Regenerative and fibrotic pathways in canine hepatic portosystemic shunt and portal vein hypoplasia, new models for clinical hepatocyte growth factor treatment

No fibrosis was seen in liver biopsies of CPSS dogs. In the PPVH dogs histological examination revealed slight portal fibrosis in one dog, slight to moderate portal fibrosis associated with slight to moderate centrolobular fibrosis in four dogs, and marked portal fibrosis with biliary proliferation in three dogs. The control dogs showed a normal liver without fibrosis. Examples of histological examination of CPSS and PPVH are included as Figures 1A and 1B, respectively.

One of the main in vivo events during regeneration and growth is the signaling via phosphorylation of the HGF receptor c-MET. Q-PCR analysis revealed that HGF mRNA levels in both CPSS and PPVH were decreased three-fold in comparison with healthy dogs (Figure 2). Moreover, the c-MET levels in CPSS and PPVH were significantly decreased (two- and three-fold, respectively). The levels of the mRNAs for TGFα (proliferation) were decreased six-fold in both CPSS and PPVH. The serine-protease HGF activator mRNA was doubled in dogs with CPSS. In contrast, it was halved in dogs with PPVH. Although not significantly in dogs with CPSS, the cell-cycle inhibitor p27kip mRNA was decreased in both conditions.

The fibrosis signaling pathway is activated through bindings of the active TGF-β1 dimer to the heteromeric type-I and type-II serine/threonine receptor kinases. As shown in Figure 3, TGF-β1 mRNA levels were increased two-fold in dogs with PPVH, whereas the levels in dogs with CPSS were not changed significantly. The receptor type-I, was induced in both liver diseases but only significantly in PPVH. Receptor type-II was increased in both CPSS and PPVH (4- and 5-fold, respectively), indicating an increased binding capacity. One of the proteolytic enzymes involved in activation of TGF-β1 is urokinase plasminogen activator (uPA). The uPA mRNA level was decreased two-fold in dogs with CPSS and, in contrast, doubled in dogs with PPVH.

We measured three well-known basic apoptotic components of which two are pro-apoptotic (caspase-3 and Fas ligand) and one is anti-apoptotic (Bcl-2). Figure 4 shows that pro-apoptotic mediator Fas ligand was severely inhibited in both dogs with CPSS and in dogs with PPVH (14- and 8-fold, respectively). Moreover, caspase-3 was halved in both CPSS and PPVH. On the other hand, no induction of the anti-apoptotic Bcl-2 was seen in dogs with CPSS, whereas Bcl-2 in dogs with PPVH was doubled. The mechanisms underlying progressive fibrosis are unknown, but fibrosis and hypoxia could have been a fibrogenic stimulus. Hypoxia coordinately up-regulates matrix production and hypoxia induced factor 1 alpha (HIF1α) [18]. These direct hypoxic effects on the expression of genes involved in fibrogenesis was shown in our dogs with PPVH which indeed had elevated levels of HIF1α.

The analysis of ECM expression was performed on three collagens (I, III and IV) and one glycoprotein (fibronectin). Interstitial collagens types I and III are the most commonly found collagens, collagen type IV is a basal membrane collagen. In Figure 5, collagen I was shown to be significantly increased in PPVH (two-fold), whereas CPSS was unchanged. Collagen III and IV were not significantly changed in both groups. Fibronectin showed to be halved in the CPSS group where PPVH remained normal.

PKB plays a pivotal role in liver regeneration and growth upon activation of the c-MET-HGF signaling pathway [10]. Western blot analysis of HGF showed an immunoreactive band at 82 kDa with no apparent quantitative differences (Figure 6A). Non-phosphorylated c-MET was detected in all samples, where it was present as an immunoreactive band of 145 kDa. Results showed a decrease in the amount of c-MET in both diseases. On the other hand, the anti-phosphorylated c-MET antibody showed an immunoreactive band in all samples with no apparent quantitative differences. Non-phosphorylated PKB was detected in all samples, where it was present as a single band of 60 kDa. The anti-phosphorylated PKB antibody showed an immunoreactive band in all samples. Two immunoreactive bands at 42 and 44 kDa representing the MAP kinase ERK1/2 showed to be equally present at the protein level between the diseased groups and healthy controls. Interestingly, this also applied for the phosphorylated form where no apparent quantitative differences were found. The 80 kDa STAT3 protein showed a similar result with no apparent quantitative differences in the non-phosphorylated form; however, the STAT3 protein seemed to be somewhat less phosphorylated at the serine 727 residue in the PPVH group. TGF-β1 exerts its actions through complex intracellular signaling pathways. All downstream signaling routes following binding of an active TGF-β1 to its receptors type-I and II elicit phosphorylation of Smad2. TGF-β1 was seen in all diseases as a single band of 25 kDa under non-denaturing conditions (Figure 6B). Interestingly, the amount of TGF-β1 was induced in PPVH compared to CPSS and controls. Non-phosphorylated Smad2 was detected in all samples, where it was present as a single band of 58 kDa, with no apparent changes in quantity. Also interestingly, the anti-phosphorylated Smad2 antibody showed a slight band in CPSS whereas in PPVH a phosphorylated Smad2 is clearly present. Moreover, anti-collagen I showed an increase in the amount of protein in PPVH compared to CPSS and healthy controls, all together emphasizing the differences in fibrosis between CPSS and PPVH. Although reduced in the CPSS and PPVH group, inactive or uncleaved caspase-3 was detected in all samples (Figure 6C), where it was present as a single band of 34 kDa. Finally, the processed forms of 20 and 13 kDa showed to be increased in CPSS and PPVH towards healthy controls.

All samples are obtained from different dog breeds appearing in the clinic with spontaneous diseases. Samples were randomly chosen and aimed to encompass different dog-breeds and both sexes in each group. The procedures were approved by the Ethical Committee as required under Dutch legislation.

The congenital portosystemic shunt (n = 11 dogs) and primary portal vein hypoplasia group (n = 8 dogs) were compared with a group of healthy dogs (n = 11 dogs). The inclusion criteria for CPSS were increased fasting plasma ammonia concentration, abnormal ammonia tolerance test (peak ammonia ≥ 150 μmol/l plasma) and ultrasonographic visualization of a small liver and a congenital portosystemic shunt with a diameter as wide as the portal vein trunk. The presence of the shunt was further confirmed with surgery, during which a wedge liver biopsy was taken and immediately put in liquid nitrogen and stored at -70°C, until analysis. In CPSS there is no portal hypertension. The inclusion criteria for PPVH were the visualization of a small liver with ultrasonography, presence of multiple small acquired portosystemic collaterals due to portal hypertension, and an abnormal ammonia tolerance test (peak ammonia ≥ 150 μmol/l plasma). Liver tissue of dogs with PPVH was obtained under local anaesthesia by ultrasound-guided biopsy with a true cut 16G biopsy needle. Two biopsies were immediately immersed in liquid nitrogen, and stored at -70°C, until analysis. The healthy control dogs were age-matched, and had AP, ALT, and fasting bile acids in plasma within the reference range. Ultrasonographically the control dog livers had a normal size, shape, and structure, and there were no histological abnormalities in stained histological sections.

Liver samples were fixed in 10% buffered formalin and routinely embedded in paraffin. Sections (4 μm) were stained with haematoxylin-eosin, the Van Gieson stain, and the reticulin stain according to Gordon and Sweet. Histologically, the presence of fibrosis was evaluated semi-quantitatively (absent, slight, moderate, or marked) as well as with respect to its localization. Fibrosis scoring was performed according to Scheuer, a defined scoring method for fibrosis in hepatitis. The slides were independently examined by one certified veterinary pathologist.

Total cellular RNA was isolated from each frozen canine liver tissue in duplicate, using the RNeasy Mini Kit (Qiagen, Leusden, The Netherlands) according to the manufacturer's instructions. The RNA samples were treated with Dnase-I (Qiagen Rnase-free DNase kit). In total 3 μg of RNA was incubated with poly(dT) primers at 42°C for 45 min, in a 60 μl reaction volume, using the Reverse Transcription System from Promega (Promega Benelux, Leiden, The Netherlands).

Q-PCR based on the high affinity double-stranded DNA-binding dye SYBR^® green I (BMA, Rockland, ME) was performed in triplicate in a spectrofluorimetric thermal iCycler^® (BioRad, Veenendaal, The Netherlands). Data were collected and analyzed with the provided application software. For each Q-PCR, 2 μl (of the 2 times diluted stock) of cDNA was used in a reaction volume of 50 μl containing 1× manufacturer's buffer, 2 mM MgCl₂, 0.5 × SYBR^® green I, 200 μM dNTP's, 20 pmol of both primers, 1.25 units of AmpliTaq Gold (Applied Biosystems, Nieuwerkerk a/d IJssel, The Netherlands), on 96-well iCycler iQ plates (BioRad). Primer pairs, depicted in Table 1, were designed using PrimerSelect software (DNASTAR Inc., Madison, WI). All PCR protocols included a 5-minute polymerase activation step and continued for 40 cycles at 95°C denaturation for 20 sec, annealing for 30 sec and elongation at 72°C for 30 sec with a final extension for 5 min at 72°C. Annealing temperatures were optimized at various levels ranging from 56°C till 67°C (Table 1). Melt curves (iCycler, BioRad), agarose gel electrophoresis, and standard sequencing procedures were used to examine each sample for purity and specificity (ABI PRISM 3100 Genetic Analyser, Applied Biosystems). Standard curves constructed by plotting the relative starting amount versus threshold cycles were generated using serial 4-fold dilutions of pooled cDNA fractions from both healthy and diseased liver tissues. The amplification efficiency, E (%) = (10^(1/-s)-1) * 100 (s = slope), of each standard curve was determined and appeared to be > 95%, and < 105%, over a wide dynamic range. For each experimental sample, the amount of the gene of interest, and of the endogenous references glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and hypoxanthine phosphoribosyl transferase (HPRT) were determined from the appropriate standard curve in autonomous experiments. If relative amounts of GAPDH and HPRT were constant for a sample, data were considered valid and the average amount was included in the study (data not shown). Results were normalized according to the average amount of the endogenous references. The normalized values were divided by the normalized values of the calibrator (healthy group) to generate relative expression levels [40].

Used antibodies are described in Table 2. For Western blot analysis 30 mg of liver tissue from at least six samples of each group (n = 6 dogs per group, randomly chosen from original group) were pooled and analyzed. Liver tissues were homogenized in RIPA buffer containing 1% Igepal, 0.6 mM phenylmethylsulfonyl-fluoride, 17 μg/ml aprotinine, and 1 mM sodium-orthovanadate (Sigma chemical Co., Zwijndrecht, The Netherlands). Protein concentrations were obtained using a Lowry-based assay (DC Protein Assay, BioRad). Twenty μg of protein of the supernatant was denatured for 3 min at 95°C and electroforesed on 7.5% Tris-HCl polyacrylamide gels (BioRad) and the proteins were transferred onto Hybond-C Extra Nitrocellulose membranes (Amersham Biosciences Europe, Roosendaal, The Netherlands) using a Mini Trans-Blot^® Cell blot-apparatus (BioRad). Immunodetection was based on an ECL Western blot analysis system, performed according to the manufacturer's instructions (Amersham Biosciences Europe). The membranes were incubated with 4% ECL blocking solution in TBS for 1 hour under gentle shaking. The incubation of the primary antibody was performed at 4°C over-night for all antibodies (see Table 2) in TBS with 0.1% Tween-20 (Boom B.V., Meppel, The Netherlands). After washing, the membranes were incubated with their respective horseradish peroxidase-conjugated secondary antibody (R&D systems, Europe Ltd., Abingdon, UK) at room temperature for 1 h and exposed to Kodak BioMax Light-1 films (Sigma chemical Co.). Densitometric analysis of immunoreactive bands was performed with a Gel Doc 2000 system coupled to the Quantity One 4.3.0 Software (BioRad).