Between August 1, 2014 and November 1, 2014, 30 healthy control subjects (35–70 years) and 69 patients who met the diagnostic criteria (17,18) of type 2 diabetes mellitus (T2DM; 50–80 years) were recruited following informed consent. Control subjects were healthy individuals without large artery atherosclerosis, diabetics or hypertension. Carotid ultrasonography was performed on each patient. A high-resolution B-mode ultrasound machine (E8, X300PE; GE Healthcare, Little Chalfont, UK) was used to examine the common carotid arteries. The region from the beginning of the bifurcation bulb to 15 mm distal was examined. Atherosclerotic plaques were defined as focal structures encroaching into the arterial lumen by 0.5 mm or 50% of the surrounding intima-media thickness (IMT) value or an IMT value >1.5 mm. Carotid atherosclerosis was defined as the presence of atherosclerotic plaques in any of the aforementioned arterial segments. A total of 69 patients were divided into the diabetes with atherosclerosis group (n=42) and diabetes without atherosclerosis group (n=27) according to the results of sonography. The present study was approved by the Institutional Review Board and the Ethics Committee of Peking University First Hospital (Beijing, China). The patients were administered NXT (Shandong Buchang Pharmaceutical Co., Ltd., Xian, China) orally at a dose of 1.2 g per day for 3 months. Healthy subjects without cardiovascular risk factors or disease were included in the study as baseline control and were not subjected to any intervention.

NXT (batch no. Z20025001) is composed of 16 herbs. The herbal materials were powdered and assembled into capsules. Five major compounds of NXT were analyzed by ultra-performance liquid chromatography for intestinal absorption, these were hydroxysafflor yellow A, paeoniflorin, ferulic acid, salvianolic acid B and ligustilide (19,20).

Following a 12-h fast peripheral venous blood samples (4 ml) were drawn from the healthy control subjects and the patients into ethylenediamine tetraacetic acid (EDTA) tubes and kept in a refrigerator at −80°C prior to the NXT intervention. In addition, samples were drawn following 90 days of treatment following a 12-h fast. Plasma (2 ml) was isolated by centrifugation at 4°C and 1,000 x g for 15 min. HDL (1.063–1.210 g/ml) were isolated by ultracentrifugation as previously described (21). Briefly, the plasma density was adjusted to 1.3 g/ml with KBr and normal saline (1.006 g/ml) was layered over the adjusted plasma to form a discontinuous NaCl/KBr density gradient. The tubes loaded with the sample and gradient were placed in a P40ST rotor of an ultracentrifuge (CP70MX; Hitachi, Tokyo, Japan) and were centrifuged at 350,000 × g for 3.5 h at 4°C. The HDL layer was collected. The purity of HDL was evaluated by 12% SDS-PAGE and western blot analysis using goat anti-apoA-I polyclonal antibody (DiaSorin, Stillwater, OK, USA) and quantified through the measurement of apoA-I content by nephelometry (Dimension XPand; Dade Behring, Marburg, Germany). HDL was dialyzed with 0.01 M phosphate-buffered saline (PBS; 0.1% EDTA-Na₂) for 72 h. The purity of HDL and the level of apolipoprotein A-I (apoA-I) in HDL was determined.

HUVECs were isolated from umbilical veins of fresh umbilical cords as previously described (22). Umbilical cords were donated by pregnant volunteers (n=3; 26–28 years) following written informed consent on June 20, 2014 from the Department of Gynecology (Peking University First Hospital). The umbilical vein was washed with PBS three times and perfused with 100 U/ml collagenase type IA (Sigma-Aldrich, St. Louis, MO, USA). Following incubation at 37°C for 15 min, HUVECs were collected and added to complete EC medium (ECM; Sciencell Research Laboratories, Carslbad, CA, USA). HUVECs were resuspended in ECM following centrifugation for 5 min at 25°C and 100 × g. HUVECs were detached with 0.05% trypsin (Sciencell Research Laboratories), at a ratio of 1:3. Cells were cultured in a humidified atmosphere of 5% CO₂ at 37°C for 3–5 days until confluence.

The bromodeoxyuridine (BrdU; (Roche, Basel, Switzerland) assay was performed as described previously (15). HUVECs incubated with different types of HDL at 100 µg/ml apoA-I concentration for 12, 24, 36 or 48 h. The cells were labeled with BrdU labeling solution (Roche) and then fixed with paraformaldehyde (Sigma-Aldrich). Following incubation with peroxidase-conjugated anti-BrdU working solution (Roche) for 90 min at 37°C, the cells were washed with washing buffer three times and substrate solution, 3,3′,5,5′-tetramethylbenzidine (Roche) was added. The absorbance of each well was measured at a wavelength of 450 nm with an ELISA plate reader (Model 550; Bio-Rad Laboratories, Inc., Hercules, CA, USA).

Wound healing and Transwell assays were used to analyze HUVEC migration. The procedure was conducted as described previously (10). The HUVEC monolayer was scraped using a pipette tip and then incubated with EC medium containing 1% bovine serum albumin (BSA) alone or with different types of HDL at an apoA-I concentration of 100 µg/ml for 18 h. Images of the migrated cells were captured and quantified in 10 random fields (CK40; Olympus Corp., Tokyo, Japan), and results were repeated for 15 individuals from each group.

Quantitative migration assays of HUVECs were performed using a modified Boyden chamber (Minicell; EMD Millipore, Billerica, MA, USA) with an 8.0-µm pore polycarbonate filter inserted. Cells were treated as described in the wound healing assay. Following migration for 5 h, migrated cells were fixed and stained with crystal violet. Migrated cells were visualized in 10 high-powerfields (magnification, ×100), and images were captured for each chamber and quantified in 5 random fields. Results were repeated for 10 individuals from each group.

HUVECs were seeded at a density of 5×10⁵ cells/well in 6-well plates and cultured to 70% confluence. HUVECs were pretreated with serum-free ECM with different types of HDL at 100 µg/ml apoA-I concentration for 18 h, and then incubated in the presence of TNF-α (200 U/ml; BD Biosciences, Franklin Lakes, NJ, USA) for 24 h. Cells were washed with PBS three times and fixed with 4% paraformaldehyde for 10 min at room temperature. Subsequently, cells were washed twice with PBS and stained with Hoechst 33258 (Beijing TransGen Biotech Co., Ltd., Beijing, China) for 20 min at room temperature. Following three washes with PBS, cells were observed using a microscope (LSM 510; Zeiss GmbH, Jena, Germany).

Caspase-3 activity in the cells was detected using a Caspase-3 Colorimetric assay kit (Beyotime Institute of Biotechnology, Haimen, China). Cells were treated as described above. Cell lysates were centrifuged at 10,000 × g for 15 min at 4°C, the supernatants were collected and the protein concentration determined by BCA Protein assay kit (Pierce Biotechnology, Inc., Rockford, IL, USA). Cellular extracts (30 µg) were then incubated in a 96-well microtitre plate with 20 ng Ac-DEVD-pNA (Beyotime Institute of Biotechnology) for 2 h at 37°C. Caspase activity was measured by cleavage of the Ac-LEVD-pNA substrate to pNA, the absorbance of which was measured at a wavelength of 405 nm. Relative caspase activity was calculated as a ratio of emission of fluorescence of treated cells to untreated cells.

The effects of different HDL treatments on in vitro angiogenesis were assessed by Matrigel tube formation assay as described previously (23). Each well of pre-chilled 96-well plates was coated with 50 µl Matrigel (BD Biosciences) and incubated at 37°C for 30 min for solidification. Following removal of any fluid, ~4×10⁴ HUVEC cells were seeded and cultured in free ECM containing different types of HDL at a concentration of 100 µg/ml apo-AI. Images were digitally captured using an Olympus microscope (Olympus Corp.) 4 h after seeding. Tube formation was assessed by counting the length of tubes from 10 random fields (magnification ×100). Results were repeated for 10 individuals from each group.

HUVECs were cultured in 12-well plates and treated with ECM or the various types of HDL at an apoA-I concentration of 100 µg/ml for 15 min for the indicated times. Following administration of HDL for 15 min, the cells were lysed with radioimmunoprecipation assay buffer [(Beijing TransGen Biotech Co., Ltd.) 50 mM Tris, pH 7.4, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS), 0.1% EDTA and 1% Triton X-100]. Cell protein samples were separated by centrifugation (4°C, 10,000 × g for 20 min), and the concentration was detected by Coomassie brilliant blue (Sigma-Aldrich). The protein samples (40 µg) were subjected to 10% SDS-polyacrylamide gel (Beijing TransGen Biotech Co., Ltd.) electrophoresis (100 V for 120 min) and transferred onto nitrocellulose membranes (EMD Millipore) according to standard protocols. The membranes were blocked for 2 h with 1% BSA containing 0.05% Tween 20 in Tris-buffered saline (Beijing TransGen Biotech Co., Ltd.). Membranes were incubated with primary antibodies, including monoclonal rabbit anti-human phosphorylated (p)-Akt (phospho S47; cat. no. ab81283; Abcam, Cambridge, MA, USA; 1:1,000), monoclonal rabbit p-extracellular signal-regulated kinase (ERK)1/2 (Cell Signaling Technology, Inc.; 1:1,000; T202/Y204; cat. no. 4377), monoclonal rabbit anti-ERK1/2 (cat. no. 9102; Cell Signaling Technology, Inc.; 1:1,000) and rabbit monoclonal anti-AKT1/2/3 antibody (cat. no. ab179463; Abcam) overnight at 4°C followed by horseradish peroxidase-conjugated goat-anti-rabbit IgG secondary antibody (cat. no. SC-2004; Santa Cruz Biotechnology, Inc., Dallas, TX, USA; 1:1,000). Protein bands were detected by electrochemiluminescence using the Super Signal West Pico kit (Pierce Biotechnology, Inc.).

All experiments were reproduced in triplicate. In the majority of experiments, the control sample was defined as 100%, and for each sample, increase or decrease of the percentage in other samples compared with the control was calculated. Data were presented as the mean ± standard error of the mean. Differences were compared with two-tailed Student's t-test using GraphPad Prism software (version 5.0; GraphPad Software, Inc., La Jolla, CA, USA) and P<0.05 was considered to indicate a statistically significant difference.

All the patients underwent ultrasonography of the carotid and intracranial arteries. The diabetic patients were divided into two groups, depending on these ultrasound findings: i) individuals free of atherosclerotic stenosis, the diabetic group and ii) individuals with atherosclerotic stenosis, the diabetic-atherosclerosis group (24). HDLs obtained from diabetic groups prior to NXT therapy were designated D-HDL baseline (BL) and D-atherosclerosis (As)-HDL BL. HDL obtained from the two groups following 3 months of NXT therapy were designated D-HDL M3 and D-As-HDL M3. Baseline demographic and clinical characteristics of the diabetic patients and healthy subjects are presented in Table I. No significant difference was observed in age, gender, body mass index, MAP and lipid profile between the diabetic and diabetic-atherosclerosis groups. The NXT treatment was well-tolerated by all subjects in the two groups and there were no significant side-effects observed during therapy. All the patients underwent the anti-diabetic therapy (insulin or oral drugs, including metformin, sulfonylurea, α-glucosidase and thiazolidinediones). Statins were prescribed to 5 patients of the diabetic-atherosclerosis group and 4 of the diabetic group. Anti-hypertensive therapy was prescribed to 15 patients of the diabetic-atherosclerosis group and to 14 of the diabetic group. Characteristics of the patients in the two groups prior to and following treatment are presented in Table II. Compared with the baseline characteristics, there were no significant changes observed in the lipid profile except for the low density lipoprotein (LDL)-c level in the diabetic-atherosclerosis group, which was decreased (P=0.043). In addition, the hemoglobin A1c level in patients of the two groups decreased significantly (P<0.01).

The HUVECs were stimulated by different types of HDL in a time-dependent manner and similar patterns of action were observed in cell proliferation. Normal (N)-HDL (100 µg/ml apo-AI) demonstrated an increase in relative cell number for 36 h, while D-HDL BL and D-As-HDL BL demonstrated diminished increases. HUVEC proliferation was increased by 27% in the D-HDL M3 group when compared with the D-HDL BL group at 48 h (Fig. 1). While the D-As-HDL M3 group demonstrated an increased ability (by 10%) to stimulate cell proliferation compared with D-As-HDL BL at 48 h (Fig. 1).

Wound healing and Transwell assays were performed to analyze HUVEC migration (Fig. 2). Compared with N-HDL, D-HDL BL and D-As-HDL BL were less effective (38 and 50%, respectively) at stimulating HUVEC migration in the wound healing assay (Fig. 2A and C). D-HDL M3 stimulated 22% greater cell migration compared with D-HDL BL, and the D-As-HDL M3 group demonstrated 15% more migration when compared with D-As-HDL BL group (Fig. 2C).

Results of the Transwell assay were similar. D-HDL M3 had a greater ability (by 35%) to stimulate cell migration compared with D-HDL BL (Fig. 2B and D). The D-As-HDL BL group stimulated less cell proliferation than the D-As-HDL M3 group (Fig. 2D).

Compared with N-HDL, D-HDL BL and D-As-HDL BL were 51 and 61%, respectively, less effective in tube formation (Fig. 3A and B). Following 3 months of NXT therapy, the D-HDL M3 and D-As-HDL M3 groups demonstrated increased ability to stimulate cell tube formation (increase of 54 and 30%, respectively) when compared with the D-HDL BL and D-As-HDL BL groups (Fig. 3B).

HUVECs exhibited marked apoptosis following TNF-α treatment. N-HDL treatment reduced rates of apoptosis by 45% (Fig. 4A and B). D-HDL BL and D-As-HDL BL demonstrated reduced ability to protect HUVECs from apoptosis; D-HDL BL and D-As-HDL BL were 48 and 35%, respectively, less effective when compared with N-HDL (Fig. 4B). Following 3 months of NXT therapy, the D-HDL M3 group exhibited an increased ability (by 34%) to inhibit cell apoptosis when compared with D-HDL BL (Fig. 4B). While, the increasing ability in inhibiting cell apoptosis of D-As-HDL M3 was 23% (Fig. 4B).

HUVECs treated with N-HDL + TNF-α exhibited 48% less caspase-3 activity in cell lysates than cells treated with TNF-α alone (Fig. 4C). The D-HDL BL and D-As-HDL BL pretreatment groups exhibited 72 and 64%, respectively more caspase-3 activity when compared with N-HDL (Fig. 4C). Following the 3-month NXT therapy, the D-HDL M3 group exhibited 53% less caspase-3 activity when compared with D-HDL BL, while no change in the caspase-3 activity level was exhibited in the D-As-HDL M3 group (Fig. 4C).

Akt and ERK1/2 phosphorylation had been demonstrated to be important in signal transduction for EC proliferation, migration and angiogenesis (25,26). The present study analyzed whether NXT therapy increased D-HDL function by increasing Akt and ERK1/2 phosphorylation (Fig. 5). Compared with N-HDL, D-HDL BL and D-As-HDL BL demonstrated reduced phosphorylation of Akt (D-HDL BL, 28%; D-As-HDL BL, 51%). HUVECs stimulated by D-HDL M3 exhibited 36% more Akt phosphorylation than those stimulated by D-HDL BL (Fig. 5A and C). HUVECs stimulated by D-HDL M3 exhibited 26% more Akt phosphorylation than those stimulated by D-HDL BL. However, no significant difference in Akt phosphorylation was observed between HUVECs stimulated by D-As-HDL M3 and those stimulated by D-As-HDL BL (Fig. 5A and C).

ERK1/2 phosphorylation was also observed. Similarly, compared with N-HDL, D-HDL BL and D-As-HDL BL reduced phosphorylation of ERK1/2 by 34 and 26%, respectively (Fig. 5B and D). HUVECs stimulated by D-HDL M3 demonstrated 80% greater ERK1/2 phosphorylation than those stimulated by D-HDL BL (Fig. 5B and D). HUVECs stimulated with D-As-HDL M3 exhibited 16% more ERK1/2 phosphorylation when compared with those stimulated by D-As-HDL BL (Fig. 5B and D).