Serum pigment epithelium-derived factor levels are independently correlated with the presence of coronary artery disease

The pigment epithelium-derived factor (PEDF) belongs to the super family of serine protease inhibitors, and the major sources of circulating PEDF are liver and adipose tissues [1‐3]. PEDF is characterized as a multifunctional protein possessing anti-angiogenic, anti-tumorigenic, anti-oxidant, anti-inflammatory, anti-thrombotic, and neuroprotective properties. Demonstrated as a highly effective anti-angiogenic factor, PEDF not only is capable of inhibiting vascular endothelial growth and migration, but can also suppress secretion of angiogenic factors [4, 5] as well as activate the FAS-FAS ligand death pathway to stimulate endothelial cell apoptosis [6]. Nakamura et al. reported anti-oxidative effects of PEDF, showing that PEDF-mediated suppression of NADPH oxidase inhibited generation of reactive oxygen species and the subsequent neointimal hyperplasia induced by balloon injury [7]. Takenaka et al. also showed a cardio-protective function by which PEDF inhibited occlusive thrombus formation in the carotis artery of a rat model [8]. Finally, various studies have demonstrated the strong anti-inflammatory activities of PEDF; for example, PEDF was characterized both as a regulator of cytokines’ expression such as monocyte chemoattractant protein-1 and tumor necrosis factor-α as well as a mediator of macrophage and T cell function [9‐11].

PEDF may rely on any of these protective properties to manifest a counteractive mechanism during the development of atherosclerosis and cardiovascular disease. Recent clinical findings have revealed that PEDF levels are closely associated with the presence of cardiovascular disease. Circulating PEDF levels were shown to be higher in subjects with metabolic syndrome (MetS) and to be correlated with the extent of MetS components [12‐14]. The fact that MetS itself, and each of its components, are important risk factors of cardiovascular disease which has led researchers to hypothesize that increased PEDF levels might have occurred as a counter-regulatory response to the presence of vascular injury. A recent study showed that serum PEDF level was independently correlated with intima-media thickness and vascular inflammation, suggesting an association of PEDF with subclinical atherosclerosis in at least two aspects: morphological abnormalities of the vessel, and inflammation in the plaques [15].

The above clinical findings revealed that PEDF levels were closely associated with the presence of cardiovascular disease, however, no study to date has focused on the relationship between serum PEDF and coronary artery disease (CAD). Since such data may also help to identify PEDF as a promising therapeutic target for atherosclerosis and cardiovascular disease [16, 17], the present study was carried out with Chinese patients who underwent coronary angiography to investigate the association between serum PEDF and CAD.

MetS is well recognized as an aggregate of cardio-metabolic risk factors subsequent cardiovascular disease and type 2 diabetes mellitus [23, 24], and the association of PEDF with MetS and its components has been verified in previous studies [12‐14]. In a 10-year prospective study, enhanced PEDF level was identified as an independent predictor for the development of MetS in men [13], which further affirmed the prediction effect of PEDF on MetS. However, these studies focused solely on subjects with CAD risk factors but no clinical evidence of cardiovascular disease, which may have limited their ability to extend this relation of serum PEDF with MetS to the subsequent onset of cardiovascular disease. To overcome this limitation, the present study enrolled CAD and non-CAD subjects (~3:4 ratio) as well as MetS and non-MetS subjects (~2:3 ratio) and still found that serum PEDF level was significantly increased in the MetS patient group. This finding agrees with previous functional studies that indicated enhanced levels of PEDF might exert a counteractive activity against simultaneous enhancement of CAD risk factors [13, 14].

In the current study of the broad panel of MetS pathogenic components, PEDF was found to be significantly positively correlated with hypertriglyceridemia, and TG was identified as an independent influencing factor of serum PEDF levels. Dyslipidemia is known to play an important role in the development of atherosclerosis and CAD. Specifically, long-term dyslipidemia has been shown to cause physical injury to the vascular intima, and the accompanying pro-inflammatory state of the tissue promotes atherosclerosis and CAD. In clinical studies, patients with elevated TG levels developed CAD more frequently than their counterparts with TG levels in the normal range and TG has been identified as an independent risk factor for coronary events [25, 26]. In the current study, TG was found to be strongly and independently correlated with serum PEDF in patients with CAD or at high risk of CAD. Moreover, these findings also agree with the previous studies in patients without CAD history [12, 13]. It is possible that the enhanced levels of serum PEDF in CAD or CAD-risk patients reflect a response to hypertriglyceridemia, whereby the body is attempting to correct the perturbance in the lipid metabolism system [27]. Intriguingly, we failed to find any significant association of serum PEDF with the other two principal components of MetS, central obesity and hypertension. These negative findings may be explained by the fact that a large portion of our subjects had been receiving therapeutic treatments for hyperglycemia, hypertension, or hyperlipidemia for long periods prior to study participation; to account for this potential confounding aspect, hypoglycemia therapy was entered the equation in the multiple regression analysis of PEDF levels. Body weight and W in subjects of this study might have changed in response to a long-term influence of medications; therefore, the relation of PEDF with obesity might be masked by other determinants.

Substantial evidence exists to support the hypothesis that PEDF plays a protective role against microangiopathy; moreover, PEDF has been proposed as a potential protective factor against diabetic microvascular complications [28‐30]. Yet, few studies to date have investigated the association of PEDF with macroangiopathy, as was done in the current study which demonstrated for the first time that serum PEDF was independently positively correlated with CAD in a Chinese population. This finding contradicts a previous study by Shiga et al., which showed no significant correlation between PEDF and the presence of CAD in a Japanese population [31]; however, this apparent inconsistency may merely indicate the different inclusion subjects.

Atherosclerosis, which is considered the pathological basis of CAD, is a chronic inflammatory disorder [32]. Previous studies have shown that CRP was positively correlated with PEDF, and suggested that this relation may reflect the body’s attempt to suppress a detrimental inflammatory reaction involving the endothelial cells [14, 33]. In our study, we found no difference in the CRP levels of CAD and non-CAD patients, possibly because the majority of those non-CAD patients actually possessed many of the CAD risk factors and >60% of them were MetS patients. Regardless, the multivariate stepwise regression analysis performed for the study population identified CRP as an independent influencing factor of serum PEDF levels, and this finding agrees with the proposed association of PEDF with chronic inflammation. Furthermore, oxidative stress, which is elevated in patients with metabolic disorder or CAD, could be one of the triggers in the liver (an important source organ of circulating PEDF), as hydrogen peroxide has been shown to induce PEDF expression in human hepatocytes [34].

Given the fact that PEDF exerts anti-angiogenic, anti-oxidative, anti-thrombotic, and anti-inflammatory properties on vascular tissues, researchers have speculated that PEDF levels might be elevated to counteract generation of a pro-atherosclerotic environment induced by vascular injuries [16, 17]. Ueda et al. reported that PEDF injection suppressed cardiac fibrosis, inhibited tissue remodeling and improved cardiac function in a rat model of acute myocardial infarction and suggested that PEDF may be a novel therapeutic strategy for human acute myocardial infarction [35]. Since PEDF exerts a number of protective effects on vascular and myocardial tissues, elevated serum PEDF levels in CAD patients may also play a counter-regulatory and protective role against vascular damage caused by hypertriglyceridemia, hyperglycemia, and chronic inflammation.

Moreover, Rychli et al. found that PEDF was significantly associated with CAD (for trend, P = 0.037) and correlated with rehospitalization for heart failure (HF) worsening, with a more prominent risk increase association in CAD patients; these previous findings further support our current results. Rychli et al. also suggested that PEDF was associated with chronic deterioration of cardiomyopathy and played a role in the progression of HF by inducing apoptosis of human cardiac myocytes and fibroblasts [36]. We believe that the elevated PEDF observed in CAD patients occurs in response to vascular injuries, chronic inflammation, and oxidative stress, and that its function involves preventing CAD deterioration.

It has been reported that cardiovascular disease was closely regulated through the signaling pathways of the mammalian target of rapamycin (mTOR) which was associated with endothelial cell survival and growth, as well as cardiomyocyte proliferation [37]. Wang et al. confirmed the antiangiogenic property of PEDF and firstly reported that insulin could down-regulate PEDF expression, which at least partly depended on mTOR kinase as the inhibitory effect of insulin on PEDF expression would disappear when added the mTOR inhibitor rapamycin [38]. It is possible that PEDF would offer exciting prospects for the development of new therapies for cardiovascular disease.

Another physiological factor that was correlated with the elevated PEDF levels in our study population was eGFR (independently negatively correlated). This finding is not surprising since renal filtration is known to affect serum PEDF levels, and is in line with findings from previous studies [15, 31].

When interpreting the results from our present study, it is important to consider some of the features of the study design that may limit the generalizability of our findings. Firstly, the cross-sectional design of the current study precluded our ability to determine the cause-effect relationship for PEDF and CAD. Secondly, the study population was relatively small, and larger prospective studies are necessary to confirm the role of PEDF in the development of CAD. Moreover, the study population was composed of individuals who were suspected of CAD (admitted to hospital to undergo coronary angiography) and most had several of the known risk factors of CAD, which may have biased the results.

Nonetheless, the population of Chinese patients assessed in this study demonstrated a significantly independent correlation of serum PEDF with the presence of CAD. Biologically, PEDF is capable of exerting a number of protective effects under conditions of metabolic disorder, and the mechanism fits with the observation of elevated serum PEDF in patients with MetS and CAD. In particular, elevated serum PEDF levels may represent a counter-regulatory mechanism that acts as a protective response against vascular damage and subsequent CAD. Molecular targeting of PEDF might be a promising therapeutic strategy for treatment of cardiometabolic disorders and represent a useful predictive index for treatment effectiveness. Further laboratory and clinical studies are warranted to confirm the protective effect of PEDF on cardiovascular disease and to assess its potential as a therapeutic target.