FormalPara Key Summary Points

Why carry out this study?

Approximately 463 million people suffer from diabetes worldwide.

Circulating CD34+KDR+ endothelial progenitor cells (EPCs) correlate with severity of diabetic peripheral arterial disease in patients with type 2 diabetes.

This study aimed to investigate the alteration of circulating CD34+KDR+CD133+ EPCs in patients with newly diagnosed type 2 diabetes and the mechanism of the effect of early intensive insulin therapy.

What was learned from the study?

Type 2 diabetes is associated with an increase of circulating CD34+KDR+CD133+ EPCs at the onset of diabetes.

Early intensive insulin therapy exerts a preserving effect on CD34+KDR+CD133+ EPC level partly through improving inflammation status and oxidative stress.

Introduction

Endothelial progenitor cells (EPCs) are immature bone marrow-derived cells which undergo differentiation into endothelial cells and contribute to endothelial repair and angiogenesis [1]. The number of circulating EPCs correlates with cardiovascular risk factors, as well as being an independent predictor of cardiovascular death [2, 3]. Different types of EPCs exist, differing by their maturity, clonogenic potential, and angiogenic effects [4]. In humans, EPCs are primarily identified by the expression of combinations of cell-surface markers, including CD34, CD133, and kinase insert domain-containing receptor (KDR). Although CD34+KDR+ EPCs constitute the major component of the circulating EPC pool, the CD34+KDR+CD133+ EPCs have a more potent vasoregenerative capacity and are mainly mobilized from bone marrow by chemokines and growth factors, such as vascular endothelial growth factor (VEGF) [5,6,7].

Diabetic vasculopathy is characterized by high prevalence, early development, and rapid progression. Circulating CD34+KDR+ EPCs were significantly reduced in subjects with impaired glucose tolerance and further decreased in new-onset type 2 diabetes as well [8, 9]. Moreover, the number of circulating CD34+KDR+ EPCs correlates with severity of disease, especially with the degree of diabetic peripheral arterial disease in patients with type 2 diabetes [10, 11]. However, the alteration of CD34+KDR+CD133+ EPCs at the onset of diabetes remains unknown. In addition, in patients with newly diagnosed type 2 diabetes, short-term intensive insulin therapy was shown to achieve long-term glycemic control, and significantly decreases the risk of microvascular complications [12, 13]; however, whether intensive insulin therapy could affect CD34+KDR+CD133+ EPCs is unclear.

In this study, we aimed to evaluate the alteration of CD34+KDR+CD133+ EPCs in patients with newly diagnosed type 2 diabetes and investigate the effect of intensive insulin therapy on these cells. Patients with newly diagnosed type 2 diabetes and control subjects matched by age and gender were enrolled. All of the patients with diabetes received short-term intensive insulin therapy. Blood samples were collected to detect EPCs by flow cytometry based on the expression of CD34, CD133, and KDR.

Methods

Patients and Research Design

In this study, 36 patients with newly diagnosed type 2 diabetes, according to the American Diabetes Association criteria [14], were enrolled from April to September 2017 at Nanjing Drum Tower Hospital Affiliated to Nanjing University Medical School (Nanjing, Jiangsu province, China). A further 22 control subjects who participated in health examinations at the hospital were randomly recruited. Medical history, physical examination, blood biochemistry, and clinical diagnosis were evaluated in all participants. Subjects were excluded from this study for the following reasons: history of taking antihypertensive drugs, hypolipidemic agents, estrogen supplements, herbal drugs, or antioxidants for 3 months before enrolling in the study, medical history of cardiovascular events; acute illness or infection, recent surgery or vascular intervention, ketosis, ketoacidosis, and other diseases requiring continuous medical treatment and cigarette smoking.

Patients with diabetes were treated with continuous subcutaneous insulin infusion by receiving Novolin-R (Novo Nordisk, Bagsvaerd, Denmark) with an insulin pump. Initial insulin dose was adjusted every day depending on the values from the seven-point glucose profile obtained 1 day before and was divided into 50% of basal and 50% of bolus injection. The glycemic control target was defined as fasting plasma glucose below 6.1 mmol/l and prandial plasma glucose below 8.0 mmol/l. Insulin treatment was discontinued 2 weeks later after the target goal was achieved and blood samples were re-collected for further detections.

This study was approved by the Ethics Review Committee of Nanjing Drum Tower Hospital Affiliated to Nanjing University Medical School (No. 2017-068). All patients signed written informed consent before enrollment.

Subjects’ Characteristics

Patients’ height was measured by a portable stadiometer with shoes off, to the nearest centimeter. The weight was measured on a digital scale with identical light clothing on, to the nearest 0.1 kg. BMI was calculated as weight divided by height squared. Blood pressure was measured using a mercury sphygmomanometer after the subject rested for at least 5 min. The mean of three measurements was recorded.

Blood Sample Analysis

Fasting blood samples were collected from the antecubital vein for fasting plasma glucose, insulin, glycosylated hemoglobin (HbA1c), triglycerides (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), urea nitrogen, creatinine, and uric acid measurements. The fasting plasma glucose levels were assessed by the hexokinase method using TBA-200FR analyzer (Toshiba Medical Systems, Tokyo, Japan). Insulin concentration was evaluated by a chemiluminescence immunoassay (Cobas e601; Roche, Basel, Switzerland). HbA1c was quantitated by high-performance liquid chromatography (Bio-Rad Laboratories, Inc., Hercules, CA, USA). TG, TC, HDL-C, and LDL-C concentrations were assessed by specific immunoassays (Cobas e601; Roche, Basel, Switzerland). Serum free fatty acids (FFAs) were detected via an enzymatic method (DiaSys Diagnostic Systems GmbH, Holzheim, Germany). Homeostasis model assessment of insulin resistance (HOMA-IR) was derived as follows: [fasting serum insulin (mU/l) × fasting plasma glucose (mmol/l)]/22.5 [15]. Serum 1.5-anhydroglucitol (1.5-AG) was measured using an enzymatic assay kit (GlycoMark, Japan).

Quantification of EPCs by Flow Cytometry

Peripheral blood mononuclear cells (PBMCs) were isolated with vacutainer cell preparation tubes (CPTs) (Becton Dickinson, Franklin Lakes, NJ, USA) from fasting blood samples. After centrifugation at 1800×g for 30 min, cells in the resulting supernatants were collected, washed with phosphate-buffered saline (PBS), and gradually frozen within 8 h in freezing medium containing 10% dimethyl sulfoxide (DMSO) and 90% fetal bovine serum. Samples were stored in liquid nitrogen until analysis of EPCs.

PBMCs were stained with the following antibodies: fluorescein isothiocyanate (FITC)-labeled anti-CD34 (BD Biosciences, USA), phycoerythrin (PE)-labeled anti-CD133 (Becton Dickinson, Franklin Lakes, NJ, USA), and allophycocyanin (APC)-labeled anti-KDR (Becton Dickinson, Franklin Lakes, NJ, USA). In brief, 100 μl of PBMC samples was incubated at 4 °C for 20 min in the dark with 5 μl of FITC-CD34, 10 μl of PE-CD133, and APC-KDR. Cells labeled with FITC-, PE-, and APC-conjugated isotypic monoclonal antibodies were used as controls to determine the background of fluorescence. After staining, cells were washed in 1 ml PBS and eventually re-suspended in 500 μl PBS. Analysis was then performed within 30 min. Absolute cell numbers were measured using an LSR II Fortessa flow cytometer (Becton Dickinson, Franklin Lakes, NJ, USA); data were analyzed with FACSDiva software (Becton Dickinson, Franklin Lakes, NJ, USA) or FlowJo 9.6.4 software (Tree Star Inc., Ashland, OR, USA). All results from flow cytometry were expressed as number of cells/106 total cytofluorimetric events.

Cytokine Detection and Oxidative Activity Analysis

Serum levels of tumor necrosis factor alpha (TNFα), interferon gamma (IFNγ), and interleukin-1 beta (IL-1β) were determined with a CBA Flex set (BD Biosciences, San Jose, CA). The plasma levels of VEGF, malonyldialdehyde (MDA), glutathione (GSH), and superoxide dismutase (SOD) were respectively determined in duplicate using commercially available ELISA kits according to the manufacturer’s guidelines (USCN Business Co. Ltd., Wuhan, China). The plasma levels of reactive oxygen species (ROS) were determined using the ROS colorimetric assay kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China).

Statistical Analysis

Data were expressed as mean ± standard deviation (SD) or n (%), unless otherwise specified. Differences in continuous variables were determined by the Student’s t test, paired sample t test, or Mann–Whitney U test, and differences in categorical variables were determined by the χ2 analysis or Fisher’s exact test. Correlation analyses were assessed by Spearman’s rank correlation test. P values < 0.05 were considered statistically significant.

Results

Baseline Characteristics of Study Population

As summarized in Table 1, there were no significant differences in age and sex between patients with type 2 diabetes and controls. Compared with controls, the patients with type 2 diabetes had significantly higher BMI, HbA1c, fasting plasma glucose, fasting insulin, HOMA-IR, TG, TC, FFA, and LDL-C levels. In contrast, HDL-C and 1.5-AG levels were significantly lower in the patients with type 2 diabetes than those of controls. In addition, levels of blood urea nitrogen, creatinine, uric acid, and estimated glomerular filtration rate were not significantly different between the two groups.

Table 1 Characteristics of control subjects and patients with diabetes before and after short-term intensive insulin treatment
Full size table

After the intensive insulin therapy, patients with type 2 diabetes showed significant decreases in fasting plasma glucose, TG, TC, FFA, and LDL-C levels compared with those before therapy. Although HbA1c level was not decreased, 1.5-AG, a sensitive glycemic marker, which reflects short-term glycemic changes [16], was notably increased in patients with diabetes after intensive insulin therapy. Moreover, HOMA-IR was significantly decreased, while fasting insulin level was not changed.

CD34+KDR+CD133+ EPCs are Increased in Patients with Newly Diagnosed Type 2 Diabetes and Reversed After Therapy

Compared with control subjects, circulating CD34+ cells, CD133+ cells, CD34+CD133+ cells, and CD34+KDR+ cells significantly decreased in patients with newly diagnosed diabetes, and remained unchanged after intensive insulin therapy (Table 1 in the supplementary material). Importantly, the circulating CD34+KDR+CD133+ EPC count was significantly higher in the diabetic group compared to the control group (35 ± 4 vs. 16 ± 2, P < 0.01; Fig. 1). After the intensive insulin therapy, the circulating CD34+KDR+CD133+ EPC count significantly decreased in the diabetic group, as compared to baseline values (15 ± 2 vs. 35 ± 4, P < 0.01, Fig. 1).

Fig. 1
figure 1

a Representative identification of circulating CD34+KDR+CD133+ EPCs. After gating lymphocytes and monocytes in the FSC versus SSC morphologic plot, total CD34+ cells were identified and were then examined for expression of CD133 and KDR. b Circulating CD34+KDR+CD133+ EPCs for control subjects and patients with newly diagnosed type 2 diabetes. Data are mean ± standard error. **P < 0.01 vs. control group. ##P < 0.01 vs. T2DM baseline

Full size image

Intensive Insulin Therapy Reverses Elevation of VEGF Level in Patients with Newly Diagnosed Type 2 Diabetes

The serum VEGF level in the diabetic group was significantly higher than that in the control group (29.2 ± 4.9 vs.17.9 ± 4.6, P < 0.01; Fig. 2a). After the insulin therapy, the serum VEGF level was remarkably reduced, as compared with baseline values (21.0 ± 3.6 vs. 29.2 ± 4.9, P < 0.01; Fig. 2a). Moreover, in all subjects, the CD34+KDR+CD133+ EPC count was positively correlated with the serum VEGF level (r = 0.392, P = 0.002; Fig. 2b).

Fig. 2
figure 2

a VEGF levels in control subjects and patients with newly diagnosed type 2 diabetes before and after intensive insulin therapy. Data are mean ± standard error. b Correlation between circulating CD34+KDR+CD133+ EPCs and VEGF levels. **P < 0.01 vs. control group. ##P < 0.01 vs. T2DM baseline

Full size image

Intensive Insulin Therapy Decreases Inflammation and Oxidative Stress in Patients with Diabetes

Pro-inflammatory cytokines were significantly higher in patients with diabetes (TNFα 18.9 ± 1.4 vs. 15.7 ± 2.1, P < 0.01; IL-1β 14.2 ± 5.1 vs. 9.41 ± 1.9, P < 0.05; IFNγ 16.1 ± 3.1 vs. 11.7 ± 3.3, P < 0.05, Table 2) compared to control subjects, and markedly decreased after therapy compared with baseline values (TNFα 17.1 ± 1.6 vs.18.9 ± 1.4, P = 0.009; IL-1β 10.9 ± 1.4 vs.14.2 ± 5.1, P < 0.001; IFNγ 13.3 ± 3.4 vs.16.1 ± 3.1, P < 0.01, Table 2).

Table 2 Inflammation status and oxidative stress of control subjects and patients with diabetes before and after short-term intensive insulin treatment
Full size table

Markers of oxidative stress were further detected in all subjects. The serum levels of ROS (2242.9 ± 444.2 vs. 1211.4 ± 510.1, P < 0.01, Table 2) and MDA (8.1 ± 2.4 vs. 4.8 ± 1.2, P < 0.01, Table 2) in the diabetic group were remarkably higher compared to those of the control group, and were significantly decreased after intensive insulin therapy (ROS 1542.0 ± 265.5 vs. 2242.9 ± 444.2, P < 0.01; MDA 5.7 ± 0.8 vs. 8.1 ± 2.4, P < 0.01). In contrast, the serum levels of SOD (81.8 ± 9.1 vs. 103.5 ± 3.3, P < 0.01, Table 2) and GSH (202.1 ± 14.3 vs. 266.3 ± 23.5, P < 0.01, Table 2) in the diabetic group were significantly lower compared to the control group, and were notably restored after intensive insulin therapy (SOD 97.9 ± 6.1 vs. 81.8 ± 9.1, P < 0.01; GSH 237.7 ± 11.9 vs. 202.1 ± 14.3, P < 0.01).

Discussion

In this study, our data showed that patients with newly diagnosed type 2 diabetes have significantly higher levels of circulating CD34+KDR+CD133+ EPCs, VEGF, oxidative stress, and inflammation. Increased levels of VEGF correlated with circulating CD34+KDR+CD133+ EPCs in these patients. Moreover, short-term intensive insulin therapy restored EPCs level through attenuating oxidative stress, inflammation, and VEGF level.

Our data firstly showed that circulating CD34+CD133+KDR+ EPCs were notably higher in the patients with newly diagnosed type 2 diabetes compared to controls. Among the CD34+KDR+ EPC pools, CD133+ immature EPCs are recently mobilized from bone marrow and have a more potent vasoregenerative capacity [5]. Importantly, the level of CD34+KDR+CD133+ EPCs is associated with the degree of endothelial damage, and the elevation of CD34+KDR+CD133+ EPCs represents the ability to perform neoangiogenesis and maintain the homeostasis of vascular endothelium [17]. Therefore, increased circulating CD34+KDR+CD133+ EPCs at the onset of diabetes may indicate a compensatory EPC mobilization from bone marrow into peripheral blood to improve the early vessel defect. Although appearing beneficial on first sight, this effect may be impaired after exhaustion of CD34+KDR+CD133+ EPCs during long-term hyperglycemia, thereby accelerating the development of microvascular and macrovascular complications. In support of this notion, levels of CD34+KDR+CD133+ EPCs were found to be significantly lower in patients with a long history of type 2 diabetes [18]. Importantly, our results further demonstrated that intensive insulin therapy significantly decreased circulating CD34+KDR+CD133+ EPCs in patients with newly diagnosed type 2 diabetes, indicating a protective role in preventing the excessive exhaustion of CD34+KDR+CD133+ EPCs. Indeed, intensive insulin therapy was shown to achieve long-term glycemic control and significantly decreased the risk of microvascular complications in patients with newly diagnosed type 2 diabetes [12, 13]. Notably, to exclude the effects of medications on circulating CD34+KDR+CD133+ EPCs and to elucidate the alteration of these cells at the onset of type 2 diabetes, we enrolled drug-naïve patients with newly diagnosed diabetes other than patients with long-standing diabetes in this study. Adding another long-standing diabetic group will better clarify the change in course of circulating CD34+CD133+KDR+ EPCs.

Another major finding of this study is that VEGF levels were increased in patients with newly diagnosed type 2 diabetes and were positively correlated with circulating CD34+KDR+CD133+ EPCs. During angiogenesis or for repairing injured blood vessel, EPCs have to be mobilized from the bone marrow to the peripheral blood. VEGF is a cytokine that plays a key role in EPC mobilization [6]. CD34+KDR+CD133+ immature EPCs, which mobilized form bone marrow [5], are tightly regulated by VEGF [6, 7, 19]. In agreement with previous studies, our data showed that EPCs correlated with VEGF levels; thus a higher level of VEGF may contribute to the increased number of CD34+KDR+CD133+ EPCs in patients with newly diagnosed type 2 diabetes. Of note, VEGF expression was shown to be upregulated by oxidative stress [20, 21], while inflammatory cytokines, such as TNF, could promote oxidative stress through inducing the generation of ROS by mitochondria [22,23,24]. In this study, activated oxidative stress was observed in patients with newly diagnosed type 2 diabetes, and was characterized by higher levels of ROS and MDA and lower levels of GSH and SOD. Furthermore, levels of pro-inflammatory cytokines, such as TNFα, IL-1β, and IFNγ, were found to be significantly higher in patients with newly diagnosed type 2 diabetes, and notably reduced after intensive insulin therapy. Thus, inflammation and oxidative stress may contribute to increased VEGF and subsequent EPCs at the onset of diabetes. Additionally, early intensive insulin therapy exerts a preserving effect on VEGF level partly through improving inflammation status and oxidative stress.

This study has several limitations. First, the number of patients was limited and the results need further confirmation in a larger cohort. Second, the function of CD34+KDR+CD133+ EPCs and effect of intensive insulin therapy on vascular function and integrity were not assessed in the present study. Using a mice model, which mimics the effect of intensive insulin therapy on metabolic improvement [25], we aim to determine the causative role of intensive insulin therapy in EPC regulation in the future, accompanied by investigating the vascular integrity and function through analyzing pericyte coverage of vessels, detecting fluorescein-conjugated dextran extravasation, and measuring vessel relaxation and contraction activity in response to stimulus. Of note, when circulating EPCs were detected by flow cytometry, EPCs of whole blood samples were gated in the lymphoid and mononuclear cell populations [26, 27], and many studies isolated the mononuclear cells from whole blood samples before analyzing EPCs [28,29,30,31]. In this study, PBMC were isolated to detect the alteration of CD34+KDR+CD133+ EPCs in patients with newly diagnosed diabetes. Although usage of cell preparation tubes guaranteed the optimal yield of mononuclear cells from the blood samples, it is important to point out that selective depletion caused by PBMC separation before analysis of stem/progenitor cells might exist.

Notwithstanding these limitations, we suggest that our data may have pathophysiological implications. Intensive insulin therapy may be suggested in the early phases after diagnosis of type 2 diabetes in order to suspend excessive CD34+KDR+CD133+ EPC exhaustion and preserve marrow mobilization function longer, thereby causing a long-term beneficial effect on vascular integrity and decreasing the risk of microvascular complications.

Conclusions

Patients with type 2 diabetes were associated with an increase of circulating CD34+KDR+CD133+ EPCs at the onset of diabetes, indicating increased compensatory mobilization. Importantly, early intensive insulin therapy exerts a preserving effect on CD34+KDR+CD133+ EPCs partly through improving oxidative stress and inflammation status, thereby implying a putative long-term beneficial effect on vascular integrity via suspending excessive EPC exhaustion.