Heme carrier protein 1 (HCP1) genetic variants in the Hemochromatosis and Iron Overload Screening (HEIRS) Study participants

Abstract

Heme carrier protein 1 (HCP1) has been identified as a possible heme carrier by in vitro analysis. To determine the association of mutations within the HCP1 gene with iron phenotypes, we examined the entire coding region of the HCP1 gene in 788 US and Canadian participants selected from the Hemochromatosis and Iron Overload Screening (HEIRS) Study using denaturing high-performance liquid chromatography. We sequenced the exon and flanking intronic regions if variants were detected. We tested 298 non-C282Y homozygotes from four racial/ethnic backgrounds (White, Black, Asian, and Hispanic) selected because they had high serum ferritin (SF) and transferrin saturations (TS). As controls, we chose 300 other random participants of the same racial/ethnic backgrounds from the same geographic locations. From the 333 HEIRS Study C282Y homozygotes, we selected 75 based on high SF and TS, 75 based on low SF and TS; 75 were selected randomly as controls. Thirty-five of the randomly selected C282Y homozygotes were also included in the high and the low SF and TS groups due to numerical limitations. We identified eight different HCP1 genetic variants; each occurred in a heterozygous state. Except one, each was found in a single HEIRS Study participant. Thus, HCP1 variants are infrequent in the populations that we tested. Five HEIRS Study participants had non-synonymous, coding region HCP1 variants. Each of these five had TS above the 84^th gender- and ethnic/racial group-specific percentile (TS percentiles: 84.7, 91.3, 97.9, 99.5, and 99.9).

INTRODUCTION

Increased iron absorption is a primary defect in hereditary hemochromatosis. The common mutation C282Y of the HFE gene on chromosome 6p occurs almost exclusively in Whites of European descent. Homozygosity for the C282Y mutation of the HFE gene on chromosome 6p accounts for most cases of hereditary hemochromatosis in this group [1;2]. H63D is another common HFE polymorphism that is found in most race/ethnicity groups. Iron overload occurs in some persons who are compound heterozygotes for C282Y and H63D and rarely in individuals without either H63D or C282Y mutations [1;2]. Other genes or mutations within HFE, could account for the marked variability in iron overload among HFE C282Y homozygotes [2–4;4–12]. No gene or mutation that accounts for a significant proportion of primary iron overload cases in non-Whites has been reported.

Most dietary iron occurs as either inorganic or heme moieties. Divalent metal ion transporter 1 (DMT1) facilitates the uptake of dietary non-heme iron across the microvillous membranes of absorptive enterocytes, but the transport system for heme iron absorption has remained elusive. Recently, Shayeghi and colleagues [13] reported discovery of a gene they called heme carrier protein 1 (HCP1; MGC9564) whose expression appeared to facilitate heme absorption in the duodenal brush border membrane. HCP1 expression appears to be regulated by a post-translational mechanism that responds to changes in body iron stores. Andrews suggested that HCP1 might be a major mediator of heme iron absorption by intestinal enterocytes, and that mutations within HCP1 may influence body iron stores, either independently or as a modifier of clinical expression in individuals with other genetic configurations such as HFE C282Y homozygosity [14].

To test whether mutations in the HCP1 gene influence serum iron measures either in those without HFE mutations or in HFE C282Y homozygotes, we used denaturing high-performance liquid chromatography (DHPLC) to screen the entire coding region of HCP1 for mutations in 788 participants from the Hemochromatosis and Iron Overload Screening (HEIRS) Study [2;7].

METHODS

Study Subjects

All subjects were participants in the Hemochromatosis and Iron Overload Screening (HEIRS) Study. A detailed description of this study’s design and its 101,168 participants’ general characteristics have been previously reported [4;15]. Each participant was classified into one of seven separate racial/ethnic groups based on self-report: White, Black, Asian, Hispanic (regardless of race reported), Pacific Islanders, American Indian, and multiple/unknown race/ethnicity. We included only Whites, Blacks, Asians, and Hispanics in the present HCP1 study, because the numbers of HEIRS Study subjects who reported Pacific Islander, Native American, or multiple/unknown racial/ethnic groups were relatively small. Blood samples were obtained from all HEIRS Study participants at initial screening without regard to fasting, for measurement of serum transferrin saturation (TS), serum ferritin (SF), and genotyping to detect the HFE missense mutations C282Y and H63D. The HEIRS Study participants tested for HCP1 gene mutations were chosen to achieve equal numbers for each study group (Table 1) and to satisfy budgetary constraints.

Table 1.

Characteristics of HEIRS Study participants studied by DHPLC.

Groups	Serum Ferritin (SF) (μg/L)				Transferrin Saturation (TS) (%)				HEIRS Population Pool
	Mean SF	SD	Range of SF		Mean TS	SD	Range of TS
Asian (cases) n=73	1,000.7	952.6	215	6,506	69.7	16.2	46	99
Asian (controls) n=75	246.3	202.0	8	970	32.2	10.8	6	63	n=13,130
Black (cases) n=75	1,602.7	1260.9	205	6,450	76.4	17.6	46	100
Black (controls) n=75	146.7	129.0	7.5	589	24.9	10.8	6	56	n=27,224
Hispanics (cases) n=75	763.5	721.2	205	3,800	62.6	15.9	46	99
Hispanics (controls) n=75	123.8	135.1	8	781	25.8	10.3	5	61	n=12,696
White (cases) n=75	1,081.8	1072.4	209	5,300	71.5	18.3	30	100
White (controls) n=75	160.3	218.1	8	1,650	28.5	13.9	5	100	n=44,808
C282Y homo/high iron n=75	1,065.4	970.4	40	5,200	90.8	12.6	38	100
C282Y homo/low iron n=75	171.7	217.5	8	840	41.3	19.0	4	91	n=333
C282Y homo/controls n=75	500.8	685.0	8	5,200	63.5	24.8	4	100

Selection of non-HFE C282Y homozygotes for DHPLC testing

We selected 75 non-C282Y homozygotes with very high SF and TS who reported either White, Black, Asian, or Hispanic race/ethnicity (Table 1). For comparison and to estimate gene frequency of HCP1 variants, we also randomly selected 75 “control” participants from each of these four racial/ethnic groups (Table 1).

High TS/SF, non-C282Y participants

We first converted all HEIRS Study participants’ TS and SF levels into gender-specific and HFE C282Y and H63D genotype-specific percentile values. We selected as cases for DHPLC testing the 75 participants within each racial/ethnic group who had the highest percentile for either TS or SF because we had no a priori reason to know whether HCP1 mutations would influence SF or TS more strongly. Thus, the high TS/SF case group included participants if they had either a very high TS percentile or a very high SF percentile.

Racial/ethnic groups “control” selection

We selected 75 participants at random from each of the racial/ethnic groups, but we constrained the number of controls selected from each field center to match each field center’s fraction of subjects in the high TS/SF cases groups.

Selection of HFE C282Y homozygotes for DHPLC testing

High TS/SF and low TS/SF C282Y homozygotes

From the 333 HFE C282Y homozygotes, we selected 75 with the highest SF and TS and another 75 with the lowest SF and TS using aforementioned gender- and field center-specific percentile-based transformation of the participants’ SF and TS concentrations, without regard to HFE genotype.

HFE C282Y homozygote controls

We selected a control group of 75 subjects from the 333 HEIRS Study HFE C282Y homozygotes randomly with no constraints on field center, gender, or race/ethnicity (Table 1). All those selected were non-Hispanic Whites. We did not exclude those chosen for either the high or low TS/SF C282Y homozygote case groups from also being selected for the random control group. Accordingly, 35 of the 75 participants selected to be random C282Y homozygote controls were simultaneously members of high and low TS/SF groups. Altogether, 190 different HFE C282Y homozygotes were screened for HCP1 variants.

Genotyping

DNA for DHPLC screening was either isolated directly from EDTA anticoagulated blood using the Puregene System (Gentra Systems, Minneapolis, MN) or by whole genome amplification with the GenomiPhi™ kit (GE-Amersham Biosciences Corp., Piscataway, NJ) from deproteinized buffy coat that had been spotted onto FTA® paper (Whatman, Clifton, NJ).

Primers were designed using IDT OligoAnalyzer (http://www.idtdna.com-/analyzer/Applications/OligoAnalyzer/) or Primer 3 (http://www-genome.wi.mit.edu/cgi-bin/primer/primer3_www.cgi) (Table 2). Each 50 μL PCR reaction contained 50–150 ng DNA, 200 μmol/L of each deoxynucleotide triphosphate, 10 mmol/L Tris-hydrochloride buffer (pH 8.3), 50 mmol/L potassium chloride, 2.5 mmol/L magnesium chloride, 1.5 U of HotstarTaq® DNA polymerase (QIAGEN, Valencia, CA), and 10–30 pmol of each primer. The reaction mixtures were brought to 94 °C for 15 min followed by 40 cycles of 94 °C for 30 sec, 30 sec at the appropriate annealing temperatures (Table 2), and an extension at 72 °C for 40 sec, followed by a final extension at 72 °C for 5 min using the PTC-225 Peltier thermal cycler (MJ Research, Inc., Waltham, MA).

Table 2.

Primers for HCP1 exons and PCR conditions

	Primers	PCR Tm	DMSO
Exon 1	Forward: CGCCGGACATTTAAGGAGGGA	63°C	4%
	Reverse: TCCAGTTACCCGCCACTACCATTA
Exon 2.1	Forward: CAGAGTGAGGAACCGCAC	68°C	No
	Reverse: TGCGACTGGAGCTGACATCTG
Exon 2.2	Forward: GTGGCCTTCTGGCTGCTA	65°C	No
	Reverse: GTCCCTGAGCCCACACAC
Exon 3	Forward: CCCCATTTTCCTGATGAGTG	62°C	No
	Reverse: GCAGCACTGGACATGAGG
Exon 4	Forward: CGCCCGCCGCCGCCCCAGAGTATGGTATTTAT	64°C	No
	Reverse: CATTATGAGCATGGGTTCAG
Exon 5	Forward: CCTGGATATTGTCCTCCAGC	64°C	No
	Reverse: GTCTAACGTATGGCAGGCAG

Mutation screening was performed on a Transgenomic WAVE® 3500 HT fragment analysis system equipped with a DNASep HT cartridge. Eight μL of PCR product was injected onto the column and eluted with a linear gradient of buffer A and buffer B obtained from the manufacturer (Transgenomic, Omaha, NE., USA). The eluted amplicons were detected at 260 nm with a deuterium lamp. All chromatograms were analyzed using WAVE Navigator™ software version 1.5.3. Testing of samples for the possible presence of homozygous mutations was done by adding a sequence-determined wild-type amplicon to the amplicons prior to injection.

All DNA samples that appeared to have a HCP1 variant were subjected to DNA sequencing using the same PCR amplicons used for DHPLC screening. Prior to sequencing, the DHPLC amplicons were purified using the Qiaquick® gel extraction kit according to the manufacturer’s instructions (Qiagen, Valencia CA.). PCR amplicons were sequenced in both directions using 3.2 pmole of the same PCR primers with Big-dye Terminator sequencing kits (Applied Biosystems, Foster City, CA, USA). All sequencing was performed at the University of Minnesota’s Advanced Genetic Analysis Center.

RESULTS

DNA samples were obtained from 790 HEIRS Study participants: 190 C282Y homozygotes and 600 with other HFE genotypes. Two of the 790 DNA samples could not be amplified by PCR. Among the 788 samples subjected to DHLPC screening of the HCP1 gene’s entire coding and partial intronic regions, we detected nine samples with HCP1 genetic variants (Table 3). Each was present in a heterozygous configuration. Each variant was confirmed by sequencing in both directions using the same PCR DHPLC primers. No HCP1 variants were found in the HFE C282Y homozygote high TS/SF group (Table 4). One HCP1 synonymous coding region genetic variant was found in the C282Y homozygote low TS/SF group, and two different missense mutations were found in the C282Y homozygote random control specimens (Table 4). All five participants with HCP1 coding region missense mutations had high TS, expressed either in absolute terms or as racial/ethnic group-, gender-, age-, field center-, and HFE genotype-specific percentile values. Two of these five participants had SF below the median (20.9^th and 39.4^th percentiles).

Table 3.

Distribution of HCP1 genetic variants in HEIRS Study participants

nm_080669.2 (HCP1 mRNA)	Racial/ethnic groups
Position and variation (Codon Frame)	Asian		African American		Hispanic		White
	High N=73*	Random N=75	High N=75	Random N=75	High N=75	Random N=75	High N=75	Random N=75
c.85 G>A (GTC > ATC) (p.V29I)	0	0	0	0	1	0	0	0
c.189 G>C (AGG > AGC) (p.R63S)	0	0	1	0	0	0	0	0
c.228+17 G>A (IVS1+17)	1	0	0	0	0	0	0	0
c.228+76 G>A (IVS1+76)	0	0	0	1	0	0	0	0
c.756 C>G (GTC > GTG) (p.V252V)	0	0	0	0	0	1	0	0
c.883 A>G (ACA > GCA) (p.T295A)	0	0	0	1	0	0	0	0
	C282Y homozygotes from the study population
	With highest TS/SF		With lowest TS/SF		Random selected homozygotes
	N=75		N=75		N=75
c.294 C>A (CTC > CTA) (p.L98L)	0		1		0
c.1366 C>T (CCC > TCC) (p.P456S)	0		0		2

^*Two samples had insufficient or poor quality DNA and could not be PCR amplified.

Table 4.

Serum iron parameters in HEIRS Study participants with heterozygous HCP1 variants.

HCP1 Variants	HFE C282Y status	Serum Ferritin (SF) (μg/L)	SF Percentile in HEIRS subgroups1	Transferrin Saturation (TS) (%)	TS Percentile in HEIRS subgroups1	Age group2	Gender3	Ethnic group	HEIRS subgroup
Non-synonymous
R63S	C/C	1380	99.9	96	99.9	40s	F	Black	case
T295A	C/C	589	92.0	56	99.5	40s	M	Black	control
V29I	C/C	368	97.7	51	91.3	50s	F	Hispanic	case
P456S	Y/Y	250	20.9	96	97.9	50s	M	White	control
P456S	Y/Y	172	39.4	59	84.7	50s	F	White	control
Intronic
c.228+17G>A	C/C	1450	99.3	78	99.8	60s	M	Asian	case
c.228+76G>A	C/C	192	82.9	25	58.3	50s	F	Black	control
Synonymous
L98L	Y/Y	131	9.5	36	1.9	60s	M	White	case
V252V	C/C4	79	15.6	39	68.6	30s	M	Hispanic	control

¹Age-, gender-, field center-, and HFE C282Y and H63D genotype-specific percentile for TS or SF within the HEIRS Study [4]

²To preserve HEIRS participants anonymity, the HEIRS Steering Committee and NHLBI Project Office allows only reporting of participants ages by decade, not specific ages.

³F=female; M=male

⁴This individual is HFE H63D heterozygous.

DISCUSSION

HCP1 genetic variants were very uncommon in the HEIRS Study participants we evaluated. The prevalences of HCP1 variants in each of the four major racial/ethnic groups were similar in participants selected for high SF/TS and randomly selected controls. Each of the five participants with non-synonymous HCP1 variants had TS above the race-, gender-, field center-, and HFE C282Y and H63D genotype-adjusted TS median values (Table 4). Only three of these had SF above the adjusted SF median. The TS percentiles of those with non-synonymous, coding region variants suggest a possible HCP1 genotype-phenotype relationship, because it seems unlikely that all participants with a non-synonymous HCP1 variant would be in the upper half of the TS percentile distribution by chance alone. Nevertheless, the prevalence of HCP1 variants in our study population was low, and a much larger study would be needed to determine any significant association of HCP1 variants with iron phenotypes.

Qiu and colleagues [16] have more recently suggested a function for the HCP1 protein that is not related to heme or iron transport. They expressed a cloned HCP1 in Xenopus oocytes and in human hepatoma cells and observed that the gene product acts as a proton-coupled folate transporter (PCFT). Although Qiu et al. [16] amended the name of this gene’s product to PCFT/HCP1, HCP1 is called SLC46A1 in GeneBank. However, some of the original findings of Shayeghi and colleagues[13] seem unexplained if HCP1 is solely a folate transporter. For example, PCFT/HCP1 localizes to the plasma membrane in response to iron deficiency and its messenger RNA is up-regulated with hypoxia [13;14].

Part of Qiu and colleagues’s rationale for classifying the gene product of HCP1 as a folate transporter was based on their finding that two members of a single family who had phenotypes of Hereditary Folate Malabsorption (HFM, OMIM 229050) and who were homozygous for an intron2/exon3 boundary mutation in HCP1 (nm_080669.2 c.1082-1 G>A) [16]. Recently, Zhao and colleagues from the same research group described five more HFM patients with homozygosity or compound heterozygosity for six additional PCFT/HCP1 mutations [17]. They found no evidence of iron abnormalities in these patients, although each was an infant at the time of the study. No information on the iron status of the HFE patients’ parents was reported. Because iron overload is usually a progressive, late-onset condition, it would be interesting to study the parents of these HFM patients for evidence of any abnormal iron phenotypes. We did not detect any of the specific PCFT/HCP1 mutations previously reported by Zhao and colleagues [17] in the 788 HEIRS subjects that we tested. The HEIRS Study did not include measurements of serum or red cell folate concentrations, so we are unable to comment on whether the PCFT/HCP1 mutations identified in HEIRS participants affect folate absorption.

We conclude that HCP1 variants are uncommon in US or Canadian Whites, Blacks, Hispanics, and Asians. Although no single HCP1 variant was associated with high TS or SF in our study, a disproportionate fraction of participants with non-synonymous, coding region mutations appear to have high TS.