Iron genes, iron load and risk of Alzheimer's disease.

BACKGROUND: Compound heterozygotes of the haemochromatosis gene (HFE) variants, H63D and C282Y, have raised transferrin saturation compared with that in the wild type. In the cohort of the Oxford Project To Investigate Memory and Ageing (OPTIMA), bicarriers of the HFE C282Y and the transferrin C2 gene variants are at five times greater risk of developing Alzheimer's disease; the addition of HFE H63D may raise the risk still further. OBJECTIVE: To investigate transferrin saturation by HFE and transferrin genotype among people without dementia-that is, controls and those with mild cognitive impairment (MCI)-and also among those with Alzheimer's disease. METHODS: Serum iron status and genotype were examined of 177 patients with Alzheimer's disease, 69 patients with MCI and 197 controls from the OPTIMA cohort. RESULTS: Although each of these variants alone had relatively little effect on iron status, the combination of either HFE C282Y and HFE H63D or of HFE C282Y and transferrin C2 markedly raised transferrin saturation in those without dementia, but had little effect in those with mature Alzheimer's disease. CONCLUSIONS: These combinations may raise the risk for Alzheimer's disease, owing to higher iron loads and therefore oxidative stress in the preclinical phase. If replicated, these findings will have implications for the prevention of Alzheimer's disease.

Targeted screening for genetic haemochromatosis: a combined phenotype/genotype approach.

OBJECTIVE: To evaluate the clinical utility of a targeted screening approach for the detection of genetic haemochromatosis. METHODS: Screening by measuring fasting serum transferrin saturation (TS) and gene testing was carried out in patients in whom a raised serum alanine amino transferase (ALT) activity and raised random serum TS had been found on routine blood testing. RESULTS: During the 29 month study period, 32 patients homozygous for the C282Y genotype were detected from a catchment population of 330,000 by screening blood samples referred initially for routine laboratory liver function tests. By comparison, during the same period of time and within the same population, only seven patients were found by clinical suspicion alone. The patients in the study, after treatment by venesection, have shown both clinical and biochemical improvement. CONCLUSIONS: The study shows that from a population of patients in whom a routine liver function profile had been requested, it is possible to detect subjects homozygous for the C282Y HFE genotype who have clinical or biochemical markers of iron overload.

The effects of wild-type and mutant HFE expression upon cellular iron uptake in transfected human embryonic kidney cells.

In order to measure the effects of HFE (haemochromatosis) upon iron uptake, stable expression of wild-type and C282Y, H63D and S65C mutant HFE cDNA was established in HEK 293 cells. Control cells were transfected with empty vector. Expression of HFE mRNA and protein was detected in the cell lines transfected with HFE cDNA, but not in the control cell line. The ferritin concentration in wild-type cells cultured in 40 microM ferric ammonium citrate was 69% of that in control cells and 81% of that in C282Y cells. The ferritin concentration in H63D cells was intermediate between wild-type and C282Y and the ferritin concentration in S65C cells was similar to wild-type cells. Uptake of transferrin-iron in wild-type, C282Y and control cells was measured over 45 min. The Hill coefficients for transferrin-iron uptake were similar. The V(max) for transferrin-iron uptake in wild-type cells was 59.5% of control cells and 69.5% of C282Y cells. Estimates of K(m) were 232 nM for wild-type cells, 338 nM for C282Y cells and 570 nM for controls. Transferrin receptor levels were lowered, but not significantly, in the HFE transfected cells. The results show that HFE reduces transferrin-iron uptake, probably as an uncompetitive inhibitor.

The 16189 variant of mitochondrial DNA occurs more frequently in C282Y homozygotes with haemochromatosis than those without iron loading.

BACKGROUND: Patients with hereditary haemochromatosis (HH) are usually homozygous for the C282Y mutation in the HFE gene. They have variable expression of iron overload and present with a variety of complications, including liver disease, diabetes, arthropathy, fatigue, and cardiomyopathy. The mitochondrial 16189 variant is associated with diabetes, dilated cardiomyopathy, and low body fat at birth, and might contribute to genetic predisposition in further multifactorial disorders. The objective of this study was to determine the frequency of the 16189 variant in a range of patients with haemochromatosis, who had mutations in the HFE gene. METHODS: Blood DNA was analysed for the presence of the 16189 variant in British, French, and Australian C282Y homozygotes and controls, with known iron status, and in birth cohorts. RESULTS: The frequency of the mitochondrial 16189 variant was found to be elevated in individuals with haemochromatosis who were homozygous for the C282Y allele, compared with population controls and with C282Y homozygotes who were asymptomatic (42/292 (14.4%); 102/1186 (8.6%) (p = 0.003); and 2/64 (3.1%) (p = 0.023), respectively). CONCLUSIONS: Iron loading in C282Y homozygotes with HH was exacerbated by the presence of the mitochondrial 16189 variant.

Ferritin.

The iron storage protein ferritin is found in all cells of the body as multiple isoferritins composed of 24 sub units of two types. The structure is well understood from increasingly detailed analysis by X-ray crystallography. Genes for the principal subunits (called H and L) have been cloned and are located on chromosomes 11q and 19q respectively. The production of H24 and L24 recombinant molecules is making it possible to explore the relationship between structure and function. The control of ferritin synthesis by iron at the level of translation is providing a model for understanding the control of protein synthesis. H-rich isoferritins are of considerable interest in haematology as they appear to be implicated in control of haemopoiesis and development of malignancy. Whether or not an abnormality in ferritin is the cause of hereditary haemochromatosis is not yet resolved.

Co-inheritance of mutations in the uroporphyrinogen decarboxylase and hemochromatosis genes accelerates the onset of porphyria cutanea tarda.

Porphyria cutanea tarda is a skin disease caused by photosensitization by porphyrins whose accumulation is caused by deficiency of hepatic uroporphyrin- ogen decarboxylase activity. Mutations in the uroporphyrinogen decarboxylase gene are present in the low-penetrant, autosomal dominant familial form but not in the commoner sporadic form of porphyria cutanea tarda. We have investigated the relationship between age of onset of skin lesions and mutations (C282Y, H63D) in the hemochromatosis gene in familial (19 patients) and sporadic porphyria cutanea tarda (65 patients). Familial porphyria cutanea tarda was identified by mutational analysis of the uroporphyrinogen decarboxylase gene. Five previously described and eight novel mutations (A80S, R144P, L216Q, E218K, L282R, G303S, 402-403delGT, IVS2 + 2 delTAA) were identified. Homozygosity for the C282Y hemochromatosis mutation was associated with an earlier onset of skin lesions in both familial and sporadic porphyria cutanea tarda, the effect being more marked in familial porphyria cutanea tarda where anticipation was demonstrated in family studies. Analysis of the frequencies of hemochromatosis genotypes in each type of porphyria cutanea tarda indicated that C282Y homozygosity is an important susceptibility factor in both types but suggested that heterozygosity for this mutation has much less effect on the development of the disease.

A 6p22 reference map of leukocyte DNA: exclusion of rearrangement in four cases of atypical haemochromatosis.

We describe a 4 Mb reference map of the haemochromatosis gene region in leukocyte DNA from seven controls and four atypical haemochromatosis patients. Three patients had normal coding sequence for HFE, the candidate gene for genetic haemochromatosis (GH). The fourth patient had classical GH but was heterozygous for Cys282Tyr with otherwise normal coding sequence. The genomic DNA was mapped by pulsed-field gel electrophoresis (PFGE) using five rare-cutting enzymes. Seventeen probes including HFE were positioned on the map. Despite proximity to the highly polymorphic major histocompatibility complex (MHC), no polymorphism was observed in the control group with these telomeric probes. Furthermore, major rearrangement of the HFE region was excluded as a mutation contributing to iron overload in these atypical patients. Maps of cloned DNA are linked through genes and other probes to this reference map of the HFE region in uncloned genomic DNA.

A molecular analysis of the telomeric end of the major histocompatibility complex. DNA typing of HLA-A3 subtypes and -B7.

A molecular investigation of the HLA-A, -B, -C, I82, D6S128, and D6S265 loci was carried out using RFLP and PCR analyses in 331 healthy individuals, 21 HLA-A3B7 homozygous subjects, and 17 HLA homozygous cell lines. New polymorphic RFLP patterns were noted at the HLA-A and -B loci which were useful for distinguishing subtypes of HLA-A3 and -B44. Strictly specific bands were observed for HLA-A3, -A11, -B7, -B57. and -Cw4. Molecular identification of two HLA-A3 subtypes was possible by HLA-A RFLP and D6S265 PCR analyses. The two alleles of the I82 locus were associated with different HLA-A3 subtypes. It was shown that the serologically HLA-A3 homozygous cell line EHM was heterozygous for the two subtypes. The common subtype formed 91% of HLA-A3 antigens. Some serologically HLA-A3 homozygous healthy individuals were heterozygous for the subtypes. Using these new polymorphisms, a thorough molecular analysis of the haplotype HLA-A3B7DR15 at the three regions of the MHC was performed. The results have implications on HLA matching in transplantation, population genetics of the MHC, and disease association studies.

The Ha-ras polymorphism in myelodysplasia and acute myeloid leukaemia.

We have assessed the possibility that rare allelic variants of the c-Ha-ras-1 locus may be linked to a susceptibility to malignancy [1]. c-Ha-ras-1 genotypes were scored in 41 patients with myelodysplasia (MDS), 51 patients with acute myeloid leukaemia (AML) and 52 normal subjects. The incidence of rare alleles in the MDS patients was 4.8% and in AML an incidence of 15.7% was found. No rare alleles were found in the normal subjects. We conclude that rare alleles in MDS are not a common predisposing factor.

Biochemical and immunological characteristics of ferritin from HL-60, U-937 and K-562 cell lines: implications for haemopoietic regulation.

It has been suggested that acidic isoferritins play a major role in the regulation of human granulocyte-macrophage (CFU-GM) proliferation. Such regulatory isoferritins are said to occur in some human tissues including both leukaemic and normal leucocytes, and certain established cell lines. They are glycosylated and bind to Con A. Extracts from HL-60, U-937 and K-562 cell lines are said to contain inhibitory activity but only in HL60 cells is the activity neutralized by antibody to acidic isoferritins. The properties of ferritin from these cell lines have been investigated. HL60 and U-937 extracts contained predominantly basic (spleen-type) isoferritins and K-562 extracts predominantly acidic (heart-type) isoferritins on immunoradiometric assay. None of the ferritin bound to Con A. On anion exchange chromatography ferritin from U-937 was basic in character, that from K-562 was acidic and HL-60 contained a variety of isoferritins. The identification of glycosylated acidic isoferritins as inhibitors of granulopoiesis appears to be an oversimplification and the classification of isoferritins on the basis of two subunit types may also need revision.

An automated immunoradiometric assay for ferritin.

A semi-automated method for the immunoradiometric assay of ferritin has been developed. Results obtained with serum samples show excellent correlation between this method and the previously described assay. A further useful reduction in the assay time may be achieved by incubating for 3 hours at 37 degrees C instead of 24 hours at 4 degrees C.

Serum ferritin concentration and iron stores in normal subjects.

The relationship between serum ferritin concentration and the amount of storage iron has been studied in normal subjects. A high degree of correlation was demonstrated between serum ferritin concentration and storage iron measured by quantitative phlebotomy. The possible advantages of assessing iron stores by using the serum ferritin concentration are discussed.

Postmortem blood ferritin concentrations in sudden infant death syndrome.

AIMS--To confirm the observation of extremely high concentrations of ferritin in postmortem serum samples in sudden infant death syndrome (SIDS); to examine the factors influencing blood ferritin concentrations postmortem; to determine whether or not these high blood ferritin concentrations are characteristic of SIDS. METHODS--Postmortem samples of cardiac blood were obtained from 58 full term infants who died of SIDS and 14 full-term infants who died of a variety of other causes. Whole blood and serum ferritin concentrations were determined and compared with age at death, liver iron concentration, serum iron concentration, and serum lactate dehydrogenase activity. RESULTS--The median postmortem blood ferritin concentration for all infants was 18,600 micrograms/l, which is about 200 times the concentration found in the serum of normal, live infants. Serum iron concentrations were high and there was a highly significant correlation between serum ferritin and iron concentrations suggesting that much of the serum iron was contributed by ferritin. There was no significant difference between serum and whole blood ferritin concentrations. H to L type ferritin ratios were higher in blood from the left than the right ventricle of the heart but the ferritin was always predominantly L type. Blood ferritin concentrations rose rapidly after death but in samples collected at postmortem examination there was a significant correlation with liver iron concentration and an inverse correlation with age. Median values for blood ferritin were higher in SIDS (22,500; n = 58) than in control cases (6900; n = 7) dying under one year of age; however, in both groups ferritin concentrations decreased with age. CONCLUSIONS--Release of ferritin into the blood postmortem seems to be characteristic of infants dying before the age of one year rather than characteristic of SIDS. Two factors may cause such ferritin release postmortem: tissue breakdown and the high level of storage iron in cells of the reticuloendothelial system (including endothelial cells lining vessel walls). SIDS occurs when tissue iron concentrations are higher than at any other time of life. It is possible that the ready availability of iron enhances free radical damage which might be implicated in SIDS.

Serum ferritin in children with thalassaemia regularly transfused.

A controlled trial of continuous chelation therapy in regularly transfused children with homozygous beta-thalassaemia has been in progress at the Hospital for Sick Children since April 1966. In the sixth and seventh years of the trial the effect of this treatment on iron overload has been assessed by estimating serum ferritin levels and liver iron concentrations in both chelator-treated and control groups. When compared with non-chelated controls, results of both these estimations were invariably lower in the chelated group. However, all the results in both groups were very high, and fell within the ranges observed in untreated idiopathic haemochromatosis. A close correlation was found between serum ferritin levels and liver iron concentrations in these children, indicating that serum ferritin is a valuable alternative to liver iron concentration in the assessment of visceral iron overload, even when massive tissue siderosis is present.

Blood ferritin concentrations in newborn infants and the sudden infant death syndrome.

Liver iron concentrations have been shown to be higher in victims of SIDS than in postmortem controls suggesting that high levels of tissue iron may be implicated in SIDS. To determine whether infants who subsequently die from SIDS are born with greater iron stores than those who do not, the iron stores in newborn infants were assessed retrospectively by measuring blood ferritin concentration in spots from Guthrie cards (collected from almost all infants born in the UK in the first week of life). A method for extracting and measuring ferritin from stored blood spots is described. Eighteen cases of SIDS were identified in South Glamorgan along with four controls for each case. Ferritin concentrations did not differ in SIDS victims and controls suggesting that victims of SIDS are not born with abnormal concentrations of stored iron. If iron stores are found to be higher in SIDS victims than in healthy live infants of the same age then it is more likely that the iron will have been acquired after birth.

Human major histocompatibility complex contains several leukemia susceptibility genes.

In mice, homozygosity for the Mhc haplotype H-2k is associated with increased susceptibility to spontaneous and virus-induced leukemia, lymphoma and other neoplasms in the predisposed host. The influence of the Mhc on malignant development in these models is to shorten the latency after virus inoculation. Here, we present evidence that a similar phenomenon results in early-onset of human leukemia. A molecular analysis of the MHC in 112 CML patients showed that those who developed the disease when aged less than 35 years (early-onset group) had higher homozygosity rates for the DOA1, HSP70 and C4 alleles of the DR53 group of ancestral haplotypes, for a subtype of HLA-A3, and a higher allele frequency of BfFb compared to the late-onset group. The oldest patient (n = 13) homozygous for DR53 was 52-years-old (p = 0.004), and all HLA-A3 homozygous patients (n = 4) were in the early-onset group (p = 0.01). The relative risk for early-onset CML yielded by HLA-A3 homozygosity was 17.6. The well-known serological HLA-Cw4 association was not confirmed at the DNA level and thought to be due to linkage disequilibrium with BfFb. The factor B association was sex-limited. The DR52 group haplotypes appeared to be protective. The HLA-identical sibling frequency was increased only in the early-onset group (p < 0.01). Our findings agree with the concept of an MHC influence on the development of malignancies. The similarity in the location of the susceptibility loci and the serological cross-reaction between H-2Ek and DR53 raise the possibility that the mouse and human MHC share the same leukemia susceptibility genes.

Biochemical aids to the staging of breast cancer.

Up to fifteen plasma proteins were measured before treatment in 249 women presenting with lumps in the breast. Concentrations showed considerable overlap between the various clinical stages, and were often normal even in metastatic disease. A discriminant function is proposed, based on measurement of C-reactive protein, beta 2-microglobulin, carcinoembryonic antigen and ferritin and calculation of a score for each subject. High-risk scores resulted for all 18 patients with Stage 4 (i.e., metastatic) disease, and the number of Stage 1 patients attaining high scores was consistent with the reported incidence of development of metastases in such a group. Follow-up studies are in progress.

The laboratory assessment of iron status--an update.

The description by Ramsay in 1957 of a practical way of determining the total iron binding capacity of serum (a measure of transferrin concentration) provided a diagnostic test for both iron deficiency and iron overload. Since 1957 the introduction of the assay for serum ferritin (in 1972) has made it possible to assess the levels of storage iron in normal subjects and assays for free erythrocyte protoporphyrin and the circulating transferrin receptor methods to evaluate iron supply for erythropoiesis. In 1957 iron metabolism in man was already well understood but its evaluation relied on measurement of tissue iron concentrations and the use of radioisotopes of iron to measure rates of erythropoiesis. The evaluation can now be carried out using the various blood assays along with the measurement of haemoglobin concentration but interpretation of the measurements in disease still requires an understanding of the way in which these measures are influenced by pathological processes.

An immunoradiometric assay for the acidic ferritin of human heart: application to human tissues, cells and serum.

Human tissues contain ferritin molecules with a range of isoelectric points but immunoassays for detecting serum ferritin have generally employed antibodies to the more basic liver or spleen proteins. To study the distribution of more acidic ferritins in tissues and serum acidic ferritin has been isolated from normal human heart and a two-site immunoradiometric assay for this protein developed. This assay gives little cross-reaction with spleen ferritin. Tissue ferritins have been fractionated by anion exchange chromatography and assayed with both spleen and heart antibodies. The spleen ferritin assay detects the more basic ferritin and the heart ferritin assay the more acidic ferritin. Acidic ferritins were found in heart, kidney, reticulocytes and HeLa cells. In sera from normal subjects and patients with iron overload, myocardial infarction, leukaemia and carcinoma only low concentrations of heart ferritin were found, although in the pathological sera spleen ferritin concentrations were generally raised. Circulating ferritin contains only a small proportion of molecules with the immunological characteristics of acidic heart ferritin.