👤 Mathilde Di Filippo

Primary hypobetalipoproteinemia (HBL) is mostly due to a polygenic origin or to monogenic disorders including loss of function (LOF) variants in APOB, much less frequently Angiopoietin-like 3 gene (ANGPTL3). A new heterozygous variant of uncertain significance (VUS), p.H343R missense variant in ANGPTL3 cosegregated with HBL in a family. The aim of the present study was to assess in vitro the functionality of this variant and to establish its causality in this family. Targeted next-generation sequencing was performed in the proband to assess monogenic and polygenic origins using an LDL-C-dedicated polygenic risk score (PRS All 8 HBL subjects had PRS This study shows that the novel ANGPTL3-p.H343R variant decreases ANGPTL3 secretion in vitro and can now be considered as a LOF variant. The lipid phenotype in this family results from a synergistic combination of the p.H343R ANGPTL3 variant and a polygenic HBL predisposition. Show less

Familial hypobetalipoproteinemia 1 (FHBL-SD2) is the most common monogenic form of primary hypocholesterolaemia, related to truncating variants in the APOB gene encoding apolipoprotein B. Due to its high level of complexity, variants of uncertain significance (VUS) require further investigations. This study aims to demonstrate the value of setting minigene assays in the FHBL-SD2's genetic diagnosis. Four APOB VUS occurring in patients with a FHBL-SD2 phenotype were considered. In silico analysis were performed with six software programs supposed to predict the potential splicing effect. Then, functional consequences were studied in vitro using a minigene splicing reporter assay. An effect on splicing was predicted in silico for the 4 variants, with the activation of a cryptic acceptor site for c.694-13A>G and c.1471-6A>G variants, and the use of a cryptic donor site for c.1123A>G and c.1470G>A variants. Minigene study showed a complete effect on splicing for 3 mutations, confirming the in silico predictions. All of these transcripts result in premature truncated variants. Therefore, these variants were reclassified as likely pathogenic and causative of FHBL-SD2. However, no effect was shown either in HeLA and HuH7 cells for the c.1470G>A variant. Minigene study appears to be a promising and valuable tool to enhance the diagnostic accuracy of FHBL-SD2. It emphasizes the challenge in interpreting VUS and underscores the importance of establishing a clear strategy to assess their significance. Therefore, promoting minigene studies would be beneficial to understand precisely the impact of splicing variants. Show less

Mobile elements (ME) can transpose by copy-and-paste mechanisms. A heterozygous insertion in APOB exon 3 coding sequence was suspected in a patient with hypobetalipoproteinemia (HBL), by gel electrophoresis of the PCR products. An insertion of a 85 bp fragment flanked by a polyA stretch and a target replication site duplication was identified as a ME insertion (MEI) from the AluYa5 subfamily, NM₀₀₀₃₈₄.3(APOB):c.135₁₃₆ins(160). Then, the DNA was reanalyzed using our NGS custom panel. Routine analysis did not reveal any causative variant, but manual inspection of the alignments and MELT enabled us to detect this MEI from NGS data. A functional study revealed that this MEI introduces a stop codon p.(Phe46Alafs*2) and additionally leads to p.(Lys41Serfs*2) due to an exon skipping. This is the first report of a MEI into APOB, as a cause of HBL. Furthermore, our study highlights the value of including MEI-callers in routine pipelines to improve primary dyslipidemia diagnosis. Show less

Dysbetalipoproteinemia (DBL) is a combined dyslipidemia associated with an increased risk of atherosclerotic cardiovascular diseases mostly occurring in ε2ε2 subjects and infrequently in subjects with rare Patients were divided into 3 groups according to their Total cholesterol (TC)/ApoB and NHDLC/ApoB are the best ratios to suspect DBL. In ε2ε2, according to their likelihood ratios (LR), the most clinically efficient algorithms were the HCL, Sniderman and De Graaf's. In APOEmut, Sniderman's algorithm exhibited the lowest negative LR (0.07) whereas the HCL's exhibited the highest positive LR (29). In both cohorts, the HCL algorithm had the best LR. We proposed a powerful algorithm based on ApoB concentration and the routine lipid profile, which performs remarkably well in detecting ε2ε2 or Show less

LIPC encodes hepatic lipase (HL), a liver-bound protein with both phospholipase and triglyceride lipase activity, and involved in the catabolism of circulating lipoproteins. We recently identified the gain-of-function variant HL-E97G, with selectively increased phospholipase activity, as a new genetic cause of familial combined hypocholesterolaemia in humans. The role of HL in the development of atherosclerosis remains controversial. In this context, the action of HL-E97G on the development of atherosclerosis remains unknown. To evaluate the lipid-lowering and anti-atherogenic properties of HL-E97G vs. wildtype HL (HL-WT) in hypercholesterolaemic APOE*3-Leiden.CETP mice, a well-established model for human-like lipoprotein metabolism, and to assess dependence of these effects on the LDL receptor (LDLR) pathway in LDLR-deficient (Ldlr-/-) mice. APOE*3.Leiden.CETP mice or Ldlr-/- mice received an intravenous injection of AAV8 expressing either eGFP (control), HL-WT or HL-E97G (3 × 1011 GC/mouse) while being fed pro-atherogenic diets. Plasma cholesterol levels were measured monthly, and aortic atherosclerotic lesion sizes were assessed at termination. HL-E97G largely decreased plasma total cholesterol exposure in APOE*3-Leiden.CETP mice (-63% vs. control; -58% vs. HL-WT), resulting at least in part from increased uptake of (V)LDL by the liver, accompanied by a marked decrease in atherosclerotic lesion size (-98% vs. control; -97% vs. HL-WT) in the aortic root. Importantly, HL-E97G also strongly reduced plasma cholesterol exposure in Ldlr-/- mice (-80% vs. control; -77% vs. HL-WT), and decreased atherosclerotic lesion size in the aortic root (-54% vs. control; -41% vs. HL-WT) and the aortic arch (-73% vs. control; -70% vs. HL-WT). HL-E97G strongly reduces plasma cholesterol levels, by increasing the uptake of (V)LDL, to decrease atherosclerosis development in mice independently of the LDLR pathway. These data suggest that modulating HL function is a promising tool in patients with familial hypercholesterolaemia. Show less

Atherosclerotic cardiovascular disease is the main cause of mortality worldwide and is strongly influenced by circulating low-density lipoprotein (LDL) cholesterol levels. Only a few genes causally related to plasma LDL cholesterol levels have been identified so far, and only 1 gene, Using next-generation sequencing, we identified a novel dominant rare variant in the Family members carrying the We identified and characterized a novel rare variant in the Show less

The LMF1 (lipase maturation factor 1) gene encodes a protein involved in lipoprotein lipase and hepatic lipase maturation. Homozygous mutations in LMF1 leading to severe hypertriglyceridemia (SHTG) are rare in the literature. A few additional rare LMF1 variants have been described with poor functional studies. The aim of this study was to assess the frequency of LMF1 variants in a cohort of 385 patients with SHTG, without homozygous or compound heterozygous deleterious mutations identified in lipoprotein lipase (LPL), apolipoprotein A5 (APOA5), apolipoprotein C2 (APOC2), glycosylphosphatidylinositol-anchored high-density lipoprotein binding protein 1 (GPIHBP1) genes, and to determine their functionality. LMF1 coding variants were screened using denaturing high-performance liquid chromatography followed by direct sequencing. In silico studies of LMF1 variants were performed, followed by in vitro functional studies using human embryonic kidney 293T (HEK-293T) cells cotransfected with vectors encoding human LPL and LMF1 cDNA. LPL activity was measured in cell culture medium after heparin addition using human VLDL-TG as substrate. Nineteen nonsynonymous coding LMF1 variants were identified in 65 patients; 10 variants were newly described in SHTG. In vitro, p.Gly172Arg, p.Arg354Trp, p.Arg364Gln, and p.Arg537Trp LMF1 variants decreased LPL activity, and the p.Trp464Ter variant completely abolished LPL activity. We identified a young girl heterozygote for the p.Trp464Ter variant and a homozygote carrier of the p.Gly172Arg variant with a near 50% decreased LPL activity in vitro and in vivo. The study confirms the rarity of LMF1 variants in a large cohort of patients with SHTG. LMF1 variants are likely to be involved in multifactorial hyperchylomicronemia. Partial LMF1 defects could be associated with intermittent phenotype as described for p.Gly172Arg homozygous and p.Trp464Ter heterozygous carriers. Show less

Optimal molecular diagnosis of primary dyslipidemia is challenging to confirm the diagnosis, test and identify at risk relatives. The aim of this study was to test the application of a single targeted next-generation sequencing (NGS) panel for hypercholesterolemia, hypocholesterolemia, and hypertriglyceridemia molecular diagnosis. NGS workflow based on a custom AmpliSeq panel was designed for sequencing the most prevalent dyslipidemia-causing genes (ANGPTL3, APOA5, APOC2, APOB, GPIHBP1, LDLR, LMF1, LPL, PCSK9) on the Ion PGM Sequencer. One hundred and forty patients without molecular diagnosis were studied. In silico analyses were performed using the NextGENe software and homemade tools for detection of copy number variations (CNV). All mutations were confirmed using appropriate tools. Eighty seven variations and 4 CNV were identified, allowing a molecular diagnosis for 40/116 hypercholesterolemic patients, 5/13 hypocholesterolemic patients, and 2/11, hypertriglyceridemic patients respectively. This workflow allowed the detection of CNV contrary to our previous strategy. Some variations were found in previously unexplored regions providing an added value for genotype-phenotype correlation and familial screening. In conclusion, this new NGS process is an effective mutation detection method and allows better understanding of phenotype. Consequently this assay meets the medical need for individualized diagnosis of dyslipidemia. Show less

The heterogeneity and mechanisms of multifactorial chylomicronemia (MCM) remain poorly understood. To gain new insights, post heparin lipolysis measured at 60 min (PHLA60), in addition to the more commonly used 10 min (PHLA10), was assessed in patients with history of MCM. 62 consecutive MCM patients were studied. The evaluation included LPL, APOC2, APOA5, GPIHBP1, LMF1 and APOE gene sequencing, as well as pre- and post-heparin injection biochemical analysis, including lipid profiles, determination of apolipoprotein B, B-48, CII, CIII, lipoprotein lipase (LPL) concentrations (LPLC0, LPLC10 and LPLC60) and post-heparin LPL activity (PHLA10 and PHLA60). In controls, PHLA60 did not differ from PHLA10, while in MCM patients, PHLA60 was significantly lower than PHLA10 (p<0.001). PHLA60 showed a bimodal distribution in MCM patients (p=0.03). One subgroup exhibited PHLA60 similar to controls, with persistent lipoprotein remodeling and, paradoxically, the highest basal plasma TG concentration. APOE ε4 was over-represented compared to the European population (p<0.05) and Apo CIII/Apo B ratio was increased (p<0.01). The other subgroup exhibited low PHLA60 (p<0.001) compared to both controls and the other MCM subgroup with a lipoprotein profile consistent with fast and transient remodeling. LMF1 p. Arg364Gln was over-represented compared to the European population (p<0.05). The study showed that PHLA60 identifies a subgroup of MCM with a defect in lipolysability and/or hepatic clearance of triglycerides-rich lipoproteins, and a larger one with a defect in LPL availability. These findings provide new insights into the heterogeneity of MCM and might contribute to adjust treatment targeting. Show less

APOC3 is a major regulator of triglycerides metabolism. Several APOC3 variants are associated with hypertriglyceridemia (HTG). Our aim was to establish the potential regulation of APOC3 3'UTR variants associated with HTG by liver or intestinal miRNAs. We sequenced APOC3 3'UTR in 100 type 2 diabetic (TD2) patients with severe HTG (TG > 15 mmol/L) (HTG group) compared to 100 normotriglyceridemic patients (NTG group). We performed in silico studies to identify potential loss of miRNA binding induced by APOC3 3'UTR variants. We also performed in vitro studies to test the functionality of miRNA/APOC3 variants interactions: APOC3 3'UTR plasmids coupled with a firefly luciferase reporter were transfected in HepG2, HuH-7 and Caco-2 cells. We identified only two variants: SstI (rs5128) and BbvI (rs5225) in APOC3 3'UTR in the 2 groups of patients. Only the SstI-S2 rare allele was significantly associated with HTG (allele frequency 19,5% in HTG group vs. 9,5% in NTG group, p = 0.0045). In silico studies predicted a potential loss in the binding of 5 miRNAs induced by the S2 variant. These 5 miRNAs are all endogenously expressed in human liver and intestine, as well as in the cell models studied. However, in vitro, the S2 variant did not modulate APOC3 3'UTR reporter gene expression in HepG2, HuH-7 and Caco-2 cells. Our results do not confirm the hypothesis of a direct regulation of the APOC3 SstI variant by hepatic or intestinal miRNAs. Show less

Determination of lipoprotein lipase (LPL) activity is important for hyperchylomicronemia diagnosis, but remains both unreliable and cumbersome with current methods. Consequently by using human VLDL as substrate we developed a new LPL assay which does not require sonication, radioactive or fluorescent particles. Post-heparin plasma was added to the VLDL substrate prepared by ultracentrifugation of heat inactivated normolipidemic human serums, diluted in buffer, pH 8.15. Following incubation at 37°c, the NEFA (non esterified fatty acids) produced were assayed hourly for 4 hours. LPL activity was expressed as µmol/l/min after subtraction of hepatic lipase (HL) activity, obtained following LPL inhibition with NaCl 1.5 mmol/l. Molecular analysis of LPL, GPIHBP1, APOA5, APOC2, APOE genes was available for 62 patients. Our method was reproducible (coefficient of variation (CV): intra-assay 5.6%, inter-assay 7.1%), and tightly correlated with the conventional radiolabelled triolein emulsion method (n = 26, r = 0.88). Normal values were established at 34.8 ± 12.8 µmol/l/min (mean ± SD) from 20 control subjects. LPL activities obtained from 71 patients with documented history of major hypertriglyceridemia showed a trimodal distribution. Among the 11 patients with a very low LPL activity (< 10 µmol/l/min), 5 were homozygous or compound heterozygous for LPL or GPIHBP1 deleterious mutations, 3 were compound heterozygous for APOA5 deleterious mutations and the p.S19W APOA5 susceptibility variant, and 2 were free of any mutations in the usual candidate genes. No homozygous gene alteration in LPL, GPIHBP1 and APOC2 genes was found in any of the patients with LPL activity > 10 µmol/l/min. This new reproducible method is a valuable tool for routine diagnosis and reliably identifies LPL activity defects. Show less

Determination of lipoprotein lipase (LPL) activity is important for hyperchylomicronemia diagnosis, but remains both unreliable and cumbersome with current methods. Consequently by using human VLDL as substrate we developed a new LPL assay which does not require sonication, radioactive or fluorescent particles. Post-heparin plasma was added to the VLDL substrate prepared by ultracentrifugation of heat inactivated normolipidemic human serums, diluted in buffer, pH 8.15. Following incubation at 37°c, the NEFA (non esterified fatty acids) produced were assayed hourly for 4 hours. LPL activity was expressed as µmol/l/min after subtraction of hepatic lipase (HL) activity, obtained following LPL inhibition with NaCl 1.5 mmol/l. Molecular analysis of LPL, GPIHBP1, APOA5, APOC2, APOE genes was available for 62 patients. Our method was reproducible (coefficient of variation (CV): intra-assay 5.6%, inter-assay 7.1%), and tightly correlated with the conventional radiolabelled triolein emulsion method (n = 26, r = 0.88). Normal values were established at 34.8 ± 12.8 µmol/l/min (mean ± SD) from 20 control subjects. LPL activities obtained from 71 patients with documented history of major hypertriglyceridemia showed a trimodal distribution. Among the 11 patients with a very low LPL activity (<10 µmol/l/min), 5 were homozygous or compound heterozygous for LPL or GPIHBP1 deleterious mutations, 3 were compound heterozygous for APOA5 deleterious mutations and the p.S19W APOA5 susceptibility variant, and 2 were free of any mutations in the usual candidate genes. No homozygous gene alteration in LPL, GPIHBP1 and APOC2 genes was found in any of the patients with LPL activity >10 µmol/l/min. This new reproducible method is a valuable tool for routine diagnosis and reliably identifies LPL activity defects. Show less

APOA5 c.*158C>T (rs2266788), located in the 3' UTR, belongs to APOA5 haplotype 2 (APOA5*2), which is strongly associated with plasma triglyceride levels and modulates the occurrence of both moderate and severe hypertriglyceridemia. Individuals with APOA5*2 display reduced APOA5 expression at the posttranscriptional level. However, the functionality of this haplotype remains unclear. We hypothesized that the hypertriglyceridemic effects of APOA5*2 could involve miRNA regulation in the APOA5 3' UTR. Bioinformatic studies have identified the creation of a potential miRNA binding site for liver-expressed miR-485-5p (MIRN485-5p) in the mutant APOA5 3' UTR with the c.*158C allele. In human embryonic kidney 293T (HEK293T) cells cotransfected with an APOA5 3' UTR luciferase reporter vector and a miR485-5p precursor, c.*158C allele expression was significantly decreased. Moreover, in HuH-7 cells endogenously expressing miR-485-5p, we observed that luciferase activity was significantly lower in the presence of the c.*158C allele than in the presence of the c.*158T allele, which was completely reversed by a miR-485-5p inhibitor. We demonstrated that the rare c.*158C APOA5 allele creates a functional target site for liver-expressed miR-485-5p. Therefore, we propose that the well-documented hypertriglyceridemic effect of APOA5*2 involves an APOA5 posttranscriptional downregulation mediated by miR-485-5p. Show less

GPIHBP1 is a new endothelial binding site for lipoprotein lipase (LPL), the key enzyme for intravascular lipolysis of triglyceride-rich lipoproteins (TGRL). We have identified two new missense mutations of the GPIHBP1 gene, C89F and G175R, by systematic sequencing in a cohort of 376 hyperchylomicronemic patients without mutations on the LPL, APOC2, or APOA5 gene. Phenotypic expression and functional consequences of these two mutations were studied. We performed clinical and genotypic studies of probands and their families. GPIHBP1 functional alterations were studied in CHO pgsA-745 transfected cells. Probands are an adult with a homozygous G175R mutation and a child with a hemizygous C89F neomutation and a deletion of the second allele. C89F mutation was associated with a C14F signal peptide polymorphism on the same haplotype. Both patients had resistant hyperchylomicronemia, low LPL activity, and history of acute pancreatitis. In CHO pgsA-745 cells, both G175R and C14F variants reduce the expression of GPIHBP1 at the cell surface. C89F mutation is responsible for a drastic LPL-binding defect to GPIHBP1. C14F may further potentiate C89F effect. The emergence of hyperchylomicronemia in the generation after a neomutation further establishes a critical role for GPIHBP1 in TGRL physiopathology in humans. Our results highlight the crucial role of C65-C89 disulfide bond in LPL binding by GPIHBP1 Ly6 domain. Furthermore, we first report a mutation of the hydrophobic C-terminal domain that impairs GPIHBP1 membrane targeting. Show less