Effects of Caffeine Treatment on Hepatopulmonary Syndrome in Biliary Cirrhotic Rats

Hepatopulmonary syndrome (HPS) is a lethal complication of cirrhosis characterized by hypoxia and overt intrapulmonary shunting. In this study, we investigated the effect of caffeine in rats with common bile duct ligation (CBDL)-induced liver cirrhosis and HPS. CBDL rats were randomly allocated to receive caffeine or vehicle for 14 days. On the 28th day after CBDL, mortality rate, hemodynamics, liver, and renal biochemistry parameters and arterial blood gas analysis were evaluated. Lung and liver were dissected for the evaluation of inflammation, angiogenesis and protein expressions. In another series with parallel groups, the intrapulmonary shunting was determined. Caffeine significantly reduced portal pressure (caffeine vs. control: 10.0 ± 3.7 vs. 17.0 ± 8.1 mmHg, p < 0.05) in CBDL rats. The mortality rate, mean arterial pressure, biochemistry data and hypoxia were similar between caffeine-treated and control groups. Caffeine alleviated liver fibrosis and intrahepatic angiogenesis but intrapulmonary inflammation and angiogenesis were not ameliorated. The hepatic VEGF/Rho-A protein expressions were down-regulated but the pulmonary inflammation- and angiogenesis-related protein expressions were not significantly altered by caffeine. Caffeine did not reduce the intrapulmonary shunting, either. Caffeine has been shown to significantly improve liver fibrosis, intrahepatic angiogenesis and portal hypertension in cirrhotic rats, however, it does not ameliorate HPS.


Introduction
Coffee, a worldwide consumed beverage, has caught attention for its potential benefits in liver diseases [1]. Increased coffee consumption is associated with the reduction of liver fibrosis in patients with chronic liver disease and the incidence of hepatocellular carcinoma [2,3]. Caffeine (3,7-dihydro-1,3,7-trimethyl-1H-purine-2,6-dione) is the major compound in coffee and many energy drinks. The anti-angiogenesis effect of caffeine has been noticed [4]. In addition, the anti-inflammatory effects of caffeine have been reported in rats challenged with lipopolysaccharide [5]. In our previous study, we found that caffeine reduced portal pressure and ameliorated portal-systemic shunting, mesenteric angiogenesis, and liver fibrosis in cirrhotic rats [6].
Hepatopulmonary syndrome (HPS) is characterized by hypoxia in patients with chronic liver diseases, especially in those with cirrhosis [7]. Emerging data show that the prevalence of HPS ranges from 5 to 32% in cirrhotic patients who need liver transplantation [8]. The prognosis of HPS is ominous, evidenced by a median survival of 24 months and a five-year survival rate of 23% in cirrhotic patients developing HPS [9]. HPS is characterized by overt intrapulmonary shunting due to vasodilatation and abnormal angiogenesis [10,11]. Intrapulmonary shunting is the main condition causing arterial hypoxia in HPS [12]: Unoxygenated blood can escape through the pulmonary shunting directly into the systemic circulation without gas exchange. The severity of arterial hypoxemia is related to the extent of ventilation-perfusion mismatch, intrapulmonary shunting and diffusion impairment [13]. Abnormal pulmonary angiogenesis plays a crucial role in the development and maintenance of HPS. Zhang et al. reported that pathological pulmonary angiogenesis and enhancement of vascular endothelial growth factor (VEGF) production by intrapulmonary monocytes were associated with the development of HPS in experimental animals [10]. Our previous study also showed that sorafenib alleviated experimental HPS by attenuating intrapulmonary angiogenesis through down-regulating the VEGF/VEGF-receptor 2 pathway [14]. In addition to abnormal angiogenesis, the initiation of HPS comes from overwhelming intrapulmonary inflammation. The chemokine fractalkine (CX 3 CL 1 ) expression and signaling are increased in HPS rats, which contributes to pulmonary intravascular monocyte accumulation, angiogenesis and the development of experimental HPS [15]. Moreover, Thenappan et al. found that experimental HPS resulted from the intrapulmonary accumulation of CD68-positive macrophages, and the depletion of macrophages might exert a therapeutic potential [16]. Consistent with those findings, we previously reported that rosuvastatin alleviated experimental HPS through the inhibition of pulmonary inflammatory angiogenesis, which was related to the down-regulation of the tumor necrosis factor-α (TNF-α)/nuclear factor kappa B (NFκB) and VEGF/Rho-A pathways [17].
The common bile duct ligation (CBDL) murine model is an established animal model of liver cirrhosis and HPS [10,[14][15][16][17]. Targeting the mechanism of HPS involving abnormal intrapulmonary angiogenesis and inflammation, in this study, we aimed to investigate the effects of caffeine on HPS in rats with CBDL-induced liver cirrhosis and the underlying mechanisms.

Mortality Rates of Caffeine-and Vehicle-Treated CBDL Rats
There was no significant difference in mortality rates between caffeine-and vehicle-treated (control) CBDL rats (control vs. caffeine: 25% (3/12) vs. 33.3% (4/12), p > 0.05). Table 1 displays body weight, hemodynamic change, liver and renal biochemistry parameters of control (n = 9) and caffeine-treated (n = 8) CBDL rats. In our previous report [17], CBDL rats had significantly higher portal pressure, elevated total bilirubin (TB), aspartate aminotransferase (AST), alanine aminotransferase (ALT), decreased partial pressure of oxygen (PaO 2 ) and increased alveolar-arterial oxygen gradient (AaPO 2 ) compared to the sham-operated rats, indicating the typical presentation of liver cirrhosis and HPS (see Supplementary Table S1). In the present study, body weight and heart rate were not significantly different between the caffeine-treated and control CBDL rats. Caffeine significantly decreased portal pressure (control vs. caffeine: 17.0 ± 8.1 vs. 10.0 ± 3.7 mmHg, p < 0.05). The plasma levels of creatinine, TB, AST, ALT were not significantly influenced by caffeine. The PaO 2 , partial pressure of carbon dioxide (PaCO 2 ) and AaPO 2 in the arterial blood gas analysis were not significantly different either.

Histopathological Change and Immunochemical Staining of Liver
The hepatic hematoxylin and eosin (H&E) staining of CBDL rats showed mononuclear cells infiltration, ballooning change of hepatocytes and destruction of the lobular structure, indicating the inflammatory change of the livers. Sirius red staining revealed the obvious fibrosis of the livers (stained in red), which was significantly attenuated by caffeine. The livers of the control CBDL rats had many CD31-positive staining cells (brown color), which was also attenuated by caffeine ( Figure 1A).

Discussion
We had previously reported that both prophylactic caffeine treatment (started from the first day of CBDL) and therapeutic caffeine treatment (started from the 14th day post-CBDL) ameliorated liver fibrosis and portal hypertension [6]. Consistent with those findings, the current study also found that therapeutic caffeine treatment with the same dose for two weeks ameliorated liver fibrosis, alleviated intrahepatic angiogenesis and reduced portal pressure in rats with CBDL-induced cirrhosis. The mean arterial pressure was not significantly altered by caffeine, implicating that caffeine did not exert adverse effects on systemic hemodynamics. Although caffeine ameliorated liver injury as evidenced by decreased ALT, AST and TB levels in rats with thioacetamide-or carbon tetrachloride-induced liver damage [18,19], our present and previous studies [6] show that the plasma levels of ALT, AST and TB were not influenced by caffeine in CBDL rats. This might be ascribed to the severe and relatively non-modifiable bile stasis and inflammation induced by common bile duct ligation and section in CBDL model. Nevertheless, caffeine still ameliorated the liver fibrosis and portal hypertension in CBDL rats with biliary cirrhosis. The anti-inflammatory impact of caffeine on various types of liver damage, indeed, awaits further investigation. On the other hand, the intrapulmonary inflammation and angiogenesis were not reversed by caffeine. The hypoxia and intrapulmonary shunting were not ameliorated either. These findings suggest that caffeine, although capable of alleviating portal hypertension and liver fibrosis, is not able to mitigate HPS in CBDL rats.
Caffeine is an adenosine A 1 and A 2A receptor antagonist [20]. Adenosine is a potent endogenous regulator of inflammation and tissue repair. Chan et al. showed that caffeine reversed hepatic fibrosis through inhibition of adenosine receptors in cirrhotic mice [21]. In the present study, we found that caffeine significantly ameliorated liver fibrosis and reduced portal pressure in cirrhotic rats, supporting the benefit of using caffeine treatment in cirrhotic patients.
Emerging evidence shows that the elevated adenosine level observed in ischemic tissue contributes to hypoxia-induced angiogenesis [22]. Extracellular adenosine activates A 2A receptors, which stimulates the release of VEGF from the parenchymal cells, followed by endothelial cells proliferation and migration [23]. Through antagonizing the adenosine receptors, caffeine attenuates angiogenesis. In the current study, we demonstrated that caffeine reduced the extent of hepatic CD31-positive staining cells, indicating the attenuation of intrahepatic angiogenesis. Furthermore, caffeine attenuated the hepatic VEGF/Rho-A kinase protein expressions, which are important mediators modulating angiogenesis. Accordingly, our findings suggest that caffeine ameliorates abnormal intrahepatic angiogenesis through, at least in part, inhibition of VEGF/Rho-A kinase pathway. Nevertheless, caffeine did not reduce intrapulmonary vWF-positive staining cells and pulmonary VEGF protein expression. In addition, the intrapulmonary shunts were not influenced by caffeine. According to the discrepant results from lungs and livers, our data indicate that caffeine-related anti-angiogenesis effect is organ-specific in cirrhosis.
In addition to anti-angiogenesis effects, caffeine has been reported to exert beneficial effects on pulmonary inflammation as well as airway remodeling [24]. In a large US cohort study, the inverse association of coffee consumption was observed for deaths from heart disease, chronic respiratory diseases, diabetes, pneumonia and influenza [25]. The researchers postulated that coffee may reduce mortality risk by affecting inflammation, lung function, insulin sensitivity, and depression. The anti-inflammatory effect of caffeine varies with treatment protocols and doses [26]. Li et al. found that chronic caffeine treatment for two weeks or acute high doses (50 mg/kg, 30 min before acute lung injury) significantly attenuated lung edema, hemorrhage, neutrophil recruitment and the inflammatory cytokine expression in mice with acute lung injury. In contrast, acute caffeine treatment at relatively low doses (5, 15, 30 mg/kg) enhanced pulmonary inflammation and damage [26]. However, the present study applying high doses of caffeine (50 mg/kg) to cirrhotic rats showed that the pulmonary polymorphonuclear cells and CD68-positive staining macrophages were not attenuated, indicating a lack of therapeutic effect on pulmonary inflammation in cirrhosis. Since CD68-positive staining macrophages participate actively in initiating pulmonary inflammation and developing HPS [16,17,27], this might be a crucial factor responsible for the failure of caffeine to treat HPS.
The coffee or caffeine consumption has many other effects on the liver and lung. We summarize these emerging findings in the

References Experimental Model Major Findings
Wiltberger et al. [28] Human Coffee consumption was associated with a decreased risk of hepatocellular carcinoma recurrence and increased survival following orthotopic liver transplantation.

Nishitsuji et al. [29] Mice
Caffeine would affect the gut dysbiosis and the disrupted plasma short-chain fatty acid profile, then subsequently prevented nonalcoholic steatohepatitis.
Veronese et al. [30] Human The consumption of coffee was not associated with liver steatosis either nonalcoholic fatty liver disease or alcoholic fatty liver disease.

Kennedy et al. [31] Human
Increased consumption of caffeinated and decaffeinated coffee is associated with reduced risk of hepatocellular carcinoma, including in pre-existing liver disease.

Arauz et al. [31] Rats
Coffee prevented liver cirrhosis by attenuating the oxidant process, blocking hepatic stellate cell activation, and down-regulated the profibrotic molecules.
Wijarnpreecha et al. [33] Human Coffee consumption showed a decreased risk of advanced liver fibrosis and inflammation among hepatitis C-infected patients.

Lammert et al. [34] Human
Coffee consumption was lower among patients with primary sclerosing cholangitis, but no primary biliary cholangitis, compared with controls.
Barcelos et al. [35] Rats Caffeine modified the hepatic responses associated to exercise-induced oxidative stress of trained rats.

Sinha et al. [36] Mice
The caffeine's lipolytic action was through autophage in mammalian liver and it was a potent stimulator of hepatic autophagic flux.

Li et al. [37] Cells
The caffeine-enhanced autophagic flux in hepatic stellate cell (HSC) was stimulated by endoplasmic reticulum stress, which further weakened HSC viability via the induction of apoptosis.

Ikeda et al. [38] Human
There was an inverse association between coffee consumption and elevated aminotransferase in men, and it was more evident in those with high alcohol consumption and in those with low body mass index.
Ong et al. [39] Human Caffeine intake does not affect liver stiffness in chronic hepatitis B-infected patients.
Kurozawa et al. [40] Human An inverse association between coffee consumption and HCC mortality was found in a large cohort study in Japan.

Ohta et al. [41] Rats
A high dose of caffeine (100 mg/kg) completely blocked both liver damage and proinflammatory cytokine responses through an A2AR-independent mechanism.
In conclusion, caffeine ameliorated liver fibrosis and reduced portal pressure without adversely affecting systemic hemodynamics in cirrhotic rats. However, caffeine did not ameliorate HPS. Further study of longer treatment durations, prophylactic caffeine treatment applied at the beginning of liver injury or co-treatment with other anti-inflammatory drugs, such as rosuvastatin or TNF-α blockade agents, might be useful to address this issue.  In conclusion, caffeine ameliorated liver fibrosis and reduced portal pressure without adversely affecting systemic hemodynamics in cirrhotic rats. However, caffeine did not ameliorate HPS. Further study of longer treatment durations, prophylactic caffeine treatment applied at the beginning of liver injury or co-treatment with other anti-inflammatory drugs, such as rosuvastatin or TNF-α blockade agents, might be useful to address this issue.  Table 3. Emerging evidence of coffee or caffeine treatment on various lung diseases.

References Experimental Model Major Findings
Amaral et al. [42] Mice Caffeine enhanced the frequency and number of parenchymal CD4+ T-cell and stimulated immune function in severe tuberculosis.
Doyle et al. [43] Human Caffeine reduced the incidence of brochopulmonary dysplasia in the new-born period and improved pulmonary function at follow-up.
Jing et al. [44] Rats Early caffeine treatment could protect immature lungs from hyperoxia-induced lung injury.

Fehrholz et al. [24] Cells
Glucocorticoid had adverse effects on long-term remodeling by induction of connective tissue growth factor in lung cells, however, co-treatment with caffeine attenuated connective tissue growth factor expression and promoting restoration of lung homeostatsis.
Tatler et al. [45] Cells Caffeine owned anti-fibrotic capacities for the epithelial cells and fibroblasts in the lung.
Chou et al. [46] Rats Caffeine could mitigate lung inflammation induced by ischemia-reperfusion of the lower limbs.
Guertin et al. [47] Human Coffee drinking was positively associated with lung cancer.
Wang et al. [48] Cells Caffeine administration increased the cisplatin-induced lung cancer killings and cellular apoptosis.

Li et al. [26] Mice
Caffeine either enhanced lung damage by antagonizing A2A receptor or exerted protection against lung damage via A2A receptor-independent mechanisms, depending on the timing of exposure and dose of administration.
Lu et al. [49] Cells Caffeine and tea polyphenos inhibited the progression of lung adenoma to adenocarcinoma.
Yoder et al. [50] Baboons Early caffeine treatments were associated with better lung function in immature baboons.
On the other hand, we drew the postulated mechanisms regarding the effect of caffeine on the liver ( Figure 3) and lung ( Figure 4).  In conclusion, caffeine ameliorated liver fibrosis and reduced portal pressure without adversely affecting systemic hemodynamics in cirrhotic rats. However, caffeine did not ameliorate HPS. Further study of longer treatment durations, prophylactic caffeine treatment applied at the beginning of liver injury or co-treatment with other anti-inflammatory drugs, such as rosuvastatin or TNF-α blockade agents, might be useful to address this issue.

Animal Model
Male Sprague-Dawley rats weighing 240-270 g at the time of surgery were used. Rats were housed in plastic cages and allowed free access to food and water. Common bile duct ligation (CBDL) was applied as an animal model of liver cirrhosis with HPS according to previous reports [10,11,14]. The rats were fasted for 12 h before the operation. Under Zoletil 50 (Tiletamine + Zolezepam) anesthesia (0.8 mg/kg, intramuscularly), CBDL was performed. A high yield of secondary biliary cirrhosis was noted four weeks after the ligation [51]. To avoid coagulation defects, CBDL rats received weekly vitamin K injections (50 µg/kg intramuscularly). In all experiments, the principles of laboratory animal care (NIH publication no. 86-23, revised 1985) were followed. This study was approved by the Taipei Veterans General Hospital Animal Committee (IACUC 2017-060) (21/6/2017).

Experimental Design
In humans with moderate to high caffeine intake, doses approximate 10-20 mg/kg/day [52]. Based on a report from Ohta et al. that the amount of caffeine in a single cup of coffee (100 mg) is estimated to be equivalent to a 5 mg/kg dose in rodents, 5 and 15 mg/kg can be considered low-dose caffeine treatment, while 50 mg/kg caffeine administration is a high-dose caffeine treatment in mouse model [41]. In the present study, we used caffeine 50 mg/kg/ day as a high-dose caffeine treatment to investigate the beneficial effect of cirrhotic rats. CBDL rats were fed via oral gavage with 50 mg/kg/day caffeine or vehicle (distilled water 5 mL/day, control group) from the 15th to 28th day after CBDL. On the 28th day after CBDL, the body weight, mortality rate, and hemodynamic data were measured. Arterial blood was collected for gas analysis. Venous blood was collected for determination of TB, AST, ALT, and creatinine. Furthermore, lung and liver were dissected for histological examination and immunohistochemical staining to evaluate liver fibrosis, intrahepatic and intrapulmonary angiogenesis and inflammation. Protein expressions were analyzed with Western blotting. In another series with parallel groups, the intrapulmonary shunting was determined with the color microsphere method.

Systemic and Portal Hemodynamic Measurements
The right internal carotid artery was cannulated with a PE-50 catheter that was connected to a Spectramed DTX transducer (Spectramed Inc., Oxnard, CA, USA). Continuous recordings of mean arterial pressure were performed on a multi-channel recorder (model RS 3400, Gould Inc., Cupertino, CA, USA). The external zero reference was placed at the level of the mid-portion of the rat. The abdomen was then opened with a midline incision, and the superior mesenteric vein was cannulated with a PE-50 catheter connected to a Spectramed DTX transducer. The abdominal cavity was closed, and the PP was recorded on a Gould model RS 3400 recorder.

Biochemistry and Blood Gas Analysis
The femoral artery and vein were cannulated with PE-50 catheters, one day before the experiments. Both catheters were fixed over the back and flushed with a heparin-contained solution. Blood was withdrawn from the femoral vein to determine plasma concentrations of ALT, AST, TB, and creatinine. Arterial blood was withdrawn from the femoral artery for blood gas analysis. Arterial gas exchange was evaluated by the alveolar-arterial oxygen gradient, which was calculated as 150-(PaCO 2 /0.8)-PaO 2 .

Histopathological and Immunochemical Staining
Liver and lung were dissected free and fixed in 10% neutral buffered formalin. The sections were stained with H&E and examined by light microscopy. The liver sections were stained with Sirius red for 8 min according to the manufacturer's instructions (Polysciences Inc., Warrington, PA, USA) to determine the extent of collagen deposition [53]. The immunochemical staining was performed with anti-CD68 antibody (diluted 1:200, ab31630, Abcam Cambridge, UK) to detect pulmonary CD68-positive macrophages, indicating the severity of intrapulmonary inflammation. To determine the intrahepatic angiogenesis, immunochemical staining with anti-CD31 antibody (1:200, Serotec and Pharmingen, San Diego, CA, USA) was performed [6]. Anti-vWF antibody (1:100, MCA127T, AbD Serotec, UK) was applied to survey the intrapulmonary angiogenesis [54]. The immunochemical staining was followed by the biotinylated anti-mouse IgG (H+L) (Vector Laboratories, Burlingame, CA, USA) 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 color at the antigen site.

Western Blot Analysis for Protein Expressions
Lung and liver were immediately frozen in liquid nitrogen and stored at −80 • C until required. Protein was extracted by pulverization in grinder with liquid nitrogen, followed by 1 mL of lysis buffer (phosphate-buffered solution containing 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS)) and 0.05% protease inhibitor cocktail solution (Roche Diagnostics GmbH, Penzberg, Germany) for each 100 mg powdered sample. Protein concentration was determined for each sample by the Bradford method [55]. An aliquot of 20-40 µg protein from each sample that dissolved in sample buffer (63 mmol/l of Tris-HCL, pH 6.8, containing 2% SDS, 10% glycerol, 5% 2-mercaptoethanol, and 0.005% bomphenol blue) and 10 µg positive control were separated on denaturing SDS-10% polyacrylamide gels by electrophoresis (Mini-PROTEAN ® 3 Cell, Bio-Rad Laboratories, Hercules, CA, USA). Pre-stained proteins markers (SDS-PAGE Standards, Bio-Rad) were used for molecular weight determination. Proteins were then transferred to a polyvinylidene difluoride membrane (Immum-Blot TM PVDF Membrane, Bio-Rad) by a semi-dry electroblotting system (Trans-Blot ® SD Semi-dry Electrophoretic Transfer Cell, Bio-Rad) for 1.5 h at 4 • C. To block non-specific binding, membranes were blocked for 30 min with 3% non-fat dry milk in TBS-T, pH 7.4 (25 mmol/l Tris base-137 mmol/l NaCl-2.7 mmol/l KCL-1% Tween 20). Blots were incubated with the primary antibody anti-ERK, -phosphorylated ERK (1:3000, Millipore Corporation), diluted with 3% non-fat dry milk in TBS-T] for 90 min at room temperature and washed. Then the blots were incubated for 90 min with the secondary antibody (horseradish peroxidase-conjugated goat anti-mouse IgG antibody, diluted with 3% non-fat dry milk in TBS-T, Sigma Chemical Co., St. Louis, MO, USA) and washed. The specific proteins were detected by enhanced chemiluminescence (Immobilon Western Chemiluminescent HRP Substrate, Merk Millipore Co., Billerica, MA, USA). With a computer-assisted video densitometer and digitalized system (BioSpectrum ® 600 Imaging System, Ultra-Violet Products Ltd., Upland, CA, USA), the blots were scanned and photographed, after which the signal intensity (integral volume) of the appropriate band was analyzed.

Intrapulmonary Shunting Analysis
The degree of intrapulmonary shunting was determined using the color microsphere method [14]. Before microsphere injection, indwelling PE-50 femoral arterial and venous catheters were placed. On the day of measurement, 2.5 × 10 6 custom mixed and counted cross-linked polystyrene-divinylbenzene microspheres labeled red (size range 6.5-10 µm; Interactive Medical Technologies, Los Angeles, CA) in 0.20 mL of sterile PBS were injected over 2-4 s through the femoral vein catheter, which was immediately flushed with 0.2 mL of sterile PBS over 2-4 s. A reference blood sample was withdrawn from the femoral arterial catheter started 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 PBS. Samples of beads before venous injection and reference blood samples were coded. The numbers of colored microspheres in the blood preparations were determined using a hemacytometer counting slide having a known volume. Because the diameter of normal pulmonary microvasculature was less than 5µm, most part of microsphere beads injected from femoral vein would be trapped in the lung. However, microspheres would pass through the intrapulmonary shunting vessels in the HPS animal and were collected in the reference blood from the femoral artery. Total numbers of microspheres passing through the pulmonary microcirculation were calculated as reference blood sample microspheres per mL times estimated blood volume. Estimated blood volume (mL) was calculated as: 0.06 x body wt (g) + 0.77 [56]. Intrapulmonary shunting (%) was calculated as: (total number of microspheres passing through the pulmonary microcirculation/total beads injected into the venous circulation) x 100.

Drugs
Caffeine was purchased from Sigma Chemical Co. (St. Louis, MO, USA).

Data Analysis
The results are expressed as mean ± standard deviation. Statistical analyses were performed with Student's t test. The survival rate was analyzed with log-rank test. Results were considered statistically significant at a p-value less than 0.05.