Effects of acute insulin‐induced hypoglycaemia on endothelial microparticles in adults with and without type 2 diabetes

To assess whether endothelial microparticles (EMPs), novel surrogate markers of endothelial injury and dysfunction, are differentially produced in response to acute insulin‐induced hypoglycaemia in adults with and without type 2 diabetes.

reach up to 31% following ≥10 years of insulin therapy initiation. [7][8][9] The frequency of asymptomatic mild hypoglycaemia is also high; approximately one in two individuals with type 2 diabetes experience at least one event over a 3-day period. 10 Hypoglycaemia induces stress responses, which include the sympatho-adrenal activation and the release of glucagon, epinephrine, cortisol and growth hormone. 11 Haemodynamic alterations occur to maintain glucose supply to the brain and promote glucose generation from the liver; these alterations include increases in heart rate, systolic blood pressure, myocardial contractility and cardiac output. 11 Blood viscosity increases, leading to an elevation in platelet count, aggregation and coagulation. 11 At a molecular level, hypoglycaemia causes increased markers of inflammation, leukocytosis, lipid peroxidation, oxidative stress and platelet-monocyte aggregation. [12][13][14][15] These hypoglycaemia-induced changes may result in the generation of biomarkers that are able to identify a hypoglycaemic event after blood glucose levels have reversed to normal.
Endothelial microparticles (EMPs) are surrogate markers of endothelial injury and dysfunction released by activated or apoptotic endothelial cells. 16,17 Microparticles (MPs) are key regulators of cell to cell interactions and, by carrying specific membrane antigens from their source cells, they act as diffusible vectors in the transcellular exchange of biological information. 17 EMPs play an important role in maintaining vascular homeostasis, and elevated EMP levels are implicated in the pathogenesis of vascular diseases, cancer, inflammatory, endocrine and metabolic disorders. [17][18][19] Several studies have reported increased EMPs in individuals with diabetes mellitus, as compared to controls without diabetes, [20][21][22] and have explored EMPs as biomarkers of vascular injury, and as potential predictors of cardiovascular outcomes in patients with or without diabetes mellitus. [23][24][25][26] Given that EMPs are produced at the initial stages of cell injury or as part of membrane remodelling, these markers may be also useful in characterizing endothelial responses to hypoglycaemia; however, their potential as a biomarker in this condition has not been previously investigated in type 2 diabetes.
The aim of the present study was to explore the effects of acute insulin-induced hypoglycaemia on EMPs in adults with and without type 2 diabetes.

A prospective parallel study was performed in the Diabetes Research
Centre at Hull Royal Infirmary in adults with type 2 diabetes (n = 25) and controls without diabetes (n = 25). All participants provided their written informed consent before taking part. Weighing Machine Group Ltd, Rotherham, UK) and height was taken barefoot using a wall-mounted stadiometer. Blood pressure was measured using a sphygmomanometer (Datascope Duo Masimo Set; Mindray Ltd, Huntingdon, UK) and a blood sample was collected in the fasted state before insulin infusion and used as baseline. Continuous insulin infusion was performed to induce hypoglycaemia. Blood samples were taken at 0, 30, 60, 120 and 240 minutes after hypoglycaemia. After 240 minutes the participants were provided with lunch and were allowed their (morning) diabetes medications. The participants took their evening medication as prescribed that day. For Visit 3 (24 hours from the induction of hypoglycaemia), patients were also allowed to take their medication, once they completed the blood tests in the fasted state, after which breakfast was provided. Prior to discharge, blood glucose was checked using a glucose analyser (HemoCue glucose 201+) to ensure normal levels, together with other vital signs.

| Insulin infusion
After an overnight fast, bilateral ante-cubital fossa indwelling cannulae were inserted 30 to 60 minutes prior to the commencement of the test  27 The blood sample schedule was timed subsequently with respect to the time point when hypoglycaemia occurred. Following the identification of hypoglycaemia, intravenous glucose was given in the form of 150 mL of 10% dextrose and a repeat blood glucose check was performed after 5 minutes if blood glucose was still <4.0 mmol/L. All patients achieved a blood glucose of ≤2.0 mmol/L (36 mg/dL), although the median duration to severe hypoglycaemia was significantly greater in participants with type 2 diabetes compared with controls (54 vs. 30 minutes; Supporting Information Table S1); however, the duration of hypoglycaemia was the same in both groups.

| EMP assessment and characterization
Platelet-free plasma was prepared within 2 hours of blood sample collection using an initial centrifugation at 1000 g for 10 minutes, followed by a second centrifugation of the supernatant at 12 000 g for 10 minutes. All assays were performed on a BD Accuri

| Statistical analysis
All variables were checked for extreme outliers (>3 times interquartile range above the third quartile or <3 times interquartile range below the first quartile) graphically. Participants who were indicated as extreme outliers for >3 EMPs (out of six EMPs studied) at least at one time point for each EMP were excluded from analysis (type 2 diabetes, n = 2; control group, n = 3). Total analysis was performed using the data from individuals with type 2 diabetes (n = 23) and controls (n = 22). All data were checked for normality according to the Shapiro-Wilk test. A two-way analysis of variance with repeated measures was used to determine main and interaction effects for EMPs' responses to hypoglycaemia. Non-normally distributed data were logtransformed prior to this analysis. Significant main or interaction effects were followed by Bonferroni's post hoc analysis. By using the percentage data from 0 minutes following hypoglycaemia for each time point, total and partial areas under the curve (AUC 0min-24h and AUC 0-240min ) were calculated. An independent t test or the Mann-Whitney test were used to detect differences in baseline characteristics and AUCs between groups. A step-wise multiple regression analysis was performed to explore whether significant overall responses (AUC) were predicted by age, sex, weight, height, duration of diabetes, BMI, systolic blood pressure, diastolic blood pressure, HbA1c, total cholesterol, LDL cholesterol, HDL cholesterol, triglycerides, insulin levels and hsCRP. Statistical significance was set at P ≤ 0.05. We performed additional statistical analyses to examine the clinical utility of EMPs in predicting hypoglycaemia. Our data showed that, after hypoglycaemia, the levels of EMPs increased in both patients with diabetes and controls ( Figure 1). We hypothesized that this increase reflects an endothelial injury and that those with diabetes are likely to have greater elevations after acute hypoglycaemia, hence, this should be useful in detecting hypoglycaemic episodes among these patients. We used the highest elevation in each EMP within 240 minutes after insulin-induced hypoglycaemia, and calculated the percentage rise from 0 minutes after hypoglycaemia in both cases and controls. This percentage change for each EMP was then modelled using a regression model with the following independent variables: diabetes status, age, sex, BMI, baseline HbA1c, insulin and total cholesterol levels. All statistical analyses were performed using IBM-SPSS version 24.0 (Chicago, Illinois) and R version 3.4.1.

| Demographic and clinical characteristics
The main demographic and clinical characteristics of the individuals with and without type 2 diabetes are presented in Table 1.

| EMP responses to insulin-induced hypoglycaemia
There were no significant differences in the baseline concentrations of any EMPs between individuals with type 2 diabetes and controls (all P values from 0.11 to 0.93; Figure 1). The percent rise in CD31 + EMPs (P = 0.03) was significantly higher in participants with diabetes compared with controls ( Table 2).
There was an increase in CD54 + EMPs at 120 minutes compared with baseline (P = 0.009), 0 minutes (P = 0.002) and 240 minutes (P = 0.001) following hypoglycaemia. A higher number of CD54 + EMPs was shown at 240 minutes after hypoglycaemia compared with all other time points (all P values <0.0001). CD54 + EMP responses after hypoglycaemia did not differ between groups for any time point (time × group interaction effect; P = 0.75). CD54 + EMPs AUC 0-240min (P = 0.62) or AUC 0min-24h (P = 0.91) were not different between groups ( Figure 2). The percent rise in CD54 + EMPs (P = 0.04) was significantly higher in those with diabetes compared with controls (Table 2).

| DISCUSSION
In the present study we characterized and compared the effects of acute insulin-induced hypoglycaemia on EMPs in individuals with and without type 2 diabetes. A similar pattern of changes was reported in both groups; EMP levels were increased at 240 minutes following hypoglycaemia and returned to their baseline values within 24 hours.
The elevations (% rise from 0 minutes hypoglycaemia) seen in CD31 + ,   CD54 + , CD62 + , CD105 + and CD142 + EMPs within 240 minutes after hypoglycaemia were associated with diabetes status after adjustments for covariables, indicating that their assessment within this timeframe would identify a hypoglycaemic event in this clinically relevant population. Furthermore, overall responses to hypoglycaemia over time (AUCs) were greater for CD31 + and CD105 + EMPs in individuals with type 2 diabetes compared with controls. Taken together, our findings indicate that hypoglycaemia exerts endothelial stress in individuals with and without diabetes, but this stress may be more pronounced in type 2 diabetes. When data were expressed as AUCs, overall responses for CD31 + and CD105 + EMPs to hypoglycaemia were more marked in participants with type 2 diabetes compared with healthy controls, perhaps a sign of increased apoptosis of endothelial cells and atherosclerosis in this group. 16 These results suggest that the endothelium in type 2 diabetes may be more susceptible to injury and dysfunction, and it is speculated that increased EMPs may provide a mechanistic link and T-cell subpopulations at sites of inflammation. 39 CD105 regulates TGF-β signalling in endothelial cells and is involved in haematopoiesis, angiogenesis and nitric oxide-dependent vasodilatation. 40 It has a key role in cellular transmigration, this notion supported by studies showing that CD105 also regulates the expression of extracellular matrix molecules such as fibronectin, collagen, PAI-1 and lumican. 40 CD106 is a major regulator of leukocyte transmigration and a modulator of endothelial signalling through NADPH oxidase-generated reactive oxygen species. 41 Finally, CD142, expressed by endothelial cells and leukocytes, initiates the extrinsic pathway of blood coagulation, and increased CD142 levels have been associated with thrombotic events. 42 Taken together, the EMPs that were elevated in response to hypoglycaemia in the present study play a critical role in vascular inflammation and affect the coagulation pathway. Available literature suggests the roles of EMPs are more complex than initially thought and it remains uncertain whether these EMP-mediated alterations aim to maintain vascular homeostasis in response to stimuli such as hypoglycaemia or if they contribute to endothelial dysfunction and the development of both macro-and microvascular complications in individuals with diabetes. 16,22,23,25 In conclusion, acute hypoglycaemia increased EMP levels, indicating the induction of endothelial stress, and their appearance was maximal at 240 minutes, suggesting that these EMPs, alone or in combination, may have utility as biomarkers for post hypoglycaemia, especially in patients with impaired hypoglycaemic awareness. The greater overall responses of CD31 + and CD105 + EMPs (AUCs) to hypoglycaemia in adults with type 2 diabetes suggest that the endothelium in diabetes may be sensitive to hypoglycaemia-induced injury and dysfunction and could provide a mechanistic link between hypoglycaemia and increased risk of vascular complications; however, clarity is needed on the mechanisms mediating EMP expression and the associated EMP effects related to hypoglycaemia duration and severity.