The prevalence of abdominal obesity and (T2DM) has been raised worldwide Basterra-Gortari et al., 2017Flegal et al., 2010. Obesity associates with substantially increased all cause and cardiovascular (CV) morbidity and mortality, as well as relates to highest risk of type 2 diabetes mellitus (T2DM) Rey-López et al., 2014Vanuzzo et al., 2008. Recent epidemiological investigations, clinical studies and some registers have shown higher prevalence of CV risk factors in obese individuals especially with morbid obesity (body mass index [BMI] more than 40 kg/m⁺) Antwi et al., 2012Finucane et al., 2011Sturm, 2007Valdés et al., 2014. Because progressive increases in yearly pre-diabetes / T2DM prevalence has observed for all classes of BMI irrespective age and sex Sterling et al., 2016, there is reason to expect that abdominal obesity mediates CV risk and risk of T2DM through underlying co-morbidities independently BMI Grundy et al., 2005Grundy et al., 2004.

In this context, obese individuals with similar BMI may be protected or opposite predisposed to obesity-related complications (i.e. T2DM, insulin resistance (IR), dyslipidaemia, hypertension) and CV disease Cuthbertson et al., 2017. The heterogeneity of obesity leads to understanding of being of emerging metabolic phenotypes e.g. metabolically unhealthy obesity (Met-UHO) and metabolically healthy obesity (Met-HO) distinguished from each other for CV risk Kim et al., 2015. Moreover, it is suggesting that Met-HO is a transient state in the pathway to cardiometabolic disease, i.e. Met-UHO and T2DM Mongraw-Chaffin et al., 2016.

Based on the Adult Treatment Panel-III criteria subjects with established obesity and co-existing other metabolic abnormalities including dyslipidemia, insulin resistance (IR), increased fasting glucose and impaired glucose tolerance, are referred Met-UHO, whereas obese individuals without these abnormalities might be defined as Met-HO Grundy et al., 2005Ryden et al., 2013. The mechanisms underlying the change in phenotype from metabolically healthy to metabolically unhealthy obesity are still unclear Ryden et al., 2013.

Micro particles (MPs) are defined a heterogeneous sub-population of extracellular vesicles with diameter average from 100 to1000 nm originated from plasma membranes of mother’ cells Berezin et al., 2015a. As a derivate of cellular membrane MPs are discussed powerful paracrine regulators of target cell structure and functions. MP released by apoptotic endothelial cells posse a wide spectrum of biological effects on intercellular communication by transferring different active molecules (proteins, peptides, hormones, growth factors, microRNAs) exhibiting coagulation activity, mediating cell growth and tissue differentiation Alexandru et al., 2016. Additionally, apoptotic endothelial cell-derived MPs (EMPs) may directly worse endothelial integrity and vascular function playing a pivotal role in development of microvascular inflammation and IR Alexandru et al., 2016Berezin et al., 2016b.

Recent clinical studies have shown that the circulating levels of apoptotic EMPs were significantly increased in T2DM patients as compared with healthy volunteers Berezin et al., 2016a and they mediated CV risk in patients with established metabolic syndrome (MetS) and T2DM Agouni et al., 2014Berezin et al., 2015bBerezin et al., 2016c. Whether apoptotic EMPs are involved in the transformation of Met-HO into Met-UHO determining the risk of T2DM and CV disease is not fully clear. The aim of the study: to investigate whether increased number of circulating apoptotic EMPs may predict transformation of Met-HO into Met-UHO.

The study was retrospectively evolved 89 patients with established abdominal obesity (47 patients with Met-UHO determined as MetS and 42 subjects with Met-HO) from the large cohort of abdominal obesity patients (n=268) who were examined between February 2012 and July 2016. We have enrolled obese subjects (body mass index was more 30 kg/m⁺) without known CV disease including angina pectoris, previous myocardial infarction / stroke, heart failure, and asymptomatic atherosclerosis (defined by negative result of the contrast-enhanced multi-spiral tomography angiography). Thirty five healthy volunteers matched for age and sex were involved in the study as a control cohort. All patients have given their informed written consent for a participation in the study.

MetS was diagnosed based on the National Cholesterol Education Program Adult Treatment Panel III criteria Williams, 2002. Patients were enrolled in the MetS cohort when at least three of the following components were defined: waist circumference ≥90 cm or ≥80 cm in men and women respectively; high density lipoprotein (HDL) cholesterol <1.03 mmol/L or <1.3 mmol/L in men and women respectively; triglycerides ≥1.7 mmol/L; blood pressure ≥130/85 mmHg or current exposure of antihypertensive drugs; fasting plasma glucose ≥5.6 mmol/L. Participants who had less than 2 criteria of MetS were classified as Met-HO patients. Those who had 2 or more criteria of MetS were classified as metabolically abnormal Williams, 2002 and were not considered candidates for this study. Participants with abdominal obesity who had less than 3 criteria of MetS were classified those who had Met-HO. Therefore, individuals with non-alcoholic fatty liver disease, polycystic ovary syndrome, and those who had higher levels of HBV / HCV antibodies, were not enrolled in the study. The flow chart with inclusion / exclusion criteria is reported in Figure 1 .

The demographic and anthropometric characteristics, the prevalence of CV risk factors are reported in Table 1 . There was not a significant difference between healthy volunteers and entire cohort, as well as between subjects with metabolically unhealthy obesity (Met-UHO) and metabolically healthy obesity (Met-HO) in age, sex, adherence to smoke and haemodynamic performances. Abdominal obesity subjects exhibited higher BMI, WHR, HOMA-IR, as well as they had increased frequency of LVH, hypertension and IR presentation. Met-UHO patients had higher HOMA-IR, than those with Met-HO, while BMI, systolic and diastolic blood pressure, heart rate were similar in both cohorts. Additionally, hypertension, IR and dyslipidemia presentation were found frequently in Met-UHO patients compared to Met-HO individuals.

Healthy volunteers and abdominal obesity individuals from entire cohort had similar level of GFR, hemoglobin, and creatinine. Consequently, abdominal obesity individuals had higher level of fasting glucose, HbA1c, total cholesterol, LDL cholesterol, triglycerides, hs-CRP, vistafin, leptin and lower adiponectin ( Table 2 ). No difference was seen in GFR, hemoglobin, fasting glucose, creatinine, and lipids’ level between Met-UHO and Met-HO patients. However, Met-UHO patients had higher HbA1c, vistafin, leptin and lower hs-CRP than those with Met-HO.

Healthy volunteers had significantly decreased level of EMPs labeled CD31⁺/Annexin V⁺ and CD144⁺/ Annexin V⁺ than those with abdominal obesity ( Figure 2 ). However, there was not found a significant difference between numbers of EMPs labeled CD31⁺/Annexin V⁺ in Met-UHO and Met-HO patients. In contrast, Met-UHO patients had a significantly increased level of circulating CD144⁺/ Annexin V⁺ compared with Met-HO individuals.

The univariate linear regression analysis between numerous of EPMs with immune phenotypes determined CD31⁺/ Annexin V⁺ and CD144⁺/ Annexin V⁺, CV risk factors, hemodynamic performances, and other biomarkers was performed. In Met-UHO patients the number of CD31⁺/ Annexin V⁺ EMPs received from peripheral blood positively related to HOMA-IR (r =0.35, P = 0.003), hs-CRP (r = 0.33, P = 0.001), number of CV risk factors (r = 0.32, P = 0.001), BMI (r = 0.31, P = 0.001), serum leptin (r = 0.31, P < 0.001), fasting glucose (r = 0.30, P < 0.001), serum vistafin (r = 0.29, P < 0.001), LDL cholesterol (r = 0.27, P = 0.003), but inversely associated with serum adiponectin (r = -0.31, P < 0.001). In contrast, in Met-HO individuals HOMA-IR (r =0.32, P = 0.001) and number of CV risk factors (r =0.31, P = 0.001) significantly related to number of CD31⁺/ Annexin V⁺ EMPs. There were not sufficient relations between number of CD31⁺/ Annexin V⁺ EMPs and biomarkers of obesity, i.e. leptin and vistafin, whereas between adiponectin and number of CD31⁺/ Annexin V⁺ EMPs an inversely weak association was found (r = -0.25, P = 0.01). Table 2 . The biomarkers of the patients enrolled in the study. Therefore, the number of CD144⁺/ Annexin V⁺ EMPs inversely related to a level of serum adiponectin (r = -0.28, P = 0.001) and positively associated with HOMA-IR (r =0.36, P = 0.001), hs-CRP (r = 0.31, P = 0.001), serum vistafin (r = 0.33, P = 0.001), number of CV risk factors (r = 0.32, P = 0.001), serum leptin (r = 0.32, P = 0.001), fasting glucose (r = 0.29, P = 0.012) in Met-UHO patients. In contrast, there was not a significant association between number of CD144⁺/Annexin V⁺ EMPs and metabolic biomarkers of obesity, such as adiponectin, leptin and vistafin in Met-HO individuals, whereas they were related to HOMA-IR (r =0.36, P = 0.001), BMI (r = 0.32, P = 0.001) and hs-CRP (r = 0.30, P = 0.001).

Multivariate unadjusted linear regression analysis has shown that in Met-UHO patients the numbers of CD31⁺/ Annexin V⁺ and CD144⁺/ Annexin V⁺ EMPs related to HOMA-IR (r = 0.32, P = 0.001 and r =0.33, P = 0.001), hs-CRP (r = 0.30, P = 0.001 and r =0.31, P = 0.001), number of CV risk factors (r =0.29, P = 0.001 and r =0.30, P = 0.001), serum leptin (r = 0.30, P = 0.001 and r = 0.31, P = 0.001), serum vistafin (r = 0.28, P = 0.001 and r = 0.30, P = 0.001), serum adiponectin (r = -0.26, P = 0.001 and r = -0.29, P = 0.001) and LDL cholesterol (r = 0.27, P = 0.001 and r =0.28, P = 0.003) respectively. In Met-HO individuals the numbers of CD31⁺/ Annexin V+ and CD144⁺/ Annexin V+ EMPs related significantly to HOMA-IR (r =0.32, P = 0.001 and r =0.36; P<0.001), BMI (r = 0.27, P = 0.001 and r =0.31; P<0.001) and hs-CRP (r = 0.26, P = 0.001 and r = 0.30, P = 0.001) respectively.

After adjustment for BMI HOMA-IR remained the most profound factor related to numbers of CD31⁺/ Annexin V⁺ and CD144⁺/ Annexin V⁺ EMPs (r = 0.33, P = 0.001 and r =0.36, P = 0.001) in Met-UHO. We also determined a relation between numbers of CD31⁺/ Annexin V⁺ and CD144⁺/ Annexin V⁺ EMPs and number of CV risk factors (r =0.27, P = 0.001 and r =0.30, P = 0.001), serum leptin (r = 0.27, P = 0.001 and r = 0.29, P = 0.001), serum vistafin (r = 0.28, P = 0.001 and r = 0.30, P = 0.001). Therefore, mild association of CD31⁺/ Annexin V⁺ and CD144⁺/ Annexin V⁺ EMPs with HOMA-IR (r = 0.34, P = 0.001 and r =0.36, P = 0.001) was found in Met-HO patients. However, the multivariate adjusted for BMI linear regression analysis has shown that in Met-UHO numbers of CD31⁺/Annexin V⁺ and CD144⁺/ Annexin V⁺ EMPs associated with serum hs-CRP (r = 0.27, P = 0.001 and r = 0.31, P = 0.001) respectively. In multivariate logistic regression analysis we found that HOMA-IR, number of CV risk factors, serum leptin and hs-CRP were independent predictors for increased numbers of circulating CD31⁺/ Annexin V⁺ and CD144⁺/ Annexin V⁺ EMPs in Met-UHO ( Table 3 ). In Met-HO patients HOMA-IR remained an independent predictor of increased numbers of circulating CD31⁺/ Annexin V⁺ and CD144⁺/ Annexin V⁺ EMPs.

Statistics for model fit for the prediction of Met-UHO development is reported in Table 4 . One can see increased number of CD31⁺/ Annexin V⁺ and CD144⁺/Annexin V⁺ EMPs added to the based predictive model (Met-HO + HOMA-IR) may sufficiently improve prognostication of based model regarding development of Met-UHO.

In this study we reported that increased number of CD31⁺/ Annexin V⁺ and CD144⁺/ Annexin V⁺ EMPs much more pretty accurate predict Met-UHO than based model (Met-HO). Taking into consideration that Met-UHO is considered a powerful risk factor of type 2 diabetes mellitus, we suggest that increased number of CD31⁺/ Annexin V⁺ and CD144⁺/ Annexin V⁺ EMPs related to HOMA-IR in both Met-UHO and Met-HO may predict probably obesity-related complications including higher risk of t2dm. Thus, early detection of abnormality in circulating levels of apoptotic EPCs may be a biomarker of IR and predictor of Met-UHO in patients with abdominal obesity when other metabolic disturbances are absent or evidence regarding them is fare limited.

Recently it has been suggested that Met-HO is a stage of development of Met-UHO and T2DM Mongraw-Chaffin et al., 2016. However, our results revealed that the most important factor that affects metabolic dysregulation in obesity is IR, which probably appears to be predominantly early stage of the Met-HO. There is evidence that an accumulation of visceral adiposity tissue (VAT) might associate with over-production of pro-inflammatory cytokines including hs-CRP, leptin and vistafin and induce IR Brede et al., 2016. Therefore, infiltration of the sub-intima by LDL cholesterol may induce production of free radicals, oxidation of cytoskeleton and membrane vesiculation of endothelial cells. Finally, membrane vesiculation of endothelial cells is enhanced by inflammatory cytokines in conveying of VAT accumulation Mause and Weber, 2010. Interestingly, the circulating number of apoptotic EMPs has well associated with conventionally obesity biomarkers (adiponectin, leptin, vistafin) in Met0UHO patients, but did not in Met-HO individuals. Indeed, in Met-HO patients we did not find severe metabolic abnormalities apart from leptin elevation compared with Met-UHO, however, IR was determined as common finding for both Met-UHO and Met-HO individuals without a difference in BMI.

The increased amount of VAT together with a chronic inflammation and IR predisposes to the development of endothelial dysfunction through attenuation synthesis and secretion of apoptotic EMPs. Indeed, pro-inflammatory cytokines, i.e. interleukin-6, tumor factor necrosis-alpha, leptin, and vistafin, may directly influence structure of endothelial cells and trigger a secretion of apoptotic EMPs Berezin, 2016bRautou et al., 2011. The main biological function of this process is attenuation of endothelial cell repair and recovery of vascular function Jansen et al., 2015. Unfortunately, co-existing IR affects endothelial progenitor cells and they are not able to differentiate into functionally mature endothelial cells even after stimulation by apoptotic EMPs Martinez and Andriantsitohaina, 2011Tetta et al., 2011. As a result, apoptotic EMP-induced endothelial dysfunction and IR may become an early predictor of shaping Met-UHO.

Recently we have reported that apoptotic EMPs may independent predict asymptomatic atherosclerosis and CV disease in T2DM patients Berezin, 2016aBerezin et al., 2016a, while their role in individuals with different phenotypes of obesity has remained controversial Montoro-García et al., 2011. First, it is not clear whether increased number of apoptotic EMPs is adaptive mechanism of vascular repair or factor of endothelial injury. Indeed, circulating EPMs, which are enhanced in a large number of metabolic disorders including abdominal obesity, associated with IR and this has been linked to deleterious effects on endothelial cells Martinez and Andriantsitohaina, 2011. At the same time, apoptotic EPMs are powerful factor contributing in endothelial progenitor cell mobbing and differentiation Montoro-García et al., 2011. Secondary, it is not fully understand the innate molecular mechanisms, which correspond to triggers of secretion of these apoptotic MPs. Apoptotic MPs as cargo microvesicles consist of a variety of biomolecules including regulated proteins, DNA, mRNA, and non-coding RNA. The proportion of these components as well as an entire secretome is under a tight control of autocrine / paracrine mechanisms and inflammatory factors (i.e. tumor necrosis factor-alpha, interleukin-2, -6), which induces EMP formation in a time-dependent manner Lee et al., 2014. Consequently, the final reply of the recipient cells, such as endothelial progenitor cells, is depends on epigenetic regulation of secretome secretion and primary trigger, which affects vesiculation Berezin, 2016bBerezin et al., 2015c. Obviously, an ability of apoptotic EMPs to modulate immune and inflammatory processes, coagulation and vascular function, angiogenesis and vascular injury may interact with other regulatory mechanisms the role of witch in the pathogenesis of abdominal obesity requires still being determined. It is no excluded that release of apoptotic EMPs might act as a direct endogenous survival signal for target cells Lichtenauer et al., 2015.

The present results exhibited first the interrelationship between increased number of circulating apoptotic EMPs and IR in patient with Met-HO. Extrapolating these findings on entire obese population, we can suggest that a clinical diagnosis of abdominal obesity irrespective to its phenotypes (Met-UHO or Met-HO) is probably not sufficient to assess a risk of T2DM and CV disease Berezin et al., 2015c. In this context, measurement of circulating apoptotic EMP number would be useful tool for stratification amongst obesity individuals at higher risk of T2DM and CV Berezin, 2016a, bMontoro-García et al., 2011, especially when conventional biomarkers of obesity are not detected in appropriate diagnostic level. Large investigations are required to understand the role of apoptotic EMPs in pathogenesis of different phenotypes of abdominal obesity, because they may be a target of the therapy as well as predictive biomarkers.

This study has some limitations. The first limitation is low number of the patient involved in the investigation. Another limitation is lack of standardization of MP measurements, while commercial flow cytometers are existed. It is necessary to note that a large pool of MPs might be produced after blood sampling due to destruction of platelets and blood cells. In this study we used platelet free plasma to prevent of contamination of samples with MPs originated from platelets. Therefore, preparation of MP isolates from samples is the most sophisticated step for further examination. The next limitation might relate to complicated assay and suffers from resolution of MP detection technique that is worth considering. Indeed, there were several technical-related difficulties in the measurement of MPs affected centrifugation of samples, labeling of MPs, using HD-FACS methodology and final assay of results obtained. Although HD-FACS methodology is widely used, theoretically overlap between two or more fluorochromes might reflect some obstacles for further interpretation of obtained results, especially including size gating in MP determination. Therefore, rotor type and centrifugation time theoretically may influence on purity of MPs. However, flow cytometery and HD-FACS methodology are commonly used procedure to determine and measure MPs.

Another limitation of the present study is that a specific role of MPs is also possible and has not been characterized. However, the authors suppose that these optionally technically restrictions might have no significant impact on the study data interpretation. Additionally, retrospective, relative small sample size may limit the significance of the present study.

In this investigation we first determined the increased apoptotic EMP number may predict transformation of Met-HO into Met-UHO. The evidence allow suggesting the apoptotic EMPs might serve as a biomarker higher risk of T2DM and CV disease. Future investigations are needed to confirm these suggestions and clear situation around discriminative value of apoptotic EMP number in abdominal obesity individuals.

AUC: area under the curve; BMI: body mass index; CI: confidence interval; CV: cardiovascular: GFR: glomerular filtration rate; EMPs: endothelial cell-derived mocroparticles; HDL-C: high-density lipoprotein cholesterol; hs-CRP: high sensitive C-reactive protein; IDI: integrated discrimination index; LDL-C: low-density lipoprotein cholesterol; Met-HO: metabolically healthy obesity; MetS: metabolic syndrome; Met-UHO: metabolically unhealthy obesity; MPs: microparticles; NCEP: National Cholesterol Education Program; NRI: net reclassification index; OR: odds ratio.

AEB gave the concept, designed experiment, analyzed data and gave final approval of the manuscript and was the supervisor of the project. AAK performed experiment, analyzed data and wrote draft of manuscript. TAB helped in data acquisition and experiment conditions optimization. YVM and TAS helped in editing and literature review for the paper TAS did data acquisition, organization and data analysis and helped in write up of final draft.