The hepatopulmonary syndrome (HPS) is characterized by hypoxia and increased intrapulmonary shunts in cirrhotic patients. Emerging evidence showed promising results of treating HPS by abolishment of intrapulmonary inflammation and angiogenesis. Rosuvastatin is a kind of 3-hydroxy-methyl-3-glutamyl coenzyme A reductase inhibitor. In addition to lipid-lowering effects, it has anti-inflammation and anti-angiogenesis properties. We postulated that rosuvastatin treatment can ameliorate HPS. Common bile duct ligation (CBDL) was applied in an experimental HPS animal model. CBDL rats received 2-week rosuvastatin (20 mg/kg/day) treatments from the fifteenth day after operation. The haemodynamic data, blood gas analysis, liver biochemistries, tumour necrosis factor-α (TNF-α) and vascular endothelial growth factor (VEGF) were examined after rosuvastatin treatment. The liver and lung tissues were dissected for histopathological studies and protein analyses. In the parallel groups, intrapulmonary shunts were determined. The haemodynamic and liver biochemistries were not changed after rosuvastatin treatment in CBDL rats, but the alveolar-arterial oxygen pressure gradient was significantly decreased, implying that HPS-induced hypoxia was reversed after rosuvastatin treatment. In addition, rosuvastatin treatment reduced intrapulmonary shunts and plasma levels of VEGF and TNF-α. Besides, the intrapulmonary protein expression of nuclear factor kappa B (NF-κB), VEGF receptor (VEGFR)-1,2 and Rho-associated A kinase were significantly down-regulated and the intrapulmonary angiogenesis was ameliorated. We concluded that rosuvastatin alleviates experimental HPS through blockade of pulmonary inflammatory angiogenesis via TNF-α/NF-κB and VEGF/Rho-associated A kinase pathways down-regulation.
HPS is a lethal complication of liver cirrhosis. The prognosis of HPS is ominous with no effective medical therapy except liver transplantation available in the past.
However, we found that a 2-week rosuvastatin treatment improves HPS in biliary cirrhotic rats through, at least in part, dual pathways of anti-inflammation by inhibition of TNF-α/NF-κB/iNOS and anti-angiogenesis by attenuation of VEGF/VEGFR-1,2/Rho-A kinase. Besides, rosuvastatin did not affect liver biochemistry, it altered neither systemic nor portal haemodynamics.
This has clinical implications since rosuvastatin may work in patients with marked fibrosis or cirrhosis.
Hepatopulmonary syndrome (HPS) is a lethal complication of liver cirrhosis . Data from liver-transplantation centres show that the prevalence of HPS ranges from 5% to 32% in cirrhotic patients . The prognosis for HPS is ominous, as indicated by a median survival of 24 months and a 5-year survival rate of 23% in cirrhotic patients developing HPS . Three important components of HPS are hypoxia with increased alveolar arterial oxygen pressure gradient (AaPO2), intrapulmonary vasodilatation with increasing shunts and chronic liver disease with portal hypertension. In the past, no effective medical therapy is documented for HPS patients except liver transplantation. However, evidence shows promising results of medical treatment in HPS through blockade of abnormal pulmonary inflammation and angiogenesis [4,5].
More and more studies have pointed out that monocyte accumulation in the pulmonary vessels plays an important role in triggering HPS [4,6]. Zhang et al.  demonstrated that CX3CL1 chemokines directly mediate monocytes adhesion and activate vascular endothelial growth factor (VEGF) and angiogenesis, suggesting that chemokine alterations resulting from liver disease may contribute to the development of HPS. In addition, bacterial translocation in cirrhosis can lead to increased macrophage adherence to pulmonary vessels and further induces HPS . Pentoxifylline reduces inflammation of lung and enhances arterial oxygenation via tumour necrosis factor α (TNF-α) and leukotriene inhibition in an experimental HPS model . Liu et al.  reported that neutralization of TNF-α, using monoclonal antibody, successfully improved HPS in cirrhotic rats. Therefore, abnormal intrapulmonary inflammation plays a key role in HPS development and alleviation of overwhelming inflammation may modulate the development of HPS.
The 3-hydroxy-methyl-3-glutamyl coenzyme A reductase inhibitors, commonly called statins, have been widely used to lower cholesterol in patients with hyperlipidaemia. In addition to lipid-lowering effect, there is growing evidence that statins can modulate portal hypertension and reduce inflammation of liver [10,11]. Statins have potent antioxidant and anti-inflammation properties. Cai et al.  reported that lovastatin treatment improved liver function after extensive hepatic resection of rats. In a double-blind randomized controlled trial, a dose of 20 mg/day of simvastatin could lower the portal pressure (PP) without deleterious effects in systemic haemodynamics in patients with liver cirrhosis , indicating that statin therapy may be beneficial in patients with liver cirrhosis and portal hypertension.
Rosuvastatin is a kind of hydrophilic statin which has been documented to significantly reduce the incidence of major cardiovascular events in healthy persons because of its anti-inflammatory propensities . Nezasa et al.  reported that rosuvastatin is taken up by hepatic cells more selectively and more efficiently than pravastatin and simvastatin. Besides, a study from Sironi et al.  showed that rosuvastatin provided end-organ protection in stroke-prone rats by anti-inflammatory effects. These findings provide evidence that rosuvastatin is a powerfully anti-inflammatory agent for patients with chronic liver disease.
Focusing on abnormal intrapulmonary inflammations of HPS, we postulated that rosuvastatin therapy could improve HPS via attenuating pulmonary inflammation in cirrhotic animals. In the present study, we evaluated the effects of rosuvastatin treatment on sham-operated and common bile duct ligated-rats to investigate its capacity to treat HPS. In addition, the mechanism was surveyed.
MATERIALS AND METHODS
Male Sprague–Dawley rats were used for experiments. Under ketamine anaesthesia (100 mg/kg, intramuscularly), rats with secondary biliary cirrhosis were induced by common bile duct ligation (CBDL). A high yield of secondary biliary cirrhosis was noted 4 weeks after CBDL . In the past, several studies have used the CBDL rat as an experimental model for investigating the mechanism of HPS [4–6,8,9]. The control group received sham operation. In all experiments, the authors adhered to the American Physiological Society guiding principles for the care and use of laboratory animals (NIH publication no. 86-23, revised 1985). The present study was approved by the Taipei Veteran General Hospital Animal Committee (IACUC 2011-106).
The first series: effects of rosuvastatin in sham-operated rats
Rosuvastatin (20 mg/kg/day; n=11) or vehicle (0.9% sodium chloride; n=10) was administered orally by gavage for 2 weeks in sham-operated rats. Treatments began from 2 weeks after sham-operation. The haemodynamic data were measured and blood samples were collected at the end of treatments. In the two parallel groups (n=6 compared with 7), colour microsphere study was conducted to determine intrapulmonary shunts.
The second series: effects of rosuvastatin in CBDL rats
Rosuvastatin (20 mg/kg/day; n=11) or vehicle (0.9% sodium chloride; n=11) was administered orally by gavage for 2 weeks in CBDL rats. Treatments began from 2 weeks after CBDL. The haemodynamic data were measured and blood samples were collected at the end of treatments. Lungs and livers were dissected for histopathological and protein expression studies including inducible nitric oxide synthase (iNOS), constitutive nitric oxide synthase (cNOS), VEGF receptor 1, 2 (VEGFR-1, VEGFR-2), nuclear factor kappa B (NF-κB), nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor α (IκBα), phosphoinositide 3-kinase (PI3K), Rho-associated (Rho-A) kinase, extracellular signal-regulated kinase 1/2 (ERK 1/2), Akt (serine/threonine kinase) and nuclear protein P65. In the two parallel groups (n=7 compared with 8), colour microsphere study was conducted to determine intrapulmonary shunts.
Systemic and portal haemodynamic measurements
The right internal carotid artery was cannulated with a PE-50 catheter that was connected to a Spectramed DTX transducer (Spectramed Inc.). Continuous recordings of mean arterial pressure (MAP) and heart rate (HR) were performed on a multi-channel recorder (model RS 3400, Gould Inc.). The mesenteric vein was cannulated to record the PP.
Body weights, biochemistry and blood gas analysis
The body weights of rats in differently treated groups were measured before CBDL or sham operation and 4 weeks after operations. The femoral artery and vein of rat were cannulated with PE-50 catheters before one day of haemodynamic experiments. Both catheters were fixed over the back and flushed with a solution of heparin. Blood was withdrawn from femoral vein for determining plasma concentration of alanine transaminase (ALT), aspartate aminotransferase (AST), total bilirubin, TNF-α and VEGF levels. Besides, the arterial blood was withdrawn from femoral artery for arterial blood gas analysis. Arterial blood gas was analysed and the AaPO2 was calculated as 150-(PCO2/0.8)-PO2.
Determination of plasma VEGF and TNF-α levels
The plasma levels of VEGF and TNF-α were measured by using commercially available ELISA kits (R&D Systems) according to the manufacturer's instructions.
Western analysis for protein expression
Lung tissues were frozen in liquid nitrogen and stored at −80°C until required. The protein extracts were incubated with the first antibody [anti-cNOS, -iNOS (1:1000; Millipore Corporation); anti-phosphorylated cNOS (1:1000; Cell Signaling Technology), anti-phosphorylated iNOS (1:1000; Abcam plc); anti-PI3K (1:1000; Cell Signaling Technology); anti-NF-κB (1:200; Santa Cruz Biotechnology); anti-IκBα (1:10000; Abcam plc); anti-phosphorylated IκBα (1:1000; Cell Signaling Technology); anti-VEGFR-1 (1:1000; Abcam plc); anti-VEGFR-2 (1:500; Millipore Corporation); anti-RhoA kinase (1:1000; Cell Signaling); anti-Akt (1:500, Cell Signaling Technology); anti-phosphorylated Akt (1:2000, Cell Signaling Technology); anti-ERK, -phosphorylated ERK (1:3000, Millipore Corporation)]. Then the blots were incubated with the secondary antibody (horseradish peroxidase-conjugated goat anti-mouse IgG antibody, Sigma Chemical Co.). With a computer assisted video densitometer and digitalized software (Kodak Digital Science™ ID Image Analysis Software, Eastman Kodak Co.), the blots were scanned, photographed then the signal intensities (integral volume) of the appropriate bands were analysed.
Western analysis for nuclear NF-κB p65 protein expression
Nuclear proteins were extracted using NE-PER nuclear and cytoplasmic extraction reagents (Thermo Scientific) according to the manufacturer's instructions. The protein extracts were transferred to membrane and then incubated with mouse anti-human total and phosphorylated NF-κB p65 antibody [anti-p65 (1:200; Santa Cruz Biotechnology); anti-phosphorylated-p65 (1:1000; Abcam plc)].
Histopathological, immunohistochemical staining and immunofluorescence staining studies of the liver and lung
The liver and lung tissues of rats were dissected free and fixed in 10% formalin solution. The sections were stained with Hematoxylin and Eosin (H&E) and examined by light microscopy. Liver paraffin section was stained with Sirius Red staining kit (Polysciences Inc.) to determine liver fibrosis . The von Willebrand factor (vWF) released by endothelial cells has been routinely used to identify angiogenesis in tissue sections .
The immunohistochemical staining with anti-CD68 (cluster of differentiation 68) antibody (diluted 1:200, Abcam) was performed to determine CD68-positive macrophages in the lung tissue; in addition, anti-vWF antibody (diluted 1:1000, Abcam) was performed to determine intrahepatic angiogenesis. The paraffin sections were heated in a microwave oven for antigen retrieval. Endogenous peroxidase was inhibited using 3% H2O2 in methanol. Immunostaining was performed using the primary antibodies, followed by the biotinylated anti-mouse IgG (H+L; Vector Laboratories) as the second antibody. Detection of biotinylated antibody was performed using the VECTASTAIN®-Elite ABC kit from Vector Laboratories. For the chromogen, DAB (3′-diaminobenzidine tetrahydrochloride) was used, which resulted in a brown colour at the antigen site. Finally, the sections were counterstained with Mayer's haematoxylin and they were covered with mounting medium.
The immunofluorescence staining technique, using anti-vWF antibody, was performed to determine angiogenesis in the lung tissue. Immunofluorescent staining was performed using the primary antibodies against vWF (diluted 1:50, mouse anti-human monoclonal, AbD Serotec). A fluorescent-labelled secondary antibody was used (diluted 1:100, Alexa Fluor® 488 flourescent, Jackson ImmunoResearch Laboratories). Finally, slides were cover slipped with carbonate-buffered glycerol and evaluated with an Olympus AX 80 microscope (Olympus) equipped with fluorescence illumination and digital cameras.
Analysis of intrapulmonary shunts
Intrapulmonary shunts were determined using the colour-microsphere techniques [4,18]. Cross-linked (2.5 × 106) coloured microspheres (size range 6.5–10 μm; Interactive Medical Technologies) were injected through the femoral vein catheter, which were immediately flushed with 0.2 ml of sterile normal saline. A reference blood sample was withdrawn from the femoral arterial catheter beginning at the time of femoral vein injection for a total of 90 s at a constant rate of 1.0 ml/min. The volume removed was replaced with an equal volume of sterile normal saline. Samples of beads before venous injection and reference blood samples were coded. The numbers of coloured microspheres in the blood preparations were determined using a haemacytometer counting slide having a known cell volume. Total numbers of microspheres passing through the pulmonary microcirculation were calculated as reference blood sample microspheres/millilitre × estimated blood volume. The estimated blood volume of each animal was derived from the following formula: blood volume (ml)=0.06 × body weight (g)+0.77 . Intrapulmonary shunting was calculated as an intrapulmonary shunt fraction (%)=(total numbers of microspheres passing through the pulmonary microcirculation/total beads injected into the venous circulation) × 100.
Rosuvastatin was purchased from Pharmaceutic AstraZeneca.
The results are expressed as mean ± S.D. Statistical analyses were performed by using one-way analysis of variance test with post hoc test by Tukey analysis or Student's t test when appropriate. Results are considered statistically significant at P<0.05.
Mortality rates and adverse effects of rosuvastatin
None of the rosuvastatin-treated or control CBDL and sham-operated rats died during the experimental period. In addition, no obviously adverse effects were noted in the rosuvastatin-treated CBDL and sham-operated rats.
Haemodynamic, liver biochemistry and arterial blood gas analysis of sham-operated and CBDL rats receiving rosuvastatin or vehicle treatment (
The baseline body weights were not significantly different among differently treated groups. The body weight gain was significantly lower in rosuvastatin-treated CBDL rats compared with other groups. After CBDL operation, the MAP significantly decreased and PP increased in rats treated with vehicle (CBDL compared with sham: MAP 110±14 compared with 122±14 mmHg; PP 15.9±2.4 compared with 7.5±1.1 mmHg; both P<0.05). The plasma levels of total bilirubin, ALT and AST were also obviously elevated in the CBDL rats compared with sham-operated rats (all P<0.05). Compared with sham-operated rats, CBDL rats had lower partial pressure of oxygen (PaO2) and higher AaPO2 (PaO2 85.4±5.6 compared with 91.9±4.9 mmHg; AaPO2 18.5±4.2 compared with 9.3±2.4 mmHg; both P<0.05). The haemodynamic data and liver biochemistry were not significantly different between rosuvastatin-treated and control CBDL rats. However, the AaPO2 was reduced after rosuvastatin treatment in CBDL rats (AaPO2=12.6±2.2 compared with 18.5±4.2 mmHg; P<0.05). The PaO2 had also a trend to rise after rosuvastatin treatment in CBDL rats (PaO2=90.7±4.3 compared with 85.4±5.6 mmHg P = 0.061). Besides, the plasma levels of VEGF and TNF-α were significantly decreased by rosuvastatin treatment in CBDL rats compared with the control group (VEGF=17.0±2.9 compared with 20.7±4.9 pg/ml; TNF-α=20.1±4.4 compared with 24.9±3.7 pg/ml; both P<0.05).
|.||Sham rats .||CBDL rats .|
|.||Rosuvastatin (n=11) .||Control (n=10) .||Rosuvastatin (n=11) .||Control (n=10) .|
|BW prior (g)||264±20||279±70||260±16||269±12|
|BW post (g)||413±24||410±23||337±27*†||371±27|
|.||Sham rats .||CBDL rats .|
|.||Rosuvastatin (n=11) .||Control (n=10) .||Rosuvastatin (n=11) .||Control (n=10) .|
|BW prior (g)||264±20||279±70||260±16||269±12|
|BW post (g)||413±24||410±23||337±27*†||371±27|
Degree of intrapulmonary shunts of CBDL and sham-operated rats
Figure 1 depicts the degrees of intrapulmonary shunts in differently treated CBDL and sham rats. The intrapulmonary shunts significantly increased in CBDL rats compared with sham-operated rats receiving vehicle treatment (CBDL compared with sham: 21.4±7.5% compared with 8.0±3.0%, P<0.05). Rosuvastatin significantly decreased intrapulmonary shunts of CBDL-induced HPS rats (rosuvastatin compared with controls: 13.2±4.0% compared with 21.4±7.5%, P<0.05). In contrast, rosuvastatin treatment did not influence intrapulmonary shunts in the sham-operated rats (rosuvastatin compared with controls: 9.5±5.3% compared with 8.0±3.0%, P > 0.05).
Intrapulmonary shunts of sham-operated and CBDL rats
The protein expression of phosphorylation of cNOS, iNOS and IκBα, PI3K, NF-κB and nuclear P65 in the pulmonary tissue of CBDL rats treated with rosuvastatin or vehicle
In CBDL rats, the phosphorylated protein level of iNOS was significantly down-regulated after rosuvastatin treatment, but phosphorylated cNOS levels were not significantly different between rosuvastatin-treated and control groups (rosuvastatin compared with control: phosphorylated-iNOS/iNOS: 0.54±0.08 compared with 0.71±0.14; P = 0.028; phosphorylated-cNOS/cNOS: 0.80±0.12 compared with 1.01±0.37, P>0.05; Figure 2A). The levels of PI3K were not significantly different between rosuvastatin-treated and control groups (0.58±0.34 compared with 0.69±0.46; P > 0.05) but the protein level of NF-κB was significantly decreased after rosuvastatin treatment (NF-κB: 0.16±0.17 compared with 0.32±0.15; P = 0.03; Figure 2B). Rosuvastatin significantly down-regulated the nuclear phosphorylated-p65 protein levels and phosphorylation of IκBα protein levels (phosphorylated-p65/p65: 0.66±0.09 compared with 0.81±0.14; phosphorylated-IκBα/IκBα: 0.56±0.28 compared with 1.12±0.42, both P<0.05; Figures 2C and 2D).
Intrapulmonary protein expression of cNOS, iNOS, NF-κB, PI3K, P65 and IκB
The protein expression of VEGFR-1, VEGFR-2, Rho-A kinase and phosphorylated Akt and ERK in the pulmonary tissue of CBDL rats treated with rosuvastatin or vehicle
The protein levels of VEGFR-1 and VEGFR-2 were significantly decreased after rosuvastatin treatment (rosuvastatin compared with control: VEGFR-1=0.09±0.04 compared with 0.31±0.27, P = 0.01; VEGFR-2=0.07±0.04 compared with 0.15±0.10; P = 0.03; Figure 3A). The protein levels of Rho-A kinase were significantly decreased after rosuvastatin treatment (6.25±3.23 compared with 9.34±3.01; P = 0.03; Figure 3B). The phosphorylation of ERK1/2 and Akt were similar between rosuvastatin-treated and control groups (both P > 0.05; Figures 3C and 3D).
Intrapulmonary protein expressions of VEGFR-1, 2, Rho-A, phosphorylated ERK 1/2 and Akt
Effects of rosuvastatin treatment on the histopathological change and angiogenesis of lung in CBDL rats
The pulmonary tissues of sham-operated rats have normal alveolar and endothelial cells scattering around the alveoli (Figure 4A). The inflammatory cells accumulate in the lung tissues of CBDL rats, which causes tissue necrosis and disruption of normal alveoli (Figure 4B). The infiltrative inflammatory cells are significantly reduced after rosuvastatin treatment (Figure 4C). These inflammatory cells, especially prominent are CD68-positive staining macrophages (dark brown colour), are significantly attenuated by rosuvastatin (Figures 4D and 4E), indicating that rosuvastatin decreased CD68-positive macrophage infiltrations in the HPS lung tissues. Besides, rosuvastatin significantly attenuates the vWF-containing cellular immunofluorescence densities in the lung tissue, suggesting the amelioration of pulmonary angiogenesis in CBDL rats (Figures 4F and 4G).
Pulmonary tissues of H&E stains, immunohistochemical and immunofluorescence stains
Effects of rosuvastatin on hepatic fibrosis and angiogenesis in CBDL rats
Under Sirius Red staining, the liver of CBDL rats shows significant fibrotic change (red colour), which was ameliorated by rosuvastatin (Figures 5A and 5B). The vWF-staining images of liver reveal that the extent of vWF-stained endothelial cells (brown colour) is similar between rosuvastatin-treated and control CBDL rats (Figures 5C and 5D), indicating that rosuvastatin did not affect intrahepatic angiogenesis.
Immunohistochemical stains of liver
In the present study, we confirm that CBDL rat is a feasible experimental animal model for investigating HPS. This is consistent with other researchers' reports [4,6,8]. In the present study, CBDL rats had a decrease in MAP accompanied by an increase in PP which indicated the existence of portal hypertension. Profound jaundice and impaired liver biochemistry parameters were noted. The Sirius Red staining also showed significant collagen deposition, typical of fibrotic change of liver in CBDL rats. These findings are consistent with cirrhotic change with hyperdynamic circulation. The decreased PaO2 and increased AaPO2 meant deoxygenation of CBDL rats. In addition, compared with sham-operated rats, CBDL rats had significantly more intrapulmonary shunts. Taken together, these features are compatible with the clinical manifestations of cirrhotic patients with HPS.
Our present novel finding demonstrates that rosuvastatin treatment can improve HPS in the experimental biliary cirrhotic rats. We found that rosuvastatin significantly decreased the AaPO2 level, indicating the improvement of hypoxia of CBDL rats. Besides, the intrapulmonary shunts were obviously ameliorated by rosuvastatin. In addition, rosuvastatin treatment significantly decreased the plasma levels of TNF-α and VEGF, which were accompanied by pulmonary iNOS phosphorylation, NF-κB, VEGFR-1, VEGFR-2 and Rho-A kinase protein expression down-regulation. These data point out that rosuvastatin improves HPS through both anti-inflammation (TNF-α/NF-κB/iNOS) and anti-angiogenesis (VEGF/VEGFR-1,2/Rho-A kinase).
There was no significantly adverse effect or mortality in rosuvastatin-treated rats. The plasma levels of AST were modestly elevated in sham-operated rats and significantly elevated in CBDL rats, which might be attributed to operation and cholestasis-related tissue injury, since levels of ALT, the parameter more specific for liver injury, were not altered. Nonetheless, rosuvastatin did not exacerbate liver injuries in both sham-operated and CBDL rats, indicating the safety of this drug in cirrhotic status. It is worth noting that the body weight gain was significantly lower in the group of CBDL rats treated with rosuvastatin, as compared with the other three groups. The two CBDL groups were similar in terms of survival, infection rate and presence of ascites. Therefore, bias from these parameters is not likely. Furthermore, consistent with our finding, the previous studies revealed that statins reduced body weight gain in experimental animals . Another investigation reported that statins might induce body weight loss via decreased appetite . Based on these findings, we may infer that rosuvastatin reduced body weight gain in cirrhotic rats via decreased oral intake.
The MAP of sham-operated and CBDL rats was not influenced by rosuvastatin in the present study. Surveys on portal hypertension and cirrhosis, either in human or in rats, acute or chronic administration with various doses of statins did not affect MAP in previous studies [22,23]. Although a meta-analysis of randomized controlled trials showed that statins have a blood pressure-lowering effect , the meta-analysis was restricted to patients with normal or high systolic blood pressure. However, in cirrhotic patients with hyperdynamic circulation and lower MAP, statins do not affect blood pressure  and are thus safe in clinical setting.
Rosuvastatin did not influence PP of CBDL rats in the present study, either. In a previous study, we found that simvastatin treatment significantly lowered PP in portal hypertensive rats . Atorvastatin also reduced PP in CBDL-induced cirrhotic rats, which was associated with a reduction in intrahepatic resistance . However, a contrary report showed that chronic simvastatin treatment for 4 weeks did not modify PP in CBDL rats . These divergent results might be ascribed to different doses, kinds and durations of statin treatments.
Rosuvastatin decreased plasma levels of TNF-α and down-regulated pulmonary NF-κB protein expression as well, implicating its anti-inflammatory effect on cirrhotic HPS rats. The effect of statins on acute lung injury and pulmonary fibrosis has been documented. Pravastatin inhibited neutrophil accumulation and reduced TNF-α concentrations in bronchoalveolar lavage fluid of bleomycin-induced lung injury in mice . In addition, atorvastatin attenuated inflammation and pulmonary vascular remodelling in hypercholesterolaemic rabbits . A 16-week simvastatin treatment decreased macrophage accumulations and reduced TNF-α levels in bronchial alveolar lavage fluid in rats . In addition to animal studies, pre-treatment with simvastatin reduced lipopolysaccharide-induced systemic and pulmonary inflammation in healthy volunteers . The anti-inflammatory effect of statins was mediated by cytokine synthesis reduction through inhibition of the mevalonic cascade followed by RhoA activation in the human bronchial epithelial cells . This finding is in agreement with our data that rosuvastatin treatment abolished the pulmonary expression of RhoA kinase. Thenappan et al.  have postulated that HPS may be derived from intrapulmonary accumulation of CD68-positive macrophages, which up-regulate the NF-κB and VEGF pathways which then initiates vasodilatation and angiogenesis in the lung. Our data totally support the crucial findings. We found that rosuvastatin attenuated the intrapulmonary CD68-positive macrophages accumulation and down-regulated inflammatory and angiogenesis pathways which then ameliorated the experimental HPS.
The effect of statins on angiogenesis has been extensively studied. Angiogenesis is an important factor in portal hypertension and liver cirrhosis. Fernandez et al.  found that combined VEGF and platelet-derived growth factor blockade reversed portal hypertension in portal hypertensive rats. The driving forces of pathological angiogenesis come from hypoxia and inflammation. Interestingly, statins have dual effects, i.e. either pro-angiogenesis or anti-angiogenesis . At a low concentration, statins induce angiogenesis through the PI3K and Akt pathways [35,36]. However, at a high concentration, statins exert anti-angiogenesis effect through an apoptotic mechanism . In addition, the effect of statins also varies with different cell type. At the same concentration, statins down-regulate VEGF in human smooth muscle cells and micro-vascular endothelial cells but up-regulate VEGF in macro-vascular endothelial cells . Furthermore, Zhu et al.  have demonstrated that statins have disparate effects on angiogenesis during hypoxia and inflammation. They found that statin treatments can enhance angiogenesis in human umbilical vein endothelial cells during hypoxia but conversely decrease it during inflammation. In the present study, we found that rosuvastatin down-regulated pulmonary RhoA kinase/VEGFR-1,2 protein expression and decreased intrapulmonary angiogenesis evidenced by decreased intrapulmonary vWF containing cells. Similar to our findings, other researchers also found that statins interfered with angiogenesis via inhibition of RhoA kinase [40,41].
NF-κB regulates the expression of TNF-α to modulate the generation of inflammatory cytokines . NF-κB is present in the cytoplasm as a heterotrimer consisting of p50, p65 and IκBα subunits. Upon activation of the complex, phosphorylation and degradation of IκBα expose nuclear localization signals on the p50–p65 complex, leading to nuclear translocation and binding to specific regulated sequences in the DNA, thus controlling gene transcription . We found that rosuvastatin treatments down regulated the phosphorylation of IkBα and decreased nuclear protein levels of P65 which indicated the attenuation of NF-kB activation on the pulmonary tissue of CBDL rats. Increased production of TNF-α and activation of NF-κB are key mediators of inflammation. As shown in the present study, the plasma levels of TNF-α and pulmonary NF-κB expression were significantly inhibited and HPS was reversed by rosuvastatin, suggesting that inflammation may play an important role in HPS. In addition, the role of NF-κB in angiogenesis has aroused much attention. Kiriakidis et al.  investigated the role of NF-κB in the induction of VEGF of human macrophages and found that, like TNF-α, VEGF is also regulated by NF-κB. The blockade of endogenous TNF-α activity resulted in a significant inhibition of VEGF release in macrophages. In view of the role of NF-κB in inflammation and cell survival, NF-κB inhibition may contribute to anti-inflammation and anti-angiogenesis effects of statins . Such evidence demonstrates that the NF-κB-mediated inflammatory pathway participates in the development of angiogenesis, which can be reversed by statins.
One may concern is whether the benefits of rosuvastatin in HPS are secondary to improvement of liver condition or are they are unique to lung. Lots of studies have indicated the beneficial effects of statins in liver fibrosis and cirrhosis [11,22–23]. In the present study, we found that rosuvastatin treatment decreased fibrotic change of liver but the intrahepatic angiogenesis was not affected. Furthermore, PP was not significantly influenced after rosuvastatin treatment, so the beneficial effects of rosuvastatin in HPS may be ascribed mainly to pulmonary anti-inflammation and anti-angiogenesis properties.
The influences of statins on liver biochemistry have been controversial. Olteanu et al.  reported that early administration of rosuvastatin at the beginning of CBDL for 7 days increased hepatic oxidative stress and exacerbated liver inflammation of rats. However, Demirbilek et al.  showed that a 10-day fluvastatin treatment initiated after 3 days of CBDL had beneficial effects on hepatic inflammation and cell apoptosis. In the current study, rosuvastatin treatment started at the third week post CBDL, at which time fibrotic or cirrhotic stage of liver is more prominent than inflammation. This is of clinical relevance since rosuvastatin may work in patients with marked fibrosis or cirrhosis.
In conclusion, a 2-week rosuvastatin treatment improves HPS in biliary cirrhotic rats without alteration of systemic and portal haemodynamics. Besides, rosuvastatin did not affect liver biochemistry. The mechanism of rosuvastatin therapy in HPS may be through, at least in part, dual pathways of anti-inflammation by inhibition of TNF-α/NF-κB/iNOS and anti-angiogenesis by attenuations of VEGF/VEGFR-1,2/Rho-A kinase. Although the extrapolation of experimental data to clinical use must be done cautiously, the beneficial effect of rosuvastatin in HPS through blockade of inflammatory angiogenesis may be a new way to treat cirrhotic patients with HPS and necessitates further clinical trials.
Ching-Chih Chang and Fa-Yauh Lee designed the present study and wrote the manuscript. Chiao-Lin Chuang and Wen-Shin Lee assisted with data analysis. Han-Chieh Lin, Sun-Sang Wang and Shou-Dong Lee conceived the project. Hsian-Guey Hsieh provided the technical assistances. Hui-Chun Huang conceived the project and finally revised the manuscript. All the authors contributed to the final version of the manuscript.
We gratefully acknowledge Chiu-Chu Wu and Yi-Chou Chen for their excellent technical assistances.
This work was supported by the Taipei Veterans General Hospital [grant number V101B-030]; and the Szu-Zuan Research Foundation of Internal Medicine [grant number 103002].
alveolar arterial oxygen pressure gradient
common bile duct ligation
constitutive nitric oxide synthase
- ERK 1/2
extracellular signal-regulated kinase 1/2
inducible nitric oxide synthase
nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor α
mean arterial pressure
nuclear factor kappa B
partial pressure of oxygen
- Rho-A kinase
tumour necrosis factor α
vascular endothelial growth factor
von Willebrand factor
These authors contributed equally to this paper.