Association between the PPP1R3B polymorphisms and serum lipid traits, the risk of coronary artery disease and ischemic stroke in a southern Chinese Han population.

BACKGROUND: Little is known about the association of the protein phosphatase 1 regulatory subunit 3B gene (PPP1R3B) single nucleotide polymorphisms (SNPs) and serum lipid levels, the risk of coronary artery disease (CAD) and ischemic stroke (IS) in the Chinese populations. This study detected such association in a Southern Chinese Han population.
METHODS: Genotypes of 4 novel PPP1R3B SNPs (rs12785, rs330910, rs330915 and rs9949) in 1704 Han Chinese (CAD, 556; IS, 531 and control, 617) were determined by the Snapshot technology.
RESULTS: The rs12785A and rs9949A allele frequency was higher in both CAD/IS patients than in controls. The rs330910T and rs330915T allele frequency was also higher in CAD patients than in controls. The rs330910T allele carriers in controls had lower serum low-density lipoprotein cholesterol (LDL-C) levels than the rs330910T allele non-carriers (P < 0.0014). The rs12785A, rs9949A and rs330910T allele carriers were associated with an increased risk of CAD (P = 0.008-0.004). There was strong linkage disequilibrium among the 4 SNPs in the controls and CAD/IS patients. The T-A-A-G haplotype was associated with a decreased risk of CAD and IS, whereas the A-A-T-A haplotype was associated with an increased risk for IS. Haplotype-environment interactions on the risk of CAD and IS were also observed.
CONCLUSIONS: Several PPP1R3B polymorphisms were associated with serum LDL-C levels, the risk of CAD and IS in the Southern Chinese Han population. But these findings still need to be confirmed in the other populations with larger sample sizes.

Background

The mortality and morbidity of coronary artery disease (CAD) and ischemic stroke (IS) remain very high in different districts [1-3]. Although both CAD and IS are complex multifactorial disorder, it is widely accepted that the pathological basis of the two diseases is atherosclerosis, and dyslipidemia plays an important role in process of coronary atherosclerosis [4]. Recently, Sabatine et al. [5] reported an inspiration result that a lipid-lowering drug, evolocumab (a monoclonal antibody) lowers low-density lipoprotein cholesterol (LDL-C) levels by approximately 60%. CAD and IS also share some common genetic and environmental determinants such as sex, age, obesity, cigarette smoking, hyperlipidemia, and hypertension [6-11]. Previous genome-wide association studies (GWASes) identified many genetic variants influenced the risk of CAD or IS [11] and it was reported that heritability estimates of coronary artery calcification are approximately 50% [12, 13].

The protein phosphatase 1 regulatory subunit 3B gene (PPP1R3B) is located on chromosome 8p. The gene which encoding a regulatory subunit of protein phosphatase 1 is involved in modulation of glycogen synthesis in liver and skeletal muscle [14-16]. Several previous studies have also found the association of variants in PPP1R3B locus with other cardiometabolic risk factors such as fasting glucose, fasting insulin, ferritin, liver enzymes, and nonalcoholic fatty liver disease [17-20]. Manning et al. [17] reported that variants in the PPP1R3B were associated with fasting glucose and fasting insulin at genome-wide levels of significance (P < 5 × 10− 8). Mehta et al. [18] found that PPP1R3B promoted hepatic glycogen synthesis and thereby regulated fasting energy homeostasis in mice those with liver-specific overexpression of PPP1R3B. Kahali et al. [19] showed that higher expression of variant falling in noncoding regions of PPP1R3B promoted nonalcoholic fatty liver disease and suggesting that this could be a functional expression quantitative trait locus (eQTL) variant [20]. Although the regulatory mechanism of the PPP1R3B expression is not well-known, it was acknowledged that the 3′-untraslated region plays an important role in governing spatial and temporal translation of a mRNA [21]. Two previous GWASes showed that single nucleotide polymorphisms (SNPs) in the PPP1R3B were associated with plasma lipid levels including high-density lipoprotein cholesterol (HDL-C), LDL-C, and total cholesterol (TC) and the risk of CAD [22, 23]. A study showed that higher expression of PPP1R3B by adenovirus administration would lower HDL-C levels in mouse liver [22]. In addition, a previous study in the Chinese Han population has showed that three PPP1R3B SNPs were significantly associated with plasma LDL-C (rs2126259, rs9987289 and rs19334) and TC (rs2126259 and rs9987289), and two SNPs with C-reactive protein (CRP; rs189798 and rs330919) but not with plasma lipid levels [24]. These findings suggest that the other SNPs in the 3’UTR of PPP1R3B may also influence serum lipid levels and the risk of CAD and IS. Therefore, the purpose of the present study was to detect the association of four novel PPP1R3B SNPs (rs12785, rs330910, rs330915 and rs9949) and their haplotypes with serum lipid levels and the risk of CAD and IS in a Southern Chinese Han population.

Methods

Study patients

The study samples contained 1087 unrelated patients (CAD, n = 556 and IS, n = 531). All of them were the hospitalized patients in the First Affiliated Hospital, Guangxi Medical University. The diagnosis of CAD based on typical clinical, discomfort, electrocardiographic changes, as well as cardiac makers (creatinine kinase-MB and troponin T). CAD was defined by coronary angiography which two independent angiographers were both blinded to the outcome of the genotypes. Coronary stenosis ≥50% in at least one of the three main coronary vessels was considered significant [25]. All of the IS patients received a strict neurological examination, and brain magnetic resonance imaging (MRI) was performed. The diagnosis of IS was according to TOAST (Trial of Org 10,712 in Acute Stroke Treatment) criteria, and patients included met one or two criteria: large-artery thrombosis and small-vessel occlusion [26]. All of the participants with a history of autoimmune, hematologic, neoplastic, liver, renal, thyroid, and type 1 diabetes were rejected. CAD subjects with a history of IS or IS patients with a history of CAD were excluded.

Control subjects

A total of 617 control subjects matched by age, gender, ethnic group (Han Chinese in Guangxi, China) were also included. All of the individuals were randomly selected from the Physical Examination Center of the First Affiliated Hospital during the same period. Questionnaires, history-taking, strict clinical examination and image examinations (computed tomography or MRI) were used to insure all the participants free of CAD and IS. All participants have provided their written informed consents and the study protocol was approved by the Ethics Committee of the First Affiliated Hospital, Guangxi Medical University (No: Lunshen-2011-KY-Guoji-001; Mar. 7, 2011). The reported investigations were in accordance with the principles of the Declaration of Helsinki.

Biochemical measurements

Before venous blood samples were obtained, all participants fasted at least 12 h. The levels of TC, triglyceride (TG), HDL-C, and LDL-C were determined by enzymatic methods with commercially available kits (RANDOX Laboratories). Serum apolipoprotein (Apo) A1 and ApoB levels were detected by the immunoturbidimetric immunoassay. The normal values in our Clinical Science Experiment Center were 3.10–5.17 mmol/L for TC, 0.56–1.70 mmol/L for TG, 0.91–1.81 mmol/L for HDL-C, 2.70–3.20 mmol/L for LDL-C, 1.00–1.78 g/L for ApoA1, 0.63–1.14 g/L for ApoB levels, and 1.00–2.50 for the ApoA1/ApoB ratio. The participants with TC > 5.17 mmol/L, and/or TG > 1.70 mmol/L were defined as hyperlipidemic [27, 28]. Hypertension was defined as a systolic blood pressure (SBP) of 140 mmHg or greater, and/or a diastolic blood pressure (DBP) of 90 mmHg or higher [29]. Drinking based on alcohol consumption (yes or no). Individuals’ age was divided into > 60- or ≤ 60-year subgroups. Body mass index (BMI) was calculated according to the values of weight divided by height squared (kg/m2). A BMI of ≤24, 24–28, and > 28 kg/m2 was defined as normal weight, overweight and obesity, respectively. Smoking was defined as current smoking (yes or no).

SNPs selection

The selection of SNPs was according to the following assumption: (1) Variants in the PPP1R3B were proved to be associated with plasma lipid levels and cardiovascular disease according to previous GWASes [22, 30–32]. (2) Selected SNPs were established by Haploview (Broad Institute or MIT and Harvard, Cambridge, MA, USA, version 4.2). (3) Variant Effect Predictor from online resource (1000 Genome Project Database) predicted that selected SNPs participate in protein-coding and lead to serum lipid changes (Additional file 1 Figure. S1). (4) SNPs information was obtained from NCBI dbSNP Build 132 (http://www.Ncbi.nlm. nih.gov/SNP/). (5) SNPs were restricted to minor allele frequency (MAF) > 1%. (6) The PPP1R3B rs12785, rs330910, rs330915 and rs9949 SNPs were selected by the block-based approach. This strategy was enabled by the correlations between tagging SNPs manifested as linkage disequilibrium (LD). Although classic tagging is not the goal of SNP selection, with innovative tagging SNP selection bias is inevitable [33-37].

Genotyping

Genomic DNA was extracted from leucocytes of venous blood using the phenol-chloroform method. Genotyping of the SNPs was accomplished by the Snapshot technology platform in the Center for Human Genetics Research, Shanghai Genesky Bio-Tech Co. Ltd. [27, 28, 38–42]. The restriction enzymes for the SNPs were SAP (Promega) and Exonuclease I (Epicentre), respectively. The forward and reverse primers were 5’-GGGGCAACCTGGGAAAGATTC-3′ and 5’-GCCTACACACTTCAGAGGGTGACA-3′ for rs12785; 5’-TTGCTTCCATTTGAGTTCGATTTATG-3′ and 5’-CCTCTCCCAGGTGGGTAACACTCT-3′ for rs330910; 5’-CGGCCCCAGAGGTCTCTTTTAC-3′ and 5’-TCAGAAAATGGTTTTATCGTGACTGTG-3′ for rs330915; and 5’-GGAAGATCCAGAAAATGGGCAGT-3′ and 5’-ACAATGTCAGAGTCAATGGGAGAATTT-3′ for rs9949 SNPs, respectively.

Statistical analyses

The statistical software of SPSS22.0 (SPSS Inc., Chicago, IL, USA) was used to carry out the statistical analyses. Quantitative variables were expressed as mean ± standard deviation (serum TG levels were expressed as medians and interquartile ranges). Qualitative variables were expressed as percentage. Allele frequency was determined via direct counting, and the standard goodness-of-fit test was used for the testing of Handy-Weinberg equilibrium (HWE). Sex ratio and the genotype distribution were evaluated by a chi-square analysis. The student’s unpaired t-test was used to test the general characteristics between patients and controls. Analysis of covariance (ANCOVA) was used to test the association of genotypes and cardiometabolic traits such as serum lipid parameters, BMI, SBP, DBP and pulse pressure (PP). Bonferroni correction was employed for variants associated with the cardiometabolic traits, and a P < 0.0014 (0.05/5 × 7) for serum lipid parameters, and P < 0.0025 (0.05/5 × 4) for BMI, SBP, DBP and PP were considered statistical significant. The associations of genotypes and the risk of CAD and IS, also the gene-environment interactions on the risk of CAD and IS were tested by the unconditional logistic regression after gender, age, BMI, smoking, alcohol consumption, hypertension and hyperlipidemia were adjusted [6, 7, 27, 28]. The correlation risk was estimated by odds ratio (OR) and 95% confidence interval (95%CI). The pattern of pair-wise LD between the four SNPs was measured by D’ and r2 using the SHEsis software [43]. A two-tailed P value less than 0.05 was considered statistically significant for the remaining variables.

Results

General characteristics of the participants

The general characteristics of the participants are presented in Table 1. The mean levels of BMI, SBP, PP, TC, TG and ApoB were higher but the values of HDL-C and LDL-C were lower in CAD patients than in controls (P < 0.05 for all). The values of BMI, SBP, PP, TC and TG were higher whereas those of DBP, HDL-C, LDL-C and ApoB were lower in IS patients than in controls (P < 0.05 for all).

Table 1

General characteristic of the participants

Characteristic	Control	CAD	IS	P CAD	P IS
Number	617	556	531
Male/female	445/172	410/146	384/147	0.534	0.942
Age, year	62.57 ± 11.79	62.22 ± 10.56	62.78 ± 12.38	0.174	0.083
Body mass index, kg/m2	22.60 ± 2.82	23.86 ± 3.24	23.44 ± 3.51	0.002	0.000
Systolic blood pressure, mmHg	128.23 ± 19.06	133.18 ± 23.42	147.78 ± 21.92	0.000	0.007
Diastolic blood pressure, mmHg	80.59 ± 11.47	79.14 ± 13.40	83.82 ± 12.92	0.058	0.006
Pulse pressure, mmHg	47.63 ± 13.70	54.05 ± 16.95	63.96 ± 17.81	0.000	0.000
Cigarette smoking, n (%)	242(39.2)	240(43.2)	221(41.6)	0.170	0.409
Alcohol consumption, n (%)	193(31.3)	154(27.7)	150(28.2)	0.179	0.263
Total cholesterol, mmol/L	4.88 ± 1.06	4.51 ± 1.16	4.52 ± 1.14	0.007	0.035
Triglyceride, mmol/L	1.22 (0.80)	1.64 (1.08)	1.62 (1.06)	0.000	0.000
HDL-C, mmol/L	1.89 ± 0.49	1.14 ± 0.34	1.22 ± 0.40	0.000	0.000
LDL-C, mmol/L	2.72 ± 0.77	2.70 ± 0.99	2.68 ± 0.89	0.000	0.000
Apolipoprotein (Apo) A1, g/L	1.41 ± 0.25	1.02 ± 0.31	1.03 ± 0.22	0.983	0.051
ApoB, g/L	0.90 ± 0.21	0.91 ± 0.26	0.89 ± 0.25	0.000	0.000
ApoA1/ApoB	1.64 ± 0.52	1.38 ± 2.48	1.26 ± 0.60	0.043	0.791

CAD coronary artery disease, IS ischemic stroke, HDL-C high-density lipoprotein cholesterol, LDL-C low-denstity lipoprotein cholesterol. The value of triglyceride was presented as median (interquartile tange), the difference between CAD/IS patients and controls was determined by the Wilcoxon-Mann-Whitney test

Genotypic and allelic frequencies in controls and patients

The genotypic and allelic frequencies of the four PPP1R3B SNPs are summarized in Additional file 1 Table S1. The genotype distribution of the four SNPs was consistent with the HWE in patients and controls (PHWE > 0.05 for all). The frequency of the rs12785A and rs9949A alleles was higher in CAD/IS patients than in controls (P < 0.05 for all). The frequency of the rs330910T and rs330915T alleles was also higher in CAD patients than in controls, but no difference was found between IS patients and controls. There was no difference in the genotype frequencies of the four SNPs between CAD/IS patients and controls.

Genotypes and cardiometabolic traits

The associations of the four PPP1R3B SNPs and cardiometabolic traits such as serum lipid parameters, BMI, SBP, DBP and PP in controls are presented in Table 2 and Additional file 1 Table S2. The rs330910T allele carriers had lower serum LDL-C levels than the rs330910T allele non-carriers (rs330910AA homozygotes; P < 0.0014). There were no differences in serum TC, TG, HDL-C, ApoA1, ApoB levels, and the ApoA1/ApoB ratio among the genotypes of the remaining 3 SNPs. There were also no significant associations between the four SNPs and BMI, SBP, DBP and PP (P > 0.0025 for all; Additional file 1 Table S2).

Table 2

Association of the PPP1R3B genotypes and serum lipid levels in controls

Genotype	n	TC(mmol/L)	TG(mmol/L)	HDL-C(mmol/L)	LDL-C(mmol/L)	ApoA1(g/L)	ApoB(g/L)	ApoA1/ApoB
rs12785
TT	356	4.88 ± 1.02	1.25 (0.89)	1.89 ± 0.48	2.79 ± 0.80	1.41 ± 0.26	0.90 ± 0.22	1.66 ± 0.56
TA	222	4.89 ± 0.87	1.21 (0.68)	1.92 ± 0.49	2.69 ± 0.72	1.4 1 ± 0.23	0.91 ± 0.23	1.64 ± 0.47
AA	39	4.47 ± 1.05	1.05 (0.82)	1.87 ± 0.52	2.41 ± 0.71	1.34 ± 0.30	0.88 ± 0.24	1.59 ± 0.41
P		0.036	0.320	0.623	0.010	0.231	0.798	0.714
TT	356	4.88 ± 1.02	1.25 (0.89)	1.89 ± 0.48	2.79 ± 0.80	1.41 ± 0.26	0.90 ± 0.22	1.66 ± 0.56
TA + AA	261	4.83 ± 0.91	1.19 (0.70)	1.92 ± 0.50	2.65 ± 0.81	1.40 ± 0.24	0.90 ± 0.23	1.63 ± 0.46
P		0.036	0.119	0.610	0.011	0.398	0.560	0.089
rs330910
AA	396	4.86 ± 1.04	1.26 (0.87)	1.87 ± 0.48	2.77 ± 0.83	1.41 ± 0.26	0.91 ± 0.23	1.65 ± 0.55
AT	194	4.88 ± 0.86	1.18 (0.62)	1.95 ± 0.51	2.67 ± 0.66	1.42 ± 0.24	0.90 ± 0.22	1.66 ± 0.46
TT	27	4.64 ± 0.85	1.14 (0.95)	1.92 ± 0.43	2.50 ± 0.59	1.34 ± 0.19	0.93 ± 0.21	1.52 ± 0.40
P		0.522	0.492	0.171	0.092	0.381	0.730	0.416
AA	396	4.86 ± 1.04	1.26 (0.87)	1.87 ± 0.48	2.77 ± 0.83	1.41 ± 0.26	0.91 ± 0.23	1.65 ± 0.55
AT+TT	221	4.84 ± 0.86	1.17 (0.67)	1.95 ± 0.50	2.64 ± 0.65	1.41 ± 0.23	0.90 ± 0.22	1.64 ± 0.45
P		0.014	0.064	0.383	0.000	0.299	0.518	0.170
rs330915
AA	409	4.89 ± 1.00	1.25 (0.85)	1.90 ± 0.48	2.77 ± 0.78	1.42 ± 0.26	0.90 ± 0.22	1.66 ± 0.54
AT	187	4.86 ± 0.87	1.22 (0.73)	1.92 ± 0.48	2.70 ± 0.74	1.40 ± 0.22	0.91 ± 0.26	1.62 ± 0.48
TT	21	4.16 ± 1.19	0.85 (0.37)	1.78 ± 0.59	2.24 ± 0.79	1.30 ± 0.37	0.80 ± 0.22	1.65 ± 0.38
P		0.003	0.080	0.488	0.009	0.119	0.095	0.765
AA	409	4.89 ± 1.00	1.25 (0.85)	1.90 ± 0.48	2.77 ± 0.78	1.42 ± 0.26	0.90 ± 0.22	1.66 ± 0.54
AT+TT	208	4.79 ± 0.93	1.18 (0.71)	1.91 ± 0.50	2.65 ± 0.75	1.39 ± 0.24	0.90 ± 0.25	1.63 ± 0.47
P		0.172	0.232	0.753	0.325	0.531	0.744	0.266
rs9949
GG	357	4.87 ± 1.02	1.25 (0.87)	1.89 ± 0.48	2.79 ± 0.80	1.41 ± 0.26	0.90 ± 0.22	1.66 ± 0.55
GA	222	4.89 ± 0.87	1.22 (0.68)	1.92 ± 0.49	2.69 ± 0.72	1.41 ± 0.23	0.91 ± 0.23	1.64 ± 0.47
AA	38	4.47 ± 1.07	1.06 (0.80)	1.88 ± 0.53	2.40 ± 0.72	1.34 ± 0.30	0.88 ± 0.25	1.58 ± 0.41
P		0.043	0.354	0.695	0.010	0.182	0.783	0.705
GG	357	4.87 ± 1.02	1.25 (0.87)	1.89 ± 0.48	2.79 ± 0.80	1.41 ± 0.26	0.90 ± 0.22	1.66 ± 0.55
GA + AA	260	4.82 ± 0.91	1.19 (0.70)	1.01 ± 0.50	2.65 ± 0.73	1.40 ± 0.24	0.90 ± 0.23	1.63 ± 0.46
P		0.044	0.126	0.585	0.011	0.401	0.594	0.104

TC total cholesterol, TG triglyceride, HDL-C high-density lipoprotein cholesterol, LDL-C low-denstity lipoprotein cholesterol, ApoA1 apolipoprotein A1, ApoB apolipoprotein B. The value of triglyceride was presented as median (interquartile range), and the difference among or between the genotypes was determined by the Kruskal-Wallis test or Wilcoxon-Mann-Whitney test. A P < 0.0014 was considered statistically significant after Bonferroni correction

PPP1R3B genotypes and the risk of CAD and IS

The associations of the four PPP1R3B SNPs and the risk of CAD and IS are shown in Table 3. After controlling for potential confounders including gender, sex, BMI, cigarette smoking, alcohol consumption, hypertension and hyperlipidemia, the genotypes of rs12785, rs330910 and rs9949 SNPs were associated with the risk of CAD (P = 0.008–0.004). The subjects with rs12785TA/AA genotypes (Dominant: OR = 1.32, 95%CI = 1.02–1.71, P = 0.009; Log-additive: OR = 1.25, 95%CI = 1.02–1.54, P = 0.007); rs330910AT/TT genotypes (Dominant: OR = 1.35, 95%CI = 1.04–1.76, P = 0.006; Log-additive: OR = 1.25, 95%CI = 1.04–1.60, P = 0.004); and rs9949GA/AA genotypes (Dominant: OR = 1.31, 95%CI = 1.02–1.73, P = 0.010; Log-additive: OR = 1.25, 95%CI = 1.02–1.53, P = 0.008) had higher risk of CAD, respectively. No significant association between the four SNPs and the risk of IS was observed.

Table 3

Association between the PPP1R3B polymorphisms and the risk of CAD and IS

SNP/Model	Genotype	OR (95%CI)CAD	P CAD	OR (95%CI)IS	P IS
rs12785
Dominant	TT	1.00		1.00
	TA + AA	1.32(1.02–1.71)	0.009	1.12(0.86–1.46)	0.405
Log-additive	–	1.25(1.02–1.54)	0.007	1.16(0.94–1.42)	0.178
rs330910
Dominant	AA	1.00		1.00
	AT+TT	1.35(1.04–1.76)	0.006	1.03(0.79–1.36)	0.812
Log-additive	–	1.29(1.04–1.60)	0.004	1.08(0.87–1.35)	0.489
rs330915
Dominant	AA	1.00		1.00
	AT+TT	1.15(0.88–1.51)	0.291	1.15(0.87–1.51)	0.330
Log-additive	–	1.19(0.95–1.49)	0.122	1.16(0.92–1.47)	0.218
rs9949
Dominant	GG	1.00		1.00
	GA + AA	1.31(1.02–1.73)	0.010	1.14(0.87–1.48)	0.341
Log-additive	–	1.25(1.02–1.53)	0.008	1.17(0.95–1.44)	0.137

SNP single nucleotide polymorphisms CAD coronary artery disease IS ischemic disease

LD analyses

As shown in Additional file 1 Table S3 and Fig. 1, there was strong LD among the rs12785, rs330910, rs330915 and rs9949 SNPs in the controls and CAD/IS patients. The LD among the rs12785, rs9949 and rs330910 was strong (D’ = 0.997–0.998, r2 = 0.751–0.991). The LD among the rs12785, rs9949 and rs330915 was also strong (D’ = 0.975–0.998, r2 = 0.669–0.991). The SNPs of rs12785 and 9949 almost reached a complete LD (r2 = 0.993), whereas the LD between the rs330910 and rs330915 SNPs was weak (D’ = 0.673).

Fig. 1

The linkage disequilibrium (LD) analyses among the four PPP1R3B SNPs of rs12785, rs330910, rs330915 and rs9949. (a) D’; (b) r2

Haplotypes and the risk of CAD and IS

As shown in Additional file 1 Table S4, the commonest haplotype was T-A-A-G (in the order of rs12785-rs330910-rs330915-rs9949), it represented more than 70% of the participants. The haplotype of T-A-A-G was also associated with a reduced risk for CAD (P = 0.014) and IS (P = 0.023). The haplotype of A-A-T-A was associated with an increased risk for IS (P = 0.036).

Gene-environment interactions on the risk of CAD and IS

As shown in Table 4, there were no significant gene-environment interactions after Bonferroni correction in this study (P > 0.0017 for all).

Table 4

The PPP1R3B SNP-environment interactions on the risk of CAD and IS

Factor	rs12785		rs330910		rs330915		rs9949
	P CAD	P IS	P CAD	P IS	P CAD	P IS	P CAD	P IS
Sex (female vs. male)	0.015	0.80	0.006	0.30	0.79	0.45	0.017	0.80
Age (≤ 60 vs. >  60 year)	0.84	0.41	0.75	0.12	0.50	0.71	0.85	0.42
BMI (≤ 24 vs. >  24 kg/m2)	0.89	0.12	0.78	0.65	0.78	0.48	0.92	0.12
Smoking (yes vs. no)	0.64	0.010	0.57	0.15	0.29	0.026	0.62	0.010
Drinking (yes vs. no)	0.61	0.92	0.41	0.41	0.62	0.87	0.57	0.96
hypertension (yes vs. no)	0.009	0.22	0.007	0.77	0.014	0.51	0.009	0.23
hyperlipidemia (yes vs. no)	0.25	0.12	0.32	0.092	0.28	0.37	0.24	0.14

PCAD/PIS, the P value between patients and control, BMI body mass index. The P values for interactions of genotypes and gender, age, BMI, drinking, smoking, hypertension, hyperlipidemia on the risk of CAD/IS were obtained from unconditional logistic regression, and a P < 0.0017 was considered statistically significant after Bonferroni correction

Stratified analyses of the PPP1R3B SNPs on the risk of CAD and IS

As described in Table 5, stratified analysis between the PPP1R3B polymorphisms and the risk of CAD and IS in dominant model showed that the rs12785A (OR = 1.99, 95% CI = 1.28–3.12), rs330910T (OR = 2.07, 95% CI = 1.32–3.26) and rs9949A allele (OR = 1.74, 95% CI = 1.12–2.71) carriers in females were associated with an increased risk of CAD compared with their own homozygotes (rs12785TT/rs330910AA/rs9949GG carriers); the rs12785A (OR = 1.82, 95% CI = 1.25–2.69), rs330910T (OR = 1.90, 95% CI = 1.29–2.80), rs330915T (OR = 1.69, 95% CI = 1.14–2.49); and rs9949A allele (OR = 1.82, 95% CI = 1.25–2.66) carriers in normotensives were associated with an increased risk of CAD compared with their own homozygotes (rs12785TT/rs330910AA/rs330915AA/rs9949GG carriers); and the rs12785A (OR = 1.57, 95% CI = 1.05–2.40) and rs9949A allele (OR = 1.56, 95% CI = 1.06–2.39) carriers in smokers were associated with an increased risk of IS compared with their own homozygotes (rs12785TT/rs9949GG carriers).

Table 5

Stratified analysis of the PPP1R3B genotypes and the risk of CAD or IS

Factor	Genotype	OR(95%CI)CAD	P CAD	OR(95%CI)IS	P IS
rs12785
Gender
Male	TT	1		1
	TA + AA	1.06(0.68–1.44)	0.830	1.15(0.80–1.56)	0.470
Female	TT	1		1
	TA + AA	1.76(1.13–2.73)	0.007	1.00(0.63–1.56)	0.970
Age, year
≤ 60	TT	1		1
	TA + AA	1.34(0.73–1.89)	0.280	1.03(0.70–1.50)	0.980
>  60	TT	1		1
	TA + AA	1.24(0.85–1.82)	0.260	1.23(0.84–1.80)	0.280
BMI, kg/m2
≤ 24	TT	1		1
	TA + AA	1.27(0.93–1.74)	0.970	1.25(0.91–1.71)	0.170
>  24	TT	1		1
	TA + AA	1.23(0.79–1.91)	0.370	0.80(0.49–1.32)	0.400
Smoking
No	TT	1		1
	TA + AA	1.34(0.97–1.87)	0.090	0.85(0.60–1.19)	0.330
Yes	TT	1		1
	TA + AA	1.15(0.78–1.71)	0.520	1.56(1.05–2.39)	0.009
Drinking
No	TT	1		1
	TA + AA	1.19(0.88–1.63)	0.270	1.06(0.77–1.45)	0.720
Yes	TT	1		1
	TA + AA	1.46(0.93–2.27)	0.100	1.03(0.66–1.61)	0.910
Hyperlipidemia
No	TT	1		1
	TA + AA	1.47(1.05–2.08)	0.090	0.92(0.66–1.32)	0.700
Yes	TT	1		1
	TA + AA	1.07(0.71–1.60)	0.740	1.39(0.94–2.05)	0.100
Hypertension
No	TT	1		1
	TA + AA	1.82(1.25–2.69)	0.002	1.28(0.91–1.80)	0.160
Yes	TT	1		1
	TA + AA	0.94(0.66–1.34)	0.750	0.92(0.59–1.34)	0.610
rs330910
Gender
Male	AA	1		1
	AT+TT	0.99(0.72–1.36)	0.960	0.91(0.66–1.27)	0.580
Female	AA	1		1
	AT+TT	2.07(1.32–3.26)	0.002	1.16(0.73–1.85)	0.530
Age, year
≤ 60	AA	1		1
	AT+TT	1.33(0.94–1.87)	0.110	0.84(0.58–1.21)	0.350
>  60	AA	1		1
	AT+TT	1.24(0.84–1.84)	0.280	1.27(0.86–1.88)	0.220
BMI, kg/m2
≤ 24	AA	1		1
	AT+TT	1.25(0.91–1.72)	0.170	1.04(0.75–1.43)	0.820
>  24	AA	1		1
	AT+TT	1.30(0.83–2.06)	0.260	0.91(0.57–1.46)	0.690
Smoking
No	AA	1		1
	AT+TT	1.37(0.97–1.92)	0.070	0.85(0.60–1.20)	0.360
Yes	AA	1		1
	AT+TT	1.16(0.77–1.74)	0.470	1.21(0.79–1.86)	0.390
Drinking
No	AA	1		1
	AT+TT	1.18(0.86–1.62)	0.300	1.06(0.76–1.48)	0.720
Yes	AA	1		1
	AT+TT	1.01(0.74–1.38)	0.960	0.82(0.52–1.30)	0.390
Hyperlipidemia
No	AA	1		1
	AT+TT	1.48(1.04–2.10)	0.030	0.81(0.56–1.16)	0.250
Yes	AA	1		1
	AT+TT	1.10(0.73–1.64)	0.650	1.30(0.87–1.93)	0.280
Hypertension
No	AA	1		1
	AT+TT	1.90(1.29–2.80)	0.001	1.05(0.74–1.49)	0.770
Yes	AA	1		1
	AT+TT	0.93(0.64–1.34)	0.680	0.94(0.62–1.42)	0.770
rs330915
Gender
Male	AA	1		1
	AT+TT	1.20(0.87–1.66)	0.270	1.22(0.87–1.70)	0.250
Female	AA	1		1
	AT+TT	1.12(0.71–1.77)	0.610	0.89(0.56–1.42)	0.630
Age, year
≤ 60	AA	1		1
	AT+TT	1.20(0.87–1.66)	0.270	1.09(0.75–1.57)	0.670
>  60	AA	1		1
	AT+TT	1.30(0.88–1.94)	0.190	1.18(0.79–1.75)	0.410
BMI, kg/m2
≤ 24	AA	1		1
	AT+TT	1.19(0.86–1.65)	0.290	1.20(0.86–1.66)	0.280
>  24	AA	1		1
	AT+TT	1.09(0.69–1.73)	0.700	0.98(0.61–1.57)	0.940
Smoking
No	AA	1		1
	AT+TT	1.31(0.93–1.84)	0.130	0.89(0.63–1.25)	0.510
Yes	AA	1		1
	AT+TT	0.97(0.64–1.48)	0.910	1.54(0.99–2.40)	0.060
Drinking
No	AA	1		1
	AT+TT	1.21(0.88–1.68)	0.240	1.09(0.78–1.53)	0.610
Yes	AA	1		1
	AT+TT	1.10(0.69–1.76)	0.680	1.04(0.66–1.65)	0.870
Hyperlipidemia
No	AA	1		1
	AT+TT	1.37(0.96–1.96)	0.070	1.03(0.72–1.48)	0.870
Yes	AA	1		1
	AT+TT	1.04(0.69–1.57)	0.860	1.31(0.87–1.97)	0.190
Hypertension
No	AA	1		1
	AT+TT	1.69(1.14–2.49)	0.008	1.24(0.87–1.77)	0.240
Yes	AA	1		1
	AT+TT	0.89(0.61–1.30)	0.530	1.00(0.66–1.51)	0.990
rs9949
Gender
Male	GG	1		1
	GA + AA	1.07(0.79–1.46)	0.660	1.13(0.82–1.55)	0.470
Female	GG	1		1
	GA + AA	1.74(1.12–2.71)	0.008	1.00(0.63–1.56)	0.980
Age, year
≤ 60	GG	1		1
	GA + AA	1.28(0.91–1.80)	0.150	1.03(0.72–1.47)	0.960
>  60	GG	1		1
	GA + AA	1.24(0.85–1.82)	0.260	1.24(0.84–1.81)	0.280
BMI, kg/m2
≤ 24	GG	1		1
	GA + AA	1.27(0.93–1.73)	0.960	1.25(0.91–1.71)	0.170
>  24	GG	1		1
	GA + AA	1.23(0.79–1.92)	0.37	0.82(0.52–1.31)	0.410
Smoking
No	GG	1		1
	GA + AA	1.34(0.96–1.86)	0.086	0.85(0.60–1.19)	0.330
Yes	GG	1		1
	GA + AA	1.15(0.77–1.71)	0.520	1.56(1.06–2.39)	0.009
Drinking
No	GG	1		1
	GA + AA	1.190.88–1.63)	0.260	1.06(0.77–1.46)	0.720
Yes	GG	1		1
	GA + AA	1.46(0.93–2.27)	0.100	1.03(0.66–1.61)	0.910
Hyperlipidemia
No	GG	1		1
	GA + AA	1.48(1.05–2.09)	0.030	0.93(0.66–1.33)	0.700
Yes	GG	1		1
	GA + AA	1.07(0.72–1.59)	0.740	1.39(0.94–2.05)	0.100
Hypertension
No	GG	1		1
	GA + AA	1.82(1.25–2.66)	0.002	1.28(0.91–1.80)	0.160
Yes	GG	1		1
	GA + AA	0.94(0.66–1.35)	0.750	0.91(0.61–1.35)	0.630

P, the P value between patients and control, BMI body mass index. The P values for stratified analysis between the PPP1R3B polymorphisms on the risk of CAD/IS, were obtained from unconditional logistic regression, when one of the environmental factors was detected, the rest six factors in gender, age, BMI, smoking, alcohol consumption, hypertension and hyperlipidemia were adjusted

Haplotype-environment interactions on the risk of CAD and IS

The interactions of several haplotypes and envirnmental factors on the risk of CAD were also noted in this study. When one of the environmental factors was detected, the rest six factors of gender, age, BMI, smoking, alcohol consumption, hypertension and hyperlipidemia were adjusted.

For the interactions of haplotype-gender on the risk of CAD, the haplotypes of A-T-T-A (OR = 1.70, 95% CI = 1.09–2.65) and A-T-A-A (OR = 3.49, 95%CI = 1.70–7.17) in females were associated with an increased risk of CAD as compared to the T-A-A-G haplotypes (interacted P-value = 0.018). As compared with the same haplotype, the haplotypes of T-A-A-G (OR = 3.02, 95%CI = 2.07–4.39), A-T-T-A (OR = 2.07, 95%CI = 1.32–3.26), and A-A-T-A (OR = 4.09, 95%CI = 1.62–9.31) in males were associated with an increased risk for CAD.

For the interactions of the haplotype-hypertension on the risk of CAD, the haplotype of A-T-T-A (OR = 1.68, 95% CI = 1.18–2.39) in normotensives was associated with an increased risk for CAD compared to the T-A-A-G haplotype. As compared with the haplotype in normotensives, the haplotypes of T-A-A-G (OR = 0.61, 95% CI = 0.42–0.88) and A-T-T-A (OR = 0.35, 95% CI = 0.22–0.55) in hypertension participants were associated with a decreased risk for CAD.

Discussion

Previous study has showed that the PPP1RB polymorphisms were associated with hepatic steatosis [26] and plasma serum CRP traits [24, 44]. Ligthart et al. [45] reported that the PPP1R3B regions had a pleiotropic effect on CRP independent of the effects on the cardiometabolic phenotypes. Two previous GWASes showed that the PPP1R3B polymorphisms were associated with serum lipid traits, and higher expression of PPP1R3B would influence the lipid traits in animal studies [22, 23]. Furthermore, the PPP1R3B variants were found to influence cardiovascular events and atherosclerosis. Zhang et al. [24] reported that the PPP1R3B polymorphisms were associated with LDL-C and serum CRP levels. In a recent study, however, López-Mejías et al. [46] reported that the PPP1R3B SNPs were not associated with cardiovascular disease in rheumatoid arthritis patients. In the present study, we firstly disclose that the genetic polymorphisms in the 3’UTR of PPP1R3B were associated with serum LDL-C levels, the risk of CAD and IS in a Southern Chinese Han population.

In the present study, we showed that the A and T allele frequencies of rs12785 in controls were 24% and 76%; the A and G allele frequencies of rs9949 were 24% and 76%; the T and A allele frequencies of rs330910 were 20% and 80%; and the T and A allele frequencies of rs330915 were 19% and 81%, respectively. However, the data from the International HapMap project showed that the allele frequencies of the four SNPs were different. For SNPs of rs12785 and rs9949, which were in complete LD in CHB (Han Chinese in Beijing, China) and CEU (Utah residents with ancestry from northern and western Europe) both with an r2 = 1, in strong LD in CLM (Colombians from Medellin, Colombia) with an r2 = 0.992. The A and T allele frequencies of rs12785 were 29.6% and 70.4% in CHB (A < T); 76.8% and 23.2 in CEU (A > T); 65.9% and 34.1% in CLM (A < T). The A and G allele frequencies of rs9949 were 29.6% and 70.4% in CHB (A < G), 76.8% and 23.2% in CEU (A > G), 65.4% and 34.6% in CLM (A > G). For the rs330910 SNP, the T and A allele frequencies of rs330910 were 21.4% and 78.6% in CHB (T < A), 76.3% and 23.7% in CEU (T > A), and 55.4% and 44.6% in ITU (Indian Telugu from the UK). For the rs330915 SNP, meanwhile, the T and A allele frequencies were different among controls in our study, CHB, CEU and CLM. The allele frequencies were different between case and control in our research, just like rs330915 associated with CAD, rs12785/rs9949 associated with IS, even though it confirmed that the allele might not be associated with the risk of diseases after controlling potential confounders. These differences might due to racial-specificity and indicated that variants in the 3’UTR of PPP1R3B interacted with environment factors on the risk of diseases. More and larger samples of epidemiological investigation are necessary to confirm the interactions.

The results of our present study showed that serum LDL-C levels were different among the three genotypes of the rs330910 SNP. The rs330910T allele carriers in controls had lower serum LDL-C levels than the rs330910T allele non-carriers (P < 0.0014). In a previous GWAS, Teslovich et al. showed that the SNP in PPP1R3B was associated with HDL-C, TC and LDL-C [22]. But in our current study, no significant association was found in TC and HDL-C among the PPP1R3B genotypes. These findings might due to SNP chosen in GWAS located in different functional consequence and different sample sizes. Thus, the association between the PPP1R3B SNPs and serum lipid levels need to be confirmed by proteomics study in further research.

In the current study, we firstly showed that the rs12785, rs330910 and rs9949 SNPs were associated with CAD but not with IS. Haplotype analysis of the four SNPs showed that T-A-A-G was associated with a decreased risk for CAD and IS, and A-A-T-A was associated with an increased risk for IS. Due to rs12785 and rs9949 SNPs were in complete LD, the results of two SNPs were exactly similar to each other. The interactions of the PPP1R3B haplotypes and some environmental factors on the risk of CAD and IS remain unknown. In the present study, we firstly found that the haplotypes of A-T-T-A (OR = 1.70, 95% CI = 1.09–2.65) and A-T-A-A (OR = 3.49, 95%CI = 1.70–7.17) were associated with an increased risk of CAD. The haplotype of A-T-T-A (OR = 1.68, 95% CI = 1.18–2.39) was associated with an increased risk of CAD in normotensives. However, these findings need to be confirmed in the other populations with larger sample sizes.

Several potential limitations should be pointed out in this study. Firstly, the sample size was relatively small compared to many GWASes and molecular epidemiological investigations. Therefore, larger sample sizes are needed to confirm our results. Secondly, although several confounders have been adjusted for the statistical analyses in this study, individuals exposed to other genetic and environmental risk factors probably modify the association of genetic polymorphisms and risk of CAD and IS. Thirdly, we could not analyze the association of the PPP1R3B SNPs and serum lipid levels in the CAD/IS patients because of the influence of lipid-lowering drugs. Finally, it is well known that both CAD and IS are complex multifactorial disorder. Although we have detected the association of four PPP1R3B SNPs and their haplotypes with serum lipid levels and the risk of CAD and IS in this study, there are still many unmeasured genetic and environmental factors and their interactions. Thus, our findings still need to be confirmed in the other populations with larger sample sizes.

Conclusions

Our present study suggests that the SNPs of rs12785, rs330910, and rs9949 were associated with the risk of CAD. The rs330910 SNP was associated with LDL-C levels. The rs330910T allele carriers in controls had lower LDL-C levels than the homozygote of rs330910AA. The carriers of rs12785A, rs330910T, rs9949A were associated with increased risk of CAD. The haplotype of T-A-A-G was associated with decreased risk of CAD and IS, and haplotype of A-A-T-A was associated with an increased risk of IS. These results suggest that the detected PPP1R3B SNPs were associated with serum LDL-C levels and the risk of CAD and IS in the Southern Chinese Han population. Detection of these PPP1R3B SNPs in our study population may have important significance for early diagnosis and future individualized treatment of dyslipidemia, CAD and IS.

Table S1. Genotypic and allelic frequencies of the PPP1R3B polymorphisms in control and patients [n (%)]. Table S2. Association between the PPP1R3B genotypes and BMI or blood pressure in controls. Table S3. Linkage disequilibrium analysis of the 4 PPP1R3B SNPs. Table S4. Haplotype analysis of the 4 PPP1R3B SNPs. Figure S1. Annotation of SNPs in UCSC. Non-coding (ncRNA) and untranslated region (3’UTR) are colored blue. (DOC 392 kb)