Since the discovery of the HFE gene and the C282Y mutation (Type 1 haemochromatosis), new genes involved in iron metabolism have been described. Juvenile haemochromatosis has been related to HFE2 and HAMP mutations (Type 2A and 2B) and is described as severe iron overload affecting patients before the age of 30 years (Brissot et al, 2011). Mutations in the TFR2 gene lead to Type 3 haemochromatosis whose clinical picture mimics Type 1. However, rare cases affecting young patients have been reported (Brissot et al, 2011). The ferroportin disease has been linked to SLC40A1 mutation and is described as iron overload affecting patients at any age (Le Lan et al, 2011). The basic mechanism accounting for iron overload in types 1, 2 and 3 haemochromatosis is decreased hepcidin synthesis. TFR2 is also involved in hepcidin synthesis regulation (Wallace et al, 2005; Gao et al, 2010) but its definite mode of action remains to be determined. In Type 1 haemochromatosis, cofactors play an important role because clinical penetrance of C282Y homozygosity is low (Allen et al, 2008). In Type 3 haemochromatosis, clinical expression seems high but reported cases are too scarce to definitely assess penetrance. Here, we report seven new cases of Type 3 haemochromatosis. Transferrin receptor 2 mutations were screened as part of the diagnostic activity of the French Reference Centre for Rare Iron Overload Diseases of Genetic Origin and the associated network of expertise Centres. Written informed consent was obtained and the study was performed in accordance with the Declaration of Helsinki and with the French regulations on medical genetic diagnosis. Patients were tested for HFE mutations (p.Cys282Tyr, p.His63Asp) (Jouanolle et al, 1996; Aguilar Martinez et al, 1997). The entire coding region and intronic flanking sequences of the TFR2 gene (NCBI NM_003227.3, NP_003218.2) were sequenced. To exclude other mutation(s), analyses of the haemochromatosis type 2 (juvenile) (HFE2), hepcidin (HAMP), and ferroportin (SLC40A1) coding sequences were performed. To determine the potential consequences of mutations on the protein, TFR2 amino acid sequence and mutations were input as required into the following algorithms: Scale-Invariant Feature Transform (SIFT; http://sift.jcvi.org/), Polymorphism Phenotyping v2 (Polyphen-2; http://genetics.bwh.harvard.edu/pph2/bgi.shtml), point mutant (Pmut; http://mmb2.pcb.ub.es:8080/PMut/), SNPs3D (http://www.snps3d.org/), and Scalable Nucleotide Alignment Program (SNAP; http://www.rostlab.org/services/snap/). Seven unrelated patients were diagnosed with Type 3 haemochromatosis. Three were homozygous for the previously undescribed p.Asn412Ile (c.1235A>T), p.Gly430Arg (c.1288G>A), p.Arg678Pro (c.2033G>C) TFR2 mutations. Consanguinity was likely only for Patient 1. Four patients were compound heterozygotes for at least one new TFR2 mutation each: p.Leu85_A a96delinsPro (c.254_286 + 9del), p.Met705Hisfs*87 (c.211 2dup), p.Arg730Cys (c.2188C>T), p.Gly735Ser (c.220 3G>A), p.Trp781* (c.2343G>A). One patient, who was a compound heterozygote for two previously described TFR2 mutations (p.Ala444Thr, p.Gly792Arg) (Lee & Barton, 2006; Biasiotto et al, 2008) was found to carry the p.Gly204Ser mutation in the SLC40A1 gene coding for ferroportin. The main clinical and biological features of the patients are summarized in Table 1. No mutations were found in the HAMP and HFE2 genes of the patients. Sequencing of the HFE gene revealed no other mutations than the H63D or C282Y (Table 1). p.Leu85_Ala96delinsPro+/- (p.Gly735Ser +/-) p.Met705Hisfs*87 +/- (p.Gly792Arg+/-) p.Ala444Thr +/- (p.Gly792Arg+/-) p.Arg730Cys +/- (p.Trp781+/-*) Patients 1 and 2 were diagnosed earlier than usually described for Type 3 haemochromatosis. Patient 1 was referred for major asthenia at the age of 10 years. Biological workup found high transferrin saturation leading to the diagnosis of haemochromatosis; an echocardiogram revealed no abnormalities. Patient 2 was diagnosed as a result of elevated transferrin saturation in the context of α1-antitrypsin deficiency. Patients 3, 4 and 5 were diagnosed between 20 and 30 years of age. Patient 3 originated from North Africa and was diagnosed upon arrival in France, which could explain the older age at diagnosis. The daughter of Patient 3 was heterozygous for the TFR2:p.Gly430Arg mutation and had normal iron parameters. The parents of Patient 4 were heterozygous for mutation p.Gly735Ser and p.Leu85_Ala96delinsPro respectively, and had normal iron parameters. Patient 6 was diagnosed with non-HFE related haemochromatosis at the age of 28 years. Diagnosis of ferroportin disease was made later by finding the SLC40A1:p.Gly204Ser mutation. The clinical presentation was unusual with elevated transferrin saturation and arthropathy. Moreover, phlebotomies were very well tolerated and removed 19·5 g iron. For these reasons, sequencing of other genes related to iron metabolism was performed and revealed two already described mutations in TFR2. Patient 7 was diagnosed during the diagnostic workup of post-partum infection. She had elevated liver enzymes, and transient elastography (Fibroscan®, Echosens, Paris, France) indicated liver cirrhosis (16·9 kPa). She underwent weekly phlebotomies. Oral chelation was added due to poor compliance, but had to be stopped because of side effects. None of the patients had overt clinical cardiomyopathy. Family studies in available relatives confirmed that all mutations were inherited in trans. Evaluation of the potential consequences of mutations at the protein level was performed in silico. Results are reported in Table 2. Mutations were predicted to be deleterious, expect for the p.Ala444Thr. However, this variant was not described in the exome variant server, thus excluding a rare variant; moreover the clinical picture clearly advocated for a deleterious impact. All mutations were located in the putative extra-cellular domain of the protein. The p.Arg678Pro mutation is of special interest because of its location on a potential site of interaction with transferrin: the RGD domain at position 678 of TFR2 corresponds to the RGD domain of TFR1 which has been shown to be required for transferrin binding to TFR1 (Dubljevic et al, 1999). Our data further document the clinical description of this disease. Six out of seven patients described here were diagnosed before the age of 30 years, two of them were aged less than 18 years. Of note, the earliest diagnosis, at the age of 10 years, was related to a mutation of the putative fixation site of transferrin on TFR2. Considering that TFR2 is thought to play an iron sensor role by iron-transferrin sensing, mutations hampering this critical mechanism may lead to a more severe phenotype. It is therefore possible that the phenotypic expression of Type 3 haemochromatosis is a spectrum with variable severity and age of onset, depending on the phenotypic impact of the mutation. Accordingly, we confirm that the search for TFR2 mutations should be part of the screening process in case of adult as well as juvenile haemochromatosis.