- Email: [email protected]
A novel N491S mutation in the human SLC11A2 gene impairs protein trafficking and in association with the G212V mutation leads to microcytic anemia and liver iron overload
Blood Cells, Molecules, and Diseases 47 (2011) 243–248
Contents lists available at ScienceDirect
Blood Cells, Molecules, and Diseases j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / y b c m d
A novel N491S mutation in the human SLC11A2 gene impairs protein trafficking and in association with the G212V mutation leads to microcytic anemia and liver iron overload Edouard Bardou-Jacquet a, b, c, Marie-Laure Island a, Anne-Marie Jouanolle c, d, Lénaïck Détivaud a, c, Nadia Fatih a, Martine Ropert c, e, Eolia Brissot f, Annick Mosser c, d, Hervé Maisonneuve f, Pierre Brissot a, b, c, Olivier Loréal a, c,⁎ a
Inserm U991, IFR140, University of Rennes 1, France Liver Disease Department, University Hospital Pontchaillou, Rennes, France c French National Center for Rare Genetic Iron Overload Diseases, France d Molecular Genetic Department, University Hospital Pontchaillou, Rennes, France e Biochemical Department, University Hospital Pontchaillou, Rennes, France f Oncology and Hematology Department, La Roche-Sur-Yon Hospital, France b
a r t i c l e
i n f o
Article history: Submitted 30 March 2010 Revised 1 July 2011 Available online 26 August 2011 (Communicated by A. Townsend, M.D., Ph.D., F.R.C.P., F.R.S., 12 July 2011) Keywords: Iron Liver Hemochromatosis DMT1 Anemia
a b s t r a c t Background: DMT1 is a transmembrane iron transporter involved in iron duodenal absorption and cellular iron uptake. Mutations in the human SLC11A2 gene coding DMT1 lead to microcytic anemia and hepatic iron overload, with unexpectedly low levels of plasma ferritin in the presence of iron stores. Design and methods: We report a patient with a similar phenotype due to two mutations in the SLC11A2 gene, the known p.Gly212Val (G212V) mutation and a novel one, p.Asn491Ser (N491S). To assess the expression of DMT1 in human liver, we studied the expression of the four DMT1 mRNA isoforms by real-time quantitative PCR in control human liver samples. We also studied the effect of G212V and N491S DMT1 mutations on RNA splicing in blood leukocytes and cellular trafficking of dsRed2-tagged-DMT1 protein in the human hepatic cell line HuH7. Results: Our results showed that i) only the isoforms 1B-IRE and 1B-nonIRE were significantly expressed in human liver; ii) the G212V mutation did not seem to affect mRNA splicing and the N491S mutation induced a splicing alteration leading to a truncated protein, which seemed quantitatively of low relevance; and iii) the N491S mutation, in contrast to the G212V mutation, led to abnormal protein trafficking. Conclusions: Our data confirm the major role of DMT1 in the maintenance of iron homeostasis in humans and demonstrate that the N491S mutation, through its deleterious effect on protein trafficking, contributes together with the G212V mutation to the development of anemia and hepatic iron overload. © 2011 Elsevier Inc. All rights reserved.
Introduction The divalent metal transporter 1 gene (SLC11A2, DMT1) encodes a transmembrane cation transporter that ensures Fe 2+ transport from the apical membrane to the cytosol in enterocytes [1], and from acidified endosomes to the cytosol in other cells [2]. SLC11A2 gene transcription leads to four distinct mRNA isoforms. Two transcription initiation sites yield two mRNAs differing at their 5′ ends, leading to proteins that differ in their N terminal end (DMT1-1A, DMT1-1B) [3]. Furthermore, alternative RNA splicing and 3′ end processing yield two isoforms differing at their 3′ ends, with and without an iron-response element [4]. The proteins translated from the IRE or non-IRE forms of
⁎ Corresponding author at: Inserm U991, CHRU Pontchaillou, Rennes France. E-mail address: [email protected] (O. Loréal). 1079-9796/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.bcmd.2011.07.004
the mRNA differ in their C terminal ends and are referred to as DMT1IRE and DMT1-nonIRE proteins as used in the literature. The different DMT1 isoforms exhibit tissue-specific expression [1,3] but equivalent transport efficiency [5]. However, IRE and nonIRE DMT1 protein isoforms have functional differences regarding their protein trafficking. They are both targeted to the plasma membrane and internalized in the early endosomes. Subsequently, the IRE isoform is accumulated in the late endosomes and in lysosomes [6–8], whereas the nonIRE isoform, which is internalized with transferrin receptor 1 and transferrin complex [6,9,10], returns to the plasma membrane via recycling endosomes [8,10]. Therefore, the IRE and nonIRE isoforms could play different roles in iron homeostasis. The IRE isoform could act as an iron transporter at the membrane of epithelial cells for transporting extracellular iron directly into cytosol, whereas the nonIRE isoform could act as the endosomal iron transporter in the transferrin–transferrin receptor pathway [8].
244
E. Bardou-Jacquet et al. / Blood Cells, Molecules, and Diseases 47 (2011) 243–248
Four patients presenting with SLC11A2 mutations have been reported [11–14]. They all showed hypochromic, microcytic anemia, and three of them presented with massive iron overload as mimicked by a rat model of SLC11A2 mutation [15]. Unexpectedly [16], plasma ferritin levels were either normal or only slightly elevated. Here, we report a new case of SLC11A2 mutations presenting with hypochromic microcytic anemia and hepatic iron overload, related to two heterozygous mutations of SLC11A2, one of which has not been previously described. The liver plays a key role in iron homeostasis, especially through the secretion of hepcidin, which regulates circulating iron content, and through the storage of excessive iron within ferritin. However, little data are available on the role of DMT1 in the liver. The aims of this study were to describe the expression of the different isoforms of DMT1 in human liver and to assess in a hepatic cell line the consequences of the two described mutations on both SLC11A2 mRNA splicing and protein trafficking. Design and methods Patient The patient was a 27-year-old woman diagnosed with anemia at the age of 13 years during a metrorrhagia work-up. Gynecologic examination was normal. Anemia (Hb: 8.6 g/dL) was microcytic (MCV: 58 fL) and hypochromic (MCH: 17.8 pg/cell). Hematological parameters showed no improvement under oral ferrous sulfate supplementation. Clinical examination was normal but showed a presentation compatible with childhood chronic anemia. Inflammation was ruled out, as were hemolysis and paroxysmal nocturnal hemoglobinuria. In plasma, transferrin saturation and serum iron were high (79%, normal range: 20–50%; and 250 μg/dL, normal range: 50–170 μg/dL, respectively), and transferrin was normal (2.3 g/L, normal range: 2–3.6 g/L). Ferritin was mildly elevated (300 μg/L, normal range: 13–76 μg/L). Hemoglobin electrophoresis was normal. Bone marrow aspiration showed no ring sideroblasts, 4% type I sideroblasts, and hemosiderin-laden macrophages. Hematological parameters of both parents and brother were normal. When the patient was 17 years old, a radioactive iron study showed altered iron incorporation in red blood cells with a maximal incorporation rate of 29% at day 14 (normal range, 70–90%). Hematological parameters were unchanged, plasma soluble transferrin receptor was highly elevated (66 nmol/L, normal b 28 nmol/L), and erythropoietin was elevated (32.6 IU/L, normal range, 12–18 IU/L). Follow up since that time showed a stable disease. The patient had undergone three rounds of deferoxamine treatment on the basis of moderately elevated ferritin levels (highest level: 574 μg/L): 2 months at age 17 years, 5 months at age 21, and 3 months at age 24. At age 27, she was referred to our reference center for iron overload evaluation. Complete blood cell counts were unchanged, and liver enzymes and C-reactive protein were normal. Plasma ferritin was 274 μg/L, serum iron was 37 μmol/L, transferrin was 1.8 g/L (normal range: 2–3.8), transferrin saturation was 83% (normal range: b45%), and serum transferrin receptor was 2.93 mg/L (normal range: 0.83–1.76 mg/L), and plasma ceruloplasmin level was normal. Liver iron concentration, as evaluated by magnetic resonance imaging (MRI), was increased: 300 μmol/g dry weight liver (N b 36). Non-transferrin-bound iron was elevated at 1.1 μmol/L (normal range: undetectable), and there was no detectable labile plasma iron (both iron species were determined as previously described [17]). HFE genotyping showed homozygosity for the p.His63Asp mutation. SLC11A2 alteration was suspected and confirmed by genetic testing, which revealed compound heterozygosity. Other genetic causes of major iron overload were ruled out by sequencing, and no mutation was identified in SLC40A1 (the ferroportin gene), HJV (hemojuvelin), HAMP (hepcidin), or TFR2 (transferrin receptor 2). HFE
sequencing confirmed homozygosity for the p.His63Asp mutation. Parents declared no known consanguinity. A family study could not be performed as DNA was not available. SLC11A2 sequencing and transcript analysis After receiving informed consent for genetic testing according to usual diagnostic procedures, patient DNA was extracted from a total blood sample collected in EDTA. Each exon of the SLC11A2 gene was amplified by PCR and then sequenced on both strands using the BigDye Terminator v3.1 Kit and ABI Prism 3130xl DNA sequencer (Applied Biosystems) following the manufacturer's instructions. Mutations were analyzed using the multiple criteria software Polyphen (http://genetics.bwh.harvard.edu/pph2/index.shtml) to determine effects at the protein level. Mutations were also input into the algorithm GeneSplicer (http://www.cbcb.umd.edu/software/ GeneSplicer/), and MaxEntScan (http://genes.mit.edu/burgelab/ maxent/Xmaxentscan_scoreseq.html) to determine potential effects on splicing. Furthermore, for RNA analysis, a blood sample was collected and total RNA was extracted using the PaxGene Blood RNA kit (PreAnalytiX QIAGEN) following manufacturer's instructions. Total RNA was reverse transcribed using SuperScript II® (Invitrogen) following the manufacturer's instructions. The surrounding sequence of mutations was amplified by PCR with the following set of primers: c.635 G N T: 5′-TGCGGAAGCTAGAAGCATTT and 5′-CCACAGCCAGTGTCGAGTTA, and c.1472A N G: 5′-ACAGCTTCCCTTTGCTCTCA and 5′-CCAACCAACGGTTGAGTCAT for the IRE isoform; and 5′ACAGCTTCCCTTTGCTCTCA and 5′-CAGGCTGAGCTGTCAATCCCAGAT for the nonIRE isoform. PCR products were checked by electrophoresis on an agarose gel and analyzed by sequencing. DNA from the abnormal band was sequenced after purification using NucleoSpin Extract® (Macherey–Nagel) and following the manufacturer's instructions. Cell culture and transfection HUH7 cells were grown in Dulbecco's Minimal Essential Medium (Invitrogen) supplemented with 10% fetal bovine serum, 100 μg/mL streptomycin, and 100 U/mL penicillin. Caco-2 differentiated cells, which are of enterocytic origin and express all SLC11A2 mRNA isoforms [5], were grown in decomplemented Dulbecco's Minimal Essential Medium, supplemented with 20% fetal bovine serum, 1% non-essential amino acid, 100 μg/mL streptomycin, and 100 U/mL penicillin, and used as control. HUH7 cells were grown in chamber slides (Lab-Tek®) and transfected using Transfectin (Bio-Rad) with 1.5 μg of DMT1 construct, following the manufacturer's instructions. Liver samples To assess SLC11A2 gene expression in human liver, we used control liver samples obtained through the Rennes Biological Resource Center, after informed consent, from five patients having undergone liver surgery for hepatic metastases originating from other organs. Only non-tumorous and histologically normal tissues were used. The study was conducted in accordance with the Helsinki declaration and the local ethics committee. RNA extraction, reverse transcription and real-time quantitative PCR RNA was extracted from liver biopsies, HUH7 cells, and Caco-2 cells using the SV Total RNA Isolation System Kit (Promega) following the manufacturer's instructions. RNA was reverse transcribed using the M-MLV-RT kit (Promega) and 1 μg of total RNA, with random primers for real-time quantitative PCR or a DMT1-IRE–specific (CACCCTAATCCAGTTCTAAGG) primer for DMT1-IRE cloning.
E. Bardou-Jacquet et al. / Blood Cells, Molecules, and Diseases 47 (2011) 243–248
245
cDNA was subjected to real-time quantitative PCR in an ABI Prism 7000 Sequence detection system (Applied Biosystems) using the qPCR® Mastermix Plus for SYBR® Green I kit (Eurogentec), following the manufacturer's instructions. The following specific primer sets were used: DMT1 nonIRE, forward: 5′-CTTGGGTTGGCAATGTTTGA, reverse: 5′-CAGCGTCCATGGTGTTCAGA; DMT1 IRE, forward: 5′GAGGAGAGCCATCTATTTTGTTCC, reverse: 5′-ATGACGATTCTGCTGAGAGGTG; DMT1 1A, forward: 5′-GCTGAAGACGGAGGCAGC, reverse: 5′-GTTCTTAGAATATGATTTTAGTTCACAGTGTG; and DMT1 1B, forward: 5′-GTTGCGGAGCTGGTAAGAATC, reverse 5′-GGAGATCTTCTCATTAAAGTAAG. Results were normalized to the quantification of 18S. Each sample was examined in triplicate. Results were expressed relative to quantification of DMT1 expression in Caco-2 with the ΔΔCt method.
Our results are in accordance with this prediction. They showed no detection of any splicing alteration for the c.635 G N T mutation. Conversely, gel electrophoresis of the nonIRE PCR product analyzing the effect of the c.1472A N G mutation revealed a weaker band approximately 100 bp smaller than control (Fig. 1). Sequencing of this abnormal band showed that the mutation created a new donor site. When this site is used, the last 104 nucleotides of exon 15 are excised, modifying the open reading frame and leading to a premature stop codon downstream. Sequencing of the PCR products overlapping the cDNA regions carrying the c.635 G N T and c.1472A N G mutations showed that the mutations were present in the heterozygous state, suggesting that both allelic forms of transcripts were present.
Plasmid constructs
Fig. 2 presents results of the quantified expression of the four SLC11A2 transcript isoforms in Caco-2 cells, human liver, and HuH7 cells. The SLC11A2 mRNA level was weaker in human liver than in Caco-2 cells. Isoform1A was not significantly expressed either in human liver or in HUH7 cells. IRE and nonIRE isoforms were both expressed in liver and HuH7 cells.
DMT1 1B-IRE was cloned from cDNA obtained from HUH7 cells, and the DMT1 1B-nonIRE wild-type and the mutated form deleted for exon 12 (del12 DMT1) were generous gifts from Drs. Horvathova and Ponka [18]. These sequences were inserted into the Xho1–Hind3 site of the pdsRed2-C1 expression vector (Clontech). Site-directed mutagenesis was performed with both IRE and nonIRE isoforms, using the QuickChange II kit (Stratagene) with the following specific primers: G212V, 5′GCTAGAAGCATTTTTTGTCTTTCTCATCACTATTATGGCCC, 5′GGGCCATAATAGTGATGAGAAAGACAAAAATGCTTCTAGC; and N491S, 5′ GGTCCTTATCATCTGTTCCATCAGTATGTACTTTGTAGTGG, 5′CCACTACAAAGTACATACTGATGGAACAGATGATAAGGACC. The identity of the construct was confirmed by DNA sequencing. Immunolocalization HUH7 cells grown in chamber slides were transfected with the DMT1 construct, fixed in 4% paraformaldehyde, permeabilized (10% fetal bovine serum, 0.1% saponin in PBS) and incubated (1 h, room temperature) with the following specific antibodies: chicken anticalreticulin (1:200) (Abcam) to label endoplasmic reticulum; rabbit anti-LAMP1 1 μg/mL (Abcam) to label late endosome/lysosome; and rabbit anti-EEA1 1 μg/mL (Abcam) to label early endosome. Cells were then washed and incubated (1 h, room temperature) with FITCtagged secondary antibodies (donkey anti-chicken, 1:200; Jackson or donkey anti-rabbit 1:200; Jackson according to primary antibody). Nuclei were stained with Hoechst 5 μg/mL. Cells were visualized using a Leica SP2 confocal microscope with a 63× oil-immersion objective (IFR 140 platform). Images were acquired using the Leica Confocal Software.
DMT1 expression in human liver and HuH7 cells
Subcellular localization of DMT1 1B-IRE, 1B-nonIRE wild type, and mutants in HuH7 cells To assess DMT1 subcellular localization, we transiently transfected the cells with the corresponding expression vectors and used antibodies against calreticulin, EEA1, and LAMP1 to label endoplasmic reticulum, early endosomes, and late endosomes/lysosomes, respectively (Figs. 3, 4; Supplementary Figs. 1, 2) to check for co-localization of DMT1 with organelles. DMT1 IRE co-localized partially with early endosomes and co-localized with late endosomes/lysosomes, whereas DMT1 nonIRE co-localized only with early endosomes (Fig. 3; Supplementary Fig. 1A, B). Moreover, the del12 DMT1 nonIRE form co-localized with endoplasmic reticulum and late endosomes in addition to early endosomes (Fig. 3; Supplementary Fig. 1C). The G212V DMT1 IRE and nonIRE forms showed the same subcellular localization as the wild-type form (Fig. 4; Supplementary Fig. 2A, B). G212 V
N491S IRE
N 491S nonIRE
Results SLC11A2 sequencing and transcript analysis Sequencing of the SLC11A2 gene revealed two mutations, c.635 G N T and c.1472A N G (sequence from GenBank NM_000617.2). Mutation c.635 G N T is located in exon 8, leading to the substitution of a glycine for valine at position 212 (G212V), in the fifth transmembrane domain of the protein. This change has already been reported as potentially deleterious [14]. Mutation c.1472A N G is located in exon 15, leading to the substitution of asparagine for serine at position 491 (N491S), in the eleventh transmembrane domain. Both amino acids are highly conserved through orthologs. Polyphen predicted no consequence for the c.635 G N T mutation. Conversely, the c.1472A N G mutation was predicted to be damaging. MaxEntScan and GeneSplicer predicted no consequence for splicing from the c.635 G N T mutation, but predicted potential new acceptor and donor splicing sites for the c.1472A N G mutation.
Fig. 1. Analysis of the effect of the mutations on mRNA splicing. Total RNA was obtained from the peripheral blood of the patient and controls and then transcribed. Before electrophoresis on an agarose gel, PCR amplification was performed using primers encompassing the c.635 G N T and c.1472AN G mutations. For the c.1472A N G mutation, IRE- and nonIRE-specific primers were used because of the differing 5′ ends near the mutation.
246
E. Bardou-Jacquet et al. / Blood Cells, Molecules, and Diseases 47 (2011) 243–248
N491S DMT1 IRE and nonIRE showed a disturbed subcellular localization: co-localization with endoplasmic reticulum and early and late endosome, similar to that obtained for the del12 DMT1 nonIRE isoform (Fig. 4; Supplementary Fig. 2C, D). Discussion
Fig. 2. Real-time quantitative PCR quantification of DMT1 isoforms in Caco-2 cells, human liver, and HuH7 cells. Liver represents the mean of five different human liver biopsies. Caco-2 and HuH7 represent means for each respective cell line. Each experiment was done in triplicate. Results were normalized to the amount of 18S and are expressed as percent of Caco-2 cells with DMT1 expression.
Ab anti-EEA1
Ab anti-LAMP1
nonIRE Del12
IRE WT
nonIRE WT
Ab anti-Calreticulin
DMT1 function was first reported in enterocytes where it transports iron from the gut lumen to cytosol. The discovery of DMT1 mutations in animal models of chronic anemia led to investigation of its role in erythropoiesis. It is currently thought that the DMT1 IRE isoform is predominantly expressed in epithelial cells where it is essential for iron uptake, whereas the DMT1 nonIRE isoform is expressed in other cells where it transports iron issued from the transferrin pathway across the endoplasmic membrane toward the cytosol. However, despite the report of unexpected hepatic iron overload and the key role of DMT1 in iron metabolism, few data are available regarding the expression of DMT1 in human liver. Here, we report a fifth case of chronic microcytic anemia with iron overload resulting from SLC11A2 mutations in a compound heterozygous state, with one previously undescribed mutation (p.Asn491Ser). The phenotypic expression was similar to the previously described patients [11–14], except for one patient who had no liver iron overload. However, in that exceptional case, an increase in plasma iron levels and transferrin saturation over a two-year follow-up did suggest that iron overload could ultimately occur [12]. Our patient was diagnosed at the age of 27, showing a long-term natural history of stable microcytic anemia with major iron overload, despite three courses of deferoxamine, and with poorly elevated plasma ferritin levels in regard to iron overload. Altogether, these clinical observations are in agreement with a
Fig. 3. Subcellular localization of wild-type DMT1 1B-IRE and 1B-nonIRE and del12 DMT1 1B-nonIRE in HuH7 cells. HuH7 cells were transiently transfected with the pdsRed2-C1 construct with DMT1 1B-IRE wild type, DMT1 1B-nonIRE wild type, and del12 DMT1 1B-nonIRE. Cells were incubated with FITC-tagged primary antibodies: chicken anti-calreticulin to label endoplasmic reticulum, rabbit anti-LAMP1 to label late endosome/lysosome, and rabbit anti-EEA1 to label early endosome. Images were acquired using a Leica SP2 confocal microscope. Bar = 10 μm. The figure shows a merged image representative of each experiment, and split images are available as Supplementary Fig. 1A–C.
E. Bardou-Jacquet et al. / Blood Cells, Molecules, and Diseases 47 (2011) 243–248
Ab anti-EEA1
Ab anti-LAMP1
nonIRE G212V
IRE G212V
NonIRE N491S
IRE N491S
Ab anti-Calreticulin
247
Fig. 4. Subcellular localization of N491S and G212V mutants of DMT1 1B-IRE and 1B-nonIRE isoforms in Huh7 cells. HuH7 cells were transiently transfected with a pdsRed2-C1 construct with G212V and N491S mutants of isoform IRE and nonIRE. Cells were incubated with FITC-tagged primary antibodies: chicken anti-calreticulin to label endoplasmic reticulum, rabbit anti-LAMP1 to label late endosome/lysosome, and rabbit anti-EEA1 to label early endosome. Images were acquired using a Leica SP2 confocal microscope. Bar = 10 μm. Figure shows merged image representative of each experiment, and split images are available as Supplementary Fig. 2A–D.
major role of DMT1 in erythropoiesis and suggest either a compensatory pathway for iron absorption or a minimal residual activity that could participate in the development of iron overload. They also emphasize, from a practical viewpoint, the importance of resorting to direct evaluation of tissue iron concentration by MRI in patients with suspected or proved SLC11A2 mutations. In mouse liver, SLC11A2 mRNA expression is lower than in intestine and kidney, and isoform 1A has not been detected [3]. Whether IRE or nonIRE forms are predominant remains controversial [3,19]. In human liver, our results confirm that DMT1 is significantly expressed and demonstrate that the 1B mRNA isoform is the only significantly expressed form. In addition, we found no predominance in the expression of IRE or nonIRE isoforms in the liver. The DMT1 IRE isoform of DMT1 could play a role, in addition to Zip14 [20], in the high hepatocyte ability for taking up non-transferrin-bound iron. In contrast to duodenum, liver DMT1 expression is increased by iron overload [21,22], indicating that DMT1 could contribute to the increased uptake of non-transferrin-bound iron by the iron overloaded hepatocyte [23,24]. Peripheral blood mRNA analyses showed no mutation effect on splicing for the mutation G212V and a quantitatively very low splicing
modification for the N491S mutation, which was detectable in the nonIRE form and not in the IRE form likely because of a lower amount of mRNA that limited detection. Therefore, the only significant proteins present in the patient are likely those carrying the N491S and G212V missense mutations. The sequencing of the PCR products of the cDNA regions overlapping the mutations showed the presence of the two mutations in the heterozygous state, suggesting that the two allelic forms of transcripts were present. We did not perform a quantitative assessment of each mutated PCR product; however, global assessment was similar to control patients. Because the 1A isoform of DMT1 was not significantly expressed in the liver and the 1B-IRE and 1B-nonIRE DMT1 isoforms were expressed in human liver, we studied the consequences of the two mutations on cellular trafficking in the HUH7 hepatic cell line expressing both IRE and nonIRE isoforms. Assessing cellular trafficking of the protein by co-localization studies, we found that the subcellular localization of DMT1 wild-type isoforms was in agreement with previously published data [6,8–10]. In addition, the DMT1 deleted protein (del12 DMT1 1B-nonIRE), used as control, led to an abnormal protein trafficking with co-localization to endoplasmic reticulum, as previously described [18]. Our data showed that the
248
E. Bardou-Jacquet et al. / Blood Cells, Molecules, and Diseases 47 (2011) 243–248
N491S mutation caused improper protein trafficking similar to findings reported with the del12 DMT1 1B-nonIRE. They suggest a loss of function resulting from disturbed protein trafficking. Further experiments are warranted to define the underlying mechanism explaining the critical role of the asparagine residue at position 491. Our results suggested that the G212V did not affect cellular trafficking of the protein, supporting an altered iron transport ability and implying that the glycine residue at position 212 plays an important role in the iron transport function of DMT1. The report by Vecchi et al. [25] in 2010 showing an HFE mutation within the HUH7 cell line could be relevant to the mutation findings in our data. However, no direct interaction has been identified between HFE and DMT1 proteins. Moreover, the identified biological role of HFE is to participate in the transduction of signals linked to iron status and regulating hepcidin expression. In our work studying the effect of mutations within SLC11A2, we did not use experimental conditions with modulation of iron status, which could alter the results. Therefore, it is likely that the HFE gene mutation within the HuH7 cell line genome did not alter the outcome of our experiments. The development of iron overload in the present case of SLC11A2 mutations raises concerns about a potential residual activity of DMT1 which, up-regulated in this situation, could lead to iron overload [12]. Thus, assessment of residual transport activity, which could participate in iron overload, should be carried on G212V DMT1 forms for which the protein localization is not disturbed. Alternatively, a complete loss of function could suggest that a compensatory pathway for iron absorption may exist that has yet to be identified. In addition, the fact that non-transferrin-bound iron was found in plasma suggests that this form of iron, which is avidly taken up by hepatocytes [23], could favor the development of hepatic iron overload. In this case, the iron route likely involves a ZIP14 pathway and not an endocytic pathway. Therefore, an iron sequestration in endocytic vesicles could not explain the low ferritin increase, as previously suggested. In conclusion, our data confirm the major role of DMT1 in the maintenance of iron homeostasis in humans and demonstrate that the N491S mutation, through a deleterious effect on protein trafficking, contributes together with the G212V mutation to the development of anemia and hepatic iron overload. Supplementary materials related to this article can be found online at doi:10.1016/j.bcmd.2011.07.004. Fundings EBJ was granted by the French National Academy of Medicine. This work was supported by the LSHM-CT-2006-037296 European Community Grant (Euroiron1), the Association Fer et Foie. Acknowledgments The author would like to thank Dr. Monika Horvathova, Faculty of Medicine Palacky University Olomouc, Czech Republic, and Pr. Prem Ponka, McGill University, Montréal, Canada, for the generous gift of plasmid constructs. We would also like to thank Stéphanie Dutertre from the microscopy platform, IFR GFAS, Rennes.
References [1] H. Gunshin, B. Mackenzie, U.V. Berger, et al., Cloning and characterization of a mammalian proton-coupled metal-ion transporter, Nature 388 (1997) 482–488. [2] V. Picard, G. Govoni, N. Jabado, P. Gros, Nramp 2 (DCT1/DMT1) expressed at the plasma membrane transports iron and other divalent cations into a calceinaccessible cytoplasmic pool, J Biol Chem 275 (2000) 35738–35745. [3] N. Hubert, M.W. Hentze, Previously uncharacterized isoforms of divalent metal transporter (DMT)-1: implications for regulation and cellular function, Proc Natl Acad Sci U S A 99 (2002) 12345–12350. [4] P.L. Lee, T. Gelbart, C. West, et al., The human Nramp2 gene: characterization of the gene structure, alternative splicing, promoter region and polymorphisms, Blood Cells Mol Dis 24 (1998) 199–215. [5] B. Mackenzie, H. Takanaga, N. Hubert, et al., Functional properties of multiple isoforms of human divalent metal-ion transporter 1 (DMT1), Biochem J 403 (2007) 59–69. [6] M. Tabuchi, N. Tanaka, J. Nishida-Kitayama, et al., Alternative splicing regulates the subcellular localization of divalent metal transporter 1 isoforms, Mol Biol Cell 13 (2002) 4371–4387. [7] F. Canonne-Hergaux, S. Gruenheid, P. Ponka, P. Gros, Cellular and subcellular localization of the Nramp2 iron transporter in the intestinal brush border and regulation by dietary iron, Blood 93 (1999) 4406–4417. [8] S. Lam-Yuk-Tseung, P. Gros, Distinct targeting and recycling properties of two isoforms of the iron transporter DMT1 (NRAMP2, Slc11A2), Biochemistry 45 (2006) 2294–2301. [9] F. Canonne-Hergaux, A.S. Zhang, P. Ponka, P. Gros, Characterization of the iron transporter DMT1 (NRAMP2/DCT1) in red blood cells of normal and anemic mk/ mk mice, Blood 98 (2001) 3823–3830. [10] N. Touret, W. Furuya, J. Forbes, et al., Dynamic traffic through the recycling compartment couples the metal transporter Nramp2 (DMT1) with the transferrin receptor, J Biol Chem 278 (2003) 25548–25557. [11] M.P. Mims, Y. Guan, D. Pospisilova, et al., Identification of a human mutation of DMT1 in a patient with microcytic anemia and iron overload, Blood 105 (2005) 1337–1342. [12] E. Blanco, C. Kannengiesser, B. Grandchamp, et al., Not all DMT1 mutations lead to iron overload, Blood Cells Mol Dis 43 (2009) 199–201. [13] A. Iolascon, M. d'Apolito, V. Servedio, et al., Microcytic anemia and hepatic iron overload in a child with compound heterozygous mutations in DMT1 (SCL11A2), Blood 107 (2006) 349–354. [14] C. Beaumont, J. Delaunay, G. Hetet, et al., Two new human DMT1 gene mutations in a patient with microcytic anemia, low ferritinemia, and liver iron overload, Blood 107 (2006) 4168–4170. [15] K. Thompson, R.M. Molina, J.D. Brain, M. Wessling-Resnick, Belgrade rats display liver iron loading, J Nutr 136 (2006) 3010–3014. [16] P. Brissot, M. Bourel, D. Herry, et al., Assessment of liver iron content in 271 patients: a reevaluation of direct and indirect methods, Gastroenterology 80 (1981) 557–565. [17] C. Le Lan, O. Loreal, T. Cohen, et al., Redox active plasma iron in C282Y/C282Y hemochromatosis, Blood 105 (2005) 4527–4531. [18] M. Priwitzerova, G. Nie, A.D. Sheftel, et al., Functional consequences of the human DMT1 (SLC11A2) mutation on protein expression and iron uptake, Blood 106 (2005) 3985–3987. [19] D. Tchernitchko, M. Bourgeois, M.E. Martin, C. Beaumont, Expression of the two mRNA isoforms of the iron transporter Nramp2/DMTI in mice and function of the iron responsive element, Biochem J 363 (2002) 449–455. [20] J.P. Liuzzi, F. Aydemir, H. Nam, et al., Zip14 (Slc39a14) mediates non-transferrinbound iron uptake into cells, Proc Natl Acad Sci U S A 103 (2006) 13612–13617. [21] H. Zoller, R.O. Koch, I. Theurl, et al., Expression of the duodenal iron transporters divalent-metal transporter 1 and ferroportin 1 in iron deficiency and iron overload, Gastroenterology 120 (2001) 1412–1419. [22] D. Trinder, P.S. Oates, C. Thomas, et al., Localisation of divalent metal transporter 1 (DMT1) to the microvillus membrane of rat duodenal enterocytes in iron deficiency, but to hepatocytes in iron overload, Gut 46 (2000) 270–276. [23] P. Brissot, T.L. Wright, W.L. Ma, R.A. Weisiger, Efficient clearance of nontransferrin-bound iron by rat liver. Implications for hepatic iron loading in iron overload states, J Clin Invest 76 (1985) 1463–1470. [24] T.L. Wright, P. Brissot, W.L. Ma, R.A. Weisiger, Characterization of non-transferrinbound iron clearance by rat liver, J Biol Chem 261 (1986) 10909–10914. [25] C. Vecchi, G. Montosi, A. Pietrangelo, Huh-7: a human “hemochromatotic” cell line, Hepatology 51 (2010) 654–659.
Contents lists available at ScienceDirect
Blood Cells, Molecules, and Diseases j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / y b c m d
A novel N491S mutation in the human SLC11A2 gene impairs protein trafficking and in association with the G212V mutation leads to microcytic anemia and liver iron overload Edouard Bardou-Jacquet a, b, c, Marie-Laure Island a, Anne-Marie Jouanolle c, d, Lénaïck Détivaud a, c, Nadia Fatih a, Martine Ropert c, e, Eolia Brissot f, Annick Mosser c, d, Hervé Maisonneuve f, Pierre Brissot a, b, c, Olivier Loréal a, c,⁎ a
Inserm U991, IFR140, University of Rennes 1, France Liver Disease Department, University Hospital Pontchaillou, Rennes, France c French National Center for Rare Genetic Iron Overload Diseases, France d Molecular Genetic Department, University Hospital Pontchaillou, Rennes, France e Biochemical Department, University Hospital Pontchaillou, Rennes, France f Oncology and Hematology Department, La Roche-Sur-Yon Hospital, France b
a r t i c l e
i n f o
Article history: Submitted 30 March 2010 Revised 1 July 2011 Available online 26 August 2011 (Communicated by A. Townsend, M.D., Ph.D., F.R.C.P., F.R.S., 12 July 2011) Keywords: Iron Liver Hemochromatosis DMT1 Anemia
a b s t r a c t Background: DMT1 is a transmembrane iron transporter involved in iron duodenal absorption and cellular iron uptake. Mutations in the human SLC11A2 gene coding DMT1 lead to microcytic anemia and hepatic iron overload, with unexpectedly low levels of plasma ferritin in the presence of iron stores. Design and methods: We report a patient with a similar phenotype due to two mutations in the SLC11A2 gene, the known p.Gly212Val (G212V) mutation and a novel one, p.Asn491Ser (N491S). To assess the expression of DMT1 in human liver, we studied the expression of the four DMT1 mRNA isoforms by real-time quantitative PCR in control human liver samples. We also studied the effect of G212V and N491S DMT1 mutations on RNA splicing in blood leukocytes and cellular trafficking of dsRed2-tagged-DMT1 protein in the human hepatic cell line HuH7. Results: Our results showed that i) only the isoforms 1B-IRE and 1B-nonIRE were significantly expressed in human liver; ii) the G212V mutation did not seem to affect mRNA splicing and the N491S mutation induced a splicing alteration leading to a truncated protein, which seemed quantitatively of low relevance; and iii) the N491S mutation, in contrast to the G212V mutation, led to abnormal protein trafficking. Conclusions: Our data confirm the major role of DMT1 in the maintenance of iron homeostasis in humans and demonstrate that the N491S mutation, through its deleterious effect on protein trafficking, contributes together with the G212V mutation to the development of anemia and hepatic iron overload. © 2011 Elsevier Inc. All rights reserved.
Introduction The divalent metal transporter 1 gene (SLC11A2, DMT1) encodes a transmembrane cation transporter that ensures Fe 2+ transport from the apical membrane to the cytosol in enterocytes [1], and from acidified endosomes to the cytosol in other cells [2]. SLC11A2 gene transcription leads to four distinct mRNA isoforms. Two transcription initiation sites yield two mRNAs differing at their 5′ ends, leading to proteins that differ in their N terminal end (DMT1-1A, DMT1-1B) [3]. Furthermore, alternative RNA splicing and 3′ end processing yield two isoforms differing at their 3′ ends, with and without an iron-response element [4]. The proteins translated from the IRE or non-IRE forms of
⁎ Corresponding author at: Inserm U991, CHRU Pontchaillou, Rennes France. E-mail address: [email protected] (O. Loréal). 1079-9796/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.bcmd.2011.07.004
the mRNA differ in their C terminal ends and are referred to as DMT1IRE and DMT1-nonIRE proteins as used in the literature. The different DMT1 isoforms exhibit tissue-specific expression [1,3] but equivalent transport efficiency [5]. However, IRE and nonIRE DMT1 protein isoforms have functional differences regarding their protein trafficking. They are both targeted to the plasma membrane and internalized in the early endosomes. Subsequently, the IRE isoform is accumulated in the late endosomes and in lysosomes [6–8], whereas the nonIRE isoform, which is internalized with transferrin receptor 1 and transferrin complex [6,9,10], returns to the plasma membrane via recycling endosomes [8,10]. Therefore, the IRE and nonIRE isoforms could play different roles in iron homeostasis. The IRE isoform could act as an iron transporter at the membrane of epithelial cells for transporting extracellular iron directly into cytosol, whereas the nonIRE isoform could act as the endosomal iron transporter in the transferrin–transferrin receptor pathway [8].
244
E. Bardou-Jacquet et al. / Blood Cells, Molecules, and Diseases 47 (2011) 243–248
Four patients presenting with SLC11A2 mutations have been reported [11–14]. They all showed hypochromic, microcytic anemia, and three of them presented with massive iron overload as mimicked by a rat model of SLC11A2 mutation [15]. Unexpectedly [16], plasma ferritin levels were either normal or only slightly elevated. Here, we report a new case of SLC11A2 mutations presenting with hypochromic microcytic anemia and hepatic iron overload, related to two heterozygous mutations of SLC11A2, one of which has not been previously described. The liver plays a key role in iron homeostasis, especially through the secretion of hepcidin, which regulates circulating iron content, and through the storage of excessive iron within ferritin. However, little data are available on the role of DMT1 in the liver. The aims of this study were to describe the expression of the different isoforms of DMT1 in human liver and to assess in a hepatic cell line the consequences of the two described mutations on both SLC11A2 mRNA splicing and protein trafficking. Design and methods Patient The patient was a 27-year-old woman diagnosed with anemia at the age of 13 years during a metrorrhagia work-up. Gynecologic examination was normal. Anemia (Hb: 8.6 g/dL) was microcytic (MCV: 58 fL) and hypochromic (MCH: 17.8 pg/cell). Hematological parameters showed no improvement under oral ferrous sulfate supplementation. Clinical examination was normal but showed a presentation compatible with childhood chronic anemia. Inflammation was ruled out, as were hemolysis and paroxysmal nocturnal hemoglobinuria. In plasma, transferrin saturation and serum iron were high (79%, normal range: 20–50%; and 250 μg/dL, normal range: 50–170 μg/dL, respectively), and transferrin was normal (2.3 g/L, normal range: 2–3.6 g/L). Ferritin was mildly elevated (300 μg/L, normal range: 13–76 μg/L). Hemoglobin electrophoresis was normal. Bone marrow aspiration showed no ring sideroblasts, 4% type I sideroblasts, and hemosiderin-laden macrophages. Hematological parameters of both parents and brother were normal. When the patient was 17 years old, a radioactive iron study showed altered iron incorporation in red blood cells with a maximal incorporation rate of 29% at day 14 (normal range, 70–90%). Hematological parameters were unchanged, plasma soluble transferrin receptor was highly elevated (66 nmol/L, normal b 28 nmol/L), and erythropoietin was elevated (32.6 IU/L, normal range, 12–18 IU/L). Follow up since that time showed a stable disease. The patient had undergone three rounds of deferoxamine treatment on the basis of moderately elevated ferritin levels (highest level: 574 μg/L): 2 months at age 17 years, 5 months at age 21, and 3 months at age 24. At age 27, she was referred to our reference center for iron overload evaluation. Complete blood cell counts were unchanged, and liver enzymes and C-reactive protein were normal. Plasma ferritin was 274 μg/L, serum iron was 37 μmol/L, transferrin was 1.8 g/L (normal range: 2–3.8), transferrin saturation was 83% (normal range: b45%), and serum transferrin receptor was 2.93 mg/L (normal range: 0.83–1.76 mg/L), and plasma ceruloplasmin level was normal. Liver iron concentration, as evaluated by magnetic resonance imaging (MRI), was increased: 300 μmol/g dry weight liver (N b 36). Non-transferrin-bound iron was elevated at 1.1 μmol/L (normal range: undetectable), and there was no detectable labile plasma iron (both iron species were determined as previously described [17]). HFE genotyping showed homozygosity for the p.His63Asp mutation. SLC11A2 alteration was suspected and confirmed by genetic testing, which revealed compound heterozygosity. Other genetic causes of major iron overload were ruled out by sequencing, and no mutation was identified in SLC40A1 (the ferroportin gene), HJV (hemojuvelin), HAMP (hepcidin), or TFR2 (transferrin receptor 2). HFE
sequencing confirmed homozygosity for the p.His63Asp mutation. Parents declared no known consanguinity. A family study could not be performed as DNA was not available. SLC11A2 sequencing and transcript analysis After receiving informed consent for genetic testing according to usual diagnostic procedures, patient DNA was extracted from a total blood sample collected in EDTA. Each exon of the SLC11A2 gene was amplified by PCR and then sequenced on both strands using the BigDye Terminator v3.1 Kit and ABI Prism 3130xl DNA sequencer (Applied Biosystems) following the manufacturer's instructions. Mutations were analyzed using the multiple criteria software Polyphen (http://genetics.bwh.harvard.edu/pph2/index.shtml) to determine effects at the protein level. Mutations were also input into the algorithm GeneSplicer (http://www.cbcb.umd.edu/software/ GeneSplicer/), and MaxEntScan (http://genes.mit.edu/burgelab/ maxent/Xmaxentscan_scoreseq.html) to determine potential effects on splicing. Furthermore, for RNA analysis, a blood sample was collected and total RNA was extracted using the PaxGene Blood RNA kit (PreAnalytiX QIAGEN) following manufacturer's instructions. Total RNA was reverse transcribed using SuperScript II® (Invitrogen) following the manufacturer's instructions. The surrounding sequence of mutations was amplified by PCR with the following set of primers: c.635 G N T: 5′-TGCGGAAGCTAGAAGCATTT and 5′-CCACAGCCAGTGTCGAGTTA, and c.1472A N G: 5′-ACAGCTTCCCTTTGCTCTCA and 5′-CCAACCAACGGTTGAGTCAT for the IRE isoform; and 5′ACAGCTTCCCTTTGCTCTCA and 5′-CAGGCTGAGCTGTCAATCCCAGAT for the nonIRE isoform. PCR products were checked by electrophoresis on an agarose gel and analyzed by sequencing. DNA from the abnormal band was sequenced after purification using NucleoSpin Extract® (Macherey–Nagel) and following the manufacturer's instructions. Cell culture and transfection HUH7 cells were grown in Dulbecco's Minimal Essential Medium (Invitrogen) supplemented with 10% fetal bovine serum, 100 μg/mL streptomycin, and 100 U/mL penicillin. Caco-2 differentiated cells, which are of enterocytic origin and express all SLC11A2 mRNA isoforms [5], were grown in decomplemented Dulbecco's Minimal Essential Medium, supplemented with 20% fetal bovine serum, 1% non-essential amino acid, 100 μg/mL streptomycin, and 100 U/mL penicillin, and used as control. HUH7 cells were grown in chamber slides (Lab-Tek®) and transfected using Transfectin (Bio-Rad) with 1.5 μg of DMT1 construct, following the manufacturer's instructions. Liver samples To assess SLC11A2 gene expression in human liver, we used control liver samples obtained through the Rennes Biological Resource Center, after informed consent, from five patients having undergone liver surgery for hepatic metastases originating from other organs. Only non-tumorous and histologically normal tissues were used. The study was conducted in accordance with the Helsinki declaration and the local ethics committee. RNA extraction, reverse transcription and real-time quantitative PCR RNA was extracted from liver biopsies, HUH7 cells, and Caco-2 cells using the SV Total RNA Isolation System Kit (Promega) following the manufacturer's instructions. RNA was reverse transcribed using the M-MLV-RT kit (Promega) and 1 μg of total RNA, with random primers for real-time quantitative PCR or a DMT1-IRE–specific (CACCCTAATCCAGTTCTAAGG) primer for DMT1-IRE cloning.
E. Bardou-Jacquet et al. / Blood Cells, Molecules, and Diseases 47 (2011) 243–248
245
cDNA was subjected to real-time quantitative PCR in an ABI Prism 7000 Sequence detection system (Applied Biosystems) using the qPCR® Mastermix Plus for SYBR® Green I kit (Eurogentec), following the manufacturer's instructions. The following specific primer sets were used: DMT1 nonIRE, forward: 5′-CTTGGGTTGGCAATGTTTGA, reverse: 5′-CAGCGTCCATGGTGTTCAGA; DMT1 IRE, forward: 5′GAGGAGAGCCATCTATTTTGTTCC, reverse: 5′-ATGACGATTCTGCTGAGAGGTG; DMT1 1A, forward: 5′-GCTGAAGACGGAGGCAGC, reverse: 5′-GTTCTTAGAATATGATTTTAGTTCACAGTGTG; and DMT1 1B, forward: 5′-GTTGCGGAGCTGGTAAGAATC, reverse 5′-GGAGATCTTCTCATTAAAGTAAG. Results were normalized to the quantification of 18S. Each sample was examined in triplicate. Results were expressed relative to quantification of DMT1 expression in Caco-2 with the ΔΔCt method.
Our results are in accordance with this prediction. They showed no detection of any splicing alteration for the c.635 G N T mutation. Conversely, gel electrophoresis of the nonIRE PCR product analyzing the effect of the c.1472A N G mutation revealed a weaker band approximately 100 bp smaller than control (Fig. 1). Sequencing of this abnormal band showed that the mutation created a new donor site. When this site is used, the last 104 nucleotides of exon 15 are excised, modifying the open reading frame and leading to a premature stop codon downstream. Sequencing of the PCR products overlapping the cDNA regions carrying the c.635 G N T and c.1472A N G mutations showed that the mutations were present in the heterozygous state, suggesting that both allelic forms of transcripts were present.
Plasmid constructs
Fig. 2 presents results of the quantified expression of the four SLC11A2 transcript isoforms in Caco-2 cells, human liver, and HuH7 cells. The SLC11A2 mRNA level was weaker in human liver than in Caco-2 cells. Isoform1A was not significantly expressed either in human liver or in HUH7 cells. IRE and nonIRE isoforms were both expressed in liver and HuH7 cells.
DMT1 1B-IRE was cloned from cDNA obtained from HUH7 cells, and the DMT1 1B-nonIRE wild-type and the mutated form deleted for exon 12 (del12 DMT1) were generous gifts from Drs. Horvathova and Ponka [18]. These sequences were inserted into the Xho1–Hind3 site of the pdsRed2-C1 expression vector (Clontech). Site-directed mutagenesis was performed with both IRE and nonIRE isoforms, using the QuickChange II kit (Stratagene) with the following specific primers: G212V, 5′GCTAGAAGCATTTTTTGTCTTTCTCATCACTATTATGGCCC, 5′GGGCCATAATAGTGATGAGAAAGACAAAAATGCTTCTAGC; and N491S, 5′ GGTCCTTATCATCTGTTCCATCAGTATGTACTTTGTAGTGG, 5′CCACTACAAAGTACATACTGATGGAACAGATGATAAGGACC. The identity of the construct was confirmed by DNA sequencing. Immunolocalization HUH7 cells grown in chamber slides were transfected with the DMT1 construct, fixed in 4% paraformaldehyde, permeabilized (10% fetal bovine serum, 0.1% saponin in PBS) and incubated (1 h, room temperature) with the following specific antibodies: chicken anticalreticulin (1:200) (Abcam) to label endoplasmic reticulum; rabbit anti-LAMP1 1 μg/mL (Abcam) to label late endosome/lysosome; and rabbit anti-EEA1 1 μg/mL (Abcam) to label early endosome. Cells were then washed and incubated (1 h, room temperature) with FITCtagged secondary antibodies (donkey anti-chicken, 1:200; Jackson or donkey anti-rabbit 1:200; Jackson according to primary antibody). Nuclei were stained with Hoechst 5 μg/mL. Cells were visualized using a Leica SP2 confocal microscope with a 63× oil-immersion objective (IFR 140 platform). Images were acquired using the Leica Confocal Software.
DMT1 expression in human liver and HuH7 cells
Subcellular localization of DMT1 1B-IRE, 1B-nonIRE wild type, and mutants in HuH7 cells To assess DMT1 subcellular localization, we transiently transfected the cells with the corresponding expression vectors and used antibodies against calreticulin, EEA1, and LAMP1 to label endoplasmic reticulum, early endosomes, and late endosomes/lysosomes, respectively (Figs. 3, 4; Supplementary Figs. 1, 2) to check for co-localization of DMT1 with organelles. DMT1 IRE co-localized partially with early endosomes and co-localized with late endosomes/lysosomes, whereas DMT1 nonIRE co-localized only with early endosomes (Fig. 3; Supplementary Fig. 1A, B). Moreover, the del12 DMT1 nonIRE form co-localized with endoplasmic reticulum and late endosomes in addition to early endosomes (Fig. 3; Supplementary Fig. 1C). The G212V DMT1 IRE and nonIRE forms showed the same subcellular localization as the wild-type form (Fig. 4; Supplementary Fig. 2A, B). G212 V
N491S IRE
N 491S nonIRE
Results SLC11A2 sequencing and transcript analysis Sequencing of the SLC11A2 gene revealed two mutations, c.635 G N T and c.1472A N G (sequence from GenBank NM_000617.2). Mutation c.635 G N T is located in exon 8, leading to the substitution of a glycine for valine at position 212 (G212V), in the fifth transmembrane domain of the protein. This change has already been reported as potentially deleterious [14]. Mutation c.1472A N G is located in exon 15, leading to the substitution of asparagine for serine at position 491 (N491S), in the eleventh transmembrane domain. Both amino acids are highly conserved through orthologs. Polyphen predicted no consequence for the c.635 G N T mutation. Conversely, the c.1472A N G mutation was predicted to be damaging. MaxEntScan and GeneSplicer predicted no consequence for splicing from the c.635 G N T mutation, but predicted potential new acceptor and donor splicing sites for the c.1472A N G mutation.
Fig. 1. Analysis of the effect of the mutations on mRNA splicing. Total RNA was obtained from the peripheral blood of the patient and controls and then transcribed. Before electrophoresis on an agarose gel, PCR amplification was performed using primers encompassing the c.635 G N T and c.1472AN G mutations. For the c.1472A N G mutation, IRE- and nonIRE-specific primers were used because of the differing 5′ ends near the mutation.
246
E. Bardou-Jacquet et al. / Blood Cells, Molecules, and Diseases 47 (2011) 243–248
N491S DMT1 IRE and nonIRE showed a disturbed subcellular localization: co-localization with endoplasmic reticulum and early and late endosome, similar to that obtained for the del12 DMT1 nonIRE isoform (Fig. 4; Supplementary Fig. 2C, D). Discussion
Fig. 2. Real-time quantitative PCR quantification of DMT1 isoforms in Caco-2 cells, human liver, and HuH7 cells. Liver represents the mean of five different human liver biopsies. Caco-2 and HuH7 represent means for each respective cell line. Each experiment was done in triplicate. Results were normalized to the amount of 18S and are expressed as percent of Caco-2 cells with DMT1 expression.
Ab anti-EEA1
Ab anti-LAMP1
nonIRE Del12
IRE WT
nonIRE WT
Ab anti-Calreticulin
DMT1 function was first reported in enterocytes where it transports iron from the gut lumen to cytosol. The discovery of DMT1 mutations in animal models of chronic anemia led to investigation of its role in erythropoiesis. It is currently thought that the DMT1 IRE isoform is predominantly expressed in epithelial cells where it is essential for iron uptake, whereas the DMT1 nonIRE isoform is expressed in other cells where it transports iron issued from the transferrin pathway across the endoplasmic membrane toward the cytosol. However, despite the report of unexpected hepatic iron overload and the key role of DMT1 in iron metabolism, few data are available regarding the expression of DMT1 in human liver. Here, we report a fifth case of chronic microcytic anemia with iron overload resulting from SLC11A2 mutations in a compound heterozygous state, with one previously undescribed mutation (p.Asn491Ser). The phenotypic expression was similar to the previously described patients [11–14], except for one patient who had no liver iron overload. However, in that exceptional case, an increase in plasma iron levels and transferrin saturation over a two-year follow-up did suggest that iron overload could ultimately occur [12]. Our patient was diagnosed at the age of 27, showing a long-term natural history of stable microcytic anemia with major iron overload, despite three courses of deferoxamine, and with poorly elevated plasma ferritin levels in regard to iron overload. Altogether, these clinical observations are in agreement with a
Fig. 3. Subcellular localization of wild-type DMT1 1B-IRE and 1B-nonIRE and del12 DMT1 1B-nonIRE in HuH7 cells. HuH7 cells were transiently transfected with the pdsRed2-C1 construct with DMT1 1B-IRE wild type, DMT1 1B-nonIRE wild type, and del12 DMT1 1B-nonIRE. Cells were incubated with FITC-tagged primary antibodies: chicken anti-calreticulin to label endoplasmic reticulum, rabbit anti-LAMP1 to label late endosome/lysosome, and rabbit anti-EEA1 to label early endosome. Images were acquired using a Leica SP2 confocal microscope. Bar = 10 μm. The figure shows a merged image representative of each experiment, and split images are available as Supplementary Fig. 1A–C.
E. Bardou-Jacquet et al. / Blood Cells, Molecules, and Diseases 47 (2011) 243–248
Ab anti-EEA1
Ab anti-LAMP1
nonIRE G212V
IRE G212V
NonIRE N491S
IRE N491S
Ab anti-Calreticulin
247
Fig. 4. Subcellular localization of N491S and G212V mutants of DMT1 1B-IRE and 1B-nonIRE isoforms in Huh7 cells. HuH7 cells were transiently transfected with a pdsRed2-C1 construct with G212V and N491S mutants of isoform IRE and nonIRE. Cells were incubated with FITC-tagged primary antibodies: chicken anti-calreticulin to label endoplasmic reticulum, rabbit anti-LAMP1 to label late endosome/lysosome, and rabbit anti-EEA1 to label early endosome. Images were acquired using a Leica SP2 confocal microscope. Bar = 10 μm. Figure shows merged image representative of each experiment, and split images are available as Supplementary Fig. 2A–D.
major role of DMT1 in erythropoiesis and suggest either a compensatory pathway for iron absorption or a minimal residual activity that could participate in the development of iron overload. They also emphasize, from a practical viewpoint, the importance of resorting to direct evaluation of tissue iron concentration by MRI in patients with suspected or proved SLC11A2 mutations. In mouse liver, SLC11A2 mRNA expression is lower than in intestine and kidney, and isoform 1A has not been detected [3]. Whether IRE or nonIRE forms are predominant remains controversial [3,19]. In human liver, our results confirm that DMT1 is significantly expressed and demonstrate that the 1B mRNA isoform is the only significantly expressed form. In addition, we found no predominance in the expression of IRE or nonIRE isoforms in the liver. The DMT1 IRE isoform of DMT1 could play a role, in addition to Zip14 [20], in the high hepatocyte ability for taking up non-transferrin-bound iron. In contrast to duodenum, liver DMT1 expression is increased by iron overload [21,22], indicating that DMT1 could contribute to the increased uptake of non-transferrin-bound iron by the iron overloaded hepatocyte [23,24]. Peripheral blood mRNA analyses showed no mutation effect on splicing for the mutation G212V and a quantitatively very low splicing
modification for the N491S mutation, which was detectable in the nonIRE form and not in the IRE form likely because of a lower amount of mRNA that limited detection. Therefore, the only significant proteins present in the patient are likely those carrying the N491S and G212V missense mutations. The sequencing of the PCR products of the cDNA regions overlapping the mutations showed the presence of the two mutations in the heterozygous state, suggesting that the two allelic forms of transcripts were present. We did not perform a quantitative assessment of each mutated PCR product; however, global assessment was similar to control patients. Because the 1A isoform of DMT1 was not significantly expressed in the liver and the 1B-IRE and 1B-nonIRE DMT1 isoforms were expressed in human liver, we studied the consequences of the two mutations on cellular trafficking in the HUH7 hepatic cell line expressing both IRE and nonIRE isoforms. Assessing cellular trafficking of the protein by co-localization studies, we found that the subcellular localization of DMT1 wild-type isoforms was in agreement with previously published data [6,8–10]. In addition, the DMT1 deleted protein (del12 DMT1 1B-nonIRE), used as control, led to an abnormal protein trafficking with co-localization to endoplasmic reticulum, as previously described [18]. Our data showed that the
248
E. Bardou-Jacquet et al. / Blood Cells, Molecules, and Diseases 47 (2011) 243–248
N491S mutation caused improper protein trafficking similar to findings reported with the del12 DMT1 1B-nonIRE. They suggest a loss of function resulting from disturbed protein trafficking. Further experiments are warranted to define the underlying mechanism explaining the critical role of the asparagine residue at position 491. Our results suggested that the G212V did not affect cellular trafficking of the protein, supporting an altered iron transport ability and implying that the glycine residue at position 212 plays an important role in the iron transport function of DMT1. The report by Vecchi et al. [25] in 2010 showing an HFE mutation within the HUH7 cell line could be relevant to the mutation findings in our data. However, no direct interaction has been identified between HFE and DMT1 proteins. Moreover, the identified biological role of HFE is to participate in the transduction of signals linked to iron status and regulating hepcidin expression. In our work studying the effect of mutations within SLC11A2, we did not use experimental conditions with modulation of iron status, which could alter the results. Therefore, it is likely that the HFE gene mutation within the HuH7 cell line genome did not alter the outcome of our experiments. The development of iron overload in the present case of SLC11A2 mutations raises concerns about a potential residual activity of DMT1 which, up-regulated in this situation, could lead to iron overload [12]. Thus, assessment of residual transport activity, which could participate in iron overload, should be carried on G212V DMT1 forms for which the protein localization is not disturbed. Alternatively, a complete loss of function could suggest that a compensatory pathway for iron absorption may exist that has yet to be identified. In addition, the fact that non-transferrin-bound iron was found in plasma suggests that this form of iron, which is avidly taken up by hepatocytes [23], could favor the development of hepatic iron overload. In this case, the iron route likely involves a ZIP14 pathway and not an endocytic pathway. Therefore, an iron sequestration in endocytic vesicles could not explain the low ferritin increase, as previously suggested. In conclusion, our data confirm the major role of DMT1 in the maintenance of iron homeostasis in humans and demonstrate that the N491S mutation, through a deleterious effect on protein trafficking, contributes together with the G212V mutation to the development of anemia and hepatic iron overload. Supplementary materials related to this article can be found online at doi:10.1016/j.bcmd.2011.07.004. Fundings EBJ was granted by the French National Academy of Medicine. This work was supported by the LSHM-CT-2006-037296 European Community Grant (Euroiron1), the Association Fer et Foie. Acknowledgments The author would like to thank Dr. Monika Horvathova, Faculty of Medicine Palacky University Olomouc, Czech Republic, and Pr. Prem Ponka, McGill University, Montréal, Canada, for the generous gift of plasmid constructs. We would also like to thank Stéphanie Dutertre from the microscopy platform, IFR GFAS, Rennes.
References [1] H. Gunshin, B. Mackenzie, U.V. Berger, et al., Cloning and characterization of a mammalian proton-coupled metal-ion transporter, Nature 388 (1997) 482–488. [2] V. Picard, G. Govoni, N. Jabado, P. Gros, Nramp 2 (DCT1/DMT1) expressed at the plasma membrane transports iron and other divalent cations into a calceinaccessible cytoplasmic pool, J Biol Chem 275 (2000) 35738–35745. [3] N. Hubert, M.W. Hentze, Previously uncharacterized isoforms of divalent metal transporter (DMT)-1: implications for regulation and cellular function, Proc Natl Acad Sci U S A 99 (2002) 12345–12350. [4] P.L. Lee, T. Gelbart, C. West, et al., The human Nramp2 gene: characterization of the gene structure, alternative splicing, promoter region and polymorphisms, Blood Cells Mol Dis 24 (1998) 199–215. [5] B. Mackenzie, H. Takanaga, N. Hubert, et al., Functional properties of multiple isoforms of human divalent metal-ion transporter 1 (DMT1), Biochem J 403 (2007) 59–69. [6] M. Tabuchi, N. Tanaka, J. Nishida-Kitayama, et al., Alternative splicing regulates the subcellular localization of divalent metal transporter 1 isoforms, Mol Biol Cell 13 (2002) 4371–4387. [7] F. Canonne-Hergaux, S. Gruenheid, P. Ponka, P. Gros, Cellular and subcellular localization of the Nramp2 iron transporter in the intestinal brush border and regulation by dietary iron, Blood 93 (1999) 4406–4417. [8] S. Lam-Yuk-Tseung, P. Gros, Distinct targeting and recycling properties of two isoforms of the iron transporter DMT1 (NRAMP2, Slc11A2), Biochemistry 45 (2006) 2294–2301. [9] F. Canonne-Hergaux, A.S. Zhang, P. Ponka, P. Gros, Characterization of the iron transporter DMT1 (NRAMP2/DCT1) in red blood cells of normal and anemic mk/ mk mice, Blood 98 (2001) 3823–3830. [10] N. Touret, W. Furuya, J. Forbes, et al., Dynamic traffic through the recycling compartment couples the metal transporter Nramp2 (DMT1) with the transferrin receptor, J Biol Chem 278 (2003) 25548–25557. [11] M.P. Mims, Y. Guan, D. Pospisilova, et al., Identification of a human mutation of DMT1 in a patient with microcytic anemia and iron overload, Blood 105 (2005) 1337–1342. [12] E. Blanco, C. Kannengiesser, B. Grandchamp, et al., Not all DMT1 mutations lead to iron overload, Blood Cells Mol Dis 43 (2009) 199–201. [13] A. Iolascon, M. d'Apolito, V. Servedio, et al., Microcytic anemia and hepatic iron overload in a child with compound heterozygous mutations in DMT1 (SCL11A2), Blood 107 (2006) 349–354. [14] C. Beaumont, J. Delaunay, G. Hetet, et al., Two new human DMT1 gene mutations in a patient with microcytic anemia, low ferritinemia, and liver iron overload, Blood 107 (2006) 4168–4170. [15] K. Thompson, R.M. Molina, J.D. Brain, M. Wessling-Resnick, Belgrade rats display liver iron loading, J Nutr 136 (2006) 3010–3014. [16] P. Brissot, M. Bourel, D. Herry, et al., Assessment of liver iron content in 271 patients: a reevaluation of direct and indirect methods, Gastroenterology 80 (1981) 557–565. [17] C. Le Lan, O. Loreal, T. Cohen, et al., Redox active plasma iron in C282Y/C282Y hemochromatosis, Blood 105 (2005) 4527–4531. [18] M. Priwitzerova, G. Nie, A.D. Sheftel, et al., Functional consequences of the human DMT1 (SLC11A2) mutation on protein expression and iron uptake, Blood 106 (2005) 3985–3987. [19] D. Tchernitchko, M. Bourgeois, M.E. Martin, C. Beaumont, Expression of the two mRNA isoforms of the iron transporter Nramp2/DMTI in mice and function of the iron responsive element, Biochem J 363 (2002) 449–455. [20] J.P. Liuzzi, F. Aydemir, H. Nam, et al., Zip14 (Slc39a14) mediates non-transferrinbound iron uptake into cells, Proc Natl Acad Sci U S A 103 (2006) 13612–13617. [21] H. Zoller, R.O. Koch, I. Theurl, et al., Expression of the duodenal iron transporters divalent-metal transporter 1 and ferroportin 1 in iron deficiency and iron overload, Gastroenterology 120 (2001) 1412–1419. [22] D. Trinder, P.S. Oates, C. Thomas, et al., Localisation of divalent metal transporter 1 (DMT1) to the microvillus membrane of rat duodenal enterocytes in iron deficiency, but to hepatocytes in iron overload, Gut 46 (2000) 270–276. [23] P. Brissot, T.L. Wright, W.L. Ma, R.A. Weisiger, Efficient clearance of nontransferrin-bound iron by rat liver. Implications for hepatic iron loading in iron overload states, J Clin Invest 76 (1985) 1463–1470. [24] T.L. Wright, P. Brissot, W.L. Ma, R.A. Weisiger, Characterization of non-transferrinbound iron clearance by rat liver, J Biol Chem 261 (1986) 10909–10914. [25] C. Vecchi, G. Montosi, A. Pietrangelo, Huh-7: a human “hemochromatotic” cell line, Hepatology 51 (2010) 654–659.