SEARCH

SEARCH BY CITATION

Keywords:

  • alcohol;
  • hepatic stellate cells;
  • multidisciplinary approaches;
  • protein networks

Summary

  1. Top of page
  2. Summary
  3. Disease
  4. Cells
  5. Fibrillar collagen
  6. Alcohol, a causative factor
  7. Systems biology
  8. Connecting with collagen fibrils
  9. Still in need of fine tuning to determine the contribution of each specific cell type
  10. Is there a way out?
  11. Human studies
  12. Drawbacks
  13. Conclusion
  14. Acknowledgements
  15. References

Among the pathogenesis and risk factors of alcoholic liver disease (ALD) are the source of dietary fat, obesity, insulin resistance, adipokines and acetaldehyde. Translocation of Gram-negative bacteria from the gut, the subsequent effects mediated by endotoxin, and the increased production of matricellular proteins, cytokines, chemokines and growth factors, actively participate in the progression of liver injury. In addition, generation of reactive oxygen and nitrogen species and the activation of non-parenchymal cells also contribute to the pathophysiology of ALD. A key event leading to liver damage is the transition of quiescent hepatic stellate cells into activated myofibroblasts, with the consequent deposition of fibrillar collagen I resulting in significant scarring. Thus, it is becoming clearer that matricellular proteins are critical players in the pathophysiology of liver disease; however, additional mechanistic insight is needed to understand the signalling pathways involved in the up-regulation of collagen I protein. At present, systems biology approaches are helping to answer the many unresolved questions in this field and are allowing to more comprehensively identify protein networks regulating pathological collagen I deposition in hopes of determining how to prevent the onset of liver fibrosis and/or to slow disease progression. Thus, this review article provides a snapshot on current efforts for identifying pathological protein regulatory networks in the liver using systems biology tools. These approaches hold great promise for future research in liver disease.


Abbreviations
ALD,

alcoholic liver disease;

ECM,

extracellular matrix;

HSC,

hepatic stellate cells.

Disease

  1. Top of page
  2. Summary
  3. Disease
  4. Cells
  5. Fibrillar collagen
  6. Alcohol, a causative factor
  7. Systems biology
  8. Connecting with collagen fibrils
  9. Still in need of fine tuning to determine the contribution of each specific cell type
  10. Is there a way out?
  11. Human studies
  12. Drawbacks
  13. Conclusion
  14. Acknowledgements
  15. References

The incidence of liver disease is growing worldwide; thus, there is a pressing need for new therapies to avoid disease progression. Alcoholic liver disease (ALD) is one of the major causes of liver disease in Western countries. It arises from binge drinking or chronic alcohol consumption. Because the liver is the chief organ responsible for metabolizing alcohol, it is especially vulnerable to alcohol-driven injury. ALD includes several pathological conditions such as fatty liver, alcoholic hepatitis, fibrosis and cirrhosis.

Fatty liver, which occurs after acute alcohol ingestion, is generally reversible with abstinence and is not believed to predispose to any chronic form of liver disease if abstinence or moderation is maintained. Alcoholic hepatitis is an acute form of alcohol-induced liver injury that occurs with the consumption of a large quantity of alcohol over a prolonged period of time; it encompasses a spectrum of severity ranging from asymptomatic derangement of biochemistries to fulminant liver failure and death.

Fibrosis is the liver wound-healing response involving a wide array of cell types, multiple unknown pathways and mediators. Although acute liver injury can set in motion fibrogenesis, the continuous signalling related to chronic liver disease caused by alcohol consumption, viral infection, drug toxicity, metabolic disorders or immune hits is necessary for considerable fibrosis to occur. Fibrosis initiates in response to hepatocellular damage, with inflammatory infiltrates further magnifying disease progression. The onset of liver fibrosis usually takes years because of the regenerative potential of the organ.

Cirrhosis because of chronic liver disease is characterized by increased scar formation replacing the normal tissue, disarray of the hepatic parenchyma, altered blood flow, portal hypertension, ascites, splenomegaly, hepatic encephalopathy and liver failure. Cirrhosis results from a disordered development of the regenerative process leading to significant hepatocyte proliferation in regenerative nodules. Hepatocellular carcinoma may ultimately emerge within a background of persistent liver injury, inflammation and hepatocellular proliferation, as it occurs in chronic hepatitis and in cirrhosis.

Cells

  1. Top of page
  2. Summary
  3. Disease
  4. Cells
  5. Fibrillar collagen
  6. Alcohol, a causative factor
  7. Systems biology
  8. Connecting with collagen fibrils
  9. Still in need of fine tuning to determine the contribution of each specific cell type
  10. Is there a way out?
  11. Human studies
  12. Drawbacks
  13. Conclusion
  14. Acknowledgements
  15. References

Although it is recognized that cells such as portal fibroblasts, circulating fibrocytes and bone marrow-derived cells play a role in liver fibrosis, hepatic stellate cells (HSC) are still considered the main source of extracellular matrix (ECM) in liver fibrosis accounting for as much as 80% of the total fibrillar collagen in the pathological state. While HSC are involved during the first stages of disease development, portal fibroblasts may intervene later when the regenerative processes are completely exhausted. The quality of the ECM deposited by both cell types is not identical because along with cross-linked collagen there is portal fibroblast-derived elastin, a good marker for irreversibility of the disease (1, 2). Nevertheless, HSC play a central role in the response to liver injury in vivo by undergoing a transition from a quiescent vitamin-A storing cell to a contractile, proliferative and highly profibrogenic myofibroblast with significant ability to produce fibrillar collagen I, the central protein in the pathogenesis of liver fibrosis.

Fibrillar collagen

  1. Top of page
  2. Summary
  3. Disease
  4. Cells
  5. Fibrillar collagen
  6. Alcohol, a causative factor
  7. Systems biology
  8. Connecting with collagen fibrils
  9. Still in need of fine tuning to determine the contribution of each specific cell type
  10. Is there a way out?
  11. Human studies
  12. Drawbacks
  13. Conclusion
  14. Acknowledgements
  15. References

Fibrillar collagen I is the most abundant protein of the human body. The tropocollagen subunits (or molecular components of the collagen fibre, consisting of three polypeptide chains coiled around each other) self-assemble with regularly staggered ends, into even larger arrays in the extracellular space of tissues. It is highly present in the scar, the end-product when tissue heals by repair.

Pathological collagen I production by HSC and other cells, a key feature in liver fibrosis, occurs following multiple pro-fibrogenic stimuli such as alcohol (3), reactive oxygen species (4–8), lipopolysaccharide (9, 10), phagocytosis of apoptotic bodies (11, 12), cytokines and growth factors (13, 14). Detailed analyses for understanding the molecular basis of the collagen I gene regulation have revealed a complex process involving reactive oxygen species as key mediators (4–8). Less is known, however, on the contribution of reactive nitrogen species (15). In addition, a series of cytokines, growth factors and chemokines, which activate ECM-producing cells through paracrine and autocrine loops, contribute to the fibrogenic response. For a summary of key papers describing the molecular mechanisms that regulate collagen synthesis and turnover the reader is referred to a comprehensive review (16) and to other publications (3, 5, 17–29). A snapshot of the main mechanisms is provided in Figure 1.

image

Figure 1.  Mechanisms that regulate collagen synthesis and turnover in alcoholic liver disease (ALD). The cascade of proteases leading to collagen I regulation is shown. A key role for metalloproteinase (MMP)1, 2, 9 and 13 along with MT1-MMP is suggested. In addition, during ALD multiple mediators participate in regulating collagen I expression both at the transcriptional and translational level. Among the profibrogenic factors are acetaldehyde and reactive oxygen species (ROS) generated during alcohol metabolism, multiple cytokines and growth factors, fee iron, which enhances lipid peroxidation (LPO) reactions particularly in the presence of polyunsaturated fatty acids (PUFAs), apoptotic bodies, cyclooxigenase (COX) metabolites, leptin, endotoxin and angiotensin II. Conversely, reactive nitrogen species, tumour necrosis factor (TNF)-α, S-adenosyl-methionine and polyenylphosphatidylcholine have been shown to prevent collagen accumulation both at the transcriptional and translational level. TGF, transforming growth factors; TIMP, tissue inhibitors of metalloproteinases; PDGF, platelet-derived growth factor; PAI, plasminogen activator inhibitor.

Download figure to PowerPoint

Alcohol, a causative factor

  1. Top of page
  2. Summary
  3. Disease
  4. Cells
  5. Fibrillar collagen
  6. Alcohol, a causative factor
  7. Systems biology
  8. Connecting with collagen fibrils
  9. Still in need of fine tuning to determine the contribution of each specific cell type
  10. Is there a way out?
  11. Human studies
  12. Drawbacks
  13. Conclusion
  14. Acknowledgements
  15. References

Although, as indicated above, many conditions may lead to a fibrogenic response, this review focuses mostly on the effects mediated by alcohol. The hepatic metabolism of alcohol occurs largely by several enzymatic systems: alcohol dehydrogenase, the microsomal ethanol oxidizing system catalysed by cytochrome P450 2E1, catalase and a nonoxidative pathway catalyzed by fatty acid ethyl ester synthase.

Changes in the nicotinamide adenine dinucleotide (NAD+)/NADH ratio because of alcohol metabolism via alcohol dehydrogenase play a major role in the intermediary metabolism in the liver. As the liver adapts to continued ethanol exposure, the functional consequences of ethanol intake involve more than simply changes in the hepatocyte redox state (30). Cytochrome P450 2E1-derived reactive oxygen species increase, hence promoting liver damage. In addition, alcohol also stabilizes cytochrome P450 2E1 protein against degradation; thus, generating a vicious circle (30). Continuous ethanol consumption also impairs liver function by independent pathways, via direct interaction of ethanol with membrane proteins and receptors involved in signal transduction, or alternatively by the solvent actions of ethanol per se, which affect membrane fluidity (31). Each of these complex mechanisms translates into changes in gene expression within the liver, leading to the onset and progression of ALD.

Alcohol-induced liver injury involves both parenchymal and non-parenchymal cells as well as recruitment of inflammatory cells. Damaged hepatocytes, Kupffer cells and biliary epithelial cells release pro-inflammatory cytokines and soluble factors that promote Kupffer cell activation and stimulate the recruitment of activated T cells. The pro-inflammatory environment leads to a profibrogenic phenotype mostly in HSC but also in other cells (Fig. 2). If liver injury persists, accumulation of activated HSC, portal fibroblasts and other cells occurs, with the subsequent secretion of large amounts of ECM proteins, and although there is active matrix remodelling, it is still insufficient, and thus significant scarring occurs because of accumulation of fibrillar collagen and other ECM proteins (32). Thus, ethanol metabolism disarrays the well-regulated wound healing response, resulting in continued tissue damage, inflammation and eventually fibrosis.

image

Figure 2.  Model for the crosstalk among liver cells leading to a fibrogenic response. Hepatocytes, Kupffer cells and biliary epithelial cells release a significant amount of cytokines, growth factors and reactive oxygen species (ROS), all of which can contribute to recruitment of activated T cells. In addition, these mediators secreted by these cells can further amplify the activated state of Kupffer cells. In liver injury, hepatocytes, activated Kupffer cells and biliary epithelial cells affect hepatic stellate cells (HSC), contributing to their activation, transition to myofibroblasts and to extracellular matrix (ECM) deposition.

Download figure to PowerPoint

Alcohol-mediated liver fibrosis can be progressive and fatal, especially if there is no cessation of alcohol drinking. In the past few years, there has been remarkable progress in our understanding of the biological, pathological, genetical and environmental factors that increase the occurrence of ALD, and how they affect the progression to fibrosis and cirrhosis. For comprehensive reviews on these areas, the reader is referred to work by Day and colleagues (33–37). However, there is still a pressing need for further clarification of the molecular events, the involved signalling pathways and the protein regulatory networks, essential for the development of liver fibrosis with particular emphasis on collagen I deposition, so that efficient pharmacological targeting can be designed; thus, new state-of-the-art approaches should be useful to gain additional insight. In this review article, we describe as an example the use of a systems biology strategy.

Systems biology

  1. Top of page
  2. Summary
  3. Disease
  4. Cells
  5. Fibrillar collagen
  6. Alcohol, a causative factor
  7. Systems biology
  8. Connecting with collagen fibrils
  9. Still in need of fine tuning to determine the contribution of each specific cell type
  10. Is there a way out?
  11. Human studies
  12. Drawbacks
  13. Conclusion
  14. Acknowledgements
  15. References

A new concept

Systems biology is the study of an organism or a disease, viewed as an integrated and interacting network of genes, proteins and biochemical reactions. Instead of analysing individual components or aspects of the organism or disease, systems biologists focus on all the components and the interactions among them, all as part of one system. These interactions are ultimately responsible for an organism or disease alterations.

Systems biology is an emergent field of research that intends to create a novel interactive and interdisciplinary scientific approach, in which data are integrated in new formulations that allow predicting or hypothesizing the basis of disease. Thus, systems biology refers to multidisciplinary approaches designed to uncover emergent properties or deregulations of biological systems.

Tools

A central tool in systems biology is the use of computational modelling to reconstruct complex systems from a wealth of reductionist molecular data (e.g. gene/protein expression, signal transduction pathways, metabolic activity, etc.). A number of deterministic, probabilistic and statistical learning models are used to understand sophisticated cellular behaviors such as protein expression during disease and the activity of signalling networks. The benefits of the systems biology approach include technology development, advances in basic concepts of biology and disease, and real-world practical applications such as predictive and preventive medicine. The goal of the example listed in this review is to target a particular health problem (i.e. the protein network leading to enhanced fibrillar collagen I in ALD) with a novel integrative approach.

Connecting with collagen fibrils

  1. Top of page
  2. Summary
  3. Disease
  4. Cells
  5. Fibrillar collagen
  6. Alcohol, a causative factor
  7. Systems biology
  8. Connecting with collagen fibrils
  9. Still in need of fine tuning to determine the contribution of each specific cell type
  10. Is there a way out?
  11. Human studies
  12. Drawbacks
  13. Conclusion
  14. Acknowledgements
  15. References

There is limited information on how potential protein interaction networks activated by chronic ethanol consumption may affect collagen I expression and lead to liver injury. This review provides an example of an approach that may hold great potential for uncovering mechanisms and signalling pathways in many liver diseases associated with collagen I deposition or with any other factors.

Over the past years, some laboratories including ours have contributed to understanding the mechanisms underlying collagen I regulation in ALD and in liver fibrosis. Several studies have been carried out on the crosstalk among liver cells and the fibrogenic response to liver injury by using cellular and molecular biology tools (4–6, 38) (Fig. 2).

The field is now moving into more holistic approaches using systems biology strategies in combination with computer platforms. As indicated above, in this review we provide a snapshot of current efforts to determine the protein regulatory network that regulates pathological fibrillar collagen I deposition by HSC under ethanol consumption. Specifically, proteins participating in collagen I induction in HSC following chronic ethanol consumption, as well as the protein regulatory network and the signalling cascades involved have not been identified and systematically characterized; thus, a database of this relevant and biologically active subproteome is still missing. Access to such database, validated also in other laboratories, would be of paramount relevance for pharmacological design.

To establish an appropriate in vivo model for building the systems biology database, our laboratory has used the Lieber–DeCarli feeding model, which utilizes equicaloric liquid diets with the same composition with regard to fat and protein, but differ in the amount of ethanol-derived calories. These diets have been used for decades and are a well established model for early alcohol-induced liver injury (39).

Hepatic stellate cells were isolated by in situ liver perfusion and gradient centrifugation (4). After confirming that HSC from chronic ethanol-fed rats were activated, profibrogenic and generated more reactive oxygen species than those from control rats, we analyzed the differentially expressed proteins by using isotope-coded affinity tags, proteolytic digestion of the protein mixture into shorter peptide fragments and sample fractionation into multiple fractions using a strong cation-exchange column. The fractionated peptides, whose cysteine residues were alkylated by the reagents carrying biotin groups, were purified by an avidin cartridge followed by acid cleavage to remove the biotin-containing linker for better identification in the mass spectrometry-based quantitative proteomics analysis.

The cleaved peptides mixtures were subjected to microcapillary HPLC and analysed by linear ion trap mass spectrometry. Software tools available at the Seattle Proteome Center (Institute for Systems Biology, Seattle) were used at each step of the data analysis. The steps that were taken were: (i) peptide tandem mass spectra database search using sequest (The Scripps Research Institute, La Jolla, CA, USA) (40, 41); (ii) assignment of a probability score to the identified peptide sequence using peptideprophet (Institute for Systems Biology, Seattle, WA, USA) (41); (iii) assignment of an overall probability to the protein identified using proteinprophet (Institute for Systems Biology) (41); (iv) calculation of isotope-coded affinity tags ratio using the automated statistical analysis on protein ratio and express software; (v) use of the systems-based data analysis Gaggle platform (42) for comprehensive understanding of biological networks.

The ability to globally profile changes in protein abundance is essential for elucidating the events that occur during cellular processes and diseased states. With the acquired data from the global comparative quantitative protein profiling analysis, we are now gearing towards visualizing and evaluating the identified and quantified proteins in the context of biological pathways and protein interaction networks related to collagen I in ALD, the downstream target in this study.

We are using the systems-based data analysis platform http://pipe.systemsbiology.net/pipe/, an integrated computational platform consisting of multiple independently developed software tools including cytoscape (Institute for Systems Biology; http://www.cytoscape.org), the standard software for the graphical exploration of networks, r statistical environment (Institute for Systems Biology) and data matrix viewer (Institute for Systems Biology) for loading, navigating and plotting the high-throughput proteomics data. This platform integrates multiple features of the data mining approaches from various international bioinformatics resources such as KEGG (http://www.genome.jp/kegg/), embl string (http://string.embl.de/), niaid david (http://david.abcc.ncifcrf.gov/ease/ease.jsp), and the Gene Ontology Miner (http://discover.nci.nih.gov/gominer/).

The acquired experimental data in this study was then examined on a network of protein–protein interactions, using the embl's string. Additional protein associations were added from BOND, (http://bond.unleashedinformatics.com/) and HPRD (http://www.hprd.org/). Metabolic pathways were learnt from the KEGG and Gene Ontology, and by combining the Bioconductor ‘Category’ package with cytoscape, statistically significant enrichment for a specific biological function was found and explored. Thus, with the judicious combination of data from many sources, a list of ethanol-regulated proteins in HSC can be understood as a network rich in biological meaning, such as the effect of ethanol on the liver's profibrogenic response. In this way, the protein regulatory network for the ethanol-mediated effects on HSC was generated and the downstream target was visualized. Once key proteins or biomarkers are identified, their role in the development of liver fibrosis could be validated in vivo using null mice.

As a logical corollary, the major goal in protein network modelling is the ability to model and/or predict downstream effects on the specific protein network when one node (representing one protein or one protein group) in the network is perturbed (e.g. as it may occur after chronic ethanol consumption), ultimately aiming at improved targeting. If we can predict the effect of such perturbation, then we can evaluate the virtue of a potential target benefited from the effect (i.e. fibrillar collagen I), when the entire system is considered. In other words, if we wish to elicit certain long-term behaviour from a network (e.g. the collagen I network in HSC), what proteins would make the best candidates for intervention to increase the likelihood of this behavior? If we suppose that the network is operating in a certain ‘undesirable’ state (e.g. increased fibrillar collagen I deposition by HSC under alcohol consumption), and we wish to transition it into a ‘desirable’ state (e.g. collagen I down-regulation in HSC) by perturbing some protein; for practical reasons, it would be a much better option to intervene with as few proteins as possible in order to achieve the goal. Such approach can expedite the systematic search and the identification of potential drug targets in therapy.

Ultimately, predictions will need to be tested in an experimental system using a biochemical approach, whereby by knocking down the expression of a ‘critical’ protein (e.g. with a neutralizing antibody, a specific chemical inhibitor, RNA interference or using cells isolated from null mice), we could validate whether the identified protein and/or signalling pathway is really ‘critical’. Ultimately, work with archived tissue samples from healthy individuals, alcoholic and/or fibrotic patients at different disease stages will help to identify biomarkers and even predict outcome.

A systems biology approach would add clarity as it would facilitate the identification of pathways worth targeting, and at the same time reflect the impact of the target on the overall process. We expect that proteomics expression profiling in combination with studies using human liver tissue by laser capture microdissection, confocal microscopy, genomics and other state-of-the-art strategies will help to address key questions in the near future.

Thus, we provide here an example where multidisciplinary approaches involving in vivo models of disease (ALD), proteomics, theoretical and computational analysis and engineered systems merge with challenging experiments, using molecular and cellular biology techniques, to tackle an unanswered question (e.g. the pathways leading to fibrillar collagen I up-regulation in ALD). There are still unresolved issues outlined below related to the fibrogenic response in ALD, which may eventually be addressed using tools that lie within the newly developed encouraging field of systems biology.

Still in need of fine tuning to determine the contribution of each specific cell type

  1. Top of page
  2. Summary
  3. Disease
  4. Cells
  5. Fibrillar collagen
  6. Alcohol, a causative factor
  7. Systems biology
  8. Connecting with collagen fibrils
  9. Still in need of fine tuning to determine the contribution of each specific cell type
  10. Is there a way out?
  11. Human studies
  12. Drawbacks
  13. Conclusion
  14. Acknowledgements
  15. References

The discovery of the phenomenon of HSC activation contributed significantly to our understanding of liver fibrosis. A growing body of evidence suggests that HSC, although still considered the main fibrillar collagen-producing cell, with a pivotal role in orchestrating hepatic immune responses, is not the only cell type with fibrogenic potential; thus, pharmacological targeting is becoming more challenging.

There are a number of subclasses of potential hepatic fibrogenic cells such as portal fibroblasts, bone marrow-derived cells, and circulating fibrocytes that could play a role in the fibrogenic response. The relative contribution of each cellular source varies with disease progression and aetiology. In this regard, for example, portal fibroblasts appear to be relevant in cholestatic liver disease and in ischaemia as well as in advanced disease and they have been described to support a prominent role in scar formation (1).

The potential for bone marrow origin of liver myofibroblasts, with the increasing evidence of their plasticity, brings additional concepts to this intricate field. Bone marrow-derived cells are widely distributed within the scar in advanced fibrosis, and replace the local recruitment of myofibroblasts over time (43).

Circulating fibrocytes constitute an additional population of cells that has raised significant attention in the past few years (44). They express haematopoietic and myeloid markers (45, 46) and produce fibrillar collagen I and III as well as fibronectin, and their role has been studied in bleomycin-induced pulmonary fibrosis (47, 48). Whether fibrocytes, a small fraction of the collagen I producing cells in liver injury, play a role in fibrosis, remains an open question.

In vivo models have now demonstrated that there is marked and transient proliferation of bile duct epithelial cells, which are associated with activation and proliferation of portal periductular fibroblasts, resulting in the accumulation of peribiliary myofibroblasts. Bile duct epithelial cells will contribute to myofibroblastic trans-differentiation and development of a biliary type of fibrogenic response that can also affect HSC (49, 50).

Lastly, although still rather controversial, importance is now given to epithelial-to-mesenchymal transitions, an orchestrated series of events in which cell–cell and cell–ECM interactions are altered to release epithelial cells from the basement membrane and surrounding tissue, the cytoskeleton is re-organized to confer the ability to move through a three-dimensional ECM, and a new transcriptional programme is induced to maintain the mesenchymal phenotype (51, 52). In addition, there are local mesenchymal stem cells in a niche together with epithelial stem cells (close to the canals of Hering). Additional work is required to dissect whether this mechanism really plays role in liver fibrosis.

Such heterogeneity of hepatic profibrogenic cells suggests that there is more than one target cell type for developing efficient therapies to prevent liver fibrosis. The integration of systems biology data would help to decipher the commonalities underlying the multiple origins of myofibroblasts. Transcriptomics, genomics and even multiple-reaction targeted proteomics studies will reveal potential markers that may be relevant towards identification of the key cells involved in the hepatic fibrogenic response. Thus, novel systems biology approaches at the genomic, transcriptomic and proteomic level are needed.

Is there a way out?

  1. Top of page
  2. Summary
  3. Disease
  4. Cells
  5. Fibrillar collagen
  6. Alcohol, a causative factor
  7. Systems biology
  8. Connecting with collagen fibrils
  9. Still in need of fine tuning to determine the contribution of each specific cell type
  10. Is there a way out?
  11. Human studies
  12. Drawbacks
  13. Conclusion
  14. Acknowledgements
  15. References

The major focus of management of ALD is abstinence from alcohol, supportive care, treatment of complications from infection and portal hypertension, and maintenance of positive nitrogen balance through nutritional support. Corticosteroid therapy should be considered in patients with severe alcoholic hepatitis. Several recent advances in understanding the pathogenesis of ALD may lead to novel future treatment approaches, including inhibition of tumour necrosis factor-α, antioxidant therapy, stimulation of liver regeneration and promoting of collagen degradation.

Currently, the therapeutic repertoire for the treatment of liver fibrosis and cirrhosis seems limited. For severe end-stage disease, liver transplant is the only effective therapy. However, it has disadvantages such as shortage of donors and the presence of concurrent diseases affecting other tissues in the potential recipient. Furthermore, patients may be excluded from transplantation because of medical comorbidities. In the case of hepatitis C-induced cirrhosis, liver transplant is far from curative, as recurrence of hepatitis C in the graft is quite frequent. Hence, there is a pressing need for successful therapy.

Liver injury virtually sets in motion a flow of events whose ultimate aim is repair. Inflammation and subsequent repair are strongly regulated and multifaceted. Significant progress has been made in unraveling the contribution of inflammation to the liver wound-healing response and how inflammatory cell types cross-talk in the evolution and resolution of scarring. Hitherto, it was believed that the replacement of the normal hepatic tissue, with a scar-like matrix composed of cross-linked collagen I was irreversible. However, numerous clinical data suggest that even cirrhosis is potentially reversible.

Apoptosis

New trends of thought suggest that one possibility to improve liver fibrosis is clearing HSC by apoptosis, although a potential concern is that HSC also represent an important source of active proteases. HSC susceptibility to apoptosis increases with their level of activation. Therefore, there is great interest on the pathways and the potential new drugs that initiate HSC apoptosis in a rather selective manner. Systems biology approaches to integrate current available information along with new discoveries may allow advancement in this particular field that remains open.

A number of studies have investigated agents that induce HSC apoptosis such as gliotoxin (53, 54) and sulphasalazine (54). Other known agents such as benzodiazepine ligands, curcumin and tanshinone induce HSC apoptosis, and their therapeutic potential is currently being explored. Degradation of fibrillar collagen is critical for induction of HSC apoptosis during recovery from liver fibrosis (55, 56). Inhibition of integrin αVβ3 results in HSC apoptosis (56), therefore providing essential survival signals for activated HSC.

Animal models using either bile duct ligation or carbon tetrachloride treatment demonstrate that during spontaneous recovery of the normal liver architecture, there is a net reduction of the number of myofibroblasts. Tissue inhibitor of metalloprotease-1 is a potential candidate protein that promotes the survival of activated myofibroblasts (57–59). In liver injury, activated HSC secrete large amounts of tissue inhibitor of metalloprotease-1, which can protect HSC against apoptosis and maintain the fibrogenic environment through the action of integrins. However, this does not detract from the fact that the default pathway for HSC may be apoptosis. With resolution of fibrosis, it is possible that HSC survival signals are obliterated and the cells undergo apoptosis. This then, removes the source of tissue inhibitor of metalloprotease-1, leading to an increase in matrix metalloprotease activity and additional removal of HSC survival signals; thus, further HSC apoptosis may occur, and ultimately histological resolution.

Senescence

A recently discovered mechanism that could participate in the resolution of liver fibrosis is senescence of activated HSC, which poses a barrier to malignant transformation (60). In mice lacking key senescence regulators, HSC continue to proliferate, leading to an enhanced profibrogenic response. Natural killer cells preferentially kill senescent activated HSC in vitro and in vivo, thereby facilitating the resolution of fibrosis (60). Cellular senescence limits the extent of fibrosis following liver damage, as it limits HSC proliferation, up-regulates matrix metalloproteases with fibrolytic activity and facilitates their clearance from the liver, as they produce signals that attract immune cells into fibrotic lesions. All these events underscore the interplay between senescent cells and the tissue microenvironment (60).

Human studies

  1. Top of page
  2. Summary
  3. Disease
  4. Cells
  5. Fibrillar collagen
  6. Alcohol, a causative factor
  7. Systems biology
  8. Connecting with collagen fibrils
  9. Still in need of fine tuning to determine the contribution of each specific cell type
  10. Is there a way out?
  11. Human studies
  12. Drawbacks
  13. Conclusion
  14. Acknowledgements
  15. References

Much of the proteomics effort in the field of hepatology has been focused on studying human liver cancer (61–64), cholangiocarcinoma (65–67) acute liver failure (68), ischaemia-reperfusion injury (69), and non-alcoholic fatty liver disease (70, 71); however, there are limited proteomics studies on human liver fibrosis, and only partial information is available from murine models (72). Recent findings in patients with a variety of hepatic diseases, including hepatitis C virus infection, suggest a clear decline in fatty acid oxidation enzymes (61–64, 73–75). Although a promising avenue, it remains to be resolved if the molecular basis underlying this perturbation is conserved among the various causative agents, leading to chronic liver damage and hepatocellular carcinoma. The apparent link with progression of liver injury and fibrosis make these cellular pathways attractive targets for further investigation. Proteomics is also gaining considerable attention in the area of hepatotoxicity and hepatoprotection (76, 77), with particular emphasis on the role of subsets of proteins from specific organelles such as mitochondria (78).

Serial liver biopsy specimens acquired from patients offers the unique opportunity to study liver disease of different aetiologies. The low protein yields associated with small clinical specimens requires the use of ultrasensitive nanoproteomics tools and platforms. The combination of high-resolution Fourier-transform ion cyclotron resonance mass spectrometry, along with the accurate mass and time-tag strategy, has already proven to be useful in offering the ultra high-sensitivity necessary for analysis of biopsy specimens of post-transplantation livers from hepatitis C virus-infected patients (73). The identification of disease biomarkers or signatures has the potential to improve patient diagnosis, treatment and outcome. In this regard, the use of targeted proteomics may advance the field significantly (79).

Proteomic tactics that characterize the functional state of proteins and their post-translational modifications are also a major avenue of current efforts to understand liver function and disease (80). Oxidative and nitrosative modifications that impair the activity of detoxifying enzymes, contribute to mitochondrial dysfunction and liver injury and have been detected in chronic alcohol consumption (81–83).

Integrated system-wide approaches, including transcriptomics, metabolomics, targeted proteomics and high-throughput techniques, combined with mathematical modelling and computational biology, have great potential for understanding the complexity of liver disease, and translate it into preventive and personalized medicine. A key goal of many of these studies is to utilize the right pipelines to integrate and analyze high-throughput datasets. By collecting information both, at the proteomic and at the genomic level, we can increase the chances of identifying biomarkers of the broad spectrum of liver diseases and of uncovering mechanistic pathways. The wealth of knowledge already acquired from quantitative and functional proteomic studies of the liver, will provide vital assistance in our efforts to identify circulating markers possibly resulting from leakage, secretion or shedding of proteins from diseased liver (84). Ultimately, the data generated from proteomics studies will be very useful when viewed in the context of a more integrated picture.

Drawbacks

  1. Top of page
  2. Summary
  3. Disease
  4. Cells
  5. Fibrillar collagen
  6. Alcohol, a causative factor
  7. Systems biology
  8. Connecting with collagen fibrils
  9. Still in need of fine tuning to determine the contribution of each specific cell type
  10. Is there a way out?
  11. Human studies
  12. Drawbacks
  13. Conclusion
  14. Acknowledgements
  15. References

At present, systems biology approaches suffer from several experimental and computational drawbacks. High-throughput analysis is sensitive to the way in which samples are collected and handled (84). Variability across platforms and between laboratories remains an issue to integrate datasets from different sources (84). An additional challenge entails translating high-throughput data into digested results that can be easily interpreted by a broader audience, including clinicians, government officials and other scientists.

Conclusion

  1. Top of page
  2. Summary
  3. Disease
  4. Cells
  5. Fibrillar collagen
  6. Alcohol, a causative factor
  7. Systems biology
  8. Connecting with collagen fibrils
  9. Still in need of fine tuning to determine the contribution of each specific cell type
  10. Is there a way out?
  11. Human studies
  12. Drawbacks
  13. Conclusion
  14. Acknowledgements
  15. References

The specific example described in this review article supports the advantage of using systems biology approaches. This strategy will allow the scientific community to identify mediator molecules, biomarkers and signalling networks that account for specific modes of liver injury. This approach will continue helping this field of research as a tool for answering many unresolved questions. Future studies will provide information of target proteins and specific cell types involved in liver disease, so that treatments can be developed to prevent and/or reverse liver fibrosis. With the new optimism, that envisions fibrosis as a programmed response to injury, which is dynamic and reversible, we hope to move in a direction that contributes to better development of antifibrotic therapies in the near future, benefiting many patients worldwide.

Acknowledgements

  1. Top of page
  2. Summary
  3. Disease
  4. Cells
  5. Fibrillar collagen
  6. Alcohol, a causative factor
  7. Systems biology
  8. Connecting with collagen fibrils
  9. Still in need of fine tuning to determine the contribution of each specific cell type
  10. Is there a way out?
  11. Human studies
  12. Drawbacks
  13. Conclusion
  14. Acknowledgements
  15. References

Financial support from the US Public Health Service Grants 5R01DK069286-05 and 2R56DK069286-06 from the National Institute of Diabetes and Digestive and Kidney Diseases; 3R01AA017733-03, 3R01AA017733-02S1, 1P20AA017067-03 and 1P20AA017067-02S1 from the National Institute on Alcohol Abuse and Alcoholism (to NN).

References

  1. Top of page
  2. Summary
  3. Disease
  4. Cells
  5. Fibrillar collagen
  6. Alcohol, a causative factor
  7. Systems biology
  8. Connecting with collagen fibrils
  9. Still in need of fine tuning to determine the contribution of each specific cell type
  10. Is there a way out?
  11. Human studies
  12. Drawbacks
  13. Conclusion
  14. Acknowledgements
  15. References