PCSK9 increases plasma LDL levels by increasing LDLR degradation [14, 15]. Gain of function mutations in the PCSK9 gene may cause an increase in plasma LDL-C and CHD risk by decreasing LDLR levels in the liver [9], while loss of function mutations may cause an increase in LDLR levels and a decrease in plasma LDL-C levels and protect against CHD formation [29, 30]. In our study, gain of function mutations of PCSK9 gene E670G (rs505151), R496W (rs374603772), and the loss of function mutation Y142X (rs6768943) which were all thought to affect serum LDL-C levels and PCSK9 rs11206510 polymorphism [31, 32] and rs2182833 [33] which have been associated with LDL-C levels previously and latter have been shown to possess no effect on PCSK9 function were examined to show the effects of variations on serum lipid and PCSK9 levels and restenosis risk in CHD patients.
According to RT-PCR results of the screening for R496W and Y142X mutations, all of the subjects in our study groups (191 people in total) had CC genotype, and no mutant allele was observed. In the study of Eroglu et al. which was performed in Aegean region of Turkey, the genotype distributions of PCSK9 R496W mutation were reported as CT 6%, TT 0.5% in dyslipidemia group (n=200), CT 0.5%, TT 0% in the control group (n=201). They also reported that total-C (p= 0.021), TG (p= 0.0001), HDL-C (p= 0.028), and LDL-C (p= 0.028) levels were higher in individuals with mutations of PCSK9-rs374603772 (R496W) and -rs137852912 (D374Y) than those without these mutations [34]. Kaya et al. conducted a study with 80 patients with familial hypercholesterolemia, and R496W mutation was observed in seven people (8.7%) and one of them was found to be homozygous. In this study, no difference was found in terms of cardiovascular disease risk between the patient group with any of the gain-of-function mutations (F216L, R496W, S127R, and D374Y) in the PCSK9 gene and individuals without the mutation [35]. Cohen et al. reported low LDL-C plasma levels in 128 individuals with loss of function mutations (Y142X and C679X) [29].
In our study, although statistical significance could not be obtained in S- and OS-CAD groups, it was observed that LDL-C levels were higher in G allele carriers than those who did not. In the study of Postmus et al., the variation of rs2182833 was found to be associated with LDL-C levels in elderly individuals (70-82 years of age) with vascular disease or at risk of vascular disease (p< 0.01). Moreover, LDL-C levels were observed as high to low homozygous> heterozygous, respectively [33]. In a study which gene-environment interaction in cardiovascular diseases was assessed with the STANISLAS cohort consisting of ~861 people who were followed up for 18-23 years, no relationship was observed between PCSK9-rs2182833 and rs11206510 variations and LDL-C and PCSK9 levels [36]. Similarly, in our study, no relationship was found between the PCSK9-rs11206510 polymorphism and LDL-C and PCSK9 levels. Genotype and allele frequencies of this polymorphism were similar among our study groups (p< 0.05). Similar to our findings, the rs11206510 polymorphism was not associated with CHD in the Asian population in a meta-analysis study conducted on 813 (290 CHD, 193 non-CHD patients, and 330 healthy controls) subjects [37]. In another study conducted in 1880 individuals in the Italian population that investigated PCSK9 gene variations and LDL-C levels and the risk of myocardial infarction (MI), rs2182833 was not associated with LDL-C, total-C and MI risk, while rs11206510 minor allele (C) was associated with with low levels of LDL-C (OR=0.82, 95%CI= 0.73-0.93, p= 1.89×10-3) and total-C (OR= 0.80, 95%CI= 0.72-0.89, p= 8.12×10-5) but not with the risk of MI [32]. In the study of Qi et al., the rs11206510 polymorphism was found to be associated with CHD in diabetic patients [38].
In our study, the PCSK9-rs11206510 T> C polymorphism in the S-CAD group was observed to be associated with lower HbA1c levels compared to the rare C allele TT genotype (p= 0.021). Similarly, in the study of Chen et al., a relationship was found between the rs11206510 minor C allele and the risk of MI [39]. In the Mendelian randomization study of Schmidt et al. which included 550,000 individuals (50,000 of whom had diabetes), rs11206510 rare C allele has been associated with higher fasting blood glucose, body mass index (BMI) and diabetes risk, although it was not found to be associated with low LDL-C levels and HbA1c levels [40]. Dysglycemia may be the result of lowering LDL-C. However, it was not yet known whether lowering LDL-C with PCSK9 inhibitors causes diabetes or not.
No difference was found between the patient and control groups in terms of TG, LDL-C and HDL-C in the study of Wang et al. [41], whereas in our study, LDL-C was lower and HDL-C was higher in OS-CAD group comparing with S-CAD group (p=0.042, p=0.008, respectively). Qi et al. reported that the HbA1c levels and the percentages of hypertension and hypercholesterolemia in diabetic patients who developed CHD were higher than those who did not develop CHD [38]. In our study, the serum levels of HbA1c and the frequencies of hyperlipidemia and Type 2 DM were found to be higher in S-CAD group compared to the OS-CAD group (p=0.015, p=0.001, p=0.007, respectively).
Decreased EF percentages (%) are one of the signs of CAD, and changes in left ventricular EF% has been reported to be associated with the severity of CAD in previous studies [42, 43]. In our study, a significant decrease was found in S-CAD group than OS-CAD group in terms of EF% (p=0.007). Similar to our findings, Bal et al., reported that EF% was found to be lower in patients with restenosis compared to patients without restenosis after stenting (p= 0.023) [44].
In our study, the frequency of the E670G mutant G allele was found to be higher in S-CAD group than in OS-CAD (p= 0.015). Besides, hyperlipidemia (p< 0.001) and the PCSK9 E670G (rs505151) G allele (p= 0.02) were determined for the risk of restenosis between S-CAD and OS-CAD groups in the logistic regression analysis showing that the PCSK9 E670G (rs505151) G allele may be an effective risk factor in the development of restenosis. Slimani et al. reported that the frequency of the E670G-G allele in the Tunisian population was higher in CAD patients than controls (0.132 vs. 0.068, p= 0.030). They also stated that plasma total-C and LDL-C levels were higher in 670G carriers than those without mutations (Total-C: 6.78 vs. 4.92 mmol/l, p< 0.0001, LDL-C: 4.60 vs. 3.00 mmol/l p= 0.001) [45]. Norata et al. suggested that the E670G-G allele is associated with high LDL-C and total-C [26]. In the meta-analysis study of Qiu et al., the E670G-G allele was found to be associated with high TG, LDL-C and cardiovascular disease risk [46]. Chen et al. suggested the association between E670G mutant allele and LDL-C and risk of atherosclerosis [47]. Similarly, in our study, LDL-C levels were found to be higher in E670G-G allele carriers compared to individuals with AA genotype in S-CAD group (125.32±49.43 vs. 107.90±34.18, p= 0.036). Unlike our results and other studies suggesting the association between E670G mutation and CAD [45, 46]. Hsu et al. reported in their study with 202 patients with CAD and 614 controls in Taiwan-China population that E670G-G allele frequency was found to be similar in CAD and control groups (p> 0.05) and LDL-C levels in E670G-G allele carriers were significantly low (2.78±0.82 mmol/L vs. 3.02±0.85 mmol/L, p=0.029) [48]. In our study, borderline significance (p= 0.05) and lower TG levels were detected in E670G-G allele carriers of OS-CAD patients compared to individuals with AA genotype in the same group.
ApoE, another ligand of LDLR, is the apoprotein component of LDL and VLDL and is a polymorphic glycoprotein that acts as a ligand for chylomicron residue receptors [17]. Since ApoE2 contains cysteine instead of arginine in E3 and E4 at the 158th position, it was weakly bound to LDLR, causing an increase in VLDL levels in plasma. ApoE4 causes less gene expression (downregulation) of LDL receptors, leading to an increase in LDL levels in plasma. Most studies have found an association between CHD and the ApoE4 allele [22, 23]. Xu et al. reported that the risk of CHD in the Caucasian population is high in ApoE4 carriers and low in ApoE2 carriers [49].
In the study conducted by Wang et al. with CHD patients in the Chinese population, it was reported that total-C, TG and LDL-C levels were higher in patients with ApoE4 allele compared to those who had homozygous ApoE3, and ApoE4 allele was associated with hypertriglyceridemia when ApoE3/E3 genotype was selected as reference in the multivariate logistic regression analysis [50]. In the study of Liu et al., it was reported that ApoE4 allele carriers have a 1.64 times higher risk of type 2 DM and 1.80 times higher risk of cardiovascular disease in Chinese Hakka population [51]. In our study, the frequency of the ApoE epsilon polymorphism E3E4 genotype was found to be higher in the S-CAD group compared to the OS-CAD group (15.6% vs 9.8%), however, it was not statistically significant. In the study conducted by Atis et al. with CAD patients in the Turkish population, no significant difference was found between E2E2, E2E3, E2E4, E3E3, E3E4, E4E4 genotypes [52]. In our study, higher TG levels (p< 0.001) were found in patients with the ApoE4 allele in the S-CAD group compared to the OS-CAD group. There was no significant relationship between the ApoE4 allele and serum lipid parameters in the OS-CAD group. In the study conducted by Karahan et al. in the Turkish population, it was stated that CAD patients carrying ApoE4 allele had higher levels of LDL-C (p= 0.001) and total-C (p= 0.03) compared to those who did not [53]. In the study of Attila et al., it was shown that the ApoE4 allele was associated with the development of CAD in the Turkish population, but the effect of E2 and E4 alleles on lipid levels was not detected [54].
In our study, TG levels were found to be higher in individuals with ApoE2 allele in OS-CAD patients compared to those without E2 allele, but no significant difference was observed (p> 0.05). Serum VLDL-C levels in OS-CAD group were found to be significantly higher (p= 0.022) in patients with E2 allele compared to those without E2 allele. E2 allele was not found to be associated with serum lipid levels in S-CAD patients. Although serum TG levels were high in E2 allele carriers in S-CAD group, the difference was not statistically significant (223.07±169.38 vs. 167.97±11.46, p> 0.05). In the study of Elmadbouh et al., hypercholesterolemia (p= 0.034) and high LDL-C (p= 0.003) were reported in patients carrying ApoE4 compared to CAD patients carrying ApoE3, and high TG levels (p= 0.037) were reported in those carrying ApoE2 [55].
ApoER (LRP8) is an important molecule in lipoprotein metabolism. It functions in the removal of LDL-C and VLDL-C from plasma and has been associated with the risk of MI in western societies. The R952Q (rs5174) mutation converts arginine to glutamine. It has been associated with familial risk of CAD and premature MI in Caucasian populations [56]. In the study of Lieb et al. which was conducted with a German population, no association was observed between R952Q mutation and familial MI or CAD [57]. In our study, just fasting blood glucose level was found to be higher in those with the ApoER-rs5174 [G> A (C> T)] G allele (GG+GA) in OS-CAD patients compared to those with homozygous AA genotype (p< 0.001) in the same group.
In our study, a statistically significant difference was observed between the groups with and without the ApoE3 allele in the S-CAD group in terms of glucose and HbA1c (p< 0.001 and p= 0.044, respectively). In a study conducted with the Han Chinese population with individuals aged 70 and over, higher fasting blood glucose levels were found in individuals with the E3E3 genotype compared to individuals with the E2E2 and E4E4 genotypes (p= 0.047) and the ApoE3E3 genotype was found to be associated with diabetes [58]. In the study of El-Lebedy et al., diabetic patients with the E3E4 genotype showed a 2.4 fold increased risk of cardiovascular disease (95%CI: 1.14--5.19, p= 0.02) and E3E4 genotype was found as an independent risk factor for cardiovascular disease (OR= 2.3, p= 0.009) but not for type 2 DM (OR= 1.7, p= 0.28). However, the E4 allele was reported as an independent risk factor for both type 2 DM (OR= 2.2, p= 0.04) and cardiovascular disease (OR= 3.0, p= 0.018) and type 2 DM patients with E4 allele have 5.9 times increased risk of cardiovascular disease development [59].
In conclusion, while our findings revealed that PCSK9-rs505151 (E670G)- G allele and hyperlipidemia may be effective risk factors in the development of restenosis. Further studies on larger study group with expression of PCSK9 gene and determination of the role of E670G substitution and in different ethnic groups are needed to confirm the association of this polymorphism with risk of restenosis.