Salvianolic Acid B Reducing Portal Hypertension Depends on Macrophages in Isolated Portal Perfused Rat Livers with Chronic Hepatitis

This study is aimed to investigate the effects of Sal B on portal hypertension (PH). PH with chronic hepatitis was induced by carbon tetrachloride (CCl4) in rats. The model was confirmed with elevated portal pressures and increased serum CD163 levels. The inducible nitric oxide synthase (iNOS) or heme oxygenase-1 (HO-1) in portal triads was assessed. The isolated portal perfused rat liver (IPPRL) was performed at d 0, d 28, d 56 , and d 84 in the progression of chronic hepatitis. After constricting with phenylephrine, the portal veins were relaxed with Sal B. The EC50 of Sal B for relaxing portal veins was −2.04 × 10−9, 7.28 × 10−11, 1.52 × 10−11, and 8.44 × 10−11 mol/L at d 0, d 28, d 56, and d 84, respectively. More macrophages infiltrated in portal triads and expressed more iNOS or HO-1 as PH advanced. The areas under the curve (AUCs) of Sal B for reducing PH were positively correlated with the levels of iNOS or HO-1 in portal triads, and so did with serum CD163 levels. Sal B reduces PH in IPPRL with chronic hepatitis, via promoting portal relaxation due to macrophage-originated NO or CO in portal triads, partly at least.


Introduction
Portal hypertension (PH) is a common complication in the patients with advanced chronic hepatitis [1]. The increased hepatic vascular resistance and portal hyperemia are involved in the reversible pathogenesis as the potent therapy targets [2].
Salvianolic acid B (Sal B) is a molecule from the root of Salvia miltiorrhiza (Danshen), which is a traditional Chinese medicine widely used for cardiovascular diseases [3]. Sal B is effective for liver fibrosis and PH in patients [4] or animals [5]. In the endothelin −1 -induced PH rats, Sal B could inhibit the constriction of hepatic stellate cells [5]. However, our previous study indicated that Sal B constrict portal veins of the isolated portal perfused rat livers (IPPRLs) at physiological status [6]. The underlying mechanisms of Sal B for PH remain unclear.
It is reported that nitric oxide (NO) and carbon monoxide (CO) play key roles in the pathogenesis of PH [7]. Both signal molecules directly relax portal veins through upregulation of cGMP via guanylate cyclase [8]. NO from endothelial NO synthase (eNOS) aggravates PH through systemic hyperemia [9], and inducible NO synthase (iNOS) exacerbates PH by producing peroxynitrite (ONOO − ) [10]. It has been reported that the reduced NO bioavailability is involved in the increased hepatic vascular resistance [11]. There is an increase of superoxide release by NADPH oxidase in liver with chronic hepatitis [12] and an overproduction of iNOS from macrophages [10]. The iNOS-derived NO reacts with superoxide, leading to ONOO − formation, with a decrease in NO bioavailability [10]. Heme oxygenase-1(HO-1) is a rate-limiting enzyme catalyzing heme to CO, iron, and biliverdin. Biliverdin is then converted to bilirubin, which acts as a highly effective antioxidant and free radical scavenger against oxidation [13]. HO-1 also showed hepatoprotection against ischemia-reperfusion injury, endotoxemia, hyperoxia-induced hepatic injury, and immunemediated apoptotic liver damage [14]. Furthermore, HO-1/CO activation downregulates the inflammatory response by blocking the formation of ONOO − from iNOS [13]. While the ONOO − induces HO-1 protein expression but mediating its inactivation [15].
Sal B has an effect on [3] production of NO or CO from activated macrophages [16] under inflammatory cytokines [17]. In addition, Sal B could protect endothelia from the oxidation by blocking PI3K/Akt signal pathway [18]. Therefore, Sal B was proposed to rescue NO bioavailability or to maintain CO potency from the macrophage at portal triads in advanced chronic hepatitis.
The purpose of present study is to investigate the effects of Sal B on PH in IPPRL with chronic hepatitis and to analyze further the NO or/and CO signals through the relationship between the Sal B potency and the existed iNOS or HO-1 from the macrophages in portal triads.

Reagents.
Carbon tetrachloride (CCl 4 ), olive oil, and heparin sodium were purchased from Sinopharm Chemical Reagent Company. Acetylcholine chloride and phenylephrine hydrochloride were obtained from Sigma (USA). Salvianolic acid B (purity >99%) was purchased from Shanghai Institute of Materia Medica, Chinese Academy of Sciences.

Animals.
Thirty two six-week-old male Wistar rats (180-200 g) were purchased from Animal Centre of the Chinese Academy of Medical Sciences. All rats were kept under a 12 h/12 h light/dark cycle, temperature (25.0 ± 0.2 • C), and humidity (45 ± 2%) controlled SPF environment. The rats were fed standard rodent pellets and allowed free access to filtered water. All experiment procedures were performed in accordance with the Guidelines of Animal Experiments from the Committee of Medical Ethics, National Health Department of China.

Induction of Portal Hypertension with Chronic Hepatitis.
PH with chronic hepatitis was induced by CCl 4 in rat as described previously (Figure 1) [19]. Rats were injected subcutaneously with a mixture of 40% (v/v) CCl 4 in olive oil (3 mL/kg) two times a week for 0, 28, 56, and 84 days, respectively [20,21]; olive oil was the vehicle for age-matched control. Eighty four hours after the last CCl 4 injection, rats were anesthetized with a subcutaneous injection of sodium pentobarbital (50 mg/kg). A midline incision was made to open abdominal cavity, and ascitic samples were collected and quantified as described preciously [22]. The exuded liquid ratios were calculated as exuded liquid weight/body weight ×100. The portal pressure in vivo was recorded. The blood sample was collected for analyzing the serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), and albumin (Alb) levels by biochemistry and CD163 levels by immunoassay. Then the hepatic artery, portal vein, and hepatic vein were canalized [20,21]. The liver, spleen, and kidneys were harvested, and the organ indexes were calculated as organ weight/body weight ×100.

CD163
Immunoassay. The blood samples were centrifuged at 1200 g for 10 min at 4 • C, and serum was stored at −80 • C until the assays. Serum CD163 levels was measured using enzyme-linked immunosorbent assay kits (R&D Systems, Wiesbaden, Germany) following the manufacturer's guidelines.

Histological and Morphometry
2.5.1. Histological Quantification. Formalin-fixed, paraffinembedded liver sections were cut at a 6 µm thickness and then stained with Masson's trichrome (Masson) [19]. Images were obtained using NanoZoomer Digital Pathology system (Hamamatsu, Japan). The collagen density was quantified using Image ProPlus analysis system 7.0.1 (no. 41N70000-60555, Media Cybernetics, USA) at 100 × magnification. The data were expressed as the one-ten thousandth of collagen (the ratio of collagen area per total analyzed field area ×1000%). Values are expressed as the average of ten fields taken from each section.  [20,21]. Phenylephrine hydrochloride was determined as 1.69 × 10 −10 , 2.64 × 10 −10 , 5.82 × 10 −10 , and 8.24 × 10 −10 mol/L, respectively, to constrict portal veins [21]. After the phenylephrine constriction, Sal B (10 −13 -10 −7 mol/L) was added into the recirculating perfusate. Dose-response curves were obtained from the Sal B concentrations and the changed percentage of the perfused pressure from the baseline of phenylephrine constriction.

Statistical Analysis.
All data are expressed as mean ± S.E.M. Comparisons between groups were performed using Student's t-test or Mann-Whitney. Significant differences were established at the 0.05 level. The equation, the EC 50 with its 95% confidence intervals of Sal B, and the area under the curve (AUC) of Sal B were analyzed using GraphPad Prism 4 (GraphPad Software). The EC 50 of Sal B (y) was regressed with the durations (0, 28, 56, and 84 days) and serum CD163 levels (x) in the progression of chronic hepatitis, and the AUC of Sal B (y) was regressed with the amounts of existed iNOS or HO-1 from immunohistochemical staining and the serum CD163 levels (x).

General Characterization of Rats.
The model of rat PH was confirmed by ascite levels, organ index, and serum biomarker levels ( Figure 2). The exuded liquid ratios significantly elevated from d 0 to d 84 (P < 0.01) as the progression of chronic hepatitis ( Figure 2). Hepatic indexes were the lowest at d 0 , the highest at d 28 , and reduced at d 56 and d 84 gradually ( Figure 2). The splenic or renal indexes increased gradually from d 0 to d 84 ( Figure 2). Serum ALT and AST levels increased from d 0 to d 28 (Figure 3). (n i = 8). a P < 0.05, aa P < 0.01 compared with rats at d 0 ; b P < 0.05 and bb P < 0.01 compared with rats at d 28 ; c P < 0.05 and cc P < 0.01 compared with rats at d 56 .

Serum CD163 Levels.
CD163 is a biomarker of the activated macrophages in PH. The serum CD163 levels were increased gradually from d 0 to d 84 (P < 0.01) as the procession of chronic hepatitis ( Figure 4). observed, and the hepatic sinusoid was severely narrowed without obvious collagen ( Figure 5(b)). At d 56 , the hepatic fibrosis was observed and the collagen increased and mainly deposited in lobules. The relived enlarged hepatic cords led to the hepatic sinusoid widen obviously. Some deposited collagen in interlobular had extended into and separated lobules incompletely, thus the directions of blood flow were not changed in hepatic sinusoid ( Figure 5(c)). At d 84 , the hepatic cirrhosis was evident. The lobules were completely destructed by deposited collagen and the formation of pseudolobules was observed, so the directions of blood flow were completely changed in hepatic sinusoid ( Figure 5(d)).

Collagen
Ratio. Quantification of Masson staining by morphometry analysis showed that collagen ratios were increased along with the progression of chronic hepatitis ( Figure 6).

iNOS Cellular Localization.
The iNOS positive cells were the hepatocytes and scattered stellates in the lobules of the normal rats at d 0 ( Figure 5(e)). The iNOS expression was reduced in the scattered hepatocytes and mainly observed in stellates cells in the lobules of the rats with chronic hepatitis at d 28 ( Figure 5(f)). The expression of iNOS was completely quenched in the hepatocytes; the positive cells were the macrophages in the portal triads and the stellates in the lobules at d 56 ( Figure 5(g)). The main thick positive cells were the macrophages in the fibrous interval pseudolobules around vessels and the stellates with thin granules in the lobules at d 84 ( Figure 5(h)).

iNOS Quantification.
The iNOS-IHC OD per volume ( Figure 6) in the portal triads of the rats with chronic hepatitis was significantly increased at d 28 (2-fold), d 56 (1.5fold), and d 84 (3-fold) compared with that at d 0 , respectively, (P < 0.01); these were decreased at d 56 and increased at d 84 compared with that at d 28 , respectively (P < 0.01); so did increased that at d 84 compared with that at d 56 (P < 0.01).

HO-1 Cellular Localization.
The main HO-1 positive cells were the hepatocytes near the central vein and the scattered stellates in the lobules, but the hepatocytes next to portal triads were absolutely negative in the normal rats at d 0 ( Figure 5(i)). Besides of the thinner granules in the hepatocytes, the main positive staining cells were the stellates in the lobules at d 28 ( Figure 5(j)). The thin granules have completely disappeared in the hepatocytes, while the positive cells were the macrophages in portal triads and the stellates in the lobules at d 56 ( Figure 5(k)). The main thick positive cells were the macrophages in the fibrous intervals out pseudolobules at d 84 ( Figure 5(l)).

HO-1 Cellular Quantification.
The total HO-1-IHC OD per volume ( Figure 6) in the rats with chronic hepatitis was significantly increased at d 28 (1.6-fold), d 56 (2-fold), and d 84 (3-fold) compared with that at d 0 , respectively (P < 0.01); these at d 56 and d 84 significantly increased compared with that at d 28 , respectively (P < 0.01); so did that at d 84 compared with that at d 56 (P < 0.01).

Discussion
It was demonstrated in the present study that Sal B relaxed portal veins in IPPRLs of CCl 4 -induced chronic hepatitis. Its mechanisms are related to the inhibition of oxidative stress from macrophages and the increase of NO bioavailability or CO potency in portal triads. Sal B is the most active antioxidant extracted from Danshen and has obvious effects for liver fibrosis, chronic hepatitis, or PH in clinic [4].The mechanisms responsible for the protective effects of Sal B in PH remain unclear. It has been reported that the portal resistance is mainly located at the terminal portal venules (TPV) in portal triads [23]. The activated macrophages release vasoactive substances concomitantly and increase the perfusion resistance [24]. Accordingly, we have previously demonstrated that the macrophages out lobules express more iNOS, produce more NO, and generate ONOO − to further reduce NO bioavailability and aggregate PH [22]. The HO-1/CO activation decreases iNOS expression, enhances antioxidative effect, and upregulates extracellular superoxide dismutase (ecSOD) [13]. The local ecSOD could scavenge superoxide and block ONOO − generation [11]. Therefore, macrophagederived NO or CO in portal triads was considered as the most effective target. Sal B, a molecule from medical plants [3] for PH [4-6, 20, 21, 25], has benefits to elevate NO bioavailability and to maintain CO potency [17], and to inhibit oxidation [18], especially in macrophages [16]. We hypothesized that Sal B relaxes portal veins through antioxidative effects on NO or CO potency.
In present study, PH model in rat was replicated by CCl 4induced chronic hepatitis and was confirmed with increased portal pressure and pathological changes, such as the peritoneal exudation, enlarged hepatic indexes, serum hepatic biomarker levels, collagen deposition, and pseudolobule formation [1,2]. The serum CD163 levels increased as the procession of chronic hepatitis, indicating the activation of macrophages, which was consistent with the PH patients [24]. We found Sal B reduced PH as a candidate from a medical plant for PH patients. Sal B increased the portal pressure of the IPPRLs at physiological status and reduced the PH of the IPPRLs at chronic hepatitis status in this study. EC 50 of Sal B relaxation was positively correlated with the duration of CCl 4 -induced chronic hepatitis, indicating the action of Sal B which was pathological dependent. Our results demonstrated that increased iNOS or HO-1 levels in the macrophages infiltrated in portal triads are involved in the mechanism of Sal B relaxation. The existed levels of iNOS or HO-1 in lobules disappeared gradually, these in portal triads strengthened continuously along with the progression of CCl 4 -induced chronic hepatitis, especially in the infiltrated macrophages. We also reported here that iNOS and HO-1 levels in portal triads are correlated positively with the AUCs of Sal B for reducing PH.
The IPPRL was used in this study to evaluate the effect of Sal B on PH. The hepatic artery was ligated to ensure that the portal resistance originated mainly from the smooth muscle cells in terminal portal venule (TPV) and the sphincter-like endothelia at hepatic sinusoid inlets [23]. In PH rodents, the TPVs were the major resistance in portal microcirculation without enough collateral (like pre-TPV) or sinusoidal (post-TPV) networks to compensate a blood pressure increase [26]. Furthermore, the infiltrated activated macrophages in portal triads were next to TPVs in the rats with oxidative chronic hepatitis (Figure 7 inserted micrographs). Sal B relived endothelin-induced elevated portal pressure in physiological rats [27] or mice [28]; these did not agree with the data in this research that Sal B increased further phenylephrine-induced elevated portal pressure in the rats without chronic inflammation [6,20,21]. It suggested that the macrophages infiltrated in portal triads being the indirect cellular targets of Sal B to reduce PH in the rats with chronic hepatitis. There were at least four possible pathways for Sal B decreasing PH from oxidative chronic hepatitis (Figure 7). (1) NO signal: Sal B inhibited oxidative stress of activated macrophages [17], blocked ONOO − generation [3], rescued iNOS activity from the inactivation by nitrate modification [10], and consequently increased local NO level to relax the TPV. Especially the AUC of Sal B for reducing PH correlated positively with the existed level of iNOS from the macrophages [8][9][10][11]23]. Sal B relaxed indirectly portal vein via restoring NO bioavailability [29].
(2) CO signal: Sal B increased the expression of HO-1 from activated macrophages [16] and elevated local CO level to dilate TPV [8,17]. Especially the AUC of Sal B for reducing PH correlated positively with the existed level of HO-1 from the macrophages in the portal triads. Meanwhile, HO-1-derived bilirubin directly inhibits NADPH oxidase and increases ecSOD and then decreases superoxide production and ONOO − formation [13]. (3) EcSOD protection: Sal B might indirectly upregulate ecSOD expression, which converts superoxide to hydrogen peroxide and blocks ONOO − generation from NO [11,30]. Then the hydrogen peroxide could enhance iNOS, HO-1, and ecSOD expression itself to against the vicious cycle in PH. (4) Calcium signal: being considered as a cardiovascular protective agent [31],  Figure 7: Salvianolic acid B indirectly reduced portal hypertension via NO/CO potency. In portal triads, the smooth muscle cells in terminal portal venules and the sphincter-like endothelia at hepatic sinusoid inlets were relaxed by both of the nitric oxide (NO) from inducible NO synthase (iNOS) or endothelial NO synthase (eNOS) and the carbon monoxide (CO) from heme oxygenase-1 (HO-1). The resistance of intrahepatic portal vein was mainly originated from both cells, since the expanding portal vein in portal triads indicated its backward location, and the narrowed hepatic sinusoids between hepatic cords did its forward location. It was the infiltrated macrophages that make salvianolic acid B (Sal B) effect switching, from its constricting in physiological status into its relaxing terminal portal venules in chronic hepatitis. The mechanisms of Sal B might be exploited from its elevating NO bioavailability and CO potency. Sal B acted on TPV endothelia and smooth muscles. On human endothelia, Sal B activated transcription factor 4 or 6, consequently regulated upwards glucose-regulated protein 78, to protect the cellular damage from oxidative stress [30]; Sal B suppressed JAK/STAT1 activation in endothelia to relive vessel inflammation [32]. On human vascular smooth muscle, Sal B limited calcium channel to decrease Ca 2+ influx [33]. It's a challenge that the exact mechanism of Sal B actions from the physiological constriction switches to pathological relaxation. The clinical aspects of heme oxygenase hinted its pharmacological actions in pathological status [13]. Further research on Sal B mechanisms might go on the way of systems biology [34].
Sal B for reducing PH might be used to explain the actions of medical plants in Chinese prescription for the ascitic patients with chronic hepatitis. It is an interesting clue to discover more effective candidates depending on the macrophage iNOS or HO-1 in portal triads, at least partly. Consequently, Sal B or its derivative might be exploited as a candidate to increase NO bioavailability or CO potency, especially from free radical damages in inflammatory diseases.