A total of 90 Sprague Dawley specific pathogen-free rats (weight, 200–300 g; age, 2–3 months; female, n=45; male, n=45) were purchased from Shanghai Jiesijie Laboratory Animal Co. Ltd. (Shanghai, China). The rats were housed at 26±1°C in a standalone environment under an alternating day and night cycle of 12/12 h with free access to water and food. Rats were made to fast for 24 h prior to model construction and the PVT model was constructed through IPVO combined with endangium destruction. The rats were anesthetized by an intraperitoneal injection of 5% chloral hydrate (350 mg/kg) and fixed on an operating table. The abdominal cavity was opened by making a longitudinal incision (4–5 cm) along the epigastric midline (Fig. 1A and B). The intestinal canal was removed from the abdominal cavity with a wet cotton swab and covered with gauze containing physiological saline (Fig. 1C). Connective tissue around the portal vein was bluntly separated and 1.5–2 cm of the free portal vein was isolated (Fig. 1D). Next, the proximal and distal ends of the free portal vein (including the epidural catheter) were ligated using a suture 4-0 (Fig. 1E). Toothless tweezers were used to clamp the portal vein at the proximal and distal ends for 5 sec every 5 min (Fig. 1F and G). Portal vein ligation was stopped after 20 min of treatment and then blood flow was allowed for 10 min. These steps (ligation and clamp) were repeated five times. Finally, 2 ml 3% ceftazidime solution was injected into the abdominal cavity and the abdominal cavity was sutured (Fig. 1H). The formation of PVT (including the thrombus size and vessel diameter) was continuously observed prior to and following the surgery using a B-mode ultrasound instrument (Philips Healthcare, Amsterdam, The Netherlands). The sham operation group was used as a control; in this group the portal vein was only dissociated, and no ligation and clamping was performed.

RT-qPCR analyses were performed to detect the mRNA expression level of P-selectin, a biomarker of PVT (20) in rats of the model and control groups 24 h following surgery. Total RNA was extracted from blood samples using RNA Extraction kit and reverse transcribed into cDNA using the First Strand cDNA Synthesis kit (both Beyotime Institute of Biotechnology, Shanghai, China). The following thermocycling conditions were used for cDNA synthesis: 45°C for 60 min and 70°C for 10 min. qPCR was subsequently performed using the BeyoFast^™ SYBR Green qPCR Mix (2X) kit (Beyotime Institute of Biotechnology) using the following specific primer pairs: P-selectin forward, 5′-GAGGCAGAGACCTCACAGCCAG-3′ and reverse, 5′-GTCAGGTAAGTGGCCAATG-3′; and β-actin forward, 5′-ACACCTTCTACAATGAGCTG-3′ and reverse, 5′-CTGCTTGCTGATCCCATCT-3′. The following thermocycling conditions were used for qPCR: Initial denaturation at 94°C for 3 min; 30 cycles at 94°C for 30 sec, 62°C for 30 sec and 72°C for 30 sec; and a final extension at 72°C for 5 min. The relative P-selectin mRNA levels were quantified using the 2^−ΔΔCq method (21) and normalized to the internal reference gene β-actin.

Rats were divided into three groups (10 rats/group): Model, Control and rPSGL-Ig. rPSGL-Ig was prepared by Hangzhou S-Evans Biosciences Co., Ltd. (Hangzhou, China). A total of 4 mg/kg rPSGL-Ig was intraperitoneally injected into PVT rats at 1 h after model construction (the rPSGL-Ig group). The formation of PVT in the rPSGL-Ig group was further evaluated by B-scan ultrasonography 6, 12 and 24 h after the surgeries, and by HE staining and TEM.

HE staining was performed on the portal vein, central hepatic vein and vasa intestinae tenuis of rats in different groups 24 h following surgery. Tissue samples were fixed in 2% paraformaldehyde overnight at 4°C. Tissue samples were dehydrated by ascending ethanol series (50, 70, 80 and 90% ethanol each for 15 min, 70% ethanol overnight, and 100% ethanol for 20 min), soaked in acetone twice for 15 min and embedded in Araldite. Following 48-h polymerization at 65°C, the embedded tissue samples were cut into ultrathin slices (70 nm) using an ultramicrotome (EM UC7; Leica Microsystems GmbH, Wetzlar, Germany). Tissue sections were stained with hematoxylene for 4 min at room temperature and couter stained with eosin for 90 sec at room temperature, and observed under a light microscope (YYS-190E; Shanghai Optical Instrument, Shanghai, China) at a magnification ×100 and ×400.

TEM were performed on the portal vein of rats in different groups. Tissue slices were prepared as described above. Following staining with uranyl acetate and lead citrate (both Sinopharm Chemical Reagent Co., Ltd., Beijing, China) each for 10 min at room temperature, the samples were observed by TEM (JEM-1230; JEOL, Ltd., Tokyo, Japan) 6, 12 and 24 h after the surgeries.

PVT model SD rats were divided randomly into four groups (n=6/group): Control, 4 mg/kg rPSGL-Ig, 6 mg/kg rPSGL-Ig and 8 mg/kg rPSGL-Ig. Whole blood was collected from the portal vein at 48 h to detect the mRNA changes in P-selectin by RT-qPCR. Next, the PVT model SD rats (n=10/group) were injected with saline, 8 mg/kg rPSGL-Ig or 2×10⁴ U/kg urokinase (URO; Livzon Pharmaceutical Group Inc., Zhuhai, China), a commonly used thrombolytic drug (22). The thrombus size pretherapy, and 12, 24, and 48 h after rPSGL-Ig treatment were observed using a B-mode ultrasound instrument.

Quantitative data were expressed as mean ± standard deviation. The quantitative data between groups were analyzed by a one-way analysis of variance followed by Bonferroni post hoc tests, which were analyzed with SPSS version 17.0 (SPSS, Inc., Chicago, IL, USA). P<0.05 indicated that the difference between groups was statistically significant.

A rat model of PVT was successfully constructed (Fig. 1). To examine the stability of the thrombus, the vascular wall with visible thrombus was punctured at 6, 12 and 24-h post-surgery. An unstable red thrombus, which occurred following a hemorrhage, was formed 6 h after surgery and became enlarged 12 h after surgery (Fig. 2A and B). At 24 h after surgery, a stable black thrombus developed, exhibiting a hardened vascular wall and venous stenosis (Fig. 2C). B-Scan ultrasonography demonstrated that the thrombus size and vessel diameter in the model group were significantly increased starting 6 h post-surgery compared with the control group in what appeared to be a time-dependent manner (all P<0.05; Table I; Fig. 2D-I). Additionally, a 2.5-fold higher expression of P-selectin was observed in the model group compared with the control group (P<0.01; Fig. 2J).

The inhibitory effects of rPSGL-Ig on the formation of PVT were evaluated by B-scan ultrasonography. Thrombus size and vessel diameter were increased in the model group (Table I). Intervention with rPSGL-Ig significantly inhibited PVT formation by lowering the thrombus size and vessel diameter at 6–24 h post-surgery compared with the model group (all P<0.05).

Histopathological changes in the portal vein, central hepatic vein and vasa intestinae tenuis in rats of different groups were evaluated by HE staining. A normal structure of endangium with flattened endothelial cells and oriented typical media smooth muscle cells in portal vein was presented in control group (Fig. 3A and B). As shown in Fig. 3C and D, the portal vein in the model group exhibited a damaged endangium, exfoliated endothelial cells and thickened media smooth muscle cells. A thrombus containing an evident fibrin network, platelet trabeculae, adipocytes and leukocytes was observed in the portal vein of the model group. Following intervention with rPSGL-Ig, thrombus formation in the portal vein was markedly inhibited, revealing an imperceptible fibrin network and platelet trabeculae, and reduced adipocytes (Fig. 3E and F). Additionally, erythrocytes were not observed in the central hepatic veins of the control group, which also presented with a compact venous wall structure without cellular swelling (Fig. 4A and B). However, an expanded venous wall, deposited erythrocytes and slightly swollen liver cells were observed in the central hepatic veins of the model group (Fig. 4C and D). Although blood extravasation was also observed in the rPSGL-Ig group, other histopathological changes were markedly relieved (Fig. 4E and F). Furthermore, erythrocytes were not present and normal venous structure with compact intestinal wall structures was observed in the vasa intestinae tenuis of the control group (Fig. 5A and B). Similar histopathological changes were observed in the vasa intestinae tenuis of the model and rPSGL-Ig groups (Fig. 5C-F).

TEM was performed to evaluate histopathological changes in the portal veins in rats of different groups. As shown in Fig. 6A-C, a normal ultrastructure of the portal vein was observed in the control group with structurally complete granulocytes and vascular endothelial cells. In the model group, massive red blood cell deposition along with destruction of the surrounding histiocytic structure was observed 6-h post-surgery (Fig. 6D). The surrounding histiocytic structure was destroyed and necrosed 12 and 24-h post-surgery, respectively, with numerous cell fragments and a small number of granulocytes remaining (Fig. 6E and F). Following treatment with rPSGL-Ig, there were many deposited erythrocytes and slight thrombosis in the portal veins at 6-h post-surgery, with partial destruction of surrounding veins and histiocytic cell structures (Fig. 6G). There were no obvious changes at 12-h post-surgery (Fig. 6H), however, a reduction in the number of erythrocytes deposited and the lack of thrombus in the portal vein, as well as intact surrounding veins and histiocytic cell structures was observed at 24-h post-surgery following treatment with rPSGL-Ig (Fig. 6I).

Evaluation of the thrombolytic effect of rPSGL-Ig revealed a decreased thrombus length (Table II), while the level of P-selectin mRNA was significantly upregulated compared with the control group (P<0.01; Fig. 7A) at a dose of 8 mg/kg. Thus, 8 mg/kg rPSGL-Ig was used to detect thrombolytic effects. No significant change in thrombus size in the rPSGL-Ig group was observed, while thrombus size decreased gradually in the URO group with time (Fig. 7B and C). However, the saline group showed an increasing trend. Thrombus size was significantly decreased in the URO and rPSGL-Ig groups compared with the saline group at 24 and 48-h post-surgery (all P<0.05). The best thrombolytic effect was observed in the URO group followed by the rPSGL-Ig group, while the saline group showed the lowest effects. Therefore, rPSGL-Ig intervention exhibited a thrombolytic effect on PVT rats, however it was less effective than URO.