Introduction
Familial hypercholesterolemia (FH) is a common monogenic disorder characterized by elevated low-density lipoprotein cholesterol (LDL-C) levels and an increased risk of premature cardiovascular disease.1 To date, 3 different genes have been causally linked to autosomal dominant FH: the low-density lipoprotein receptor (LDLR) gene, the apolipoprotein B-100 (APOB) gene, and the proprotein convertase subtilisin / kexin type 9 (PCSK9) gene. To date, more than 1700 different causal mutations in the LDLR gene have been identified, so screening for these mutations to establish a molecular diagnosis represents a methodological challenge. Various mutations in the APOB gene, especially mutations in the LDLR-binding domain of APOB, are also related to FH. Mutations in other regions of the APOB gene are much rarer and usually difficult to screen for using DNA sequencing.2,3 There is evidence that FH patients with monogenic mutations, even if treated, have a significantly higher mortality rate than those with polygenic hypercholesterolemia.4 In light of these findings, early diagnosis and treatment of patients with FH are necessary to prevent cardiovascular complications. Therefore, comprehensive screening for known as well as rare or unknown mutations is required. Moreover, the detection rate of pathogenic mutations in the LDLR, APOB, and PCSK9 genes ranges from 40% to 60%, so there is a need for identifying new genes responsible for FH.5
The increasing availability and decreasing cost of next generation sequencing (NGS) make it a suitable tool for diagnosing monogenic dyslipidemias. NGS is also useful in identifying novel genes in LDL metabolism that possibly contribute to the FH phenotype, which may open new possibilities in the diagnosis and treatment of patients with FH.5-8
In this study, we employed NGS to describe genetic variants affecting cholesterol levels associated with FH and variants in other genes affecting LDL-C levels in a group of patients from the Małopolska province in Poland.
Patients and methods
The study comprised patients with a clinical diagnosis of FH based on the Simon Broome Register.9 According to these criteria, definite FH is defined as a total cholesterol (TC) level greater than 6.7 mmol/l or LDL-C level greater than 4 mmol/l in a child younger than 16 years, or a TC level greater than 7.5 mmol/l or LDL-C level greater than 4.9 mmol/l in an adult (levels either before treatment or highest on treatment), plus tendon xanthomas in the patient or a first-degree (parent, sibling, child) or second-degree relative (grandparent, uncle, aunt). Alternatively, it can be defined as the presence of DNA-based evidence of a LDLR mutation, familial defective APOB-100, or a PCSK9 mutation.
Possible FH is defined as a TC level greater than 6.7 mmol/l or LDL-C level greater than 4 mmol/l in a child younger than 16 years, or a TC level greater than 7.5 mmol/l or LDL-C level greater than 4.9 mmol/l in an adult (levels either before treatment or highest on treatment), and meeting at least 1 of the following criteria: family history of myocardial infarction (at the age <50 years in a second-degree relative or <60 years in a first-degree relative) or family history of hypercholesterolemia (>7.5 mmol/l in an adult first- or second-degree relative or >6.7 mmol/l in a child or sibling aged <16 years).
Patients with either a definite or a probable diagnosis of FH were included in the study. They were recruited from among the inhabitants of the Małopolska province in Poland who had been referred to an outpatient lipid clinic at the University Hospital of the Jagiellonian University Medical College due to severe hypercholesterolemia. We also included the data of 2 hypercholesterolemic children, members of family 418 (Supplementary material, Table S1), treated at the Department of Endocrinology, Institute of Pediatrics of the Jagiellonian University Medical College, who were diagnosed with FH according to the criteria proposed by Kwiterowich.10 Secondary causes of hyperlipidemia were excluded in all patients. The participants filled out a standardized questionnaire concerning family history, past medical history, smoking habits, and treatment. All patients gave their written informed consent prior to the inclusion, and the study was approved by the Jagiellonian University Bioethical Committee (KBET/34/B/2012).
Blood samples were collected after an overnight fast. In all patients, the fasting serum lipid levels were analyzed by enzymatic methods. The apolipoprotein (apo) A1 level was determined by immunoturbidimetry (APTEC Diagnostics nv, Sint-Niklaas, Belgium), and the level of apo B-100, by sandwich enzyme-linked immunosorbent assay kits (Human apo B-100: SEA603Hu Cloud-Clone Corp., Houston, Massachusetts, United States), according to the manufacturers’ instructions. Serum lipoprotein (Lp) (a) concentration was determined by immunoturbidimetry.
DNA sequencing
The DNA was extracted from whole blood samples and collected into ethylenediaminetetraacetic acid tubes with the Maxwell 16 Blood DNA Purification Kit on a Maxwell device (Promega, Madison, Wisconsin, United States). The isolated DNA quantity was measured on the Quantus fluorometer (Promega). To detect mutations, we used a SureSelect custom-designed sequencing panel (Agilent, Santa Clara, California, United States) consisting of probes targeting exons of genes causative for monogenic FH: LDLR, APOB, and PCSK9, as well as genes with a recently proved causality or having smaller, quantitative effects on FH: apolipoprotein E (APOE), ATP-binding cassette subfamily G member 5 (ABCG5), ATP-binding cassette subfamily G member 8 (ABCG8), lipoprotein lipase (LPL), NPC intracellular cholesterol transporter 1 (NPC1), low-density lipoprotein receptor adaptor protein 1 (LDLRAP1), lipase C, hepatic type (LIPC), cadherin EGF LAG seven-pass G-type receptor 2 (CELSR2), as well as the signal transducing adaptor protein 1 gene (STAP1), whose causality in FH has been recently negated.11 Library preparation was performed using SureSelect Target Enrichment (Agilent) reagents, according to the manufacturer’s protocol. Pooled DNA libraries were prepared using pair-end sequencing (75 bp) with v3 reagents on the MiSeq platform (Illumina, San Diego, California, United States), following the manufacturer’s instructions.
Sequencing data analysis
Raw reads were processed with the Illumina software via generating demultiplexed fast files with base calls and corresponding base-call quality scores. These files were then processed through a custom pipeline that had been described in detail in our previous publication.12 Briefly, the quality of the reads was assessed with FastQC v.11.5 (Babraham Bioinformatics, Cambridge, United Kingdom). Raw reads were aligned to the human reference genome GRCh37 (hg19) using the BWA-MEM algorithm v.0.7.5. Realignment across indels and base quality recalibration were performed with a Genome Analysis Toolkit v.3.7 (GATK). Duplicated reads were filtered out with a SAMtools v.0.1.19 (https://www.htslib.org/).
Variants were called using a GATK Haplotypecaller and filtered out using recommended hard filtering parameters based on GATK Best Practices. The filtered variants were annotated with SnpEff and VEP (https://pcingola.github.io/SnpEff), then prioritized according to the GEnome MINIng (GEMINI, version 0.18.3) impact severity classification. Subsequently, the variants were filtered against the frequencies in known databases (minor allele frequency [MAF] <1%): 1000 Genomes, ESP, and ExAC. Pathogenicity prediction scores, such as CADD, Fitcons, Sorting Intolerant From Tolerant (SIFT2), Polyphen 2, MutationTaster2, Protein Variation Effect Analyzer (PROVEAN), and FATHMM-XF,13-15 as well as clinical significance (ClinVar) were used to evaluate pathogenicity of the detected variants.16-18 The variants identified as pathogenic according to at least 3 of the 5 metrics described above were considered pathogenic. As a control of NGS reliability, DNA from FH patients with a known mutation in APOB (1 person) and LDLR (6 persons) were included.
Statistical analysis
Statistical analysis included calculations of means and SD for variables with a normal distribution, and medians and interquartile ranges for those with a nonnormal distribution. The t test and the Mann–Whitney test were used for comparisons between sexes. Qualitative characteristics were presented as numbers and percentages. The data of pediatric patients are not included in the Table presenting the characteristics of the study group (Table 1). The calculations were performed using the statistical package STATISTICA v. 13.3 (StatSoft, Kraków, Poland). P values lower than 0.05 were considered significant.
Parameter | Total (n = 90) | Men (n = 35) | Women (n = 55) | P value | |||
---|---|---|---|---|---|---|---|
n | Mean (SD) or median (IQR) | n | Mean (SD) or median (IQR) | n | Mean (SD) or median (IQR) | ||
Age, y | 87 | 43.8 (14.1) | 34 | 39.2 (12.03) | 53 | 46.7 (14.66) | 0.01 |
TC, mmol/l | 88 | 7.22 (2.07) | 33 | 6.88 (2.32) | 55 | 7.42 (1.9) | 0.23 |
HDL-C, mmol/l | 88 | 1.57 (0.46) | 33 | 1.38 (0.44) | 55 | 1.69 (0.43) | 0.002 |
LDL-C, mmol/l | 88 | 4.95 (2.01) | 33 | 4.94 (2.34) | 55 | 4.95 (1.81) | 0.99 |
TG, mmol/l | 87 | 1.24 (0.94–2.04) | 32 | 1.35 (1.01–2.1) | 55 | 1.19 (0.91–1.95) | 0.38 |
Apo A1, mg/dl | 60 | 143.64 (41.69) | 24 | 141.55 (39.22) | 36 | 145.03 (43.76) | 0.75 |
Apo B, mg/dl | 60 | 104.3 (19.93) | 24 | 103.65 (20.62) | 36 | 104.72 (19.74) | 0.84 |
Lp(a), g/l | 60 | 0.05 (0.03–0.16) | 24 | 0.03 (0.03–0.08) | 36 | 0.08 (0.03–0.22) | 0.01 |
Glucose, mmol/l | 78 | 5.01 (4.5–5.4) | 29 | 5.02 (4.7–5.52) | 49 | 5 (4.5–5.37) | 0.63 |
Creatinine, mg/l | 82 | 65 (58–75.5) | 33 | 74 (62–78) | 49 | 63 (58–68) | 0.02 |
BMI, kg/m2 | 89 | 26.18 (4.48) | 35 | 26.11 (3.69) | 54 | 26.23 (4.97) | 0.89 |
Waist circumference, cm | 88 | 86.98 (12.56) | 35 | 90.63 (11.76) | 53 | 84.57 (12.6) | 0.02 |
WHR | 86 | 0.87 (0.09) | 35 | 0.91 (0.07) | 51 | 0.84 (0.08) | <0.001 |
ALT, U/l | 80 | 26 (19–41) | 30 | 43.5 (26–53) | 50 | 23 (16–28) | <0.001 |
GGTP, U/l | 65 | 21 (14.5–39) | 24 | 38.55 (26–68) | 41 | 18 (12–22.6) | <0.001 |
SBP, mm Hg | 87 | 132.7 (17.38) | 34 | 136 (15.77) | 53 | 130.5 (18.17) | 0.15 |
DBP, mm Hg | 87 | 83.2 (12.19) | 34 | 83 (12.21) | 53 | 83.3 (12.3) | 0.9 |
Variables with a nonnormal distribution are presented as median with interquartile range (IQR) and were compared using the Mann–Whitney test. Variables with a normal distribution are presented as mean (SD) and were compared using the t test. a Data of pediatric patients analyzed in the study (n = 2) are not presented in the table. SI conversion factors: to convert apo A1 and apo B to g/l, multiply by 0.01; creatinine to μmol/l, by 88.4, ALT and GGTP to µkat/l, by 0.0167. Abbreviations: ALT, alanine aminotransferase; apo, apolipoprotein; BMI, body mass index; DPB, diastolic blood pressure; GGTP, γ-glutamyltranspeptidase; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; Lp(a), lipoprotein (a); SBP, systolic blood pressure; TC, total cholesterol; TG, triglycerides; WHR, waist-to-hip ratio |
Results
We examined a total of 90 unrelated patients, including 35 men (38.9%) and 55 woment (61.1%). Among them, 12 patients (13.5%) had coronary artery disease and 7 (7.9%) had a history of myocardial infarction. Table 1 presents the characteristics of the study population according to sex. The mean (SD) age of the patients was 43.8 (14.1) years, and the mean (SD) LDL-C level was 4.95 (2.1) mmol/l. Overall, 60.1% of the patients were treated with statins.
Table 2 presents variants in the LDLR and APOB genes found in the examined group. All the known control LDLR and APOB gene mutations were identified in the NGS analysis, thus confirming the reliability of this procedure in our center. We found 13 known pathogenic variants and 3 polymorphisms in the LDLR gene, as well as 3 variants that were probably disease-causing. The most common mutation was p.Cys34Gly in exon 2, found in 5 patients. Three pathogenic variants were observed in exon 4, and 3 in exon 9 of the LDLR gene. One variant of uncertain significance (VUS) / pathogenic was found in exon 10. Table 2 also presents variants that were not previously linked to FH in ClinVar, which were detected in the LDLR gene in 3 patients and in the APOB gene in 1 patient. Lipid characteristics of the single patient harboring the new variant of APOB and his family are presented in Supplementary material, Table S2. The mutation c:119dupT (pSer41fs) in exon 2 is predicted to result in a frame shift and change of serine to leucine in the 11th position of the LDLR protein.
Gene | Mutation | Amino acid change | Exon | Patients with allele, n | dbSNP | ClinVar |
---|---|---|---|---|---|---|
LDLR | NM_000 527.4:c.58G>A | p.Gly20Arg | 1 | 1 | rs147509697 | LB, VUS |
LDLR | NM_000 527.4:c.100T>G | p.Cys34Gly | 2 | 5 | rs879254405 | LP, P |
LDLR | NM_000 527.4:c.313+1G>T | – | 3 | 1 | rs112029328 | LP |
LDLR | NM_000 527.4:c.442T>C | p.Cys148Arg | 4 | 1 | rs879254525 | LP, P |
LDLR | NM_000 527.4:c.530C>T | p.Ser177Leu | 4 | 1 | rs121908026 | LP, P |
LDLR | NM_000 527.4:c.666C>A | p.Cys222* | 4 | 1 | rs756613387 | P |
LDLR | NM_000 527.4:c.782G>T | p.Cys261Phe | 5 | 1 | rs121908040 | LP, P |
LDLR | NM_000 527.4:c.798T>A | p.Asp266Glu | 5 | 1 | rs139043155 | LB, LP, P |
LDLR | NM_000 527.4:c.986G>T | p.Cys329Phe | 7 | 2 | rs761954844 | LP |
LDLR | NM_000 527.4:c.1 061–8T>C | – | 7 | 1 | rs72658861 | B, LB, LP, VUS |
LDLR | NM_000 527.4:c.1222G>A | p.Glu408Lys | 9 | 1 | rs137943601 | LP, P |
LDLR | NM_000 527.4:c.1223A>T | p.Glu408Val | 9 | 1 | rs879254838 | LP |
LDLR | NM_000 527.4:c.1328G>A | p.Trp443* | 9 | 1 | rs879254866 | P |
LDLR | NM_000 527.4:c.1449G>T | p.Trp483Cys | 10 | 1 | rs879254907 | P, VUS |
LDLR | NM_000 527.4:c.1 705+1G>A | – | 11 | 1 | rs875989926 | LP, P |
LDLR | NM_000 527.4:c.1775G>A | p.Gly592Glu | 12 | 2 | rs137929307 | LP, P |
LDLR | NM_000 527.4:c.1862C>G | p.Thr621Arg | 13 | 1 | rs879255058 | LP, P |
LDLR | NM_000 527.4:c.2 390–16G>A | – | 16 | 1 | rs183496025 | LB |
LDLR | NM_000 527.4:c.119dupT | p.Ser41fs | 2 | 1 | NA | DC (P >0.99)a |
LDLR | NM_000 527.4:c.787G>A | p.Asp263Asn | 5 | 1 | rs750900506 | DC (P = 0.99)a |
LDLR | NM_000 527.4:c.1486G>A | p.Gly496Ser | 10 | 1 | NA | DC (P = 0.73)a |
PCSK9 | NM_174 936.3:c.60_65dupGCTGCT | p.Leu21_Leu22dup | 1 | 1 | rs35574083 | B |
APOB | NM_000 384.2:c.12382G>A | p.Val4128Met | 29 | 1 | rs1801703 | LB, VUS |
APOB | NM_000 384.2:c.11833A>G | p.Thr3945Ala | 27 | 1 | rs1801698 | VUS |
APOB | NM_000 384.2:c.10708C>T | p.His3570Tyr | 26 | 1 | rs201736972 | P |
APOB | NM_000 384.2:c.10580G>A | p.Arg3527Gln | 26 | 9 | rs5742904 | LP, P |
APOB | NM_000 384.2:c.10579C>T | p.Arg3527Trp | 26 | 1 | rs144467873 | LP, P, VUS |
APOB | NM_000 384.2:c.10131G>A | p.Leu3377Leu | 26 | 1 | rs1799812 | VUS |
APOB | NM_000 384.2:c.8462C>T | p.Pro2821Leu | 26 | 2 | rs72653095 | LB, VUS |
APOB | NM_000 384.2:c.8353A>C | p.Asn2785His | 26 | 1 | rs2163204 | B, LB, VUS |
APOB | NM_000 384.2:c.7696G>A | p.Glu2566Lys | 26 | 10 | rs1801696 | LB, VUS |
APOB | NM_000 384.2:c.7615G>A | p.Val2539Ile | 26 | 1 | rs148170480 | VUS |
APOB | NM_000 384.2:c.6639_6641delTGA | p.Asp2213del | 26 | 1 | rs541497967 | LB, VUS |
APOB | NM_000 384.2:c.3 122–6G>A | – | 20 | 1 | rs72653071 | VUS |
APOB | NM_000 384.2:c.2 068–4T>A | – | 14 | 1 | rs41291161 | LB |
APOB | NM_000 384.2:c.1594C>T | p.Arg532Trp | 12 | 1 | rs13306194 | B, LB, VUS |
APOB | NM_000 384.2:c.538–9C>T | – | 5 | 1 | rs1800478 | B, LB |
APOB | NM_000 384.2:c.13181T>C | p.Val4394Ala | 29 | 1 | rs12720843 | VUS |
APOB | NM_000 384.2:c.11087T>C | p.Ile3696Thr | 26 | 1 | rs370096275 | VUS |
APOB | NM_000 384.2:c.10032A>C | p.Lys3344Asn | 26 | 1 | rs757857092 | VUS |
APOB | NM_000 384.2:c.7640T>C | p.Val2547Ala | 26 | 1 | NA | Polymorphism (P = 0.99)a |
APOB | NM_000 384.2:c.3851G>A | p.Arg1284Gln | 25 | 1 | rs372154910 | VUS |
APOB | NM_000 384.2:c.64_66delCTG | p.Leu22del | 1 | 1 | rs773844839 | B |
a Variants that were not previously linked to familiar hypercholesterolemia in ClinVar, with pathogenicity probability calculated by MutationTaster2 Abbreviations: APOB, apolipoprotein B-100 gene; B, benign; dbSNP, database of single-nucleotide polymorphism; DC, disease-causing; LB, likely benign; LDLR, low-density lipoprotein receptor gene; LP, likely pathogenic; NA, not available; P, pathogenic; PCSK9, proprotein convertase subtilisin / kexin type 9 gene; VUS, variant of unknown significance |
Interestingly, we found the same mutation (c.100T>G [p.Cys34Gly]) in exon 2 of the LDLR gene in 2 patients, and homozygosity for this mutation in their children (a son and a daughter). The son’s lipid values on irregular drug treatment were as follows: TC, 11.59 mmol/l; non–high-density lipoprotein cholesterol (HDL-C), 10.83 mmol/l; HDL-C, 0.76 mmol/l; LDL-C, 10.41 mmol/l; and triglycerides, 0.92 mmol/l. High cholesterol levels were detected at the age of 4 years due to knee xanthomas, which disappeared after hypolipidemic treatment. Physical examination revealed corneal arcus (Supplementary material, Figure S1) and Achilles tendon xanthoma, whereas echocardiography showed thickening of the right leaflet of the aortic valve. He and his father had also increased Lp(a) levels (1.04 g/l). Lipid values of the father during treatment with atorvastatin 40 mg and ezetimibe 10 mg were as follows: TC, 6.6 mmol/l; LDL-C, 5.71 mmol/l; HDL-C, 0.77 mmol/l; triglycerides, 0.77 mmol/l; Lp(a), 0.97g/l; apo B, 1.64 g/l; and apo A1, 0.89 g/l. The genealogical tree of family 418 and lipid levels of individual family members are presented in Supplementary material, Figure S2 and Table S1, respectively.
In the APOB gene, we observed 3 pathogenic variants, 14 VUSs, and 3 polymorphisms. We also found 1 variant in exon 26 not previously annotated in ClinVar that was probably disease-causing. The most common pathogenic mutation in the APOB gene, p.Arg3527Gln (rs5742904) in exon 26, was found in 9 patients. Four variants in the APOB gene were found, including 3 in exon 26 (p.Arg3527Trp, p.Leu3377Leu, and p.Asp2213del) in the LDLR-binding region, and 1 in exon 14 (c.2068-4T>A). We also observed other APOB gene variants, outside of exon 14 and the LDLR-binding region, in exons 1, 5, 12, 20, and 25. In exon 12 of the APOB gene, a VUS / polymorphism (p.Arg532Trp) was found. A c64_66delCTG (p.Leu22del) variant in exon 1, probably not pathogenic, was detected in 1 patient. A p.Glu2566Lys polymorphism in exon 26, probably not pathogenic, was found in 10 patients.
Table 3 presents selected variants in other genes associated with LDL-C levels. Interestingly, we observed a high frequency of pathogenic variants of the APOE gene, potentially influencing LDL-C levels. The APOE variant rs7412 (p.Arg176Cys//c.526C>T; Cgc/Tgc), probably pathogenic according to in silico prediction, was present in 4 patients. The most common APOE gene variant, p.Cys130Arg/c.388T>C (rs429358; Tgc/Cgc), probably benign, was observed in 16 patients and was also located in exon 4. This single-nucleotide polymorphism (SNP) affects the amino acid at position 130 of the resulting protein (Table 3).
Gene | Mutation | Amino acid change | Exon | Patients with allele, n | MutationTaster2 | Polyphen-2 | SIFT | dbSNP |
---|---|---|---|---|---|---|---|---|
LDLRAP1 | NM_015 627.2:c.672C>T | p.Ser224Ser | 7 | 1 | DC | – | T | rs41291054 |
STAP1 | NM_012 108.2:c.120+6T>C | – | 1 | 2 | DC | – | – | rs187909999 |
STAP1 | NM_012 108.2:c.619G>A | p.Asp207Asn | 6 | 2 | DC | B | T | rs146545610 |
ABCG5 | NM_022 436.2:c.1336C>T | p.Arg446* | 10 | 1 | DC | – | – | rs199689137 |
ABCG5 | NM_022 436.2:c.593G>A | p.Arg198Gln | 5 | 2 | DC | ProD | D | rs141828689 |
ABCG5 | NM_022 436.2:c.293C>G | p.Ala98Gly | 3 | 2 | DC | PosD | D | rs145164937 |
ABCG8 | NM_022 437.2:c.1083G>A | p.Trp361* | 7 | 1 | DC | – | – | rs137852987 |
ABCG8 | NM_022 437.2:c.1715T>C | p.Leu572Pro | 11 | 1 | DC | ProD | D | rs769576789 |
APOE | NM_000 041.2:c.388T>C | p.Cys130Arg | 4 | 16 | – | ProD | D | rs429358 |
APOE | NM_000 041.2:c.526C>T | p.Arg176Cys | 4 | 4 | – | B | T | rs7412 |
Abbreviations: ABCG5, ATP-binding cassette subfamily G member 5 gene; ABCG8, ATP-binding cassette subfamily G member 8 gene; APOE, apolipoprotein E gene; D, damaging; LDLRAP1, low-density lipoprotein receptor adaptor protein 1 gene; PosD, possibly damaging; ProD, probably damaging; SIFT, Sorting Intolerant from Tolerant; STAP1, signal transducing adaptor protein 1 gene; T, tolerated; others, see Table 2 |
Three of our patients with a FH phenotype carried no variants in the FH genes, but instead were carriers of rare variants of the ABCG5 gene. We also found 2 variants of the ABCG8 gene (Table 3). The 3 ABCG5 variants, namely, pArg446* in exon 10, p.Arg198Gln in exon 5, and p.Ala98Gly in exon 3, are likely to be pathogenic. Variants in gene ABCG8, p.Trp361* in exon 7 and p.Leu572Pro in exon 11, are also potentially pathogenic.
Interestingly, in 2 patients, we found a likely pathogenic missense variant (pathogenic score of 0.933 according to MutationTaster2) of the STAP1 gene: c.120+6T>C (Table 3). This variant was located in exon 1, and affected a highly-conserved amino acid. The LDL-C concentration in this STAP1 variant carrier was 4.5 mmol/l. We also found a missense variant in exon 6 of the STAP1 gene, p.Asp207Asn (c.619G>A), probably not pathogenic. This variant was predicted to possibly affect the protein function.
Of note, in the majority of patients with mutations in the LDLR or APOB gene, variants of other genes affecting the LDL-C level, such as ABCG5, ABCG8, LPL, LRP1, and APOE were also detected, which might influence the response to treatment.
Discussion
NGS is replacing older forms of genetic diagnosis in clinical practice. In the present study, targeting the genes responsible for FH was carried out by designing probe panels targeting the sequence of genes of interest, such as LDLR, APOB, and PCSK9, as well as monogenic mutations in other genes related to high LDL-C levels, such as APOE, sterol regulatory element binding transcription factor 2 (SREBP2), and lipase A, lysosomal acid type (LIPA).19
In accordance with our previous report,20 we found a great heterogeneity of mutations in the LDLR gene. In the present study population, most mutations in the LDLR gene were located in exon 4, similarly to what had been observed in the Czech population.21 In our previous cohort from the Małopolska province, the majority of mutations were located in exon 2.20 The most common mutation in the LDLR gene in our patients was p.Gly592Glu, which was also most frequently found in another Polish study22 and in the Slovak population.23
In our study, we confirmed that the most common mutation responsible for FH was the mutation p.Arg3527Gln in the APOB gene. It was observed in 10% of the examined group, which is in line with our previous report.20 In other publications analyzing the Polish population this mutation was observed in about 4% of the examined patients,21,24 and the frequency was similar in other Slavic populations25 as well as other European countries.26-29 We did not find the Thr3492Ile variant of the APOB gene in any of our patients.24 However, we identified APOB gene variants that had not been previously linked to FH in ClinVar. Routine analysis tests for the presence of mutations in the LDLR-binding region of the APOB gene (amino acids 3441–3615), while NGS enabled the analysis of the whole APOB gene, thus allowing us to describe new APOB variants in FH patients.
We observed APOB variants in exons 29 (2 persons), 25, 27, 20, 14, 12, 5, and 1. The APOB gene mutations located outside the LDLR-binding region had been described previously.30-32 Alves et al30 used a flow cytometry assay and fluorescently-labelled LDL from individuals with APOB gene mutations in exons 22 and 29 and showed that carriers of p.Arg1164Thr (exon 22) and p.Gln4494del (exon 29) were marked by a 40% decrease in internalization of LDL in lymphocytes and HepG2 cells, similarly to persons with an APOB-3527 mutation.
No FH-causing gain-of-function mutations in the PCSK9 gene were found in the group of patients analyzed in the present study.
Mutations in the STAP1 gene were described by Fouchier et al33 as being associated with FH. However, the causative role of STAP1 gene mutations in the context of FH is still under debate and, according to some authors, these mutations should not be considered disease-causing until more data appear.34-36 In a study by Loaiza et al,35 STAP1 gene variants did not alter plasma LDL-C levels. We found a variant in the STAP1 gene that was pathogenic according to in silico prediction, and it was associated with high LDL-C levels in the affected individuals. However, a family segregation study and functional tests are required to confirm its pathogenicity.
In our cohort of patients, we observed a relatively high frequency of APOE gene variants. APOE variants rs429358 and rs7412 are included by the Global Lipid Genetics Consortium among 12 SNP alleles showing a LDL-C–rising effect.37 A benign APOE gene variant rs429358 was found in 16 patients, which corresponded to almost 20% of our group. We did not find the APOE Leu167del mutation, which had been identified as disease-causing in the Italian and French populations.38-41
Interestingly, NGS data showed that a few patients with FH-like phenotypes carried mutations in the ABCG5, APOE, and LIPA genes, which are the causative genes for classical sitosterolemia, dysbetalipoproteinemia, and lysosomal acid lipase deficiency, respectively.41-44 Tada et al43 observed variants of the ABCG5 gene in FH patients without mutations in the genes associated with the FH phenotype; however, in our study, all the patients with an ABCG5 mutation carried mutations also in other LDL-C–rising genes, such as APOE or LDLRAP1.43 The available literature data suggest that deleterious mutations in the ABCG5/ABCG8 genes could contribute to worsening of FH or mimicking of the disease.43,44
The main limitation of the study is that we did not perform functional tests for the newly detected potentially pathogenic variants. Of note, all the existing algorithms predicting pathogenicity (including those used in the present study) had been reported to have limited sensitivity and / or specificity. Therefore, a lot of effort is still required from researchers to prove that the variants described as pathogenic indeed play a functional role in cholesterol metabolism.
Małgorzata Waluś-Miarka, MD, Department of Metabolic Diseases, Jagiellonian University Medical College, ul. Jakubowskiego 2, 31-507 Kraków, Poland, phone: +48 12 400 29 50, email: m.walus-miarka@uj.edu.pl.
November 25, 2022.
January 9, 2023.
January 17, 2023.
None.
The study was funded by a grant IP2011058771 from the Ministry of Science and Higher Education (to MW-M).
MW-M conceived the concept of the study, performed clinical examinations, analyzed the data, and wrote the manuscript. JT-Ż and MW-M contributed to the design of the research. JT-Ż and MK analyzed the data. EK performed statistical analysis. All authors were involved in data collection and contributed to manuscript revision.
None declared.
Totoń-Żurańska J, Wołkow P, Kapusta M, et al. Targeted sequencing of a gene panel in patients with familial hypercholesterolemia from Southern Poland. Pol Arch Intern Med. 2023; 133: 16417. doi:10.20452/pamw.16417
- 1.
- Dobrowolski P, Kabat M, Kępka C, et al. Atherosclerotic cardiovascular disease burden in patients with familial hypercholesterolemia: interpretation of data on involvement of different vascular beds. Pol Arch Intern Med. 2022; 132: 16248.Crossref
- 2.
- Benito-Vicente A, Uribe KB, Jebari S, et al. Familial hypercholesterolemia: the most frequent cholesterol metabolism disorder caused disease. Int J Mol Sci. 2018; 19: 3426.Crossref
- 3.
- Santos RD. Screening and management of familial hypercholesterolemia. Curr Opin Cardiol. 2019; 34: 526-530.Crossref
- 4.
- Sharifi M, Futema M, Nair D, Humphries SE. Polygenic hypercholesterolemia and cardiovascular disease risk. Curr Cardiol Rep. 2019; 21: 43.Crossref
- 5.
- Reiman A, Pandey S, Lloyd KL, et al. Molecular testing for familial hypercholesterolaemia-associated mutations in a UK-based cohort: development of an NGS-based method and comparison with multiplex polymerase chain reaction and oligonucleotide arrays. Ann Clin Biochem. 2016; 53: 654-662.Crossref
- 6.
- Hinchcliffe M, Le H, Fimmel A, et al. Diagnostic validation of a familial hypercholesterolaemia cohort provides a model for using targeted next generation DNA sequencing in the clinical setting. Pathology. 2014; 46: 60-68.Crossref
- 7.
- Johansen CT, Dubé JB, Loyzer MN, et al. LipidSeq: a next-generation clinical resequencing panel for monogenic dyslipidemias. J Lipid Res. 2014; 55: 765-772.Crossref
- 8.
- Hegele RA, Ban MR, Cao H, et al. Targeted next-generation sequencing in monogenic dyslipidemias. Curr Opin Lipidol. 2015; 26: 103-113.Crossref
- 9.
- DeMott K, Nherera L, Shaw EJ, et al. Clinical Guidelines and Evidence Review for Familial Hypercholesterolaemia: The Identification and Management of Adults and Children with Familial Hypercholesterolaemia. London: National Collaborating Centre for Primary Care and Royal College of General Practitioners; 2008.
- 10.
- Kwiterovich PO Jr. Recognition and management of dyslipidemia in children and adolescents. J Clin Endocrinol Metab. 2008; 93: 4200-4209.Crossref
- 11.
- Hegele RA, Joshua W, Knowles WJ, Horton JD. Delisting STAP1. The rise and fall of a putative hypercholesterolemia gene. Arterioscler Thromb Vasc Biol. 2020; 40: 847-849.Crossref
- 12.
- Totoń-Żurańska J, Kapusta P, Rybak-Krzyszkowska M, et al. Contribution of a novel B3GLCT variant to Peters plus syndrome discovered by a combination of next-generation sequencing and automated text mining. Int J Mol Sci. 2019; 28: 6006.Crossref
- 13.
- Ng PC, Henikoff S. SIFT: predicting amino acid changes that affect protein function. Nucleic Acids Res. 2003; 31: 3812-3814.Crossref
- 14.
- Choi Y, Chan AP. PROVEAN web server: a tool to predict the functional effect of amino acid substitutions and indels. Bioinformatics. 2015; 31: 2745-2747.Crossref
- 15.
- Rogers MF, Shihab HA, Mort M, et al. FATHMM-XF: accurate prediction of pathogenic point mutations via extended features. Bioinformatics. 2018; 34: 511-513.Crossref
- 16.
- Schwarz JM, Rödelsperger C, Schuelke M, Seelow D. MutationTaster evaluates disease-causing potential of sequence alterations. Nat Methods. 2010; 7: 575-576.Crossref
- 17.
- Landrum MJ, Lee JM, Benson M, et al. ClinVar: public archive of interpretations of clinically relevant variants. Nucleic Acids Res. 2016; 44: D862-D868.Crossref
- 18.
- Adzhubei IA, Schmidt S, Peshkin L, et al. A method and server for predicting damaging missense mutations. Nat Methods. 2010; 7: 248-249.Crossref
- 19.
- Henderson R, O’Kane M, McGilligan V, et al. The genetics and screening of familial hypercholesterolemia. J Biomed Sci. 2016; 23: 39.Crossref
- 20.
- Sharifi M, Walus-Miarka M, Idzior-Waluś B, et al. The genetic spectrum of familial hypercholesterolemia in south-eastern Poland. Metabolism. 2016; 65: 48-53.Crossref
- 21.
- Tichý L, Fajkusová L, Zapletalová P, et al. Molecular genetic background of an autosomal dominant hypercholesterolemia in the Czech Republic. Physiol Res. 2017; 66 (Suppl 1): S47-S54.Crossref
- 22.
- Chmara M, Wasag B, Zuk M, et al. Molecular characterization of Polish patients with familial hypercholesterolemia: novel and recurrent LDLR mutations. J Appl Genet. 2010; 51: 95-106.Crossref
- 23.
- Gabčová D, Vohnout B, Staníková D, et al. The molecular genetic background of familial hypercholesterolemia: data from the Slovak nation-wide survey. Physiol Res. 2017; 31: 75-84.Crossref
- 24.
- Bednarska-Makaruk M, Bisko M, Pulawska MF, et al. Familial defective apolipoprotein B-100 in a group of hypercholesterolaemic patients in Poland. Identification of a new mutation Thr3492Ile in the apolipoprotein B gene. Eur J Hum Genet. 2001; 9: 836-842.Crossref
- 25.
- Tichý L, Freiberger T, Zapletalová P, et al. The molecular basis of familial hypercholesterolemia in the Czech Republic: spectrum of LDLR mutations and genotype-phenotype correlations. Atherosclerosis. 2012; 223: 401-408.Crossref
- 26.
- Tichý L, Freiberger T, Zapletalová P, et al. The molecular basis of familial hypercholesterolemia in the Czech Republic: spectrum of LDLR mutations and genotype-phenotype correlations. Atherosclerosis. 2012; 223: 401-408.Crossref
- 27.
- Iacocca MA, Chora JR, Carrié A, et al; ClinGen FH Variant Curation Expert Panel. ClinVar database of global familial hypercholesterolemia-associated DNA variants. Hum Mutat. 2018; 39: 1631-1640.Crossref
- 28.
- Widhalm K, Dirisamer A, Lindemayr A, Kostner G. Diagnosis of families with familial hypercholesterolaemia and / or Apo B-100 defect by means of DNA analysis of LDL-receptor gene mutations. J Inherit Metab Dis. 2007; 30: 239-247.Crossref
- 29.
- Grenkowitz T, Kassner U, Wühle-Demuth M, et al. Clinical characterization and mutation spectrum of German patients with familial hypercholesterolemia. Atherosclerosis. 2016; 253: 88-93.Crossref
- 30.
- Alves AC, Etxebarria A, Soutar AK, et al. Novel functional APOB mutations outside LDL-binding region causing familial hypercholesterolaemia. Hum Mol Genet. 2014; 23: 1817-1828.Crossref
- 31.
- Alves AC, Benito-Vicente A, Medeiros AM, et al. Further evidence of novel APOB mutations as a cause of familial hypercholesterolaemia. Atherosclerosis. 2018; 277: 448-456.Crossref
- 32.
- Motazacker MM, Pirruccello J, Huijgen R, et al. Advances in genetics show the need for extending screening strategies for autosomal dominant hypercholesterolaemia. Eur Heart J. 2012; 33: 1360-1366.Crossref
- 33.
- Fouchier SW, Dallinga-Thie GM, Meijers JC, et al. Mutations in STAP1 are associated with autosomal dominant hypercholesterolemia. Circ Res. 2014; 115: 552-555.Crossref
- 34.
- Hegele RA, Borén J, Ginsberg HN, et al. Rare dyslipidaemias, from phenotype to genotype to management: a European Atherosclerosis Society Task Force consensus statement. Lancet Diabetes Endocrinol. 2020; 8: 50-67.Crossref
- 35.
- Loaiza N, Hartgers ML, Reeskamp LF, et al. Taking one step back in familial hypercholesterolemia: STAP1 does not alter plasma LDL (low-density lipoprotein) cholesterol in mice and humans. Arterioscler Thromb Vasc Biol. 2020; 40: 973-985.Crossref
- 36.
- Danyel M, Ott CE, Grenkowitz T, et al. Evaluation of the role of STAP1 in familial hypercholesterolemia. Sci Rep. 2019; 9: 11995.Crossref
- 37.
- Teslovich TM, Musunuru K, Smith AV, et al. Biological, clinical and population relevance of 95 loci for blood lipids. Nature. 2010; 466: 707-713.
- 38.
- Marduel M, Ouguerram K, Serre V, at al. Description of a large family with autosomal dominant hypercholesterolemia associated with the APOE p.Leu167del mutation. Hum Mutat. 2013; 34: 83-87.Crossref
- 39.
- Wintjens R, Bozon D, Belabbas K, et al. Global molecular analysis and APOE mutations in a cohort of autosomal dominant hypercholesterolemia patients in France. J Lipid Res. 2016; 57: 482-491.Crossref
- 40.
- Almigbal TH, Batais MA, Hasanato RM, et al. Role of apolipoprotein E gene polymorphism in the risk of familial hypercholesterolemia: a case-control study. Acta Biochim Pol. 2018; 65: 415-420.Crossref
- 41.
- Awan Z, Choi HY, Stitziel N, et al. APOE p.Leu167del mutation in familial hypercholesterolemia. Atherosclerosis. 2013; 231: 218-222.Crossref
- 42.
- Rios J, Stein E, Shendure J, et al. Identification by whole-genome resequencing of gene defect responsible for severe hypercholesterolemia. Hum Mol Genet. 2010; 19: 4313-4318.Crossref
- 43.
- Tada H, Okada H, Nomura A, et al. Rare and deleterious mutations in ABCG5/ABCG8 genes contribute to mimicking and worsening of familial hypercholesterolemia phenotype. Circ J. 2019; 83: 1917-1924.Crossref
- 44.
- Reeskamp LF, Volta A, Zuurbier L, et al. ABCG5 and ABCG8 genetic variants in familial hypercholesterolemia. J Clin Lipidol. 2020; 14: 207-217.Crossref