Correlation Between High-Density Lipoprotein and Monocyte Subsets in Patients with Stable Coronary Heart Disease.

BACKGROUND: High-density lipoprotein (HDL) consists of heterogeneous particles with a variety of structures and functions. Its role in atherosclerosis has been gradually recognized. Studies have shown dysfunction of small HDL in patients with coronary artery disease (CAD). Monocytes play an important role in atherosclerosis, which can be divided into 3 subgroups based on the expression of surface markers CD14 and CD16. This study aimed to investigate the association between HDL and monocyte subsets in CAD patients.
MATERIAL AND METHODS: A total of 90 patients with stable CAD were selected in this study. Monocytes were divided into classical monocytes (CM, CD14++CD16-), intermediate monocytes (IM, CD14++CD16+), and non-classical monocytes (NCM, CD14+CD16++). HDL components in serum were determined by high-resolution polyacrylamide gel electrophoresis (detected by Quantimetrix HDL Lipoprint system, referring to HDL subfractions analysis: A new laboratory diagnostic assay for patients with cardiovascular diseases and dyslipoproteinemia).
RESULTS: Serum level of small HDL was positively correlated with circulating proinflammatory NCM (r=0.30; p=0.004), negatively correlated with CM, and not correlated with IM. We also found that disease severity was not associated with diabetes mellitus, glycosylated hemoglobin, hypertension, smoking history, or statin dosage.
CONCLUSIONS: Our study confirmed that small HDL level is associated with an increase in NCM and a decrease in CM, suggesting the proinflammatory relationship between small HDL and intrinsic immune function during the progression of stable CAD.

Background

Cardiovascular disease remains the leading cause of death in elderly patients worldwide despite the rapid development of cardiovascular drugs. Several clinical and epidemiological researches have indicated that the level of high-density lipoprotein (HDL) is closely related to cardiovascular disease. Studies on the treatment of such cardiovascular disease have been focused on how to improve HDL level [1]. Nevertheless, increasing HDL level by cholesteryl ester transfer protein (CETP) inhibitor has not reduced the risk of coronary heart disease, but increased the morbidity and mortality rate in patients with such disease [2,3]. Some researchers have suggested that the function of HDL might have been damaged in this pathological environment; therefore, it is crucial to ensure its function. HDL consists of heterogeneous lipoprotein particles characterized by specific structures, metabolic functions, and atherosclerosis-resistance. Small HDL in healthy people has been confirmed to impact the progression of atherosclerosis by increasing the consumption of cholesterol, as well as antioxidant and anti-inflammatory responses. However, its function is abnormal in patients with atherosclerosis and dyslipidemia [4-6]. Clinical research shows that small HDL is associated with the incidence and severity of CAD, whereas large molecular HDL showed a negative correlation.

Monocytes play an important role in the inflammatory reaction during atherosclerosis. Monocytes can be divided into 3 subgroups based on the expression of surface CD14 and CD16: classical monocytes (CM, CD14++CD16−), intermediate monocytes (IM, CD14++CD16+), and non-classical monocytes (NCM, CD14+CD16++). Both IM and NCM are proinflammatory cells, whose proportion is related to the occurrence of CAD, intima-media thickness, and plaque stability. Total cholesterol, low-density lipoprotein (LDL) cholesterol, and triglycerides are related to proinflammatory NCM, whereas HDL has a negative correlation. A previous study on 900 cases of CAD patients showed that IM has a predictive value for cardiovascular disease [7-11]. In the current study, we explored the correlation between different HDL components and monocytes subsets in CAD patients.

Material and Methods

Patients and study design

All subjects were CAD patients treated in our hospital between September 2009 and August 2010. Inclusion criteria were: age above 18 years and, with stable CAD diagnosed by selective coronary angiography. Exclusion criteria were: recent occurrence of acute coronary syndrome with ST segment elevation myocardial infarction, non-ST elevation myocardial infarction, or unstable angina, a history of percutaneous coronary intervention (PCI), cardiac failure, cancer, and acute or chronic liver or renal failure. All selected patients provided signed informed consent. The study was approved by the Institutional Ethics Committee in our hospital.

Blood sample

Before the selective coronary arteriography, a blood sample was collected from the antecubital vein of each patient, using 3.8% sodium citrate tube, serum separation tube, and EDTA tube (Greiner Bio-One, Frickenhausen, Germany) and centrifuged at 3000 rpm at 4°C for 15 min.

Laboratory detection

The levels of granulocyte colony-stimulating factor (g-CSF), macrophage colony-stimulating factor (M-CSF), and interleukin-6 (IL-6) were determined by enzyme-linked immunosorbent assay (ELISA; R&D Systems, Minneapolis, MN). The levels of IL-10 and granulocytes-macrophage colony-stimulating factor (GM-CSF) were determined by Luminex assay (R&D Systems, catalog number FCST03).

Flow cytometry

White blood cells and monocyte subsets were detected by flow cytometry. The staining and gating of cells is shown in Figure 1. CD45-PerCP monoclonal antibody (BD Biosciences, catalog number 345809, San Diego, CA, USA), CD14-FITC monoclonal antibody (BD Biosciences, catalog number 345784), CD16-APC-H7 monoclonal antibody (BD Biosciences, catalog number 560195), CD3-APC monoclonal antibody (BD Biosciences, catalog number 345767), CD19 (BD Biosciences, catalog number 345791), CD56 (Beckton Dickinson, catalog number 341027), and isotype control were used for staining. The cells were incubated in the dark for 15 min, mixed with 1.5 ml of lysate (BD FACS lysing solution, BD Biosciences), incubated in the dark for another 15 min, and washed 3 times with PBS. Cells were resuspended in 3 ml of fixing solution and analyzed by flow cytometry. Data analyses were performed using FACS Canto II and FACS Diva software (BD Biosciences). CD45+CD3− and CD19–CD56− cells with specific forward scatter (FSC) and side scatter (SSC) were monocytes. The monocytes were divided into CM, IM, and NCM as previously described. The absolute number of monocytes was calculated based on the number of white blood cells and CD45+ cells detected in flow cytometry.

Figure 1

Identification of monocyte subsets. Monocytes were defined as CD45+ cells (B) with typical FSC and SSC (A). CD3−, CD19−, and CD56− labeled cells (C) were excluded for possible T cell contamination. CD45+CD3/19/56− cells were monocytes (D).

Detection of lipid

The serum levels of total cholesterol, HDL, LDL, and triglyceride were determined. HDL subgroup components were quantified using the Quantimetrix HDL Lipoprint System® (Quantimetrix Corporation, Redondo Beach, CA, USA). HDL was divided into 10 subgroups based on their locations on SDS-PAGE: 1–3, 4–7, and 8–10 represented large, medium, and small HDL particles, respectively (Figure 2).

Figure 2

Quantimetrix HDL Lipoprint system separate and quantify HDL subgroup component.

Statistical analysis

All statistical analyses were performed using SPSS20.0 software (Chicago, IL, USA). Classified variables are expressed as count or percentage and compared by χ2 or Fisher exact tests. Numerical data are presented as mean±standard deviation (χ̄±S) and analyzed by one-way ANOVA. Deflection data were compared by ANOVA after logarithmic transformation. Relevance was determined by Pearson correlation analysis. Three subgroups (CM, IM, and NCM) were incorporated into the linear regression model. These subgroups were also incorporated into the model when the clinical features, statins usage, or lipid parameters were correlated with monocyte subgroups or small HDL level (p<0.2). P values smaller than 0.05 were considered significantly different.

Results

Patient information

A total of 90 patients diagnosed as having stable CAD by angiography were enrolled in this study, including 72 males (80%) and 21 smokers (23%). The mean age of the patients was 64.1±10.0 years old. Among these patients, a total of 25, 36, and 29 cases suffered from single-, 2-, and 3-vessel coronary arterial disease, respectively; 31% and 52% patients received high- and low-dose statin treatment, respectively and 17% patients received no statin therapy.

Correlation between monocytes subgroups and serum level of small HDL

Monocyte subsets were analyzed by flow cytometry (Figure 1). The mean density of CM, NCM, and IM was 270.5±142.7 cells/μl (82.1±6.7%), 39.7±28.9 cells/μl (12.3±5.9%), and 18.7±15.1 cells/μl (5.6±3.3%), respectively. Serum level of small HDL was negatively correlated with circulating CM (r=−0.33, p=0.001; Figure 3A), positively correlated with NCM (r=0.30, p=0.004; Figure 3B), and not correlated with IM (r=0.14, p=0.20; Figure 3C). As shown in Table 1, no correlation was observed between monocyte subsets and medium HDL, large HDL, and total HDL. Linear regression analysis revealed that small HDL was independently correlated with proinflammatory NCM and circulating CM compared with other lipid parameters, risk factors, and statin. As shown in Table 2, IM was only correlated with total cholesterol and LDL.

Figure 3

Correlation between serum level of small HDL and monocyte subsets.

Table 1

Correlation between HDL subgroups and circulating monocyte subsets.

Monocyte subgroup	CM CD14++CD16−		IM CD14++CD16+		NCM CD14+CD16++
	R	P value	R	P value	R	P value
Total HDL	−0.08	0.45	−0.06	0.60	0.12	0.25
Small HDL	−0.33	0.001	0.14	0.20	0.30	0.004
Medium HDL	−0.05	0.66	−0.32	0.76	0.07	0.50
Large HDL	0.06	0.55	−0.12	0.26	−0.01	0.96

The significant differences are indicated in bold.

Table 2

Multivariable regression model for the correlation between small HDL and monocyte subsets.

	CM CD14++CD16−			IM CD14++CD16+			NCM CD14+CD16++
	Univariate P value	β	P value	Univariate P value	β	P value	Univariate P value	β	P value
Small HDL	0.001	−0.33	0.006	0.20	0.21	0.08	0.004	0.27	0.002
Statin	0.21	0.14	0.19	0.45	−0.07	0.49	0.31	−0.12	0.25
Serum	0.36	−0.11	0.35	0.19	−0.18	0.12	0.07	0.23	0.05
VLDL	0.21	0.18	0.48	0.22	0.07	0.78	0.47	−0.25	0.32
Age	0.37	0.09	0.50	0.98	0.23	0.09	−0.23	−0.21	0.08
LDL	0.02	−0.19	0.56	0.05	0.80	0.01	0.11	−0.23	0.45
BMI	0.84	0.04	0.71	0.73	0.02	0.84	0.67	−0.06	0.58
Total cholesterol	0.02	−0.10	0.77	0.18	−0.86	0.01	0.06	0.60	0.08
Smoking history	0.51	−0.03	0.78	0.02	0.32	0.006	0.61	−0.14	0.21
Hypertension	0.49	−0.03	0.81	0.12	0.16	0.15	0.94	−0.06	0.60
Triglyceride	0.71	0.06	0.81	0.74	0.22	0.31	0.82	−0.19	0.39
Gender	0.96	0.01	0.99	0.12	0.14	0.28	0.42	−0.08	0.54
Total model R2		21%	0.040		24%	0.035		25%	0.031

CM value was the lowest (79.3±7% vs. 83.7±6% and 83.9±6%; p=0.004; Figure 4A) with the serum level of small HDL at the high tertiles (13–20 mg/L) compared with that at medium (9–12 mg/L) and low tertiles (2–8 mg/L). Furthermore, the density of proinflammatory NCM was highest (14.7±7% vs. 10.7±5% and 10.8±5%; p=0.006; Figure 4B) in patients with small HDL at the high tertiles, whereas IM was not correlated with the tertiles of small HDL (5.9±3% vs. 5.6±3% vs. 5.3±3%; p=0.54; Figure 4C).

Figure 4

Correlation between level grade of small HDL and monocyte subsets.

Correlation between HDL subsets with lipid parameters and cardiovascular risk factors

Small HDL level was significantly correlated with triglycerides, VLDL, LDL, and total cholesterol (Table 3), but no association was observed between small HDL level and total HDL level. Large HDL was negatively correlated with small HDL, VLDL, and triglycerides, and was highly associated with total cholesterol. Medium HDL was correlated with large HDL, LDL, and VLDL, but not with triglycerides. Medium HDL level (28.8±7.1 vs. 23.7±5.8 mg/dl; p=0.002) and large HDL level (19.3±11.1 vs. 10.7±5.7 mg/dl; p=0.005) in female patients was significantly higher compared with male patients, whereas there was no significant difference in small HDL between female and male patients (11.3±3.6 vs. 11.3±4.3 mg/dl; p=0.99). We also found that large HDL was negatively correlated with weight (r=−0.28; p=0.008). In addition, diabetes mellitus, glycosylated hemoglobin, hypertension, and smoking history were not related to HDL subgroup, statin dosage, or severity of disease.

Table 3

Correlation between HDL subgroups and lipid index.

	Small HDL		Medium HDL		Large HDL
	R	P value	R	P value	R	P value
HDL	0.060	0.56	0.68	<0.0001	0.80	<0.0001
Total cholesterol	0.39	<0.0005	0.60	<0.0001	0.24	0.023
LDL	0.30	<0.005	0.43	<0.0001	0.11	0.30
VLDL	0.42	<0.0001	0.22	0.042	−0.22	0.035
Triglycerides	0.36	<0.005	0.02	0.86	−0.37	<0.0005
Small HDL	–		0.01	0.91	−0.26	0.014
Medium HDL	0.01	0.91	–		0.63	<0.0001
Large HDL	−0.26	0.014	0.63	<0.0001	–

HDL – high density lipoprotein; LDL – low-density lipoprotein; VLDL – very low density lipoprotein. The significant differences are indicated in bold.

Association between HDL subgroups with colony-stimulating factor and inflammation markers

Serum level of small HDL was significantly correlated with granulocyte colony-stimulating factor (G-CSF; r=0.22, p=0.05) but not with granulocyte macrophage colony-stimulating factor (GM-CSF; r=0.05, p=0.66) or macrophage colony-stimulating factor (M-CSF; r=−0.09, p=0.37). Medium, large, and total HDLs were not related to any of the 3 types of CSF. Serum level of small HDL was not associated with proinflammatory marker hsCRP (r=−0.05; p=0.64), IL-6 (r=−0.10; p=0.38) or IL-10 (r=0.06; p=0.62). In addition, intermediate-density lipoprotein, large HDL, and total HDL were not related to hsCRP, IL-6, or IL-10. Neither of the monocyte subsets were correlated with hsCRP (CM: R=−0.11, p=0.32; IM: R=0.14, p=0.19; NCM: R=0.04, p=0.68), IL-6 (CM: R=0.06, p=0.59; IM: R=0.13, p=0.24; NCM: R=−0.14, p=0.24), or IL-10 (CM: R=0.01,p=0.94; IM: R=−0.01, p=0.98; NCM: R=−0.01, p=0.94).

Discussion

Manyepidemiological and prospective studies have clearly shown that serum HDL level is negatively correlated with the risk of coronary heart disease. HDL exerts a variety of protective effects on arteries, including cholesterol outflow, antioxidation, anti-inflammation, cell protection, vasodilator, and antithrombosis [12]. Moreover, several studies have also confirmed that small HDL particles can potentially prevent atherosclerosis. During dyslipidemia, including elevation of triglycerides or total cholesterol, small HDL level was significantly increased, whereas the number of large HDL particles was substantially reduced, leading to significant change in HDL metabolism and distribution of subsets. This study confirmed that serum level of small HDL was correlated with lipid index, such as total cholesterol, LDL, VLDL, and triglyceride, but not with total HDL, lipoprotein, or statin usage in 90 patients with stable CAD diagnosed by angiography.

The level of small HDL changes in patients with dyslipidemia or obesity and in patients with cardiovascular disease [13]. A study of 115 patients with CAD suggested that large HDL level was significantly increased, as revealed by coronary angiography. Another 10-year follow-up study covering 1000 patients revealed that small and large HDLs have prognostic value to the progression of ischemic heart disease [14-16]. The incidence of CAD in females is more closely related to smaller HDL particles. Furthermore, small HDL level in patients with acute ischemic shock is significantly higher than that in healthy populations. A clinical study of 60 patients has confirmed that small HDL is associated with non-calcified plaques by coronary artery CT and intravascular ultrasound. In another study, covering 102 patients with myocardial infarction and 200 healthy controls, large and middle HDL are negatively correlated with early-stage acute myocardial infarction, whereas small HDL level is up-regulated in young patients with acute myocardial infarction. Furthermore, small HDL level is also increased in patients with acute coronary syndrome, whereas large HDL level is reduced [17,18].

The heterogeneity of monocytes and its association with atherosclerosis has been confirmed by the detection of expression of surface markers CD14 and CD16. The number of CD16+ monocytes is increased in acute and chronic inflammation and atherosclerosis, which are rapidly activated by the stimulation of inflammation [19]. Our study demonstrated that the increase in small HDL level was associated with the distribution of proinflammatory monocyte subsets in patients with stable CAD. Specifically, serum level of small HDL was positively correlated with cell density of NCM (CD14+CD16++), negatively correlated with CM (CD14++CD16−), and not correlated with IM (CD14++CD16+). Several studies have shown that the correlation between total HDL or LDL and monocyte subsets disappears after adjusting for BMI. In this study, the correlation between small HDL and monocyte subsets was independent of age, sex, smoking history, BMI, diabetes, and statin usage.

The patients were divided into 3 subgroups based on the level of small HDL. Monocyte subsets significantly promoted atherosclerosis and other inflammatory responses in the group with the highest level of small HDL. In addition, the proportion of NCM was increased while that of CM was decreased. However, monocyte subsets were not relevant to serum CRP, IL-6, or IL-10. Currently, the association between CD16+ monocytes and hsCRP level remains controversial. While some studies report that CD16+ monocytes were correlated with hsCRP level in patients with unstable angina, others have shown that CD16+ monocytes were related to TNF-α level instead of the level of hsCRP or IL-6 [20]. Furthermore, it has been confirmed that several colony-stimulating factors (CSFs) are expressed in various vascular cells, which can affect the progression of atherosclerosis by regulating macrophage phenotype and cholesterol intake [21]. In this study, the plasma level of G-CSFs was significantly correlated with small HDL. The association between HDL subgroups and monocyte subsets was only analyzed at a certain time point; therefore, their association with functional changes in the process of atherosclerosis cannot be determined.

Conclusions

Our study has demonstrated that small HDL level is associated with an elevation in IM and a reduction in CM, revealing the proinflammatory correlation between small HDL and intrinsic immune function in stable CAD.