| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
LIVER AND BILIARY TRACT
Departments of 1Hepatology, 2Gastroenterology, and 3Histopathology, University Hospital Gasthuisberg, Leuven, Belgium
Submitted 8 May 2006 ; accepted in final form 22 September 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
portal hypertension; RhoA; protein kinase C; calcium-calmodulin; stress-relaxed matrix contraction; intrahepatic vascular resistance
Since the contributions of these different pathways have not yet been fully elucidated for HSC contraction, we therefore aimed to 1) define the relative role of Ca2+-dependent and -independent pathways in the actomyosin interaction of HSCs (Fig. 1), using a three-dimensional stress-relaxed collagen lattice contraction model; 2) compare the degree of Ca2+ dependency of vascular smooth muscle cells (VSMCs), cardiac myofibroblasts, and activated HSCs; and 3) determine the relevance of these in vitro findings in the increased IHVR associated with thioacetamide (TAA)-induced cirrhosis in the rat.
|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
HSCs were isolated from male Wistar rats, weighing 300–400 g (Animal House, Leuven, Belgium), as approved by the local Ethical Committee on Animal Research and as described previously (3). In brief, following in situ perfusion of the liver with collagenase type IV (Sigma, St. Louis, MO) and pronase E (Merck, Darmstadt, Germany), the resulting cell suspension was fractionated by density gradient centrifugation using Optiprep (Nycomed). Cells were harvested at densities of <1.053 (9% Optiprep) according to Alpini et al. (1). Viability and purity was systematically >95%, as determined by trypan blue exclusion and morphological characterization. Cells were seeded on uncoated plastic culture dishes and cultured in William's E medium supplemented with 10% FCS, 0.6 IU/ml insulin, 2 mM glutamine, and 1% antibiotic-antimycotic solution (Invitrogen, Merelbeke, Belgium). The medium was renewed every 48–72 h. Characterization of rat liver-derived myofibroblast-like cultures, established by culturing enriched HSC fractions on plastic, was performed by staining with anti-
-smooth muscle actin (-SMA), anti-desmin, and anti-synaptophysin, as previously described (3, 4). Experiments were performed between the first and third passages (1:3 split ratio) using three cell cultures from independent isolations. Rat cardiac ventricular myofibroblasts (a gift from Dr. V. Petrov, Department of Molecular and Cardiovascular Research, University of Leuven, Leuven, Belgium) and rat aortic VSMCs (a gift from Prof. Dr. S. Janssen, Center for Transgene Technology and Gene Therapy, University of Leuven, Leuven, Belgium) were isolated and cultivated as previously described (9, 20).
Three-Dimensional Stress-Relaxed Collagen Lattice Contraction Model
The ability of HSCs to contract three-dimensional collagen matrixes was assessed as previously described with some slight modifications (29). In brief, hydrated collagen gels were prepared using rat tail tendon collagen I (Becton Dickinson Labware, Becton Dickson, Bedford, MA) and adjusted to physiological strength and pH with 1 N NaOH and 10x PBS at 4°C. Afterward, the collagen solution was mixed with a HSC suspension so that the final solution resulted in a collagen concentration of 1.5 mg/ml and 250,000 cells/ml. A 500-µl aliquot of the collagen solution was then cast into each well of a 24-well tissue culture plate (Falcon, Meylan, France) and, after 1 h, was covered with complete culture medium (1 ml/well) to ascertain adequate gelation. Afterward, the polymerized collagen matrix containing HSCs remained attached to the culture dish for 24 h, leading to mechanical loading ("stressed") (16). After 24 h, stabilized lattices were washed twice with 1x PBS, followed by the addition of 1 ml serum-free culture medium/well containing 1 µCi 3H2O (Amersham Biosciences, Roosendaal, The Netherlands). Depending on the experiment, presumed agonists and/or inhibitors were added. To initiate matrix contraction, mechanically stressed matrixes were released by gentle circumferential dislodgement of the lattice using a micropipette tip ("relaxation"). Cell-mediated contraction was measured by determining the relative partioning of 3H2O between the gel phase and surrounding medium following 24 h of contraction, thereby allowing the estimation of gel phase volumes. More specifically, the separate tritium activities of the medium and gel phase were measured in 10 ml oscillation fluid (Perkin-Elmer) using a Beckmann liquid scintillation spectrometer. Control cell-free gels provided estimates for the precontraction volume and allowed us to determine relative changes in volume (percent contraction). All data presented here are from experiments using at least three sets of three collagen lattices using culture-activated HSCs from three different rat HSC isolations.
The same setup was used in the experiments comparing contractions of VSMCs, HSCs, and cardiac myofibroblasts. In these specific experiments, the extent of contraction was expressed relative to FCS-induced contraction to allow comparisons between cell types.
Inhibitors and Agonists Affecting Contractility
Depending on the experiment, FCS (10%), the 1-adrenergic agonist methoxamine (10–4 M, Sigma), or the Ca2+ ionophore A-23187 (Sigma) were used as agonists. Mostly, FCS was used because its ease of use, high availability, and ability to elicit as powerful a contraction as endothelin-1 (25). The actomyosin interaction, the end target of both Ca2+-dependent and Ca2+-independent pathways, was studied with the use of 2,3-butanedione 2-monoxime (BDM), a nonmuscle myosin ATPase inhibitor (Sigma). The following kinase inhibitors were used: the calmodulin-mediated MLCK inhibitor W-7 (Sigma), the Rho kinase inhibitor Y-27632 (Calbiochem, La Jolla, CA), and the PKC inhibitor staurosporin (Sigma). Kinase inhibitors and BDM were added to agonist-free collagen lattices 5 min prior to the addition of an agonist and release. To evaluate the effect of increased intracellular Ca2+ on contraction in the different cell types, the Ca2+ ionophore A-23187 was used under agonist-free conditions. To investigate the effect of inhibition of MLCP, calyculin (Calbiochem) was added to Ca2+-depleted and normal (Ca2+ containing) lattices. The former lattices were obtained by washing lattices with 1x Ca2+-free PBS (3 x 2 min each), followed by 1x Ca2+-free PBS + 3 mM EGTA (2 x 5 min each and then 14 min) and a final set of washes with 1x Ca2+-free PBS + 3 mM EGTA + 0.01 µM A-23187 (3 x 2 min each). The final washing step allowed depletion of intracellular Ca2+.
Western Blot Analysis for rMLC and Phosphorylated rMLC
HSCs, cultured for 24 h in the presence of W-7, Y-27632, staurosporine, or FCS alone, were harvested and homogenized in lysis buffer (0.125 M Tris·HCl, 4% SDS, 20% glycerol, 0.02% bromophenol blue, and 0.2 M DTT). After heat denaturation (100°C for 5 min) and centrifugation (13,000 g for 10 min), equal amounts of protein (6.5 µg) were run on a 7% SDS-PAGE gel and then transferred onto a nitrocellulose Protran membrane (Schleicher & Schuell, Dassel, Germany). Membranes were blocked with 5% blocking solution (milk powder) in 1x PBS for 1.5 h at room temperature to avoid nonspecific binding. Thereafter, blots were incubated overnight with primary antibodies against rMLC and phosphorylated rMLC (1:200, Santa Cruz Biotechnology, Santa Cruz, CA). The next day, membranes were incubated with horseradish peroxidase (HRP)-conjugated secondary antibody (SwaR-HRP, 1:500, Prosan, Merelbeke, Belgium), and immunoreactivity was visualized using chemiluminescence detection (ECL-plus, Amersham Biosciences). Membranes were stained with Ponceau staining to confirm equal protein loading and transfer between lanes. Densitometric quantification of Western blot signal intensity was performed with Un-Scan-IT Gel software (Silk Scientific) and calculated as percentages to mean values of HSCs treated only with FCS.
Fluo-4 Visualization of Intracellular Ca2+
HSCs, grown on glass coverslips, were incubated with 5 µM fluo-4 + 0.025% pleurionic acid (Molecular Probes) under FCS-free conditions. Coverslips were subsequently transferred to the coverglass chamber of a confocal scanning microscope (Nikon TE 300, Noran Oz). Cells were observed and photographed before, during, and after the addition of FCS-free medium containing 0.01 µM A-23187.
Animal Model of Cirrhosis
Male Wistar rats, weighing 200–250 g, were intoxicated with TAA in drinking water. The TAA concentration was adapted weekly to changes in body weight, leading to homogenous macronodular cirrhosis with all the typical characteristics of portal hypertension after 18 wk, as previously described (13, 14).
In Situ Liver Perfusion
In TAA-treated cirrhotic and control rats (n = 25 each), the effect of the 1-agonist methoxamine was studied on IHVR with and without preincubation with W-7, Y-27632, and staurosporin by in situ liver perfusion, as previously described (13, 14). Briefly, the portal vein was cannulated and perfused through a 14-gauge angiocath with oxygenated Krebs solution at 37°C. After the inferior vena cava was transsected, allowing the perfusate to escape, a thoracotomy was performed to cannulate the suprahepatic inferior vena cava. Once the effluent was clear, recirculation was set up with a volume of 150 ml of buffer at a constant flow of 35 ml/min. Perfusion pressure was continuously monitored (Dataq, Akron, OH). Criteria of liver viability included the gross appearance of the liver, stable perfusion pressure (starting value ± 1 mmHg), and stable buffer pH (7.4 ± 0.1) during the initial 30-min stabilization period.
Statistical Analysis
Statistical analysis was performed using SigmaStat 2.0 (Jandel, San Rafael, CA). An unpaired Student's t-test, Mann-Whitney rank-sum test, or ANOVA was used when appropriate. Data are given as means ± SE. P 
0.05 was considered statistically significant.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
In Vitro Results
Relevance of intracellular Ca2+ and Ca2+-dependent pathways in HSC-mediated gel contraction. Increasing Ca2+ intracellularly under agonist-free conditions with A-23187, a Ca2+ ionophore, was demonstrated by means of confocal Ca2+ imaging, which showed a maximal rise in intracellular Ca2+ within seconds (Fig. 2). A dose-response curve with A-23187 showed maximal contraction of 54.7 ± 1.8% at 0.01 µM. Since this amounts to only 75.6 ± 1.8% of FCS-promoted contraction, increasing intracellular Ca2+ appears insufficient to exclusively cause contraction (Figs. 3 and 4). The equivalent maximal contraction relative to FCS-induced contraction generated in VSMCs and cardiac myofibroblasts amounted to 85.1 ± 2.3% and 34.6 ± 2.1%, respectively (P < 0.001 vs. the equivalent effect in HSCs; Fig. 4).
|
|
|
|
Relevance of Ca2+-independent pathways in HSC-mediated gel contraction. First, MLCP, the supposed converging end point of Ca2+-independent pathways, was inhibited using calyculin, a type 1 phosphatase inhibitor. Calyculin (10–9 M) was able to induce contraction of HSC-embedded collagen lattices in the absence of FCS (53.7% vs. 15.3% under FCS-free conditions, P < 0.05). To further specify the relative role of MLCP activity in HSC-mediated contraction, HSC contraction was examined under Ca2+-depleted and Ca2+-containing conditions in the presence or absence of calyculin (10–9 M). Under similar Ca2+-depleted conditions, contraction occurred more efficient when MLCP was inhibited (4 ± 2.3% vs. 30 ± 2.7%, P = 0.002; Fig. 6). In Ca2+-containing lattices, a comparable effect was observed when MLCP was inhibited (15.3 ± 3.8% vs. 53.7 ± 5.2%, P = 0.004), but the degree of contraction was increased compared with Ca2+-depleted conditions. These data not only illustrate an auxiliary role for MLCP in the regulation of HSC contraction in addition to increasing levels of intracellular Ca2+ but also document its involvement in the phenomenon of "Ca2+ sensitization."
|
|
|
1-adrenergic agonist methoxamine.
To finally refute the possibility that these pathways were relevant exclusively to FCS stimulation, we repeated the experiments with W-7, Y-27632, and staurosporin in the presence of the 1-adrenergic agonist methoxamine, for which activated HSCs express the receptors and biosynthetic enzymes (18). Except for under basal conditions comparing only both agonists, we observed no differences to FCS-induced contraction in the presence of W-7, Y-27632, and staurosporin (Fig. 9).
|
In agonist-free conditions, basal IHVR of cirrhotic rat livers was already increased compared with normal rat livers (0.22 ± 0.01 vs. 0.13 ± 0.01 mmHg·min·ml–1, respectively, P < 0.001). The addition of methoxamine (10–4 M) led to an increase in IHVR in both cirrhotic and normal perfused rat livers (0.65 ± 0.02 vs. 0.32 ± 0.02 mmHg·min·ml–1, respectively, P < 0.001). In the cirrhotic perfused rat liver, the response to methoxamine was aggravated compared with the normal perfused rat liver (change in increase in IHVR: 0.39 ± 0.03 mmHg·min·ml–1 for TAA vs. 0.21 ± 0.02 mmHg·min·ml–1 for control, P = 0.001), which is consistent with the phenomenon of "hyperresponsiveness" to vasoconstrictors in cirrhosis. Preincubation with Y-27632 (10–4 M), staurosporin (25 nM), and W-7 (10–4 M) in the cirrhotic rat liver significantly decreased this hyperresponsiveness to methoxamine by –66.8 ± 1.3%, –52.4 ± 2.7%, and –28.7 ± 2.8%, respectively, suggesting a predominant involvement of Ca2+-independent pathways in the increased active IHVR associated with cirrhosis (Fig. 10). In the normal liver, we noted a reduction in IHVR only after inhibition of Ca2+-independent pathways and to a lesser extent than that found in cirrhosis: –43.1 ± 4.2% (P = 0.008 vs. TAA + Y-27632) and –40.2 ± 4.2% (P = 0.038 vs. TAA + staurosporin) (Fig. 10).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
-SMA and myosin II) (5, 26); 3) the expression of L-type voltage-operated Ca2+ channels (2); and 4) the fact that agonists that are known to cause contraction in HSCs are associated with increases in intracellular Ca2+ (22–24). Although persuasive, these observations are challenged heretofore by a lack of direct evidence proving that an increase in intracellular Ca2+ indeed mediates force generation and by an apparent contradiction in terminology, as witnessed in the smooth muscle cell-like contraction pattern of a myofibroblast-like activated HSC. In contrast to previous studies, we applied a three-dimensional stress-relaxed collagen lattice contraction model to study the contractility of HSCs (16, 19, 29, 31). The reasons herefore are twofold. First, cell adhesion, analogous to the in vivo situation, occurs three dimensionally to attachment sites made upon protein fibrils rather than two dimensionally along a protein-coated interface (like for contraction studies using confocal Ca2+ imaging or when HSCs are grown on top of polymerized collagen lattices). This model therefore mimics the reciprocal geometric and mechanical relationships with the surrounding matrix that HSCs entertain in vivo (7, 16). Second, the relevance of stressed gels follows from the observation that in the cirrhotic liver, the injured liver tissue is tethered in a way that cell contraction will inevitably increase stress in the surrounding matrix, resulting in a mechanical feedback that is missing in classically used unrestrained gels ("floating gels") as well as in the classical two-dimensional assay, as mentioned earlier.
Using this in vitro model, we first tested the premise that force generation in HSCs is Ca2+ dependent. We demonstrated that, although Ca2+/calmodulin-mediated MLCK activity is necessary for HSC contraction, elevation of intracellular Ca2+ was insufficient to cause maximal contraction. These results contrasted with the demonstrated properties of VSMCs, in which gradual elevation of intracellular Ca2+ ultimately led to a similarly efficient contraction as obtained after agonist stimulation, proving the dominance of Ca2+-dependent pathways in these cells (19, 21, 27, 28). In cardiac myofibroblasts, quite the reverse phenomenon was observed, since these cells, like other types of myofibroblasts (19), appeared almost Ca2+ insensitive. These observations clearly refute any comparison for HSCs with smooth muscle cells and myofibroblasts, making terms like "smooth muscle cell like" and "myofibroblast like" inappropriate in this context.
The inability of increased intracellular Ca2+ to promote maximal contraction suggests that activated HSCs have an additional mechanism to Ca2+/calmodulin-dependent MLCK that is critical in regulating contractility. We therefore tested Ca2+-independent pathways. Since these pathways are thought to converge in the inactivation of MLCP, resulting in delayed degradation of phosphorylated rMLC and thefore decreased contraction, we first focused on this enzyme (8, 11, 19, 26, 27). We demonstrated that the phosphatase inhibitor calyculin, in the absence of any agonist, could promote HSC contraction, proving its involvement in the regulation of force generation additive to the Ca2+-dependent pathway. To further specify its role in this regulation, we examined HSC contraction under Ca2+-depleted or Ca2+-containing conditions in the presence or absence of calyculin. These experiments showed that inhibition of MLCP increased the amount of contraction independently of the level of intracellular Ca2+, suggestive of a phenomenon known as Ca2+ sensitization (21, 27). In smooth muscle cells, this phenomenon has been extensively described and is mainly attributed to the inhibition of MLCP (21). Several mechanisms have been identified in these cells to inactivate MLCP, the most important of which are the phosphorylation of the regulatory subunit of MLCP by the Rho/Rho kinase pathway and the inhibition of the catalytic subunit of MLCP mediated by the PKC-dependent pathway (8, 17, 21, 27). For HSCs, to our knowledge, the concept of Ca2+ sensitization is novel. Reviewing the two aformentioned mechanisms, which are considered to be the main regulating factors in smooth muscle cells in this process, we observed surprising parallels in HSCs. First, we and others (10, 15, 30, 31), under different conditions, could document a role for the RhoA signaling pathway since preincubation with the Rho kinase inhibitor Y-27632 attenuated contraction and decreased phosphorylation of rMLC, the rate-limiting step in the actomyosin interaction. Furthermore, we also proved that the RhoA signaling pathway acts through MLCP, since preincubation with calyculin, a phosphatase inhibitor, largely counteracted inhibition of contraction by Y-27632. This suggests that Rho kinase, a serine/threonine kinase, phosphorylates MLCP and inhibits phosphatase activity, resulting in delayed degradation of phosphorylated rMLC and thus protracted contraction. Second, we also revealed a role for the PKC-dependent pathway. Staurosporin, a PKC inhibitor, prevented FCS- and methoxamine-promoted contraction and decreased phosphorylation of rMLC. Similarly as for RhoA inhibition, calyculin also almost completely opposed inhibition of contraction by staurosporin, identifying MLCP as the end target of this pathway as well. These results indicate an auxiliary role for PKC to the RhoA pathway as a G protein-coupled effector of Ca2+ sensitization.
It should be taken into account that the inhibitors used to block the different pathways are highly specific but not exclusive to their intended target. In addition, a role for MLC phosphorylation-independent regulatory mechanisms is postulated in smooth muscle cells (21). Our data show that Ca2+-dependent (MLCK related) and Ca2+-independent (MLCP related) pathways are important in HSC contraction, but these findings do not exclude the presence or absence of alternative, more discrete pathways. The study of these will require more selective experimental tools and models that are lacking at the moment.
To correlate our in vitro findings to the in vivo situation, we reevaluated these pathways in the in situ perfused liver. We first confirmed the presence of an exaggerated response to methoxamine in the cirrhotic rat liver compared with the normal liver, illustrating the known phenomenon of intrahepatic hyperresponsiveness to vasoconstrictors in the cirrhotic liver (6, 14). Both Ca2+-dependent and -independent pathways were shown to be involved in this hyperresponsiveness of the cirrhotic liver since inhibition of these pathways attenuated the increased IHVR. A predominance of the Ca2+-independent pathways, mediated through PKC- and RhoA-mediated signaling cascades, was demonstrated.
In conclusion, we have shown that both Ca2+-dependent and Ca2+-independent pathways are necessary to raise a HSC-specific contraction pattern, which distinguishes them from both typical myofibroblasts and smooth muscle cells. These in vitro findings correlated with the data obtained in the cirrhotic liver where both pathways were involved, with Ca+-independent pathways predominating the picture. A better understanding of the intracellular signal transduction mechanisms leading to HSC contraction and the demonstration of their relevance in the establishment of the increased IHVR in the cirrhotic liver might lead to the identification of novel potential targets for the treatment of portal hypertension.
|
GRANTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. L. Friedman Hepatic Stellate Cells: Protean, Multifunctional, and Enigmatic Cells of the Liver Physiol Rev, January 1, 2008; 88(1): 125 - 172. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
P < 0.05 vs. TAA + MTX; 
P < 0.05 vs. TAA + W-7; 
P < 0.05 vs. TAA + Y-27632; 
P < 0.05 vs. normal rat liver + MTX.