In the present study, we assessed the relation between AT quantity, AT distribution and metabolic parameters of AT (dys) function and plasma levels of four CVD-associated EV-markers. Furthermore, the relation between these EV-markers and metabolic syndrome or incident type 2 diabetes in patients with clinically manifest vascular disease was investigated. We show that EV-cystatin C was positively related to metabolic complications of obesity, including low-grade systemic inflammation, low HDL-cholesterol levels and metabolic syndrome. In contrast, EV-CD14 was inversely related to AT abundance and dyslipidaemia, and was moreover related to a relative risk reduction for the development of type 2 diabetes.
Cystatin C has previously been shown to be elevated in obese subjects , to be secreted by AT in vitro and to be associated with the metabolic syndrome . Furthermore, cystatin C has been associated with prediabetes and cardiovascular disease, independent of renal function [18, 38, 39]. In contrast to other studies performed in patients without known cardiovascular disease, we could not demonstrate a relation between obesity and circulating EV-cystatin C-levels, nor were HOMA-IR levels significantly related to EV-cystatin C levels. As we have specifically studied a cohort of patients with clinically manifest vascular disease, a difference in cohort characteristics between these studies may account for discordant results. Nonetheless, we did find strong relations between alternative parameters of adipose tissue dysfunction with EV-cystatin C levels, such as low-grade systemic inflammation and low HDL levels. These findings may suggest that AT dysfunction rather than AT abundance is a more important determinant of EV-cystatin C levels, at least in patients with cardiovascular disease. Furthermore, in concordance with studies performed in healthy individuals , we observed a strong relation between EV-cystatin C and the metabolic syndrome in patients with clinically manifest vascular disease. Thus, the EV-marker cystatin C may be an important biomarker for CVD not only in healthy individuals but importantly also in patients with manifest vascular disease.
Since obesity is associated with a pro-thrombotic state, we hypothesized that obesity could contribute to circulating EV-serpin G1 and EV-serpin F2, both pro-coagulant markers. Serpin G1, better known as C1 inhibitor, is primarily involved in the inhibition of coagulation and atherosclerotic plaque formation , though its role in obesity has not been investigated. Serpin F2, or α-2-antiplasmin, is a major inhibitor of plasmin and thereby controls the coagulation system . Previous studies reported a negative relation of VAT thickness and soluble plasma serpin F2 levels . In our study, obesity was not related to EV-serpin F2 nor to EV-serpin G1, and both markers were not related to metabolic complications of obesity. However, EV-serpin G1 and EV-serpin F2 did show a strong positive relation with low-grade inflammation. These data suggest that in patients with manifest vascular disease, these markers might contribute to low grade inflammation, though both EV-serpin G1 and EV-serpin F2 appear to play no role in the pathophysiology of metabolic complications.
Surprisingly, we observed an inverse relation between EV-CD14 levels with obesity and obesity-related metabolic complications in patients with clinically manifest vascular disease. CD14 is expressed primarily by monocytes, which play important roles in obesity, obesity-induced AT inflammation and insulin resistance [5, 42]. CD14 has furthermore been associated with the development of atherosclerosis and the recurrence of vascular events [18, 43]. EVs secreted by monocytes express CD14, and these EVs are capable of inducing endothelial damage in vitro. However, conflicting results have been reported by others, as soluble CD14 did not relate to endothelial damage in type 2 diabetic subjects , and a recent study showed that lower soluble CD14 levels were associated with an increase in BMI in both obese and non-obese patients . Possibly, two different forms of CD14 studied (soluble CD14 versus membrane-bound CD14) may account for differences in observed relations between obesity and CD14. Although both soluble and membrane bound CD14 are involved in inflammatory signalling pathways, CD14 is only part of a receptor complex and soluble CD14 needs binding to a cellular signal-transducing receptor in order for cell activation to occur . Furthermore, an excess of circulating soluble CD14 is believed to inhibit monocyte responses to inflammatory signals for membrane CD14 [42, 46]. It is unclear whether EVs contain soluble or membrane bound CD14, and whether EV-associated CD14 is functional. Nonetheless, as EV-CD14 levels were associated with low grade inflammation mirrored by circulating hsCRP levels, it is tempting to speculate that high EV-CD14 levels might contribute to vascular risk via inflammatory pathways, but not via obesity-induced metabolic complications.
The mechanisms by which EV-markers could influence the development of type 2 diabetes remain elusive. EVs are regarded as tailor-made messengers for intercellular communication, as their unique composition allows transfer of signalling molecules to a wide variety of target tissues . However, even though a negative relation was observed between EV-CD14 and incident type 2 diabetes in this study, this does not necessarily imply a direct role for CD14 in reduced progression of development of type 2 diabetes. Considering that CD14 is only part of a receptor complex, the functional role of either soluble or membrane CD14 in EVs remains elusive. It could be hypothesized that EVs containing high levels of CV14 are also enriched for insulin sensitizing molecules like adiponectin, as circulating adiponectin levels were positively related to circulating EV-CD14 levels in this study (Additional file 1: Table S2). Further studies are needed to evaluate the pathophysiological role of EV-CD14 in the development of metabolic complications of obesity.
As levels of EV-cystatin C and EV-CD14 can be partly explained by AT abundance and AT (dys) function, we questioned whether AT could actively secrete both EV-markers or whether AT (dysfunction) triggers other tissues for their release. Cell types potentially involved in the release of EV-cystatin C and EV-CD14 include activated monocytes, endothelial cells and platelets, which are all present in high numbers in atherosclerosis lesions [43, 47]. However, evidence suggests that these cell types also play active roles in AT dysfunction, in which hypertrophic adipocytes induce endothelial stress and the recruitment of monocytes . Furthermore, production of both CD14 and cystatin C is increased in AT of obese compared to lean subjects [36, 42]. Therefore, AT itself could be capable of secreting these markers due to adipocyte hypertrophy, hypoxia and increased influx of pro-inflammatory cells such as macrophages. As AT is capable of secreting functional EVs as shown in mice and humans [48, 49], it will be interesting to study whether these AT EVs contain the EV-markers assessed in the present study.
Strengths of this study include the large sample size of a well-characterized and relevant patient population, which allowed for adjustment of multiple relevant potential confounding factors. Furthermore, measurement of different fat compartments with ultrasound allowed for the assessment of the contribution of the different adipose tissues depots to the levels of circulating EV-markers, known to be associated with CVD. Limitations of this study include the fact that, due to the cross-sectional design, causality in the relationships remain unknown. Second, the study population consisted solely of patients with clinically manifest vascular disease, which may limit the generalization of the results to other cohorts.