Mutant antimicrobial peptide hepcidin is associated with severe juvenile hemochromatosis

brief communications

Mutant antimicrobial peptide hepcidin is associated with severe juvenile hemochromatosis

Published online 9 December 2002; doi:10.1038/ng1053

Animal models indicate that the antimicrobial peptide hepcidin (HAMP; OMIM 606464) is probably a key regulator of iron absorption in mammals. Here we report the identification of two mutations (93delG and 166C→T) in HAMP on 19q13 in two families with a new type of juvenile hemochromatosis.

Animal models of iron overload include mice deficient in Usf2 that do not express the antimicrobial peptide Hamp10. Constitutive overexpression of hepcidin in Hamp transgenic mice leads to iron-deficient anemia11. These findings indicate a key role for HAMP in regulation of iron absorption in mammals and make HAMP a functional candidate for association with juvenile hereditary hemochromatosis that is not linked to 1q.

HAMP is a pro-peptide of 84 amino acids that undergoes enzymatic cleavage into mature peptides of 20, 22 and 25 amino acids12,13. Active peptides are rich in cysteines that form intramolecular bonds and stabilize the β-sheet structure14. The peptide of 25 amino acids, isolated from human blood12 and urine13, has prevalent hepatic expression14.

To determine whether HAMP was associated with juvenile hereditary hemochromatosis not linked to 1q, we genotyped affected individuals from two families. All affected individuals were less than 30 years old at the onset of clinical symptoms and

family 2

c

normal

homozygous

166C→T

(proband)

heterozygous 166C→T

e

© 2003 Nature Publishing Group http://www.nature.com/naturegenetics

The regulation of intestinal iron absorption is crucial to avoid toxicity. Disruption of this regulation in hereditary hemochromatosis leads to iron overload, cirrhosis, cardiomyopathy, arthritis and endocrine failure. The most common form of hereditary hemochromatosis (OMIM 235200) is caused by mutations in the gene HFE1, which may encode a duodenal crypt body iron sensor. Type 3 hereditary hemochromatosis (OMIM 604250) is associated with mutations of a low-affinity transferrin

a

D19S425 2 1 22
D10S208 1 2 1 2
93delG + wt + wt
D19S228 11 1 2

I

3rd cousins

1

2

II1 23

D19S425 1222 22 D10S208 2211 11 93delG wtwt + + ++ D19S228 12 11 11

Fig. 1 Detection of mutations in HAMP in families with juvenile hemochromatosis not linked to 1q. a, Pedigree of family I and 19q13 haplotype analysis. Roman numerals, generations; arabic numerals, individuals; filled symbols, affected individuals; open symbols, unaffected individuals; 1,2, different size alleles; +, 93delG deletion present; wt, wild-type. Regions of homozygosity are boxed. b, Sequence chromatographs of the HAMP gene region spanning the 93delG deletion (reverse sequence shown) from the indicated individuals. The asterisk indicates the position of the deletion in the 93delG heterozygote followed by the resulting allelic slippage. c, Sequence chromatographs of the HAMP gene region spanning the 166C→T mutation (forward sequence shown) from the indicated individuals. d, Single-strand conformation polymorphism pattern of amplified exon

2. Lane 1, molecular weight marker; lane 2, 93delG homozygote; lane 3, heterozygote; lanes 4–14, normal controls. e, Family segregation of 166C→T as analyzed by HphI digestion. The undigested fragment is 352 bp. Two fragments (257 and 95 bp) are produced from the mutated allele. P, proband; M, mother; S1 and S2, heterozygous siblings; MWM, molecular weight marker.

receptor (TFR2; refs. 2,3) and type 4 hereditary hemochromatosis (OMIM 606069) associated with the iron exporter ferroportin 1 (refs. 4,5). Juvenile or type 2 hereditary hemochromatosis (OMIM 602390) has the most severe phenotype and can be lethal at a young age6,7. Although the gene associated with this disorder has not been identified, the locus maps to chromosome 1q21 (ref. 8). A single inbred pedigree with juvenile hereditary hemochromatosis that is not linked to 1q was recently reported9.

b family 1
normal (individual II-1)
homozygous 93delG (proband II-2)
heterozygous 93delG (individual I-1)
d

nature genetics • volume 33 • january 2003

brief communications

© 2003 Nature Publishing Group http://www.nature.com/naturegenetics

Fig. 2 Comparison of the predicted mutant peptides resulting from the 93delG and 166C→T mutations with normal pro-hepcidin.

had severe iron overload with liver fibrosis or cirrhosis and hypogonadism6,9, meeting the diagnostic criteria for juvenile hemochromatosis7. One affected individual also had cardiomyopathy6.

We analyzed microsatellites encompassing a region of 2.7 cM on chromosome 19q13 in family I (Fig. 1a) and identified a region of homozygosity identical in both probands. We then sequenced the HAMP coding region, exon-intron boundaries and 5′ and 3′ untranslated regions in probands of family I and in a previously reported affected individualin family II and identified two mutations. The first was deletion of a guanine in exon 2 at position 93 of HAMP cDNA (93delG). Probands of family I were homozygous with respect to this deletion, and obligate carriers were heterozygous (Fig. 1a,b). Single-strand conformation polymorphism analysis excluded the mutation in 50 unaffected controls (Fig. 1d). The 93delG deletion results in a frameshift, and, if mutated RNA reaches translation, generates an elongated (179 residues) abnormal pro-hepcidin peptide. Because the frameshift occurs after residue 31, the active peptides and the cysteine motif are completely disordered.

The second mutation, which we identified in the proband of family II, was a C→T substitution at position 166 in exon 3 of HAMP cDNA (166C→T), which changes arginine at position 56 to a stop codon (R56X; Fig. 1c). HphI restriction analysis showed the expected intrafamilial segregation of 166C→T (Fig. 1e) and its absence in 50 controls (data not shown). The R56X amino-acid change occurs in a penta-arginine (residues 55–59) basic domain, which is probably the recognition site for pro-hormone convertases12, and produces a truncated pro-hepcidin lacking all mature peptide sequences (Fig. 2).

The iron overload resulting from mutations in HAMP suggests that HAMP plays a role in maintaining iron balance in humans, and adds HAMP to the list of genes associated with hereditary hemochromatosis. The severity of the phenotype of juvenile hemochromatosis relative to the other types of hereditary hemochromatosissuggests that HAMP may very well be a principal component of the iron regulatory machinery. This conclusion is supported by the fact that wild-type HFE and TFR2 in individuals with mutations in HAMP are unable to inhibit iron absorption. Our results are in agreement with recent data on hypotransferrinemic mice indicating that Hamp may mediate the regulation of iron absorption according to the needs of both erythropoeiesis and body iron stores15. Hemochromatosis associated with HAMP inactivation is the first example of a genetic disorder associated with an antimicrobial pep-tide, but the apparent lack of susceptibility to infections in affected individuals suggests that the antimicrobial role of HAMP is not critical for staving off infection.

Our results identify a new form of juvenile hereditary hemochromatosis, increasing the genetic heterogeneity of iron-overload diseases. They may also facilitate identification of the gene on 1q that is associated with juvenile hereditary hemochromatosis, as it may encode a molecule involved in the HAMP signaling pathway.

Acknowledgments

We thank D. Mavrogianni, I. Jibreel, F. Daraio and P. Porporato for technical support. We also thank the individuals with hemochromatosis and their families for their participation. This work was supported by grants from Telethon ONLUS Foundation, European Community and Italian Ministry of Instruction and University (to C.C.) and from the University of Athens (to D.L.).

Competing interests statement

The authors declare that they have no competing financial interests.

Antonella Roetto1*, George Papanikolaou2*, Marianna Politou2, Federica Alberti1, Domenico Girelli3, John Christakis4, Dimitris Loukopoulos& Clara Camaschella1

*These authors contributed equally to this work.

Department of Hematology, Theagenio Cancer Center, Thessaloniki, Greece. Correspondence should be addressed to C.C. (e-mail: [email protected]).

Received 7 October; accepted 29 October 2002.

  1. Feder, J.N. et al. Nat. Genet. 13, 399–408 (1996).
  2. Camaschella, C. et al. Nat. Genet. 25, 14–15 (2000).
  3. Roetto, A. et al. Blood 97, 2555–2560 (2001).
  4. Montosi, G. et al. J. Clin. Invest. 108, 619–623 (2001).
  5. Njajou, O.T. et al. Nat. Genet. 28, 213–214 (2001).
  6. Camaschella, C. et al. Eur. J. Hum. Genet. 5, 371–375 (1997).
  7. De Gobbi, M. et al. Br. J. Haematol. 117, 973–979 (2002).
  8. Roetto, A. et al. Am. J. Hum. Genet. 64, 1388–1393 (1999).
  9. Papanikolaou, G. et al. Blood Cells Mol. Dis. 29, 168–173 (2002).
  10. Nicolas, G. et al. Proc. Natl. Acad. Sci. USA 98, 8780–8785 (2001).
  11. Nicolas, G. et al. Proc. Natl. Acad. Sci. USA 99, 4596–4601 (2002).
  12. Park, C.H., Valore, E.V., Waring, A.J. & Ganz, T. J. Biol. Chem. 276, 7806–7810 (2001).
  13. Krause, A. et al. FEBS Lett. 480, 147–150 (2000).
  14. Pigeon, C. et al. J. Biol. Chem. 276, 7811–7819 (2001).
  15. Weinstein, D.A. et al. Blood 100, 3776–3781 (2002).

22

nature genetics • volume 33 • january 2003