- Email: [email protected]
Mutations in the SORT1 gene are unlikely to cause autosomal dominant hypercholesterolemia
Atherosclerosis 225 (2012) 370e375
Contents lists available at SciVerse ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
Mutations in the SORT1 gene are unlikely to cause autosomal dominant hypercholesterolemia Kristian Tveten, Thea Bismo Strøm, Jamie Cameron, Knut Erik Berge, Trond P. Leren* Unit for Cardiac and Cardiovascular Genetics, Department of Medical Genetics, Oslo University Hospital Rikshospitalet, Oslo, Norway
a r t i c l e i n f o
a b s t r a c t
Article history: Received 2 August 2012 Received in revised form 25 September 2012 Accepted 1 October 2012 Available online 11 October 2012
Objective: To study whether mutations in the SORT1 gene could be a cause of autosomal dominant hypercholesterolemia and to study the effect of sortilin on the binding and internalization of low density lipoprotein (LDL). Methods: 842 unrelated hypercholesterolemic subjects without mutations in genes known to cause autosomal dominant hypercholesterolemia, were screened for mutations in the SORT1 gene by DNA sequencing. Transfections of wild-type or mutant SORT1 plasmids in HeLa T-REx cells and the use of siRNA were used to study the effect of sortilin on the number of cell-surface LDL receptors and on the binding and internalization of LDL. Results: A total of 45 mutations in the SORT1 gene were identified of which 15 were missense mutations. Eight of these were selected for in vitro studies, of which none had a major impact on the amount of LDL bound to the cell surface. There was a positive correlation between the amount of sortilin on the cell surface and the amount of LDL bound. The observation that a mutant sortilin which is predominantly found on the cell surface rather than in post-Golgi compartments, bound very high amounts of LDL, indicates that sortilin does not increase the binding of LDL through an intracellular mechanism. Rather, our data indicate that sortilin binds LDL on the cell surface. Conclusion: Even though sortilin binds and internalizes LDL by receptor-mediated endocytosis, mutations in the SORT1 gene are unlikely to cause autosomal dominant hypercholesterolemia and may only have a marginal effect on plasma LDL cholesterol levels. Ó 2012 Elsevier Ireland Ltd. All rights reserved.
Keywords: Cholesterol Endocytosis LDL receptor Mutation SORT1 Sortilin
1. Introduction Sortilin is a Type 1 transmembrane protein which acts as a multiligand receptor [1,2]. It is predominantly found in the transGolgi network and in early endosomes and less than 10% is on the plasma membrane [3]. The main function of sortilin is to transport ligands between the trans-Golgi network and the early endosomes, but also to bind and internalize various ligands across the cell membrane by receptor-mediated endocytosis [2,4]. Sortilin is a proprotein encoded by the SORT1 gene on chromosome 1p13.3 [1]. The 800 amino acid precursor is cleaved by furin in the Golgi apparatus to release a 44 amino acid propeptide [1]. The mature protein consists of a vacuolar protein sorting 10 protein (Vps10p) domain which is the ligand-binding domain, a 22 amino acid transmembrane domain and a 53 amino acid * Corresponding author. Unit for Cardiac and Cardiovascular Genetics, Department of Medical Genetics, Oslo University Hospital Rikshospitalet, P.O. Box 4950, Nydalen, NO-0424 Oslo, Norway. Tel.: þ47 23075552; fax: þ47 23075561. E-mail addresses: [email protected], trond.leren@ rikshospitalet.no (T.P. Leren). 0021-9150/$ e see front matter Ó 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.atherosclerosis.2012.10.026
cytoplasmic domain [1]. The Vps10p domain consists of a 10-bladed b-propeller that is also found in four other structurally related proteins involved in intracellular transport. It is the tunnel of the b-propeller that interacts with ligands, but this interaction is prevented by the uncleaved prodomain which masks the tunnel until sortilin reaches the Golgi apparatus [2]. Thus, it is only after the prodomain has been released in the Golgi apparatus, that sortilin may bind ligands for transport to the early endosomes. The cytoplasmic domain contains several motifs that interact with adapter proteins to shuttle sortilin from the trans-Golgi network to the endosomes and to concentrate sortilin in clathrin-coated pits on the cell membrane [1]. Whereas, sortilin first was implicated in intracellular transport in neurons of the central and peripheral nervous system, it also functions in several other tissues, such as the liver, and has recently been found to play a role in lipid metabolism. The first clues for a role in lipid metabolism came from genome-wide association studies showing that a locus on chromosome 1p13.3 was associated with levels of low density lipoprotein (LDL) cholesterol and the risk of myocardial infarction [5,6].
K. Tveten et al. / Atherosclerosis 225 (2012) 370e375
This locus on chromosome 1p13.3 contains the four genes cadherin EGF LAG seven-pass G-type receptor 2 (CELSR2), proline/ serine-rich coiled-coil 1 (PSRC1), myosin binding protein H-like (MYBPHL) and SORT1, and subsequent studies have shown that it is SORT1 that affects lipid metabolism [7]. A key finding in this respect was that the single nucleotide polymorphism rs12740374 between the CELSR2 and PSRC1 genes alters a predicted binding site for the transcription factor CCAAT/enhancer binding protein (C/EBP), with the major allele disrupting the binding site and the minor allele creating it [7]. As a consequence, the minor allele increases the expression of the SORT1 gene [7]. The effect of sortilin on very low density lipoprotein (VLDL) synthesis, as determined by overexpression and knockdown studies in mouse models, is conflicting [4,7e9]. Musunuru et al. [7] found that sortilin reduced the synthesis of VLDL and thereby reduced the levels of LDL cholesterol, whereas Kjolby et al. [8] found that sortilin increased the synthesis of VLDL and thereby increased the levels of LDL cholesterol. Differences in the genetic make-up of the two mouse models have been suggested as a possible explanation for this discrepancy. Moreover, LinselNitschke et al. [10] found increased amounts of LDL internalized in HEK293 cells transfected with a SORT1 plasmid. However, whether this reflects sortilin-mediated binding and internalization of LDL at the cell surface or if sortilin increases the number or functionality of the LDL receptors by some intracellular effect, is unknown. Thus, the exact mechanism for the effect of sortilin on lipid metabolism remains to be determined. Mutations in the LDL receptor (LDLR) gene cause familial hypercholesterolemia which is an autosomal dominant form of hypercholesterolemia characterized by xanthomas and premature coronary heart disease [11]. From the data of Linsel-Nitschke et al. [10] showing that overexpression of sortilin increased the amount of LDL internalized, one may speculate that mutations in the SORT1 gene could reduce the amount of LDL internalized and thereby cause autosomal dominant hypercholesterolemia. In this study we have screened severe hypercholesterolemic subjects without mutations in genes known to cause autosomal dominant hypercholesterolemia, for mutations in the SORT1 gene. Moreover, the impact of wild-type (WT) SORT1 and selected mutations in the SORT1 gene on the binding and internalization of LDL, has been studied. 2. Material and methods
371
listed in Supplementary Table S1. The conditions for thermal cycling are available upon request. 2.3. Cell culture HeLa T-REx cells (Invitrogen, Carlsbad, CA) were cultured in Modified Eagle’s medium (PAA Laboratories GmbH, Pasching, Austria) containing streptomycin (50 mg/ml), penicillin (50 U/ml), Lglutamine (2 mM) and 10% fetal calf serum (Invitrogen), in a humidified atmosphere (37 C, 5% CO2). 2.4. Mutageneses and transfections The human SORT1 cDNA (OriGene, Rockville, MD) was cloned into pIRES2-AcGFP1 plasmid (Clontech Laboratories, Inc., Mountain View, CA) using SacI and XmaI restriction sites to generate pIRES2WT-SORT1. The pIRES2-AcGFP1 is a bicistronic vector where the gene of interest and the Aequoirea coerulescens green fluorescent protein 1 (AcGFP1) gene are translated from a single mRNA. This allows selection of only transfected cells for the flow cytometric analyses. pIRES2-WT-SORT1 was used as a template for mutagenesis to generate mutant SORT1 plasmids. The primer sequences used for mutageneses are listed in Supplementary Table S2 and mutagenesis was carried out using QuickChange XL Mutagenesis Kit (Stratagene, La Jolla, CA) according to the manufacturer’s instructions. The integrity of the plasmids was confirmed by DNA sequencing. HeLa T-REx cells were transiently transfected using FuGENE HD (Roche Diagnostics GmbH, Mannheim, Germany) according to the manufacturer’s instructions. Small interfering (si)RNAs targeting the SORT1 gene or the LDLR gene as well as control siRNAs were obtained from Qiagen (Qiagen GmbH, Hilden, Germany) and transfections were performed using Lipofectamine RNAiMAX (Invitrogen) according to the manufacturer’s instructions. 2.5. Western blot analysis of sortilin Western blot analysis was carried out essentially as previously described [11]. Briefly, cell lysates were obtained and run on 4e20% TriseHCl Criterion Precast Gels (Bio-Rad, Hercules, CA) and blotted onto Immuno-Blot PVDF Membranes (Bio-Rad, Hercules, CA). The membranes were immunostained with a rabbit anti-sortilin antibody (Abcam, Cambridge, UK).
2.1. Subjects 2.6. Quantitation of the amount of cell-surface sortilin and LDLR 842 unrelated hypercholesterolemic subjects referred for genetic testing with respect to familial hypercholesterolemia were included. None of these subjects were heterozygous for a mutation in the LDLR gene or in the proprotein convertase subtilisin/kexin type 9 (PCSK9) gene or were heterozygous for mutation R3500Q in the apolipoprotein B-100 gene, as determined by standard Sanger DNA sequencing. There were 537 females and 305 males and their mean (SD) age was 54.5 (11.6) years. Their mean (SD) levels of total serum cholesterol before lipid-lowering therapy was started, were 10.2 (1.3) mmol/l.
The amount of cell-surface LDLR on HeLa T-REx cells was determined by flow cytometry as previously described [12]. Briefly, the cells were cultured in medium containing 5 mg/ml lipoproteindeficient serum for 24 h to increase the expression of the LDLR. The cells were then incubated with a mouse anti-LDLR monoclonal antibody IgG-C7 (Progen Biotechnik GmbH, Heidelberg, Germany) or a goat anti-sortilin antibody (R&D Systems, Inc., Minneapolis, MN) for 1 h at 4 C, washed and incubated with Alexa FluorÒ 647 goat anti-mouse IgG or Alexa FluorÒ 647 donkey anti-goat IgG (Molecular Probes, Eugene OR), respectively for 30 min at 4 C.
2.2. DNA sequencing of the SORT1 gene 2.7. Quantitation of the amount of LDL bound and internalized A total of 19 amplicons that together spanned the 20 exons with flanking intron sequences of the SORT1 gene were amplified by polymerase chain reaction and subjected to DNA sequencing using BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems Inc., Foster City, CA) and analyzed on a 3730XL Genetic Analyzer (Applied Biosystems Inc., Foster City, CA). The primer sequences are
The amount of LDL bound and internalized in HeLa T-REx cells was determined by flow cytometry. Briefly, the cells were cultured in medium containing 5 mg/ml lipoprotein-deficient serum for 24 h to increase the expression of the LDLR. The cells were then washed and incubated with 10 mg/ml of 1,10
372
K. Tveten et al. / Atherosclerosis 225 (2012) 370e375
-dioctadecyl-3,3,30 ,30 -tetramethylindodicarbocyanine perchlorate (DiD)-labeled LDL for 2 h at 4 C to study the amount of LDL bound and for 2 h at 37 C to study the amount of LDL internalized. 3. Results 3.1. Identification of mutations in the SORT1 gene among hypercholesterolemic subjects 842 unrelated subjects referred for genetic testing with respect to autosomal dominant hypercholesterolemia were screened for mutations in the SORT1 gene by DNA sequencing. A total of 45 mutations were identified of which 23 were located in exons and 22 were located in introns (Table 1). 23 of the 45 mutations have been reported in the database dbSNP [13] and one additional mutation has been reported in the Exome Sequencing project [14]. The remaining 21 mutations have, to our knowledge, not previously been reported. Of the exonic mutations, 15 were missense mutations and 8 were silent mutations (Table 1). Of the intronic
mutations, two were at positions þ5 and 4, respectively, and the rest were located deeper in introns. No non-sense mutations and no mutations altering the reading frame were identified. Pathogenicity of the missense mutations was assessed by the use of the software programs PolyPhen2 (http://genetics.bwh. harvard.edu/pph2/) and SIFT (http://sift.jcvi.org/) (Table 1). Mutations G310E, D358Y, V407M, R411C and H822R were predicted to be pathogenic by both programs. However, such predictions should be interpreted with care. Splice-site mutation 832 þ 5G > A was predicted to affect RNA splicing by the use of three software programs (NNSPLICE, MaxEntScan and SpliceSiteFinder-like) but not by a fourth (HumanSplicingFinder). Splice-site mutation 1475 4G > T was not predicted to affect RNA splicing by any of the four software programs. Because relevant family members were not available to study whether mutations in the SORT1 gene segregated with hypercholesterolemia, the clinical significance of the identified mutations is uncertain. A total of 51 subjects carried a least one missense mutation in the SORT1 gene. The mean level of total serum cholesterol among these subjects was 10.2 (1.4) mmol/l and the corresponding level among the remaining
Table 1 Mutations detected by DNA sequencing of the SORT1 gene. PolyPhen2 and SIFT scores are indicated as well as refSNP cluster ID (rs) numbers [13] and Exome sequencing project (ESP) numbers [14]. Mutation Exon/intron
p.-Notation
c.-Notation
Exon 1 Exon 1 Exon 1 Exon 1 Exon 1 Exon 2 Intron 2 Exon 3 Intron 3 Intron 3 Intron 3 Exon 4 Intron 4 Intron 4 Exon 5 Exon 5 Exon 5 Intron 6 Intron 7 Intron 7 Exon 8 Exon 8 Exon 9 Exon 9 Intron 9 Exon 10 Exon 10 Exon 10 Exon 10 Intron 10 Exon 11 Exon 11 Intron 11 Intron 11 Intron 12 Intron 13 Exon 15 Intron 15 Intron 15 Intron 16 Intron 17 Exon 19 Intron 19 Intron 19 Intron 19
P4P G10G P15S P70P S78R S111S
c.12C > G c.30C > T c.43C > T c.210C > G c.234C > A c.333A > G c.366 þ 37A > G c.370A > G c.440 þ 17G > A c.440 þ 65T > A c.441 108C > T c.442G > A c.543 þ 46T > C c.543 þ 52G > C c.594C > T c.597A > G c.653C > A c.782 þ 52C > T c.832þ5G > A c.833 62A > G c.904A > G c.929G > A c.969C > A c.1072G > T c.1109-76_1109-75insAT c.1133C > T c.1206G > A c.1219G > A c.1231C > T c.1265 37A > G c.1330G > C c.1340A > G c.1371 þ 38T > G c.1372 182A > G c.1475 4G > T c.1643 þ 42G > A c.1888T > C c.2024 þ 35T > G c.2025 84A > G c.2141 þ 53G > A c.2251 26T > C c.2465A > G c.2481 þ 16C > T c.2481 þ 40C > A c.2481 þ 96C > G
I124V
E148K
F198F R199R T218N
K302E G310E T323T D358Y T378I T402T V407M R411C E444Q E447G
Y630H
H822R
PolyPhen2
SIFT
rs/ESP
e e Benign e Benign e e Benign e e e Benign e e e e Benign e e e Benign Possibly damaging e Possibly damaging e Benign e Probably damaging Probably damaging e Benign Benign e e e e Benign e e e e Possibly damaging e e e
e e Tolerated e Not tolerated e e Tolerated e e e Tolerated e e e e Tolerated e e e Tolerated Not tolerated e Not Tolerated e Tolerated e Not tolerated Not tolerated e Not tolerated Not tolerated e e e e Tolerated e e e e Not tolerated e e e
72646553 150163924
145159492 61797119 11581665 72646558
11142 72646560
141749679 2228604 78583579 3217347 ESP_1_109883404 200994282 72646566 72646568 144141753 67195750 11102970 201962979
200737042 72646581 146391412
661278
K. Tveten et al. / Atherosclerosis 225 (2012) 370e375
791 subjects was 10.2 (1.3) mmol/l. Thus, there was no apparent effect of these missense mutations on the level of total serum cholesterol. 3.2. Effect of mutations in the SORT1 gene on the binding of LDL If mutations in the SORT1 gene cause autosomal dominant hypercholesterolemia, they would be expected to the increase the synthesis of LDL or to reduce the binding and internalization of LDL. To determine the effect of the different mutant sortilins on the binding of LDL, HeLa T-REx cells were transiently transfected with WT or mutant SORT1 plasmids and the amount of LDL bound to the cells at 4 C was determined by flow cytometry. Eight missense mutations with various combinations of PolyPhen2 and SIFT scores were selected for these analyses (Table 2). The amount of sortilin on the cell surface of cells transfected with the WT-SORT1 plasmid was approximately 25 fold higher than that of cells transfected with empty plasmid, and the amount of LDL bound was 2e3-fold higher in cells transfected with the WTSORT1 plasmid (Table 2). Thus, sortilin increases the binding of LDL. For the eight missense mutations in the SORT1 gene, the amounts of sortilin on the cell-surface relative to that of cells transfected with the WT-SORT1 plasmid, varied from 0.11 to 1.06 (Table 2). The corresponding variation in the amount of LDL bound, was 0.53e 1.14. Considering all mutants, there was a significant positive correlation between the amounts of sortilin on the cell surface and the amount of LDL bound (r ¼ 0.79, p < 0.01). Thus, the effect of the different mutations in the SORT1 gene on the amount of LDL bound, may reflect differences in the amount of sortilin on the cell surface. The observation that cells transfected with an LDLR plasmid bound more than 6-fold more LDL than cells transfected with the WT-SORT1 plasmid (Table 2), may suggest that sortilin plays a minor role in the binding of LDL as compared to that of the LDLR. 3.3. Effect of silencing the SORT1 gene on internalization of LDL To further study the effect of sortilin on the amount of LDL internalized, siRNA was used to study the effect of silencing the SORT1 gene. Except for SORT1_5 which resulted in a 36% reduction in the amount of LDL internalized, the other siRNAs had only a marginal effect (Fig. 1). Moreover, no apparent correlation was observed between the level of sortilin knockdown and the amount of LDL internalized. For comparison, siRNA against the LDLR gene resulted in a 71% reduction in the amount of LDL internalized (Fig. 1).
Table 2 Effect of selected SORT1 mutations on the amount of cell-surface sortilin and on the amount of LDL bound (mean SD). Mutation/controls
Cell-surface sortilin
Bound LDL
Mutation WT-SORT1 P15S G310E D358Y T378I V407M R411C Y630H H822R
1.00 0.19 0.54 0.11 0.82 0.73 1.06 0.80 0.47
1.00 0.65 1.02 0.53 1.12 1.04 1.11 1.14 1.14
Controls Untransfected Empty plasmid LDLR
0.06 (0.03) 0.04 (0.02) 0.07 (0.04)
(0.09) (0.08) (0.08) (0.05) (0.11) (0.09) (0.18) (0.11)
(0.19) (0.27) (0.17) (0.26) (0.32) (0.26 (0.34) (0.43)
0.76 (0.22) 0.39 (0.12) 6.51 (1.64)
373
3.4. Effect of overexpression of sortilin on internalization of LDL As a complementary strategy to study the role of sortilin in the internalization of LDL, HeLa T-REx cells were transfected with the WT-SORT1 plasmid. As compared to cells transfected with empty plasmid, cells transfected with the WT-SORT1 plasmid internalized 20% more LDL, whereas cells transfected with the LDLR plasmid internalized 156% more LDL (Fig. 2). 3.5. Effect of a mutant sortilin lacking the cytoplasmic domain on the binding of LDL To study whether sortilin increased the amount of LDL bound through some intracellular effect, HeLa T-REx cells were transfected with a SORT1 plasmid lacking the sequence for the cytoplasmic domain (SORT1DCD). By deleting the cytoplasmic domain, sortilin is no longer transported from the trans-Golgi network to the endosomes/lysosomes and is predominantly located on the cell surface [14]. The amounts of cell-surface sortilin and LDLR, as well as the amount of LDL bound, were determined by flow cytometry of transfected cells. As compared to cells transfected with the WTSORT1 plasmid, cells transfected with the SORT1DCD plasmid exhibited 19-fold more sortilin on the cell surface and they bound a 3-fold higher amount of LDL (Table 3). Because it is unlikely that the increased amount of LDL bound to cells transfected with the SORT1DCD plasmid was due to the 21% increased amount of cellsurface LDLRs, it appears that sortilin itself binds LDL without acting through an intracellular mechanism. For comparison, HeLa T-REx cells transfected with the LDLR plasmid bound 7-fold more LDL than cells transfected with the WT-SORT1 plasmid (Table 3). Similar findings were observed when these experiments were performed in HepG2 cells (data not shown). 4. Discussion In this study we have screened 842 unrelated hypercholesterolemic subjects for mutations in the SORT1 gene. None of the subjects were heterozygous for a mutation in the LDLR gene or in the PCSK9 gene and none carried the R3500Q mutation in the apolipoprotein B-100 gene. 45 mutations were identified in the SORT1 gene of which 15 were missense mutations. Subjects carrying missense mutations had similar levels of total serum cholesterol as the non-carriers. Thus, there was no apparent effect of these missense mutations on the levels of total serum cholesterol. Eight of the missense mutations were selected for in vitro studies to study their effect on the binding of LDL. Overexpression of sortilin markedly increased the amount of sortilin on the cell surface and the amount of LDL bound. Even though the eight sortilin mutants affected the binding of LDL differently, this variation seemed to result from differences in the amount of sortilin on the cell surface which could be due to differences in the level of expression. The positive correlation between the amount of sortilin on the cell surface and the amount of LDL bound, suggests that sortilin binds LDL on the cell surface. Overexpression of sortilin in HeLa T-REx cells resulted in a 20% increased amount of LDL internalized. This finding is in agreement with the findings of Linsel-Nitschke et al. [10] who observed a 23% increased amount of LDL internalized in HEK293 cells. They reasoned that because sortilin is primarily located intracellularly in post-Golgi compartments, the most likely mechanism for the increased amount of LDL internalized was through some indirect effects. However, our data showing that high amounts of LDL were bound to a mutant sortilin lacking the cytoplasmic domain, which fails to be shuttled from the trans-Golgi network to endosomes/
374
K. Tveten et al. / Atherosclerosis 225 (2012) 370e375
Fig. 1. Effect of knocking down sortilin on internalization of LDL. Four different siRNAs against the SORT1 gene were used to knock down sortilin expression in HeLa T-REx cells. The amount of LDL internalized was then determined by flow cytometry. The amount of LDL internalized in cells transfected with control siRNA was assigned a value of 1. Mean (SD) values of three separate experiments are shown. The amount of sortilin in cell lysates was determined by Western blot analysis. One representative Western blot is shown.
3.50
LDL internalization
3.00
2.56
2.50 2.00 1.50
1.23
1.20 1.00
1.00 0.50 0.00 Untransfected
Empty
SORT1
LDLR
Fig. 2. Effect of overexpression of sortilin on internalization of LDL. HeLa T-REx cells were transiently transfected with a bicistronic plasmid containing cDNA of the SORT1 gene, the LDLR gene or with an empty plasmid, and the amount of LDL internalized was determined by flow cytometry. The amount of LDL internalized in cells transfected with empty plasmid was assigned a value of 1. Mean (SD) values of three separate experiments are shown.
lysosomes [15], makes it unlikely that sortilin acts by some intracellular mechanism to increase the number of LDLRs. The ability of sortilin to bind LDL has also been shown by studies of plasmon surface resonance [8]. Moreover, a recent study that was published during preparation of this manuscript has also shown that sortilin is able to bind and internalize LDL [16]. Thus, sortilin binds and internalizes LDL by receptor-mediated endocytosis. The question then regards the role of sortilinmediated endocytosis of LDL in the regulation of plasma LDL
Table 3 Effect of SORT1 lacking the cytoplasmic domain (SORT1DCD) on the amount of cellsurface sortilin and LDLR and on the amount of LDL bound (mean SD). Plasmid
Cell-surface sortilin
Cell-surface LDLR
Bound LDL
WT-SORT1 SORT1DCD LDLR Untransfected Empty plasmid
1.00 19.1 0.07 0.03 0.02
1.00 1.22 (0.30) 159.5 (6.65) 1.23 (0.22) 0.65 (0.03)
1.00 3.33 6.75 0.90 0.35
(1.58) (0.004) (0.002) (0.004)
(0.25) (1.27) (0.14) (0.09)
cholesterol levels. From our data where the SORT1 gene or the LDLR gene have been silenced or overexpressed, it appears that the LDLR plays a much more significant role in receptor-mediated endocytosis of LDL than sortilin. This notion is also supported by the very high levels of LDL cholesterol of approximately 20 mmol/l in familial hypercholesterolemia homozygotes who lack functioning LDLRs [11], but who are presumed to have normal amounts of sortilin. Heterozygosity for a loss-of-function mutation in the SORT1 gene would be expected to result in a 50% reduction in the amount of functional sortilin on the cell surface. Based upon the results from genome-wide association studies, where homozygosity for the major allele at the chromosome 1p13.3 locus was associated with a more than 90% reduced expression of sortilin in human liver as compared to homozygosity for the minor allele [7], and a mere 0.4 mmol/l increased level of LDL cholesterol [7], it is unlikely that a 50% reduction in the amount of sortilin by a loss-of-function mutation would have much impact on the levels of LDL cholesterol. Based upon this reasoning, our mutation screening of the SORT1 gene and subsequent in vitro testing of identified missense mutations, we consider it unlikely that mutations in the SORT1 gene cause autosomal dominant hypercholesterolemia. Rather, loss-of-function mutations in the SORT1 gene may have only a marginal effect on the levels of LDL cholesterol. Appendix A. Supplementary material Supplementary material associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j. atherosclerosis.2012.10.026. References [1] Petersen CM, Nielsen M, Nykjær A, et al. Molecular identification of a novel candidate sorting receptor purified from human brain by receptor-associated protein affinity chromatography. J Biol Chem 1997;272:3599e605. [2] Quistgaard E, Madsen P, Grøftehauge MK, Nissen P, Petersen C, Thirup SS. Ligands bind to sortilin in the tunnel of a ten-bladed b-propeller domain. Nat Struct Mol Biol 2009;16:96e8.
K. Tveten et al. / Atherosclerosis 225 (2012) 370e375 [3] Willnow TE, Petersen CM, Nykjær A. VPS10P-domain receptors e regulators of neuronal viability and function. Nat Rev Neurosci 2008;9:899e909. [4] Willnow TE, Kjølby M, Nykjær A. Sortilins: new players in lipoprotein metabolism. Curr Opin Lipidol 2011;22:79e85. [5] Samani NJ, Erdmann J, Hall AS, et al. Genomewide association analysis of coronary artery disease. N Engl J Med 2007;357:443e53. [6] Myocardial Infarction Genetics Consortium. Genome-wide association of early-onset myocardial infarction with single nucleotide polymorphisms and copy number variants. Nat Genet 2009;41:334e41. [7] Musunuru K, Strong A, Frank-Kamenetsky, et al. From noncoding variant to phenotype via SORT1 at the 1p13 cholesterol locus. Nature 2010;466:714e9. [8] Kjolby M, Andersen OM, Breiderhoff T, et al. Sort1, encoded by the cardiovascular risk locus 1p13.3, is a regulator of hepatic lipoprotein export. Cell Metab 2010;12:213e23. [9] Dubé JB, Johansen CT, Hegele RA. Sortilin: an unusual suspect in cholesterol metabolism. Bioessays 2011;33:430e7. [10] Linsel-Nitschke P, Heeren J, Aherrahrou Z, et al. Genetic variation at chromosome 1p13.3 affects sortilin mRNA expression, cellular LDL-uptake and
[11]
[12]
[13] [14] [15]
[16]
375
serum LDL levels which translates to the risk of coronary heart disease. Atherosclerosis 2010;208:183e9. Goldstein JL, Hobbs HH, Brown MS. Familial hypercholesterolemia. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The metabolic & molecular basis of inherited disease. New York: McGraw-Hill; 2001. p. 2863e914. Cameron J, Holla ØL, Ranheim T, Kulseth MA, Berge KE, Leren TP. Effect of mutations in the PCSK9 gene on the cell surface LDL receptors. Hum Mol Genet 2006;15:1551e8. Sherry ST, Ward MH, Kholodov M, et al. dbSNP: the NCBI database of genetic variation. Nucleic Acids Res 2001;29:308e11. http://browser.1000genomes.org/Homo_sapiens/Gene/Variation_Gene/Table? db¼core;g¼ENSG00000134243;r¼1:109852192-109940573 (Table 1). Nielsen M, Madsen P, Christensen EI, et al. The sortilin cytoplasmic tail conveys Golgi-endosome transport and binds the VHS domain of the GGA2 sorting protein. EMBO J 2001;20:2180e90. Strong A, Ding Q, Edmonson A, et al. Hepatic sortilin regulates both apolipoprotein B secretion and LDL catabolism. J Clin Invest 2012;122:2807e16.
Contents lists available at SciVerse ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
Mutations in the SORT1 gene are unlikely to cause autosomal dominant hypercholesterolemia Kristian Tveten, Thea Bismo Strøm, Jamie Cameron, Knut Erik Berge, Trond P. Leren* Unit for Cardiac and Cardiovascular Genetics, Department of Medical Genetics, Oslo University Hospital Rikshospitalet, Oslo, Norway
a r t i c l e i n f o
a b s t r a c t
Article history: Received 2 August 2012 Received in revised form 25 September 2012 Accepted 1 October 2012 Available online 11 October 2012
Objective: To study whether mutations in the SORT1 gene could be a cause of autosomal dominant hypercholesterolemia and to study the effect of sortilin on the binding and internalization of low density lipoprotein (LDL). Methods: 842 unrelated hypercholesterolemic subjects without mutations in genes known to cause autosomal dominant hypercholesterolemia, were screened for mutations in the SORT1 gene by DNA sequencing. Transfections of wild-type or mutant SORT1 plasmids in HeLa T-REx cells and the use of siRNA were used to study the effect of sortilin on the number of cell-surface LDL receptors and on the binding and internalization of LDL. Results: A total of 45 mutations in the SORT1 gene were identified of which 15 were missense mutations. Eight of these were selected for in vitro studies, of which none had a major impact on the amount of LDL bound to the cell surface. There was a positive correlation between the amount of sortilin on the cell surface and the amount of LDL bound. The observation that a mutant sortilin which is predominantly found on the cell surface rather than in post-Golgi compartments, bound very high amounts of LDL, indicates that sortilin does not increase the binding of LDL through an intracellular mechanism. Rather, our data indicate that sortilin binds LDL on the cell surface. Conclusion: Even though sortilin binds and internalizes LDL by receptor-mediated endocytosis, mutations in the SORT1 gene are unlikely to cause autosomal dominant hypercholesterolemia and may only have a marginal effect on plasma LDL cholesterol levels. Ó 2012 Elsevier Ireland Ltd. All rights reserved.
Keywords: Cholesterol Endocytosis LDL receptor Mutation SORT1 Sortilin
1. Introduction Sortilin is a Type 1 transmembrane protein which acts as a multiligand receptor [1,2]. It is predominantly found in the transGolgi network and in early endosomes and less than 10% is on the plasma membrane [3]. The main function of sortilin is to transport ligands between the trans-Golgi network and the early endosomes, but also to bind and internalize various ligands across the cell membrane by receptor-mediated endocytosis [2,4]. Sortilin is a proprotein encoded by the SORT1 gene on chromosome 1p13.3 [1]. The 800 amino acid precursor is cleaved by furin in the Golgi apparatus to release a 44 amino acid propeptide [1]. The mature protein consists of a vacuolar protein sorting 10 protein (Vps10p) domain which is the ligand-binding domain, a 22 amino acid transmembrane domain and a 53 amino acid * Corresponding author. Unit for Cardiac and Cardiovascular Genetics, Department of Medical Genetics, Oslo University Hospital Rikshospitalet, P.O. Box 4950, Nydalen, NO-0424 Oslo, Norway. Tel.: þ47 23075552; fax: þ47 23075561. E-mail addresses: [email protected], trond.leren@ rikshospitalet.no (T.P. Leren). 0021-9150/$ e see front matter Ó 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.atherosclerosis.2012.10.026
cytoplasmic domain [1]. The Vps10p domain consists of a 10-bladed b-propeller that is also found in four other structurally related proteins involved in intracellular transport. It is the tunnel of the b-propeller that interacts with ligands, but this interaction is prevented by the uncleaved prodomain which masks the tunnel until sortilin reaches the Golgi apparatus [2]. Thus, it is only after the prodomain has been released in the Golgi apparatus, that sortilin may bind ligands for transport to the early endosomes. The cytoplasmic domain contains several motifs that interact with adapter proteins to shuttle sortilin from the trans-Golgi network to the endosomes and to concentrate sortilin in clathrin-coated pits on the cell membrane [1]. Whereas, sortilin first was implicated in intracellular transport in neurons of the central and peripheral nervous system, it also functions in several other tissues, such as the liver, and has recently been found to play a role in lipid metabolism. The first clues for a role in lipid metabolism came from genome-wide association studies showing that a locus on chromosome 1p13.3 was associated with levels of low density lipoprotein (LDL) cholesterol and the risk of myocardial infarction [5,6].
K. Tveten et al. / Atherosclerosis 225 (2012) 370e375
This locus on chromosome 1p13.3 contains the four genes cadherin EGF LAG seven-pass G-type receptor 2 (CELSR2), proline/ serine-rich coiled-coil 1 (PSRC1), myosin binding protein H-like (MYBPHL) and SORT1, and subsequent studies have shown that it is SORT1 that affects lipid metabolism [7]. A key finding in this respect was that the single nucleotide polymorphism rs12740374 between the CELSR2 and PSRC1 genes alters a predicted binding site for the transcription factor CCAAT/enhancer binding protein (C/EBP), with the major allele disrupting the binding site and the minor allele creating it [7]. As a consequence, the minor allele increases the expression of the SORT1 gene [7]. The effect of sortilin on very low density lipoprotein (VLDL) synthesis, as determined by overexpression and knockdown studies in mouse models, is conflicting [4,7e9]. Musunuru et al. [7] found that sortilin reduced the synthesis of VLDL and thereby reduced the levels of LDL cholesterol, whereas Kjolby et al. [8] found that sortilin increased the synthesis of VLDL and thereby increased the levels of LDL cholesterol. Differences in the genetic make-up of the two mouse models have been suggested as a possible explanation for this discrepancy. Moreover, LinselNitschke et al. [10] found increased amounts of LDL internalized in HEK293 cells transfected with a SORT1 plasmid. However, whether this reflects sortilin-mediated binding and internalization of LDL at the cell surface or if sortilin increases the number or functionality of the LDL receptors by some intracellular effect, is unknown. Thus, the exact mechanism for the effect of sortilin on lipid metabolism remains to be determined. Mutations in the LDL receptor (LDLR) gene cause familial hypercholesterolemia which is an autosomal dominant form of hypercholesterolemia characterized by xanthomas and premature coronary heart disease [11]. From the data of Linsel-Nitschke et al. [10] showing that overexpression of sortilin increased the amount of LDL internalized, one may speculate that mutations in the SORT1 gene could reduce the amount of LDL internalized and thereby cause autosomal dominant hypercholesterolemia. In this study we have screened severe hypercholesterolemic subjects without mutations in genes known to cause autosomal dominant hypercholesterolemia, for mutations in the SORT1 gene. Moreover, the impact of wild-type (WT) SORT1 and selected mutations in the SORT1 gene on the binding and internalization of LDL, has been studied. 2. Material and methods
371
listed in Supplementary Table S1. The conditions for thermal cycling are available upon request. 2.3. Cell culture HeLa T-REx cells (Invitrogen, Carlsbad, CA) were cultured in Modified Eagle’s medium (PAA Laboratories GmbH, Pasching, Austria) containing streptomycin (50 mg/ml), penicillin (50 U/ml), Lglutamine (2 mM) and 10% fetal calf serum (Invitrogen), in a humidified atmosphere (37 C, 5% CO2). 2.4. Mutageneses and transfections The human SORT1 cDNA (OriGene, Rockville, MD) was cloned into pIRES2-AcGFP1 plasmid (Clontech Laboratories, Inc., Mountain View, CA) using SacI and XmaI restriction sites to generate pIRES2WT-SORT1. The pIRES2-AcGFP1 is a bicistronic vector where the gene of interest and the Aequoirea coerulescens green fluorescent protein 1 (AcGFP1) gene are translated from a single mRNA. This allows selection of only transfected cells for the flow cytometric analyses. pIRES2-WT-SORT1 was used as a template for mutagenesis to generate mutant SORT1 plasmids. The primer sequences used for mutageneses are listed in Supplementary Table S2 and mutagenesis was carried out using QuickChange XL Mutagenesis Kit (Stratagene, La Jolla, CA) according to the manufacturer’s instructions. The integrity of the plasmids was confirmed by DNA sequencing. HeLa T-REx cells were transiently transfected using FuGENE HD (Roche Diagnostics GmbH, Mannheim, Germany) according to the manufacturer’s instructions. Small interfering (si)RNAs targeting the SORT1 gene or the LDLR gene as well as control siRNAs were obtained from Qiagen (Qiagen GmbH, Hilden, Germany) and transfections were performed using Lipofectamine RNAiMAX (Invitrogen) according to the manufacturer’s instructions. 2.5. Western blot analysis of sortilin Western blot analysis was carried out essentially as previously described [11]. Briefly, cell lysates were obtained and run on 4e20% TriseHCl Criterion Precast Gels (Bio-Rad, Hercules, CA) and blotted onto Immuno-Blot PVDF Membranes (Bio-Rad, Hercules, CA). The membranes were immunostained with a rabbit anti-sortilin antibody (Abcam, Cambridge, UK).
2.1. Subjects 2.6. Quantitation of the amount of cell-surface sortilin and LDLR 842 unrelated hypercholesterolemic subjects referred for genetic testing with respect to familial hypercholesterolemia were included. None of these subjects were heterozygous for a mutation in the LDLR gene or in the proprotein convertase subtilisin/kexin type 9 (PCSK9) gene or were heterozygous for mutation R3500Q in the apolipoprotein B-100 gene, as determined by standard Sanger DNA sequencing. There were 537 females and 305 males and their mean (SD) age was 54.5 (11.6) years. Their mean (SD) levels of total serum cholesterol before lipid-lowering therapy was started, were 10.2 (1.3) mmol/l.
The amount of cell-surface LDLR on HeLa T-REx cells was determined by flow cytometry as previously described [12]. Briefly, the cells were cultured in medium containing 5 mg/ml lipoproteindeficient serum for 24 h to increase the expression of the LDLR. The cells were then incubated with a mouse anti-LDLR monoclonal antibody IgG-C7 (Progen Biotechnik GmbH, Heidelberg, Germany) or a goat anti-sortilin antibody (R&D Systems, Inc., Minneapolis, MN) for 1 h at 4 C, washed and incubated with Alexa FluorÒ 647 goat anti-mouse IgG or Alexa FluorÒ 647 donkey anti-goat IgG (Molecular Probes, Eugene OR), respectively for 30 min at 4 C.
2.2. DNA sequencing of the SORT1 gene 2.7. Quantitation of the amount of LDL bound and internalized A total of 19 amplicons that together spanned the 20 exons with flanking intron sequences of the SORT1 gene were amplified by polymerase chain reaction and subjected to DNA sequencing using BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems Inc., Foster City, CA) and analyzed on a 3730XL Genetic Analyzer (Applied Biosystems Inc., Foster City, CA). The primer sequences are
The amount of LDL bound and internalized in HeLa T-REx cells was determined by flow cytometry. Briefly, the cells were cultured in medium containing 5 mg/ml lipoprotein-deficient serum for 24 h to increase the expression of the LDLR. The cells were then washed and incubated with 10 mg/ml of 1,10
372
K. Tveten et al. / Atherosclerosis 225 (2012) 370e375
-dioctadecyl-3,3,30 ,30 -tetramethylindodicarbocyanine perchlorate (DiD)-labeled LDL for 2 h at 4 C to study the amount of LDL bound and for 2 h at 37 C to study the amount of LDL internalized. 3. Results 3.1. Identification of mutations in the SORT1 gene among hypercholesterolemic subjects 842 unrelated subjects referred for genetic testing with respect to autosomal dominant hypercholesterolemia were screened for mutations in the SORT1 gene by DNA sequencing. A total of 45 mutations were identified of which 23 were located in exons and 22 were located in introns (Table 1). 23 of the 45 mutations have been reported in the database dbSNP [13] and one additional mutation has been reported in the Exome Sequencing project [14]. The remaining 21 mutations have, to our knowledge, not previously been reported. Of the exonic mutations, 15 were missense mutations and 8 were silent mutations (Table 1). Of the intronic
mutations, two were at positions þ5 and 4, respectively, and the rest were located deeper in introns. No non-sense mutations and no mutations altering the reading frame were identified. Pathogenicity of the missense mutations was assessed by the use of the software programs PolyPhen2 (http://genetics.bwh. harvard.edu/pph2/) and SIFT (http://sift.jcvi.org/) (Table 1). Mutations G310E, D358Y, V407M, R411C and H822R were predicted to be pathogenic by both programs. However, such predictions should be interpreted with care. Splice-site mutation 832 þ 5G > A was predicted to affect RNA splicing by the use of three software programs (NNSPLICE, MaxEntScan and SpliceSiteFinder-like) but not by a fourth (HumanSplicingFinder). Splice-site mutation 1475 4G > T was not predicted to affect RNA splicing by any of the four software programs. Because relevant family members were not available to study whether mutations in the SORT1 gene segregated with hypercholesterolemia, the clinical significance of the identified mutations is uncertain. A total of 51 subjects carried a least one missense mutation in the SORT1 gene. The mean level of total serum cholesterol among these subjects was 10.2 (1.4) mmol/l and the corresponding level among the remaining
Table 1 Mutations detected by DNA sequencing of the SORT1 gene. PolyPhen2 and SIFT scores are indicated as well as refSNP cluster ID (rs) numbers [13] and Exome sequencing project (ESP) numbers [14]. Mutation Exon/intron
p.-Notation
c.-Notation
Exon 1 Exon 1 Exon 1 Exon 1 Exon 1 Exon 2 Intron 2 Exon 3 Intron 3 Intron 3 Intron 3 Exon 4 Intron 4 Intron 4 Exon 5 Exon 5 Exon 5 Intron 6 Intron 7 Intron 7 Exon 8 Exon 8 Exon 9 Exon 9 Intron 9 Exon 10 Exon 10 Exon 10 Exon 10 Intron 10 Exon 11 Exon 11 Intron 11 Intron 11 Intron 12 Intron 13 Exon 15 Intron 15 Intron 15 Intron 16 Intron 17 Exon 19 Intron 19 Intron 19 Intron 19
P4P G10G P15S P70P S78R S111S
c.12C > G c.30C > T c.43C > T c.210C > G c.234C > A c.333A > G c.366 þ 37A > G c.370A > G c.440 þ 17G > A c.440 þ 65T > A c.441 108C > T c.442G > A c.543 þ 46T > C c.543 þ 52G > C c.594C > T c.597A > G c.653C > A c.782 þ 52C > T c.832þ5G > A c.833 62A > G c.904A > G c.929G > A c.969C > A c.1072G > T c.1109-76_1109-75insAT c.1133C > T c.1206G > A c.1219G > A c.1231C > T c.1265 37A > G c.1330G > C c.1340A > G c.1371 þ 38T > G c.1372 182A > G c.1475 4G > T c.1643 þ 42G > A c.1888T > C c.2024 þ 35T > G c.2025 84A > G c.2141 þ 53G > A c.2251 26T > C c.2465A > G c.2481 þ 16C > T c.2481 þ 40C > A c.2481 þ 96C > G
I124V
E148K
F198F R199R T218N
K302E G310E T323T D358Y T378I T402T V407M R411C E444Q E447G
Y630H
H822R
PolyPhen2
SIFT
rs/ESP
e e Benign e Benign e e Benign e e e Benign e e e e Benign e e e Benign Possibly damaging e Possibly damaging e Benign e Probably damaging Probably damaging e Benign Benign e e e e Benign e e e e Possibly damaging e e e
e e Tolerated e Not tolerated e e Tolerated e e e Tolerated e e e e Tolerated e e e Tolerated Not tolerated e Not Tolerated e Tolerated e Not tolerated Not tolerated e Not tolerated Not tolerated e e e e Tolerated e e e e Not tolerated e e e
72646553 150163924
145159492 61797119 11581665 72646558
11142 72646560
141749679 2228604 78583579 3217347 ESP_1_109883404 200994282 72646566 72646568 144141753 67195750 11102970 201962979
200737042 72646581 146391412
661278
K. Tveten et al. / Atherosclerosis 225 (2012) 370e375
791 subjects was 10.2 (1.3) mmol/l. Thus, there was no apparent effect of these missense mutations on the level of total serum cholesterol. 3.2. Effect of mutations in the SORT1 gene on the binding of LDL If mutations in the SORT1 gene cause autosomal dominant hypercholesterolemia, they would be expected to the increase the synthesis of LDL or to reduce the binding and internalization of LDL. To determine the effect of the different mutant sortilins on the binding of LDL, HeLa T-REx cells were transiently transfected with WT or mutant SORT1 plasmids and the amount of LDL bound to the cells at 4 C was determined by flow cytometry. Eight missense mutations with various combinations of PolyPhen2 and SIFT scores were selected for these analyses (Table 2). The amount of sortilin on the cell surface of cells transfected with the WT-SORT1 plasmid was approximately 25 fold higher than that of cells transfected with empty plasmid, and the amount of LDL bound was 2e3-fold higher in cells transfected with the WTSORT1 plasmid (Table 2). Thus, sortilin increases the binding of LDL. For the eight missense mutations in the SORT1 gene, the amounts of sortilin on the cell-surface relative to that of cells transfected with the WT-SORT1 plasmid, varied from 0.11 to 1.06 (Table 2). The corresponding variation in the amount of LDL bound, was 0.53e 1.14. Considering all mutants, there was a significant positive correlation between the amounts of sortilin on the cell surface and the amount of LDL bound (r ¼ 0.79, p < 0.01). Thus, the effect of the different mutations in the SORT1 gene on the amount of LDL bound, may reflect differences in the amount of sortilin on the cell surface. The observation that cells transfected with an LDLR plasmid bound more than 6-fold more LDL than cells transfected with the WT-SORT1 plasmid (Table 2), may suggest that sortilin plays a minor role in the binding of LDL as compared to that of the LDLR. 3.3. Effect of silencing the SORT1 gene on internalization of LDL To further study the effect of sortilin on the amount of LDL internalized, siRNA was used to study the effect of silencing the SORT1 gene. Except for SORT1_5 which resulted in a 36% reduction in the amount of LDL internalized, the other siRNAs had only a marginal effect (Fig. 1). Moreover, no apparent correlation was observed between the level of sortilin knockdown and the amount of LDL internalized. For comparison, siRNA against the LDLR gene resulted in a 71% reduction in the amount of LDL internalized (Fig. 1).
Table 2 Effect of selected SORT1 mutations on the amount of cell-surface sortilin and on the amount of LDL bound (mean SD). Mutation/controls
Cell-surface sortilin
Bound LDL
Mutation WT-SORT1 P15S G310E D358Y T378I V407M R411C Y630H H822R
1.00 0.19 0.54 0.11 0.82 0.73 1.06 0.80 0.47
1.00 0.65 1.02 0.53 1.12 1.04 1.11 1.14 1.14
Controls Untransfected Empty plasmid LDLR
0.06 (0.03) 0.04 (0.02) 0.07 (0.04)
(0.09) (0.08) (0.08) (0.05) (0.11) (0.09) (0.18) (0.11)
(0.19) (0.27) (0.17) (0.26) (0.32) (0.26 (0.34) (0.43)
0.76 (0.22) 0.39 (0.12) 6.51 (1.64)
373
3.4. Effect of overexpression of sortilin on internalization of LDL As a complementary strategy to study the role of sortilin in the internalization of LDL, HeLa T-REx cells were transfected with the WT-SORT1 plasmid. As compared to cells transfected with empty plasmid, cells transfected with the WT-SORT1 plasmid internalized 20% more LDL, whereas cells transfected with the LDLR plasmid internalized 156% more LDL (Fig. 2). 3.5. Effect of a mutant sortilin lacking the cytoplasmic domain on the binding of LDL To study whether sortilin increased the amount of LDL bound through some intracellular effect, HeLa T-REx cells were transfected with a SORT1 plasmid lacking the sequence for the cytoplasmic domain (SORT1DCD). By deleting the cytoplasmic domain, sortilin is no longer transported from the trans-Golgi network to the endosomes/lysosomes and is predominantly located on the cell surface [14]. The amounts of cell-surface sortilin and LDLR, as well as the amount of LDL bound, were determined by flow cytometry of transfected cells. As compared to cells transfected with the WTSORT1 plasmid, cells transfected with the SORT1DCD plasmid exhibited 19-fold more sortilin on the cell surface and they bound a 3-fold higher amount of LDL (Table 3). Because it is unlikely that the increased amount of LDL bound to cells transfected with the SORT1DCD plasmid was due to the 21% increased amount of cellsurface LDLRs, it appears that sortilin itself binds LDL without acting through an intracellular mechanism. For comparison, HeLa T-REx cells transfected with the LDLR plasmid bound 7-fold more LDL than cells transfected with the WT-SORT1 plasmid (Table 3). Similar findings were observed when these experiments were performed in HepG2 cells (data not shown). 4. Discussion In this study we have screened 842 unrelated hypercholesterolemic subjects for mutations in the SORT1 gene. None of the subjects were heterozygous for a mutation in the LDLR gene or in the PCSK9 gene and none carried the R3500Q mutation in the apolipoprotein B-100 gene. 45 mutations were identified in the SORT1 gene of which 15 were missense mutations. Subjects carrying missense mutations had similar levels of total serum cholesterol as the non-carriers. Thus, there was no apparent effect of these missense mutations on the levels of total serum cholesterol. Eight of the missense mutations were selected for in vitro studies to study their effect on the binding of LDL. Overexpression of sortilin markedly increased the amount of sortilin on the cell surface and the amount of LDL bound. Even though the eight sortilin mutants affected the binding of LDL differently, this variation seemed to result from differences in the amount of sortilin on the cell surface which could be due to differences in the level of expression. The positive correlation between the amount of sortilin on the cell surface and the amount of LDL bound, suggests that sortilin binds LDL on the cell surface. Overexpression of sortilin in HeLa T-REx cells resulted in a 20% increased amount of LDL internalized. This finding is in agreement with the findings of Linsel-Nitschke et al. [10] who observed a 23% increased amount of LDL internalized in HEK293 cells. They reasoned that because sortilin is primarily located intracellularly in post-Golgi compartments, the most likely mechanism for the increased amount of LDL internalized was through some indirect effects. However, our data showing that high amounts of LDL were bound to a mutant sortilin lacking the cytoplasmic domain, which fails to be shuttled from the trans-Golgi network to endosomes/
374
K. Tveten et al. / Atherosclerosis 225 (2012) 370e375
Fig. 1. Effect of knocking down sortilin on internalization of LDL. Four different siRNAs against the SORT1 gene were used to knock down sortilin expression in HeLa T-REx cells. The amount of LDL internalized was then determined by flow cytometry. The amount of LDL internalized in cells transfected with control siRNA was assigned a value of 1. Mean (SD) values of three separate experiments are shown. The amount of sortilin in cell lysates was determined by Western blot analysis. One representative Western blot is shown.
3.50
LDL internalization
3.00
2.56
2.50 2.00 1.50
1.23
1.20 1.00
1.00 0.50 0.00 Untransfected
Empty
SORT1
LDLR
Fig. 2. Effect of overexpression of sortilin on internalization of LDL. HeLa T-REx cells were transiently transfected with a bicistronic plasmid containing cDNA of the SORT1 gene, the LDLR gene or with an empty plasmid, and the amount of LDL internalized was determined by flow cytometry. The amount of LDL internalized in cells transfected with empty plasmid was assigned a value of 1. Mean (SD) values of three separate experiments are shown.
lysosomes [15], makes it unlikely that sortilin acts by some intracellular mechanism to increase the number of LDLRs. The ability of sortilin to bind LDL has also been shown by studies of plasmon surface resonance [8]. Moreover, a recent study that was published during preparation of this manuscript has also shown that sortilin is able to bind and internalize LDL [16]. Thus, sortilin binds and internalizes LDL by receptor-mediated endocytosis. The question then regards the role of sortilinmediated endocytosis of LDL in the regulation of plasma LDL
Table 3 Effect of SORT1 lacking the cytoplasmic domain (SORT1DCD) on the amount of cellsurface sortilin and LDLR and on the amount of LDL bound (mean SD). Plasmid
Cell-surface sortilin
Cell-surface LDLR
Bound LDL
WT-SORT1 SORT1DCD LDLR Untransfected Empty plasmid
1.00 19.1 0.07 0.03 0.02
1.00 1.22 (0.30) 159.5 (6.65) 1.23 (0.22) 0.65 (0.03)
1.00 3.33 6.75 0.90 0.35
(1.58) (0.004) (0.002) (0.004)
(0.25) (1.27) (0.14) (0.09)
cholesterol levels. From our data where the SORT1 gene or the LDLR gene have been silenced or overexpressed, it appears that the LDLR plays a much more significant role in receptor-mediated endocytosis of LDL than sortilin. This notion is also supported by the very high levels of LDL cholesterol of approximately 20 mmol/l in familial hypercholesterolemia homozygotes who lack functioning LDLRs [11], but who are presumed to have normal amounts of sortilin. Heterozygosity for a loss-of-function mutation in the SORT1 gene would be expected to result in a 50% reduction in the amount of functional sortilin on the cell surface. Based upon the results from genome-wide association studies, where homozygosity for the major allele at the chromosome 1p13.3 locus was associated with a more than 90% reduced expression of sortilin in human liver as compared to homozygosity for the minor allele [7], and a mere 0.4 mmol/l increased level of LDL cholesterol [7], it is unlikely that a 50% reduction in the amount of sortilin by a loss-of-function mutation would have much impact on the levels of LDL cholesterol. Based upon this reasoning, our mutation screening of the SORT1 gene and subsequent in vitro testing of identified missense mutations, we consider it unlikely that mutations in the SORT1 gene cause autosomal dominant hypercholesterolemia. Rather, loss-of-function mutations in the SORT1 gene may have only a marginal effect on the levels of LDL cholesterol. Appendix A. Supplementary material Supplementary material associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j. atherosclerosis.2012.10.026. References [1] Petersen CM, Nielsen M, Nykjær A, et al. Molecular identification of a novel candidate sorting receptor purified from human brain by receptor-associated protein affinity chromatography. J Biol Chem 1997;272:3599e605. [2] Quistgaard E, Madsen P, Grøftehauge MK, Nissen P, Petersen C, Thirup SS. Ligands bind to sortilin in the tunnel of a ten-bladed b-propeller domain. Nat Struct Mol Biol 2009;16:96e8.
K. Tveten et al. / Atherosclerosis 225 (2012) 370e375 [3] Willnow TE, Petersen CM, Nykjær A. VPS10P-domain receptors e regulators of neuronal viability and function. Nat Rev Neurosci 2008;9:899e909. [4] Willnow TE, Kjølby M, Nykjær A. Sortilins: new players in lipoprotein metabolism. Curr Opin Lipidol 2011;22:79e85. [5] Samani NJ, Erdmann J, Hall AS, et al. Genomewide association analysis of coronary artery disease. N Engl J Med 2007;357:443e53. [6] Myocardial Infarction Genetics Consortium. Genome-wide association of early-onset myocardial infarction with single nucleotide polymorphisms and copy number variants. Nat Genet 2009;41:334e41. [7] Musunuru K, Strong A, Frank-Kamenetsky, et al. From noncoding variant to phenotype via SORT1 at the 1p13 cholesterol locus. Nature 2010;466:714e9. [8] Kjolby M, Andersen OM, Breiderhoff T, et al. Sort1, encoded by the cardiovascular risk locus 1p13.3, is a regulator of hepatic lipoprotein export. Cell Metab 2010;12:213e23. [9] Dubé JB, Johansen CT, Hegele RA. Sortilin: an unusual suspect in cholesterol metabolism. Bioessays 2011;33:430e7. [10] Linsel-Nitschke P, Heeren J, Aherrahrou Z, et al. Genetic variation at chromosome 1p13.3 affects sortilin mRNA expression, cellular LDL-uptake and
[11]
[12]
[13] [14] [15]
[16]
375
serum LDL levels which translates to the risk of coronary heart disease. Atherosclerosis 2010;208:183e9. Goldstein JL, Hobbs HH, Brown MS. Familial hypercholesterolemia. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The metabolic & molecular basis of inherited disease. New York: McGraw-Hill; 2001. p. 2863e914. Cameron J, Holla ØL, Ranheim T, Kulseth MA, Berge KE, Leren TP. Effect of mutations in the PCSK9 gene on the cell surface LDL receptors. Hum Mol Genet 2006;15:1551e8. Sherry ST, Ward MH, Kholodov M, et al. dbSNP: the NCBI database of genetic variation. Nucleic Acids Res 2001;29:308e11. http://browser.1000genomes.org/Homo_sapiens/Gene/Variation_Gene/Table? db¼core;g¼ENSG00000134243;r¼1:109852192-109940573 (Table 1). Nielsen M, Madsen P, Christensen EI, et al. The sortilin cytoplasmic tail conveys Golgi-endosome transport and binds the VHS domain of the GGA2 sorting protein. EMBO J 2001;20:2180e90. Strong A, Ding Q, Edmonson A, et al. Hepatic sortilin regulates both apolipoprotein B secretion and LDL catabolism. J Clin Invest 2012;122:2807e16.