Hepatocyte collagen production in vivo in normal rats.

Although hepatocytes produce collagen in vitro, their contribution to hepatic collagen synthesis in vivo is unknown. To answer this question, we injected rats intraperitoneally with [3H]proline and [14C]ornithine. [3H]Proline labeled prolyl-t-RNA in both hepatocytes and nonparenchymal cells. In contrast, [14C]ornithine was rapidly converted to [14C]arginine via the urea cycle only in hepatocytes, labeling arginyl-t-RNA. Approximately 60% of the 14C in albumin and transferrin was present as arginine while the remainder was found in proline and related amino acids. As expected for proteins that have the same proline/arginine ratio and that are produced solely by the hepatocyte, the [3H]proline/[14C]arginine ratio was very similar in albumin and transferrin. Conversely, in nonparenchymal cells a negligible percentage of 14C was present as arginine. A sizeable percentage of the 14C in hepatic collagen was present as arginine; given the greater proline(+hydroxyproline)/arginine ratio in hepatic collagen, our data indicate that in normal rats, hepatocytes contribute most of newly synthesized hepatic collagen.

d-alpha-tocopherol inhibits collagen alpha 1(I) gene expression in cultured human fibroblasts. Modulation of constitutive collagen gene expression by lipid peroxidation.

Ascorbic acid stimulates collagen gene transcription in cultured fibroblasts, and this effect is mediated through the induction of lipid peroxidation by ascorbic acid. Quiescent cultured fibroblasts in the absence of ascorbic acid have a high constitutive level of collagen production, but the mechanisms of collagen gene regulation in this unstimulated state are not known. Because lipid peroxidation also occurs in normal cells, we wondered if lipid peroxidation plays a role in the regulation of basal collagen gene expression. Inhibition of lipid peroxidation in cultured human fibroblasts with d-alpha-tocopherol or methylene blue decreased the synthesis of collagen, the steady-state levels of procollagen alpha 1(I) mRNA and the transcription of the procollagen alpha 1(I) gene. This effect on collagen gene expression was selective and not associated with cellular toxicity. Thus, these experiments suggest a role for lipid peroxidation in the modulation of constitutive collagen gene expression.

Specifically decreased collagen biosynthesis in scurvy dissociated from an effect on proline hydroxylation and correlated with body weight loss. In vitro studies in guinea pig calvarial bones.

The question whether ascorbate regulates collagen production solely through its direct role in proline hydroxylation was investigated. Proteins in calvarial bones from control and scorbutic weanling guinea pigs were labeled in short-term cultures with radioactive proline. Proteins were digested with purified bacterial collagenase to distinguish between effects on collagen polypeptide production and hydroxyproline formation. There was a preferential decrease in the absolute rate of collagen biosynthesis beginning after 2 wk of ascorbate deficiency, and this effect was temporally dissociated from decreased proline hydroxylation. There were no significant changes in the absolute rates of collagen degradation or noncollagen protein production. In vitro inhibition of proline hydroxylation in normal bone with alpha, alpha'-dipyridyl did not affect the relative rate of collagen synthesis, further dissociating these functions. Ascorbate added to scorbutic bone cultures reversed defective proline hydroxylation but not defective collagen synthesis, suggesting that the latter was an indirect effect of scurvy. There was a linear correlation between the extent of body weight lost during the 3rd and 4th wk of scurvy and the rate of collagen synthesis in scorbutic bone. This correlation also applied to control animals receiving ascorbate, but with weight loss induced by food restriction. These studies establish for the first time that ascorbate deficiency in guinea pigs leads to a specific decrease in collagen polypeptide synthesis and suggest that this decrease results from the reduced food intake and/or weight-loss characteristic of scurvy.

Proliferation of hepatic stellate cells is inhibited by phosphorylation of CREB on serine 133.

Proliferating, activated, hepatic stellate cells have a high level of collagen type I expression. Therefore, stellate cell proliferation is a critical step in hepatic fibrosis. Here we show that proliferation of activated primary rat stellate cells was blocked by elevation of cAMP with 8 Br-cAMP or isomethylbutyl xanthine, a phosphodiesterase inhibitor, and by stimulation of Ca2+ fluxes with the Ca2+ ionophore A-23187. Because phosphorylation of CREB on Ser133 is an important mediator of cAMP-protein kinase (PKA) and Ca2+-calmodulin kinase II (CAMK-II) activation, we tested whether CREB-PSer133 was essential for stellate cell quiescence. Nuclear extracts from quiescent, but not from activated, stellate cells contained CREB-PSer133. Moreover, the phosphorylation of CREB on Ser133 was stimulated in activated cells by inducing the activity of PKA or CAMK-II. In addition, coexpression of CREB and either a constitutively active PKA or a constitutively active CAMK-II inhibited the proliferation of activated stellate cells. In contrast, expression of CREB alone, PKA or CAMK-II alone, CREB-Ala 133 (which lacks the Ser133 phosphoacceptor) with PKA or CAMK-II, or CREB with inactive PKA or CAMK-II mutants did not affect stellate cell proliferation, suggesting that CREB-PSer133 is necessary for blocking the stellate cell cycle. Conversely, expression of a trans-dominant negative CREB-Ala 133 mutant (which competes with CREB/CREB-PSer133 for cognate DNA binding sites and presumably for protein interactions) induced a greater than fivefold entry into S-phase of quiescent stellate cells, compared with control cells expressing either beta-galactosidase or wt CREB, indicating that CREB-PSer133 may be indispensable for the quiescent stellate cell phenotype. This study suggests that PKA and CAMK-II play an essential role on stellate cell activation through the induction of CREB phosphorylation on Ser133, and provides potential approaches for the treatment of hepatic fibrogenesis in patients with chronic liver diseases.

LAP (NF-IL6) transactivates the collagen alpha 1(I) gene from a 5' regulatory region.

Although collagen is known to enhance hepatocyte differentiation and hepatocytes produce collagen in vivo, the transcriptional factors responsible for collagen type I gene expression in hepatic cells are not known. LAP (Liver Activator Protein) is a member of the C/EBP family, which in differentiated hepatocytes contributes to the high levels of liver-specific gene expression. In this study we show that LAP binds to the collagen alpha 1(I) promoter at both reverse CCAAT motifs and activates transcription. Furthermore, an upstream element, collagen element I (-370/-344), which shares homology with the LAP binding cis-element of the albumin promoter (9 of 13 bp) is described. This collagen element I stimulates transcription in both orientations and when placed in front of either a homologous or a heterologous chimeric report construct. These experiments suggest that LAP may be important in the expression of collagen in differentiated hepatocytes through both the promoter and a newly described upstream element.

Activation of hepatic stellate cells by TGF alpha and collagen type I is mediated by oxidative stress through c-myb expression.

Excessive production of collagen type I is a major contributor to hepatic fibrosis. Activated (myofibroblastic), but not quiescent, hepatic stellate cells (lipocytes) have a high level of collagen type I and alpha-smooth muscle actin expression. Therefore, stellate cell activation is a critical step in hepatic fibrosis. Here we show that quiescent stellate cells were activated by the generation of free radicals with ascorbate/FeSO4 and by malondialdehyde, a product of lipid peroxidation. In addition, stellate cell activation by collagen type I matrix and TGF alpha was blocked by antioxidants, such as d-alpha-tocopherol and butylated hydroxytoluene. Moreover, oxidative stress, TGF alpha and collagen type I markedly stimulated stellate cell entry into S-phase, NFkB activity, and c-myb expression, which were prevented by antioxidants. c-myb antisense oligonucleotide blocked the activation and proliferation of stellate cells induced by TGF alpha. Nuclear extracts from activated, but not from quiescent, stellate cells formed a complex with the critical promoter E box of the alpha-smooth muscle actin gene, which was disrupted by c-myb and NFkB65 antibodies, and competed by c-myb and NFkB cognate DNA. c-Myb expression was also stimulated in activated stellate cells in carbon tetrachloride-induced hepatic injury and fibrogenesis. This study indicates that oxidative stress plays an essential role, through the induction of c-myb and NFkB, on stellate cell activation.

Malondialdehyde and 4-hydroxynonenal protein adducts in plasma and liver of rats with iron overload.

In hepatic iron overload, iron-catalyzed lipid peroxidation has been implicated in the mechanisms of hepatocellular injury. Lipid peroxidation may produce reactive aldehydes such as malondialdehyde (MDA) and 4-hydroxynonenal (4-HNE), which may form aldehyde-protein adducts. We investigated whether lipid peroxidation occurred in rats fed a diet containing 3% carbonyl iron for 5-13 wk, and if this resulted in the formation of MDA- and 4-HNE- protein adducts. Chronic iron feeding resulted in hepatic iron overload (greater than 10-fold) and concomitantly induced a 2-fold increase in hepatic lipid peroxidation. Using an antiserum specific for MDA-lysine protein adducts, we demonstrated by immunohistochemistry the presence of aldehyde-protein adducts in the cytosol of periportal hepatocytes, which co-localized with iron. In addition, MDA- and 4-HNE-lysine adducts were found in plasma proteins of animals with iron overload. Only MDA adducts were detected in albumin, while other plasma proteins including a approximately 120-kD protein had both MDA and 4-HNE adducts. In this animal model of hepatic iron overload, injury occurs primarily in periportal hepatocytes, where MDA-lysine protein adducts and excess iron co-localized.

Two different cis-acting regulatory regions direct cell-specific transcription of the collagen alpha 1(I) gene in hepatic stellate cells and in skin and tendon fibroblasts.

The expression of the collagen alpha 1(I) gene in activated stellate cells plays an important role during liver fibrogenesis. To identify the critical cis-elements of the collagen alpha 1(I) gene in stellate cells, we used transgenic animals bearing various collagen alpha 1(I) regulatory regions directing the expression of either a human growth hormone minigene or the bacterial beta-galactosidase gene. We found that collagen alpha 1(I)-human growth hormone transgene expression was constitutively high in tendon and skin, provided the transgene contained the -2.3 to -0.44 kb collagen regulatory region. However in the liver, expression was stimulated several-fold, as was the endogeneous gene, by the fibrogenic hepatotoxin carbon tetrachloride. This stimulation occurred whether the collagen 5' regulatory region extended -2.3, -1.6 or -0.44 kb, and in the presence or absence of much of the first intron (+292 to +1607 bp). In addition, the -0.44 kb 5' region was sufficient for high-level transgene expression in stellate cells, following their activation by culture on plastic. In contrast, in skin and tendon, high-level transcription of the collagen alpha 1(I) gene required the -2.3 to -0.44 kb 5' flanking region. Thus, two different cis-regulatory regions direct cell-specific transcription of the collagen alpha 1(I) gene in stellate cells and in skin and tendon.

Tumor necrosis factor-alpha inhibits albumin gene expression in a murine model of cachexia.

The mechanisms responsible for decreased serum albumin levels in patients with cachexia-associated infection, inflammation, and cancer are unknown. Since tumor necrosis factor-alpha (TNF alpha) is elevated in cachexia-associated diseases, and chronic administration of TNF alpha induces cachexia in animal models, we assessed the regulation of albumin gene expression by TNF alpha in vivo. In this animal model of cachexia, Chinese hamster ovary cells transfected with the functional gene for human TNF alpha were inoculated into nude mice (TNF alpha mice). TNF alpha mice became cachectic and manifested decreased serum albumin levels, albumin synthesis, and albumin mRNA levels. However, even before the TNF alpha mice lost weight, their albumin mRNA steady-state levels were decreased approximately 90%, and in situ hybridization revealed a low level of albumin gene expression throughout the hepatic lobule. The mRNA levels of several other genes were unchanged. Hepatic nuclei from TNF alpha mice before the onset of weight loss were markedly less active in transcribing the albumin gene than hepatic nuclei from control mice. Therefore, TNF alpha selectively inhibits the genetic expression of albumin in this model before weight loss.

Protein kinase A and C site-specific phosphorylations of LAP (NF-IL6) modulate its binding affinity to DNA recognition elements.

LAP (NF-IL6 or C/EBP beta), is a liver transcriptional activator protein that confers liver-specific gene expression. Because LAP has a characteristic phosphoacceptor sequence for cAMP-dependent protein kinase A (PKA), we tested if in vitro phosphorylation of LAP by PKA modulates its interaction with specific DNA sequences. The major PKA phosphorylation site of LAP was identified as Ser105, which is a predicted PKA site. As expected, this PKA phosphorylation site disappears after mutation of Ser105 to Ala. Kinetic studies with LAP and LAP Asp105 (which mimics a phosphoserine residue) demonstrated that phosphorylation of Ser105 itself has no effect on DNA binding. Phosphorylation of other sites by PKA, identified in the region between Ser173 and Ser223 and at Ser240, by analysis of truncated and mutated LAP peptides, resulted in an inhibition of DNA binding. LAP was also phosphorylated by purified protein kinase C in vitro, and the major phosphoacceptor was shown to be Ser240 within the DNA-binding domain of LAP. Phosphorylation of LAP at this residue or introduction of a Ser240 to Asp mutation resulted in marked decrease in its binding to DNA. These results suggest that site-specific phosphorylations of LAP modulate transactivation of its target genes.

Ascorbic acid stimulation of collagen biosynthesis independent of hydroxylation.

Ascorbic acid stimulates collagen gene expression in cultured fibroblasts but mechanisms responsible for this effect are poorly understood. In the presence of the transitional metal iron, ascorbic acid could induce lipid peroxidation with the formation of reactive aldehydes. Because another aldehyde, acetaldehyde, the first metabolite of ethanol, also stimulates collagen transcription in cultured fibroblasts, we investigated whether ascorbic acid induces lipid peroxidation in cultured cells and if this is the mechanism by which ascorbic acid stimulates collagen gene expression. Ascorbic acid (0.2 mmol/L) induced lipid peroxidation and stimulated collagen alpha 1(I) gene transcription in cultured human fibroblasts. Inhibition of the ascorbic acid-induced lipid peroxidation in cultured human fibroblasts with alpha-tocopherol (50 mumol/L) or methylene blue (10 mumol/L) prevented the stimulation of collagen gene expression. Addition of malondialdehyde (200 mumol/L), a product of lipid peroxidation, to cultured human fibroblasts also increased two- to threefold collagen production and procollagen alpha 1(I) mRNA levels. Thus, ascorbic acid induces lipid peroxidation and reactive aldehydes, and this step may be necessary for stimulation of collagen gene expression by ascorbic acid in cultured human fibroblasts.

Inhibition of collagen synthesis in human episcleral fibroblasts by calcium ionophore a23187.

PURPOSE: To test the ability of calcium ionophore A23187 to specifically inhibit collagen biosynthesis in primary cultures of human episcleral fibroblasts obtained at the time of trabeculectomy. METHODS: Calcium ionophore A23187-treated and untreated (control) cultures were incubated with radiolabeled proline. Radioactivity of collagenasesensitive and -resistant proteins was assayed to determine the rate of collagen production. RESULTS: The relative rate of collagen production as compared to the percent of total protein production decreased by 65% in the cultures treated with 2.0 muM ionophore. Additionally, there was a 78% decrease in absolute collagen production in the treated group. CONCLUSIONS: Calcium ionophore A23187 inhibits collagen biosynthesis in cultured human episcleral fibroblasts. It is possible that inhibition of collagen biosynthesis could enhance treatment of ocular fibroproliferative disorders, including episcleral fibroproliferation, a leading cause of glaucoma filtering surgery failure.

Selective inhibition of collagen synthesis by the Ca2+ ionophore A23187 in cultured human fibroblasts.

The question of whether the Ca2+ ionophore A23187 affects collagen production relative to total protein synthesis or has possible effects on collagen degradation was investigated. Cultured normal human fibroblasts were incubated with radioactive proline, and the radioactivity of collagenase-sensitive and -resistant proteins was used to calculate the rates of protein production. The net production of collagen relative to total proteins was inhibited by A23187 in a dose-related manner, and 50% inhibition of basal collagen production was achieved with 0.6 microM A23187. There was a 70% decrease in the absolute rate of collagen production in the presence of 0.6 microM A23187 which represented a 4-fold greater inhibition of collagen production than of noncollagen protein production. The major mechanism for the decreased net production of collagen was decreased synthesis, rather than increased degradation. Ca2+ mobilization induced by cholecystokinin octapeptide was also associated with selective inhibition of collagen production in normal human fibroblasts. These studies establish that the Ca2+ ionophore A23187 induces a selective decrease in collagen polypeptide synthesis by normal human fibroblasts and suggest a modulatory role of Ca2+ on collagen metabolism.

Hepatic collagen production in the rat is unaffected by methotrexate.

Methotrexate (MTX) has been implicated in the pathogenesis of hepatic fibrosis. However, no information exists regarding the effects of MTX on hepatic collagen metabolism. Therefore, we studied the role of MTX in hepatic collagen production in vivo in rats receiving an 8-week course of varying doses of MTX. Twenty-four hours prior to sacrifice animals received an injection of [5-3H]proline. Collagen was extracted with hot trichloroacetic acid and the proteinbound [3H]hydroxyproline was used as a measure of de novo collagen production. The hepatic collagen content was essentially the same in the control and treatment groups in spite of evidence of hepatotoxicity. Similarly, no significant differences were present among the control and MTX-treated groups in the de novo absolute collagen production. In summary, we found no evidence of increased hepatic fibrogenesis in small groups of animals after 8 weeks of treatment with MTX. Data clearly supporting the claim that MTX itself is responsible for hepatic fibrosis are lacking.

[3H]tryptophan-[14C]proline dual label method for the simultaneous determination of collagen and noncollagen protein production.

A new method for the simultaneous determination of newly synthesized collagen and noncollagen proteins has been developed. Because tryptophan is not found in collagen noncollagen proteins were specifically labeled with [3H]tryptophan. [14C]Proline was used to label both groups of proteins. To calculate the 14C-labeled noncollagen protein the 3H radioactivity of the protein mixture was divided by the ratio of 3H:14C in noncollagen protein of a representative sample. This value was obtained by collagenase digestion. The remaining 14C radioactivity in the protein mixture was attributed to [14C]collagen. There was a very good correlation between the dual label method and the widely used collagenase digestion method for the measurement of collagen and noncollagen protein production and for the calculation of the relative rate of collagen synthesis. This new method provides a simple and accurate analysis of collagen production, and it is suitable for rapid processing of a large number of biological samples.

Muscle wasting and dedifferentiation induced by oxidative stress in a murine model of cachexia is prevented by inhibitors of nitric oxide synthesis and antioxidants.

Muscle wasting is a critical feature of patients afflicted by AIDS or cancer. In a murine model of muscle wasting, tumor necrosis factor alpha (TNF alpha) induces oxidative stress and nitric oxide synthase (NOS) in skeletal muscle, leading to decreased myosin creatinine phosphokinase (MCK) expression and binding activities. The impaired MCK-E box binding activities resulted from abnormal myogenin-Jun-D complexes, and were normalized by the addition of Jun-D, dithiothreitol or Ref-1, a nuclear redox protein. Treatment of skeletal muscle cells with a phorbol ester, a superoxide-generating system, an NO donor or a Jun-D antisense oligonucleotide decreased Jun-D activity and transcription from the MCK-E box, which were prevented by antioxidants, a scavenger of reducing equivalents, a NOS inhibitor and/or overexpression of Jun-D. The decreased body weight, muscle wasting and skeletal muscle molecular abnormalities of cachexia were prevented by treatment of TNF alpha mice with the antioxidants D-alpha-tocopherol of BW755c, or the NOS inhibitor nitro-L-arginine.

D-Galactosamine hepatotoxicity is associated with endotoxin sensitivity and mediated by lymphoreticular cells in mice.

Two strains of mice (C57BL/10ScN and C3H/HeJ) that carry the same mutant lipopolysaccharide gene (Lpsd) which makes them resistant to the toxic effects of endotoxin (LPS) are also partially resistant to the hepatotoxic effects of D-galactosamine. As measured by serum alanine aminotransferase, the degree of liver injury induced by D-galactosamine in the LPS-resistant strains is only 10%-30% that of closely related strains of LPS-sensitive mice. Similarly, histopathologic changes are less pronounced in the endotoxin-resistant strains than in LPS-susceptible mice. By transferring spleen cells from LPS-susceptible strains to lethally irradiated, LPS-resistant mice, we established that susceptibility to D-galactosamine is mediated by lymphoreticular cells. Radiation-resistant spleen cells transferred D-galactosamine sensitivity, suggesting a role for macrophages. We did not exclude the possibility that lymphocytes can also transfer the response to D-galactosamine. These results establish that in mice, D-galactosamine sensitivity is associated with endotoxin sensitivity and that the former is mediated by lymphoreticular cells, not by hepatocytes.

TGF-beta and collagen-alpha 1 (I) gene expression are increased in hepatic acinar zone 1 of rats with iron overload.

We have shown that lipid peroxidation stimulates collagen-alpha 1 (I) gene transcription in cultured cells. Because iron is a transitional metal known to induce lipid peroxidation, we investigated whether hepatic lipid peroxidation modulates collagen gene expression in iron-overloaded rats. In this animal model of hemochromatosis, we show colocalization with iron in the hepatic acinar zone 1 of both lipid peroxidation and increased collagen-alpha 1 (I) transcripts, using immunohistochemistry for malondialdehyde-protein adducts and in situ hybridization, respectively. Iron overload stimulated the expression of the cytokine transforming growth factor-beta (TGF-beta) in acinar zone 1, in spite of the minor degree of hepatocellular necrosis and inflammation. The formation of reactive aldehydes and TGF-beta, both inducers of collagen gene expression, may play a role in the stimulation of hepatic collagen production in iron overload. These mechanisms could be a link between iron overload and fibrosis in genetic hemochromatosis.

Acetaldehyde increases collagen gene transcription in cultured human fibroblasts.

Acetaldehyde, the first metabolite of ethanol, mediates many of the biological effects of ethanol. We have previously shown that acetaldehyde, but not ethanol, stimulates collagen production in cultured human fibroblasts (Holt, K., Bennett, M., and Chojkier, M. (1984) Hepatology 4, 843-848). Here, we examined the effects of acetaldehyde on collagen gene expression. Confluent human fetal fibroblasts were incubated for up to 4 h in the presence of ascorbate (0.2 mM) alone or with the addition of either ethanol (12 mM) or acetaldehyde (200 microM). Acetaldehyde induced the production of collagen (up to 2.5-fold) and had a small inhibitory effect on procollagen secretion (-20%). The steady-state levels of mRNAs were measured by hybridizing total cellular RNA to specific cDNA probes at high stringency. Acetaldehyde increased the steady-state level of collagen alpha 1(I) and collagen alpha 2(I) mRNAs about 3-fold and had small effects on beta-actin mRNA (+50%) and collagenase mRNA (-50%). Northern blots revealed that the RNAs were intact and that acetaldehyde preferentially increased the abundance of the longer of the two collagen alpha 1(I) transcripts. Acetaldehyde increased both collagen alpha 1(I) and collagen alpha 1(III) transcriptional activity by 2.5-fold and had small effects on beta-actin and collagenase gene transcription. The increase in both collagen production and collagen mRNA levels induced by acetaldehyde was blocked by methylene blue, a scavenger of reducing equivalents. These data indicate that reducing equivalents, which enhance the formation and stability of acetaldehyde-protein adducts, may be required for acetaldehyde-stimulated collagen production. Thus, this study suggests that acetaldehyde increases collagen production by increasing collagen gene transcription in cultured human fibroblasts.