2 month old male Wistar-Kyoto rats (WKY, 280–300 g) and spontaneous hypertension rats (SHR, 280–300 g) were purchased from Shanghai Center of Experimental Animals, Chinese Academy of Sciences. Rats were acclimatized in temperature and humidity-controlled rooms with a 12-h dark/light cycle throughout the study. After 8 week high salt diet, WKY rats that treated by saline were used as normal control (WKY, n = 10). SHR rats were randomly divided into four groups (10 per group): rats that treated by saline were used as hypertension model (SHR); rats in the other three groups were treated by 5 mg/kg Rg1 (SHR-Rg1(5)), 10 mg/kg Rg1 (SHR-Rg1(10)) or 20 mg/kg Rg1 (SHR-Rg1(20)). Saline or Rg 1 (dissolved in saline) were given once a day intraperitoneally. The purity of Rg1 that purchased from Shanghai Yousi Bio-Tech Co., Ltd. was more than 99% evaluated by high-performance liquid chromatography ( Additional file 1: Figure S1) and the chemical structure of Rg1 was elucidated by ¹³ C NMR ( Additional file 2: Figure S2Additional file 3: Figure S3), which is in agreement with those previous report 18. 8.0% high salt diet was fed during the whole research, and saline or Rg1 was given from the ninth week of experiment for 4 weeks. The whole experiment protocol was shown in Additional file 4 Figure S4. Experimental procedures were approved by the institute animal ethics committee (SIMM-AE-GDA-2010-05) and were in accordance with the National Institute of Health guidelines.

Systolic blood pressure (SBP) and diastolic pressure (DBP) were measured 0.5 h after the administration of Rg1 at the indicated time ( Additional file 4: Figure S4) using the tail-cuff method. Briefly, the rats were placed in a restrainer with heating pad. The blood pressure was continuously recorded by a tail-cuff apparatus (ALC-NIBP, Shanghai Alcott Biotech Co., China) that was controlled with a computer after stabilizing at 37°C for at least 10 min.

The rats were anesthetized, and a Mikro-tipped SPR-320 catheter (Millar Instruments Inc) was inserted through the right carotid artery into left ventricle. Heart rate, mean arterial pressure (MAP), left ventricular systolic pressure (LVSP), end-diastolic pressure (EDP) of rats were recorded by PowerLab 8/30 instrument (ADInstruments, Australia). Maximal rate of pressure development for contraction (+dP/dt_max) and maximal rate of pressure development for relaxation (−dP/dt_max) were all calculated from the continuously collected pressure signal.

After the treatment of Rg1, ascending aorta, heart and kidney of each rat were weighed, and then fixed by 4% neutral-buffered paraformaldehyde for 24 h. Heart index was calculated as that the heart weight was divided by body weight. All the specimens were paraffin-embedded, cut at 5 μm and were stained with haematoxylin and eosin. Photomicrographs were taken using an Olympus BX51 microscope plus Olympus DP71 CCD camera (Olympus Corporation, Japan). Software Image-Pro Plus version 6.0 was used to detect lumen diameter and media thickness of aorta. The aortic lumen diameter was calculated as the mean of the inner diameters through the center of vascular circle, and the media thickness defined as the distance between the internal and external elastic lamina. At least 16 values were measured from the points distributed evenly on every aorta. Paraffin-embedded slices were also stained with 0.1% picric sirius red (Sigma-Aldrich Inc, St Louis, USA) for fibrosis detection.

Rat eyes were fixed in 4% neutral-buffered paraformaldehyde and the solution was replaced every two days. Before retinal dissection, the eyes were washed by running water for 20 mins. Retinas were dissected, flattened on slides, washed by PBS, then incubated in PBS containing 0.5% Triton X-100 and 10% normal goat serum for 1 h at room temperature. After a short rinse with PBS, retinas were incubated with the α-SMA-Cy3 (Sigma-Aldrich, Germany) and lectin-FITC (Sigma-Aldrich, Germany) at 1000 times dilution at 37°C for 1 h. Retinal artery could be stained by both α-SMA-Cy3 and lectin-FITC positively. Photomicrographs were taken with Olympus BX51 fluorescence microscope and retinal vessel diameter was directly measured at the point of 0.5 mm from the middle of the arteriole.

Aorta, heart, kidney samples were dissected, cut into small pieces and fixed with 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer for 2 h in 4°C. The specimens were then rinsed, post-fixed in cacodylate -buffered 2% osmium tetroxide, and embedded as monolayers in LX-112 (Ladd Research Industries, USA). Ultrathin sections were contrasted with uranyl acetate followed by lead citrate and observed with a Tecnai 12 BioTwin transmission electron microscope (Philips Electronic Instruments, USA). Random sections were selected for analysis by an electron microscopy technician blinded to the treatments.

All quantitative values are given as mean ± S.E. Mean values of data from different treatment groups were compared using one-way ANOVA. After confirming the equal variances, unpaired Student’s t-test was used to compare difference between means of two groups using SigmaPlot software. P < 0.05 was considered to be statistically significant.

The structure of Rg1 was shown in Figure 1. Hypertension is well established as a stimulus for cardiovascular remodeling. It was therefore important to determine whether Rg1 regulated blood pressure or not. Throughout the experiments, the blood pressure in the four SHR groups were significantly higher than in the WKY group, but no regulation was found for Rg1 on blood pressure during the whole experiment using tail-cuff method. Furthermore, Rg1 did not show considerable regulation on left ventricular function demonstrated by HR, MAP, +dP/dt_max, -dP/dt_max and LVSP detected by Millar catheter (Table 1).

All the values are expressed as mean ± S.E. Systolic pressure (SBP), diastolic pressure (DBP), heart rate (HR), mean arterial pressure (MAP), maximum rate of pressure development for relaxation (−dP/dt_max), maximum rate of pressure development for contraction (+dP/dt_max), left ventricular systolic pressure (LVSP) were demonstrated. Heart weight divided by body weight (HW/BW) is also calculated. MAP, EDP, LVSP, +dp/dt_max, -dp/dt_max were acquired using Millar catheter and recorded by PowerLab 8/30 instrument (ADInstruments, Australia). Heart rate, SBP and DBP were acquired by tail-cuff apparatus. Mean values of data from different treatment groups were compared using one-way ANOVA. After confirming the equal variances, unpaired Student’s t-test was used to compare difference between means of two groups using SigmaPlot software. ^#P < 0.05, ^###P < 0.001 versus WKY; **P < 0.01, ***P < 0.001 versus SHR. n = 10 for each group.

Aortic structure was investigated by histopathological analysis using haematoxylin and eosin stain and shown in Figure 2A and Figure 2B. The cell alignment was un-regular and cell density was reduced in SHR relative to WKY, and Rg1 treatment markedly attenuated this phenomenon (Figure 2B). Vascular fibrosis that impairs vascular contractility was detected by picro sirus stain. No significant improvement of Rg1 on vascular fibrosis was found (Figure 2C, D). It is well known that long term hypertension could induce vascular remodeling including greater values of lumen diameter and media thickness. The lumen diameter was 1.86 ± 0.02 mm in SHR rats compared with 1.49 ± 0.02 mm in WKY rats (P < 0.001). Rg1 decreased this outward remodeling significantly (P < 0.001 for 5 mg/kg Rg1; P < 0.05 for 20 mg/kg Rg1; Figure 2E). Furthermore, the thickening of aortic media was clearly observed in the arteries of SHR comparing with WKY (173.7 ± 4.3 μm versus 100.0 ± 1.7 μm; P < 0.001), and the media thickness in 5 mg/kg Rg1 treated SHR decreased to 156.6 ± 2.1 μm (P < 0.01; Figure 2F).

Hypertension is a strong stimulus not only on conductance vessel such as aorta, but also on resistance vessel such as retinal arteries 19. In the present study, we investigated effects of Rg1 on the remodeling of small retinal arteries (Figure 3). Arteries and veins were distinguished by double-staining of the retinal blood vessels with lectin-FITC and an anti-α-SMA-Cy3 antibody, which resulted in a more pronounced α-SMA staining of small arteries and arterioles. The remodeling of retinal arteries was evaluated by arteries diameter at the point 0.5 mm to the middle of the arteriole. Hypertension induced considerable decrease on retinal arterial diameter in SHR (57.06 ± 3.47 μm) compared with WKY (104.65 ± 5.695 μm; P < 0.001). While Rg1 inhibited this inward remodeling induced by hypertension, the diameter values were 72.2 ± 4.2 μm (SHR-Rg1(5)), 61.5 ± 3.9 μm (SHR-Rg1(10)) and 69.8 ± 5.0 μm (SHR-Rg1(20)), respectively.

The ultra-structure of aorta was examined using transmission electron microscopy (Figure 4). Vascular smooth muscle cell is the main cell type of aorta and play important role on vascular function. In the present study, we detected the mitochondrial ultra-structure of smooth muscle cells of aorta. The mitochondria of SHR showed irregularly and spherical shaped, especially swollen with electron-lucent matrix. With Rg1 treatment, although the matrix of mitochondria was partial disruption, but no-swollen cristae was detected.

Cardial and renal deterioration are both complications in patients with hypertension. Histopathological detection was performed in order to investigate the protective effects of Rg1 on structure of heart and kidney in SHR. Cardiomyocyte hypertrophy was a significant characteristic in SHR compared with WKY, and this impairment was attenuated in SHR-Rg1 groups (Figure 5A). This finding is consistence with the result of heart index, the values were 2.9 ± 0.2 mg/kg (WKY), 3.9 ± 0.05 mg/kg (SHR), 3.8 ± 0.08 mg/kg (SHR-Rg1(5)), 3.8 ± 0.07 mg/kg (SHR-Rg1(10)), 3.6 ± 0.06 mg/kg (SHR-Rg1(20)), respectively (Table 1). The typical photomicrographs of glomeruli were shown in Figure 5B. Glomeruli of SHR showed prominent hyalinization (mesangial thickening) comparing with WKY rats. SHR treated with Rg1 showed the reduction in the development of hyalinization.

Myocardial fibrosis, the common end point of hypertensive heart disease, has been linked to the development of contractile dysfunction and cardiac failure. Picro sirus stain showed the fibrotic area was larger in SHR compared with WKY at the perivascular area. With Rg1 treatment, the fibrotic areas were considerably smaller than SHR (Figure 6A). Thicker and darker stain of collagen fiber was seen on SHR compared with WKY at the few-vascular area; while collagen fiber became thinner, weaker and more discontinuous with Rg1 treatment (Figure 6B). No significant improvement of Rg1 on renal fibrosis was detected (data not shown).

The damage on mitochondria of cardiomyocyte included swelling, disorganization or even disappearance of cristae in SHR group. While, no overt ultra-structural abnormality was observed in ventricular samples from Rg1 treated groups, with relative organized mitochondria (Figure 7A). Electron microscopy also demonstrated a heterogeneous thickening of the glomerular basement membrane, accompanying irregular arrangement or inter-adhesion of the podocytes in SHR group. With Rg1 treatment, podocytes were in very regular array (Figure 7B).