Introduction

Hereditary HFE-associated haemochromatosis is a disorder of iron metabolism, which in many individuals causes an inappropriately high intestinal absorption of dietary iron and induces progressive body iron overload [1]. The excess iron is deposited in parenchymal organs including the liver, heart and pancreas and may lead to liver cirrhosis, hepatocellular carcinoma, diabetes mellitus and cardiomyopathy [2].

The discovery of the HFE gene mutations or variants in 1996 [3] had great impact on clarification of the pathophysiology and diagnostic approach of haemochromatosis. More than 90% of patients of northern European ancestry with clinical haemochromatosis are homozygous for the C282Y variant, which is a G→A transition at nucleotide 845 in the open reading frame [4, 5].

Based on phenotypic screening (serum transferrin saturation, serum ferritin) the prevalence of haemochromatosis in Denmark has been estimated to 0.37–0.46% [6], i.e. between 1 in 217 and 1 in 270 persons of ethnic Danish origin.

Previous studies of ethnic Danes have reported a C282Y variant frequency of 5.7% [7], which is close to the estimated allele frequency of ~6.1% derived from phenotypic screening [6]. Consequently, the phenotypic/genotypic haemochromatosis allele ratio in Denmark is close to 1 [8].

HFE haemochromatosis is inherited as an autosomal recessive trait [9]. Since the identification of the HFE gene and its variants, hereditary haemochromatosis is emerging as one of the most common genetic disorders among populations of northern European descent [10]. The purpose of the present study was twofold. Firstly, to describe the frequency of the three most common HFE variants C282Y (c845G→A), H63D (c187C→G) and S65C (c193A→T) in a population of ethnic Danish men. Secondly, to assess the penetrance of the variants on the occurrence of biochemical and clinical haemochromatosis. In this paper we describe the frequencies of the HFE variants, additional results will be published separately.

Materials and methods

Participants

This prospective, epidemiologic population study was performed in 2000–2001. It was approved by The Danish Scientific Ethical Committee and fulfilled The Declaration of Helsinki. All men of ethnic Danish origin aged 30–50 years (n = 10,993) being residents in the municipalities of Næstved and Vordingborg in the southern part of Zealand were invited by letter to participate. Written informed consent was obtained from all participants.

HFE variant analyses

Saliva samples for analysis of HFE genotype were taken by the participants and sent by mail to our laboratory. Samples were obtained from 6,567/10,993 (60%) participants. In 1,064 subjects, the genotype could not be determined on saliva. They were invited to have a blood sample drawn for genetic testing, which was obtained in 517 subjects. Thus the HFE genotype was assessed in 6,020/10,993 (55%) of the population initially being invited.

Extraction of DNA from saliva and blood was performed using the method of Rudbeck and Dissing [11] with minor modifications. The stop-mixture contained cresol-red in order to ensure a visual verification, when the DNA sample was transferred to the polymerase chain reaction (PCR) tube. Cresol-red did not interfere with the PCR [12]. For each PCR, 4 µl DNA extract was used as template in a 50 µl reaction mixture in the presence of 0.5 µM of each primer, 40 µM of each dNTP, 75 mM Tris (pH 8.8), 20 mM (NH4)2SO4, 0.01% tween 20, 1.5 mM MgCl2, and 0.02 U Taq polymerase (Fermentas UAB, Lithuania). Two primer sets were used: C282Ys (5′-CCCCCAGAACATCACCATGAA)/C282Yas (5′-CCCCTAACAAAGAGCAGATCC) and H63D/S65Cs (5′-GTTGCTCTGTCTCCAGGATCACACTCTC)/H63D/S65Cas (5′-TCCCACCCTTTCAGACTCTGTCTCAG). The HFE H63D/S65C primers included internal control restriction enzyme sites for each of the enzymes used for this assay (underscore). Cycling parameters for the HFE H63D/S65C PCR were 94°C initial denaturation for 5 min, 94°C DNA denaturation for 60 s, 54°C primer annealing for 60 s, 72°C primer extension for 60 s, carried out for 35 cycles, and finally a prolonged DNA extension at 72°C for 5 min to complete the reaction. Cycling parameters for the HFE 845 PCR were 94°C initial denaturation for 5 min, 94°C for 40 s, 55°C for 60 s, 72°C for 60 s, carried out for 5 cycles, 94°C for 40 s, 56°C for 40 s, 72°C for 40 s, carried out for 30 cycles, and finally a prolonged DNA extension at 72°C for 5 min to complete the reaction. The amplified PCR fragments were digested with restriction endonucleases RsaI (C282Y) and Bsp143I and HinfI identified the H63D and S65C variants, respectively. Digest was performed according to the manufacturer’s (Fermentas UAB, Lithuania) recommendation. The digested products were electrophoresed and visualised after ethidium bromide staining.

Three cases of HFE 845A and 187G in “cis” phase have been reported [13–15]. Such a genotype was not found in our study. We consider it a rare exception and assume that all chromosomes contain only one of the three variants. The most common HFE genotypes C282, H63, and S65 are designated wild type (wt) in this paper.

Statistics

Hardy–Weinberg equilibrium in the observed genotype distribution was assessed with χ2-goodness-of-fit test. The significance level was set at 0.05. Differences in HFE variant frequencies were tested with χ2-test.

Results

The HFE genotype distribution in this population was consistent with the Hardy-Weinberg equilibrium (χ2 = 5.4, p = 0.8).

Table 1 shows the number and composition of the HFE genotypes. The allele frequency of the C282Y, H63D and S65C variants was 5.6%, 12.8% and 1.8%, respectively. In the entire series, we observed 1.4% H63D/C282Y, 0.1% S65C/C282Y and 0.4% H63D/S65C compound heterozygotes.

Table 1 Distribution of HFE genotypes in ethnic Danish men
Full size table

Discussion

Following the discovery of the HFE gene mutations in 1996 [3], many genetic population surveys have been reported [16]. The frequency of the C282Y variant is highest among people in Northern Europe, comprising Ireland, UK and Scandinavia [16]. The frequency displays a decreasing gradient from Northern to Southern Europe and from Western to Eastern Europe [16]. The prevalence of homozygosity for the C282Y variant in a series of ethnic Danish patients with clinical genetic haemochromatosis was 96% [17], stressing the importance of this variant in hereditary iron overloading disorders in subjects of Danish heritage. In Danes, there is good agreement between the genetic frequency of C282Y homozygosity of 0.36% in the present study and the phenotypic (using serum transferrin saturation and serum ferritin) frequency of ~0.37% in male blood donors [6].

Genetic studies of the HFE variants in Danes [18–21] are shown in Table 2. The study of Merryweather-Clarke et al. [18], which was performed on Guthrie cards from Danish newborns, differ from the other Danish studies [7, 19–22]. In the total series of Danes (n = 19,170) the frequency of the C282Y variant was 5.6%. The results of the present study were not significantly different from other Danish studies [7, 19–22].

Table 2 Observed HFE variant allele frequencies in Danish studies
Full size table

The allele frequencies of C282Y in the other Nordic countries [18, 23–27] are shown in Table 3 and the cumulated frequencies in Table 4. The frequency of the C282Y variant in the total series of Danes (n = 19,170) was not significantly different from the frequency in Iceland and Sweden, but significantly lower than the frequency in the Faeroe Islands (p < 0.05) and Norway (p < 0.01). The C282Y frequency was significantly higher in Danes than in Finns (p < 0.001) [26], who have the lowest frequency among the Nordic countries. However, the Finnish people are not of Nordic origin, but related to the Hungarian people, who originally came from the Ural region.

Table 3 Observed HFE variant allele frequencies in the Nordic countries
Full size table
Table 4 Cumulated HFE variant allele frequencies in the Nordic Countries
Full size table

In the total series of Danes (n = 19,170) the frequency of the H63D variant was 12.8%. The present result, derived from a male population (n = 6,020) was similar to previous Danish studies (Table 2). The allele frequency of the H63D variant in Danes was not significantly different from that reported in Iceland, Norway and Sweden, but significantly lower than the frequency in the Faeroe Islands (p < 0.01).

The S65C frequency in the total series of Danes (n = 7,039) was not significantly different from that reported in Sweden and Finland, but significantly lower than the frequency in the Faeroe Islands (p < 0.05).

The frequency of C282Y/H63D compound heterozygotes (1.44%) was similar to the frequency of 1.42% found in the previous study from Copenhagen [21] and not significantly different from the frequency of 1.72% in a series of Danish patients with clinical haemochromatosis initially diagnosed by phenotypic methods [14].

We have hypothesised that the C282Y mutation originated among the population in Southern Scandinavia or Northern Germany and was spread to other parts of Northern Europe by the Vikings in the period 700–1050 [16]. Around the turn of the first millennium, Scandinavian Vikings became marine emigrants, who conquered coastal country and established settlements nearby the sea. The highest frequencies of the C282Y variant are found in the coastal areas of Northern Europe. The Viking influence was dominant in Denmark, Iceland, the Faeroe Islands, Norway, Sweden as well as the Aaland Islands and prominent in Scotland, England, Ireland and Northern France.

In hereditary HFE C282Y haemochromatosis, women present with clinical symptoms at a higher age than men [28], probably because they are partly protected against iron overload due physiological iron losses with menstruation and pregnancy [29]. Men have significantly smaller physiological iron losses than women and a higher intake of dietary iron, alcohol and red meat [11, 12, 30–33]. Lifestyle factors, e.g. alcohol intake and red meat consumption tend to increase serum ferritin levels and body iron stores and consequently have an enhancing effect on the phenotypic frequency by increasing the biochemical penetrance i.e. elevated serum transferrin saturation and serum ferritin as well as the clinical penetrance, i.e. organ damage.

Hereditary haemochromatosis is the most prevalent hereditary disorder in people of northern European heritage. However, the biochemical and clinical penetrance of the C282Y variant in its homozygous form appears to be incomplete and display considerable variation between different populations [30–34]. This in part explains the discrepancy between the high frequency of C282Y homozygosity and the low incidence of clinical haemochromatosis. In Denmark, the expected frequency of C282Y homozygosity is 0.31%. The ethnic population in Denmark is ~5,000,000 persons, i.e. there are ~16,000 C282Y homozygous individuals. Yet in the period 1950–1985, less than 200 patients with clinical haemochromatosis were diagnosed in Denmark [2]. The situation is similar in the Faeroe Islands. The C282Y variant frequency is 8.0% and the frequency of homozygosity 0.64%. The Faeroese population consists of ~40,000 persons, i.e. there should be ~260 C282Y homozygous individuals. However, only a few families with hereditary haemochromatosis have been identified [35]. Apparently, the incomplete penetrance of C282Y homozygosity decreases the clinical impact of this variant.

In conclusion, in the Danish population, the HFE gene variants C282Y, H63D and S65C are quite common and the genetic predisposition for haemochromatosis is therefore relatively high. In daily clinical practice, the diagnostic awareness should be focused on subjects at risk for genetic iron overload by widespread assessment of biochemical iron status. Furthermore, relatives to probands with hereditary haemochromatosis should be tested by genetic methods for HFE variants and if found positive also tested by biochemical methods in order to clarify their lifetime risk of iron overload disease.