Caspase inhibitors for the treatment of liver disease: friend or foe?

1. INTRODUCTION

Because of the liver’s function in the body as both a filter for toxins and an innate immune organ, the liver is predisposed to cellular turnover. Under healthy conditions, this occurs through the removal of aging and damaged hepatocytes by apoptosis. However, in a diseased liver, pathological cell death often occurs. In the previous twenty years, there has been considerable discussion in the research community over the different types of cell death. Apoptosis has a complex signaling pathway that prototypically results in activation of caspases in hepatocytes, resulting in the degradation of cellular proteins and activation of DNases, which cleave nuclear DNA, and induce cell death. As such, inhibition of caspases has routinely been tested in multiple models of liver disease as a potential means for reducing cell death. However, caspase inhibitors have failed to protect in many of these experiments. One reason discussed extensively was the fact that apoptosis was not the major mode of cell death in these models [1]. In addition, recent studies have indicated caspase inhibition can have profound effects outside of just inhibition of apoptotic cell death and may not always interrupt the cell death process in a sustainable fashion. The purpose of this editorial is to discuss caspase inhibition both as a clinical therapeutic option, and as a component of the wider, accepted mechanisms of hepatocyte cell death.

2. CASPASE INHIBITION – PHYSIOLOGY, PHARMACOLOGY, AND BEYOND

2.1 Cell Death in Liver Disease

The prototypical pathway for understanding cell death in liver injury is the tumor necrosis factor-α (TNF-α) pathway. Binding of TNF-α to its receptor TNFR1 causes trimerization followed by the formation of a death-inducing signaling complex (DISC) composed of the intracellular part of TNFR1 (death domain), TNF receptor associated death domain (TRADD) and TNF receptor associated factor (TRAF) proteins [2]. This complex can activate cell death through a number of different mechanisms. Primarily, in the presence of pro-caspase-8, this triggers the apoptotic process. Autocatalytic activation of caspase-8 initiates truncation of bid protein, which translocates to the mitochondria and causes pore formation by activating pro-apoptotic Bcl-2 family protein Bax and Bak in the outer mitochondrial membrane. This releases mitochondrial cytochrome c, which associates in the cytosol with ATP, apoptosis associated factor-1 (APAF-1), and pro-caspase-9 to form the apoptosome [2]. Apoptosome formation leads to the autocatalytic activation of caspase-9, which then cleaves pro-caspase-3 to form active caspase-3, which, among others, cleaves the inhibitor of caspase-activated DNase (ICAD) and induces degradation of nuclear DNA by caspase-activated DNase (CAD) [3]. Thus, pan-caspase inhibitors can effectively prevent acute TNF- and Fas-receptor mediated apoptosis [4,5].

However, novel data have shown that when caspase activity is inhibited, cells can undergo an entirely different mode of cell death mediated by two proteins called receptor interacting protein kinase 1 (RIP1) and RIP3 [6,7]. After TNF-α stimulation, a complex is formed involving RIP1, RIP3, Fas associated death domain (FADD), and active caspase-8, which degrades RIP1 and RIP3 [6]. However, when caspase-8 is inhibited or genetically deleted, this complex activates downstream signaling pathways that result in necroptosis [6,8]. Necroptosis executes cell death through activation of mixed lineage kinase domain like pseudokinase (MLKL) protein, which translocates to the plasma membrane and causes pore formation and cell death [9]. This process has a molecular similarity to the more recently discovered pyroptosis pathway wherein activation of the inflammasome by damage- or pathogen-associated molecular patterns, results in cleavage of gasdermin D by caspase 11 or caspase 4/5 [10]. The cleaved form of gasdermin D also translocates to the plasma membrane and causes pore formation and cell death [10]. Regardless, it is clear that cell death is highly contextual, and it may not be as simple as inhibiting caspases to block apoptotic cell death and prevent liver injury.

2.2 Caspase Inhibitors: The Past and the Present

Apoptosis has been implicated as a modality of cell death in multiple models [1]. Some examples include in vitro exposure of rat hepatocytes to the human bile acid glycochenodeoxycholate, which results in prototypical apoptosis that is preventable by caspase inhibitors [11]. Additionally, galactosamine/endotoxin treatment in mice causes TNF-α dependent apoptosis and inflammatory liver injury that is also effectively prevented when caspase inhibitors are given before the onset of apoptosis [4]. Similarly, activation of the Fas receptor by either an agonistic antibody or Fas ligand triggers hepatocellular apoptosis that can be completely prevented by caspase inhibitors [5,12]. While there are a number of other models that have been suggested to be apoptotic in nature, many of these studies are controversial in their analysis of the data [13,14]. The bile duct ligation model of cholestasis does not result in activation of caspases or histological apoptosis; the injury is caused primarily by inflammatory necrosis mediated by neutrophils [14]. Similarly, hepatic ischemia reperfusion injury does not involve caspase activation and morphological evidence of apoptosis [13]. In both cases, caspase inhibitors do not protect against the injury [13,14]. Likewise, acetaminophen (APAP)-induced liver injury in mice cannot be attenuated by caspase inhibitors [15]. Although there is evidence of modest caspase activation in human patients with APAP hepatotoxicity [16], obstructive cholestasis [17], and alcoholic hepatitis [18], the pathophysiological relevance is likely minor and thus, caspase inhibitors may not have a major impact on the pathophysiology. A case in point was the use of the pan-caspase inhibitor IDN-6556 to attenuate reperfusion injury in human liver transplants [19]. The inhibitor was only effective when added to the storage solution but not when administered during reperfusion, i.e. at the time of the assumed apoptotic cell death [19]. This raised the concern that the effects of high concentrations of the caspase inhibitor were due to off-target effects, i.e. inhibition of other proteases such as calpains and cathepsins during ischemia, rather than inhibition of apoptosis during reperfusion. The recently developed M65/M30 ELISA kits that use differential cleavage of keratin-18 by caspases to determine apoptotic versus non-apoptotic cell death conclusively demonstrated that while there is detectable apoptosis in APAP-induced or cholestatic liver injury in human patients, the level of non-apoptotic cell death commonly exceeds these values by ratios of 10:1 or greater [16,17]. Even in diseases with some degree of apoptosis, other therapeutic options are commonly present, and may be more efficacious. Hepatitis C is a disease where there are definitive increases in apoptosis in patients, although individual Hepatitis C proteins can have varying affects on apoptosis in different experimental models (20). Despite this well-defined characteristic, caspase inhibitors as a therapeutic option are probably unwarranted at this juncture. Not only would it require chronic therapy, but newer direct-acting antivirals can actively destroy the virus by targeting non-structural proteins. This class of drugs has been largely curative, and thus while apoptosis may have once been a solid therapeutic target, it is unlikely it will have additional substantial curative benefit, even in difficult to cure genotypes. As such, the primary issues facing the current Hepatitis C burden are optimizing treatment regimens of the new direct-activing antiviral drugs and reaching sick and at-risk patient populations for early detection (21). Thus, it is critical to objectively assess the importance and the relevance of apoptotic cell death in a given pathophysiology, especially in humans, before the therapeutic use of caspase inhibitors is considered.

2.3 Caspase Inhibitors: The Future

In addition to the concern that apoptosis may not be the dominant mode of cell death in a pathophysiology, more recent findings even question if caspases may be the ideal therapeutic target for apoptosis. Consistent with previous reports, treatment of mice with a pan-caspase inhibitor effectively protected against apoptosis during the early injury phase in the galactosamine/endotoxin model, but even with this sustained inhibition of caspase activity, cells eventually undergo massive necrosis 24 hours after the initial treatment [22]. Although caspase inhibition prevented RIP1 and RIP3 degradation, this cell death was independent of traditional necroptosis. Both RIP3-deficient mice and wild type animals treated with the RIP1 inhibitor necrostatin-1 were not protected and had a poorer inflammatory profile, indicating that activation of the TNF-α pathway may have additional mechanisms of injury in vivo that also escape caspase inhibition [22]. These findings clearly demonstrate differential effects of caspase inhibitors when used short-term versus more long-term or even chronically. It is imperative to better understand these different effects as caspase inhibitors have been proposed for treatment of chronic disorders such as non-alcoholic steatohepatitis and alcoholic hepatitis, both of which feature a prominent, deleterious inflammatory component. Caspase inhibitors may worsen the injury in these patients over time and exacerbate the innate immune response.

Recently a set of signaling complexes present in immune cells known as inflammasomes were discovered and have been a major source of research focus subsequently (23). Classically, inflammasomes such as the NACHT, LRR and PYD domains-containing protein 3 (Nalp3) inflammasome serve primarily to cleave and activate intracellular pro-caspase-1 into caspase-1, which then cleaves and activates pro-IL-1ß into IL-1ß. This pro-inflammatory cytokine is secreted via either a poorly understood leaderless mechanism, or passively via necrosis/pyroptosis (reviewed in 24). Non-canonically, inflammasomes can also be activated by detection of lipopolysaccharide and other microbial pathogens through direct sensing and activation by human caspase 4/5 or rodent caspase 11, which then activate pro-caspase-1 (24). IL-1ß then acts as a major mediator of inflammation by recruiting other immune effector cells such as monocytes and neutrophils (24). Caspase inhibition has been proposed a way to prevent inflammasome activation, and thus potentially prevent inflammasome mediated inflammation in numerous liver diseases (25). Caspase inhibition is efficacious in some experimental models; however, neither the establishment of a defined pathological role nor the use of caspase inhibitors as a treatment has been successful in human liver injury models (24,25). The inflammasome is likely activated in multiple disease states in humans, although current evidence suggests a more limited role in liver disease. Research in this area is ongoing though, and may be a viable target for future disease.

3. CONCLUSIONS

The exact role and clinical relevance of apoptosis in many acute and chronic liver diseases is still controversial. On the other hand, there is no question that suicide-substrate types of caspase inhibitors are potent and highly effective in preventing apoptotic cell death. However, an emerging caveat in this area is that inhibition of caspases will prevent apoptosis but may trigger necroptosis and other caspase-independent forms of cell death. This raises concerns that caspase inhibitors, especially when used long-term, may be less effective than previously anticipated based on short-term experiments or may even be detrimental by switching to necrosis and consequently trigger a more prominent inflammatory response. Thus, a better understanding of both the disease pathophysiology and the potential long-term effects of caspase inhibition on cell death is critical before caspases can be considered as a potential therapeutic target in human liver diseases.