The pathologic hallmark of type 2 diabetes mellitus (T2DM) involves the vasculature leading to both microvascular and macrovascular complications (1, 2). Hyperglycemia is associated with long-term damage and failure of various organs mainly affecting the eyes, nerves, kidneys, and heart (1). Patients with T2DM have a markedly increased risk of developing cardiovascular disease (CVD); and reducing levels of LDL-C using statin therapy only partially diminishes the increased propensity of these patients towards CVD events (3). African Americans suffer disproportionately from many chronic diseases including T2DM (4). Prevailing statistics suggest that African American adults are 50% to 100% more likely to have diabetes than are Whites (5). Therefore, the strategies in place to prevent disease development and progression warrant further examination.

The abnormal lipid profile that often accompanies T2DM is well-known to be associated with an increased risk of atherosclerotic vascular disease (6). This disorder is typically characterized by elevated serum triglycerides, low HDL-c concentrations and increased smaller, denser LDL-c particles (7). Small dense LDL (sdLDL) particles are believed to be particularly atherogenic due to increased susceptibility to oxidation, high endothelial permeability, decreased LDL receptor affinity, and an increased interaction with matrix components. sdLDL penetrates the vascular endothelium more easily and accelerates the atherosclerotic process (8). Also, there is evidence that some crucial functions of HDL, such as endothelial protection and repair, are adversely modified by diabetes, most likely through alteration of specific components of HDL particles (9). It has been shown that HDL from patients with T2D have a significant reduction in their antioxidative and endothelium-dependent vasorelaxant properties (7). HDL isolated from patients with T2DM have an impaired ability to stimulate nitric oxide synthase activity (10) and nitric oxide (NO) production from endothelial cells compared to HDL from healthy controls (11).

Circulating HDL and LDL particles are subjected to multiple enzymatic and non-enzymatic modifications including glycation, oxidation, nitration and carbamylation (12, 13). Plasma lipoproteins may represent a perfect biological sensor for oxidative stress in the arterial wall by their close interactions with endothelial cells in the vasculature and the susceptibility of their surface lipids to oxidative modification (14). ApoA-I and apoB have been shown to be specifically targeted for nitration by myeloperoxidase (MPO) leading to dysfunction of circulating HDL and LDL particles (15, 16). Moreover, MPO levels has been shown to predict accelerated progression of coronary atherosclerosis in diabetic patients (17). The pathways leading to the nitration of tyrosine residues in apoAI-HDL and apoB-LDL have been extensively studied (18, 19). High levels of nitrated tyrosine residues on apoA-I have been found in human plasma and atherosclerotic lesions from CAD patients, resulting in profound impairment of HDL particles and ABCA1-dependent cholesterol efflux (20, 21).

The endothelium is continuously exposed to various physiological molecules that may have a direct impact on NO actions. The damaging effects of oxidative stress on the cardiovascular system determine endothelial dysfunction through reduction in NO synthesis and bioavailability, inflammatory response, and lipid peroxidation (22). Interestingly, NO can play a dual role in atherosclerosis; and some of the pathological vascular conditions were linked with excessive rather than reduced NO production (23). Altered NO levels in patients with T2D have been reported by different studies (24); however, results were controversial.

To our knowledge, there is no study reporting the direct link between plasma nitrated lipoproteins and endothelial dysfunction in diabetic microvascular and macrovascular complications. In this study, we aimed to investigate the relationship between circulating nitrated lipoproteins levels and subclinical vascular changes in African Americans diabetic patients.

The present study aimed to test the hypothesis that plasma nitrated lipoproteins levels may be positively related to vascular dysfunction in patients with T2DM. To our knowledge, this is the first study addressing the impact of nitrated lipoproteins (both NT-HDL and NT-LDL) on vascular dysfunction in T2DM patients. Here, we found that the concentrations of plasma NT-LDL were much lower than NT-HDL (~ 10 times less). This is in accordance with our previously published reports and others (16, 19, 29) showing that apoA-I containing HDL particles is indeed the most preferred target for MPO- mediated nitration (15). Moreover, despite the differences in the degree of nitration on respective apolipoproteins ApoA-I and apoB in these particles, our study revealed a strong positive correlation between NT- HDL and NT-LDL levels. Regression analysis showed that the association between NT-HDL and NT-LDL was partially attributable to age factor that was found a confounding variable in our base model. In fact, age dependent proteins nitration has been previously reported in many diseases (30). It is also known that nitration of tyrosine residues of apoB-LDL makes LDL a very potent ligand for macrophages scavenger receptors leading to cholesterol accumulation and foam cell formation (31). This process has been well recognized as a major key event in the initiation and progression of atherogenesis. Based on the structural and functional differences between HDL and LDL particles, we hypothesized that circulating NT-LDL and NT-HDL levels may differentially influence vascular function in T2DM diabetic patients.

Strict glycemic control has been proposed as an important factor to lower the risk of both microvascular and macrovascular complications of T2DM. We assessed the relationship between glycemic control (HbA1c levels) and nitrated lipoproteins and their impact on vascular function. Unexpectedly, we found an inverse correlation between HbA1c levels and nitrated lipoproteins in total population and in poorly-controlled patients with HbA1c > 7%. Furthermore, HbA1c levels correlated positively with VRI outcome in total population. HbA1c levels were also positively correlated with NO levels in well-controlled patients with HbA1c ≤ 7%. A possible explanation for these paradoxical findings could be related to intra-individual variability in HbA1c measurements among participants and/or effect of other possible contributing factors that could influence NO bioavailability in diabetes. In fact, hypertension and dyslipidemia were shown to be important modifiable risk factors contributing to the pathogenesis of cardiovascular events in T2DM (32). Also, it is not possible to exclude possible effect of other residual confounding factors for known and unknown risks such as fasting blood glucose or inflammatory markers, which were not reported in the current study. Several paradoxical relationships between glycemic control and clinical outcomes has been reported. Unexpected inverse relationship has been reported between HbA1c levels and mortality in patients with diabetes and advanced systolic heart failure (33). Positive linear relationship has been also reported between HbA1c and all-cause mortality (34). Moreover, both low normal HbA1c level and high HbA1c level (J- or U-shaped curve) have been associated as well with increased risk of all-cause mortality (35). Most of all these effects were mainly attributable to known episode of severe hypoglycemia and/or the type of medication in T2DM. In the present study, we couldn’t obtain information about frequency of hypoglycemia events from our participants. Poor glycemic control is related to vascular events in patients with T2DM, but the presence of advanced vascular disease might also influence associations between variables. HbA1c level measurements were not repeated in this current study; nevertheless, to avoid any possible bias with recording inconsistent HbA1c values, we have used an average of HbA1c values of the most recent measurements recorded within 6-12 months prior patients vascular testing.

It is well known that uncontrolled glycemia leads to impairment of NO production which may result in accelerated vascular complications in diabetic patients. NO has been shown to be involved in the defective insulin-mediated stimulation of blood flow in T2DM as well as in the pathogenesis of diabetic nephropathy (24). We found that both NT- HDL and NT-LDL negatively correlated with total NO levels, especially in patients with poor glycemic control, suggesting a possible interplay between the role of glucose homeostasis in the production of NO and circulating lipoproteins. In fact, it is well known that modified LDL (e.g. oxidized LDL) and NO exert contradictory actions within the vascular endothelium microenvironment which may influence key events in atherogenesis (36). HDL is thought to exert at least some parts of its antiatherogenic properties via stimulation of endothelial NO production, nearby inhibiting oxidative stress and inflammation (37). Based on available data linking diabetes to endothelial dysfunction, and the fact that HDL can protect endothelium by stimulating production of NO, it is possible that NT-HDL particles might not be fully functional and capable of modulating NO availability. Controversies exist concerning the effect of insulin on nitric oxide synthase (NOS) activity and NO generation. Studies with cultured endothelial cells suggest that insulin stimulates NO formation, whereas elevated glucose levels inhibit NO synthesis (38).

Several studies reported the existence of free-nitrotyrosine and nitrated lipoproteins enrichment in human atherosclerotic lesions using immunochemical methods and mass spectrometry (39). We have previously shown that higher NT-apoAI/apoAI ratio was significantly elevated in diabetic patients with CAD, and this ratio was negatively correlated with apoAI-HDL dependent macrophages cholesterol efflux (29). The present study validates and provides a rational evidence to our previous observations using an in vitro macrophages cholesterol efflux assay to assess nitrated apoAI-HDL functionality. Thus, we believe the actual report is the first physiological evidence addressing the relationship between circulating plasma nitrated lipoproteins and vascular dysfunction in diabetic patients. Nevertheless, additional investigations are warrant in order to highlight exact mechanisms underlying the observed differential effects of circulating NT-HDL and NT-LDL on vascular dysfunction in T2DM.

Diabetic micro- and macro vascular complications seem to be strongly interconnected with microvascular diseases promoting atherosclerosis development. Currently, it is still not well understood whether microvascular complications distinctly precede macrovascular complications, or both could progress simultaneously. In the current study, we found that NT-HDL independently predict VRI outcome but not PWV nor CIMT, suggesting that NT-HDL may play a role in early stages of diabetic microvascular complications. The pathology of macrovascular disease includes large vessels occlusion and transmural inflammation of the vessel wall leading to tissue damage (40). Traditional Framingham risk factors may influence the determination of vascular function (41). The assessment of CIMT has been well accepted as an indicator of atherosclerosis. Our study showed that, in contrast to NT-HDL, NT-LDL was negatively associated with CIMT. This unexpected finding is somewhat intriguing. One plausible explanation is that similar to oxidized and glycated LDLs, NT-LDL may be able to bind to other receptors than LDL receptors (LDLR), and start to be taken up by arterial cells, accumulate in the arterial intima, and display its proatherogenic effect on vascular cells for prolonged period of time. Therefore, this process may result in net depletion of circulating plasma NT-LDL levels due to faster clearance. In fact, Hamilton et al., using in vitro binding experiments, demonstrated that both native LDL and peroxinitrite-treated LDL bind to LDLR, and the uptake of NT-LDL was stronger than native LDL, suggesting an alternate mechanism for cellular uptake involving macrophages LOX-1, CD36, and SR-A receptors (42). Faster clearance of modified LDL has been well demonstrated (43). Similarly, LDL nitration has been shown to be associated with enhanced monocyte/macrophage uptake when compared with oxidized LDL isolated from rheumatoid and osteoarthritis patients with CVD (44). Furthermore, a significantly higher amount of nitrotyrosine residues containing lipoproteins has been found in atherosclerotic site than in plasma (45). Modified LDL isolated from aortic atherosclerotic intima had 90-fold higher levels of 3-nitrotyrosine as compared to healthy subjects (46). Taken together, our findings are in good agreement with our previously published work and others (12, 15, 16, 19, 45).

Unlike NT-HDL, the association between NT-LDL and CIMT outcome becomes insignificant after inclusion of other independent variables into the base model. Potential confounding variables were identified, and sex was the most significant contributing factor in the CIMT variance, especially in the poorly-controlled patients’ group. The explanation for sex difference in this report is unknown. This could be linked in part to hormonal differences between women and men. In fact, it is well known that estrogen upregulates endothelial NOS which may result in changes of NO production and association between NT-LDL and CIMT (47). Another plausible explanation is that CIMT may represent adaptive changes to biomechanical parameters such as media hypertrophy, aging and others, and therefore this parameter may not be a direct indicator of vascular atherosclerotic changes. The lack of association between NT-LDL and CIMT outcome suggest that plasma NT-LDL may not be considered a direct powerful measure of vascular atherosclerotic process in diabetic patients.

Use of drugs such as statins, beta-blockers and ACE inhibitors has been shown to reduce MPO levels in patients with acute coronary syndrome, but not in patients with stable CAD (48). In this study, we did not measure plasma MPO levels; however our study revealed that use of calcium channel blockers and ACE inhibitors significantly affected the association between NT- HDL levels and VRI outcome most likely by suppression of tyrosine nitration of proteins as shown in diabetes and chronic renal diseases (49, 50).

There are some limitations to be considered in the present study. First, the relatively small sample size from a single center may have limited the power to detect weak correlations among different stratified groups; however, the sample was sufficient to demonstrate significance among the groups. Second, the ELISA allowed us to measure total NT-LDL and NT-HDL particles; therefore, we were unable to distinguish between various nitrated subclasses of lipoproteins. In fact, most of published reports showed either circulating free nitrotyrosine or total nitrated apolipoproteins-containing lipoproteins by ELISAs and mass spectrophotometry techniques. Third, we did not exclude patients with prior lipid lowering medication which may have biased the prognostic value of nitrated lipoproteins levels. Our study did not, however, show any significant effect of known lipid modulating drugs such as statins. Lastly, the study population represents a relatively heterogeneous multi-ethnic community-based T2D cohort which may have contributed to individual intra-variability measurements. We are aware of differences in race and study population selection could account for discrepancies among studies; and therefore, this may limit clarification of causality between plasma NT-HDL and VRI outcome. This cross-sectional study provides strong evidence for differential effects between circulating NT-HDL and NT-LDL in the impairment of microvascular dysfunction in T2DM. Therefore, caution should be taken when interpreting data from quantitative measurements of nitrated lipoproteins to assess vascular disease state.

In conclusion, our study demonstrated that plasma levels of NT-HDL, but not NT-LDL, independently predicted VRI outcome in diabetic patients especially in well-controlled patients with HbA1c ≤ 7%. In addition, total NO availability does not seem to significantly influence the association between NT-HDL and VRI, suggesting that circulating NT-HDL action on the vasculature may occur via NO-independent pathway. Additionally, our data demonstrated that plasma NT-HDL is more reliable biomarker than NT-LDL and could be used as a predictor of early microvascular function in diabetes. This finding may have important clinical implication despite unknown exact mechanisms. Therefore, our study is worthy of further investigations and must be confirmed in larger cohorts.

This research was funded by The New York State Department of Health Empire Clinical Research Investigator Program award (ECRIP, 2015-2016; PI: Dr. Moro Salifu) to the State University of New York Downstate Medical Center, University Hospital of Brooklyn. All authors have read and agree to the published version of the manuscript. Conceptualization, Paul Dreizen and Moro Salifu; Data curation, Ahmed Bakillah; Formal analysis, Carl Rosenberg and Ahmed Bakillah; Funding acquisition, Moro Salifu; Investigation, Ajibola Adedayo, Ayobami Eluwole, Fasika Tedla, Arye Kremer, Muhammad Khan, Louis Salciccioli and Jason Lazar and Ahmed Bakillah; Methodology, Ayobami Ayobami Eluwole, Ajibola Adedayo, Paul Dreizen, Moro Salifu, Jason Lazar and Ahmed Bakillah; Project administration, Nicole Mastrogiovanni; Resources, Moro Salifu; Software, Moro Salifu and Jason Lazar; Supervision, Clinton Brown, Jason Lazar and Ahmed Bakillah; Validation, Ahmed Bakillah; Visualization, Ahmed Bakillah; Writing – original draft, Ahmed Bakillah; Writing – review & editing, Ajibola Adedayo, Fasika Tedla, John La Rosa, Mohamed Boutjdir, Mary Ann Banerji, Clinton Brown, Moro Salifu, Jason Lazar and Ahmed Bakillah. We thank Mrs. Lorraine Thomas for the excellent technical assistance and all study participants from SUNY Downstate and Kings County Clinics. Part of this work has been presented at the American Diabetes Association 78th Scientific Sessions (ADA18). Drs. Ajibola Adedayo, Ayobami Eluwole, Fasika Tedla and Arye Kremer were clinical fellows under ECRIP program.