Abstract

Human copper enzymes reflect, by nature of their function and tissue specificity, an essential family of proteins. These enzymes are 1) cytochrome c oxidase, which is required for mitochondrial respiration; 2) Cu, Zn-superoxide dismutase (SOD), a ubiquitous antioxidant; 3) dopamine β hydroxylase, which is essential for catecholamine production; 4) lysyl oxidase, which is integral for collagen cross-linking and elastin formation; 5) peptidyl glycine α-amidating monooxygenase, which is responsible for neuropeptide and peptide hormone processing; 6) tyrosinase, which is essential for melanin synthesis, and 7) ceruloplasmin and hephaestin, members of the multicopper oxidase family, which are critical for efficient iron efflux through their action as ferroxidases. Recent identification of the inherited disorder aceruloplasminemia—a triad of diabetes, retinal degeneration, and neurodegeneration with an absence of detectable serum ceruloplasmin—associated with an altered iron metabolism has revealed an important link between copper and iron homeostasis.1,2
A review of hepatic copper homeostasis reveals that ingested copper, either bound to albumin or histidine, enters the hepatocyte via the copper transporter (CTR1), where its fate is determined by a family of copper chaperones or is stored in a metallothionein pool. Copper may be trafficked via cyclooxygenase-17 to the mitochondria for incorporation into cytochrome oxidase. Alternatively, it may be shuttled via CCS (the copper chaperone for SOD) for copper delivery to Cu, Zn-SOD. Finally, copper may be transported by HAH1, the copper chaperone for the Wilson disease P-type adenosine triphosphatase (ATPase), to the trans-Golgi network. It is here that the Wilson disease P-type ATPase is responsible for transporting copper into the secretory pathway, where it can either be incorporated into the cuproprotein ceruloplasmin or trafficked into a vesicular compartment that is excreted into bile. 3
A mutation in the Wilson disease P-type ATPase gene prevents copper from being appropriately deposited within the secretory compartment and, hence, prevents incorporation into ceruloplasmin.4–6 Apoceruloplasmin (ceruloplasmin lacking copper) is degraded rapidly, with a half-life of 4 hours, compared with holoceruloplasmin (ceruloplasmin containing copper), with a half-life of 4 days. 7 Ceruloplasmin biosynthesis occurs independently of copper availability/incorporation: if copper is present, holoceruloplasmin is synthesized and secreted; if copper is absent, apoceruloplasmin is produced. Any disorder/mutation that would interfere with copper being successfully trafficked into the secretory compartment would result in a low level of circulating serum ceruloplasmin. As such, nutritional copper deficiency, synthetic hepatic dysfunction, and Wilson disease all manifest clinically with low circulating serum ceruloplasmin. Serum ceruloplasmin levels have been used as a sensitive marker of copper status and are clinically important in diagnosing Wilson disease.
Wilson disease is an autosomal recessive disorder that results in excess copper accumulation in the liver. Copper trafficked into hepatocytes cannot be successfully compartmentalized into the secretory pathway for either cuproprotein biosynthesis or biliary excretion and, as such, accumulates in the liver. Patients present with liver disease, Kayser-Fleischer rings, basal gangliar symptoms (tremor and dystonia), and psychiatric disorders.
Over the past decade and a half, case reports of patients with “atypical Wilson disease” have been identified. Frequently labeled as hypoceruloplasminemia, these adults present with basal gangliar neurodegeneration, retinal degeneration, diabetes mellitus, and near-absent circulating serum ceruloplasmin.8–10 Laboratory analyses further reveal a mild microcytic anemia (hemoglobin, 10-11 g/dL) and elevated ferritin (>900 μg/L). Liver biopsy reveals a lack of fibrosis or cirrhosis and abundant iron deposition throughout the parenchyma. Unlike the patient with Wilson disease, there is no abnormal copper accumulation and hepatic architecture is normal, which suggests a clinical disorder similar to but distinct from Wilson disease.
This new clinical disorder is termed “aceruloplasminemia” and represents a disruption of iron homeostasis as a result of mutations in the ceruloplasmin gene with subsequent loss of ceruloplasmin's ferroxidase activity. A link between iron and copper metabolism had been established elsewhere. Earlier work by Roser et al 11 and Osaki et al 12 revealed that copper-deficient pigs developed a profound iron deficiency anemia that could be corrected only when copper was administered and was unaffected by iron therapy. More recently, yeast biologists have identified a multicopper oxidase Fet3—a ceruloplasmin homologue—as a protein essential for iron trafficking into budding yeast.13–16 Mutations of the copper binding sites of Fet3 abrogate normal iron homeostasis in Saccharomyces cerevisiae.
Ceruloplasmin is a 132-kDa β2 glycoprotein that is synthesized and secreted predominantly by hepatocytes as a holoprotein. This member of the multicopper oxidase family contains more than 95% of the copper found in serum. The ceruloplasmin gene contains 19 exons. In addition to the plasma form of ceruloplasmin, an alternative splice product yields a glycosyl phosphatidylinositol membrane anchored form that is present in specific tissues.16–18 Multiple mutations within the ceruloplasmin gene have been identified, and more than 40 families have been studied. An estimated frequency of this autosomal recessive disorder is 1:10,000.
Patients with aceruloplasminemia present in adulthood with diabetes mellitus and, within a decade of this diagnosis, manifest the clinical symptoms of neurodegeneration and retinal degeneration. The serum chemistry and hematologic profiles reveal near-absent serum ceruloplasmin, low serum copper, and elevated serum ferritin. Patients with aceruloplasminemia are mildly anemic, with a low-to-normal transferrin saturation, and their anemia worsens with phlebotomy if it is treated inappropriately as a hereditary hemochromatosis. T2-weighted magnetic resonance imaging reveals iron deposition in the basal ganglia in aceruloplaminemia. Autopsy tissues of brain from an aceruloplasminemic patient show that, despite normal cortical architecture, both cavitary degeneration of the basal ganglia and substantia nigra and pigmentary discoloration of the basal ganglia are present. Ophthalmologic exams reveal peripheral retinal degeneration associated with iron deposition. Liver biopsy is characterized by normal histology without any evidence of cirrhosis or fibrosis and abundant iron deposition when stained with Perls stain.8,10
Unlike other disorders of iron accumulation, aceruloplasminemia results in both parenchymal and central nervous system iron deposition. Ceruloplasmin is synthesized and secreted behind the blood-brain barrier, and in situ hybridization has revealed astrocyte-specific ceruloplasmin mRNA expression specifically within those astrocytes associated with the microvasculature and surrounding melanized dopaminergic neurons. 19 Oligodendrocytes have been shown elsewhere to synthesize and secrete transferrin. We propose that iron, bound to transferrin, crosses the blood-brain barrier via transferrin receptor-mediated endocytosis, on release is oxidized by astrocyte-synthesized ceruloplasmin, and is trafficked throughout the central nervous system bound to oligodendrocyte-produced transferrin. A lack of ceruloplasmin within the central nervous system would result in neurodegeneration by three possible mechanisms 1) Fe2+-mediated oxyradical damage, 2) microglial iron overload secondary to an increase in non-transferrin-bound iron, and 3) neuronal iron starvation secondary to a lack of substrate. 19
To further evaluate the role of ceruloplasmin in iron homeostasis, a ceruloplasmin-null mouse was generated by use of homologous targeting recombinant technology. 20 Western blot analysis revealed absent serum ceruloplasmin protein. Furthermore, Northern blot analysis revealed absent ceruloplasmin mRNA from the brain and liver. Elevated serum ferritin and increased parenchymal iron in the spleen and liver were apparent by age 1 year. Ferrokinetic studies revealed normal iron absorption and distribution. As with the human patients with this disease, the liver architecture appeared normal, but Perls staining for iron revealed abundant iron accumulation in both hepatocytes and Kupffer cells. The spleen histology and iron staining were remarkable for significant macrophage iron accumulation.
Where was this increased iron accumulation coming from? Iron homeostasis is regulated at the level of absorption across the duodenal enterocyte, but this only represents 5% of the total iron cycle. The majority of iron in the body is recycled from senescent red blood cells by macrophages and stored within the reticuloendothelial system. Thus, reticuloendothelial iron trafficking represents 95% of the iron cycle. Given that the ferrokinetic studies have revealed normal iron absorption and distribution in the ceruloplasmin null mouse, whereas histological review of tissues have revealed significant macrophage iron accumulation, we hypothesized that ceruloplasmin must play a role in reticuloendothelial iron homeostasis. A lack of ceruloplasmin would be predicted to result in inefficient iron efflux and manifest as a predominantly reticuloendothelial iron overload syndrome. To make ceruloplasmin rate-limiting, we challenged the iron “recycling” machinery in our mice with two paradigms: 1) injecting ceruloplasmin-null mice with damaged red blood cells, hence increasing the reticuloendothelial red cell burden; and 2) making ceruloplasmin-null mice anemic, thus calling on the reticuloendothelial cells to release storage iron for new red cell synthesis. Both paradigms would test the ability of a ceruloplasmin-null mouse to regulate iron efflux via recycling an increased iron load or responding to an increased iron demand.
The wild-type ceruloplasmin mice, Cp+/+, were able to easily bear the burden of an increased damaged red cell load and, within 3 hours after injection, increased their serum iron, which represents transferrin-bound iron being trafficked for storage. The ceruloplasmin-null mice, Cp-/-, released no increased serum iron in response to the damaged red cell burden. However, if Cp-/- mice were injected with purified holoceruloplasmin, a brisk and rapid release of iron was identified. Similarly, when wild-type Cp+/+ mice were made anemic by serial phlebotomy, they were able to respond to the anemia (hemoglobin 8 g/dL) by increasing the amount of iron trafficked from reticuloendothelial iron stores to their bone marrow for new red cell synthesis. The Cp-/- mouse was incapable of effluxing iron unless presented with exogenous purified holoceruloplasmin.
Thus, ceruloplasmin appears to be essential in regulating the efficiency of iron efflux from reticuloendothelial cells. Consistent with its role as a ferroxidase, ceruloplasmin is critical for determining the rate of Fe3+ iron available for binding to transferrin. The exact mechanism involved in this process has yet to be determined. Furthermore, how an absence of ceruloplasmin leads to neurodegeneration is currently being investigated. The aceruloplasminemic human and mouse teach us that this copper-containing protein, ceruloplasmin, is essential in the regulation of iron homeostasis and that aceruloplasminemia represents a unique disorder of iron metabolism with systemic and central nervous system iron-associated pathology.
Footnotes
Acknowledgments
These studies were supported by Grants DK02464 and DK58086 from the National Institutes of Health. This review was presented at the AFMR symposium “New Insights into GI and Liver Diseases Based on Molecular Aspects of Transport Physiology,” at FASEB on April 3, 2001.
