|Home | About | Journals | Submit | Contact Us | Français|
Disruption of iron homeostatsis within the central nervous system (CNS) can lead to profound abnormalities during both development and aging in mammals. The radiation-induced polycythaemia (Pcm) mutation, a 58-bp microdeletion in the promoter region of ferroportin 1 (Fpn1), disrupts transcriptional and post-transcriptional regulation of this pivotal iron transporter. This regulatory mutation induces dynamic alterations in peripheral iron homeostatis such that newborn homozygous Pcm mice exhibit iron deficiency anemia with increased duodenal Fpn1 expression while adult homozygotes display decreased Fpn1 expression and anemia despite organismal iron overload. Herein we report the impact of the Pcm microdeletion on iron homeostasis in two compartments of the the central nervous system: brain and retina. At birth, Pcm homozygotes show a marked decrease in brain iron content and reduced levels of Fpn1 expression. Upregulation of transferrin receptor 1 (TfR1) in brain microvasculature appears to mediate the compensatory iron uptake during postnatal development and iron content in Pcm brain is restored to wildtype levels by 7 weeks of age. Similarly, changes in expression are transient and expression of Fpn1 and TfR1 is indistinguishable between Pcm homozygotes and wildtype by 12 weeks of age. Strikingly, the adult Pcm brain is effectively protected from the peripheral iron overload and maintains normal iron content. In contrast to Fpn1 downregulation in perinatal brain, the retina of Pcm homozygotes reveals increased levels of Fpn1 expression. While retinal morphology appears normal at birth and during early postnatal development, adult Pcm mice demonstrate a marked, age-dependent loss of photoreceptors. This phenotype demonstrates the importance of iron homeostasis in retinal health.
Iron functions as an essential co-factor for heme and multiple non-heme proteins (Hentze et al., 2004). In addition to meeting a physiological demand, tight regulation of cellular and organismal iron levels also prevents the untoward oxidation-reduction chemistry of iron overload (Aisen et al., 2001). While recent years have witnessed significant progress in elucidating the mechanisms of peripheral iron homeostasis in mammals, the regulation of iron balance in the central nervous system (CNS) remains poorly understood. Iron-dependent processes in the CNS include regulation of the amount and composition of myelin (Kwik-Uribe et al., 2000), synthesis and degradation of neurotransmitters by enzymes such as tryptophan hydroxylase and tyrosine hydroxylase (Yehuda and Youdim, 1989), and control of GABAergic activity (Taneja et al., 1986). Insights gleaned from both dietary manipulation and hereditary disorders underscore the critical role of iron and iron-related genes in brain function (Beard, 2003; Ponka, 2004). Early stages of neurodevelopment in humans and various model organisms appear particularly vulnerable to iron deficiency and, despite subsequent repletion of CNS iron balance, behavioral and biochemical abnormalities can manifest at later stages of life (Beard, 2003; Beard, 2007; Beard and Connor, 2003).
Beyond development, recent studies also implicated abnormal iron homeostasis in neurodegeneration. For example, patients with neuroferritinopathy, Friedreich's Ataxia, panthothenate kinase-associated neurodegeneration (PKAN) or aceruloplasminemia manifest iron deposits in various regions of the CNS and have severe neurological abnormalities, including extrapyramidal dysfunction or cerebellar ataxia (Curtis et al., 2001; Pandolfo, 1999; Xu et al., 2004; Zhou et al., 2001). Likewise, loss of function of the ferroxidase ceruloplasmin (Cp), which is required for iron efflux from astrocytes (Jeong and David, 2003), revealed iron accumulation in murine CNS (Jeong and David, 2006; Patel et al., 2002). Severe defects in brain iron homeostasis and degenerative movement disorder were reported for targeted disruption of Iron regulatory protein 2 (LaVaute et al., 2001).
Aside from brain and spinal cord, the retina represents another CNS region with increased susceptibility to iron overload-induced degenerative sequelae. Using Perls' histochemistry to detect iron, elevated iron levels were detected in post mortem retinas from age-related macular degeneration (AMD) donors compared to healthy age-matched controls (Hahn et al., 2003). In addition, a case of AMD with geographic atrophy had increased immunolabeling for ferritin and ferroportin 1 compared to an age-matched control (Dentchev et al., 2005). Occurrences of retinal abnormalities have also been reported in patients with PKAN, Friedreich's Ataxia, and Hereditary Hemochromatosis (He et al., 2007). Maculopathy has been demonstrated in patients with aceruloplasminemia, who accumulate iron by their fifth or sixth decade of life in retina and others tissues (Dunaief et al., 2005; Miyajima et al., 1987; Morita et al., 1995; Yamaguchi et al., 1998). Similarly, mice deficient for Cp and its homolog Heph (Hephaestin) had morphologic features reminiscent of AMD, including retinal iron accumulation, pigment epithelium hypertrophy, hyperplasia and death, photoreceptor degeneration and subretinal neovascularization (Hadziahmetovic et al., 2008; Hahn et al., 2004).
The ferroxidases Cp and Heph oxidize iron from the ferrous (Fe2+) to the ferric form (Fe3+) in support of cellular iron export by ferroportin 1 (Fpn1). Based on the functional cooperation between ferroxidases and Fpn1 in the regulation of iron efflux, it is expected that disruption of Fpn1 activity would cause aberrant iron homeostasis in the CNS, including retina. The present study tests this hypothesis in polycythaemia (Pcm) mice, which carry a radiation-induced 58-bp microdeletion in the Fpn1 promoter region (Mok et al., 2004a). The polycythaemia mutation is characterized by an erythropoietin-dependent increase in red blood cells in heterozygotes and a hypochromic, microcytic anemia in homozygotes. This regulatory mutation causes tissue-specific dysregulation of Fpn1 expression and results in the gamut of systemic iron homeostasis defects, ranging from iron deficiency at birth to tissue iron overload in adult Pcm mice (Mok et al., 2004a; Mok et al., 2004b; Mok et al., 2006).
Our results show perinatal iron deficiency in Pcm homozygous brain, which returns to normal iron levels by 7 weeks of age. Immunohistochemistry reveals a striking decrease in Fpn1 expression in smooth muscle cells of meningeal arterioles and endothelial cells of interstitial capillaries at birth, suggesting decreased iron import into the brain of Pcm homozygotes during development. Compensatory upregulation of transferrin receptor 1 (TfR1) expression in brain microvasculature appears to mediate the recovery from perinatal iron deficiency. In contrast, across the layers of the perinatal retina Pcm homozygotes demonstrate increased Fpn1 and TfR1 expression followed by an age-related degeneration of photoreceptors. Our results demonstrate that dysregulation on Fpn1 expression perturbs developmental brain iron homeostasis and leads to retinal degeneration, which is consistent with the importance of iron homeostasis in CNS health.
In the context of an embryonic iron deficiency (Mok et al., 2004a; Mok et al., 2004b), Pcm pups demonstrated a profound decrease in brain iron content at P0 (Fig. 1A). Notably, Pcm heterozygotes and homozygotes contained only approximately 70% and 25%, respectively, of wild-type brain iron levels. At 3 weeks, the iron content of Pcm brain approximated 85% of wild type, reflecting compensatory iron uptake during early postnatal development (Fig. 1A). However, iron balance in Pcm brain remained significantly lower compared with wild type, indicating a protracted recovery from the severe developmental iron deficiency. In contrast to the marked iron accumulation in visceral organs (Mok et al., 2004a), the iron content in 7- and 12-week-old Pcm brain was indistinguishable from wild type (Fig. 1A). Furthermore, Perls' Prussian Blue staining on Pcm brain sections did not reveal evidence of localized iron accumulation (data not shown). This comports with normal brain iron content in aged cohorts of Pcm homozygotes and wild-type littermates (Mok et al., 2006). Thus, whereas both peripheral tissues and brain displayed severe iron deficiency developmentally, adult brain appeared protected from the systemic iron overload.
Previous studies in Pcm mice detected tissue- and stage-specific dysregulation of Fpn1 expression (Mok et al., 2004a; Mok et al., 2004b; Mok et al., 2006). In order to assess the effects of the Pcm microdeletion on Fpn1 expression in the developing brain, Western blot analysis was performed on lysates from wild type and Pcm mutants between E14.5 and 12 weeks of age. Pcm homozygous brain manifested a decrease in Fpn1 levels between E14.5 and 3 weeks (Fig. 1B-D). By 12 weeks of age, Fpn1 expression recovered to wild-type levels (Fig. 1E). Quantitative realtime PCR revealed a significant decrease in Fpn1 mRNA expression in Pcm homozygous brain compared to wild-type at P0. In contrast, at 12 weeks of age, wild-type and Pcm homozygous brains had similar Fpn1 mRNA expression levels (Fig. 2). Thus, similar to placenta and fetal liver (Mok et al., 2004b), concordance between Fpn1 transcript and protein levels implicated transcriptional downregulation of Fpn1 expression as the primary molecular defect caused by the Pcm microdeletion in the developing brain.
By immunohistochemistry, P0 and 3-week old Pcm brain exhibited significantly decreased Fpn1 expression in smooth muscle cells of meningeal arterioles, as well as in endothelial cells of interstitial capillaries (Fig. 3A,A′,D,D′). There was no detectable difference in Fpn1 expression between wild-type and Pcm vessels at 12 weeks of age (Fig. 3G,G′). Fpn1 expression in other cell types, including neurons, glia, choroid plexus and ependymal cells, appeared unaffected by the Pcm mutation (Fig. 3A,A′,D,D′G,G′ and data not shown). Since Fpn1 is considered a likely candidate for iron transport across the abluminal membrane of endothelial cells at the blood-brain barrier (Ponka, 2004; Wu et al., 2004), decreased Fpn1-mediated iron transport into the brain could compound the embryonic systemic iron deficiency, resulting in markedly decreased brain iron levels at P0 (Fig. 1A,A′).
Normal iron balance in adult Pcm brain (Fig. 1A) implicated compensatory mechanism(s) in the recovery from developmental iron deficiency. Furthermore, adult Pcm brain appeared protected from the peripheral iron overload (Fig. 1A), which resulted from increased duodenal uptake during postnatal development (Mok et al., 2004a; Mok et al., 2006). Towards elucidation of the underlying regulatory mechanisms, Western blot analysis identified stage-specific changes in the expression of iron transporters. For example, E14.5 and P0 brain from Pcm mutants displayed elevated expression of TfR1 and divalent metal transporter 1 (Dmt1) (Fig. 1B,C). While Dmt1 expression recovered to wild-type levels in 3-week old Pcm homozygotes, TfR1 levels remained elevated (Fig. 1D). At 12 weeks of age, TfR1 expression was indistinguishable between wild-type and Pcm brain (Fig. 1E). Interestingly, Dmt1 levels were significantly lower at this stage, suggesting a potential role for Dmt1 in the protection from brain iron overload.
Quantitative real-time PCR demonstrated a statistically significant increase in TfR1 mRNA levels in Pcm homozygous brain at P0 (Fig. 2). This is consistent with increased stabilization of TfR1 mRNA due to binding of iron regulatory proteins in response to cellular iron deficiency (Hentze et al., 2004). TfR1 mRNA levels remained elevated at 12 weeks of age (Fig. 2). The significance of this finding in the context of normal brain iron content is unknown. Dmt1 mRNA levels were indistinguishable between Pcm homozygous and wild-type brain at both P0 and 12 weeks of age (Fig. 2), suggesting a predominantly post-transcriptional mechanism of Dmt1 regulation in response to brain iron content. Likewise, transcript levels of other iron metabolism genes, including Trf and Hepcidin (Hamp), were normal in P0 and 12-week-old Pcm homozygotes (Fig. 2).
Fpn1, TfR1 and Dmt1 protein expression was dynamically regulated in Pcm brain vessels. For example, immunohistochemistry showed marked upregulation of TfR1 expression in Pcm brain vessels during early postnatal development, which gradually decreased to wild-type levels by 12 weeks of age (Fig. 3B,B′E,E′,H,H′). While Pcm vessels exhibited discrete upregulation of Dmt1 at P0, expression levels at 3 weeks of age were comparable to wild-type littermates (Fig. 3C,C′,F,F′). Consistent with Western blot analysis, a subtle but reproducible decrease in Dmt1 expression was detected in Pcm vessels compared with wild type at 12 weeks of age (Fig. 3I,I′). Beyond differences in temporal regulation, compensatory changes in TfR1 and Dmt1 expression encompassed distinct subsets of brain vessels. Whereas upregulation of TfR1 in P0 and 3-week old Pcm brain was confined to endothelial cells of small interstitial vessels (Fig. 3B,B′,E,E′), Dmt1 expression was dynamically regulated in endothelial and smooth muscle cells of larger meningeal vessels (Fig. 3C,C′,E,E′). There was no evidence for significant changes in expression of both iron transporters in other cell types (Fig. 3, and data not shown). These findings conform to the notion that the vasculature represents a pivotal site for dynamic regulation of iron transporter expression in response to brain iron content (Rouault, 2001). In addition, they strongly suggest that compensatory mechanisms are operational at early stages of brain development, potentially ameliorating the developmental iron deficiency in Pcm mutants.
Deficiency in the ferroxidases Cp and Heph, which along with Fpn1 function to export iron from cells, leads to an age-related retinal degeneration (Hadziahmetovic et al., 2008; Hahn et al., 2004). Therefore, we investigated whether dysregulation of Fpn1 in Pcm homozygotes may affect retinal morphology and expression of pivotal iron metabolism genes. Compared to wild-type littermates, the retinas from P0 Pcm mice displayed stronger Fpn1 immunolabeling throughout the inner neuroblastic layer (INBL) and the rudimentary photoreceptor inner segments (Fig. 4A). A slight increase was also observed in the retinal pigment epithelium (RPE). Increased TfR1 immunolabel was found in the outer neuroblastic layer (ONBL) and the rudimentary photoreceptor inner segments of retinas from P0 Pcm mice when compared to wild-type controls. L-ferritin and H-ferritin expression were comparable between wildtype and Pcm retinas at P0 (data not shown).
Dysregulation of Fpn1 in P0 retina suggested the possibility of abnormalities in retinal development. However, at 7 weeks of age, the number of photoreceptor nuclei (within the outer nuclear layer; ONL) was indistinguishable between wild-type and Pcm homozygotes (8.59+/-0.10 nuclei per row vs. 9.01+/-0.61 respectively, p=0.51; Fig. 4B, Table 1). Additionally, the inner nuclear layer (INL), inner plexiform layer (IPL), and ganglion cell layer (GCL) appeared morphologically normal. Quantification of iron levels from 7 week wild-type and Pcm homozygotes retinas (0.058+/-0.004 μg/retina vs. 0.069+/-0.001 μg/retina, p=0.1402) and RPE/choroids (0.0065+/-0.008 μg/RPE/choroid vs. 0.060+/-0.001 μg/RPE/choroid, p=0.6688) no difference. Furthermore, at 7 weeks of age, Western blot analysis and immunofluorescence revealed no significant difference in Fpn1, TfR1, or L-ferritin levels between wildtype and Pcm homozygous retinas (data not shown). Strikingly, upon aging, the ONL became progressively thinner and at 12 weeks of age counts were trending close to statistical significance (Table 1). At 1 year of age, the number of photoreceptor nuclei per row was significantly lower in Pcm retina (2.21+/-0.20) compared with wildtype (5.01+/-0.30, p=0.02; Fig. 5B, Table 1). Furthermore, photoreceptor inner and outer segments were absent from most areas of Pcm retina by 1 year. Thus, Pcm homozygotes demonstrated an age-dependent, progressive retinal degeneration.
Previous studies have demonstrated that the Pcm mutation manifests different cellular and molecular alterations based on the age of the animal and the tissue studied (Mok et al., 2004a; Mok et al., 2004b; Mok et al., 2006). In the context of decreased placental Fpn1 protein levels, Pcm homozygotes display systemic iron deficiency and anemia at birth. During early postnatal development elevated duodenal Fpn1 expression governs increased organismal iron uptake. Concomitant low hepcidin levels, hence, an ineffective Fpn1-hepcidin homeostatic loop (Nemeth et al., 2004), result in marked organismal iron uptake and abnormally high iron levels in peripheral tissues, including liver, kidney, duodenum, and heart, reminiscent of patients with ferroportin disease (also known as hereditary hemochromatosis, type IV) (Mok et al., 2004a; Mok et al., 2004b; Mok et al., 2006).
These findings raise the question as to how two components of the CNS, the brain and the retina, respond to the dynamic changes in peripheral iron balance in Pcm mutants. Although parallels between iron homeostasis in the CNS and peripheral tissues have been noted, the blood-brain barrier (BBB) and the blood-cerebrospinal barrier effectively compartmentalize the CNS and necessitate controlled transport of molecules and ions, including iron (Ballabh et al., 2004; Bradbury, 1997). Fpn1 is considered a likely candidate for iron transport across the abluminal membrane of endothelial cells at the BBB (Ponka, 2004).
Pcm homozygotes demonstrated decreased levels of Fpn1 expression in smooth muscle cells of meningeal arterioles, as well as in endothelial cells of interstitial capillaries at P0 and 3 weeks of age (Fig. 3). Reduced levels of Fpn1 expression in cerebral blood vessels could compound the developmental brain iron deficiency by decreasing iron flux across the BBB resulting in markedly decreased brain iron levels at P0 (Fig. 1A). Pcm brain iron balance recovered to wild-type levels by 7 weeks of age, presumably due to compensatory increases in iron transport across the BBB. Indeed, Pcm brain at E14.5 and P0 displayed elevated expression of TfR1 and Dmt1 (Figs. 1B,C, ,2,2, and and3).3). However, whereas Dmt1 expression returned to wild-type levels at 3 weeks of age, TfR1 expression remained elevated in Pcm homozygous brain, most notably in capillary endothelial cells (Fig. 3). These results are consistent with a pivotal role for TfR1 and, to a lesser extent, Dmt1 compensatory brain iron uptake (Moos and Morgan, 2000; Rouault, 2001; Taylor et al., 1991).
Importantly, brain iron content at 7 and 12 weeks of age was indistinguishable between wild-type and Pcm mutant mice. This indicates the presence of effective mechanisms to protect the Pcm brain from sustained peripheral iron overload. Intact compartmentalization of the CNS iron cycle was reminiscent of normal brain iron balance in humans with secondary iron overload (Ponka, 2004). Iron accumulation in Pcm brain after 12 weeks of age appears less likely since excess peripheral iron is sequestered in RES macrophages and, hence, appears non-bioavailable, due to Hamp-mediated downregulation of Fpn1 (Mok et al., 2006). Indeed, brain iron content in aged Pcm heterozygous and homozygous animals was indistinguishable from wild-type controls (Mok et al., 2006) and there was no overt evidence of neurodegeneration (data not shown).
Due to the importance of iron homeostasis in mammalian development, perinatal iron deficiency has been extensively studied in rodents. These models vary in their methods to induce fetal iron deficiency and the length of post-natal dietary iron deficiency (Lozoff and Georgieff, 2006). In gestational/lactation models, pregnant dams are fed an iron-deficient diet starting around day 5 of gestation and continue an iron-deficient diet during lactation until weaning, at which point pups are fed a normal iron diet. This period of iron deficiency in rats is shorter than iron deficiency anemia observed in Pcm homozygotes. However, the correction to normal non-heme brain iron levels observed in 12-week-old Pcm homozygotes does not occur in iron-deficient rats (Felt and Lozoff, 1996). Thus, the Pcm mouse represents a powerful genetic model to study the regulatory mechanisms of brain iron uptake in response to developmental iron deficiency.
Within the retina, disruption of iron homeostasis can lead to cellular degeneration (He et al., 2007). At P0, the retina is still developing, with ongoing mitosis and lamination. The INBL includes post-mitotic ganglion, amacrine, horizontal cells, bipolar and Muller cell precursors (Cepko et al., 1996). Compared with wild-type control, Pcm retina at P0 displayed increased Fpn1 expression within the INBL, suggesting increased iron export in this location. Consistent with this, there was increased immunolabel for TfR1 in the Pcm retina at P0. This increase in TfR1 may be compensating for retinal iron deficiency caused by increased Fpn1 expression. While, Pcm homozygotes depicted no changes in retinal morphology or cell number in the ONL at this age(Fig. 4B), it is possible that developmental iron deficiency plays a role in the subsequent photoreceptor degeneration in older mice.
Interestingly, Pcm retinas exhibited an age-related degeneration that is significantly more severe than in wild-type controls (Fig. 4B, Table 1). As observed in this study, previous reports demonstrated that the retinas of wild-type A/J mice degenerate slowly with age (Danciger et al., 2003; Danciger et al., 2007; Gresh et al., 2003). Intercrosses of albino and pigmented mouse strains identified quantitative trait loci (QTL) involved in age-related retinal degeneration in this inbred strain (Danciger et al., 2003; Danciger et al., 2007). The QTLs mapped to several chromosomes and included Antxr1, Ruvbl1, and Eefsec (Danciger et al., 2007). Fpn1, which is located on chromosome 1 was not detected in this QTL screen and it remains to be determined whether Fpn1 interacts genetically with these loci. However, based on the progressive retinal phenotype in Pcm homozygotes, Fpn1 appears to act as a genetic modifier of age-related retinal degeneration in this albino strain of mice.
In humans, retinopathy consisting of retinal hemorrhages and small nerve fiber layer infarcts (cotton-wool spots) are occasionally observed in cases of iron-deficiency anemia (Carraro et al., 2001; Holt and Gordon-Smith, 1969). However, thinning of the ONL has not been described. In mice, mutation or disruption of genes integral to iron homeostasis often leads to anemia (Bernstein, 1987; Donovan et al., 2005; Trenor et al., 2000). Whereas retinal phenotypes might have generally been underreported, a subset of mouse mutants displays defects in this important compartment of the CNS. For example, Cp/Heph-deficient mice are anemic (Z. L. Harris and J. L. Dunaief, unpublished data) and display an age-related retinal degeneration. It is important to note Cp/Heph-deficient mice that survive to one year exhibit retinal iron accumulation in the RPE and widespread ONL thinning and RPE hypertrophy (Dunaief et al., 2005; Hadziahmetovic et al., 2008; Hahn et al., 2004). This represents a very different retinal phenotype compared with aged Pcm homozygotes, which are devoid of RPE hypertrophy or RPE iron accumulation. Thus whereas lack a functional iron export system systemically leads to an RPE iron overload in aged Cp/Heph-deficient mice, systemic dysregulation of the iron export protein Fpn1 in Pcm mice leads to slow, progressive ONL degeneration. Also supporting the importance of Fpn1 is RPE iron homeostasis is the observation that conditional disruption of Fpn1 within the RPE leads to L-ferritin accumulation within Cre recombinase positive RPE and loss of the L-ferritin loaded RPE cells in aged mice (J. Iacovelli and J. L. Dunaief, unpublished data). Although changes in retinal iron levels could not be detected in Pcm retinas by L-ferritin IHC, Western analysis or tissue iron quantification, more subtle changes in retinal iron homeostasis may account for the degeneration seen in these retinas. The retinal phenotype bears striking similarities with spleen degeneration in Pcm embryos, wherein apoptotic cell death and spleen regression ensue in the context of elevated levels of Fpn1 expression (Mok et al., 2004b). Therefore, while the specific mechanism(s) of cellular degeneration await further investigation, this study demonstrates a novel function for Fpn1 in the regulation of long-term photoreceptor survival.
This study employed Pcm mice on a fully congenic N14 A/J background. Mice were fed a standard chow with an iron content of 240 ppm and genotyped as described previously (Mok et al., 2004a). Brains from wild-type and Pcm mutants were dissected at embryonic day 14.5 (E14.5), postnatal day 0 (P0), 3 and 12 weeks of age.
Non-heme iron in brain, retina, and retinal pigment epithelium/choroids was quantified based on the methodology of Torrance and Bothwell (1980). Samples were processed as described previously (Mok et al., 2004a). To determine the abundance and localization of iron in Pcm brain, Prussian blue staining of brain sections was accomplished using the Accustain iron staining kit (Sigma, St. Louis, MO), according to the manufacturer's protocol.
As described recently (Kim et al., 2007), dissected brains were fixed in Bouin's reagent, dehydrated in a graded series of ethanol, embedded in paraffin and sectioned midsagittally at 5-μm thickness. To allow for comparison of gross morphology and protein expression levels, sections from wild-type and Pcm homozygous brains were mounted side by side on the same slide. Sections were stained with hematoxylin and eosin (H&E) according to standard protocols and immunohistochemistry was performed as described previously (Mok et al., 2004b). For Fpn1, Dmt1 and TfR1 antigen retrieval, sections were boiled in 0.01M citric acid, pH 6.0. Blocking was achieved with goat (Fpn1 and Dmt1) or horse serum (TfR1). Primary antibody incubations were performed overnight at 4°C using rabbit αFpn1 at 1:200 (Mok et al., 2006), rabbit αDmt1 at 1:100 (Alpha Diagnostic International) and mouse αTfR1 at 1:1000 dilution (Invitrogen). Secondary antibodies were peroxidase-conjugated with the Vectastain Elite ABC Kit (Vector Laboratories, Burlingame, CA) and antigen expression was detected with Vector NovaRED substrate (Vector Laboratories). Images were acquired with a Sony 085 CCD color RGB sensor digital camera mounted on a Zeiss Axioplan 2 microscope and processed with Adobe Photoshop software.
For immunofluorescence on P0 retina, pup heads were fixed in 4% parafomaldehyde in PBS and then cryoprotected in 30% Sucrose in PBS. Pup heads were embedded in Tissue-Tek OCT (Sakura Finetek, Torrance, CA). Immunofluorescence for Fpn1 and TfR1 was performed on 10 μm cryosections. For TfR1, antigen retrieval was performed as describe above prior to blocking. Sections were blocked as previously described (Dunaief et al., 2002) and incubated in rabbit αFpn1 (Mok et al., 2006) at a 1:100 dilution or rabbit αTfR1 at a 1:80 dilution. The secondary antibody used was donkey anti-rabbit-CY3 (Jackson ImmunoResearch Laboratories, West Grove, PA). Ommision of the primary antibody was used as a negative control. Sections were counterstained with Vectrashield containing DAPI (Vector Laboratories) and images were obtained using a Nikon epifluorescence microscope and Image Pro 6.0. Samples were processed in parallel and equivalent exposure times were used for all samples.
Eyes were enucleated post mortem and fixed overnight in 2% gluteraldehyde/2% paraformaldehyde in PBS. Eyecups were prepared by removal of the anterior segment and embedded in JB-4 according to the manufacture's protocol (Polysciences Inc., Warrington, PA). Semi-thin 3 μm section were cut and stained with Toluidine Blue O (Sigma). Images of sections containing optic nerve and separated by 30 μm were obtained for quantitative analysis using Nikon TE-300 epifluorescence microscope with Image Pro Plus 6.1 software (Media Cybernetics, Bethesda, MD). The number of photoreceptor nuclei per row present in the outer nuclear layer (ONL) was counted in three adjacent rows at a given distance from the optic nerve head. The average nuclei in the three adjacent rows was calculated from 5 sections from 3 eyes for each genotype at 7 wks and 1 yr and 2 eyes for each genotype at 12 wks. A P-value was generated using the generalized estimating equations with correlations from repeated measures applied. Differences of P<0.05 were considered significant.
Brains were dissected from E14.5 embryos, P0 pups, as well as 3- and 12-week-old animals and were processed for Western blot analysis as described (Mok et al., 2004a). Briefly, following homogenization, brains were lysed in RIPA buffer plus Complete®, EDTA-free protease inhibitor (Roche), extract supernatant was collected and protein was quantified using the Bio-Rad DC Protein Assay kit (Bio-Rad). Fifty μg of total protein were mixed with equal volume sample buffer with β-mercaptoethanol. For detection of Dmt1, TfR1 and actin expression, samples were boiled; non-heated samples were employed for Fpn1 analysis. Protein lysates were separated on 8% SDS-PAGE and transferred to PVDF membrane. Blocking was achieved by incubation in TRIS-buffered saline containing 5% milk and 0.1% Tween 20. Membranes were incubated overnight at 4°C with the following primary antibodies: rabbit α-Fpn1 at 1:1000 dilution (Mok et al., 2004a), rabbit α-Dmt1 at 1:2000 (Alpha Diagnostic International), mouse α-TfR1 at 1:2000 (Invitrogen, Carlsbad, CA) and goat α-actin at 1:5000 (Santa Cruz Biotechnology, Santa Cruz, CA). Following washes, membranes were incubated with horseradish peroxidase-conjugated secondary antibodies (1:10000 dilution) and signals were developed using ECL reagent (Santa Cruz Biotechnology).
Total RNA was extracted from P0 or 12-week-old wild-type and Pcm brain using Trizol reagent (Invitrogen). First-strand synthesis from 50-600 ng of total RNA employed Superscript III RTase (Invitrogen) with random hexamers. PCR was performed using the 2× TaqMan® Universal Master Mix, no UNG (Applied Biosystems) and the 7500 Fast System (Applied Biosystems) according to the manufacturer's protocol. The following Taqman MGB probes from Applied Biosystems were employed: Fpn1 (exon junction 3-4, Mm00489837_m1), TfR1 (exon junction 2-3, Mm00441941_m1), Trf (exon junction 1-2, Mm00446708_m1), Dmt1 (exon junction 15-16, Mm00435363_m1), Hamp (exon junction 1-2, Mm00519025_m1) and 18S ribosomal RNA (4319413E) as an endogenous control for each sample. The relative levels of mRNA expression were expressed as a ratio of the gene-specific probe to 18S rRNA and normalized to a single wild-type ratio set to 1.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.