Oxidative stress: Does it ‘initiate’ hepatic stellate cell activation or only ‘perpetuate’ the process?
Article first published online: 30 AUG 2002
Journal of Gastroenterology and Hepatology
Volume 17, Issue 10, pages 1045–1048, October 2002
How to Cite
Apte, M. (2002), Oxidative stress: Does it ‘initiate’ hepatic stellate cell activation or only ‘perpetuate’ the process?. Journal of Gastroenterology and Hepatology, 17: 1045–1048. doi: 10.1046/j.1440-1746.2002.02845.x
- Issue published online: 30 AUG 2002
- Article first published online: 30 AUG 2002
See article in J. Gastoenterol. Hepatol. 2002; 17: 785–90
Oxidative stress is thought to play an important role in the development of hepatic injury from a variety of causes including alcohol abuse, viral infections, biliary obstruction (cholestasis), and iron or copper over-load.1–3 The development of oxidative stress in these conditions is usually a result of increased free radical generation combined with depleted antioxidant defences in the organ. These events lead to a disturbance in cellular homeostasis including DNA damage, activation of the glutathione redox cycle, depletion of adenosine triphosphate and nicotinamide adenine dinucleotide, and peroxidation of membrane lipids.
Lipid peroxidation is a known association of liver fibrosis in chronic liver injury (regardless of etiology).4 Consequently, several substances with antioxidant properties have been examined for their potential antifibrotic effects in both animal and human studies. These include: (i) α-tocopherol (vitamin E), shown to be effective in experimental models of liver fibrosis, although its value in humans is uncertain; (ii) silymarin (a plant flavanoid), shown to be antifibrotic in a rat model of biliary obstruction and also found to be of some benefit in patients with alcoholic liver disease;5,6 (iii) polyenylphosphatidylcholine (PPC), a polyunsaturated phospholipid found to have an antifibrogenic effect in alcoholic liver disease, possibly via reduced induction of the microsomal enzyme cytochrome P4502E1 and reduced oxidative stress;7–9 (iv) S-adenosyl-L-methionine, a substrate for glutathione synthesis shown to attenuate liver fibrosis in experimental liver injury caused by alcohol, carbon tetrachloride (CCL4) or biliary obstruction,10–12 which has also been used in patients with alcoholic liver disease, drug-induced liver disease and primary biliary cirrhosis;13 and (v) compounds such as retinoids14 and the natural phenolic compounds resvatrol and quercetin,15,16 found to reduce liver fibrosis in experimental models.
It is now well-established that fibrogenesis in the liver is mediated by hepatic stellate cells (HSC) that have been activated during liver injury.17 Hepatic stellate cells are located in the perisinusoidal space of Disse and act as storage sites for vitamin A. These cells are a major source of extracellular matrix (ECM) proteins in the liver and are also a source of the enzymes (matrix metalloproteinases) that degrade ECM. Therefore, HSC are thought to play a role in the maintenance of normal ECM in the liver by regulating the balance between its synthesis and degradation.
During liver injury, activation of HSC occurs in two steps.17–19 The first is the ‘initiation’ of activation whereby the cells begin to lose vitamin A, display cytoplasmic extensions and exhibit increased expression of cell surface receptors for growth factors and cytokines. The second step is the ‘perpetuation’ of activation involving a further change in morphology to a myofibroblast-like phenotype, with further loss of vitamin A, increased proliferation, and increased matrix synthesis. The consequence of these morphological and functional changes is a significant imbalance between ECM synthesis and degradation, with deposition of excessive amounts of ECM proteins (particularly fibrillar collagen) in the interstitium.
In vitro studies have identified a number of factors that can activate HSC in culture. 13 These include various growth factors, proinflammatory cytokines, toxic compounds such as acetaldehyde (a major metabolite of ethanol) and CCL4 and pro-oxidant compounds that produce oxidative stress.
Oxidative stress-induced activation of HSC is mediated via the direct effects of reactive oxygen species and/or via highly reactive aldehydic products of lipid peroxidation such as malondialdehyde (MDA) and 4-hydroxynonenal (HNE).4 During liver injury in vivo, HSC may be exposed to these ‘activating’ compounds via two pathways (which need not be mutually exclusive): (i) endogenous exposure as a result of the generation of reactive oxygen species and aldehydes within the cells; or (ii) exogenous exposure as a result of the generation of diffusible reactive species such as HNE and H2O2 by surrounding cells (hepatocytes, Kupffer cells, inflammatory cells).
Experimental evidence exists in the literature to support both these possibilities. Studies using human and/or animal HSC in culture have demonstrated increased proliferation and increased collagen synthesis in cells exposed to pro-oxidant substances such as the ethanol metabolite acetaldehyde, ferrous sulfate/ascorbic acid or ferric nitrilotriacetate.20–22 These compounds have been shown to increase MDA levels within HSC (suggesting increased lipid peroxidation within the cells). Importantly, HSC activation by the pro-oxidant compounds is prevented by pretreatment of the cells with an antioxidant such as D-α-tocopherol (vitamin E). These findings indicate that oxidative stress can be generated endogenously within HSC.
Exogenous addition of lipid peroxidation products to cell culture medium has also been reported to activate HSC. Parola et al.23 have demonstrated increased procollagen I synthesis in human HSC exposed to 4HNE and other hydroxy alkenals of different chain lengths, at concentrations (1 µM) known to occur in vivo under conditions of mild to moderate oxidative stress.24 Similar results have been demonstrated by Tsukamoto et al. with rat HSC.25 Another aldehydic product of lipid peroxidation, MDA, has also been reported to increase collagen production by HSC in culture.26 However, the concentrations of MDA required to exert this effect are 200-fold higher than the effective concentrations of HNE.
Efforts are currently underway to delineate the molecular mechanisms responsible for oxidative stress-induced HSC activation, particularly collagen synthesis in the cells. Using nuclear extracts from culture-activated rat HSC, Lee et al. have demonstrated that oxidative stress-induced HSC proliferation and increased collagen synthesis is associated with induction of transcription factors NFkB and c-myb (which regulate gene expression in mammalian cells).27 More recently, Parola et al.28 reported that the aldehyde HNE forms adducts with the p46 and p54 isoforms of c-Jun amino-terminal kinase in human HSC, leading to the nuclear translocation and activation of these signaling molecules, and, in turn, the induction of the transcription factors c-jun and AP-1. A recent study by Svegliati-Baroni et al. has implicated the Na+/H+ exchanger (an electroneutral transporter) in HSC, in oxidative stress-induced HSC proliferation and collagen synthesis.29
While there is wide acceptance for the concept that oxidative stress-induced HSC activation plays an important role in liver fibrogenesis, what remains controversial is whether oxidative stress can ‘initiate’ the activation of HSC from their quiescent state or whether preactivation of HSC (by other factors such as proinflammatory cytokines and growth factors released during tissue necroinflammation) is essential before oxidative stress can exert its inductive effects. In other words, is oxidative stress an ‘initiator’ or only a ‘perpetuator’ of HSC activation? This issue is particularly relevant to conditions such as alcoholic liver fibrosis and metal overload-induced fibrosis because it brings into focus an important question, that is, is tissue necrosis or inflammation an absolute prerequisite for the stimulation of fibrogenesis in the liver during alcohol abuse or iron/copper overload?
Clinical (in vivo) evidence in support of a non-necroinflammatory pathway of alcohol-induced fibrosis is provided by a recent study describing HSC activation in the absence of hepatitis in liver biopsy specimens from alcoholic patients.30In vitro studies examining the direct effects of ethanol and its metabolite acetaldehyde (known pro-oxidants) also provide evidence in support of an activating effect of these compounds on cultured HSC.3,31–34 However, it must be noted that most of these studies have used pre-activated (passaged) HSC. To address the question as to whether similar effects are detectable in early culture (non-activated) HSC, Maher et al.35 exposed primary HSC cultures (at 3 and 7 days of culture) to 100 µM acetaldehyde for 24 h. The authors report that HSC in primary culture are not activated by acetaldehyde, suggesting that this pro-oxidant compound does not stimulate ‘quiescent’ HSC. However, the study is limited by the fact that only one concentration of acetaldehyde was used (a concentration lower than that known to activate cultured HSC) and that the incubation period was restricted to a relatively short period of 24 h. Our group has recently demonstrated that the pancreatic counterpart of HSC, that is, pancreatic stellate cells (PSC), are activated by exposure to ethanol (10 mM and 50 Mm) or acetaldehyde (150 µM and 200 µM) for 48 h and that this activation is mediated via oxidative stress.36 Furthermore, we have demonstrated that ethanol-induced PSC activation occurs very early during the culture period (using quiescent PSC within 18 h of isolation and culture that were exposed to ethanol for a further 24 h), indicating that preactivation of PSC is not essential for ethanol to exert its stimulatory effect.
Studies examining the direct effects of oxidative stress on cultured HSC provide similarly conflicting findings. Lee et al.27 have reported that quiescent HSC are activated on exposure to the pro-oxidant compound ascorbate/ferrous sulfate or the lipid peroxidation product MDA. The HSC in these experiments were cultured on Matrigel (a basement membrane-like substance known to maintain HSC in their quiescent phase). Similarly, Nieto et al.,37 using a coculture model of E47 hepatocytes which overexpresss cytochrome P4502E1 (an effective producer of oxidative stress) and HSC plated on Matrigel, have demonstrated increased α-smooth muscle actin (α-SMA) expression, increased collagen synthesis and increased proliferation of HSC. This activation of HSC is inhibited by antioxidants and by CYP2E1 inhibitors, suggesting that it is mediated by diffusible mediators including reactive oxygen species generated within the E47 hepatocytes. In contrast to the conclusions of these studies, Maher et al.26 assert that stimulatory effects of oxidative stress (via MDA) only occur after HSC have been preactivated by culture on plastic. However, MDA is known to be significantly less potent than 4HNE (another aldehydic product of lipid peroxidation), which was not examined in the study. It is also possible that early culture HSC are more responsive to the endogenous production of oxidant stress (via exposure to pro-oxidant compounds such as ascorbate/ferrous sulfate), than to the exogenous addition of lipid peroxidation products.
The paper by Olynyk et al. in issue number 7 of the Journal examines the fibrogenic potential of oxidative stress by using HSC in early primary culture exposed to a range of concentrations of the aldehydic lipid peroxidation products HNE and MDA over an incubation period of 3 days or 7 days.38 The authors report a surprising finding, that both MDA and HNE significantly decrease α-SMA expression in HSC at 3 days when compared to controls not incubated with the compounds. It is not clear from the paper as to whether this apparent ‘reversal to quiescence’ of the HSC is accompanied by other features of the ‘resting’ state such as prominent vitamin A-containing lipid droplets and a rounded morphology with few cytoplasmic extensions. The authors also report that, in contrast to the effects on α-SMA expression, MDA and HNE have no significant effect on collagen synthesis. However, there appears to be a trend towards an increase in collagen synthesis with MDA at all concentrations used, and this increase attains statistical significance at an MDA concentration of 10 µM. The reasons for the difference in the effects of the aldehydic products on the two indices of HSC activation (α-SMA (significant reduction) and collagen expression (no change or increase)) are unclear.
An expansion of the experimental protocol of Olynyk et al.38 may address the question as to whether oxidative stress is ‘an initiator’ or ‘a perpetuator’ of HSC activation more comprehensively. Such a protocol would need to involve: (i) the use of activated and non-activated HSC from the same cell preparation, cultured under identical conditions; (ii) the exposure of these cells to exogenous (by addition of reactive oxygen species and lipid peroxidation products to culture medium) and endogenous oxidant stress (addition of pro-oxidant compounds to culture medium) and; (iii) the assessment of a range of indices of HSC activation.
As noted earlier, the ability of oxidative stress to stimulate ‘pre-activated’ HSC is undisputed. However, on careful consideration of the literature to date and with particular reference to the elegant coculture study of hepatocytes and HSC by Nieto et al.,37 it is this author's opinion that the potential of oxidative stress to also activate HSC from their quiescent state cannot be discounted. Oxidative stress may thus play a role in both the initiation as well as the perpetuation of the activation process in HSC.
Why is it important to determine whether oxidative stress is an ‘initiator’ of HSC activation and not just a ‘perpetuator’ of the process? The distinction is clearly important from the point of view of understanding the pathogenesis of liver fibrosis. The possibility that oxidative stress can induce activation of HSC from their normal ‘resting’ state means that in liver diseases where oxidative stress is known to be an important factor, the process of fibrogenesis is likely to be stimulated early in the course of the disease, in the absence of any other ‘insults’ such as overt necroinflammatory episodes. The impact of such a distinction on the therapy of liver fibrosis is less apparent, although it may be postulated that antioxidant treatment introduced early during liver injury may prevent or retard the development of fibrosis by inhibiting oxidative stress-induced HSC activation. In this regard, a multicenter clinical trial of the antioxidant PPC13 is currently underway in patients with alcoholic liver disease, based on the ability of this compound to decrease the accumulation of myofibrobast-like HSC in alcohol-induced liver injury, to inhibit the mitogenic effect of platelet-derived growth factor on HSC, and to stimulate matrix degradation by HSC.9,39 It is possible that this PPC trial will provide the impetus for similar large-scale clinical studies to test other candidate antioxidants (which can inhibit HSC activation) for their antifibrogenic potential in chronic liver disease.