A total of 70 male Sprague-Dawley rats (7 weeks old) were purchased from the Experimental Animal Center of West China Hospital (Sichuan, China). The rats weighed 250±15 g and were randomly divided into two groups: Control group (group C) and the high-altitude acute hypoxia group (group H). The experiment was approved by the Chengdu Military General Hospital Ethics Committee (Chengdu, China).

MacConkey agar culture medium and KTA Tachypleus amebocyte lysate were purchased from Zhanjiang Bokang Marine Biological Co., Ltd. (Guangzhou, China), terminal deoxynucleotidyl transferase mediated dUTP nick end labeling (TUNEL) kit was purchased from Roche Diagnostics (cat. no. 11684817; Basel, Switzerland) and occludin rabbit monoclonal antibody (cat. no. 71-1500) was purchased from Thermo Fisher Scientific, Inc. (Waltham, MA, USA). A FLYDWC50-1A simulated high-altitude low-pressure chamber was purchased from Guizhou Fenglei Air Ordnance Ltd. Co. (Guizhou, China). A Tecnai 10 transmission electron microscope (Koninklijke Philips N.V., Amsterdam, Netherlands), S-3400 N scanning electron microscope (Hitachi, Ltd., Toyko, Japan) and ZH-PT experimental animal treadmill (Huaibei Zhenghua Biological Instrument Equipment Co. Ltd., Anhui, China) were also used in the present study.

The animals were raised in the respiratory department of Chengdu Military General Hospital (Chengdu, China) at an altitude of ~500 m and a temperature of 24°C for 1 week to allow them to adapt to the environment. Food and water were provided ad libitum and a natural light cycle (12 h light/12 h dark) was adopted.

The adaptive training began at 8:00 a.m. on day 8 and was conducted as follows: A total of 10 tracks of the ZH-PT experimental animal treadmill were set to a slope degree of 0° and a rotating speed of 8 m/min. Rats exercised for 30 min, followed by a 30 min rest. During the training, rats were free to drink and had access to food ad libitum. The training was repeated 5 times in the same day, with a total training time of 2.5 h.

The present model was established based on the high-altitude pulmonary edema (HAPE) animal model described by Bai et al (14), which did not aim to exhaust the animals. In order to induce fatigue stress, the treadmill settings were adjusted. At 8:00 a.m. on day 9, the animals were put on the treadmill, the slope of the conveyor belt was set to 0° and rotation speed was set to 8 m/min. The rats ran for 15 min to adapt to the treadmill. Following this, the slope was adjusted to 10° and the rotation speed was increased to 15 m/min. The rats exercised for 4 h, followed by a 30 min rest. Rats had ad libitum access food and water during the training. The training was repeated until 8:00 a.m. on day 10, after 24 h training. If the rats stopped, they were transported to the end of treadmill where there were some electrified pillars. Thus, appropriate electrical stimulation was applied to maintain the running state. The aim was to make the exercise as consistent as possible for all animals.

In order to stimulate acute hypoxia in group H rats, at 8:00 a.m. on day 10, rats in groups C and H were moved into separate chambers of the FLYDWC50-1A simulated high-altitude low-pressure chamber. A compression resistant door separated the chambers, with an adapter chamber in the middle and a precise control system regulated the two separate environments. Rats in group C were kept at a temperature 23–25°C, humidity of 60%, wind speed of 0.3 m/sec and a 12-h light/dark cycle. A total of 25 g/day rat food was provided for each rat (12.5 g every 12 h) and the rats were free to get water. Hypobaric hypoxia was not applied to the group C rats. For group H, the access to food and water were same to group C, but the altitude was set to 4,000 m, the chamber pressure was 40.4 kPa, the air capacity was 40 m³/h, the temperature was 25°C, the humidity was 60%, the O₂ content was 18.0%, the oxygen partial pressure was 7.8 kPa, the ascending rate was 3 m/sec and the light/dark cycle was 12-h.

Following 72 h in the separate environments, blood and abdominal visceral tissue samples from rats in the two groups were collected. Following the induction of anesthesia (10% ethyl carbamate, 1 ml/kg body weight by intraperitoneal injection; Chengdu Kelong Chemical Co., Ltd. (Chengdu, China), samples from group H were collected within 10 min by surgical resection at a simulated altitude of 4,000 m and all operations were strictly aseptic. From each group (n=35), 20 rats were selected for sacrifice by surgical resection as described above, and the blood, liver, spleen, mesenteric lymph nodes and jejunum were collected and preserved in liquid nitrogen (−195°C). The remaining 15 rats from each group were anesthetized (10% ethyl carbamate, 1 ml/kg body weight by intraperitoneal injection) and 2 cm jejunum was removed, 10 cm from the pylorus. The jejunum was sliced into 1-mm samples, which were immersed in fixing solution for pathological observation. For each group, 5 samples were immersed in 4% paraformaldehyde fixation fluid, 5 samples were immersed in 2.5% glutaraldehyde fixation fluid, and the rest were immersed in lanthanum aldehyde fixation solution (2% triformol, 2.5% glutaraldehyde. 2% lanthanum nitrate, 0.1 M sodium cacodylate trihydrate). All of the fixing solutions were at 4°C. Samples were fixed for 24 h.

To detect bacterial translocation, a homogenate was produced by adding 900 µl sterile saline solution to 100 mg samples of liver, spleen and mesenteric lymph node. A total of 0.3 ml homogenate was inoculated on MacConkey culture medium (17 g peptone, 5 g pig bile salts, 5 g NaCl, 17 g agar, 1,000 ml distilled water, 10 g lactose, 10 ml 0.01% crystal violet aqueous solution, 5 ml 0.5% neutral red aqueous solution). Following incubation at 37°C for 24 h, primary biochemical identification was applied. The bacterial translocation rate (%) = positive organ number in the same group/total number of cultured organs.

Dynamic turbidimetric analysis was used to detect the level of lipopolysaccharides (LPS) in the serum. A total of 3 ml serum was collected by apex puncture and centrifuged at 4°C at 960 × g for 15 min. From this, 100 µl serum was collected and 900 µl KTA Tachypleus amebocyte lysate was added. This was mixed evenly and kept at 65°C for 20 min. Following this, 200 µl mixture was added to the primary reagent of the enzyme reaction (from the kit) and LPS content was automatically detected following 60 min using a Bacterial Endotoxin Dynamic Testing system (EDS-99; Tianjin Wireless Electronic Co., Ltd., Tianjin, China).

Perfusion in situ combined with post-fixation was used to prepare the light microscope specimens. Animals were anesthetized (10% ethyl carbamate, 1 ml/kg body weight, by intraperitoneal injection), the right auricle was opened and the left ventricle was intubated. For perfusion, 200 ml normal saline at 4°C and 100 ml 4% triformol were used. When there was clear effluent from the right auricle and the liver was hard, the perfusion was deemed successful. The 2 cm of jejunum was cut into 1-mm sections, following fixation using 4% triformol at 4°C for 24 h, the specimens were dehydrated with 95% ethyl alcohol and embedded in paraffin. Then, the samples were cut into 4-µm sections and stained with hematoxylin and eosin (H&E). The changes in the intestinal epithelium villi and mucous layer were observed using a CX23 optical microscope (Olympus Corporation, Tokyo, Japan); 10 high-power fields were selected in every section to detect the height and surface area of villi using the following formula: S=2πrh, where r, radius of villus and h, height of villus. The mean value was noted.

The perfusion and fixation solutions were changed to 2.5% glutaradehyde solution in order to observe the samples using a scanning electron microscope, but other protocols were the same as previously stated for light microscopy. The samples were placed in 2.5% glutaradehyde solution at 4°C, and prepared as specimens according to the routine specimen preparation method (15), prior to observation under the scanning electron microscope.

In order to observe samples using a transmission electron microscope, perfusion and the fixation solutions were changed to lanthanum aldehyde fixation solution (2% triformol, 2.5% glutaraldehyde, 2% lanthanum nitrate, 0.1 M sodium cacodylate trihydrate), as lantham nitrate was used as the tracer. The samples were put into lanthanum aldehyde solution for fixation at 4°C for 24 h and prepared as specimens according to the routine specimen preparation (15), prior to observation under the transmission electron microscope.

A TUNEL assay was used to detect apoptotic epithelial cells. The samples were routinely fixed in 4% triformol at 4°C, embedded and cut into 5-µm paraffin slices, according to the manufacturer's protocol (Roche Diagnostics). A stained nucleus, in blue, was defined as positive for apoptosis. Every section was observed under a light microscope, at a magnification of ×400, to count the cells (total cells of 4 fields/4), and the apoptotic index was calculated as follows: Apoptotic cell number/total cell number.

Total proteins from the jejunum tissue were extracted from a 50-mg sample and protein concentration was detected using a bicinchoninic acid (BCA) protein assay kit (Beyotime Institute of Biotechnology, Haimen, China). The protein was diluted to 5 µg/µl and 5X SDS loading buffer was added prior to heating to 100°C for 3 min to degenerate the proteins. A total of 20 µg/µl protein solution was loaded in each well; this then underwent electrophoresis on a 5% SDS stacking gel and a 10% SDS separating gel at 90 V for 100 min. This was transferred to PVDF membrane by 100 V for 50 min, and put into 5% skimmed milk powder solution to block for 1 h at 25°C. Then, occludin primary antibodies (71–1500; dilution 1:1,000) and β-actin primary antibodies (sc47778; dilution 1:3,000; Santa Cruz Biotechnology, Inc., Dallas, TX, USA) were added and the samples were incubated at 4°C overnight. Following washing with Tris-buffered saline and Tween-20, the corresponding secondary antibodies (ZB-2301 and ZB-2305; Beijing Zhongshan Golden Bridge Biotechnology Co., Ltd., Beijing, China) for occludin and β-actin respectively, were added and the membrane was incubated at 25°C for 1 h. The proteins were then developed using a BeyoEcl Plus kit (P0018; Beyotime Institute of Biotechnology Beijing, China). The relative grey value was used to quantify occludin expression, using ImageJ software, v2.1.4.7 (National Institutes of Health, Bethesda, MD, USA) according to the following formula: Relative grey value = grey value of occludin band/grey value of β-actin band.

SPSS, version 18.0 (SPSS, Inc., Chicago, IL, USA) was used for statistical analysis. Enumeration data were presented as percentages and the measurement data were presented as mean ± standard deviation. The comparison among groups was completed by independent-sample t-test and the comparison of ratio was completed by χ² analysis. P<0.05 was considered to represent a statistically significant difference.

There was no bacterial translocation detected in group C and the bacterial translocation was significantly higher in group H (P<0.01; Table I).

The serum LPS concentration in group C was 0.051±0.011 EU/ml and the LPS concentration in group H was 0.803±0.115 EU/ml, a statistically significant difference (P<0.01; Fig. 1).

The results from light microscopy demonstrate that in group C, the epithelial mucous was smooth, the villi were arranged regularly and there was an even distribution of goblet cells among epithelial cells (Fig. 2A). In group H, the intestinal epithelial mucous was thinner, the villi reduced, the arrangement disordered and there was extracapillary inflammatory cell infiltration and red blood cell exudation (Fig. 2B). Compared with group C, the villi height, surface area and mucous thickness were significantly decreased in group H (P<0.05; Table II).

Scanning electron microscopy demonstrated that for group C, the villi were smooth, arranged regularly and morphology was plump (Fig. 3A), however, in group H, the epithelial microvilli were shorter, areas had shrunk and broken. Furthermore, there was space widening (Fig. 3B).

Transmission electron microscope results indicated that in group C, there was a high density of lanthanum granules attached to the capillary cavity and microvilli surfaces, the tight junction between epithelial cells was closed and there was no lanthanum granule deposition in the basement membrane or extracapillary mesenchyme (Fig. 4A). The cytoplasm of epithelial cells was complete and the morphology of endoplasmic reticulum and mitochondria was normal (Fig. 4B). In group H, the tight junction among epithelial cells was open and lanthanum granules permeated into the widened tight junction (Fig. 4C). There were lanthanum granules in the extracapillary space and basement membrane, indicating continuous linear distribution. Furthermore, parts of the organelles were swollen (Fig. 4D).

In group C, there were no clearly apoptotic cells in the intestinal epithelium (Fig. 5A), and the TUNEL assay indicated that the apoptotic rate was 3.4±1.1% (data not shown). In group H, there were a large number of apoptotic cells in the intestinal epithelium (Fig. 5B) and the apoptotic rate was 18.7±1.9% (data not shown). The difference in apoptotic rates was statistically significant (P<0.01).

Western blot analysis was used to investigate the expression of occludin (Fig. 6A), and the expression levels were quantified by densitometric analysis (Fig. 6B). Occludin expression was significantly downregulated in group H (0.42±0.011) compared with group C (1.31±0.084; P<0.05) in the tight junction of the epithelium in rats.