Assessment the effects of insulin on adiponectin, nitric oxide, myeloperoxidase and lipid profile in type 1 diabetic patients

Type 1 diabetes (T1DM) is well recognized risk factor cardiovascular disease (CVD). Insulin therapy is recommended for all patients with type 1 diabetes. Previous findings showed that diabetes impairs endothelial function and increased glucose level reduces nitric oxide (NO) output and increases myeloperoxidase (MPO) activity. However, adiponectin (APN) decreases serum glucose levels. The current study evaluated effects of insulin therapy on circulating levels of oxidative stress and CVD biomarkers like NO, APN, MPO, AIP and lipid profile in type 1 diabetic patients. Fifty patients with T1DM and 18 healthy people were enrolled in this study. The recruited people with T1DM were classified into two groups: 22 newly diagnosed (untreated) type 1 diabetic patients and 28 insulin treated patients. In all groups, circulating NO, APN, MPO, AIP and lipids levels were measured. Compared to control, untreated diabetes revealed a significant increase in the serum levels of APN, MPO, TG, VLDL, TC, LDL and AIP, with a marked reduction in NO and HDL levels. However, insulin therapy significantly lowered MPO, TC and LDL, with no significant changes in the other biochemical parameters. As expected, oxidative stress and CVD-associated markers were significantly increased in untreated diabetes. Insulin therapy exhibited a relatively positive effect on oxidative stress and CVD biomarkers. Accordingly, insulin plus antioxidant supplementation required to normalize these parameters.


Introduction
Type 1 diabetes is well-established risk factor for CVD with 3 to 4 fold higher risk of mortality in comparison to those without diabetes (Lind et al. 2014;Lee et al. 2019). All patients with T1DM required insulin therapy (Silver et al. 2018). NO is a critical modulator with various biological induce hypertension in mice (Komers and Cooper 1996), supporting for the significance of NO in the regulation of blood pressure. Because these actions are closely related to pathogenesis of atherosclerosis, inhibiting them by NO protects cardiovascular function. Among the biologically active adipocytokines, APN protein hormone which decreased in type 2 diabetes, obesity and after myocardial infarction (Su et al. 2011;Wang et al. 2017). In contrast, higher circulating APN levels have been shown in type 1 diabetic patients (Lindström et al. 2006).
MPO is a lysosomal enzyme activated and released by neutrophils (Khan et al. 2018). Previous studies have been reported that MPO may play a key role in the pathogenesis of CVD that may result from atherosclerotic plaque formation and endothelial dysfunction (Baldus et al. 2003;Heinecke 2003). In healthy people, high circulating MPO concentration is associated with increased risk of future CVD (Meuwese et al. 2007).
Increased circulating total cholesterol (RD&I Christchurch) and LDL levels are well-established predictors for CVD (Anand et al. 2011). The CVD occurring in around 2% per year in young age in patients with T1DM (Soedamah-Muthu et al. 2006). In addition, type 1 diabetic patients have 10-fold higher risk of cardiovascular mortality in comparison to general population (Lind et al. 2014). AIP is a critical marker for predicting CVD risk, because of the strong association between AIP and atherosclerosis (Begum et al. 2015).
Insulin therapy used to restore the absolute insulin deficiency caused by β-cell destruction in type 1 diabetic patients (American Diabetes 2009). Several findings suggested that insulin therapy increases the cardiovascular risk and mortality (Boyne and Saudek 1999;Herman et al. 2017;Braffett et al. 2019), independently of its notable glucose-lowering effects. These findings propose that insulin therapy could increase the cardiovascular risk by pleiotropic effects on inflammation, atherosclerosis, dyslipidaemia, heart failure, and arrhythmias (Herman et al. 2017). In T1DM children, two findings revealed that APN levels have been increased significantly after intensive insulin therapy (Hecht Baldauff et al. 2016;Martos-Moreno et al. 2006). Furthermore, two conflicting findings concerning the effect of neutral protamine hagedorn (NPH) insulin therapy on NO levels in alloxan-induced diabetic rats (Cerchiaro et al. 2001;Martins et al. 2010).
Although of being commonly used in treatment of T1DM, published data concerning the effect of conventional insulin therapy (NPH and regular insulin) on the concentrations of NO, APN and MPO in type 1 diabetic patients are not available. Therefore, the correlation between NO, APN, MPO, AIP and lipid profile in T1D on insulin therapy were not assessed simultaneously. Hence, the current study aims to determine levels of NO, APN, MPO, AIP and lipids in type 1 diabetic patients on insulin therapy, in addition to evaluate the correlations among these parameters. The evaluation for these parameters aid to identify patients with cardiovascular risk and thus prevention of development of CVD.

Patients
Our comparative cross-sectional study included fifty patients with type 1 diabetes and eighteen healthy subjects aged between 12 and 31 years, between December 2019 and April 2020. Subjects were classified into three experimental groups; 22 type 1 diabetic patients (newly diagnosed); 28 type 1 diabetic patients treated with insulin (short-acting insulin with intermediate-acting insulin) twice a day for a period of 3-16 months.; and 18 healthy subjects as a control group.
Pregnant or lactating women, patients receiving any other drugs, patients with diabetes complications or other clinical conditions were excluded from the study. Height and weight were directly measured for all participants to calculate the body mass index (BMI).

Laboratory analysis
After an overnight fasting, blood samples were collected from patients and incubated immediately in water bath for 10 min, then centrifuged at approximately 3500 rpms for 12 mins. After direct estimation of fasting serum glucose, samples were stored at -20 °C for later analysis of NO, APN, MPO, TG, HDL and cholesterol levels. LDL and VLDL were calculated by Friedewald's equation. Then, AIP was estimated as log (TG/HDL) (Dobiásová and Frohlich 2001).
Serum insulin and glucose levels were determined by enzyme linked immunosorbent assay (ELISA) and enzymatic colorimetric method, respectively. Then, glucose and insulin values were used for determination of insulin resistance using the following equation (Matthews et al. 1985): HOMA-IR = Insulin (µU/mL) × Glucose (mmol/L) / 22.5 Serum NO levels were estimated by Greiss reagent (Miranda et al. 2001). Briefly, we mixed equal volumes (200 μl) of supernatant and Griess reagent, and then absorbance was measured at 540 nm by ELISA. NO concentration was determined according to the standard curve of sodium nitrite. Circulating APN levels were estimated by ELISA, using a kit supplied by USBIOLOGICAL (USA) (Ouchi et al. 2004).
Enzymatic activity of MPO was determined by a method (Kumar et al. 2002) based on enzymatic oxidation of 0 -dianisidin (Inoue et al.) (reducing substrate) which catalysed by hydrogen peroxide (H 2 O 2 ) to produce a coloured substance measured at 450 nm. Serum lipid levels were measured by a colorimetric method that depends on sulfo-phospho-vanillin reaction (Chabrol and Charonnat 1937).

Data analysis
All data are presented as mean ± SD. Mann Whitney test and Kruskal-Wallis test followed by a Dunn's multiple comparisons test were used in statistical analysis of two and multiple datasets, respectively, using GraphPad Prism version 8.0 (San Diego, California, USA). Data values p < 0.05 were represented statistically significant.

Results
Clinical characteristics of control, untreated and insulin-treated groups.
Baseline characteristics of the study groups (age, body mass index and duration of therapy) are given in Table 1. No significant variations have been found among control, newly diagnosed and insulin-treated groups.

Validation of serum glucose, insulin and HOMA-IR
As compared to healthy control, untreated patients revealed a significant increase in serum glucose with concomitant decrease in insulin level. As expected, insulin therapy exhibited a significant increase in serum insulin and HO-MA-IR with a marked reduction in glucose level. (Table 2).
Values represent as mean ± SD. a represent differences between untreated and insulin-treated patients in contrast to healthy control; b represents differences between insulin and untreated patients. ***p < 0.001; ****p < 0.0001 represents statistically significant differences, as set by Kruskal-Wallis test followed by a Dunn's multiple comparison post-hoc test.

Validation of serum NO, APN, MPO, lipids and AIP
Newly diagnosed patients had significantly higher levels of circulating APN, MPO, TG, VLDL, TC, LDL and AIP, and significantly lower levels of serum NO and HDL as compared to healthy control. In insulin-treated patients, levels of circulating MPO, TC and LDL decreased significantly, with a relatively positive effect on the other biochemical parameters (Tables 3 and 4). MPO is negatively correlated with NO in type 1 diabetic group (r = -0.67) ( Table 4).
Values represent as mean ± SD. a represent differences between untreated and insulin-treated patients in contrast to healthy control; b represents differences between insulin and untreated patients. **p < 0.01; ****p < 0.0001 represents statistically significant differences, as set by Kruskal-Wallis test followed by a Dunn's multiple comparison post-hoc test.
Values represent as mean ± SD. a represent differences between untreated and insulin-treated patients in contrast to healthy control; b represents differences between insulin and untreated patients. *p< 0.05; ***p < 0.001; ****p < 0.0001 represents statistically significant differences, as set by Kruskal-Wallis test followed by a Dunn's multiple comparison post-hoc test.

Discussion
The present study was conducted to evaluate the impact of insulin on oxidative stress and CVD biomarkers in type 1 diabetic patients.
Alteration of NO level in diabetic patients has been reported by various studies; however, results are discrepant. Some findings revealed reduced serum NO in type 1 and type 2 diabetes (Daimon et al. 2000;Ayub et al. 2011), while the opposite effect has been found by others (Hoeldtke et al. 2003;Mylona-Karayanni et al. 2006;Adela et al. 2015). In addition, NO bioavailability was reported as being reduced in diet-induced obese and diabetic mice (Kim et al. 2008). Our findings revealed that serum NO level is reduced in diabetic patients as compared to control. Previous findings have revealed racial variations in NO levels across various ethnicities (Mata-Greenwood and Chen 2008). Insulin therapy revealed a non-significant increase in NO level. Ding et al. (2000) reported that insulin increased NO production in cultured endothelial cells incubated with high-glucose conc., but this stimulatory effect of insulin is inhibited by hyperglycaemia.    Several studies have been previously demonstrated that adiponectin level is higher in type 1 diabetic patients as compared to healthy control (Galler et al. 2007;Maahs et al. 2007;LeCaire and Palta 2015), which are in line of our findings. Again, insulin therapy reported a non-significant reduction in APN level. Skovso et al. (2015) revealed that insulin therapy has a relatively positive effect on APN in rats. Recently, it has been found that insulin works via two main molecular mechanisms: by reducing the inflammatory effect that produced by elevated free fatty acid in adipose tissue and by reducing the production of reactive oxygen species (ROS) caused by increased blood glucose level (Kadowaki and Yamauchi 2005). As a result of these metabolic effects, insulin maintains the function of mitochondria, increases APN production and thus activates AMP-activated protein kinase enzyme, resulting in depleting fat stores in adipose tissues and restoring glucose utilization and oxidation (Yamauchi et al. 2001). Moreover, by reducing the expression and effect of nitric oxide synthase in the endothelium, insulin can prevent microcirculatory changes and thereby, reducing cellular hypoxia (Langouche et al. 2005). However, the favourable effects of insulin therapy are dose-dependent.
Our findings revealed that serum MPO level were significantly increased in type 1 diabetic patients as compared to control. Heilman et al. (2009) reported that the serum level of MPO was significantly higher in type 1 diabetic children than that in healthy control, which is explain the elevated risk for CVD in type 1 diabetic patients. In contrast, another study by Saeed and Castle (1998) showed that MPO activity can be diminished in diabetic patients. Increased glucose levels are associated with modulation of cellular expression of adhesion molecules and cytokines that involved in the pathogenesis of atherosclerosis. Although the exact mechanism underlying is not clear, one possible mechanism is associated MPO's role as a mediator of vascular injury (Zhang et al. 2004). Our findings revealed a significant increase in MPO levels in type 1 diabetic patients in comparison to control group, resulting in increased risk for future CVD. In the present study, a strong inverse correlation was observed between plasma NO and MPO levels in diabetic patients, which is in accordance with previous study (Eiserich et al. 1998).
Recently, NO was suggested as modulator for peroxidase activity of MPO and acts as a substrate for MPO (Davies 2011). Therefore, increased MPO activity resulting in higher NO consumption and subsequent endothelial dysfunction. MPO can generate reactive nitrating oxidant species from NO (Eiserich et al. 1998), which have been identified as an important proinflammatory mediator in CVD (Shishehbor et al. 2003). As a result, MPO pathway may connect between inflammatory process and vascular complications in diabetes.
CVD is the main cause of mortality in type 1 diabetic patients (Libby et al. 2005). Dyslipidaemia has been found to be closely associated with T1DM (Grauslund et al. 2010). Therefore, it looks important to focus on dyslipidaemia in type 1 diabetic patients to reduce CVD. Several studies confirmed that circulating levels of CVD risk markers were higher type 1 diabetic patients, including increased TG, VLDA, TC, LDL and AIP, as well as reduced plasma HDL (Guy et al. 2009;Zachurzok et al. 2016;Zabeen et al. 2018). Interestingly, these studies are in line with our findings. The pathogenesis of this dyslipidaemia is not fully understood, but it is likely related to hyperglycaemia and defect in insulin actions (Vergès 2009;Fathi et al. 2020). Moreover, Goldberg (2001) suggested that insulin therapy may normalize or improve these abnormal lipid profile, and good blood glucose control may have enhanced HDL and reduced TG levels. In our study, insulin treated patients revealed a significant reduction in TC and LDL, as well as a non-significant improvement in circulating levels of TG, VLDL, HDL and AIP. Insulin plays a pivotal role in regulating metabolism of lipids, where insulin inhibits lipase in adipose tissue. Therefore, insulin increasing TG in the adipocytes and lowering the circulating levels of free fatty acids (Vergès 2001). Moreover, insulin suppresses hepatic production of VLDL and increases LDL clearance by up-regulation of LDL receptor expression (Mazzone et al. 1984). In addition, insulin regulates HDL metabolism by acting on hepatic activities of lipase (Ruotolo et al. 1994). Therefore, the low dose of insulin is probably the possible explanation for the relatively little effect of insulin on oxidative stress and CVD risk parameters.

Conclusion
Type 1 diabetes mellitus is closely linked with increased biomarkers of oxidative stress and dyslipidaemia. Insulin therapy produced relatively positive effects on oxidative stress and CVD risk factors. Thus, type 1 diabetic patients required a combined therapy of insulin and antioxidant (vitamin C and vitamin E) to normalize these parameters.