seeking candidate mutations that affect iron homeostasis

ct: veryhow

nes haveariation:

-1,,sequenced

.definedalitiestionship

of

Lee et al. Blood Cells, Molecules, and Diseases (2002)29(3) Nov/Dec: 471–487

doi:10.1006/bcmd.2002.0586

Seeking Candidate Mutations That Affect Iron HomeostasisSubmitted 10/03/02(Communicated by E. Beutler, M.D., 10/14/02)

Pauline Lee,1 Terri Gelbart,1 Carol West,1 Carol Halloran,1 and Ernest Beutler1

ABSTRACT: Hereditary hemochromatosis is characterized by marked variation of expression of the defefew homozygotes with the C282Y/C282YHFE genotype have full-blown clinical disease, a larger number sbiochemical stigmata of iron overload, and some seem normal biochemically. The following candidate gebeen examined in detail to determine whether polymorphisms in them may be responsible for this vtransferrin, transferrin receptor 1, transferrin receptor 2, ferritin-L, ferritin-H,IRP1, IRP2, HFE, �2 microglobulin,mobilferrin/calreticulin, ceruloplasmin, ferroportin,NRAMP1, NRAMP2 (DMT1), haptoglobin, heme oxygenaseheme oxygenase-2, hepcidin,USF2, ZIRTL, duodenal cytochrome b ferric reductase (DCYTB), TNF�, keratin 8and keratin 18. The coding sequence, exon–intron junctions, and promoters of each of these genes wasin DNA from 20 subjects: 5HFE C282Y/C282Y with clinical disease, 5HFE C282Y/C282Y with normal/lowferritin levels and no disease, 5wt/wt with high ferritin and transferrin saturation, and 5wt/wt normal controlsWhen coding or promoter polymorphisms were encountered, DNA from large numbers of ethnicallysubjects was examined for these polymorphisms and a relationship between their existence and abnormof iron homeostasis was sought. Only in the case of one transferrin mutation did we find a strong relabetween the polymorphism and iron deficiency anemia. The putative genes that affect the expressionHFE

In-ousra-oseere

er-seswnbeennesaythe

orinedctsus-,o-

California

INTRODUCTION

While 60 to 100% of Europeans with heretary hemochromatosis are homozygous forHFE 845A3G (C282Y) mutation, it has beeshown clearly that very few of these patieactually develop iron storage disease (1, 2).deed, about one-third of the patients homozygfor this mutation have normal transferrin satutions and normal ferritin levels, even among thwho have not been blood donors. Clearly, thare other factors that influence the expressiothe homozygous state for theHFE C282Y mutation. Moreover, in southern Europe some 40%patients who have been diagnosed as havingstorage disease do not have mutations of theHFEgene. In Africa and Asia iron storage diseexists, but theHFE C282Y mutation is virtuallabsent from the population.

Mutations in genes other thanHFE, including

Correspondence and reprint requests to: Pauline Lee. E-mail: plee@1 The Scripps Research Institute, Department of Molecular and Exp

471

ferroportin, transferrin receptor-2, transferrin, culoplasmin, and the 1q-linked gene that caujuvenile hemochromatosis, are other knocauses of iron storage disease. Our goal hasto determine to what extent mutations in geknown to be involved in iron homeostasis mcause iron storage disease and may modifyexpression ofHFE mutations.

MATERIALS AND METHODS

Subjects

In an effort to identify genes that causemodulate iron storage disease we have exama core group of 20 white subjects: five subjewith high serum ferritin levels and homozygofor the common C282YHFE mutation, five subjects with high ferritin and normalHFE genotypefive subjects with low/normal ferritin and h

s.edu.tal Medicine, MEM215, 10550 North Torrey Pines Road, La Jolla,

0A)

mutations remain elusive.© 2002 Elsevier Science (US

92014.

1079-9796/02 $35.0© 2002 Elsevier Science (US

A)

di-then

nts

n of-

ofiron

asey

scripperimen

All rights reserved.

mozygous for the C282Y HFE genotype, and fivesubjects with low/normal ferritin and normal HFEgenotype (Table 1). For some genes of particularinterest, additional HFE C282Y homozygous andHFE wildtype iron overloaded subjects of variousethnicities were analyzed.

Detection Of Mutations

We have sought mutations in the coding re-gion, promoter regions, or untranslated regionscontaining iron-responsive elements of candidategenes by sequence analysis. All genomic se-quences, except transferrin, were obtained fromGenBank. The transferrin genomic sequence wasobtained as described (3). Whenever possible,primers were designed 60–120 bp from the in-tron–exon border in order to obtain completesequence of the exons and flanking splice junc-tion. Primers were generally designed with an-nealing temperatures between 60 and 64°C. Poly-merase chain reactions were performed using a1� reaction mix as described (4) with or withoutthe addition of 5% DMSO. Automated DNA se-quencing was performed using the ABI 377 DNAsequencer. Mutations that cause an amino acidchange or that are in the promoter region werefurther studied by allele specific oligomer hybrid-ization (ASOH) on a larger group of subjects withknown transferrin saturation and serum ferritinlevels.

RESULTS

Thus far, we have completed the analyses on24 candidate genes that might alter iron ho-meostasis directly and/or modulate the penetranceof iron overload disease (Table 2). These includegenes for which mutations or loss of expression

have been shown to alter iron homeostasis, i.e.,�2-microglobulin, ceruloplasmin, USF2, and hep-cidin; genes that are involved in iron transport orstorage, i.e., transferrin, transferrin receptor-1 and-2, ferroportin, nramp1 and nramp2 (DMT1),ZIRTL, duodenal cytochrome b ferric reductase(Dcytb), mobilferrin/calreticulin; ferritin light andheavy chains; iron regulatory genes, i.e., iron reg-ulatory proteins 1 and 2 (IRP1 and IRP2); genesinvolved in iron salvage, i.e., haptoglobin, hemeoxygenase-1 and -2 (HMOX1 and HMOX2); andgenes that are associated with susceptibility tohepatic cirrhosis, i.e., TNF�, keratin 8, and kera-tin 18.

TABLE 2

Genes Involved in Iron Homeostasis

Gene Sequenced Alleles studied

Iron Overload Related Genes in Mice and Men

HFE 6 exons 81,792�2Microglobulin 3 exons 14Ceruloplasmin 19 exons 944USF2 9 of 10 exons 40Hepcidin 3 exons 1373

Genes Involved in Iron Storage and Transport

Ferritin-L 4 exons 75Ferritin-H 4 exons 1,132Transferrin Promoter and 17 exons 9,740TfR 1 19 exons 3� IREs 350TfR 2 18 exons 462Ferroportin Promoter and 8 exons 6,068ZIRTL 5 exons 42NRAMP1 Promoter and 15 exons 632NRAMP2 (DMT1) 17 exons 254DCYTB 4 exons 342Calreticulin Promoter and 9 exons 32

Iron Regulatory Genes

IRP-1 21 exons 632IRP-2 22 exons 1406

Iron Salvage Genes

Haptoglobin n/a 464Heme oxygenase 1 5 exons 2068Heme oxygenase 2 6 exons 40

Cirrhosis Susceptibility Genes

TNF � Promoter 470Keratin 8 8 exons 506Keratin 18 7 exons 14

TABLE 1

Sequencing Strategy for Finding Polymorphisms That Influencethe Expression of HFE Mutations or That Cause Non-HFE

Hemochromatosis

C282Y/C282Y wt/wt

High ferritin 5 5Low/control ferritin 5 5

Blood Cells, Molecules, and Diseases (2002) 29(3) Nov/Dec: 471–487 Lee et al.doi:10.1006/bcmd.2002.0586

472

Genes in Which Disruption or Mutations Leadto an Iron Overload Phenotype

�2-microglobulin. Targeted disruption of �2-microglobulin has been shown to cause iron over-load in mice (5), presumably because the proteinis required to transport HFE to cell membranes(6). No mutations were found (7).

Ceruloplasmin. DNA of the core group of 20patients, one fetal hemochromatosis sample, oneAsian with a high transferrin saturation, one whitesubject with hyperferritinemia, and one AfricanAmerican with iron overload was sequenced. Twocoding region mutations were identified in theceruloplasmin gene, a R367C in exon 6 and aT551I in exon 9. The R367C and the T551I mu-tations were further examined on additional sub-jects. ASOH analyses revealed that the R367Cwas a polymorphism found in African Americans(frequency 0.043, 258 alleles) and absent inwhites (268 alleles). T551I was observed inwhites at a frequency of 0.035 (370 alleles). Nei-ther mutation was found to be associated with achange in serum ferritin and transferrin saturationlevels (4).

Hepcidin. Hepcidin was first identified as asmall peptide with antimicrobial activity in urineand serum (8). Subsequently, hepcidin was foundby subtractive hybridization to be upregulated iniron-overloaded mice (9) Mice lacking hepcidinexpression following Usf2 knockout were foundto exhibit iron overload (10). Transgenic miceoverexpressing hepcidin were found to exhibitsevere iron deficiency. These data demonstrate animportant role for hepcidin in iron homeostasis.Our initial studies of the hepcidin gene identifiedone polymorphism resulting in an amino acidchange, a C3T at nt 92 (T31M) in a normalsubject (4). This mutation was present in thewhite (including Hispanic) population at an allelefrequency of 0.0026 (n � 867). It was not ob-served in African Americans (n � 400) or Asians(n � 66). Genotyping identified a subject ho-mozygous for the 31M polymorphism. This fe-male of mixed descent (white and Hispanic) had

normal serum iron, ferritin, transferrin saturation,hematocrit, MCV, and hemoglobin values (4).

USF2. Disruption of the mouse Usf2 gene inexon 7 led to the loss of hepcidin gene expression,apparently as a result of a downstream cis effect(8). Since the human hepcidin gene also liesdownstream of the human USF2 gene, we inves-tigated the possibility that mutations in USF2might have an effect on expression of humanhepcidin. Exons 2–10 of the USF2 gene wereexamined in iron-overloaded and normal subjects.Exon 1 was not sequenced due to technical diffi-culties. No coding region polymorphisms result-ing in an amino acid change were identified. Fourhaplotypes could be identified with 11 single nu-cleotide polymorphisms (SNPs) (IVS5�19,IVS9-14, IVS9-6, nt 1056, nt 1174, nt 1283, nt1332, nt 1353, nt 1476, nt 1524, nt 1527) (Fig. 1).The distribution of the haplotypes did not differsignificantly between subjects with high and lowferritin or transferrin saturation levels.

Genes Involved in Iron Transport and Storage

Transferrin. Transferrin, as demonstrated byelectrophoretic and genetic analyses, is highlypolymorphic (Fig. 2). In the course of our study,we did not identify any new mutations in thetransferrin gene coding region in our iron-over-loaded patients that would account for iron stor-age disease (11). The gene frequencies of the fivemajor transferrin variants (TFv1–5) were exam-ined in 919 normal subjects and 113 iron HFEC282Y homozygous subjects. Transferrin variant4 (TFv4, P570S) is the most prevalent polymor-phism in all ethnic groups examined, occurring atan allele frequency of 0.1492. There was no affectof TFv4 genotype on changes in serum ferritin,transferrin saturation, MCV, or hemoglobin levelsin either normal subjects or hemochromatosissubjects homozygous for the HFE C282Y muta-tion (11, 12). The transferrin G277S variant(probably TF C3 variant) occurs at a gene fre-quency of 0.0634 in whites. We found this muta-tion to be a risk factor for iron deficiency anemia(13).

DNA from one patient with congenital atrans-

Lee et al. Blood Cells, Molecules, and Diseases (2002) 29(3) Nov/Dec: 471–487

doi:10.1006/bcmd.2002.0586

473

ferrinemia was sequenced in our laboratory. Thepatient was found to have a G3C transversion atnt 1429 resulting in a A477P mutation and a 10 bpdeletion followed by a 9 bp insertion resulting ina frameshift mutation (3).

The promoter region of the transferrin genewas also investigated for an effect on iron ho-meostasis. Five SNPs numbering from the start oftranscription (-739, -617, -551, -34, �49) wereused to define seven different haplotypes. Thetotal iron binding capacity (TIBC) associated witheach haplotype was haplotype 2�1�4�3. Trans-ferrin haplotype 2 had a significantly higher meanTIBC and haplotype 3 had a significantly lowermean TIBC than the more common haplotype 1.The different haplotypes were not associated withdifferences in serum ferritin or transferrin satura-tion levels in normal, iron overload, or anemicsubjects. White subjects with Parkinson’s disease

exhibited an slight excess of transferrin haplotype3 over the normal white subjects (14).

Transferrin receptor-1. Transferrin receptor-1is a key component in the entry of iron-loadedtransferrin into the cell and its trafficking withinthe cell. A polymorphism within this receptor wasoriginally described by Tsuchihashi et al. (15).Sequencing of the entire coding region includingthe 3� UTR containing the iron-responsive ele-ments of the transferrin receptor-1 gene identified,in addition to the previously described mutation, asilent G3A mutation at nt 2124 in exon 19(T708T), five common intron polymorphisms,and two uncommon intron polymorphisms (4).The five common intronic polymorphisms IVS12�17 (C3A), IVS13 �39 (A3G), IVS16 -12(A3G), IVS16 -5 (C3T), and IVS16 -4 (A3G)were in linkage disequilibrium and occurred at an

FIG. 1. Polymorphisms in the USF2 gene. All numbering begins at the start of translation, the “A” of ATG as nucleotide1. IVS: intervening sequence.


474

allele frequency of 0.60 in whites. The two lesscommon intronic polymorphisms IVS10 �13(C3A) and IVS13 �25 (T3G) were in or nearequilibrium and occurred at an allele frequency of0.02 (n � 46). The allele frequencies of eachpolymorphism did not differ significantly betweensubjects with severe hemochromatosis and normalsubjects (4).

Transferrin receptor-2. Transferrin receptor-2is a homolog of the classical transferrin receptor,transferrin receptor 1. Mutations in humans havebeen associated with iron overload (17–19). Inpreviously reported studies (16) we sequencedthis gene in 17 whites, 10 Asians, and 8 AfricanAmericans with iron overload and a C282C/C282C HFE genotype, as well as 4 subjects with-out iron overload and homozygous for the mutantHFE C282Y genotype, 5 patients with iron over-load and homozygous for the mutant HFE C282Y

genotype, and 5 normal individuals. None of theindividuals exhibited the Sicilian mutations,Y250X in exon 6, M172K in exon 4, and E60X inexon 2. One iron-overloaded individual of Asiandescent exhibited a I238M mutation which wassubsequently found to be a polymorphism presentin the Asian population at a frequency of 0.0192.The presence of the I238M mutation was notassociated with an increase in ferritin or trans-ferrin saturation levels. Three silent polymor-phisms were also identified, nt 1770 (D590D) andnt 1851 (A617A) and a polymorphism at nt 2255in the 3� UTR.

Ferroportin. In our initial studies of the fer-roportin gene, we identified three polymorphicregions 5� of the start of translation, and an A3Gchange at nt 1681 resulting in R561G mutation inexon 8 (4). The polymorphic microsatellite in thepromoter region and the single nucleotide poly-

FIG. 2. Polymorphisms in the transferrin gene. All numbering begins at the start of translation, the “A” of ATG asnucleotide 1. C1, C2, C3, B2, Bv, D1, chi: transferrin variant C1, C2, C3, B2, Bv, D1 and chi, respectively, UTR: untranslatedregion; Pro: promoter region; IVS: intervening sequence.


doi:10.1006/bcmd.2002.0586

475

morphisms at nt -98 and -8 were examined andwere not found to have a significant affect onferritin or transferrin saturation levels on normalsubjects or subjects homozygous for the HFEC282Y mutation. The R561G mutation was foundin a normal subject.

Since our initial studies were performed (4,20), iron overload disease has been attributed tonew mutations in the ferroportin gene (21–27). Asa result, we sequenced the entire coding region offerroportin in 25 additional iron overloaded sub-jects of whom 13 are white, 5 African Americanand 7 Asian. One African American iron-over-loaded subject was identified to be heterozygousfor the R561G polymorphism, which we de-scribed previously in a normal white subject. Onewhite iron-overloaded subject was heterozygousfor a C3G change at nt 1070 resulting in a T3Schange at amino acid 357. This mutation waspresent along with a silent A3G change at nt1218 (E406E). Further analysis of the T357Spolymorphism was performed. No additional sub-jects carrying the T357S mutation were found innormal subjects (418 alleles), white subjects withferritin �500, transferrin saturation �45% andnormal HFE genotype (180 alleles) and whitesubjects with ferritin �200, transferrin saturation�40 and normal HFE genotype (342 alleles). No

mutations in the ferroportin gene were detected inany of the Asian iron overload subjects (Fig. 3).

Zinc iron regulated transporter-like (ZIRTLor SLC39A). ZIRTL is a mammalian member ofthe divalent metal transporter family includingyeast ZRT1 and plant IRT1 and ZIP 1-4 whichtransport zinc and iron and possibly other divalentmetals like Cd2�, Co2� and Mn2� (28, 29). IRT,induced with iron deficiency, and ZIP1 and 3induced by zinc deficiency are expressed in plantroots. Human ZIRTL mRNA was detected in mostadult and fetal tissues (28, 30). The human ZIRTLgene is located within the epidermal differentia-tion complex on chromosome 1q21 near, but out-side, the critical region for juvenile hemochroma-tosis (28, 31). The human ZIRTL gene spans 5exons, the first two of which correspond to 5�UTRsequence. The coding region (ex 3–5) were ex-amined in a child with microcytic anemia, a sub-ject with juvenile hemochromatosis, and the coregroup of 20 iron overload and normal subjects.No mutations or polymorphisms were found inthe ZIRTL gene.

NRAMP1 (SLC 11A1). NRAMP1 was firstidentified as a gene associated with natural resis-tance to infection expressed by macrophages of

FIG. 3. Polymorphisms in the ferroportin gene. All numbering begins at the start of translation, the “A” of ATG asnucleotide 1. MS: microsatellite; UTR: untranslated region; IVS: intervening sequence.


476

reticulo-endothelial origin (32). With the identifi-cation of its homolog, NRAMP2, as a divalentmetal transporter (33, 34), NRAMP1 was shownto also effect intracellular iron levels. Macro-phages expressing high NRAMP1 activity demon-strated increased iron efflux resulting in low in-tracellular iron levels (35, 36). Similarly lowNRAMP1 activity was associated with elevatedlevels of intracellular iron. It was interesting tonote that mice homozygous for the Nramp1 D169mutation such as C57BL6 and BALB/c expressedlow transferrin saturation (38% and 46%) whereasmice such as DBA/2 which were homozygous forthe wildtype Nramp1 gene (G169) expressed hightransferrin saturation levels (76%) (37). Thisraised the question as to whether NRAMP1 mightbe the modulator of hemochromatosis penetrancein different mouse strains and man.

The human NRAMP1 gene spans 16 exons,including an alternatively spliced exon, and con-tains a polymorphic microsatellite in the promoterregion, four silent polymorphisms in the codingregion (L39L, G65G, F66F, G249G), three poly-morphisms resulting in an amino acid change(A318V, V350M, and D543N), two single nucle-otide polymorphisms, one deletion, and one del/insertion polymorphism in the 3�UTR, as well asnumerous intron polymorphisms (Fig. 4).

Of the three coding region polymorphismsresulting in an amino acid change, only theV350M mutation was observed in the initialscreen and thus examined further. The V350Mmutation occurred at an allele frequence of 0.0051(n � 592 alleles) and was not correlated withabnormal iron values or disease.

The promoter microsatellite was examinedand the results were inconclusive. The frequen-cies of the four polymorphic sites in the 3�UTRwere not found to differ between iron-overloadedand normal subjects.

NRAMP-2 (DMT1). Spontaneously occuringmutations of Nramp2 have been found to be thecause of the mk mouse (34) and of the Belgraderat (39), animal models of a microcytic anemiawith iron deficiency. We found that in humansthere were two major splice forms of theNRAMP2 mRNA (38). Five single nucleotidepolymorphisms were identified within theNRAMP2 gene. One of these, nt 1303C3A, oc-curs in the coding region of NRAMP2 and resultsin an amino acid change from leucine to isoleu-cine. A polymorphism, nt 1254T3C, also occursin the coding region of NRAMP-2 but does notcause an amino acid change. The other three poly-morphisms are within introns (IVS2�11A/G,

FIG. 4. Polymorphisms in the nramp1 gene. All numbering begins at the start of translation, the “A” of ATG as nucleotide1. Pro MS: promoter microsatellite; IVS: intervening sequence.


doi:10.1006/bcmd.2002.0586

477

IVS4�44C/A, and IVS6�538G/Gdel). In addi-tion, a polymorphic microsatellite TATATC-TATATATC(TA)(6–7) (CA)(10–11) CCCCCTATA-(TATC)(3) (TCTG)(5) (TCCG(TCTA)(6) was iden-tified in intron 3. Analysis of cDNA derivedby direct amplification of reversed transcribedRNA or cDNA clones isolated from a libraryprovides evidence of skipping of exons 10 and 12of NRAMP2. Deletion of either of these exonswould result in a sequence that remains in frameyet would generate a protein that would lacktransmembrane spanning regions 7 or 8, respec-tively. The deletion of a single transmembranedomain would have severe topological conse-quences. The coding region of the NRAMP2 geneof severe hemochromatosis patients with or with-out mutations in the HFE gene were examinedand found to be normal. One hemochromatosispatient, with a normal HFE genotype, was het-erozygous for the 1303C3A mutation. Further-more, in an examination of hemochromatosis pa-tients with mutant HFE and normal HFE genes,we did not observe a linkage disequilibrium ofeither group with a particular NRAMP2 haplo-type.

Duodenal cytochrome b-like ferric reductase(DCYTB). The duodenal cytochrome b-like ferricreductase is postulated to reduce inorganic iron tobe taken up by transporter proteins such as

DMT1, which is specific for ferrous iron. Muta-tions in DCYTB are therefore likely to present asiron deficiency anemia. The GenBank SNP data-base identified a common polymorphism in exon4, nt 797A3G (S266N), of DCYTB (Fig. 5). Weexamined 83 nonanemic white females and 85anemic white females (ferritin �20, hemoglobin�12) to determine if this DCYTB mutation wasassociated with anemia. We found that the allelefrequency (G) was 0.3353 in the anemic groupand 0.3554 in the nonanemic, which was notstatistically significant.

In two normal subjects, a G3C polymor-phism at nt -98 and a A3G polymorphism at nt-94, apparently in linkage disequilibrium, wereidentified in the promoter region of the DCYTBgene. These polymorphisms have not been furtherinvestigated.

Mobilferrin/calreticulin. Mobilferrin, a ho-molog of calreticulin, is part of a protein complexnamed paraferritin which also includes �3 inte-grin, flavin monooxygenase, �2 microglobulin,and a large GTP binding protein that has beenshown to be important in the uptake of ferric ironin duodenal mucosal cells, K562 cells, and IEC-6rat small intestinal cells (40). Calreticulin mayalso play a role in the assembly and transport ofMHC Class I molecules, of which HFE is a mem-ber, through the endoplasmic reticulum (41). This

FIG. 5. Polymorphisms in the duodenal cytochrome b like ferric reductase gene. All numbering begins at the start oftranslation, the “A” of ATG as nucleotide 1.


478

raised the question as to whether mutations in thecalreticulin gene might affect iron homeostasis.We have examined the promoter region and the 9exons of the calreticulin gene in 5 subjects withiron overload disease, 3 who had a wildtype HFEgenotype and 2 who were heterozygous for theHFE C282Y mutation. An additional 8 HFE wild-type hemochromatosis subjects and 3 normal sub-jects were examined for mutations in the promoterregion. There were no apparent mutations thatcould account for iron overload disease (7).

Ferritin, light and heavy chains. Mutations inthe ferritin light chain gene have been associatedwith hyperferritinemia-cataract syndrome (42–45)and dominantly inherited, late-onset basal gangliadisease (46). None of the described mutationswere observed in our patient population nor wereadditional mutations found in the ferritin lightchain (4).

Mice with targeted disruptions in the ferritinheavy chain gene are not viable (47). Heterozy-gous ferritin heavy chain � mice were viable butexhibited hyperferritinemia (47). In humans, amutation in the IRE of ferritin heavy chain hasbeen described in a Japanese family with autoso-mal dominant iron overload (48). Although thedescribed IRE mutation in ferritin heavy chainwas observed in a ferritin heavy chain processedpseudogene, our analyses of the functional geneusing gene-specific primers demonstrated the ab-sence of the IRE mutation and other coding regionmutations in our patient group (4).

Iron Regulatory Proteins

Iron regulatory protein 1 (IRP1). Examina-tion of the 21 exons comprising the IRP1 gene inthe panel of iron overload and normal subjectsidentified a single C3A change at nt 1664 result-ing in a P555H mutation in a homozygous HFEC282Y subject with low/normal serum ferritinand transferrin saturation levels. Analysis of anadditional 2218 alleles did not identify additionalsubjects with this mutation. Proline 555 is con-served between IRP1 and IRP2 but not mitochon-drial aconitase. The role of the P555H mutationwas not assessed further.

Iron regulatory protein 2 (IRP2). Mice withtargeted disruptions in the IRP2 gene exhibit hy-perferritinemia and neural degeneration (49). Se-quencing of the 22 exons of the IRP2 gene iden-tified three mutations, a C3G transversion at nt816 (F272L), a C3G transversion at nt 475(L159V), and a C3T change at nt 1739 (T580I).The F272L mutation occurred in whites at anallele frequency of 0.0014 (1384 alleles). Theallele frequency was too low to determine if theF272L mutation had a significant effect on serumferritin or transferrin saturation levels (4). Fur-thermore, the F272L mutation was not found to besignificantly associated with sporadic Parkinson’sdisease (50). The L159V and T580I mutations, inlinkage disequilibrium, were found only in Afri-can Americans at a frequency of 0.102 (1236alleles). We did not find a significant correlationbetween L159V/T580I genotype versus serumferritin and transferrin saturation levels (4).

Genes Involved in Iron Salvage

Haptoglobin. Two major alleles of haptoglo-bin, type 1 and 2, are found in the general popu-lation. Type 2 haptoglobin contains an internalduplication of exons 3 and 4. It has been reportedthat subjects homozygous for type 2 haptoglobinmanifest increased levels of iron, including serumiron, ferritin, and transferrin saturation (51, 52).Furthermore, men with hemochromatosis, ho-mozygous for the HFE C282Y mutation, were ata greater risk for iron overload when they werealso homozygous for haptoglobin type 2 (53). Wehave examined the relationship between haptoglo-bin type and iron levels in 115 HFE C282Yhomozygous subjects and 117 matched controlswith an HFE wildtype genotype. We found thatthere was no correlation between haptoglobintype and serum ferritin and transferrin saturationlevels (54).

Heme oxygenases. Heme oxygenases (HO)are important enzymes in the salvage of iron fromheme. HO catabolizes heme to biliverdin and re-leases free iron and carbon monoxide. There aretwo heme oxygenase genes, HMOX1 andHMOX2. HMOX1 is markedly induced by heme


doi:10.1006/bcmd.2002.0586

479

and stress. HMOX2 is constitutively expressed.HO deficiency in humans results in growth retar-dation, hemolytic anemia, renal and hepatic irondeposition, asplenia, and vulnerability to stressinjury (55, 56) HO deficiency in mice was alsofound to have reduced serum iron levels and hightotal iron binding capacity (55).

HMOX1 transfected into HEK cells exhibitsincreased iron efflux and reduced iron uptake.Similarly, fibroblasts isolated from HMOX1�/�mice demonstrate reduced iron release and in-creased iron uptake (57). These data clearly dem-onstrate a role for heme oxygenase in iron efflux.HMOX1 and HMOX2 were examined for muta-tions that might alter iron levels in human sub-jects.

Heme oxygenase 1. Sequencing of the five

exons comprising the HMOX1 gene in humansidentified three common polymorphisms, a C3Tchange at nt -89 numbering from the initiatormethionine (A of ATG is 1), a G3C change at nt19 resulting in an amino acid change (D7H), anda deletion in intron 4 (IVS 52-64) (Fig. 6). Themicrosatellite in the promoter region was not ex-amined.

The D7H polymorphism was investigated fur-ther. Examination of 2068 alleles demonstratedthat the allele frequency was 0.0566 in the whitepopulation. Subjects homozygous for the His7allele tended toward lower ferritin and transferrinsaturation levels but the differences were not sta-tistically significant (Table 3).

Heme oxygenase 2. The HMOX2 gene is com-prised of six exons. A silent polymorphism in

TABLE 3

Effect of Hmox1 nt19 G 3 C (D7H) on Indices of Iron Metabolism

C/C C/G G/G

Males

Subjects 3 103 862Ferritin 65.85 � 1.2 118.06 � 1.1 118.83 � 1.0

(29.0–149.5) (100.7–138.5) (112.7–125.2)Transferrin saturation 28.33 � 7.5 28.22 � 0.9 26.48 � 0.3

(�4.1–60.8) (26.4–30.1) (25.8–27.1)Serum iron 96.67 � 28.0 93.88 � 3.1 91.19 � 1.1

(�23.9–217.3) (87.7–100.1) (89.0–93.4)Hemoglobin 13.6 � 0.5 15.1 � 0.1 15.1 � 0.03

(11.5–15.7) (14.9–15.3) (15.0–05.2)TIBC 336.67 � 20.3 335.27 � 4.1 347.81 � 1.5

(249.1–424.2) (327.2–343.4) (344.9–350.7)MCV 89.97 � 1.9 89.12 � 0.4 89.38 � 0.1

(81.5–98.4) (88.2–90.0) (89.1–89.7)

Females

Subjects 1 8 58Ferritin 10.99 42.86 � 1.16 54.35 � 1.01

(30.1–61.0) (42.3–69.8)Transferrin saturation 23 28.13 � 3.4 24.53 � 1.1

(20.1–26.2) (22.3–26.8)Serum iron 105 102.5 � 8.8 87.29 � 3.7

(81.6–123.4) (79.8–94.8)Hemoglobin 12.9 13.3 � 0.4 13.4 � 0.1

(12.4–14.2) (13.2–13.7)TIBC 454 373.5 � 14.1 362.29 � 7.1

(340.2–406.8) (348.1–376.5)MCV 96 92.1 � 1.3 89.9 � 0.5

(89.1–95.1) (88.9–90.9)


480

exon 3, a G3A change at nt 103 (Ser34), and aG3A polymorphism at nt 1491 in the 3� UTRwere identified (Fig. 7). These polymorphisms didnot correlate with iron overload disease and werenot investigated further.

Genes Involved in Susceptibility to Cirrhosis

It may be that penetrance of the hemochroma-tosis phenotype requires the presence of other risk

factor mutations in genes not directly related toiron, such as genes that may confer susceptibilityto hepatic cirrhosis.

TNF promoter. TNF� has been shown to par-ticipate in the regulation of iron levels by upregu-lating ferritin and transferrin receptor levels (58–61), and inhibiting iron release (62). TNF� is alsoregulated by iron in that the release of TNF� from

FIG. 6. Polymorphisms in the heme oxygenase 1 gene. All numbering begins at the start of translation, the “A” of ATGas nucleotide 1. Pro MS: promoter microsatellite; IVS: intervening sequence.

FIG. 7. Polymorphisms in the heme oxygenase gene. All numbering begins at the start of translation, the “A” of ATGas nucleotide 1.


doi:10.1006/bcmd.2002.0586

481

macrophages is enhanced following iron deple-tion and reduced following iron loading (63).Macrophages from patients with hemochromato-sis subjects release less TNF� than those of con-trol subjects (64, 65). It has been suggested byFargion et al. (65) that promoter polymorphismsin the TNF� gene influence the phenotypic ex-pression of hemochromatosis. In particular, it wassuggested that the �238 promoter major allelewas a risk factor for penetrance of the hemochro-matosis phenotype; the minor allele was proposedto be hepatoprotective. Using serum collagen IVlevels as an indicator of hepatic fibrosis, we foundno correlation between TNF� �238 genotype andserum collagen, aspartate transaminase, and fer-ritin levels in hemochromatosis patients homozy-gous for the HFE C282Y mutation (66).

Keratin 8 and keratin 18. Keratin 8 and 18 arethe major keratin pair (K8/18) found in epithelialcells of liver, intestine, and pancreas. Disruptionof the keratin 8 gene in mice results in extensiveliver hemorrhage and is lethal to the embryo (67).Heterozygous mutations in the keratin 8 gene, aT3C at nt 161 (Y54H) and a G3T at nt 184(G62C), were identified in patients with crypto-genic cirrhosis (68). We examined the eight exonsof the keratin 8 gene in patients with severe ironoverload and severe liver disease.

Of 13 subjects with iron overload and livercirrhosis or 5 control subjects, none carried theY54H or the G62C mutations in exon 1. Threepolymorphisms were identified in the 5�UTR, anA3C at nt �70, a T3A at nt �10, and a T3Cat nt �4, which did not correlate with the pres-ence or absence of liver disease. Three intronpolymorphisms, a C3G at IVS2�55, an A3Gat IVS2�60, and a C deletion at IVS7�9, werealso identified (Fig. 8).

A G3A mutation at nt 1022 (R341H) was

identified in 3 out of the 13 iron overload subjectswith liver cirrhosis but not in normal subjects.This mutation was examined in 119 subjects ho-mozygous for the HFE C282Y mutation and 116controls matched by age, sex, and ethnicity. Theeffect of the R341H mutation on serum collagenIV levels, an indicator of liver fibrosis, was ex-amined. We found that subjects heterozygous forthe keratin 8 R341H mutation tended towardlower collagen IV levels than subjects with thewildtype keratin 8 allele. These differences werenot statistically significant (Table 4).

Transgenic mice that express mutant keratin18 exhibit chronic hepatitis and an increased sen-sitivity to drug-induced liver damage. A humankeratin 18 mutation, H127L, was identified in apatient with cryptogenic cirrhosis (69). Seveniron-overloaded subjects with liver cirrhosis wereexamined for mutations in the coding region ofthe keratin 18 gene. No mutations or polymor-phisms were found in any of the seven exons ofthe keratin 18 gene.

DISCUSSION

As our ability to identify mutations at a nu-cleotide level has increased, it has become abun-dantly clear that there is enormous variability inthe clinical phenotype of patients who carry thesame disease-producing mutations. This observa-tion is so universal that it has rightly been saidthat there are no single gene diseases, and this isnowhere more true than in the case of hemochro-matosis. The variability of the clinical expressionof homozygosity for the HLA-linked hemochro-matosis mutation, now known to be the C282Ymutation of the HFE gene, has been known for along time. However, the vanishing low penetranceof this phenotype has come to be appreciated onlyin the past 2 or 3 years. Such differences could, of

TABLE 4

Effect of Keratin 8 R341H Mutation on Collagen Levels in HFE C282Y Homozygous Subjects

HFE Keratin 8 No. of Subjects Average Collagen � SE 95% Confidence Interval

C282Y/C282Y 341H/341R 9 116.85 � 19.9 70.96–162.75C282Y/C282Y 341R/341R 110 165.63 � 12.9 139.86–191.41C282C/C282C 341R/341R 116 135.54 � 7.3 121.03–150.05


482

course, be due to environmental factors. The as-sociation between hemochromatosis and the in-gestion of alcohol has long been known (70). Itcould also be a consequence of epigenetic factorssuch as retrotransposons that have been shown tocause strikingly phenotypic differences in genet-ically identical mice (71). But the possibility thatit is other genes that determine whether or not theperson homozygous for the C282Y HFE mutationactually develops hemochromatosis is attractiveand potentially testable.

The interaction of mutations influencing thecourse of genetic diseases has been demonstratedpreviously in the case of disorders of hemostasis,thalassemias, and neonatal jaundice associatedwith G6PD deficiency. In all of these instances,candidate genes were studied: in factor 5 Leidenmutations of protein C (72); in � thalassemia,those of � thalassemia (73); and in G6PD defi-ciency that the UDP glucuronosyltransferase pro-moter associated with Gilbert disease (74). In thepresent study we have sought, in a similar fashion,to identify a gene involved in iron metabolismthat accounts for a significant portion of the vari-

ability seen in homozygotes for the C282Y mu-tation.

In the present study we have examined 24candidate genes to determine whether functionalpolymorphisms in these genes affected the ex-pression of the homozygous state for the C282YHFE mutation. The genes that were selected in-cluded genes whose products are involved in irontransport (such as transferrin, transferrin receptor1, transferrin receptor 2, DMT1, ferroportin, andhephaestin), those involved in iron regulation(such as hepcidin IRP1 and IRP2), and genes thatmay modify the response to excess iron (such asthe TNF�). When mutations that were judged tohave a possible functional effect were detected,the effect of such a mutation was tested on theDNA from a large number of subjects with knowntransferrin saturation and ferritin levels.

Our approach has been straightforward. Wehave sequenced the coding regions, intron–exonboundaries, and promoters of 20 individuals tofind polymorphisms. We have tried to increase theprobability of finding a relatively low frequencypolymorphism by enriching the population with

FIG. 8. Polymorphisms in the keratin 8 gene. All numbering begins at the start of translation, the “A” of ATG asnucleotide 1. IVS: intervening sequence.


doi:10.1006/bcmd.2002.0586

483

homozygotes for the C282Y mutation who have arelatively severe phenotype and also with subjectswho have the wildtype HFE gene but have evi-dence of iron overload. The latter group, we rea-soned, could well have severe mutations of a genewhich, with minor mutations or in the heterozy-gous state, increased the expression of the C282Yhomozygotes. There was a problem, of course, inselecting severely affected individuals. In spite ofthe obvious skepticism by some that hemochro-matosis is rare, we have contacted many centers inan attempt to obtain DNA from individuals withsevere expression of the defect, and we havefound none. We have thus had to content our-selves with subjects with a less severe phenotype,i.e., persons who, in point of fact, do not expressall of the clinical stigmata of severe hereditaryhemochromatosis, cirrhosis, diabetes, cardiomy-opathy, and darkening of the skin. Instead, someof our “severely affected” subjects have littlemore than a very high ferritin or a considerableiron burden as demonstrated by serial phlebot-omy. In any case, only one mutation was foundthat appeared to influence iron metabolism, andthis mutation, the G277S of transferrin, was a riskfactor for iron deficiency, not for iron overload.

It may be that our failure to find a modifyinggene was due to the dearth of fully affected ho-mozygotes. It is possible, too, that environmentalfactors play a larger role than has generally beenappreciated or that epigenetic factors, clearly fac-tors that will be difficult to identify on an exper-imental basis, play a dominant role in the expres-sion of hemochromatosis (75). More likely, per-haps, is the possibility that a gene not yet knownto play a role in iron homeostasis is the importantone. One such gene might be the elusive onelocated on chromosome 1q that is the cause ofjuvenile hemochromatosis. In spite of the ad-vances that have been made in the past few years,our understanding of iron metabolism remainsincomplete. If the gene(s) that accounts for thevariable expression of hemochromatosis can befound, the discovery will not only be medicallyuseful, but may shed new and unexpected light onthe mechanisms that control total body iron con-tent.

ACKNOWLEDGMENTS

This is Manuscript No. 15336-MEM. Supported byNational Institutes of Health Grant DK53505 and theStein Endowment Fund. This work was presented at asymposium on “Molecular and Clinical Aspects of Hu-man Iron Metabolism” September 28–October 3, 2002,Lake Chiemsee, Bavaria, Germany.

REFERENCES

1. Åsberg, A., Hveem, K., Kruger, O., and Bjerve, K. S.(2002) Persons with screening-detected haemochro-matosis: as healthy as the general population? Scand.J. Gastroenterol. 37, 719–724.

2. Beutler, E., Felitti, V. J., Koziol, J. A., Ho, N. J., andGelbart, T. (2002) Penetrance of the 845G6A(C282Y) HFE hereditary haemochromatosis mutationin the USA. Lancet 359, 211–218.

3. Beutler, E., Gelbart, T., Lee, P., Trevino, R., Fernan-dez, M. A., and Fairbanks, V. F. (2000) Molecularcharacterization of a case of atransferrinemia. Blood96, 4071–4074.

4. Lee, P. L., Gelbart, T., West, C., Halloran, C., Felitti,V., and Beutler, E. (2001) A study of genes that maymodulate the expression of hereditary hemochromato-sis: Transferrin receptor-1, ferroportin, ceruloplasmin,ferritin light and heavy chains, iron regulatory proteins(IRP)-1 and -2, and hepcidin. Blood Cells Mol. Dis.27, 783–802.

5. de Sousa, M., Reimao, R., Lacerda, R., Hugo, P.,Kaufmann, S. H., and Porto, G. (1994) Iron overloadin beta 2-microglobulin-deficient mice. Immunol. Lett.39, 105–111.

6. Feder, J. N., Tsuchihashi, Z., Irrinki, A., Lee, V. K.,Mapa, F. A., Morikang, E., Prass, C. E., Starnes,S. M., Wolff, R. K., Parkkila, S., Sly, W. S., andSchatzman, R. C. (1997) The hemochromatosisfounder mutation in HLA-H disrupts {beta}2-micro-globulin interaction and cell surface expression.J. Biol. Chem. 272, 14025–14028.

7. Beutler, E., West, C., and Gelbart, T. (1997) HLA-Hand associated proteins in patients with hemochroma-tosis. Mol. Med. 3, 397–402.

8. Park, C. H., Valore, E. V., Waring, A. J., and Ganz, T.(2001) Hepcidin, a urinary antimicrobial peptide syn-thesized in the liver. J. Biol. Chem. 276, 7806–7810.

9. Nicolas, G., Bennoun, M., Devaux, I., Beaumont, C.,Grandchamp, B., Kahn, A., and Vaulont, S. (2001)Lack of hepcidin gene expression and severe tissueiron overload in upstream stimulatory factor 2 (USF2)knockout mice. Proc. Natl. Acad. Sci. USA 98, 8780–8785.

10. Nicolas, G., Bennoun, M., Porteu, A., Mativet, S.,


484

Beaumont, C., Grandchamp, B., Sirito, M., Sawadogo,M., Kahn, A., and Vaulont, S. (2002) Severe irondeficiency anemia in transgenic mice expressing liverhepcidin. Proc. Natl. Acad. Sci. USA 99, 4596–4601.

11. Lee, P. L., Ho, N. J., Olson, R., and Beutler, E. (1999)The effect of transferrin polymorphisms on iron me-tabolism. Blood Cells Mol. Dis. 25, 374–379.

12. Fairbanks, V. F., Fahey, J. L., and Beutler, E. (1971)Clinical Disorders of Iron Metabolism. Grune & Strat-ton, New York.

13. Lee, P. L., Halloran, C., Trevino, R., Felitti, V., andBeutler, E. (2001) Human transferrin G277S muta-tion: A risk factor for iron deficiency anaemia. Br. J.Haematol. 115, 329–333.

14. Lee, P. L., Halloran, C., and Beutler, E. (2001) Poly-morphisms in the transferrin 5� flanking region asso-ciated with differences in total iron binding capacity:Possible implications in iron homeostasis. Blood CellsMol. Dis. 27, 539–548.

15. Tsuchihashi, Z., Hansen, S. L., Quintana, L., Kronmal,G. S., Mapa, F. A., Feder, J. N., and Wolff, R. K.(1998) Transferrin receptor mutation analysis in he-reditary hemochromatosis patients. Blood Cells Mol.Dis. 24, 317–321.

16. Lee, P. L., Halloran, C., West, C., and Beutler, E.(2001) Mutation analysis of the transferrin receptor-2gene in patients with iron overload. Blood Cells Mol.Dis. 27, 285–289.

17. Camaschella, C., Roetto, A., Cali, A., De Gobbi, M.,Garozzo, G., Carella, M., Majorano, N., Totaro, A.,and Gasparini, P. (2000) The gene TFR2 is mutated ina new type of haemochromatosis mapping to 7q22.Nature Genet. 25, 14–15.

18. Girelli, D., Bozzini, C., Roetto, A., Alberti, F., Daraio,F., Colombari, R., Olivieri, O., Corrocher, R., andCamaschella, C. (2002) Clinical and pathologic find-ings in hemochromatosis type 3 due to a novel muta-tion in transferrin receptor 2 gene. Gastroenterology122, 1295–1302.

19. Roetto, A., Totaro, A., Piperno, A., Piga, A., Longo,F., Garozzo, G., Cali, A., De, G., Gasparini, P., andCamaschella, C. (2001) New mutations inactivatingtransferrin receptor 2 in hemochromatosis type 3.Blood 97, 2555–2560.

20. Lee, P. L., Gelbart, T., Halloran, C., and Beutler, E.(2000) Mutations in candidate genes that may influ-ence the hemochromatosis phenotype: Ferroportin andceruloplasmin. Blood 96, 223a.

21. Njajou, O. T., Vaessen, N., Joosse, M., Berghuis, B.,van Dongen, J. W., Breuning, M. H., Snijders, P. J.,Rutten, W. P., Sandkuijl, L. A., Oostra, B. A., VanDuijn, C. M., and Heutink, P. (2001) A mutation inSLC11A3 is associated with autosomal dominanthemochromatosis. Nature Genet. 28, 213–214.

22. Fleming, R. E. and Sly, W. S. (2001) Ferroportin

mutation in autosomal dominant hemochromatosis:loss of function, gain in understanding. J. Clin. Invest.108, 521–522.

23. Montosi, G., Donovan, A., Totaro, A., Garuti, C.,Pignatti, E., Cassanelli, S., Trenor, C. C., Gasparini,P., Andrews, N. C., and Pietrangelo, A. (2001) Auto-somal-dominant hemochromatosis is associated with amutation in the ferroportin (SLC11A3) gene. J. Clin.Invest. 108, 619–623.

24. Njajou, O. T., Vaessen, N., Oostra, B., Heutink, P.,and Van Duijn, C. M. (2002) The hemochromatosisN144H mutation of SLC11A3 gene in patients withtype 2 diabetes. Mol. Genet. Metab. 75, 290–291.

25. Roetto, A., Merryweather-Clarke, A. T., Daraio, F.,Livesey, K., Pointon, J. J., Barbabietola, G., Piga, A.,Mackie, P. H., Robson, K. J. H., and Camaschella, C.(2002) A valine deletion of ferroportin 1: A commonmutation in hemochromatosis type 4? Blood 100, 733–734.

26. Devalia, V., Carter, K., Walker, A. P., Perkins, S. J.,Worwood, M., May, A., and Dooley, J. S. (2002)Autosomal dominant reticuloendothelial iron overloadassociated with a 3-base pair deletion in the ferropor-tin 1 gene (SLC11A3). Blood 100, 695–697.

27. Wallace, D. F., Pedersen, P., Dixon, J. L., Stephenson,P., Searle, J. W., Powell, L. W., and Subramaniam,V. N. (2002) Novel mutation in ferroportin1 is asso-ciated with autosomal dominant hemochromatosis.Blood 100, 692–694.

28. Lioumi, M., Ferguson, C. A., Sharpe, P. T., Freeman,T., Marenholz, I., Mischke, D., Heizmann, C., andRagoussis, J. (1999) Isolation and characterization ofhuman and mouse ZIRTL, a member of the IRT1family of transporters, mapping within the epidermaldifferentiation complex. Genomics 62, 272–280.

29. Eng, B. H., Guerinot, M. L., Eide, D., and Saier,M. H., Jr. (1998) Sequence analyses and phylogeneticcharacterization of the ZIP family of metal ion trans-port proteins. J Membr. Biol 166, 1–7.

30. Lioumi, M., Olavesen, M. G., Nizetic, D., and Ra-goussis, J. (1998) High-resolution YAC fragmentationmap of 1q21. Genomics 49, 200–208.

31. Papanikolaou, G., Politou, M., Roetto, A., Bosio, S.,Sakelaropoulos, N., Camaschella, C., and Loukopou-los, D. (2001) Linkage to chromosome 1q in Greekfamilies with juvenile hemochromatosis. Blood CellsMol. Dis. 27, 744–749.

32. Vidal, S. M., Malo, D., Vogan, K., Skamene, E., andGros, P. (1993) Natural resistance to infection withintracellular parasites: isolation of a candidate for Bcg.Cell 73, 469–485.

33. Gunshin, H., Mackenzie, B., Berger, U. V., Gunshin,Y., Romero, M. F., Boron, W. F., Nussberger, S.,Gollan, J. L., and Hediger, M. A. (1997) Cloning and


doi:10.1006/bcmd.2002.0586

485

characterization of a mammalian proton-coupled met-al-ion transporter. Nature 388, 482–488.

34. Fleming, M. D., Trenor, C. C. 3., Su, M. A., Foernzler,D., Beier, D. R., Dietrich, W. F., and Andrews, N. C.(1997) Microcytic anaemia mice have a mutation inNramp2, a candidate iron transporter gene. NatureGenet. 16, 383–386.

35. Barton, C. H., Biggs, T. E., Baker, S. T., Bowen, H.,and Atkinson, P. G. P. (1999) Nramp1: A link betweenintracellular iron transport and innate resistance tointracellular pathogens. J. Leuk. Biol. 66, 757–762.

36. Biggs, T. E., Baker, S. T., Botham, M. S., Dhital, A.,Barton, C. H., and Perry, V. H. (2001) Nramp1 mod-ulates iron homoeostasis in vivo and in vitro: Evi-dence for a role in cellular iron release involvingde-acidification of intracellular vesicles. Eur. J. Immu-nol. 31, 2060–2070.

37. LeBoeuf, R. C., Tolson, D., and Heinecke, J. W.(1995) Dissociation between tissue iron concentra-tions and transferrin saturation among inbred mousestrains. J. Lab. Clin. Med. 126, 128–136.

38. Lee, P. L., Gelbart, T., West, C., Halloran, C., andBeutler, E. (1998) The human nramp2 gene: Charac-terization of the gene structure, alternative splicing,promoter region and polymorphisms. Blood CellsMol. Dis. 24, 199–215.

39. Fleming, M. D., Romano, M. A., Su, M. A., Garrick,L. M., Garrick, M. D., and Andrews, N. C. (1998)Nramp2 is mutated in the anemic belgrade (b) rat:evidence of a role for nramp2 in endosomal irontransport. Proc. Natl. Acad. Sci. USA 95, 1148–1153.

40. Conrad, M. E., Umbreit, J. N., and Moore, E. G.(1999) Iron absorption and transport. Am. J. Med. Sci.318, 213–229.

41. Solheim, J. C., Harris, M. R., Kindle, C. S., andHansen, T. H. (1997) Prominence of b2-microglobu-lin, class 1 heavy chain conformation, and tapasin inthe interactions of class 1 heavy chain with calreticulinand the transporter associated with antigen processing.J. Immunol. 158, 2236–2241.

42. Cazzola, M., Foglieni, B., Bergamaschi, G., Levi, S.,Lazzarino, M., and Arosio, P. (2002) A novel deletionof the l-ferritin iron-responsive element responsiblefor severe hereditary hyperferritinaemia–cataract syn-drome. Br. J. Haematol. 116, 667–670.

43. Brooks, D. G., Manova-Todorova, K., Farmer, J.,Lobmayr, L., Wilson, R. B., Eagle, R. C., Jr., St Pierre,T. G., and Stambolian, D. (2002) Ferritin crystal cat-aracts in hereditary hyperferritinemia cataract syn-drome. Invest. Ophthalmol. Vis. Sci. 43, 1121–1126.

44. Aguilar-Martinez, P., Biron, C., Masmejean, C., Jean-jean, P., and Schved, J. F. (1996) A novel mutation inthe iron responsive element of ferritin L-subunit geneas a cause for hereditary hyperferritinemia-cataractsyndrome. Blood 88, 1895.

45. Cicilano, M., Zecchina, G., Roetto, A., Bosio, S.,Infelise, V., Stefani, S., Mazza, U., and Camaschella,C. (1999) Recurrent mutations in the iron regulatoryelement of L-ferritin in hereditary hyperferritinemia-cataract syndrome. Haematologica 84, 489–492.

46. Curtis, A. R. J., Fey, C., Morris, C. M., Bindoff, L. A.,Ince, P. G., Chinnery, P. F., Coulthard, A., Jackson,M. J., Jackson, A. P., McHale, D. P., Hay, D., Barker,W. A., Markham, A. F., Bates, D., Curtis, A., andBurn, J. (2001) Mutation in the gene encoding ferritinlight polypeptide causes dominant adult-onset basalganglia disease. Nature Genet. 28, 350–354.

47. Ferreira, C., Santambrogio, P., Martin, M. E., An-drieu, V., Feldmann, G., Henin, D., and Beaumont, C.(2001) H ferritin knockout mice: A model of hyper-ferritinemia in the absence of iron overload. Blood 98,525–532.

48. Kato, J., Fujikawa, K., Kanda, M., Fukuda, N., Sasaki,K., Takayama, T., Kobune, M., Takada, K., Takimoto,R., Hamada, H., Ikeda, T., and Niitsu, Y. (2001) Amutation in the iron-responsive element of H ferritinmRNA, causing autosomal dominant iron overload.Am. J. Hum. Genet. 69, 191–197.

49. LaVaute, T., Smith, S., Cooperman, S., Iwai, K.,Land, W., Meyron-Holtz, E., Drake, S. K., Miller, G.,Abu-Asab, M., Tsokos, M., Switzer, R., Grinberg, A.,Love, P., Tresser, N., and Rouault, T. A. (2001) Tar-geted deletion of the gene encoding iron regulatoryprotein-2 causes misregulation of iron metabolism andneurodegenerative disease in mice. Nature Genet. 27,209–214.

50. Lee, P. L., Gelbart, T., West, C., Halloran, C., Sipe,J. C., and Beutler, E. (2002) Mutations in iron-regu-latory protein 2 (IRP2) and lack of association withsporadic Parkinson’s disease. Mov. Disord. 17, 1302–1304.

51. Langlois, M. R., Martin, M. E., Boelaert, J. R., Beau-mont, C., Taes, Y. E., De Buyzere, M. L., Bernard,D. R., Neels, H. M., and Delanghe, J. R. (2000) Thehaptoglobin 2–2 phenotype affects serum markers ofiron status in healthy males. Clin. Chem. 46, 1619–1625.

52. Delanghe, J. R. and Langlois, M. R. (2002) Haptoglo-bin polymorphism and body iron stores. Clin. Chem.Lab. Med. 40, 212–216.

53. Van Vlierberghe, H., Langlois, M., Delanghe, J.,Horsmans, Y., Michielsen, P., Henrion, J., Cartuyvels,R., Billiet, J., De Vos, M., and Leroux-Roels, G.(2001) Haptoglobin phenotype 2–2 overrepresentationin Cys282Tyr hemochromatotic patients. J. Hepatol.35, 707–711.

54. Beutler, E., Gelbart, T., and Lee, P. L. (2002) Hapto-globin polymorphism and iron homeostasis. Clin.Chem.

55. Kawashima, A., Oda, Y., Yachie, A., Koizumi, S., and


486

Nakanishi, I. (2002) Heme oxygenase-1 deficiency:The first autopsy case. Hum. Pathol. 33, 125–130.

56. Yachie, A., Niida, Y., Wada, T., Igarashi, N., Kaneda,H., Toma, T., Ohta, K., Kasahara, Y., and Koizumi, S.(1999) Oxidative stress causes enhanced endothelialcell injury in human heme oxygenase-1 deficiency.J. Clin. Invest. 103, 129–135.

57. Ferris, C. D., Jaffrey, S. R., Sawa, A., Takahashi, M.,Brady, S. D., Barrow, R. K., Tysoe, S. A., Wolosker,H., Baranano, D. E., Dore, S., Poss, K. D., and Snyder,S. H. (1999) Haem oxygenase-1 prevents cell death byregulating cellular iron. Nat. Cell Biol 1, 152–157.

58. Miller, L. L., Miller, S. C., Torti, S. V., Tsuji, Y., andTorti, F. M. (1991) Iron-independent induction offerritin H chain by tumor necrosis factor. Proc. Natl.Acad. Sci. USA 88, 4946–4950.

59. Warren, S., Torti, S. V., and Torti, F. M. (1993) Therole of iron in the cytotoxicity of tumor necrosisfactor. Lymph. Cytokine Res. 12, 75–80.

60. Kwak, E. L., Larochelle, D. A., Beaumont, C., Torti,S. V., and Torti, F. M. (1995) Role for NF-kappa B inthe regulation of ferritin H by tumor necrosis factor-alpha. J. Biol. Chem. 270, 15285–15293.

61. Tsuji, Y., Miller, L. L., Miller, S. C., Torti, S. V., andTorti, F. M. (1991) Tumor necrosis factor-alpha andinterleukin 1-alpha regulate transferrin receptor in hu-man diploid fibroblasts. Relationship to the inductionof ferritin heavy chain. J Biol Chem 266, 7257–7261.

62. Alvarez-Hernandez, X., Liceaga, J., McKay, I. C., andBrock, J. H. (1989) Induction of hypoferremia andmodulation of macrophage iron metabolism by tumornecrosis factor. Lab. Invest. 61, 319–322.

63. Scaccabarozzi, A., Arosio, P., Weiss, G., Valenti, L.,Dongiovanni, P., Fracanzani, A. L., Mattioli, M., Levi,S., Fiorelli, G., and Fargion, S. (2000) Relationshipbetween TNF-alpha and iron metabolism in differen-tiating human monocytic THP-1 cells. Br. J. Haema-tol. 110, 978–984.

64. Gordeuk, V. R., Ballou, S., Lozanski, G., and Britten-ham, G. M. (1992) Decreased concentrations of tumornecrosis factor-a in supernatants of monocytes fromhomozygotes for hereditary hemochromatosis. Blood79, 1855–1860.

65. Fargion, S., Valenti, L., Dongiovanni, P., Scac-

cabarozzi, A., Fracanzani, A. L., Taioli, E., Mattioli,M., Sampietro, M., and Fiorelli, G. (2001) Tumornecrosis factor alpha promoter polymorphisms influ-ence the phenotypic expression of hereditary hemo-chromatosis. Blood 97, 3707–3712.

66. Beutler, E. and Gelbart, T. (2002) Tumor necrosisfactor a promoter polymorphisms and liver abnormal-ities of homozygotes for the c.845 G6A (C282Y)hereditary hemochromatosis mutation. Blood 100,2268–2269.

67. Baribault, H., Price, J., Miyai, K., and Oshima, R. G.(1993) Mid-gestational lethality in mice lacking ker-atin 8. Genes Dev. 7, 1191–1202.

68. Ku, N.-O., Gish, R., Wright, T. L., and Omary, M. B.(2001) Keratin 8 mutations in patients with crypto-genic liver disease. N. Engl. J. Med. 344, 1580–1587.

69. Ku, N. O., Wright, T. L., Terrault, N. A., Gish, R., andOmary, M. B. (1997) Mutation of human keratin 18 inassociation with cryptogenic cirrhosis. J. Clin. Invest.99, 19–23.

70. Sheldon, J. H. (1935) Haemochromatosis. OxfordUniversity Press, London.

71. Whitelaw, E., and Martin, D. I. (2001) Retrotrans-posons as epigenetic mediators of phenotypic varia-tion in mammals. Nature Genet. 27, 361–365.

72. Rosendaal, F. R. (1999) Venous thrombosis: a multi-causal disease. Lancet 353, 1167–1173.

73. Beutler, E., Turner, E., and Kuhl, W. (1981) The effectof alpha-thalassemia on the expression of the beta-thalassemia/HPFH heterozygote in a black family.Blood 57, 1132–1134.

74. Kaplan, M., Renbaum, P., Levy-Lahad, E., Hammer-man, C., Lahad, A., and Beutler, E. (1997) Gilbertsyndrome and glucose-6-phosphate dehydrogenasedeficiency: A dose-dependent genetic interaction cru-cial to neonatal hyperbilirubinemia. Proc. Natl. Acad.Sci. USA 94, 12128–12132.

75. Roetto, A., Totaro, A., Cazzola, M., Cicilano, M.,Bosio, S., D’Ascola, G., Carella, M., Zelante, L.,Kelly, A. L., Cox, T. M., Gasparini, P., and Cam-aschella, C. (1999) Juvenile hemochromatosis locusmaps to chromosome 1q. Am. J. Hum. Genet. 64,1388–1393.


doi:10.1006/bcmd.2002.0586

487

seeking candidate mutations that affect iron homeostasis

Documents