- Email: [email protected]
The Effects of Hepatic Fibrosis on Ito Cell Gene Expression
Matrix Vol. 11/1992, pp. 36-43 © 1992 by Gustav Fischer Verlag, Stuttgart
The Effects of Hepatic Fibrosis on Ito Cell Gene Expression FRANCIS R. WEINER, ANISH SHAH, LUIS BIEMPICA 1 , MARK A. ZERN 2 and MARK J. CZAJA Department of Medicine and the Marion Bessin Liver Research Center, Albert Einstein College of Medicine, Bronx, NY 10461, USA.
Abstract While Ito cells appear to be a major source of increased matrix synthesis during hepatic fibrogenesis, the cellular changes that occur in these cells during liver fibrosis have not been well delineated. In this study we examined Ito cell gene expression in isolated cells from normal rats, and rats with carbon tetrachloride-induced fibrosis, in order to better define the changes occurring in these cells during this pathologic process. Specifically, we addressed three questions: (1) which matrix genes are over expressed in Ito cells in fibrotic liver; (2) do these cells increase their expression of the fibrogenic cytokine transforming growth factor-pI (TGF-PI); and (3) do Ito cells change their phenotype during hepatic fibrogenesis as reflected by alterations in the expression of their intermediate filament genes? Northern hybridization analysis revealed that Ito cells isolated from fibrotic livers had significant increases in mRNA levels of types I, III and IV procollagen compared to normal cells, while no increases were found in hepatocytes, and Kupffer/endothelial cells had only an increase in type I procollagen mRNA. Analysis of other matrix proteins which increase during hepatic fibrogenesis revealed elevations in laminin Band fibronectin mRNA levels only in Ito cells. Increased Ito cell matrix gene expression was also associated with a 4-fold increase in TGF-~11evels in these cells. No increase in TGF-p1 mRNA was found in hepatocytes, and less than a 2-fold increase was found in Kupffer/endothelial cells isolated from fibrotic livers. Ito cells isolated from both normal and fibrotic livers expressed vimentin, desmin and ~-actin mRNA, but no a-actin mRNA, implying that Ito cells in these fibrotic livers retained a normal Ito cell phenotype. These studies suggest that increases in Ito cell mRNA levels of matrix proteins contribute to the development of hepatic fibrosis. Furthermore, enhanced Ito cell expression ofTGF-pl mRNA may further augment this cell's matrix synthesis. Finally, the increases in the mRNA of these matrix proteins appear to occur in Ito cells which are not phenotypically altered by this pathologic process. Key words: hepatic fibrosis, Ito cells, TGF-Pl.
Introduction Although considerable knowledge has accumulated about hepatic fibrogenesis in recent years, fundamental Deceased. Current address for Dr. Zern is Department of Medicine, Roger Williams General Hospital, 825 Chalkstone Avenue, Providence, RI 02908, USA. I
2
questions about the cellular mechanisms of matrix accumulation during this pathologic process remain unanswered. Until recently, much of the uncertainty had centered around the determination of which hepatic cell type(s) is responsible for the increase in matrix synthesis during the development of liver fibrosis (Milani et al., 1989 a; Chojkier et al., 1988; Friedman et al., 1985). Cummulative data from a number of investigations now strongly suggest that
Ito Cell Gene Expression Ito cells playa major role in this process (Milani et a1., 1989 a; Friedman, 1990; Maher and McGuire, 1990). Although these studies have associated Ito cells with hepatic fibrosis, the molecular alterations which Ito cells undergo in vivo in the presence of a fibrogenic stimulus are not entirely known. Therefore, in this study Ito cells were isolated from normal rats and rats with carbon tetrachloride (CC14 )3-induced fibrosis in order to examine the changes that occur in these cells during their involvement in this fibrotic process. Specifically, studies of Ito cell gene expression were performed in order to address three questions concerning the functional alterations occurring in Ito cells during fibrosis. The first question asked which matrix genes are affected in Ito cells during the development of CCl4 -induced hepatic fibrosis. In addition, pure populations of hepatocytes and Kupffer/endothelial cells were also examined in order to put the changes in Ito cell extracellular matrix gene expression in perspective with those occurring in other hepatic cells during hepatic fibrogenesis. The second question was concerned with whether transforming growth factor-~1 (TGF-~1) expression was affected in these cells during fibrogenesis. This particular cytokine was studied because we have previously demonstrated that TGF-~1 gene expression rises in parallel with increasing collagen synthesis during CC14 -induced fibrosis (Czaja et a1., 1989). We also demonstrated that normal Ito cells possess TGF-~1 mRNA, and that TGF-~1 stimulates Ito cell type I procollagen gene expression and collagen synthesis (Weiner et a1., 1990). Therefore, Ito cell TGF-~1 gene expression was examined to explore the possibility that these cells may self-regulate the fibrotic process through their production of this cytokine. The final question addressed in this study asked if Ito cells undergo a phenotypic transformation during liver fibrogenesis. It has been suggested that Ito cells increase their matrix synthesis during hepatic fibrosis in association with their transformation into fibroblast or myofibroblast-like cells (Nakano and Lieber, 1982; Mak and Lieber, 1988). The fact that various intermediate filaments are differentially expressed in Ito cells, myofibroblasts and smooth muscle cells (Takase et aI., 1988; Tsutsumi et aI., 1987; Leavitt et aI., 1985), provides a means of determining whether or not Ito cells do undergo a phenotypic change during hepatic fibrosis. Therefore, the gene expression of various intermediate filaments was analyzed in Ito cells isolated from normal and fibrotic livers in order to define the phenotypic changes occurring in Ito cells during CC14 -induced liver fibrosis.
3 Abbreviations used in this text include: CCl 4 = carbon tetrachloride; GAPD = glyceraldehyde-3-phosphate dehydrogenase; and TGF-~l = transforming growth factor-~1.
37
Materials and Methods The study protocols were approved by the Animal Care and Use Committee of the Albert Einstein College of Medicine and followed N. 1. H. guidelines on the care and use of laboratory animals. Animal model
One-year old male Sprague-Dawley rats weighing 500-700 g were maintained under 12 h light/dark cycles and allowed food and water ad libitum. Treated rats were injected subcutaneously twice weekly with 1 mVkg of a 1: 1 (v/v) mixture of CC1 4 and mineral oil for 8 weeks (Kucharzy, 1987). Three days after the last injection of CC1 4 Ito cells, hepatocytes, and Kupffer/endothelial cells were isolated from the livers of normal and treated rats. Individual rats were used either for the isolation of nonparenchymal cells (Ito cells, and Kupffer/endothelial cells) or hepatocytes. Portions of liver from other normal rats and rats treated with CCl4 were also used for histologic examination and the isolation of total RNA. These experiments were performed on three separate groups of rats of similar size and age. Cell isolation
Ito cells were isolated from normal and CC4-treated rats as described by Hendriks and co-workers (Hendriks et aI., 1985), and modified by Blaner et a1. (1985). The liver was sequentially perfused with a series of Gey's balanced salt solutions containing various concentrations of pronase E (EM Science, Cherry Hill, NJ) and collagenase type I (Sigma Chemical Co., St. Louis, MO) as described previously (Weiner et a1., 1990). Ito cells were separated from endothelial and Kupffer cells by centrifugation of the nonparenchymal cell suspension over a discontinuous, two layer Nycodenz (Nyegaard & Co., Oslo, Norway) density gradient (Weiner et a1., 1990). Ito cells isolated from normal or fibrotic livers were located at the top of the 11 % (w/ v) Nycodenz layer, and were over 90% viable by trypan blue exclusion. The Ito cell yield ranged from 2-4 X 107 cells per liver. The purity of Ito cells isolated from normal or fibrotic livers was approximately 90% as determined by immunofluorescent staining with antidesmin antibodies (Tsutsumi et aI., 1987; Eghbali et aI., 1987), or staining with oil red 0 (Yamada et aI., 1987). Kupffer/endothelial cells from normal and fibrotic livers were isolated from the top of the 17.2% (w/v) Nycodenz layer of the same density gradient used for Ito cell isolation. Their viability was over 90% as determined by trypan blue exclusion. Kupffer and endothelial cells were estimated to make up 80% of this fraction based on their light microscopic size and lack of antidesmin staining. Peroxidatic staining revealed that our Kupffer/endothelial cell population contained at least 40%
38
F. R. Weiner et al.
Kupffer cells. The major contaminating cell type of our Kupffer/endothelial cell fraction was Ito cells. Hepatocytes were isolated from separate fibrotic and normal livers as previously described Oefferson et al., 1985) using the procedure of Berry and Friend (1969) and the perfusion mixture of Leffert et al. (1979). The viability of hepatocytes isolated from either normal or fibrotic livers was 90% as assessed by trypan blue exclusion, and their purity was about 90% as determined by light microscopic appearance and staining for peroxisomes (Wilson et al., 1988).
Liv ITO K/E N CllN CIIN C
Hep N CI
Coil
colm
RNA extraction and Northern blot hybridization Equal numbers of freshly isolated cells of each type were used for RNA isolation using a modification of the Chirgwin procedure (Chirgwin et al., 1979) as previously described (Czaja et al., 1987 a). Steady-state mRNA levels were determined by Northern blot hybridization analysis (Czaja et al., 1987b). Using 20 Ilg of total RNA per sample, RNA was hybridized with one of the following eDNA probes: rat a2(I) procollagen (Genovese et al., 1984), mouse types III (Liau et al., 1985) and IV procollagen (Wang and Gudas, 1983), rat fibronectin (Schwarzbauer et al., 1983), chicken ~-actin (Cleveland et al., 1980), human a-actin (Gunning et al., 1983), mouse TGF-~l (Derynck et al., 1986), mouse laminin B (Wang and Gudas, 1983), human glyceraldehyde-3-phosphate dehydrogenase (GAPD) (Tso et al., 1985), and hamster desmin and vimentin (Quax et al., 1984). The eDNA probes were labeled with 2 p]deoxycytidine triphosphate to a specific activity of 2 - 5 X 10 8 cpm/ug of DNA using a primer extension kit (Amersham, Arlington Heights, IL). After hybridization, the GeneScreen filters (NEN Research Products, Boston, MA) were washed under stringent conditions, exposed to Kodak XAR-5 film at -70°C, and scanned by densitometry. To ensure that equal amounts of RNA had been used for each sample, hybridized filters were stained with ethidium bromide and the amount of RNA per sample was visualized under ultraviolet light (Maniatis et al., 1982).
e
Results The livers from animals treated with CCl 4 for 8 weeks had extensive fibrosis, with fibrotic bands radiating from central veins to portal zones, and steatosis. Untreated rat livers were normal. Northern hybridization analysis revealed that this histologically detectable fibrosis was associated with increased mRNA levels for types I, III and IV procollagen in CCl 4 -treated animals compared to normals (Fig. 1). Quantitation of these increases by densitometric scanning of three separate groups of normal and CCl4 -treated rats demonstrated 13.9 ± 2.9 (mean ± SEM), and 7.5 ± 4.2-fold increases in types I and IV procollagen mRNA levels respectively, in livers from treated ani-
CoiN Lam Fibro
Fig. 1. Northern blot hybridization analysis of total RNA isolated from livers of normal (N) and CCl 4 -treated rats (C), and from various cell types isolated from similar livers. The RNA was electrophoresed, transferred to a GeneScreen filter, and hybridized with the indicated [32 PJ-labeled eDNA probes as described in Materials and Methods. The autoradiogram of the Northern blot of hepatocyte RNA hybridized to the type I procollagen eDNA probe was exposed for a longer period of time than the other type I procollagen autoradiograms in order to allow the faint signals from this hybridization to be evident. Abbreviations for the eDNA probes indicated in the figure are: a2(1) procollagen, Col I; type III procolJagen, Col III; type IV procollagen, Col IV; laminin B, Lam; and fibronectin, Fibro. Sample abbreviations are liver, Liv; Ito cells, ITO; Kupffer/endothelial cells, KIE; and hepatocytes, Hep. mals compared to normal livers. Type III procollagen was only detectable at very low levels in normal liver, but was significantly increased in the livers of rats with CCI 4 induced fibrosis. When Ito cells isolated from similarly treated rats were examined, 11.7 ± 2.0, 9.5 ± 0.3 and 8.5 ± 1.3-fold increases in their types I, III and IV procollagen mRNA levels respectively were found compared to the levels found in Ito cells isolated from normal livers (Fig. 1 and Table I). Thus Ito cell procollagen gene expression was increased at the time when histologically significant hepatic fibrosis was present. Kupffer/endothelial cells isolated from fibrotic liver had a 5.7 ± 0.2-fold increase in type I procollagen mRNA content compared to similar cells isolated from normal liver, but no significant change in their
Ito Cell Gene Expression
39
Table I. Results of densitometric scanning of autoradiograms of Northern blots".
COLI COL III COL IV FlBRO TGF-~1 ~-ACTIN
DESMIN VIMENTIN GAPD c
ITO
K/E
HEP
11.7 ± 2.0 9.5 ± 0.3 8.5 ± 1.3 -b 4.4 ± 0.2 2.5 ± 0.8 3.3 ± 0.6 3.0± 0.5 1.1
5.7± 0.2 1.2 ± 0.2 1.1 ± 0.2 1.0 ± 0.3 1.8 ± 0.2 1.4 ± 0.1 N.D. 1.2 ± 0.1 1.4
1.2±0.4 1.0 ± 0.2 1.0 ± 0.2 0.4 ± 0.2 1.1 ± 0.2 1.5 ± 0.1 N.D. N.D. 1.3
" Mean ± SEM for three experiments. Messenger RNA levels in hepatic cells isolated from fibrotic liver expressed as the fold increase in mRNA levels over the same hepatic cells isolated from normal liver. Abbreviations: ITO = Ito cells; K/E = Kupffer/ endothelial cells; HEP = hepatocytes; COL I = a2(I) procollagen; COL III = type III procollagen; COL IV = type IV procollagen; FlBRO = fibronectin; TGF-~1 = transfoming growth factor-~1; GAPD = glyceraldehyde-3-phosphate dehydrogenase; and N. D. = none detected. b No detectable mRNA in nomal cells to quantitiate the % increase in the mRNA levels in fibrotic cells. C Mean of two experiments.
types III and IV procollagen mRNA content (Fig. 1 and Table I). In contrast to Ito and Kupffer/endothelial cells, no significant increase in types I, III or IV procollagen mRNA content was found in hepatocytes isolated from fibrotic livers even with prolonged exposure of these Northern blots (Fig. 1). When these same samples were hybridized with a glyceraldehyde-3-phosphate dehydrogenase (GAPD) eDNA clone, no significant change was found in the mRNA levels for this constitutive gene (Fig. 2). Ethidium bromide staining of each filter with ultraviolet visualization of the ribosomal bands also revealed that equal amounts of total RNA had been used for each set of samples. Analysis of a noncollagenous extracellular matrix gene, laminin B, revealed that the mRNA levels of laminin B were undetectable in normal liver but present in the livers of treated rats. When isolated cells from fibrotic and normal liver were studied, laminin B mRNA content was also undetectable in Ito cells isolated from normal liver, but significantly increased in Ito cells isolated from fibrotic livers (Fig. 1). Laminin B mRNA was not found in hepatocytes or Kupffer/endothelial cells isolated from either normal or fibrotic livers. In contrast to laminin B, fibronectin, another noncollagenous extracellular matrix protein, had no significant change in its hepatic mRNA levels in CCl 4 -treated animals compared to untreated animals (Fig. 1). Despite this finding, evaluation of specific cellular populations revealed both a significant increase in this mRNA in Ito cells isolated from fibrotic livers and a 60% decrease in hepatocytes (Fig. 1 and TableI). Kupffer/endothelial cells had no significant change in their total fibronectin mRNA content (Fig. 1 and TableI).
IN
ITO K/E CIIN CII N Cl
Liv
Hep I N C
TGF-f31
Actin Desmin
Vimentin .
GAPD
Fig. 2. Northern blot hybridization analysis of total RNA isolated from livers of normal rats (N) and CC!4-treated rats (C) and from nonparenchymal and parenchymal cells isolated from similar rats. The autoradiogram of hepatocyte RNA hybridized with the TGF~1 eDNA clone was exposed for a longer time than the other samples hybridized with this eDNA probe. Abbreviations for the eDNA probes are: transforming growth factor-~1, TGF-~l; and glyceraldehyde-3-phosphate dehydrogenase, GAPD. Abbreviations also included are liver, Liv; Ito cells, ITO; Kupffer/endothelial cells, K/E; and hepatocytes, Hep.
Analysis of TGF-~1 mRNA content in fibrotic livers revealed an increase of 4.2 ± 0.2-fold in fibrotic livers compared to normal livers (Fig. 2). This increase in TGF-~1 mRNA content was associated with a 4.4 ± 0.2-fold increase in TGF-~1 mRNA in Ito cells isolated from fibrotic livers. Less than a two-fold (1.8 ± 0.2) increase in TGF-~1 mRNA was found in Kupffer/endothelial cells isolated from fibrotic livers. Hepatocytes isolated from normal rats contained only low levels of TGF-~1 mRNA and these cells failed to demonstrate any increase in their TGF-~1 mRNA content with the development of fibrosis (Fig. 2 and Table I). To investigate whether these changes in Ito cell matrix and cytokine gene expression were associated with
40
F. R. Weiner et a1.
phenotypic changes, the gene expression for a series of intermediate filaments was determined in these cells. Desmin mRNA which was undetectable in normal liver increased significantly in fibrotic liver. Vimentin and ~ actin mRNA were detectable in normal liver and were found to increase by 2.7 ± 0.5 and 1.6 ± O.l-fold, respectively in livers of rats treated with CCI 4 compared to untreated animals (Fig. 2) Desmin mRNA was only found in Ito cells, and increased 3.3 ± 0.6-fold with fibrosis. Vimentin mRNA was found in both Ito cells and Kupffer/endothelial cells, and increased 3.0 ± 0.5-fold in Ito cells isolated from fibrotic liver. No significant change in vimentin mRNA levels was found in Kupffer/endothelial cells isolated from normal and fibrotic livers. ~-actin mRNA was the only type of actin mRNA found in normal and fibrotic livers, and in all isolated cell types. Specifically, no a-actin was found in Ito cells isolated from normal or fibrotic livers using either a human a-actin eDNA probe or human eDNA probes specific for the 3' untranslated regions of skeletal or cardiac a-actin (data not shown). However, the human a-actin specific eDNA probe did demonstrate a-actin mRNA in cultured Ito cells (data not shown), consistent with the idea that cell culture alters Ito cell phenotype.
Discussion This study demonstrates that the increases in types I, III and IV procollagen mRNA content which occur in CCI4 induced hepatic fibrosis are associated with increased expression of these genes in Ito cells. It is also likely that this finding of increased procollagen mRNA levels in Ito cells correlates with increased collagen protein production by these cells, since many studies have shown that procollagen mRNA content corresponds well with collagen synthesis and content (Weiner et aI., 1987; Weiner et aI., 1990; Pierce et aI., 1987; Raghow et aI., 1987). In addition, Milani et a1. (1989a and 1990) found that the strongest in situ hybridization signals for procollagen mRNAs were generally found in areas of excessive collagen deposition. Taken together these studies support the idea that Ito cells with increased procollagen mRNA levels are actively involved in collagen synthesis during CCl 4 -induced hepatic fibrosis. Our data also suggest that Kupffer or endothelial cells are capable of increased type I collagen synthesis during fibrogenesis. However, part of this effect may be due to Ito cell contamination which would be consistent with the study of Irving et a1. (1984), who were unable to detect type I collagen in normal endothelial cells. In contrast, no increases in types I, III or IV procollagen mRNA content were found in hepatocytes isolated from fibrotic liver as compared to those from normal livers. Therefore, while this data does not exclude hepatocytes as a source of extracellular matrix synthesis in normal liver, it does demonstrate that Ito cells are an important source of increased matrix synthesis dur-
ing hepatic fibrogenesis. A recent study examining only matrix gene expression in isolated hepatic cell populations also found that the highest levels of procollagen gene expression were found in Ito cells isolated from fibrotic liver (Maher and McGuire, 1990). However, it is impossible to conclude that Ito cells are the only source of increased matrix synthesis without knowledge of the amount of RNA present in each cellular population, the percentage of mRNA extracted from each cell type, and the numbers of various hepatic cells in normal liver versus fibrotic liver. The collagens were not the only matrix genes whose expression was increased in fibrotic Ito cells. Increased laminin gene expression was also found in these cells. In contrast to Clement et a1. (1988) who found that some hepatocytes were positive for laminin in fibrotic human livers, we found no laminin B mRNA in hepatocytes isolated from normal or fibrotic rat liver. However, our results are consistent with those of several other investigators who demonstrated either immunohistochemically or by in situ hybridization that nonparenchymal cells were a major source of laminin in rats during the development of CCI 4 induced hepatic fibrosis (Martinez-Hernandez, 1985), or in humans with fibrotic livers (Milani et aI., 1989 b). Examination of fibronectin mRNA content revealed that while total fibronectin mRNA content was not significantly altered in fibrotic liver, fibronectin mRNA content was decreased in hepatocytes and increased in Ito cells. It would appear that these results are inconsistent with studies which have shown increased fibronectin in fibrotic liver by immunohistochemistry (Martinez-Hernandez, 1985; Clement et aI., 1986), however, the existence of two types of hepatic fibronectin offers a potential explanation for these findings. Hepatic fibronectin consists of plasma fibronectin which is synthesized predominantly by hepatocytes (Tamkun and Hynes, 1983) and cellular fibronectin which is produced by many cells (Clement et aI., 1988). These two forms result from alternative splicing within the coding region of the fibronectin gene (Schwarzbauer et aI., 1983), and their mRNAs are both detected by our fibronectin eDNA. Plasma fibronectin is found primarily in serum while cellular fibronectin is deposited locally in the extracellular matrix (Martinez-Hernandez, 1984). In chronic liver disease, plasma fibronectin levels decrease with advancing hepatic dysfunction (Naveau et aI., 1985), while increased cellular fibronectin continues to be synthesized as a component of the increased extracellular matrix (Martinez-Hernandez, 1985). Thus the decrease in hepatocyte fibronectin mRNA observed in fibrotic livers may be due to decreased plasma fibronectin gene expression. However, total liver fibronectin mRNA levels were unchanged possibly because of the increase in Ito cell cellular fibronectin gene expression. This increase in Ito cell fibronectin mRNA is also consistent with the immunohistochemical results of Clement et a1. (1988).
Ito Cell Gene Expression While we have previously demonstrated in CCl 4 -induced fibrosis that TGF-~1 mRNA was elevated during the development of hepatic fibrosis (Czaja et aI., 1989), the cellular source of TGF-~1 was not determined. In the present study, we have demonstrated that the increase in TGF~1 mRNA which occurs in the fibrotic livers of CCI4 treated rats is associated with a significant increase in the TGF-~1 mRNA content of Ito cells consistent with the studies of Nakatsukasa et ai. (1990a and 1990b). While Kupffer/endothelial cells also contained increased amounts of TGF-~1 mRNA, it was less than that found in Ito cells isolated from the same fibrotic liver. The fact that these cells have increased production of both procollagen and TGF-~ 1 mRNAs further supports a significant role for Ito cells in the development of hepatic fibrosis. This finding takes on even greater importance when one considers our previous finding that TGF-~1 not only enhanced collagen gene expression and protein synthesis in Ito cells, but also increased Ito cell TGF-~1 gene expression (Weiner et aI., 1990). Further studies are now required to determine whether Ito cells produce functionally active TGF-~1 during hepatic fibrogenesis. Finally, the possibility that phenotypic changes occur in Ito cells during the development of hepatic fibrosis was investigated. The presence of various intermediate filaments has been felt to distinguish Ito cells, myofibroblasts, smooth muscle cells and fibroblasts (Tsutsumi et aI., 1987; Takase et aI., 1988; Leavitt et aI., 1985). Thus the detection of changes in the expression of these proteins in Ito cells isolated from fibrotic livers compared to normal livers could indicate whether phenotypic changes occur in these cells during the development of hepatic fibrosis. Desmin mRNA was found to persist in Ito cells isolated from fibrotic livers. The presence of desmin mRNA has been suggested to be specific for Ito cells (Tsutsumi et aI., 1987) and Takase et aI., (1988) have demonstrated that myofibroblasts lack immunohistochemically detectable desmin. Thus our finding that Ito cells from fibrotic liver contained desmin mRNA suggests that these cells are phenotypically distinct from myofibroblasts. Further support for this conclusion exists from our analysis of the type of actin mRNA found in these cells. Ito cells contained only ~-actin and not a-actin mRNA regardless of whether they were isolated from normal or fibrotic livers. This lack of myoblast-associated a-actin mRNA again suggests that fibrogenic stimulation may not induce all Ito cells to become smooth musclelike cells. Therefore, in contrast to other studies which have suggested that transformed Ito cells are responsible for increased collagen deposition during hepatic fibrogenesis (Nakano and Lieber, 1982; Mak and Lieber, 1988), our data suggests that phenotypically normal Ito cells also contribute to the excessive matrix deposition in the CCl 4 model. In summary, while previous studies have examined the hepatic cell type(s) responsible for matrix synthesis (Choj-
41
kier et aI., 1988; Miliani et aI., 1989 a; Maher and McGuire, 1990), our study represents a more comprehensive analysis of the molecular changes Ito cells undergo during CCl4 -induced hepatic fibrosis. On the basis of our findings, it appears that gene expression for a number of matrix proteins increases in Ito cells at the time of hepatic fibrosis. Furthermore, these changes occur without a significant alteration in the molecular phenotype of Ito cells. The fact that Ito cells from fibrotic livers have a significant increase in their TGF-~1 gene expression also suggests that this cell may both regulate, and contribute to the overproduction of matrix proteins during hepatic fibrogenesis. Further studies are now necessary to determine the precise mechanisms responsible for the initiation and control of this Ito cell response during the development of hepatic fibrosis. Acknowledgements This investigation was supported in part by National Institutes of Health awards DK38484 (F. R. W.) and DK01793 (M.]. C.). We thank Emily Bobe and Anna Caponigro for their secretarial assistance. The GAPD cDNA clone was obtained from the American Type Culture Collection.
References Berry, M. N. and Friend, D. 5.: High-yield preparation of isolated rat liver parenchymal cells: a biochemical and fine structural study.]. Cell. Bioi. 43: 506-520, 1969. Blaner, W.S., Hendriks, H.F.j., Brouwer, A., de Leeuw, A.M., Knook, D.L. and Goodman, D.S.: Retinoids, retinoid-binding proteins and retinyl palmitate hydrolase distribution in different types of rat liver cells.]. Lipid. Res. 26: 1241-1251,1985. Chirgwin, ].M., Przbyla, A.E., MacDonald, R.]. and Rutter, W.].: Isolation of biologically active ribonucleic acid from sources enriched with ribonuclease. Biochemistry 18: 5294-5299, 1979. Chojkier, M., Lyche, K. D. and Filip, M.: Increased production of collagen in vivo by hepatocytes and nonparenchymal cells in rats with carbon tetrachloride-induced hepatic fibrosis. Hepatology 8: 808-814, 1988. Clement, B., Grimaud, ].-A., Champion, J.P., Deugnier, Y. and Guillouzo, A.: Cell types involved in collagen and fibronectin production in normal and fibrotic human liver. Hepatology 6: 225-234,1986. Clement, B., Rescan, P.-Y., Baffet, G., Loreal, 0., Lehry, D., Campion, ].-P. and Guillouzo, A.: Hepatocytes may produce laminin in fibrotic livers and in primary culture. Hepatology 8: 794-803,1988. Cleveland, D. W., Lapata, M. A., MacDonald, R.j., Cowan, N.]., Rutter, W.]. and Kirschner, M. W.: Number and evolutionary conservation of u- and ~-tubulin and cytoplasmic ~- and y-actin genes using specific cloned eDNA probes. Cell 20: 95 -105, 1980. Czaja, M.]., Weiner, F.R., Eghbali, M., Giambrone, M.-A., Eghbali, M. and Zero, M. A.: Differential effects of y-interferon on collagen and fibronectin gene expression. J. Bioi. Chern. 262: 13348-13351,1987a.
42
F. R. Weiner et a1.
Czaja, M.]., Weiner, F. R., Schwarzenberg, S.]., Sternlieb, I., Scheinberg, I.H., Van Thiel, D.H., LaRusso, N.F. and Zern, M.A.: Molecular studies of ceruloplasmin deficiency in Wilson's disease.]. Clin. Invest. 80: 1200-1204, 1987b. Czaja, M.]., Weiner, F. R., Flanders, K. e., Giambrone, M.-A., Wind, R., Biempica, L. and Zern, M.A.: In vitro and in vivo association of transforming growth factor-beta 1 with hepatic fibrosis.]. Cell. BioI. 108: 2477-2482, 1989. Derynck, R., jarrett, j.A., Chen, E. Y. and Goelddel, D. V.: The murine transforming growth factor-~ precursor. ]. BioI. Chern. 261: 4377 -4379,1986. Eghbali, M., Seifter, S., Robinson, T.F. and Blumenfeld, 0.0.: Enzyme-antibody histochemistry: a method for detection of collagens collectively. Histochemistry 87: 257- 262,1987. Friedman, S.L., Roll, F.]., Boyles, J and Bissell, D.M.: Hepatic lipocytes: the principal collagen-producing cells of normal rat liver. Proc. Natl. Acad. Sci. USA 82: 8681-8685, 1985. Friedman, S. L.: Cellular sources of collagen and regulation of collagen production in liver. Seminars in Liver Disease 10: 20-29, 1990. Genovese, e., Rowe, D. and Kream, B.: Construction of DNA sequences complementary to rat al and rat a2 collagen mRNA and their use in studying the regulation of type I collagen synthesis by 1,24-dihydroxyvitamin D. Biochemistry 23: 62106216,1984. Gunning, P., Ponte, P., Okayama, H., Engel, j., Blau, H. and Kedes, L.: Isolation and characterization of full-length eDNA clones for human a-, ~- and y-actin mRNAs: skeletal but not cytoplasmic actins have an amino-terminal cysteine that is subsequently removed. Mol. Cell. BioI. 3: 787-795, 1983. Hendriks, H.F., Verhoofstad, A.M.M., Brouwer, A., de Leeuw, A. M. and Knook, D. L.: Perisinusoidal fat-storing cells are the main vitamin A storage sites in rat liver. Exp. Cell. Res. 160: 138-149,1985. Irving, M. G., Roll, J, Huang, S. and Bissell, D. M.: Characterization and culture of sinusoidal endothelium from normal rat liver: Lipoprotein uptake and collagen phenotype. Gastroenterology87: 1233-1247, 1984. jefferson, D.M., Reid, L.M., Giambrone, M.-A., Shafritz, D.A. and Zern, M. A.: Effects of dexamethasone on albumin and collagen gene expression in primary cultures of adult rat hepatocytes. Hepatology 5: 14-20, 1985. Kucharzy, E.J: Dynamics of collagen accumulation and activity of collagen degrading enzymes in livers of rats with carbon tetrachloride-induced hepatic fibrosis. Conn. Tissue Res. 16: 143-151,1987. Leavitt, ]., Gunning, P., Kedes, L. and jariwalla, R.: Smooth muscle a-actin is a marker for mouse NIH 3T3 and Rat-2 cells. Nature 316: 840-842, 1985. Leffert, H. L., Koch, K. S., Moran, T. and Williams, M.: Liver cells. Meth. Enzymol. 58: 536-544, 1979. Liau, G., Mudryj, M. and deCrombrugghe, B.: Identification of the promoter and first exon of the mouse al (III) collagen gene. ]. BioI. Chern. 260: 3773-3777, 1985. Maher, JJ and McGuire, R. F.: Extracellular matrix gene expression increases preferentially in rat lipocytes and sinusoidal endothelial cells during hepatic fibrosis in vivo.]. Clin. Invest. 86: 1641-1648, 1990. Mak, K. M. and Lieber, e. S.: Lipocytes and transitional cells in alcoholic liver disease: a morphometric study. Hepatology 8: 1027-1033,1988. Maniatis, T., Fritsch, E. F. and Sambrook, j.: Gel electrophoresis. In: Molecular cloning: a laboratory manual, ed. by Maniatis,
T., Fritsch, E.F. and Sambrook, j., Cold Spring Harbor Laboratory, New York, 1982; pp. 161. Martinez-Hernandez, A.: The hepatic extracellular matrix. I. Electron immunohistochemical studies in normal rat liver. Lab. Invest. 51: 57-74, 1984. Martinez-Hernandez, A.: The hepatic extracellular matrix. II. Electron immunohistochemical studies in rats with CCI4 induced cirrhosis. Lab. Invest. 53: 166-186,1985. Milani, S., Herbst, H., Schuppan, D., Hahn, E. G. and Stein, H.: In situ hybridization for procollagen types I, III and IV mRNA in normal and fibrotic rat liver: Evidence for predominant expression in nonparenchymal liver cells. Hepatology 10: 84-92, 1989a. Milani, S., Herbst, H., Schuppan, D., Riecken, E. O. and Stein, H.: Cellular localization of laminin gene transcripts in normal and fibrotic human liver. Amer.]. Path. 134: 1175 -1181,1989 b. Milani, S., Herbst, H., Schuppan, D., Kim, K. Y., Riecken, E. O. and Stein, H.: Procollagen expression by nonparenchymal rat liver cells in experimental biliary fibrosis. Gastroenterology 98: 175-184,1990. Nakano, M. and Lieber, e. S.: Ultrastructure of initial stages of perivenular fibrosis in alcohol-fed baboons. Am.]. Pathol. 106: 145 -155, 1982. Nakatsukasa, H., Nagy, P., Evarts, R. P., Hsia, e.-e., Marsden, E. and Thorgeirsson, S. S.: Cellular distribution of transforming growth factor-~l and procollagen types I, III and IV transcripts in carbon tetrachloride-induced rat liver fibrosis.]. Clin. Invest. 85: 1833-1843, 1990a. Nakatsukasa, H., Evarts, R. P., Hsia, e.-e. and Thorgeirsson, S.S.: Transforming growth factor-~l and typeI procollagen transcripts during regeneration and early fibrosis of rat liver. Lab. Invest. 63: 171-181, 1990b. Naveau, S., Poynard, T., Abella, A., Pignon, J-P., Poitrine, A., Agostini, H. and Zourabichvili, 0.: Prognostic value of serum fibronectin concentration in alcoholic cirrhotic patients. Hepatology 5: 819-823, 1985. Pierce, R. A., Glaug, M. R., Greco, R. S., Mackenzie, J W., Boyd, e. D. and Deak, S.B.: Increased procollagen mRNA levels in carbon tetrachloride-induced liver fibrosis in rats. ]. BioI. Chern. 262: 1652-1658,1987. Quax, W., Van Den Heuvel, R., Egberts, W. V. and Quax-jeuken, Y.: Intermediate filament cDNAs from BHK-21 cells: Demonstration of distinct genes for desmin and vimentin in all vertebrate classes. Proc. Natl. Acad. Sci. USA 81: 5970-5974, 1984. Raghow, R., Postlethwaite, A. E., Keski-Oja,]., Moses, H. L. and Kang, A. H.: Transforming growth factor-~ increases steady state levels of type I procollagen and fibronectin messenger RNAs post-transcriptionally in cultured human dermal fibroblasts.]. Clin. Invest. 79: 1285 -1288,1987. Schwarzbauer, JE., Tamkun, J W., Lemischka, I. R. and Hynes, R.O.: Three different fibronectin mRNAs arise by alternative splicing within the coding region. Cell 35: 421-431, 1983. Takase, S., Leo, M.A., Nouchi, T. and Lieber, e.S.: Desmin distinguishes cultured fat-storing cells from myofibroblasts, smooth muscle cells and fibroblasts in the rat. J. Hepatology 6: 267-276,1988. Tamkun,j. W. and Hynes, R. 0.: Plasma fibronectin is synthesized and secreted by hepatocytes. j. BioI. Chern. 258: 4641-4647, 1983. Tso, J Y., Sun, X.-H., Kao, T.-H., Reece, K. S. and Wu, R.: Isolation and characterization of rat and human glyceraldehyde-3phosphate dehydrogenase cDNAs: genomic complexity and
Ito Cell Gene Expression molecular evolution of the gene. Nucleic Acid Research 13: 2485-2502,1985. Tsursumi, M., Takada, A. and Takase, 5.: Characterization of desmin-positive rat liver sinusoidal cells. Hepatology 7: 277-284,1987. Wang, 5.-Y. and Gudas, L.j.: Isolation of eDNA clones specific for collagen IV and laminin from mouse teratocarcinoma cells. Proc. Natl. Acado Sci. USA 80: 5880-5884, 1983. Weiner, FoR., Czaja, M.J., Giambrone, M.-A., Takahashi,S., Biempica, L. and Zern, M.A.: Transcriptional and posttranscriptional effects of dexamethasone on albumin and procollagen messenger RNAs in murine schistosomiasis. Biochemistry 26: 1557-1562,1987. Weiner, FoR., Giambrone, M.-A., Czaja, M.j., Shah, A., Annoni, G., Takahashi,S., Eghbali, M. and Zern, M.A.: Ito cell gene
43
expression and collagen regulation. Hepatology 11: 111-117, 1990. Wilson,j. M.,Jefferson, D. M., Chowdhury, R.J., Novikoff, P. M., Johnston, D.E. and Mulligan, R.C.: Retrovirus-mediated transduction of adult hepatocytes. Proc. Natl. Acad. Sci. USA 85: 1314-1318,1988. Yamada, M., Blaner, W. 5., Soprano, D. R., Dixon, j. L., Kjeldbye, HoM. and Goodman, D.S.: Biochemical characterization of isolated rat liver stellate cells. Hepatology 7: 1224-1229, 1987. Dr. Mark j. Czaja, Marion Bessin Liver Research Center, Albert Einstein Collage of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461, USA.
The Effects of Hepatic Fibrosis on Ito Cell Gene Expression FRANCIS R. WEINER, ANISH SHAH, LUIS BIEMPICA 1 , MARK A. ZERN 2 and MARK J. CZAJA Department of Medicine and the Marion Bessin Liver Research Center, Albert Einstein College of Medicine, Bronx, NY 10461, USA.
Abstract While Ito cells appear to be a major source of increased matrix synthesis during hepatic fibrogenesis, the cellular changes that occur in these cells during liver fibrosis have not been well delineated. In this study we examined Ito cell gene expression in isolated cells from normal rats, and rats with carbon tetrachloride-induced fibrosis, in order to better define the changes occurring in these cells during this pathologic process. Specifically, we addressed three questions: (1) which matrix genes are over expressed in Ito cells in fibrotic liver; (2) do these cells increase their expression of the fibrogenic cytokine transforming growth factor-pI (TGF-PI); and (3) do Ito cells change their phenotype during hepatic fibrogenesis as reflected by alterations in the expression of their intermediate filament genes? Northern hybridization analysis revealed that Ito cells isolated from fibrotic livers had significant increases in mRNA levels of types I, III and IV procollagen compared to normal cells, while no increases were found in hepatocytes, and Kupffer/endothelial cells had only an increase in type I procollagen mRNA. Analysis of other matrix proteins which increase during hepatic fibrogenesis revealed elevations in laminin Band fibronectin mRNA levels only in Ito cells. Increased Ito cell matrix gene expression was also associated with a 4-fold increase in TGF-~11evels in these cells. No increase in TGF-p1 mRNA was found in hepatocytes, and less than a 2-fold increase was found in Kupffer/endothelial cells isolated from fibrotic livers. Ito cells isolated from both normal and fibrotic livers expressed vimentin, desmin and ~-actin mRNA, but no a-actin mRNA, implying that Ito cells in these fibrotic livers retained a normal Ito cell phenotype. These studies suggest that increases in Ito cell mRNA levels of matrix proteins contribute to the development of hepatic fibrosis. Furthermore, enhanced Ito cell expression ofTGF-pl mRNA may further augment this cell's matrix synthesis. Finally, the increases in the mRNA of these matrix proteins appear to occur in Ito cells which are not phenotypically altered by this pathologic process. Key words: hepatic fibrosis, Ito cells, TGF-Pl.
Introduction Although considerable knowledge has accumulated about hepatic fibrogenesis in recent years, fundamental Deceased. Current address for Dr. Zern is Department of Medicine, Roger Williams General Hospital, 825 Chalkstone Avenue, Providence, RI 02908, USA. I
2
questions about the cellular mechanisms of matrix accumulation during this pathologic process remain unanswered. Until recently, much of the uncertainty had centered around the determination of which hepatic cell type(s) is responsible for the increase in matrix synthesis during the development of liver fibrosis (Milani et al., 1989 a; Chojkier et al., 1988; Friedman et al., 1985). Cummulative data from a number of investigations now strongly suggest that
Ito Cell Gene Expression Ito cells playa major role in this process (Milani et a1., 1989 a; Friedman, 1990; Maher and McGuire, 1990). Although these studies have associated Ito cells with hepatic fibrosis, the molecular alterations which Ito cells undergo in vivo in the presence of a fibrogenic stimulus are not entirely known. Therefore, in this study Ito cells were isolated from normal rats and rats with carbon tetrachloride (CC14 )3-induced fibrosis in order to examine the changes that occur in these cells during their involvement in this fibrotic process. Specifically, studies of Ito cell gene expression were performed in order to address three questions concerning the functional alterations occurring in Ito cells during fibrosis. The first question asked which matrix genes are affected in Ito cells during the development of CCl4 -induced hepatic fibrosis. In addition, pure populations of hepatocytes and Kupffer/endothelial cells were also examined in order to put the changes in Ito cell extracellular matrix gene expression in perspective with those occurring in other hepatic cells during hepatic fibrogenesis. The second question was concerned with whether transforming growth factor-~1 (TGF-~1) expression was affected in these cells during fibrogenesis. This particular cytokine was studied because we have previously demonstrated that TGF-~1 gene expression rises in parallel with increasing collagen synthesis during CC14 -induced fibrosis (Czaja et a1., 1989). We also demonstrated that normal Ito cells possess TGF-~1 mRNA, and that TGF-~1 stimulates Ito cell type I procollagen gene expression and collagen synthesis (Weiner et a1., 1990). Therefore, Ito cell TGF-~1 gene expression was examined to explore the possibility that these cells may self-regulate the fibrotic process through their production of this cytokine. The final question addressed in this study asked if Ito cells undergo a phenotypic transformation during liver fibrogenesis. It has been suggested that Ito cells increase their matrix synthesis during hepatic fibrosis in association with their transformation into fibroblast or myofibroblast-like cells (Nakano and Lieber, 1982; Mak and Lieber, 1988). The fact that various intermediate filaments are differentially expressed in Ito cells, myofibroblasts and smooth muscle cells (Takase et aI., 1988; Tsutsumi et aI., 1987; Leavitt et aI., 1985), provides a means of determining whether or not Ito cells do undergo a phenotypic change during hepatic fibrosis. Therefore, the gene expression of various intermediate filaments was analyzed in Ito cells isolated from normal and fibrotic livers in order to define the phenotypic changes occurring in Ito cells during CC14 -induced liver fibrosis.
3 Abbreviations used in this text include: CCl 4 = carbon tetrachloride; GAPD = glyceraldehyde-3-phosphate dehydrogenase; and TGF-~l = transforming growth factor-~1.
37
Materials and Methods The study protocols were approved by the Animal Care and Use Committee of the Albert Einstein College of Medicine and followed N. 1. H. guidelines on the care and use of laboratory animals. Animal model
One-year old male Sprague-Dawley rats weighing 500-700 g were maintained under 12 h light/dark cycles and allowed food and water ad libitum. Treated rats were injected subcutaneously twice weekly with 1 mVkg of a 1: 1 (v/v) mixture of CC1 4 and mineral oil for 8 weeks (Kucharzy, 1987). Three days after the last injection of CC1 4 Ito cells, hepatocytes, and Kupffer/endothelial cells were isolated from the livers of normal and treated rats. Individual rats were used either for the isolation of nonparenchymal cells (Ito cells, and Kupffer/endothelial cells) or hepatocytes. Portions of liver from other normal rats and rats treated with CCl4 were also used for histologic examination and the isolation of total RNA. These experiments were performed on three separate groups of rats of similar size and age. Cell isolation
Ito cells were isolated from normal and CC4-treated rats as described by Hendriks and co-workers (Hendriks et aI., 1985), and modified by Blaner et a1. (1985). The liver was sequentially perfused with a series of Gey's balanced salt solutions containing various concentrations of pronase E (EM Science, Cherry Hill, NJ) and collagenase type I (Sigma Chemical Co., St. Louis, MO) as described previously (Weiner et a1., 1990). Ito cells were separated from endothelial and Kupffer cells by centrifugation of the nonparenchymal cell suspension over a discontinuous, two layer Nycodenz (Nyegaard & Co., Oslo, Norway) density gradient (Weiner et a1., 1990). Ito cells isolated from normal or fibrotic livers were located at the top of the 11 % (w/ v) Nycodenz layer, and were over 90% viable by trypan blue exclusion. The Ito cell yield ranged from 2-4 X 107 cells per liver. The purity of Ito cells isolated from normal or fibrotic livers was approximately 90% as determined by immunofluorescent staining with antidesmin antibodies (Tsutsumi et aI., 1987; Eghbali et aI., 1987), or staining with oil red 0 (Yamada et aI., 1987). Kupffer/endothelial cells from normal and fibrotic livers were isolated from the top of the 17.2% (w/v) Nycodenz layer of the same density gradient used for Ito cell isolation. Their viability was over 90% as determined by trypan blue exclusion. Kupffer and endothelial cells were estimated to make up 80% of this fraction based on their light microscopic size and lack of antidesmin staining. Peroxidatic staining revealed that our Kupffer/endothelial cell population contained at least 40%
38
F. R. Weiner et al.
Kupffer cells. The major contaminating cell type of our Kupffer/endothelial cell fraction was Ito cells. Hepatocytes were isolated from separate fibrotic and normal livers as previously described Oefferson et al., 1985) using the procedure of Berry and Friend (1969) and the perfusion mixture of Leffert et al. (1979). The viability of hepatocytes isolated from either normal or fibrotic livers was 90% as assessed by trypan blue exclusion, and their purity was about 90% as determined by light microscopic appearance and staining for peroxisomes (Wilson et al., 1988).
Liv ITO K/E N CllN CIIN C
Hep N CI
Coil
colm
RNA extraction and Northern blot hybridization Equal numbers of freshly isolated cells of each type were used for RNA isolation using a modification of the Chirgwin procedure (Chirgwin et al., 1979) as previously described (Czaja et al., 1987 a). Steady-state mRNA levels were determined by Northern blot hybridization analysis (Czaja et al., 1987b). Using 20 Ilg of total RNA per sample, RNA was hybridized with one of the following eDNA probes: rat a2(I) procollagen (Genovese et al., 1984), mouse types III (Liau et al., 1985) and IV procollagen (Wang and Gudas, 1983), rat fibronectin (Schwarzbauer et al., 1983), chicken ~-actin (Cleveland et al., 1980), human a-actin (Gunning et al., 1983), mouse TGF-~l (Derynck et al., 1986), mouse laminin B (Wang and Gudas, 1983), human glyceraldehyde-3-phosphate dehydrogenase (GAPD) (Tso et al., 1985), and hamster desmin and vimentin (Quax et al., 1984). The eDNA probes were labeled with 2 p]deoxycytidine triphosphate to a specific activity of 2 - 5 X 10 8 cpm/ug of DNA using a primer extension kit (Amersham, Arlington Heights, IL). After hybridization, the GeneScreen filters (NEN Research Products, Boston, MA) were washed under stringent conditions, exposed to Kodak XAR-5 film at -70°C, and scanned by densitometry. To ensure that equal amounts of RNA had been used for each sample, hybridized filters were stained with ethidium bromide and the amount of RNA per sample was visualized under ultraviolet light (Maniatis et al., 1982).
e
Results The livers from animals treated with CCl 4 for 8 weeks had extensive fibrosis, with fibrotic bands radiating from central veins to portal zones, and steatosis. Untreated rat livers were normal. Northern hybridization analysis revealed that this histologically detectable fibrosis was associated with increased mRNA levels for types I, III and IV procollagen in CCl 4 -treated animals compared to normals (Fig. 1). Quantitation of these increases by densitometric scanning of three separate groups of normal and CCl4 -treated rats demonstrated 13.9 ± 2.9 (mean ± SEM), and 7.5 ± 4.2-fold increases in types I and IV procollagen mRNA levels respectively, in livers from treated ani-
CoiN Lam Fibro
Fig. 1. Northern blot hybridization analysis of total RNA isolated from livers of normal (N) and CCl 4 -treated rats (C), and from various cell types isolated from similar livers. The RNA was electrophoresed, transferred to a GeneScreen filter, and hybridized with the indicated [32 PJ-labeled eDNA probes as described in Materials and Methods. The autoradiogram of the Northern blot of hepatocyte RNA hybridized to the type I procollagen eDNA probe was exposed for a longer period of time than the other type I procollagen autoradiograms in order to allow the faint signals from this hybridization to be evident. Abbreviations for the eDNA probes indicated in the figure are: a2(1) procollagen, Col I; type III procolJagen, Col III; type IV procollagen, Col IV; laminin B, Lam; and fibronectin, Fibro. Sample abbreviations are liver, Liv; Ito cells, ITO; Kupffer/endothelial cells, KIE; and hepatocytes, Hep. mals compared to normal livers. Type III procollagen was only detectable at very low levels in normal liver, but was significantly increased in the livers of rats with CCI 4 induced fibrosis. When Ito cells isolated from similarly treated rats were examined, 11.7 ± 2.0, 9.5 ± 0.3 and 8.5 ± 1.3-fold increases in their types I, III and IV procollagen mRNA levels respectively were found compared to the levels found in Ito cells isolated from normal livers (Fig. 1 and Table I). Thus Ito cell procollagen gene expression was increased at the time when histologically significant hepatic fibrosis was present. Kupffer/endothelial cells isolated from fibrotic liver had a 5.7 ± 0.2-fold increase in type I procollagen mRNA content compared to similar cells isolated from normal liver, but no significant change in their
Ito Cell Gene Expression
39
Table I. Results of densitometric scanning of autoradiograms of Northern blots".
COLI COL III COL IV FlBRO TGF-~1 ~-ACTIN
DESMIN VIMENTIN GAPD c
ITO
K/E
HEP
11.7 ± 2.0 9.5 ± 0.3 8.5 ± 1.3 -b 4.4 ± 0.2 2.5 ± 0.8 3.3 ± 0.6 3.0± 0.5 1.1
5.7± 0.2 1.2 ± 0.2 1.1 ± 0.2 1.0 ± 0.3 1.8 ± 0.2 1.4 ± 0.1 N.D. 1.2 ± 0.1 1.4
1.2±0.4 1.0 ± 0.2 1.0 ± 0.2 0.4 ± 0.2 1.1 ± 0.2 1.5 ± 0.1 N.D. N.D. 1.3
" Mean ± SEM for three experiments. Messenger RNA levels in hepatic cells isolated from fibrotic liver expressed as the fold increase in mRNA levels over the same hepatic cells isolated from normal liver. Abbreviations: ITO = Ito cells; K/E = Kupffer/ endothelial cells; HEP = hepatocytes; COL I = a2(I) procollagen; COL III = type III procollagen; COL IV = type IV procollagen; FlBRO = fibronectin; TGF-~1 = transfoming growth factor-~1; GAPD = glyceraldehyde-3-phosphate dehydrogenase; and N. D. = none detected. b No detectable mRNA in nomal cells to quantitiate the % increase in the mRNA levels in fibrotic cells. C Mean of two experiments.
types III and IV procollagen mRNA content (Fig. 1 and Table I). In contrast to Ito and Kupffer/endothelial cells, no significant increase in types I, III or IV procollagen mRNA content was found in hepatocytes isolated from fibrotic livers even with prolonged exposure of these Northern blots (Fig. 1). When these same samples were hybridized with a glyceraldehyde-3-phosphate dehydrogenase (GAPD) eDNA clone, no significant change was found in the mRNA levels for this constitutive gene (Fig. 2). Ethidium bromide staining of each filter with ultraviolet visualization of the ribosomal bands also revealed that equal amounts of total RNA had been used for each set of samples. Analysis of a noncollagenous extracellular matrix gene, laminin B, revealed that the mRNA levels of laminin B were undetectable in normal liver but present in the livers of treated rats. When isolated cells from fibrotic and normal liver were studied, laminin B mRNA content was also undetectable in Ito cells isolated from normal liver, but significantly increased in Ito cells isolated from fibrotic livers (Fig. 1). Laminin B mRNA was not found in hepatocytes or Kupffer/endothelial cells isolated from either normal or fibrotic livers. In contrast to laminin B, fibronectin, another noncollagenous extracellular matrix protein, had no significant change in its hepatic mRNA levels in CCl 4 -treated animals compared to untreated animals (Fig. 1). Despite this finding, evaluation of specific cellular populations revealed both a significant increase in this mRNA in Ito cells isolated from fibrotic livers and a 60% decrease in hepatocytes (Fig. 1 and TableI). Kupffer/endothelial cells had no significant change in their total fibronectin mRNA content (Fig. 1 and TableI).
IN
ITO K/E CIIN CII N Cl
Liv
Hep I N C
TGF-f31
Actin Desmin
Vimentin .
GAPD
Fig. 2. Northern blot hybridization analysis of total RNA isolated from livers of normal rats (N) and CC!4-treated rats (C) and from nonparenchymal and parenchymal cells isolated from similar rats. The autoradiogram of hepatocyte RNA hybridized with the TGF~1 eDNA clone was exposed for a longer time than the other samples hybridized with this eDNA probe. Abbreviations for the eDNA probes are: transforming growth factor-~1, TGF-~l; and glyceraldehyde-3-phosphate dehydrogenase, GAPD. Abbreviations also included are liver, Liv; Ito cells, ITO; Kupffer/endothelial cells, K/E; and hepatocytes, Hep.
Analysis of TGF-~1 mRNA content in fibrotic livers revealed an increase of 4.2 ± 0.2-fold in fibrotic livers compared to normal livers (Fig. 2). This increase in TGF-~1 mRNA content was associated with a 4.4 ± 0.2-fold increase in TGF-~1 mRNA in Ito cells isolated from fibrotic livers. Less than a two-fold (1.8 ± 0.2) increase in TGF-~1 mRNA was found in Kupffer/endothelial cells isolated from fibrotic livers. Hepatocytes isolated from normal rats contained only low levels of TGF-~1 mRNA and these cells failed to demonstrate any increase in their TGF-~1 mRNA content with the development of fibrosis (Fig. 2 and Table I). To investigate whether these changes in Ito cell matrix and cytokine gene expression were associated with
40
F. R. Weiner et a1.
phenotypic changes, the gene expression for a series of intermediate filaments was determined in these cells. Desmin mRNA which was undetectable in normal liver increased significantly in fibrotic liver. Vimentin and ~ actin mRNA were detectable in normal liver and were found to increase by 2.7 ± 0.5 and 1.6 ± O.l-fold, respectively in livers of rats treated with CCI 4 compared to untreated animals (Fig. 2) Desmin mRNA was only found in Ito cells, and increased 3.3 ± 0.6-fold with fibrosis. Vimentin mRNA was found in both Ito cells and Kupffer/endothelial cells, and increased 3.0 ± 0.5-fold in Ito cells isolated from fibrotic liver. No significant change in vimentin mRNA levels was found in Kupffer/endothelial cells isolated from normal and fibrotic livers. ~-actin mRNA was the only type of actin mRNA found in normal and fibrotic livers, and in all isolated cell types. Specifically, no a-actin was found in Ito cells isolated from normal or fibrotic livers using either a human a-actin eDNA probe or human eDNA probes specific for the 3' untranslated regions of skeletal or cardiac a-actin (data not shown). However, the human a-actin specific eDNA probe did demonstrate a-actin mRNA in cultured Ito cells (data not shown), consistent with the idea that cell culture alters Ito cell phenotype.
Discussion This study demonstrates that the increases in types I, III and IV procollagen mRNA content which occur in CCI4 induced hepatic fibrosis are associated with increased expression of these genes in Ito cells. It is also likely that this finding of increased procollagen mRNA levels in Ito cells correlates with increased collagen protein production by these cells, since many studies have shown that procollagen mRNA content corresponds well with collagen synthesis and content (Weiner et aI., 1987; Weiner et aI., 1990; Pierce et aI., 1987; Raghow et aI., 1987). In addition, Milani et a1. (1989a and 1990) found that the strongest in situ hybridization signals for procollagen mRNAs were generally found in areas of excessive collagen deposition. Taken together these studies support the idea that Ito cells with increased procollagen mRNA levels are actively involved in collagen synthesis during CCl 4 -induced hepatic fibrosis. Our data also suggest that Kupffer or endothelial cells are capable of increased type I collagen synthesis during fibrogenesis. However, part of this effect may be due to Ito cell contamination which would be consistent with the study of Irving et a1. (1984), who were unable to detect type I collagen in normal endothelial cells. In contrast, no increases in types I, III or IV procollagen mRNA content were found in hepatocytes isolated from fibrotic liver as compared to those from normal livers. Therefore, while this data does not exclude hepatocytes as a source of extracellular matrix synthesis in normal liver, it does demonstrate that Ito cells are an important source of increased matrix synthesis dur-
ing hepatic fibrogenesis. A recent study examining only matrix gene expression in isolated hepatic cell populations also found that the highest levels of procollagen gene expression were found in Ito cells isolated from fibrotic liver (Maher and McGuire, 1990). However, it is impossible to conclude that Ito cells are the only source of increased matrix synthesis without knowledge of the amount of RNA present in each cellular population, the percentage of mRNA extracted from each cell type, and the numbers of various hepatic cells in normal liver versus fibrotic liver. The collagens were not the only matrix genes whose expression was increased in fibrotic Ito cells. Increased laminin gene expression was also found in these cells. In contrast to Clement et a1. (1988) who found that some hepatocytes were positive for laminin in fibrotic human livers, we found no laminin B mRNA in hepatocytes isolated from normal or fibrotic rat liver. However, our results are consistent with those of several other investigators who demonstrated either immunohistochemically or by in situ hybridization that nonparenchymal cells were a major source of laminin in rats during the development of CCI 4 induced hepatic fibrosis (Martinez-Hernandez, 1985), or in humans with fibrotic livers (Milani et aI., 1989 b). Examination of fibronectin mRNA content revealed that while total fibronectin mRNA content was not significantly altered in fibrotic liver, fibronectin mRNA content was decreased in hepatocytes and increased in Ito cells. It would appear that these results are inconsistent with studies which have shown increased fibronectin in fibrotic liver by immunohistochemistry (Martinez-Hernandez, 1985; Clement et aI., 1986), however, the existence of two types of hepatic fibronectin offers a potential explanation for these findings. Hepatic fibronectin consists of plasma fibronectin which is synthesized predominantly by hepatocytes (Tamkun and Hynes, 1983) and cellular fibronectin which is produced by many cells (Clement et aI., 1988). These two forms result from alternative splicing within the coding region of the fibronectin gene (Schwarzbauer et aI., 1983), and their mRNAs are both detected by our fibronectin eDNA. Plasma fibronectin is found primarily in serum while cellular fibronectin is deposited locally in the extracellular matrix (Martinez-Hernandez, 1984). In chronic liver disease, plasma fibronectin levels decrease with advancing hepatic dysfunction (Naveau et aI., 1985), while increased cellular fibronectin continues to be synthesized as a component of the increased extracellular matrix (Martinez-Hernandez, 1985). Thus the decrease in hepatocyte fibronectin mRNA observed in fibrotic livers may be due to decreased plasma fibronectin gene expression. However, total liver fibronectin mRNA levels were unchanged possibly because of the increase in Ito cell cellular fibronectin gene expression. This increase in Ito cell fibronectin mRNA is also consistent with the immunohistochemical results of Clement et a1. (1988).
Ito Cell Gene Expression While we have previously demonstrated in CCl 4 -induced fibrosis that TGF-~1 mRNA was elevated during the development of hepatic fibrosis (Czaja et aI., 1989), the cellular source of TGF-~1 was not determined. In the present study, we have demonstrated that the increase in TGF~1 mRNA which occurs in the fibrotic livers of CCI4 treated rats is associated with a significant increase in the TGF-~1 mRNA content of Ito cells consistent with the studies of Nakatsukasa et ai. (1990a and 1990b). While Kupffer/endothelial cells also contained increased amounts of TGF-~1 mRNA, it was less than that found in Ito cells isolated from the same fibrotic liver. The fact that these cells have increased production of both procollagen and TGF-~ 1 mRNAs further supports a significant role for Ito cells in the development of hepatic fibrosis. This finding takes on even greater importance when one considers our previous finding that TGF-~1 not only enhanced collagen gene expression and protein synthesis in Ito cells, but also increased Ito cell TGF-~1 gene expression (Weiner et aI., 1990). Further studies are now required to determine whether Ito cells produce functionally active TGF-~1 during hepatic fibrogenesis. Finally, the possibility that phenotypic changes occur in Ito cells during the development of hepatic fibrosis was investigated. The presence of various intermediate filaments has been felt to distinguish Ito cells, myofibroblasts, smooth muscle cells and fibroblasts (Tsutsumi et aI., 1987; Takase et aI., 1988; Leavitt et aI., 1985). Thus the detection of changes in the expression of these proteins in Ito cells isolated from fibrotic livers compared to normal livers could indicate whether phenotypic changes occur in these cells during the development of hepatic fibrosis. Desmin mRNA was found to persist in Ito cells isolated from fibrotic livers. The presence of desmin mRNA has been suggested to be specific for Ito cells (Tsutsumi et aI., 1987) and Takase et aI., (1988) have demonstrated that myofibroblasts lack immunohistochemically detectable desmin. Thus our finding that Ito cells from fibrotic liver contained desmin mRNA suggests that these cells are phenotypically distinct from myofibroblasts. Further support for this conclusion exists from our analysis of the type of actin mRNA found in these cells. Ito cells contained only ~-actin and not a-actin mRNA regardless of whether they were isolated from normal or fibrotic livers. This lack of myoblast-associated a-actin mRNA again suggests that fibrogenic stimulation may not induce all Ito cells to become smooth musclelike cells. Therefore, in contrast to other studies which have suggested that transformed Ito cells are responsible for increased collagen deposition during hepatic fibrogenesis (Nakano and Lieber, 1982; Mak and Lieber, 1988), our data suggests that phenotypically normal Ito cells also contribute to the excessive matrix deposition in the CCl 4 model. In summary, while previous studies have examined the hepatic cell type(s) responsible for matrix synthesis (Choj-
41
kier et aI., 1988; Miliani et aI., 1989 a; Maher and McGuire, 1990), our study represents a more comprehensive analysis of the molecular changes Ito cells undergo during CCl4 -induced hepatic fibrosis. On the basis of our findings, it appears that gene expression for a number of matrix proteins increases in Ito cells at the time of hepatic fibrosis. Furthermore, these changes occur without a significant alteration in the molecular phenotype of Ito cells. The fact that Ito cells from fibrotic livers have a significant increase in their TGF-~1 gene expression also suggests that this cell may both regulate, and contribute to the overproduction of matrix proteins during hepatic fibrogenesis. Further studies are now necessary to determine the precise mechanisms responsible for the initiation and control of this Ito cell response during the development of hepatic fibrosis. Acknowledgements This investigation was supported in part by National Institutes of Health awards DK38484 (F. R. W.) and DK01793 (M.]. C.). We thank Emily Bobe and Anna Caponigro for their secretarial assistance. The GAPD cDNA clone was obtained from the American Type Culture Collection.
References Berry, M. N. and Friend, D. 5.: High-yield preparation of isolated rat liver parenchymal cells: a biochemical and fine structural study.]. Cell. Bioi. 43: 506-520, 1969. Blaner, W.S., Hendriks, H.F.j., Brouwer, A., de Leeuw, A.M., Knook, D.L. and Goodman, D.S.: Retinoids, retinoid-binding proteins and retinyl palmitate hydrolase distribution in different types of rat liver cells.]. Lipid. Res. 26: 1241-1251,1985. Chirgwin, ].M., Przbyla, A.E., MacDonald, R.]. and Rutter, W.].: Isolation of biologically active ribonucleic acid from sources enriched with ribonuclease. Biochemistry 18: 5294-5299, 1979. Chojkier, M., Lyche, K. D. and Filip, M.: Increased production of collagen in vivo by hepatocytes and nonparenchymal cells in rats with carbon tetrachloride-induced hepatic fibrosis. Hepatology 8: 808-814, 1988. Clement, B., Grimaud, ].-A., Champion, J.P., Deugnier, Y. and Guillouzo, A.: Cell types involved in collagen and fibronectin production in normal and fibrotic human liver. Hepatology 6: 225-234,1986. Clement, B., Rescan, P.-Y., Baffet, G., Loreal, 0., Lehry, D., Campion, ].-P. and Guillouzo, A.: Hepatocytes may produce laminin in fibrotic livers and in primary culture. Hepatology 8: 794-803,1988. Cleveland, D. W., Lapata, M. A., MacDonald, R.j., Cowan, N.]., Rutter, W.]. and Kirschner, M. W.: Number and evolutionary conservation of u- and ~-tubulin and cytoplasmic ~- and y-actin genes using specific cloned eDNA probes. Cell 20: 95 -105, 1980. Czaja, M.]., Weiner, F.R., Eghbali, M., Giambrone, M.-A., Eghbali, M. and Zero, M. A.: Differential effects of y-interferon on collagen and fibronectin gene expression. J. Bioi. Chern. 262: 13348-13351,1987a.
42
F. R. Weiner et a1.
Czaja, M.]., Weiner, F. R., Schwarzenberg, S.]., Sternlieb, I., Scheinberg, I.H., Van Thiel, D.H., LaRusso, N.F. and Zern, M.A.: Molecular studies of ceruloplasmin deficiency in Wilson's disease.]. Clin. Invest. 80: 1200-1204, 1987b. Czaja, M.]., Weiner, F. R., Flanders, K. e., Giambrone, M.-A., Wind, R., Biempica, L. and Zern, M.A.: In vitro and in vivo association of transforming growth factor-beta 1 with hepatic fibrosis.]. Cell. BioI. 108: 2477-2482, 1989. Derynck, R., jarrett, j.A., Chen, E. Y. and Goelddel, D. V.: The murine transforming growth factor-~ precursor. ]. BioI. Chern. 261: 4377 -4379,1986. Eghbali, M., Seifter, S., Robinson, T.F. and Blumenfeld, 0.0.: Enzyme-antibody histochemistry: a method for detection of collagens collectively. Histochemistry 87: 257- 262,1987. Friedman, S.L., Roll, F.]., Boyles, J and Bissell, D.M.: Hepatic lipocytes: the principal collagen-producing cells of normal rat liver. Proc. Natl. Acad. Sci. USA 82: 8681-8685, 1985. Friedman, S. L.: Cellular sources of collagen and regulation of collagen production in liver. Seminars in Liver Disease 10: 20-29, 1990. Genovese, e., Rowe, D. and Kream, B.: Construction of DNA sequences complementary to rat al and rat a2 collagen mRNA and their use in studying the regulation of type I collagen synthesis by 1,24-dihydroxyvitamin D. Biochemistry 23: 62106216,1984. Gunning, P., Ponte, P., Okayama, H., Engel, j., Blau, H. and Kedes, L.: Isolation and characterization of full-length eDNA clones for human a-, ~- and y-actin mRNAs: skeletal but not cytoplasmic actins have an amino-terminal cysteine that is subsequently removed. Mol. Cell. BioI. 3: 787-795, 1983. Hendriks, H.F., Verhoofstad, A.M.M., Brouwer, A., de Leeuw, A. M. and Knook, D. L.: Perisinusoidal fat-storing cells are the main vitamin A storage sites in rat liver. Exp. Cell. Res. 160: 138-149,1985. Irving, M. G., Roll, J, Huang, S. and Bissell, D. M.: Characterization and culture of sinusoidal endothelium from normal rat liver: Lipoprotein uptake and collagen phenotype. Gastroenterology87: 1233-1247, 1984. jefferson, D.M., Reid, L.M., Giambrone, M.-A., Shafritz, D.A. and Zern, M. A.: Effects of dexamethasone on albumin and collagen gene expression in primary cultures of adult rat hepatocytes. Hepatology 5: 14-20, 1985. Kucharzy, E.J: Dynamics of collagen accumulation and activity of collagen degrading enzymes in livers of rats with carbon tetrachloride-induced hepatic fibrosis. Conn. Tissue Res. 16: 143-151,1987. Leavitt, ]., Gunning, P., Kedes, L. and jariwalla, R.: Smooth muscle a-actin is a marker for mouse NIH 3T3 and Rat-2 cells. Nature 316: 840-842, 1985. Leffert, H. L., Koch, K. S., Moran, T. and Williams, M.: Liver cells. Meth. Enzymol. 58: 536-544, 1979. Liau, G., Mudryj, M. and deCrombrugghe, B.: Identification of the promoter and first exon of the mouse al (III) collagen gene. ]. BioI. Chern. 260: 3773-3777, 1985. Maher, JJ and McGuire, R. F.: Extracellular matrix gene expression increases preferentially in rat lipocytes and sinusoidal endothelial cells during hepatic fibrosis in vivo.]. Clin. Invest. 86: 1641-1648, 1990. Mak, K. M. and Lieber, e. S.: Lipocytes and transitional cells in alcoholic liver disease: a morphometric study. Hepatology 8: 1027-1033,1988. Maniatis, T., Fritsch, E. F. and Sambrook, j.: Gel electrophoresis. In: Molecular cloning: a laboratory manual, ed. by Maniatis,
T., Fritsch, E.F. and Sambrook, j., Cold Spring Harbor Laboratory, New York, 1982; pp. 161. Martinez-Hernandez, A.: The hepatic extracellular matrix. I. Electron immunohistochemical studies in normal rat liver. Lab. Invest. 51: 57-74, 1984. Martinez-Hernandez, A.: The hepatic extracellular matrix. II. Electron immunohistochemical studies in rats with CCI4 induced cirrhosis. Lab. Invest. 53: 166-186,1985. Milani, S., Herbst, H., Schuppan, D., Hahn, E. G. and Stein, H.: In situ hybridization for procollagen types I, III and IV mRNA in normal and fibrotic rat liver: Evidence for predominant expression in nonparenchymal liver cells. Hepatology 10: 84-92, 1989a. Milani, S., Herbst, H., Schuppan, D., Riecken, E. O. and Stein, H.: Cellular localization of laminin gene transcripts in normal and fibrotic human liver. Amer.]. Path. 134: 1175 -1181,1989 b. Milani, S., Herbst, H., Schuppan, D., Kim, K. Y., Riecken, E. O. and Stein, H.: Procollagen expression by nonparenchymal rat liver cells in experimental biliary fibrosis. Gastroenterology 98: 175-184,1990. Nakano, M. and Lieber, e. S.: Ultrastructure of initial stages of perivenular fibrosis in alcohol-fed baboons. Am.]. Pathol. 106: 145 -155, 1982. Nakatsukasa, H., Nagy, P., Evarts, R. P., Hsia, e.-e., Marsden, E. and Thorgeirsson, S. S.: Cellular distribution of transforming growth factor-~l and procollagen types I, III and IV transcripts in carbon tetrachloride-induced rat liver fibrosis.]. Clin. Invest. 85: 1833-1843, 1990a. Nakatsukasa, H., Evarts, R. P., Hsia, e.-e. and Thorgeirsson, S.S.: Transforming growth factor-~l and typeI procollagen transcripts during regeneration and early fibrosis of rat liver. Lab. Invest. 63: 171-181, 1990b. Naveau, S., Poynard, T., Abella, A., Pignon, J-P., Poitrine, A., Agostini, H. and Zourabichvili, 0.: Prognostic value of serum fibronectin concentration in alcoholic cirrhotic patients. Hepatology 5: 819-823, 1985. Pierce, R. A., Glaug, M. R., Greco, R. S., Mackenzie, J W., Boyd, e. D. and Deak, S.B.: Increased procollagen mRNA levels in carbon tetrachloride-induced liver fibrosis in rats. ]. BioI. Chern. 262: 1652-1658,1987. Quax, W., Van Den Heuvel, R., Egberts, W. V. and Quax-jeuken, Y.: Intermediate filament cDNAs from BHK-21 cells: Demonstration of distinct genes for desmin and vimentin in all vertebrate classes. Proc. Natl. Acad. Sci. USA 81: 5970-5974, 1984. Raghow, R., Postlethwaite, A. E., Keski-Oja,]., Moses, H. L. and Kang, A. H.: Transforming growth factor-~ increases steady state levels of type I procollagen and fibronectin messenger RNAs post-transcriptionally in cultured human dermal fibroblasts.]. Clin. Invest. 79: 1285 -1288,1987. Schwarzbauer, JE., Tamkun, J W., Lemischka, I. R. and Hynes, R.O.: Three different fibronectin mRNAs arise by alternative splicing within the coding region. Cell 35: 421-431, 1983. Takase, S., Leo, M.A., Nouchi, T. and Lieber, e.S.: Desmin distinguishes cultured fat-storing cells from myofibroblasts, smooth muscle cells and fibroblasts in the rat. J. Hepatology 6: 267-276,1988. Tamkun,j. W. and Hynes, R. 0.: Plasma fibronectin is synthesized and secreted by hepatocytes. j. BioI. Chern. 258: 4641-4647, 1983. Tso, J Y., Sun, X.-H., Kao, T.-H., Reece, K. S. and Wu, R.: Isolation and characterization of rat and human glyceraldehyde-3phosphate dehydrogenase cDNAs: genomic complexity and
Ito Cell Gene Expression molecular evolution of the gene. Nucleic Acid Research 13: 2485-2502,1985. Tsursumi, M., Takada, A. and Takase, 5.: Characterization of desmin-positive rat liver sinusoidal cells. Hepatology 7: 277-284,1987. Wang, 5.-Y. and Gudas, L.j.: Isolation of eDNA clones specific for collagen IV and laminin from mouse teratocarcinoma cells. Proc. Natl. Acado Sci. USA 80: 5880-5884, 1983. Weiner, FoR., Czaja, M.J., Giambrone, M.-A., Takahashi,S., Biempica, L. and Zern, M.A.: Transcriptional and posttranscriptional effects of dexamethasone on albumin and procollagen messenger RNAs in murine schistosomiasis. Biochemistry 26: 1557-1562,1987. Weiner, FoR., Giambrone, M.-A., Czaja, M.j., Shah, A., Annoni, G., Takahashi,S., Eghbali, M. and Zern, M.A.: Ito cell gene
43
expression and collagen regulation. Hepatology 11: 111-117, 1990. Wilson,j. M.,Jefferson, D. M., Chowdhury, R.J., Novikoff, P. M., Johnston, D.E. and Mulligan, R.C.: Retrovirus-mediated transduction of adult hepatocytes. Proc. Natl. Acad. Sci. USA 85: 1314-1318,1988. Yamada, M., Blaner, W. 5., Soprano, D. R., Dixon, j. L., Kjeldbye, HoM. and Goodman, D.S.: Biochemical characterization of isolated rat liver stellate cells. Hepatology 7: 1224-1229, 1987. Dr. Mark j. Czaja, Marion Bessin Liver Research Center, Albert Einstein Collage of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461, USA.