Beyond Scar Formation: Portal Myofibroblast-Mediated Angiogenesis in the Fibrotic Liver

Progressive liver fibrosis, leading to cirrhosis, is the most common cause of liver failure and the hallmark of chronic liver disease. Liver fibrosis is a pathophysiological process triggered within the liver after injury to mediate tissue repair and wound-healing response.¹ Angiogenesis and vascular remodeling are strongly associated with progressive liver fibrosis.² Angiogenesis may have both beneficial effects, such as restoration of nutrient and oxygen delivery to underperfused tissue, and harmful effects, such as changes in signaling and alterations in vascular structural makeup. In any case, given that liver fibrosis and angiogenesis are coupled, it is rational to study the links between these two biological processes.

Liver myofibroblasts (MFs) are the effector cells responsible for scar formation observed during fibrosis and cirrhosis.³ In fact, the “myofibroblast hypothesis” (that MFs are the matrix-producing cells in liver injury) may rightly be deemed the “myofibroblast theory” (in the traditional language of science, in which a theory is a hypothesis verified so many times that is unlikely to be proven incorrect). It is increasingly recognized that MFs constitute a heterogeneous cell population that may derive from multiple cellular precursors. However, two cell populations predominate: MFs derived from hepatic stellate cells (HSCs) and those derived from portal fibroblasts (PFs). In the healthy liver, PFs are resident mesenchymal cells with a peribiliary distribution located in portal areas.^5,6 Functions of quiescent PFs include regulation of extracellular matrix (ECM) turnover^4,5 and maintenance of cholangiocyte cell mass.^4,5 In the fibrosing liver, PFs undergo myofibroblastic differentiation to give rise to PF-derived MFs, generally described as portal myofibroblasts (PMFs). It is worth noting that the terminology used for liver MF is often as confusing as it is descriptive, which has probably prevented its widespread adoption. Regardless, the phenotypic characterization of PMF has advanced a great deal in recent years, which has led to an increased understanding of PMF function. PMFs are typically characterized by expression of contractile alpha-smooth muscle actin filaments^6,7 and increased production of several ECM-associated proteins, such as collagen-1,^6–8 fibronectin,^8,9 fibulin-2,^8,9 and lysyl oxidase family members.¹⁰

What had not been clear in studies of PMF was the presence of any link between PMF and angiogenesis. Thus, the goal of the study¹¹ by Lemoinne et al. featured in the current issue of HEPATOLOGY was to evaluate the functional contribution of PMF to angiogenesis in vitro and in vivo. For that purpose, the strategy pursued by the investigators here was to initially establish a correlation between the spatial distribution of fibrogenic PMF and of newly formed hepatic blood vessels in fibrotic livers, then test PMF ability to promote angiogenesis in vitro and in vivo, and, finally, determine the identity of the proangiogenic mediator(s) involved and the underlying mechanisms. First, the investigators performed a transcriptome analysis on primary culture-activated rat PF and identified type 15 collagen alpha1 (Col15α1) as a novel cell marker for PMF. This interesting finding confirmed Col15α1 specificity in labeling PMF in healthy and injured livers of mice and humans in vitro and in vivo. It is important to note that expression of Col15α1 was not what would be expected if all PMFs were labeled; rather, a subset of PMF appeared to be Col15α1 positive. In fact, the distribution of Col15α1-expressing PMFs closely paralleled the appearance of the hepatic neovasculature human and experimental liver fibrosis. Overall, these results determined that PMF identified by Col15α1 positivity were consistently found in the close vicinity of new blood vessels in cirrhotic livers, suggesting a potential mechanistic relationship between those two cell types. Next, the investigators tried to assess whether and how culture-activated PMFs release proangiogenic factors promoting liver endothelium activation/recruitment and neovascularization. They demonstrated that conditioned media from PMF contains neoangiogenic factors capable of inducing chemotaxis and tube formation using primary and immortalized endothelial cells (ECs). Fittingly, coculture of PMFs with immortalized ECs led to enhanced endothelial tubulogenesis. They also demonstrated that PMFs released these neoangiogenic factors largely within microparticles (defined by the investigators as vesicles of 0.1–1.0 μm in diameter and in accord with established definitions). Next, the investigators demonstrated that the primary effector molecule in PMF microparticle-mediated angiogenesis was vascular endothelial growth factor A (VEGF-A). Because cholangiocytes are known to regulate PMF phenotype,¹² the investigators tested whether cholangiocyte-PMF interactions regulate PMF-mediated angiogenesis. There results demonstrated that PMFs treated with cholangiocyte media were indeed more angiogenic than PMFs alone. Taken together, the experimental findings provide strong evidence that at least a subset of PMF can be identified by Col15α1 positivity, that PMFs promote angiogenesis through release of microparticles containing VEGF-A, and that this process is influenced by cholangiocytes.

Two important lessons may be gleaned from this work. The first of these is that PMFs are biologically relevant. There is currently great controversy as to whether PMFs are important in the pathogenesis of liver fibrosis, with one set of recent experiments suggesting that MFs from HSCs are the only biologically relevant matrix-producing cells¹³ and another set suggesting that MFs from PFs and HSCs are both important, but may differ in relevance based on the mechanism of injury.¹⁴ What cannot be ignored is that PMFs are always present in the cirrhotic liver. As quoted by the fictional rock band in director Rob Reiner’s 1984 films This is Spinal Tap: “In ancient times, hundreds of years before the dawn of history, lived an ancient race of people: the Druids. No one knows who they were, or what they were doing.” This is comedy, to be sure, but it reminds us that, simply because we cannot define the specific role of an observed phenomenon, we cannot discard that observed phenomenon. Hence, it is very possible that the most important role of PMF is not one of matrix production, but rather of cell-cell regulation. In the case of the current work, that regulation may be angiogenic signaling to endothelia. In the case of previous studies, that regulation may be control of bile ductular proliferation, as has been established previously.

A second lesson that will require further expansion is the regulation of PMF by cholangiocytes. The investigators demonstrated that a cholangiocyte-derived factor augmented PMF angiogenesis. What must be explored is what that specific factor (or combination of factors) is (are). In layman’s terms, what’s the special sauce? Cholangiocytes have been shown to release a number of regulatory factors, including a whole complement of immune and inflammatory mediators (reviewed in Syal et al.¹⁵), and communication from cholangiocytes to PF/PMF by chemokine (C-C motif) ligand 2¹² and interleukin-6^16,17 have both been demonstrated. Perhaps more important, multiple reports have shown that cholangiocytes themselves release VEGF molecules. VEGF acts as an autocrine regulator of cholangiocyte proliferation,¹⁸ and cholangiocyte-mediated VEGF release regulates cyst growth and angiogenesis in a mouse model of autosomal-dominant polycystic kidney disease.¹⁹ Thus, it is worth asking two questions: First, is VEGF the cholangiocyte special sauce, and second, which is more biologically relevant—PMF or cholangiocyte-mediated angiogenesis?

In any case, it is evident that the current work should influence investigations into PMF function for the immediate future. Both lessons noted provide biologically relevant questions that, if answered, will dramatically increase our understanding of the pathogenesis of cirrhosis, which, in turn, will hopefully lead to novel therapies to prevent or ameliorate cirrhosis.

Abbreviations

Col15α1

type 15 collagen alpha1

ECs

endothelial cells

ECM

extracellular matrix

HSCs

hepatic stellate cells

MF

myofibroblast

PF

portal fibroblast

PMF

portal myofibroblasts

VEGF-A

vascular endothelial growth factor A