Retinal Localization and Copper-Dependent Relocalization of the Wilson and Menkes Disease Proteins

Abstract

purpose. Menkes and Wilson diseases are associated with retinal degeneration. The Menkes and Wilson genes are homologous copper transporters, but differences in their expression pattern lead to different disease manifestations. To determine whether the Wilson and Menkes genes may act locally in the retina, this study was undertaken to assess retinal Wilson and Menkes expression and localization.

methods. RT/PCR was used to test for the presence of Wilson and Menkes mRNAs in mouse and human retinas and retinal pigment epithelial cell lines. The Menkes and Wilson proteins were immunolocalized in human and mouse retinas and in the ARPE-19 cell line.

results. The Menkes mRNA and protein were present in the RPE and neurosensory retina whereas the Wilson mRNA and protein were limited to the RPE. In the RPE, both proteins localized to the Golgi. Increased copper concentration led to relocalization of the Wilson protein to a diffuse cytoplasmic distribution.

conclusions. Both the Menkes and Wilson proteins are present in the RPE. Since the RPE is a blood–brain barrier, these proteins most likely regulate not only their own copper levels but also copper levels of the overlying photoreceptors. Because the Wilson protein delivers copper to the ferroxidase ceruloplasmin in the liver, it is likely that the Wilson and/or Menkes proteins provide copper to ceruloplasmin made in the RPE. Retinopathy in Wilson and Menkes diseases may result not only from abnormal systemic copper levels but also from loss of retinal Wilson or Menkes protein.

Wilson disease (WD) is best known to ophthalmologists and eye researchers by the corneal Kayser-Fleischer ring and sunflower cataract resulting from copper deposition. WD is an autosomal recessive disorder of copper metabolism resulting from insufficient copper excretion into bile by hepatocytes, the sole mechanism of copper excretion. ¹ Excessive accumulation of copper in the hepatocyte cytoplasm causes cellular necrosis and leakage of copper into the plasma. Copper is also deposited in extrahepatic tissues, including the cornea and the brain. Patients have retinal degeneration and neurodegeneration with extrapyramidal, cerebellar, and cognitive dysfunction. 

The WD gene encodes a copper-transporting P-type ATPase with 65% amino acid identity to another copper transporter, the Menkes disease (MD) gene. ^{2 3 4} MD is an X-linked recessive disorder of copper metabolism resulting in mental retardation and progressive degeneration of the central nervous system with death in early childhood. ⁵ Patients with Menkes disease have abnormal “kinky” hair, seizures, and developmental delay and rarely survive beyond 3 years of age. ⁶ Loss of function of the MD protein results in copper deficiency in the developing fetus due to failure of copper transfer across the placenta, intestine, and blood–brain barrier, the predominant sites of MD gene expression. 

Both the MD and WD proteins are localized to the trans-Golgi network, concentrated in the perinuclear area under steady state conditions. Under these conditions, MD and WD proteins provide copper to proteins including ceruloplasmin, a ferroxidase thought to facilitate iron export from cells and enable iron binding to transferrin. ⁷ An increase in intracellular copper concentration results in rapid relocalization of MD and WD proteins to a cytoplasmic vesicular compartment ^{8 9} where they facilitate copper export from cells. 

Since Wilson and Menkes ATPases have similar activities, the significant clinical differences between Wilson and Menkes diseases are most likely due to differences in expression patterns. The WD gene is expressed in liver, kidney, and brain, whereas the MD gene is expressed in the placenta, heart, brain, testes, kidney, lung, fibroblasts, intestinal epithelial cells, and at low levels in the spleen and liver ¹⁰ as well as in the endothelium of the blood–brain barrier. ^{11 12} Both appear to be essential for copper export, as patients with WD have hepatocyte and central nervous system (CNS) copper overload. Patients with MD have elevated kidney and intestinal epithelial cell copper levels ¹³ despite the systemic copper deficiency resulting from failure of intestinal epithelial cells to export their copper into plasma. 

A splice variant of the WD gene, the pineal night-specific ATPase (PINA), is expressed in pineal and retina with 100-fold greater expression at night than during the day. PINA is expressed in cone photoreceptors, and has copper transport activity in yeast. These findings suggest a potential role of rhythmic copper metabolism in retinal circadian function. ¹⁴  

Within the CNS, the MD gene is expressed in the endothelium of the blood–brain barrier. ¹¹ It is also expressed in neurons including those of the hippocampus, where N-methyl-d-aspartate (NMDA) receptor activation results in relocalization of the Menkes protein, leading to cellular copper efflux. ¹⁵ WD protein is also expressed in some neurons and glia. Within the cerebellum, the WD gene is expressed in Purkinje cells whereas the MD gene is expressed in Bergmann glia. Studies with WD knockout mice reveal some degree of functional overlap between MD and WD proteins, ¹⁶ with ceruloplasmin expression shifting from the normal site of WD expression (neurons) to the site of MD expression (glia). 

Patients with either Menkes or Wilson disease have retinal degeneration. Patients typically have blond fundi, suggesting that these proteins may play a role in RPE or choroidal melanogenesis, potentially due to reduced activity of the copper-binding enzyme tyrosinase. Histologically, one patient with MD had marked loss of ganglion cells and swelling of mitochondria in ganglion cells and photoreceptors. There were dense RPE inclusions, including lipofuscin. An abnormal Bruch’s membrane lacked an elastic layer and contained banded structures similar to wide-spaced collagen. ¹⁷ ERGs are abnormal in patients with Menkes disease, showing profound delays in photopic b-wave implicit time and very abnormal rod-isolated wave forms. ¹⁸ Patients with WD have reduced amplitude of photopic a-waves. ¹⁹ A pigmentary maculopathy has been described in WD. ²⁰  

The retinal degenerations in Wilson and Menkes diseases could result from abnormal systemic copper levels or loss of retinal copper transporters. To determine whether Wilson and Menkes proteins may function within the retina, we assessed their retinal localization. 

Methods

Cell Culture

The RPE cell line ARPE-19 ²¹ was cultured in DMEM/F12 with 10% fetal bovine serum, penicillin-streptomycin, and l-glutamine. For copper-induced WD relocalization studies, cells were allowed to reach confluence in six-well culture dishes, and were then treated for 1 hour with 100 μM CuCl₂ (Sigma, St. Louis, MO) dissolved in culture media. Control cells received change of medium with no additional copper. 

Primary mouse RPE cells were cultured as follows: eyes from 3-week-old C57/BL6 mice were enucleated immediately after death. Globes were briefly immersed in 70% ethanol followed by removal of the anterior segment, vitreous, and retina in DMEM. The remaining eyecups were washed twice in PBS and then incubated in 0.25% trypsin in DMEM for 20 minutes at 37°C. RPE cells were then collected by pipetting with fresh DMEM supplemented with 20% fetal bovine serum, penicillin-streptomycin, and l-glutamine. Cells were maintained in this medium on laminin-coated plates at 37°C with 5% CO₂ and passaged 10 times before RNA extraction. After passaging, cells lost their pigmentation but maintained their hexagonal shape. 

RNA Purification

RNA was purified (TRIzol; Invitrogen, Carlsbad, CA) from primary cultured murine RPE and ARPE-19 cells and used to generate first-strand cDNA with a T7-(dT)24-mer primer and the reverse transcriptase (SuperScript II; Invitrogen). 

A human donor eye was obtained postmortem through the Alabama Eye Bank in accordance with the Declaration of Helsinki. Informed consent for donation and access to ophthalmic histories was obtained from eye donor next of kin. The donor was an 82-year-old man with a history free of retinal abnormalities. The postmortem interval was 3 hours. The eye was incised circumferentially at the equator, penetrating the sclera, retina, and vitreous. The retina and vitreous body were removed together after excision of the retina at the optic nerve head. The retina was dissected, frozen in liquid nitrogen, and stored at −80°C in microfuge tubes in extraction reagent (TRIzol; Invitrogen). The sample was then thawed on ice and homogenized for total RNA extraction. RNA was extracted from homogenized tissue with chloroform, precipitated with isopropanol, and dissolved in water after washing with 75% ethanol. First-strand cDNA was prepared according to the manufacturer’s protocol with oligo-(dT)12 to 18-mer primer using reverse transcriptase (SuperScript III; Invitrogen). The reaction was performed for 2 hours at 50°C, and the enzyme was inactivated by heating at 70°C for 15 minutes. The cDNA was then stored at −80°C until used for RT-PCR. 

RT-PCR

Primers for MD and WD genes were designed to span exon–intron boundaries for specific detection of cDNA rather than genomic DNA. 

Primers for the Human MD Gene, Used at an Annealing Temperature of 53°C

Forward primer 5′-TGATTGTGCTGGCAACCAC-3′ and reverse primer 5′-TCGTTTCTGTTCGGGAGATAC-3′, to amplify a 705-bp product spanning from exon 10 to exon 15. Forward primer 5′-AACCATTACTCACGGAACCC-3′ and reverse primer 5′-TCCCACATTAGCCATTGCC-3′, to amplify an 805-bp product spanning from exon 16 to exon 20. 

Primers for Human WD Gene, Used with an Annealing Temperature of 57°C

Forward 5′-GGCAAAGTCCCCACAATCAAC-3′ and reverse 5′-CCACCAGGATGACCAGAGAATAAAC-3′ primers to amplify a 824-bp product and spanning exons 3 to 6; forward 5′-AGGAGCCCTGTGACATTCTTCG-3 and reverse 5′-TGATGAGGATGCCGTTCTGC-3′ primers amplifying a 757-bp product spanning exons 8 to 13. 

Primers for the Murine MD and WD Genes

The murine Menkes forward primer was 5′-CTGCTCGGTCTATTGCTCTC-3′, with a reverse primer of 5′-CGAATCCTCTTGACTGTTTTCC-3′, used at an annealing temperature of 60°C. ²² Murine Wilson forward primer was 5′-AAACTCATGTCACTCCAAGCC-3′, with a reverse primer of 5′-CTTCGAGGACTTTCCCATCC-3′, used at an annealing temperature of 58°C. ²²  

The PCR mixture contained 1× PCR buffer, 0.2 mM dNTPs, 1 μM of each primer, 0.5 μL of first-strand cDNA as a template and 1.25 U of Taq polymerase (AmpiTaq Gold Polymerase; Roche, Indianapolis, IN). The reaction was performed with a thermal cycler (model 9800; Applied Biosystems, Inc. [ABI], Foster City, CA) with an initial 10-minute denaturation step at 95°C and 30 cycles of 94°C for 45 seconds, primer annealing at the specified temperatures for 45 seconds, and at 72°C for 1 minute with final elongation step of 10 minutes at 72°C. Amplification products were separated by electrophoresis on 1% agarose gels in 0.5× TAE (Tris-acetate-EDTA) buffer and visualized by ethidium bromide. 

Immunocytochemistry with ARPE 19 Cells

ARPE 19 cells were cultured in eight-well chamber slides (Laboratory-Tek; Nunc, Rochester, NY). After reaching full confluence, the cells were fixed with 4% paraformaldehyde for 5 minutes, and immunocytochemistry (ICC) was performed as described. ²³ Primary antibodies were rabbit anti-Wilson disease protein (1:250, a gift from Jonathan Gitlin, Washington University, St. Louis) ⁸ and sheep anti-trans-Golgi network protein (TGN) 46 (1:500, Serotec, Raleigh, NC). Control cells were treated identically, with omission of primary antibody. Cells were analyzed by wide-field epifluorescence microscopy using identical exposure times and magnification as described. ²³  

Immunohistochemistry on Mouse Retinal Sections

All procedures conformed to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the Institutional Animal Care and Use Committee of the University of Pennsylvania. BALB/c mouse eyes were immersion fixed with 4% paraformaldehyde for 2 hours, prepared as eyecups, and cryoprotected in 30% sucrose. The tissue was embedded in optimal cutting temperature compound (Tissue-Tek OCT; Sakura Finetek, Torrance, CA). Immunohistochemistry (IHC) was performed on 10-μm-thick sections, as described. ²³ Tissue was incubated with the following primary antibodies: rabbit anti-Menkes disease protein (1:100, a gift from J. Gitlin, Washington University, St. Louis) ⁵ and mouse anti-cis Golgi protein 130 g (1:50, BD Pharmingen, San Diego, CA) or rabbit anti-Menkes disease protein and goat anti-early endosomal protein EEA 1 (1:100, Santa Cruz Biotechnology, Santa Cruz, CA). 

Control sections were treated identically with omission of primary antibody. Sections were analyzed by fluorescence microscopy using identical exposure time and magnification. 

Human Retina Immunohistochemistry and Bleaching

Eyes were obtained postmortem from the Foundation Fighting Blindness (Owings Mills, MD) eye donor program. Eyes were fixed and cryosectioned as described. ²⁴ To eliminate the autofluorescence in the human retina, sections were first incubated in 0.25% KMnO₄ in PBS for 5 minutes at room temperature, rinsed with PBS twice, for 3 minutes each time and then incubated in 0.1% oxalic acid in PBS for 8 to 10 minutes at room temperature. Sections were again washed in PBS three times, 5 minutes each. Immunohistochemistry was then performed as described. ²³ Primary antibodies were rabbit anti-Menkes disease protein (1:100) and rabbit anti-Wilson disease protein (1:250). 

Results

To determine whether RPE cells express the MD and WD genes, RT/PCR was used to detect the mRNAs for each gene. To avoid amplification of genomic DNA, intron-spanning primers were selected. Both the spontaneously immortalized human ARPE-19 cell line and a primary culture of mouse RPE cells express both the MD and WD genes. The neurosensory retina (photoreceptors through ganglion cells, without RPE) from both mouse and human appeared to express the MD gene but not the WD gene (Fig. 1) . 

Within ARPE-19 cells, ICC detected WD protein in a perinuclear and cytoplasmic distribution (Fig. 2A) . The perinuclear label colocalized with TGN46, a trans-Golgi protein (Fig. 2B) . Previous studies in other cell types have shown redistribution of Wilson’s protein from the trans-Golgi to more widely distributed cytosolic vesicles ²⁵ in response to increased copper levels, most likely to transport copper into the vesicles for export. To determine whether WD protein in RPE cells behaves in a similar manner, ARPE-19 cells were exposed to elevated copper levels for 1 hour, then fixed for ICC. The cells exposed to elevated copper levels had more diffuse cytoplasmic WD protein localization (compare Figs. 2E 2F ). The specificity of the anti-WD antibody was previously demonstrated by Western blot analysis and ICC. ⁸  

To assess the localization of MD protein, cryosections of mouse retina were immunolabeled with an anti-MD antibody. The RPE, outer plexiform layer, and ganglion cell layers were positive for MD protein (Fig. 3) . Within the RPE, there was a punctate perinuclear and cytoplasmic label. This punctate label was in close proximity to the cis-Golgi protein gm130 (Fig. 3C) . MD protein does not colocalize with early endosomal protein EEA (Fig. 3D) . The specificity of the anti-MD antibody has been demonstrated by Western blot analysis and ICC, ⁵ and nonspecific adherence of the secondary antibody to the tissue was ruled out using a secondary-antibody–only control (Fig. 3A) . 

In the human RPE, IHC is technically challenging due to the pigmentation and autofluorescence of the RPE. A bleaching technique employing potassium permanganate and oxalic acid eliminated autofluorescence, but there was some distortion of the RPE morphology (Fig. 4A)making subcellular localization more difficult. Nevertheless, IHC reveals minimal label with secondary antibody alone and strong RPE label with both the anti-Menkes and anti-Wilson antibodies (Figs. 4B 4C 4D) . 

Similar to the anti-MD IHC pattern in mouse retina sections, MD protein is present in the OPL, INL, and GCL in human retina (Fig. 5) . The RPE fluorescence results from autofluorescent lipofuscin also visible in the secondary antibody alone control (Fig. 5A) . This control rules out nonspecific adherence of the secondary antibody to the tissue. Anti-WD IHC on human retinas (not shown) revealed no label in the neurosensory retina, consistent with the RT-PCR results. 

Discussion

Copper is an essential trace element facilitating electron transfer in many biochemical reactions, but can also cause oxidative damage to cells. Copper deficiency results in decreased activity of the enzymes lysyl oxidase, dopamine-β-hydroxylase, and tyrosinase, and impaired generation of collagen, elastin, keratin, ceruloplasmin, and melanin. ²⁶ Thus, copper overload or copper deficiency can lead to cellular dysfunction and death. The MD and WD proteins deliver copper to nascent proteins in the trans-Golgi, and, when copper levels are too high, the MD and WD proteins migrate to facilitate copper export. Our finding of expression of the MD and WD genes in the RPE and MD in the neurosensory retina suggest that these genes play an important role in retinal copper homeostasis. Loss of local retinal activity of MD and WD proteins may contribute to the retinal degeneration in Menkes and Wilson diseases. 

Within the RPE, the MD and WD proteins are likely to serve several functions related to copper trafficking. Because the RPE is the outer retinal blood–brain barrier, it may control copper transport to the outer retina after copper import by copper transport protein 1. ²⁷ The copper-dependent WD protein relocalization within ARPE-19 cells (Fig. 2)suggests that the WD protein may help maintain appropriate intracellular copper levels within RPE cells. WD or MD protein migration toward the apical or basal RPE may facilitate copper import or export across the blood–brain barrier from the retina. The expression of both MD and WD genes within a single cell type is unusual. The finding that these proteins can substitute for each other in cell lines ²⁸ suggests that they have overlapping functions. Our finding of expression of both MD and WD genes within RPE cells suggests that there may be some unique functions of each protein. 

Because the MD and WD proteins localized to the RPE Golgi, they are likely to provide copper transport to nascent cuproproteins. Tyrosinase, essential for melanogenesis, is one such cuproprotein. Indeed, macular mice, which harbor a mutation in the MD gene, ²⁹ have a reduced number of melanosomes within the RPE. ³⁰ RPE cells also express the copper-binding ferroxidases ceruloplasmin and hephaestin. ³¹ Because hepatic holoceruloplasmin synthesis requires copper transport by the WD protein, it is likely that the ceruloplasmin and hephaestin produced by the RPE are dependent on MD and/or WD proteins for their copper incorporation. Because mice deficient in ceruloplasmin and hephaestin accumulate iron in the RPE and have age-dependent retinal degeneration, ³¹ impairment of MD and WD function could lead not only to RPE copper abnormalities but also to iron accumulation. It would be of interest to test whether iron accumulation occurs. 

The finding of MD protein in the neurosensory retina (in the OPL and GCL) suggests that it plays a role in copper transport within the neurons residing in these layers. Within hippocampal neurons, NMDA receptor activation results in relocalization of the MD protein leading to cellular copper efflux which could influence neurotransmission. ¹⁵ It would be of interest to determine whether a similar relocalization of MD proteins occurs with NMDA receptor activation in the retina. Our study did not detect WD mRNA or protein in the neurosensory retina, which is not surprising, given the more limited tissue distribution of WD expression. Previous work indicates that a splice variant of WD (PINA) is expressed in cones, with expression most likely directed by a promoter located within an intron in the middle of the WD gene. ¹⁴ The PINA mRNA includes only the 3′ half of the WD gene and would not be amplified by the primers used in our RT-PCR experiments. PINA expression exhibits a dramatic diurnal rhythm in both pineal gland and retina with 100-fold greater expression at night than during the day. Because PINA has copper transport function when introduced into yeast, ¹⁴ it is possible the PINA functions in circadian photoreceptor copper transport. Because ceruloplasmin synthesis has not been detected in photoreceptors, PINA’s role in photoreceptors is more likely to be related to other cuproproteins or regulation of cellular copper levels. ^{32 33 34}  

The finding that oral supplementation of zinc plus copper reduced the risk of progression in patients with early age-related macular degeneration (AMD) ³⁵ suggests that metals play an important role in AMD and in retinal health. The rationale for inclusion of copper in the AREDS (Age-Related Eye Disease Study) was that zinc supplementation interferes with intestinal copper absorption and can result in copper deficiency. Copper deficiency, in turn, could result in retinal iron overload, as suggested by the ceruloplasmin/hephaestin-deficient mice. These studies highlight the interdependence of metals, suggesting that dietary deficiency of one metal, or mutation in one metal homeostasis gene, could affect levels of other metals. RPE expression of MD, WD, ceruloplasmin, and hephaestin suggests that the retina has local control mechanisms for metal homeostasis. The consequence of metal dysregulation, as in the ceruloplasmin/hephaestin-deficient mice, is retinal degeneration. 

 Supported by Research to Prevent Blindness (Career Development Award to JLD and an unrestricted award), National Eye Institute Grant EY015240, the International Retina Research Foundation, the F. M. Kirby Foundation, the Steinbach Foundation, a gift from Alma Cohen, and the Paul and Evanina Bell Mackall Foundation Trust.

 Submitted for publication December 15, 2005; revised February 3, 2006; accepted April 27, 2006.

 Disclosure: P. Krajaci c, None; Y. Qian, None; P. Hahn, None; T. Dentchev, None; N. Lukinova, None; J.L. Dunaief, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

RT-PCR amplification products of Menkes and Wilson genes separated by agarose gel electrophoresis. (A) RT-PCR products from mouse retinas and cultured mouse RPE cells. A band of the expected size (288 bp) is present in the amplification reaction using Menkes-specific primers and cDNA from both primary mouse RPE cells (mRPE) and mouse neurosensory retina without RPE (mRet). Primers specific for the Wilson gene generated a 180-bp amplification product with cDNA from mRPE but not mRet. A 50-bp increment ladder is shown for size reference. (B) RT-PCR products from ARPE-19 cells and human neurosensory retinas. Two sets of primers were used for the Menkes gene (M-1 and M-2) and two sets for the Wilson’s gene (W-1 and W-2). The cDNA template was from the source indicated at bottom. Bands of the expected size are present in the amplification reaction using Menkes-specific primers and cDNA from both ARPE-19 and human neurosensory retina without RPE (H-Ret). Primers specific for the Wilson gene generate an amplification product with cDNA from ARPE-19 but not H-Ret. A 100-bp increment ladder is shown for size reference.

Figure 2.

 

Fluorescence photomicrographs of ARPE-19 cells immunolabeled to detect WD protein. (A–C) Cells double labeled with anti-WD protein and anti-TGN 46, a trans-Golgi marker. (A) Green channel shows anti-WD immunoreactivity. Arrows: perinuclear labeling. (B) In double exposure, orange indicates colocalization (arrows) of WD and TGN46. (C) Red channel shows anti-TGN localization to the perinuclear area. (D) The no-primary-antibody control. (E) Perinuclear anti-WD immunoreactivity without Cu exposure (green, arrows). (F) Diffuse cytoplasmic anti-WD immunoreactivity (green) after a 1-hour treatment with 100 μM Cu. Nuclei are labeled with DAPI (blue). Scale bars, 25 μm.

Figure 3.

 

Fluorescence photomicrographs of BALB/c mouse retina immunolabeled with anti-MD antibody. (A) The no-primary-antibody control. (B) Anti-MD immunoreactivity (red) in the RPE, inner segments, outer plexiform layer (OPL) and ganglion cell layer (GCL). (C, bottom left) Double label with anti-MD and anti-cis-Golgi gm130. Anti-MD label (green) is in close proximity to anti-cis-Golgi gm130 (red). (D, bottom right) Double label with anti-MD and anti-early endosomal EEA. Anti-MD label (green) did not colocalize with anti-early endosomal EEA (red). Nuclei are labeled with DAPI (blue). Scale bars, 25 μm.

Figure 4.

 

Fluorescence photomicrographs of bleached human retinal sections immunolabeled to detect WD and MD protein. (A) Differential interference contrast (DIC) image of bleached RPE (arrow). (B) No-primary-antibody control. The RPE shows no autofluorescence in this bleached section. (C) Anti-WD immunoreactivity in the RPE (red). (D) Anti-MD label in the RPE (red). Nuclei are labeled with DAPI (blue). Scale bar, 50 μm.

Figure 5.

 

Fluorescence photomicrographs of human retinal sections immunolabeled to detect MD protein. (A) No primary antibody control. RPE autofluorescence appears reddish-yellow. (B) Anti-MD label (red) present in outer plexiform layer (OPL), and ganglion cell layer. Nuclei are labeled with DAPI (blue). Scale bar, 50 μm.

The authors thank Michael Marks for helpful suggestions and antibodies, and Christine Curcio for a human retina maintained in TRIzol.