER to the cytoplasm and degraded by proteasomes. To investigate whether wild-type or mutant C1QTNF5 are degraded by the proteasomes, transfected ARPE-19 cells were incubated with proteasome inhibitors MG132 or MG115, with DMSO as a control. The results showed that both wild-type and mutant protein increased markedly in both MG132 and MG115 treated cells compared with DMSO treated cells, suggesting that they are normally degraded by proteasomes (Fig. 3C). HEK-293 EBNA cells transiently expressing C-terminal His-tagged C1QTNF5 (wild-type or mutant) also showed the same result (data not shown). To determine the effect of proteasomes on the stability of wild-type versus mutant C1QTNF5, transfected cells were treated with cycloheximide (to block protein synthesis) and the levels of remaining protein analysed across time. Wild-type C1QTNF5 gradually declined with a half-life of the protein of >15 h, while the mutant was less stable with a half-life of about 8 h (Fig. 3D). These results suggest that the ER-retained mutant protein is preferentially degraded by proteasomes.
C1QTNF5 interacts with MFRP
As dicistronic transcripts show a tendency to be involved in the same functional pathway (16), we investigated whether C1QTNF5 and MFRP are also functionally related. We first used full-length C1QTNF5 as a bait to screen a human brain yeast two-hybrid cDNA library and identified 19 MFRP clones out of 363 positives (5%). Because C1QTNF5 and MFRP have functionally related gC1q and CUB domains, we tested for a direct interaction between them in protein pull-down assays. Purified His-tagged gC1q domain (wild-type or mutant) was incubated in vitro with GST-tagged fusion proteins containing both MFRP CUB domains separated by an LDLa domain (CUBT), the two CUB domains individually (CUB1, CUB2) and the cysteine-rich domain (CRD), compared with a GST-only control (Fig. 4A). After washing and elution from glutathione–Sepharose beads, interacting proteins were analysed by SDS–PAGE and immunoblotting. The analysis showed that both wild-type and mutant gC1q domains bind CUB-containing fusions, especially the CUBT fusion, but not the CRD or GST control (Fig. 4B and C).
|
Wild-type and mutant full length C1QTNF5 or its gC1q domain were used as ‘bait’ and the MFRP CUBT domain as ‘prey’ to test for direct interaction in a yeast two-hybrid analysis. A positive interaction is shown by the ability to grow on triple selection plates lacking leucine (–Leu), tryptophan (–Trp) and histidine (–His), as shown in Fig. 4D. This showed a positive interaction between the MFRP CUBT domain when paired with C1QTNF5 (both wild-type and mutant) or its gC1q domain (wild-type and mutant). The positive interactions were also confirmed by testing the ability to metabolize X-gal with ß-galactosidase and to form a blue colour (data not shown).
MFRP contains a potential membrane spanning domain (Fig. 4A) and so is predicted to be an integral membrane protein. The subcellular localization of MFRP transfected into HEK-293T cells showed MFRP to be predominantly localized to the plasma membrane (Fig. 5A). In contrast, His-tagged C1QTNF5 (wild-type or mutant) transfected into HEK-293T cells showed a diffuse cytoplasmic localization for the wild-type, while the mutant protein formed abnormal cytoplasmic aggregates (Fig. 5A). The interaction between C1QTNF5 gC1q domain and MFRP CUB domains was also supported by their co-localization within both cytoplasm and nucleus of HEK-293T cells, following co-transfection with CUBT (FLAG-tagged) and gC1q domain (His-tagged) constructs (Fig. 5B). The nuclear labelling of both proteins is possibly due to the fact that both gC1q (16.8 kDa) and CUBT (30 kDa) domains are below the size exclusion limit (<40 kDa) for diffusion through the nuclear pore.
|
To obtain further evidence for an interaction between C1QTNF5 and MFRP or its CUB domains, co-immunoprecipitation experiments were carried out in which full length MFRP or the CUBT construct (FLAG-tagged) were co-transfected with C1QTNF5 or its gC1q domain (His-tagged) into HEK-293T cells (Fig. 5C and D). Anti-FLAG antibody co-immunoprecipitated MFRP together with C1QTNF5 (wild-type and mutant) (Fig. 5C). Anti-FLAG antibody also co-immunoprecipitated the CUBT domain together with the gC1q domain (wild-type and mutant) (Fig. 5D).
Mouse retinal degeneration 6 (rd6) mutation does not affect MFRP localization or interaction with C1QTNF5
The rd6 mutant mouse carries a homozygous splice donor mutation in the Mfrp gene which results in the deletion of 58 amino acids, including three amino acids within the predicted transmembrane domain and the entire serine–threonine–proline (STP)-rich domain (14). First, we checked whether the deletion affects Mfrp localization, by transfection of HEK-293T cells using a FLAG-tagged mouse Mfrp gene. This showed that both wild-type and mutant Mfrp localize to the plasma membrane to the same extent (Fig. 6A). Secondly, to investigate whether the deletion affects the interaction between mouse Mfrp and C1qtnf5, FLAG-tagged Mfrp (wild-type or mutant) and GFP-tagged C1qtnf5 were co-transfected into HEK-293T cells, and anti-FLAG antibody used to co-immunoprecipitate Mfrp. As predicted, both wild-type and mutant Mfrp could co-immunoprecipitate with C1qtnf5 (Fig. 6B), while the GFP-only vector was not immunoprecipitated (data not shown). In addition, neither Mfrp nor C1qtnf5 was immunoprecipitated when normal mouse IgG was used as control (data not shown).
|
We next explored whether the rd6 deletion affects Mfrp structure assessed by analysis of denaturing and non-denaturing PAGE gels (Fig. 6C). This showed that the mutant protein forms increased high-molecular weight aggregates, under both denaturing and non-denaturing conditions.
Co-localization of C1qtnf5 and Mfrp in mouse retina
In order to define the localization of C1qtnf5 and Mfrp in mouse retina, albino mouse eye sections were examined for C1qtnf5 and Mfrp immunofluorescence using anti-human C1QTNF5 antibody, which recognizes cell expressed human C1QTNF5 (WT and MUT) in western blots (Supplementary Material, Fig. S2), and anti-human MFRP antibody, which recognizes cell-expressed human MFRP and mouse Mfrp in western blot (data not shown). The results (Fig. 7) showed that both C1qtnf5 and Mfrp co-localize in the RPE cell layer, together with the RPE-expressed protein Rpe65. Immunofluorecent labelling detected strong signals for both Mfrp and C1qtnf5 at both apical and basal plasma membranes.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSSUPPLEMENTARY MATERIALREFERENCES |
|---|
These results suggest two possible disease mechanisms linking the Ser163Arg mutation in C1QTNF5 with the retinal degeneration found in L-ORMD. The first is haploinsufficiency resulting from lack of secreted C1QTNF5 protein by RPE cells, which appears to be a consequence of the tendency of the mutant protein to form high-molecular weight aggregates (Fig. 2) (10
). The finding that mutant and wild-type C1QTNF5 can both self-interact and interact with each other (Fig. 1) is supported by our analysis of a conserved aromatic motif within its gC1q domain (10,17), which in collagen X is critical for the interaction between its NC1 domains and initiates assembly of multimers (18). Deletion or mutation of the aromatic motif in C1QTNF5, which spans residues 143–155 and is predicted to lie within the hydrophobic interior of the gC1q domain, results in loss of ability to assemble into normal higher-order structures (Fig. 1E). In contrast, the predicted surface location of the Ser163 residue (10) suggests that the aromatic interaction domain is functionally intact in the presence of the Ser163Arg mutation, allowing multimerization into higher-order assemblies. The observed interaction of wild-type and mutant gC1q domains also implicates a misfolded mutant protein which can inhibit the successful folding of newly synthesized wild-type protein by forming abnormal high-molecular weight aggregates, resulting in further reduction of normal protein secretion in heterozygotes. If equal amounts of wild-type and mutant C1QTNF5 are translated and associate at random, then only 25% is predicted to be present as a trimer and even less than this in the case of hexamers and higher-order multimers. Such higher-order structures are predicted to occur by analogy with the structurally closely related collagens VIII and X, which are believed to be secreted into the extracellular matrix and to form a three-dimensional hexagonal lattice (18,19).
The downstream consequences of reduced C1QTNF5 secretion are unclear, but the observation that reduced adhesion of cells stably expressing mutant compared with wild-type C1QTNF5 to laminin-coated plates is interesting. As secreted C1QTNF5 was only detectable in the cells expressing wild-type protein, and mutant protein appeared to be retained within the endoplasmic reticulum, one possibility is that haploinsufficiency of secreted C1QTNF5 results in impaired cell adhesion. Reduced secretion of wild-type C1QTNF5 could impair cellular signalling as a result of reduced interaction with MFRP. MFRP is co-expressed with C1QTNF5 as a dicistronic transcript and its product interacts with C1QTNF5 by means of its CUB and LDLa domains, on the basis of yeast two-hybrid, protein pull-down and co-immunoprecipitation experiments (Figs 4 and 5). The mouse orthologue of MFRP is both predicted and observed to show a plasma membrane localization in mouse RPE cells and co-localizes with C1QTNF5 at both apical and basal plasma membranes (Fig. 7). The ligand for the putative MFRP receptor function is unknown, but may be a Wnt or secreted frizzled molecule, while C1QTNF5 may act as a co-receptor in this presumed signalling pathway, which in the case of Wnt/frizzled receptors can influence both cell–cell and cell–matrix adhesions (20).
A second possible disease mechanism follows from the observation that mutant C1QTNF5 is retained within the ER and may give rise to ER stress which compromises RPE function in a non-specific manner (21). Many disease-causing mutations do not seriously compromise the synthesis of the polypeptide, but rather produce a protein that cannot fold normally (22). The ER has a quality-control system to monitor protein folding, maturation, transportation and activation of specific proteins post-translationally (21,22). Proteins entering the secretory pathway fold within the ER but misfolded proteins will not transit to the Golgi and instead are usually subject to ER-associated degradation, a quality-control system that recognizes and disposes misfolded or unassembled polypeptides in the ER. This involves the retro-translocation of aberrant proteins across the ER membrane to the cytosol, where they are ubiquitinated prior to degradation by the ubiquitin proteasome system. Some misfolded proteins are not degraded efficiently and are retained in the ER, where they contribute to the pathogenesis of several diseases associated with ER stress (23). Our results suggest that this mechanism is operating in L-ORMD, as in cells transfected with wild-type or mutant C1QTNF5, only the mutant is retained within the ER, where it forms high-molecular weight aggregates. Mutant C1QTNF5 also co-localizes with the resident ER chaperones calreticulin and protein disulphide isomerase (Fig. 3A and B), both of which participate in re-folding of mutant proteins in the ER. Mutant C1QTNF5 and mixed wild-type-mutant aggregates could accumulate in the ER due to their less efficient removal by proteasomal degradation (Fig. 3D). In contrast, the wild-type protein is not retained in the ER and shows a longer half-life, perhaps indicating that it is less susceptible to proteasomal degradation. The use of proteasomal inhibitors MG115 and MG132 led to an accumulation of both wild-type and mutant proteins (Fig. 3C), supporting the possibility that proteasomes are responsible for their degradation. However, there is a caveat that transfected cells over-expressing a protein may not reflect the in vivo situation, in which the capacity of the ubiquitin–proteasome system may be adequate for dealing with lower amounts of misfolded protein.
The interaction of C1QTNF5 and MFRP and the co-localization of these two proteins in the vicinity of the RPE plasma membrane raises the question as to its physiological or pathological significance. The function of MFRP is unknown, but it may contribute to both cell adhesion and intracellular signalling, via the canonical or non-canonical Wnt/ß-catenin pathways (20). The situation is complicated further by the different phenotypes associated with MFRP mutations in humans and mice. Homozygous loss-of-function mutations in the human MFRP gene cause nanophthalmos, a developmental disorder of ocular growth resulting in extreme hyperopia, associated with reduced axial length and increased scleral and choroidal vascular bed thickness (13). In contrast, an in-frame deletion of exon 4 in mouse Mfrp causes a retinal degeneration (rd6), raising the possibility of a species difference. We therefore compared the properties of wild-type and rd6 mouse Mfrp, which showed that the rd6 mutation does not affect its membrane localization, perhaps because only three amino acids of the predicted transmembrane domain are deleted (Fig. 6). The rd6 deletion, however, increased the tendency to form high-molecular weight aggregates, suggesting that this may represent a toxic gain-of-function rather than a loss-of-function, as distinct from a species difference. Deletion of the STP-rich domain in rd6, which can be important for protein interactions (24), may also be important in the degeneration phenotype.
The sub-RPE deposit of L-ORMD (7,8) is similar to that described both in AMD and in the macular dystrophy caused by mutations in TIMP-3 (25), raising the question as to whether they have similar pathogenic mechanisms. Mutations in TIMP-3 are found in Sorsby's fundus dystrophy, which shows considerable clinical and pathological similarities to L-ORMD (25). TIMP-3 is a secreted extracellular matrix protein, in which mutations cause abnormal protein interactions and abnormal cell adhesion (26,27) reminiscent of mutant C1QTNF5. Mutations in TIMP-3 also result in the accumulation of a turnover-resistant fraction of the TIMP-3 protein in ARPE-19 cells (28). Recent evidence suggests that the innate immune system plays an important role in the pathogenesis of AMD (29,30). Failure to digest or clear cellular debris extruded by RPE cells could result in the build up of sub-RPE deposits and immune attack. Our findings indicate that lack of secreted C1QTNF5 and accumulation of mutant C1QTNF5 in RPE cells, either individually or together, could lead to the build up of extracellular deposits and promote immune-mediated damage.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSSUPPLEMENTARY MATERIALREFERENCES |
|---|
Mutagenesis and DNA constructs
The S163R, F143A, G149A, F153A mutations and deletion of the aromatic motif within human C1QTNF5 gC1q domain were generated by a three-step-based PCR method. In order to generate the deletion of mouse Mfrp exon 4, the full length mouse Mfrp cDNA was cloned by RT–PCR using RNA isolated from adult mouse eye and ligated into pGEM-T easy vector (Promega) as described (31
). The deletion was introduced by a three-step PCR mutagenesis method using full-length mouse Mfrp cDNA in the pGEM-T easy vector as template. Full-length mouse C1qtnf5 cDNA was cloned by PCR using mouse brain marathon-ready cDNA (BD Biosciences) as template.
Human full-length C1QTNF5 (Ser163 or Arg163) was subcloned into the pCEP4 vector (Invitrogen) or pCEP-Pu/AC7 vector (32) with a hexaHistidine (His) tag at the C-terminal, or pEGFP-N1 vector. The gC1q domain (Ser163, wild-type or Arg163, mutant) of human C1QTNF5 was subcloned into the pCEP4 vector with a His tag at the C-terminal, or into pGEX-3T-3 (Amersham) or pET-21a (Novagen) vectors. Human MFRP full-length cDNA and CUBT domain (MFRP144–411) were cloned by PCR using Human Retinal Marathon-Ready cDNA (BD Biosciences) as template and subcloned into the pCEP4 vector with a FLAG tag at the C-terminal; Human MFRP CUB1domain (MFRP144–250), CUB2 domain (MFRP302–411), CUBT domain (MFRP144-411), CRD domain (MFRP466–579) were subcloned into the pGEX-3T-3 vector. Mouse Mfrp full-length cDNA or deletion mutant was subcloned into pCEP4 vector with a FLAG tag at the N-terminus. Mouse C1qtnf5 full-length cDNA was subcloned into the pEGFP N1 vector (BD Biosciences) with a GFP tag at the C-terminus. All constructs were verified by sequencing. All primers for the above constructs are available on request.
Protein expression in Escherichia coli
The pGEX-3T-3 vector containing human C1QTNF5 gC1q domain (Ser163 or Arg163), human MFRP CUB1, CUB2, CUBT or CRD domains (Fig. 4A), was transformed into E. coli BL21 (DE3) strain. The pET-21a vector containing human C1QTNF5 gC1q domain (Ser163, Arg163, F143A, G149A, F153A and aromatic motif deletion mutant) was also transformed into E. coli BL21 (DE3) strain. Expression was induced with 1 mM isopropyl-ß-D-thiogalactopyranoside at 18°C, and cells were centrifuged and disrupted using a French press in cell lysis buffer (150 mM NaCl, 50 mM Tris–HCl, 10% glycerol, pH 8.0) containing proteinase inhibitor (Roche). The cell lysate was centrifuged for 30 min at 13 000 rpm and the supernatant was kept at –70°C for further analysis. His-tagged gC1q fusion protein (wild-type or mutant) was purified with Ni-NTA super-flow (QIAGEN) following the manufacturer's instructions.
Cell culture and transfection
Human embryonic kidney (HEK) 293T cells and 293 EBNA cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (FCS) and penicillin/streptomycin. ARPE-19 cells were maintained in DMEM:F-12 medium supplemented with 10% FCS and penicillin/streptomycin. The cells were grown in six-well plates and transfected with the above constructs. For generation of cell lines stably expressing human C1QTNF5 (wild-type or mutant), HEK 293-EBNA cells were transfected with pCET-Pu/AC7 vector with C1QTNF5 (wild-type or mutant) and selected in medium containing 2 µg/ml puromycin. For the analysis of proteasome inhibitors MG132 (C2211, Sigma) and MG115 (C6706, Sigma) (33) and cycloheximide treatment, ARPE-19 cells were grown in six-well plates and were transfected with human full-length C1QTNF5 (Ser163 or Arg163)-EGFP constructs. MG132 (20 µg/ml, Sigma), MG115 (30 µM, Sigma) and Cycloheximide (50 µg/ml, Sigma) were applied to the cells 24 h after transfection. Cells were incubated for 16 h with the MG132 or MG115, or 0–24 h with cycloheximide before harvesting. Harvested cells were lysed with sample buffer (50 mM Tris–HCl, pH 6.8, 0.1% bromophenol blue, 10% glycerol, 1% SDS) and analysed by western blotting.
Cell adhesion assay
Human laminin (Sigma) was added to 96-well non-tissue culture-treated plates (0.5 µg/well) and allowed to dry at 37°C for 2 h. Wells were washed and incubated with 3% bovine serum albumin (BSA) in PBS for 1 h at 37°C. Subconfluent HEK-293 EBNA cells were harvested, washed, resuspended in culture medium (1x105 cells/ml) and allowed to attach to the wells for 1 h at 37°C. Non-adherent cells were removed by washing, and the attached cells were stained for 10 min with 0.2% crystal violet. The wells were washed three times with PBS, and cell-associated crystal violet was eluted by addition of 100 µl of 10% acetic acid. Cell adhesion was quantified by measuring the optical density of eluted crystal violet at a wavelength of 600 nm.
Immunocytochemistry
ARPE-19 Cells or HEK-293T cells were grown on coverslips to 
70% confluence then transfected with the constructs using GeneJuice transfection reagent (Novagen) following the manufacturer's protocol. After incubation for 48 h, cells were washed in PBS, fixed with ice-cold methanol, blocked with 2% BSA in PBS, then incubated with primary antibodies. Cells were washed in PBS, blocked again and incubated with Texas Red or FITC conjugated secondary antibodies. Cells were mounted in Vectashield (Vector Laboratories, Inc. Burlingame, CA, USA) containing DAPI (1.0 µg/ml). Images were captured using an Axioplan 2 fluorescent microscope and analysed using IPLab software.
Fluorescence staining of mouse eye sections
10 µM-thick cryosections of adult albino mouse eyes were air-dried for 1 h at room temperature (RT), then fixed in 4% paraformaldehyde in PBS for 10 min at 4°C. They were then incubated in PBS with 0.1% Triton-X-100 (PBST) for 10 min, blocked in 4% BSA in PBST for 1 h at RT and re-immersed in PBST for three times for 5 min each. The primary antibodies [anti-human C1QTNF5 antibody raised in rabbit immunized with purified thioredoxin fused gC1q domain (10), anti-human MFRP antibody raised in goat immunized with purified recombinant human MFRP C-terminal extracellular domain (R&D Systems, Cat. Number: AF1915) and anti-human RPE65 raised in mouse, kindly provided by Dr Debra A. Thompson (Kellog Eye Institute, University of Michigan)] were diluted 1:200 in the blocking solution and applied on the sections for 1 h at RT. After three 5-min washes in PBST, the sections were treated with either TxRed- or FITC-conjugated secondary antibody, diluted 1:400 in the blocking solution, for 1 h at RT in the dark. Another series of three 5-min washes followed before mounting the slides with Vectashield (Vector Labs) with DAPI. The sections were visualized by fluorescent microscopy.
Pull-down assay
The GST, GST-gC1q (wild-type or mutant), GST-CUB1, GST-CUB2, GST-CUBT and GST-CRD proteins were incubated with purified His-tagged gC1q domain (wild-type or mutant) on glutathione–Sepharose 4B beads. Bound proteins were washed and removed by boiling, then separated by SDS–PAGE and analysed by immunoblotting.
Immunoprecipitation
HEK-293T cells were co-transfected with either full length MFRP or its CUB-LDLa-CUB (CUBT) domain together with either full length C1QTNF5 (wild-type or mutant) or its gC1q domain (wild-type or mutant) and grown for 3 days. The transfected cells were harvested and lysed in a buffer containing 50 mM HEPES pH 7.4, 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate and protease inhibitors. After 20 min on ice, the cells were centrifuged at 13 000 rpm for 10 min at 4°C. The supernatant was used for co-immunoprecipitation as previously described (34). The immunoprecipitates were subjected to SDS–PAGE followed by immunoblotting with appropriate antibodies.
Immunoblotting analysis
Proteins from cell lysates or E. coli expression were separated by 12% SDS–PAGE gels and transferred to nitrocellulose membranes. Native (non-denaturing) gels were prepared by omitting SDS and dithiothreitol from the standard Laemmli SDS–PAGE protocol. Membranes were blocked in 5% non-fat dried milk in PBS. Primary antibodies were used at 1:3000 [anti-C terminal His tag (Invitrogen) and anti-FLAG tag (Sigma)] or 1:500 for anti-GFP antibody (BD Biosciences). Anti-rabbit or anti-mouse HRP-conjugated secondary antibodies (Amersham Biosciences) were used at 1:5000 dilution. Bound antibody was visualized by ECL (Amersham Biosciences). Band intensity was qualified using Quantity One-4.4.1 software (Bio-Rad).
Yeast two-hybrid analysis
Wild-type and mutant full length C1QTNF5 or its gC1q domain were inserted into the pGBKT7 or pGADT7 vector (BD Biosciences), the MFRP CUBT domain (MFRP144–411) was inserted into pGADT7 vector and transformed into yeast and grown on the selection medium lacking leucine, tryptophan and histidine (–L/T/H) as described previously (34). A colour assay for ß-galactosidase activity and autoactivation tests were carried out according as described (BD Science Yeast Protocols Handbook PT3024-1). For library screening, the pGBKT7 vector containing full length human C1QTNF5 was used as bait for screening a human brain yeast two-hybrid cDNA library (BD Biosciences) according to the manufacturer's instructions.
|
SUPPLEMENTARY MATERIAL |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSSUPPLEMENTARY MATERIALREFERENCES |
|---|
Supplementary Material is available at HMG Online.
|
ACKNOWLEDGEMENTS |
|---|
We thank UK Medical Research Council, UK Fight for Sight, EVI-GENORET (European Union), the Chief Scientist's Office of Scotland and British Retinitis Pigmentosa Society for financial support. We also thank S. Bruce for the artwork.
Conflict of Interest statement. None declared.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSSUPPLEMENTARY MATERIALREFERENCES |
|---|
-
Friedman, D.S., O'Colmain, B.J., Munoz, B., Tomany, S.C., McCarty, C., de Jong, P.T., Nemesure, B., Mitchell, P., Kempen, J. and Eye Diseases Prevalence Research Group. (2004) Prevalence of age-related macular degeneration in the United States. Arch. Ophthalmol., 122, 564–572.
[Abstract/Free Full Text] - Abdelsalam, A., Del Priore, L. and Zarbin, M.A. (1999) Drusen in age-related macular degeneration: pathogenesis, natural course, and laser photocoagulation-induced regression. Surv. Ophthalmol., 44, 1–29.[CrossRef][Web of Science][Medline]
-
Crabb, J.W., Miyagi, M., Gu, X., Shadrach, K., West, K.A., Hirokazu, S., Kamei, M., Hasan, A., Yan, L., Rayborn, M.E. et al. (2002) Drusen proteome analysis: an approach to the etiology of age-related macular degeneration. Proc. Natl. Acad. Sci. USA, 99, 14682–14687.
[Abstract/Free Full Text] -
Malek, G., Li, C.M., Guidry, C., Medeiros, N.E. and Curcio, C.A. (2003) Apolipoprotein B in cholesterol-containing drusen and basal deposits of human eyes with age-related maculopathy. Am. J. Pathol., 162, 413–425.
[Abstract/Free Full Text] -
Curcio, C.A., Millican, C.L., Bailey, T. and Kruth, H.S. (2001) Accumulation of cholesterol with age in human Bruch's membrane. Invest. Ophthalmol. Vis. Sci., 42, 265–274.
[Abstract/Free Full Text] - Hammond, C.J., Webster, A.R., Snieder, H., Bird, A.C., Gilbert, C.E. and Spector, T.D. (2002) Genetic influence on early age-related maculopathy: a twin study. Ophthalmology, 109, 730–736.[CrossRef][Web of Science][Medline]
-
Kuntz, C.A., Jacobson, S.G., Cideciyan, A.V., Li, Z.Y., Stone, E.M., Possin, D. and Milam, A.H. (1996) Sub-retinal pigment epithelial deposits in a dominant late-onset retinal degeneration. Invest. Ophthalmol. Vis. Sci., 37, 1772–1782.
[Abstract/Free Full Text] - Milam, A.H., Curcio, C.A., Cideciyan, A.V., Saxena, S., John, S.K., Kruth, H.S., Malek, G., Heckenlively, J.R., Welebber, R.G. and Jacobson, S.G. (2000) Dominant late-onset retinal degeneration with regional variation of sub-retinal pigment epithelium deposits, retinal function, and photoreceptor degeneration. Ophthalmology, 107, 2256–2266.[CrossRef][Web of Science]
-
Jacobson, S.G., Cideciyan, A.V., Wright, E. and Wright, A.F. (2001) Phenotypic marker for early disease detection in dominant late-onset retinal degeneration. Invest. Ophthalmol. Vis. Sci., 42, 1882–1890.
[Abstract/Free Full Text] -
Hayward, C., Shu, X., Cideciyan, A.V., Lennon, A., Barran, P., Zareparsi, S., Sawyer, L., Hendry, G., Dhillon, B., Milam, A.H. et al. (2003) Mutation in a short-chain collagen gene, CTRP5, results in extracellular deposit formation in late-onset retinal degeneration: a genetic model for age-related macular degeneration. Hum. Mol. Genet., 12, 2657–2667.
[Abstract/Free Full Text] -
Ayyagari, R., Griesinger, I.B., Bingham, E., Lark, K.K., Moroi, S.E. and Sieving, P.A. (2000) Autosomal dominant hemorrhagic macular dystrophy not associated with the TIMP3 gene. Arch. Ophthalmol., 118, 85–92.
[Abstract/Free Full Text] -
Ayyagari, R., Mandal, M.N., Karoukis, A.J., Chen, L., McLaren, N.C., Lichter, M., Wong, D.T., Hitchcock, P.F., Caruso, R.C., Moroi, S.E. et al. (2005) Late-onset macular degeneration and long anterior lens zonules result from a CTRP5 gene mutation. Invest. Ophthalmol. Vis. Sci., 46, 3363–3371.
[Abstract/Free Full Text] -
Sundin, O.H., Leppert, G.S., Silva, E.D., Yang, J.M., Dharmaraj, S., Maumenee, I.H., Santos, L.C., Parsa, C.F., Traboulsi, E.I., Broman, K.W. et al. (2005) Extreme hyperopia is the result of null mutations in MFRP, which encodes a Frizzled-related protein. Proc. Natl. Acad. Sci. USA, 102, 9553–9558.
[Abstract/Free Full Text] -
Kameya, S., Hawes, N.L., Chang, B., Heckenlively, J.R., Naggert, J.K. and Nishina, P.M. (2002) Mfrp, a gene encoding a frizzled related protein, is mutated in the mouse retinal degeneration 6. Hum. Mol. Genet., 11, 1879–1886.
[Abstract/Free Full Text] - Katoh, M. (2001) Molecular cloning and characterization of mFRP, a novel gene encoding a membrane-type Frizzled-related protein. Biochem. Biophys. Res. Commun., 282, 116–123.[CrossRef][Web of Science][Medline]
- Blumenthal, T. (1996) Gene clusters and polycistronic transcription in eukaryotes. Bioessays, 20, 480–487.
- Brass, A., Kadler, K.E., Thomas, J.T., Grant, M.E. and Boot-Handford, R.P. (1992) The fibrillar collagens, collagen VIII, collagen X and the C1q complement proteins share a similar domain in their C-terminal non-collagenous regions. FEBS Lett., 303, 126–128.[CrossRef][Web of Science][Medline]
-
Chan, D., Freddi, S., Weng, Y.M. and Bateman, J.F. (1999) Interaction of collagen alpha1(X) containing engineered NC1 mutations with normal alpha1(X) in vitro. Implications for the molecular basis of schmid metaphyseal chondrodysplasia. J. Biol. Chem., 274, 13091–13097.
[Abstract/Free Full Text] -
Stephan, S., Sherratt, M.J., Hodson, N., Shuttleworth, C.A. and Kielty, C.M. (2004) Expression and supramolecular assembly of recombinant
1(VIII) and 2(VIII) collagen homotrimers. J. Biol. Chem., 279, 21469–21477.[Abstract/Free Full Text] - Schambony, A., Kunz, M. and Gradl, D. (2004) Cross-regulation of Wnt signaling and cell adhesion. Differentiation, 72, 307–318.[CrossRef][Medline]
- Xu, C., Bailly-Maitre, B. and Reed, J.C. (2005) Endoplasmic reticulum stress: cell life and death decisions. J. Clin. Invest., 115, 2656–2664.[CrossRef][Web of Science][Medline]
- Ellgaard, L. and Helenius, A. (2003) Quality control in the endoplasmic reticulum. Nat. Rev. Mol. Cell. Biol., 4, 181–191.[CrossRef][Web of Science][Medline]
- Trombetta, E.S. and Parodi, A.J. (2003) Quality control and protein folding in the secretory pathway. Annu. Rev. Cell Dev. Biol., 19, 649–676.[CrossRef][Web of Science][Medline]
-
Ingham, R.J., Colwill, K., Howard, C., Dettwiler, S., Lim, C.S.H., Yu, J., Hersi, K., Raaijmakers, J., Gish, G., Mbamalu, G. et al. (2005) WW domains provide a platform for the assembly of multiprotein networks. Mol. Cell. Biol., 25, 7092–7106.
[Abstract/Free Full Text] - Weber, B.H.F., Vogt, G., Pruett, R.C., Stohr, H. and Felbor, U. (1994) Mutation in the tissue inhibitor of metalloproteinases-3 (TIMP-3) in patients with Sorsby's fundus dystrophy. Nat. Genet., 8, 352–355.[CrossRef][Web of Science][Medline]
- Yeow, K.M., Kishnani, N.S., Hutton, M., Hawkes, S.P., Murphy, G. and Edwards, D.R. (2002) Sorsby's fundus dystrophy tissue inhibitor of metalloproteinases-3 (TIM-3) mutants have unimpaired matrix metalloproteinase inhibitory activities, but affect cell adhesion to the extracellular matrix. Matrix Biol., 21, 75–88.[CrossRef][Web of Science][Medline]
-
Qi, J.H., Ebrahem, Q., Yeow, K., Edwards, D.R., Fox, P.L. and Anand-Apte, B. (2002) Expression of Sorsby's fundus dystrophy mutations in human retinal pigment epithelial cells reduces matrix metalloproteinase inhibition and may promote angiogenesis. J. Biol. Chem., 277, 13394–13400.
[Abstract/Free Full Text] - Langton, K.P., Mckie, N., Smith, B.M., Brown, N.J. and Barker, M.D. (2005) Sorsby's fundus dystrophy mutations impair turnover of TIMP-3 by retinal pigment epithelial cells. Hum. Mol. Biol., 14, 3579–3586.
-
Daiger, S.P. (2005) Genetics. Was the Human Genome Project worth the effort? Science, 308, 362–364.
[Abstract/Free Full Text] - Hageman, G.S., Luthert, P.J., Victor Chong, N.H., Johnson, L.V., Anderson, D.H. and Mullins, R.F. (2001) An integrated hypothesis that considers drusen as biomarkers of immune-mediated processes at the RPE-Bruch's membrane interface in aging and age-related macular degeneration. Prog. Retin. Eye Res., 20, 705–732.[CrossRef][Web of Science][Medline]
- Shu, X., Zeng, Z., Eckmiller, M.S., Gautier, P., Vlachantoni, D., Manson, F.D.C., Tulloch, B., Sharp, C., Gorecki, D.C. and Wright, A.F. (2006) Developmental and tissue expression of Xenopus laevis. RPGR. Invest. Ophthalmol. Vis. Sci., 47, 348–356.[CrossRef]
- Kohfeldt, E., Maurer, P., Vannahme, C. and Timpl, R. (1997) Properties of the extracellular calcium binding module of the proteoglycan testican. FEBS Lett., 414, 557–561.[CrossRef][Web of Science][Medline]
- Lee, D.H. and Goldberg, A.L. (1998) Proteasome inhibitors: valuable new tools for cell biologists. Trends Cell Biol., 8, 397–403.[CrossRef][Web of Science][Medline]
-
Shu, X., Fry, A.M., Tulloch, B., Manson, F.D.C., Crabb, J.W., Khanna, H., Faragher, A.J., Lennon, A., He, S., Trojan, P. et al. (2005) RPGR ORF15 isoform co-localizes with RPGRIP1 at centrioles and basal bodies and interacts with nucleophosmin. Hum. Mol. Genet., 14, 1183–1197.
[Abstract/Free Full Text]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S Borooah, C Collins, A Wright, and B Dhillon Late-onset retinal macular degeneration: clinical insights into an inherited retinal degeneration Postgrad. Med. J., September 1, 2009; 85(1007): 495 - 500. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S Borooah, C Collins, A Wright, and B Dhillon Late-onset retinal macular degeneration: clinical insights into an inherited retinal degeneration Br J Ophthalmol, March 1, 2009; 93(3): 284 - 289. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. H. Jeon, K.-y. Kim, J. H. Kim, A. Baek, H. Cho, Y. H. Lee, J. W. Kim, D. Kim, S. H. Han, J.-S. Lim, et al. A novel adipokine CTRP1 stimulates aldosterone production FASEB J, May 1, 2008; 22(5): 1502 - 1511. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Tuo, C. M. Bojanowski, M. Zhou, D. Shen, R. J. Ross, K. I. Rosenberg, D. J. Cameron, C. Yin, J. A. Kowalak, Z. Zhuang, et al. Murine Ccl2/Cx3cr1 Deficiency Results in Retinal Lesions Mimicking Human Age-Related Macular Degeneration Invest. Ophthalmol. Vis. Sci., August 1, 2007; 48(8): 3827 - 3836. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. N. A. Mandal, V. Vasireddy, M. M. Jablonski, X. Wang, J. R. Heckenlively, B. A. Hughes, G. B. Reddy, and R. Ayyagari Spatial and Temporal Expression of MFRP and Its Interaction with CTRP5 Invest. Ophthalmol. Vis. Sci., December 1, 2006; 47(12): 5514 - 5521. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||