**2. Hepatic ischemia-reperfusion injury associated with liver transplantation. An unresolved problem in clinical practice**

Liver transplantation (LT) dates back to 1963, when Thomas Starzl carried out the first transplant on a child suffering from biliary atresia. LT has evolved as the therapy of choice for patients with end-stage liver disease. However, I/R injury, inherent in every LT, is the main cause of both initial poor function and primary non-function of liver allograft. The latter is responsible for 81% of re-transplantations during the first week after surgery (Clavien et al., 1992; Jaeschke, 1996). I/R injury is a phenomenon whereby cellular damage in a hypoxic organ is accentuated following the restoration of oxygen delivery (Jaeschke, 1998; Teoh et al., 2003; Jaeschke, 2003). In the liver, this form of injury was recognized as a clinically important pathological disorder by Toledo-Pereyra et al. in 1975 during studies of experimental LT. However, it was not until the mid-1980s that the term reperfusion injury was generally used in the literature on LT (Teoh et al., 2003).

A variety of clinical factors including starvation, graft age, and steatosis contribute to enhance liver susceptibility to I/R injury, further increasing the patient risks related to reperfusion injury (Shah & Kamath, 2003). In clinical LT, starvation of the donor, due to prolonged intensive care unit hospitalization or lack of an adequate nutritional support, increases the incidence of hepatocellular injury and primary nonfunction (Massip-Salcedo et al., 2007).

The waiting list for LT is growing at a fast pace, whereas the number of available organs is not growing at a proportional rate. The shortage of organs has led centers to expand their criteria for the acceptance of marginal grafts, which show poor tolerance to I/R (Busuttil & Tanaka, 2003). Some of these include the use of organs from aged donors, non-heart-beating donors (NHBD), and grafts such as small-for-size or steatotic livers. However, I/R injury is the underpinning of graft dysfunction that is seen in the marginal organ (Busuttil & Tanaka, 2003). The fundamental problem with NHBD organs is the prolonged warm ischemia before cold preservation (Reddy et al., 2004). Controlled NHBDs provide organs that are far less prone to ischemic damage and tend to offer superior posttransplant function (Busuttil & Tanaka, 2003). The use of uncontrolled NHBDs is associated with a very high risk of primary nonfunction (Reddy et al., 2004).

One of the benefits of reduced-size grafts from living donors is a graft of good quality with a short ischemic time, this latter being possible because live donor procurements can be electively timed with recipient procedure (Farmer et al., 2001). On the other hand, the major concern over application of living-related liver transplantation for adults is graft-size disparity. The small graft needs regeneration to restore the liver/body ratio. It is well known that I/R significantly reduces liver regeneration after hepatectomy (Franco et al., 2004).

Donor age of more than 70 years was found to be associated with lower patient and graft survival (Busuttil & Tanaka, 2003, Casillas et al., 2006). Additionally these donors also have an increased incidence of steatosis, which may potentiate cold preservation injury (Busuttil & Tanaka, 2003). Steatotic livers are one of the most common types of organs from marginal donors. The present review will focus on this type of liver grafts. Among other factors, unhealthy lifestyles associated with the consumption of alcohol and inappropriate diets have increased the proportion of patients with steatotic livers.

Hepatic steatosis is a major risk factor for liver surgery and transplantation, and fatty livers are unsuitable for many reasons. Operative mortality associated with steatosis exceeds 14%, compared with 2% for healthy livers, and the risks of primary non-function and dysfunction after surgery are similarly higher (Casillas et al., 2006; Selzner et al., 2000). Thus, hepatic steatosis is the major cause of graft rejection after LT and exacerbates the organ shortage problem (Fernández et al., 2004). Therefore, minimizing the adverse effects of I/R injury could increase the number of both grafts suitable for transplantation and patients who successfully recover from LT. The first step towards achieving this objective is a full understanding of the mechanisms involved in I/R injury.

clinically important pathological disorder by Toledo-Pereyra et al. in 1975 during studies of experimental LT. However, it was not until the mid-1980s that the term reperfusion injury

A variety of clinical factors including starvation, graft age, and steatosis contribute to enhance liver susceptibility to I/R injury, further increasing the patient risks related to reperfusion injury (Shah & Kamath, 2003). In clinical LT, starvation of the donor, due to prolonged intensive care unit hospitalization or lack of an adequate nutritional support, increases the incidence of hepatocellular injury and primary nonfunction (Massip-Salcedo et

The waiting list for LT is growing at a fast pace, whereas the number of available organs is not growing at a proportional rate. The shortage of organs has led centers to expand their criteria for the acceptance of marginal grafts, which show poor tolerance to I/R (Busuttil & Tanaka, 2003). Some of these include the use of organs from aged donors, non-heart-beating donors (NHBD), and grafts such as small-for-size or steatotic livers. However, I/R injury is the underpinning of graft dysfunction that is seen in the marginal organ (Busuttil & Tanaka, 2003). The fundamental problem with NHBD organs is the prolonged warm ischemia before cold preservation (Reddy et al., 2004). Controlled NHBDs provide organs that are far less prone to ischemic damage and tend to offer superior posttransplant function (Busuttil & Tanaka, 2003). The use of uncontrolled NHBDs is associated with a very high risk of

One of the benefits of reduced-size grafts from living donors is a graft of good quality with a short ischemic time, this latter being possible because live donor procurements can be electively timed with recipient procedure (Farmer et al., 2001). On the other hand, the major concern over application of living-related liver transplantation for adults is graft-size disparity. The small graft needs regeneration to restore the liver/body ratio. It is well known that I/R significantly reduces liver regeneration after hepatectomy (Franco et al.,

Donor age of more than 70 years was found to be associated with lower patient and graft survival (Busuttil & Tanaka, 2003, Casillas et al., 2006). Additionally these donors also have an increased incidence of steatosis, which may potentiate cold preservation injury (Busuttil & Tanaka, 2003). Steatotic livers are one of the most common types of organs from marginal donors. The present review will focus on this type of liver grafts. Among other factors, unhealthy lifestyles associated with the consumption of alcohol and inappropriate diets

Hepatic steatosis is a major risk factor for liver surgery and transplantation, and fatty livers are unsuitable for many reasons. Operative mortality associated with steatosis exceeds 14%, compared with 2% for healthy livers, and the risks of primary non-function and dysfunction after surgery are similarly higher (Casillas et al., 2006; Selzner et al., 2000). Thus, hepatic steatosis is the major cause of graft rejection after LT and exacerbates the organ shortage problem (Fernández et al., 2004). Therefore, minimizing the adverse effects of I/R injury could increase the number of both grafts suitable for transplantation and patients who successfully recover from LT. The first step towards achieving this objective is a full

have increased the proportion of patients with steatotic livers.

understanding of the mechanisms involved in I/R injury.

was generally used in the literature on LT (Teoh et al., 2003).

primary nonfunction (Reddy et al., 2004).

al., 2007).

2004).

A large number of factors and mediators play a part in liver I/R injury (Banga et al., 2005; Casillas et al., 2006; Fan et al., 1999; Jaeschke, 2003; Lentsch et al., 2000). The relationships between the signalling pathways involved are highly complex and it is not yet possible to describe, with absolute certainty, the events that occur between the beginning of reperfusion and the final outcome of either poor function or a non-functional liver graft.

Figure 1 shows some of the mechanisms involved in the pathophysiology of I/R injury. Due to the complexity of hepatic I/R injury, the present review summarizes the established basic concepts of the mechanisms and cell types involved in this process. The lack of oxygen to hepatocytes during ischemia causes mitochondrial deenergization, ATP depletion, alterations of H+, Na+, Ca2+ homeostasis that activate hydrolytic enzymes and impair cell volume regulation and sinusoidal endothelial cells (SEC) as well as Kupffer cells (KC) swelling (Massip-Salcedo et al., 2007). This fact together with the imbalance between nitric oxide (NO) and endothelin (ET) production, contributes to narrowing of the sinusoidal lumen and thus to microcirculatory dysfunction. Capillary narrowing also contributes to hepatic neutrophil accumulation (Peralta et al., 1996; Peralta et al., 2000a). Concomitantly, the activation of KC releases reactive oxygen species (ROS) and proinflammatory cytokines, including tumour necrosis factor-α (TNF-α) and interleukin-1 (IL-1) (Bilzer & Gerber, 2000; Lentsch et al., 2000). ROS can also derive from xanthine deshydrogenase/xanthine oxidase (XDH/XOD). Cytokines release throughout the induction of adhesion molecules (intercellular cell adhesion molecule [ICAM] and vascular cell adhesion molecule [VCAM]) and chemokines promote neutrophil activation and accumulation, thereby contributing to the progression of parenchymal injury by releasing ROS and proteases (Jaeschke, 1998, 2003; Lentsch et al., 2000). Besides, IL-1 and TNF-α recruit and activate CD4+ T-lymphocytes, which produce granulocyte-macrophage colony-stimulating factor (GM-CSF), interferon gamma (INF-γ) and tumor necrosis factor beta (TNF-β). These cytokines amplify KC activation and TNF-α and IL-1 secretion and promote neutrophil recruitment and adherence into the liver sinusoids (Casillas et al., 2006; Selzner, 2003). Platelet activating factor (PAF) can prime neutrophils for superoxide generation, whereas leukotriene B4 (LTB4) contributes to the amplification of the neutrophil response (Jaeschke, 1998, 2003) (see Fig. 1).

The present review will present data from the literature about the possible sources of ROS, NO effects, mechanisms, and the role of some pro-inflammatory mediators such as TNF-α, and transcription factors, for example, nuclear factor kappa B (NFκB). These data will provide a better explanation on why hepatic I/R injury remains an unresolved problem in the clinical practice.

#### **3.1 Mechanisms responsible for ROS production**

The source of ROS in hepatic I/R has long been controversial. As regards the mechanisms responsible for ROS production, experiments with XDH/XOD inhibitors such as allopurinol suggest that this system is the main ROS generator in hepatocytes and it has also been implicated in LT-related lung damage (Casillas et al., 2006; Fernández et al., 2002). However, results obtained in experimental models of the isolated perfused liver have underestimated the importance of the XDH/XOD system, and suggest that mitochondria could be the main source of ROS (Jaeschke & Mitchell, 1989). On the other hand, some data challenge the

Fig. 1. Summary of the mechanisms involved in hepatic ischemia-reperfusion injury. (Bilzer & Gerber, 2000; Casillas et al., 2006; Jaeschke, 1998, 2003; Lentsch et al., 2000; Massip-Salcedo et al., 2007 ; Peralta et al., 1996, 2000a; Selzner, 2003)

pathophysiological relevance of intracellular oxidant stress during reperfusion (Grattagliano et al., 1999; Metzger et al., 1988). Grattagliano et al., 1999, demonstrated that mitochondria do not seem to actively participate in the reperfusion-induced oxidative stress. In addition, studies by Jaeschke et al. and Metzger et al. showed that the increased vascular oxidant stress after 30 and 60 min of ischemia was attenuated by inactivation of KC but not by high dose of allopurinol (Metzger et al., 1988). Interestingly, ROS release by KC occurs via the XDH/XOD system (Wiezorek et al., 1994). The conversion from XDH to XOD following cold storage is very slow in endothelial cells and hepatocytes, but much faster and higher in KC (Wiezorek et al., 1994). However, the KC function in I/R injury is still an area of active investigation. The elimination of KC did not modify the deleterious effects of I/R and the activation of neutrophils is not essential for reoxygenation injury (Imamura et al., 1995; Teoh et al., 2003). Clearly, then, there is a range of potentially conflicting results with regard to the mechanisms responsible for ROS generation in liver I/R injury. For instance, in our opinion, in order to clarify the importance of XDH/XOD versus mitochondria it should be taking into account that there are differences in the experimental models evaluated, including the times of ischemia. In this line, XDH/XOD play a crucial role in hepatic I/R injury only in conditions in which significant conversion of XDH to XOD occurs (80-90% of XOD) such as 16 h of cold ischemia. However, this ROS generation system does not appear to be crucial at shorter ischemic periods such as 6 h of cold ischemia (Fernández et al., 2002). Thus, even after prolonged periods of ischemia, where a significant conversion of XDH to the XOD occurs, this enzyme may only play a minor role compared with mitochondria (Jaeschke & Mitchell, 1989). In contrast with the experimental studies, the clinical reports suggest that 45-65% XOD was sufficient to induce hepatic damage (Pesonen et al., 1998). In addition, the drugs used for inhibiting XDH/XOD should be considered, since, for example allopurinol, seems to have more than one mechanisms of action. It is not only a potent inhibitor of XOD, but it may also improve ischemia-induced mitochondrial dysfunction (Casillas et al., 2006; Jeon et al., 2001). In fact, evidence for reduced mitochondrial dysfunction after high doses of allopurinol was shown in a warm hepatic I/R model (Jeon et al., 2001). Similarly, in assessing the relative contribution of intracellular versus vascular oxidant stress to hepatic I/R injury, it should also be noted that oxidative stress in hepatocytes and the stimulatory state of KC after I/R depend on the duration of ischemia, and may also differ between ischemia at 4ºC and that at 37ºC, which probably leads to different developmental mechanisms of liver damage (Casillas et al., 2006). The differences in KC function in liver I/R injury cannot be attributed to the type experiment, since most authors used an *ex vivo* model of perfused rat liver. Nor could they be explained by differences in the times of cold ischemia, since the results obtained following the same ischemic period (24 h) were completely opposed (Imamura et al., 1995). The type of drug used for KC inactivation is the most probable explanation, since most of the studies implicating KC as main source of ROS used gadolinium chloride (GdCl3) (Schauer et al., 2001; Zhong et al., 1996) whereas those that did not implicate KC used liposome-encapsulated dichloromethylene diphosphate (Imamura et al., 1995). Indeed, differences in the properties and action mechanisms of these two KC inhibitors have been reported.

#### **3.2 Mediators and transcription factors in I/R injury**

#### **3.2.1 Nitric oxide**

24 Liver Transplantation – Basic Issues

Fig. 1. Summary of the mechanisms involved in hepatic ischemia-reperfusion injury. (Bilzer & Gerber, 2000; Casillas et al., 2006; Jaeschke, 1998, 2003; Lentsch et al., 2000; Massip-Salcedo

pathophysiological relevance of intracellular oxidant stress during reperfusion (Grattagliano et al., 1999; Metzger et al., 1988). Grattagliano et al., 1999, demonstrated that mitochondria do not seem to actively participate in the reperfusion-induced oxidative stress. In addition, studies by Jaeschke et al. and Metzger et al. showed that the increased vascular oxidant stress after 30 and 60 min of ischemia was attenuated by inactivation of KC but not by high dose of allopurinol (Metzger et al., 1988). Interestingly, ROS release by KC occurs via the XDH/XOD system (Wiezorek et al., 1994). The conversion from XDH to XOD following cold storage is very slow in endothelial cells and hepatocytes, but much faster and higher in KC (Wiezorek et al., 1994). However, the KC function in I/R injury is still an area of active investigation. The elimination of KC did not modify the deleterious effects of I/R and the activation of neutrophils is not essential for reoxygenation injury (Imamura et al., 1995; Teoh et al., 2003). Clearly, then, there is a range of potentially conflicting results with regard to the mechanisms responsible for ROS generation in liver I/R injury. For instance, in our opinion, in order to clarify the importance of XDH/XOD versus mitochondria it should be taking into account that there are differences in the experimental models evaluated, including the times of ischemia. In this line, XDH/XOD play a crucial role in hepatic I/R injury only in conditions in which significant conversion of XDH to XOD occurs (80-90% of XOD) such as 16 h of cold ischemia. However, this ROS generation system does not appear to be crucial at

et al., 2007 ; Peralta et al., 1996, 2000a; Selzner, 2003)

It is difficult to distinguish between beneficial and harmful mediators in I/R injury. Some authors have found that NO exerts a beneficial effect on I/R injury in different organs, tissues and cells, whereas other studies report no effect or even a deleterious action of NO (Peralta et al., 2001a). In our opinion, in addition to the differences in animal species, experimental models of hepatic I/R tested, and the dose and timing of administration of the different pharmacological modulators of NO, these differential effects of NO could be explained, at least partially, by the different source of NO. In this context, some studies suggest that although endothelial NO synthase (eNOS)-derived NO production is protective in I/R, inducible NO synthase (iNOS)-derived NO production may contribute to I/R injury. This may be a function of the NO generation kinetics of the two isoforms in I/R. The basal, low-level NO generation by the constitutively expressed eNOS isoform may abrogate the microcirculatory stresses of engraftment and reperfusion. In contrast, iNOS-derived NO cannot be generated until several hours after stimulation because of requirements for transcriptional induction of this isoform. Excess NO production may no longer be of microcirculatory benefit at this later time (Shah & Kamath, 2003). Furthermore, the excessive levels of iNOS-derived NO production may be detrimental through the generation of NOSderived superoxide production or the generation of peroxynitrite. Additionally, whether NO is cytoprotective or cytotoxic in hepatic I/R injury may be determined at apoptosis (Casillas et al., 2006). For example, NO may promote apoptosis by inducing cytochrome c (Cyt c) release and caspase activation (Chung et al., 2001). However, NO may also upregulate the anti-apoptotic protein Bcl-2 (Genaro et al., 1995). In addition, to understand the different results in relation with the action mechanisms of NO, it is important to clarify whether the NO source is endogenous or exogenous. In this regard, although the beneficial role of endogenous NO could be related to an attenuation of leukocyte accumulation, the exogenous supplementation of NO did not modify this parameter but was associated with an inhibition of endothelin release (Peralta et al., 2001a).

#### **3.2.2 TNF and NF**κ**B**

Differential effects of NO mentioned above have also been reported for other mediators involved in hepatic I/R injury. According to the cell type and experimental or pathologic conditions, TNF-α is protective or injurious to the liver in the context of I/R injury. TNF-α may stimulate cell death or it may induce hepatoprotective effects mediated by antioxidant, antiapoptotic, and other anti-stress mediators coupled with a pro-proliferative biologic response (Casillas et al., 2006). For example, although the deleterious effect of the TNF-α in local and systemic damage associated with hepatic I/R is well established (Peralta et al., 1999), this mediator is also a key factor in hepatic regeneration (Teoh et al., 2003), an important process in reduced-size LT. Conversely, a study by our group found no correlation between TNF-α levels and liver regeneration in reduced-size LT (Franco et al., 2004), while Tian *et al*., 2006, linked disruption of TNF-α release to lower hepatic injury and increased liver regeneration. These divergent results about the role of TNF-α in liver regeneration could be explained by different TNF-α inhibitors or animal species utilized in these experiments as well as differences in the experimental models of LT used, including the times of cold ischemia. These differential effects observed for TNF-α can also be extrapolated to transcription factors.

It is well known that NFκB can regulate various downstream pathways and thus has the potential to be both pro- and antiapoptotic (Fan et al., 1999). Currently it is not clear whether the beneficial effects of NFκB activation in protection against apoptosis or its detrimental proinflammatory role predominate in liver I/R (Fan et al., 1999). Hepatic neutrophil recruitment and hepatocellular injury are significantly reduced when NFκB activation is suppressed in mice following partial hepatic I/R (Casillas et al., 2006). However, nuclear factor kappa B (NFκB) activation is essential for hepatic regeneration after rat LT, and reduces apoptosis and hepatic I/R injury (Bradham et al., 1999). To understand the role of NFκB in the context of hepatic I/R, is important to consider the differences in animal species used, for instance, mechanisms of protection from apoptosis might be different in rats and mice (Chaisson et al., 2002). In addition, the experimental design used to evaluate the role of this transcription factor may also be important. Thus, some studies using adenoviral vector containing a repressor to prevent NFκB activation may not accurately reflect the role of NFκB signalling in regenerating liver because adenoviral vectors themselves cause increased TNF-α levels, DNA synthesis, and apoptosis in the liver before partial hepatectomy (Iimuro et al., 1998). Moreover, to explain these apparently controversial effects of NFkB, the pattern of NFkB activation under cold ischemia conditions should be taken into account. Takahashi *et al.,* 2002, have demonstrated in rat LT that NFkB activation during reperfusion occurs in two phases. The early peak of NFkB DNA binding was found 1-3 h after reperfusion and represents the nuclear translocation of NFkB p50/p65 heterodimers, whereas the second peak, mainly composed of p50 homodimers, was observed at 12 h post-reperfusion. In this study, the donor liver treatment with adenovirus encoding the IkB super-repressor gene cannot affect the early peak of NFkB activation, but partially inhibited the second peak of NFkB DNA binding. The results indicated that, in contrast to early NFkB activation, inhibition of the late phase of NFkB activation was not associated with variations in levels of inflammatory mediators, but rather enhanced hepatocellular apoptosis (Takahashi et al., 2002), which reinforces the dual function of NFkB in transplanted liver. Nevertheless, this hypothesis does not fully explain the differences in the results. Indeed, Bradham *et al*., 1999, observed a marked increase in apoptosis when NFkB blockade was carried out at 3 h of reperfusion, which seems to be a reperfusion time associated with the early peak of activation of NFkB. Of course, there are differences between Takahashi´s and Bradham´s studies. For example, whereas Bradham infused the adenoviral vector by endovenous injection 24 h before liver explantation, in Takahashi's study the graft was perfused with UW solution containing the adenovirus immediately before cold storage.

#### **3.2.3 Neutrophil accumulation**

26 Liver Transplantation – Basic Issues

(Cyt c) release and caspase activation (Chung et al., 2001). However, NO may also upregulate the anti-apoptotic protein Bcl-2 (Genaro et al., 1995). In addition, to understand the different results in relation with the action mechanisms of NO, it is important to clarify whether the NO source is endogenous or exogenous. In this regard, although the beneficial role of endogenous NO could be related to an attenuation of leukocyte accumulation, the exogenous supplementation of NO did not modify this parameter but was associated with

Differential effects of NO mentioned above have also been reported for other mediators involved in hepatic I/R injury. According to the cell type and experimental or pathologic conditions, TNF-α is protective or injurious to the liver in the context of I/R injury. TNF-α may stimulate cell death or it may induce hepatoprotective effects mediated by antioxidant, antiapoptotic, and other anti-stress mediators coupled with a pro-proliferative biologic response (Casillas et al., 2006). For example, although the deleterious effect of the TNF-α in local and systemic damage associated with hepatic I/R is well established (Peralta et al., 1999), this mediator is also a key factor in hepatic regeneration (Teoh et al., 2003), an important process in reduced-size LT. Conversely, a study by our group found no correlation between TNF-α levels and liver regeneration in reduced-size LT (Franco et al., 2004), while Tian *et al*., 2006, linked disruption of TNF-α release to lower hepatic injury and increased liver regeneration. These divergent results about the role of TNF-α in liver regeneration could be explained by different TNF-α inhibitors or animal species utilized in these experiments as well as differences in the experimental models of LT used, including the times of cold ischemia. These differential effects observed for TNF-α can also be

It is well known that NFκB can regulate various downstream pathways and thus has the potential to be both pro- and antiapoptotic (Fan et al., 1999). Currently it is not clear whether the beneficial effects of NFκB activation in protection against apoptosis or its detrimental proinflammatory role predominate in liver I/R (Fan et al., 1999). Hepatic neutrophil recruitment and hepatocellular injury are significantly reduced when NFκB activation is suppressed in mice following partial hepatic I/R (Casillas et al., 2006). However, nuclear factor kappa B (NFκB) activation is essential for hepatic regeneration after rat LT, and reduces apoptosis and hepatic I/R injury (Bradham et al., 1999). To understand the role of NFκB in the context of hepatic I/R, is important to consider the differences in animal species used, for instance, mechanisms of protection from apoptosis might be different in rats and mice (Chaisson et al., 2002). In addition, the experimental design used to evaluate the role of this transcription factor may also be important. Thus, some studies using adenoviral vector containing a repressor to prevent NFκB activation may not accurately reflect the role of NFκB signalling in regenerating liver because adenoviral vectors themselves cause increased TNF-α levels, DNA synthesis, and apoptosis in the liver before partial hepatectomy (Iimuro et al., 1998). Moreover, to explain these apparently controversial effects of NFkB, the pattern of NFkB activation under cold ischemia conditions should be taken into account. Takahashi *et al.,* 2002, have demonstrated in rat LT that NFkB activation during reperfusion occurs in two phases. The early peak of NFkB DNA binding was found 1-3 h after reperfusion and represents the nuclear translocation of NFkB p50/p65 heterodimers, whereas the second

an inhibition of endothelin release (Peralta et al., 2001a).

**3.2.2 TNF and NF**κ**B** 

extrapolated to transcription factors.

Activation of neutrophils has been implicated in the hepatic microvascular dysfunction and parenchymal damage associated with I/R (Cutrin et al., 2002). Still, a controversial topic is the question of how neutrophils actually accumulate in the liver. The classical theory argues that the increased expression of adhesion molecules such as ICAM-1 and P-selectin plays a key role in neutrophil accumulation and the subsequent liver damage associated with I/R (Banga et al., 2005, Cutrin et al., 2002). In contrast, it has also been reported that neutrophil accumulation observed in the liver following I/R is not dependent on the up-regulation of either ICAM-1 or P-selectin (Peralta et al., 2001b).

To explain the results that neutrophil accumulation is not dependent on adhesion molecules, we subscribe to the theory proposed by Jaeschke, 2003. This theory argues that although Pselectin and ICAM-1 appear to be relevant for neutrophil adherence in postsinusoidal venules, the neutrophils relevant for the injury accumulate in sinusoids, which were identified as the dominant sites for neutrophil extravasation. In these capillaries, neutrophil sequestration does not depend on B2 integrins or on ICAM-1 or selectins (Essani et al., 1998; Vollmar et al., 1995; Jaeschke et al., 1996). Thus, mechanical factors such as active vasoconstriction, vascular lining cell swelling and injury, and reduced membrane flexibility after activation of the neutrophil, appear to be involved in trapping of these leukocytes in sinusoids (Jaeschke et al., 1996). The extensive vascular injury during reperfusion eliminates, in part, the sinusoidal endothelial cell barrier and the neutrophil has direct access to hepatocytes (Jaeschke, 2003; McKeown et al., 1988). Nevertheless, even with damaged but still present EC, transmigration may still be required (Jaeschke, 1998). As a consequence, I/R injury is only moderately or not at all attenuated by anti-ICAM therapies (Farhood et al., 1995; Vollmar et al., 1995). In regard with the role of P-selectin, sinusoidal EC neither contain Weibel Palade bodies nor do they transcriptionally upregulate relevant levels of Pselectin (Essani et al., 1998). However, during I/R, a number of interventions directed against selectins reduced hepatic neutrophil accumulation and cell injury (Amersi et al., 2001). Because these findings cannot be explained by the prevention of P-selectin-dependent rolling in sinusoids, it has been suggested that most liver I/R models include some degree of intestinal ischemia, which leads to neutrophil accumulation in remote organs including the liver (Casillas et al., 2006; Kubes et al., 2002). Thus the lower number of neutrophils in the liver when selectins are blocked may be a secondary effect due to the protection of antiselectin therapy against intestinal reperfusion injury (Kubes et al., 2002).

#### **3.3 Cell death in liver transplantation**

The severity of hepatocyte damage depends on the length of time the ischemia lasts. In human LT, a long ischemic period is a predicting factor for post-transplantation graft dysfunction, and some transplantation groups hesitate to transplant liver grafts preserved for more than 10 h (Fernández et al., 2002). Some studies in experimental models of LT indicate that 24 h of cold ischemia induces low survival at 24 h after LT. However, at shorter ischemic periods, LT may also result in primary organ dysfunction. The main victims of ischemic injury are the hepatocytes and SECs. These two cell types show different responses to different types of ischemia: hepatocytes are more sensitive to warm ischemia and SECs to cold ischemia (Bilzer & Gerber, 2000; McKeown et al., 1988). Although most hepatocytes remain viable after 48 h of cold preservation and reperfusion, SECs suffer severe damage following reperfusion (40% non-viable) (Caldwell et al., 1989). The result of this sinusoidal damage is the subsequent microcirculatory abnormalities upon reperfusion, resulting in hepatocyte injury and dysfunction (McKeown et al., 1988). This contributes to the development of primary nonfunction or impaired primary function after LT. However, some studies have called the importance of sinusoidal injury into question. Huet et al., 2004, have demonstrated that damage to the extracellular matrix from prolonged preservation and reperfusion appears to be the critical factor in graft failure (Banga et al., 2005). In addition, it is possible that perturbations in hepatocyte levels of adenine nucleotides during cold storage can trigger proteolytic events that contribute to damage in the liver graft and subsequently compromise hepatic functions after LT (Kukan & Haddad, 2001). Moreover, cold ischemia profoundly disturb several key hepatocellular functions, such as volume and pH homeostasis, as well as solute transport and drug metabolism, protein synthesis and mitochondrial function. This contributes to preservation injury of the liver graft. Therefore, these observations indicate that aside from reducing EC damage, LT therapy may benefit from strategies aimed at improving the maintenance of appropriate hepatocyte functions (Kukan & Haddad, 2001; Vajdova et al., 2002).

Apoptosis has been regarded as the fate of cells experiencing I/R injury (Sasaki et al., 1996). In this line, different studies have demonstrated apoptotic death in hepatocytes and/or SECs after both cold and warm ischemia of the rat liver (Gao et al., 1998; Kohli et al., 1999). All of the aforementioned studies (Gao et al., 1998; Kohli et al., 1999; Sasaki et al., 1996) used TdT-mediated dUTP-biotin nick and labelling (TUNEL staining) for DNA ladders to demonstrate apoptosis. However, the ability of TUNEL staining to distinguish between apoptosis and necrosis has been called into question. The activation of caspases has also been used to demonstrate apoptosis in rat SECs following cold I/R (Natori et al., 1999). Indeed, use of pan-caspase inhibitors protected rat liver SECs and hepatocytes against I/R injury after prolonged periods of both cold and warm ischemia. On the other hand, other groups oppose the view that the majority of cells undergo apoptosis in response to either warm or cold I/R injury, believing that necrosis is the principle form of cell death (Massip-Salcedo et al., 2007). They believe that in a number of studies the proportion of cells undergoing apoptosis is not of significant magnitude and that the degree of caspase activation does not correlate with the number of SECs and hepatocytes supposedly undergoing apoptosis. Thus, a controversy has emerged over the past years as to whether

the liver when selectins are blocked may be a secondary effect due to the protection of

The severity of hepatocyte damage depends on the length of time the ischemia lasts. In human LT, a long ischemic period is a predicting factor for post-transplantation graft dysfunction, and some transplantation groups hesitate to transplant liver grafts preserved for more than 10 h (Fernández et al., 2002). Some studies in experimental models of LT indicate that 24 h of cold ischemia induces low survival at 24 h after LT. However, at shorter ischemic periods, LT may also result in primary organ dysfunction. The main victims of ischemic injury are the hepatocytes and SECs. These two cell types show different responses to different types of ischemia: hepatocytes are more sensitive to warm ischemia and SECs to cold ischemia (Bilzer & Gerber, 2000; McKeown et al., 1988). Although most hepatocytes remain viable after 48 h of cold preservation and reperfusion, SECs suffer severe damage following reperfusion (40% non-viable) (Caldwell et al., 1989). The result of this sinusoidal damage is the subsequent microcirculatory abnormalities upon reperfusion, resulting in hepatocyte injury and dysfunction (McKeown et al., 1988). This contributes to the development of primary nonfunction or impaired primary function after LT. However, some studies have called the importance of sinusoidal injury into question. Huet et al., 2004, have demonstrated that damage to the extracellular matrix from prolonged preservation and reperfusion appears to be the critical factor in graft failure (Banga et al., 2005). In addition, it is possible that perturbations in hepatocyte levels of adenine nucleotides during cold storage can trigger proteolytic events that contribute to damage in the liver graft and subsequently compromise hepatic functions after LT (Kukan & Haddad, 2001). Moreover, cold ischemia profoundly disturb several key hepatocellular functions, such as volume and pH homeostasis, as well as solute transport and drug metabolism, protein synthesis and mitochondrial function. This contributes to preservation injury of the liver graft. Therefore, these observations indicate that aside from reducing EC damage, LT therapy may benefit from strategies aimed at improving the maintenance of appropriate hepatocyte functions

Apoptosis has been regarded as the fate of cells experiencing I/R injury (Sasaki et al., 1996). In this line, different studies have demonstrated apoptotic death in hepatocytes and/or SECs after both cold and warm ischemia of the rat liver (Gao et al., 1998; Kohli et al., 1999). All of the aforementioned studies (Gao et al., 1998; Kohli et al., 1999; Sasaki et al., 1996) used TdT-mediated dUTP-biotin nick and labelling (TUNEL staining) for DNA ladders to demonstrate apoptosis. However, the ability of TUNEL staining to distinguish between apoptosis and necrosis has been called into question. The activation of caspases has also been used to demonstrate apoptosis in rat SECs following cold I/R (Natori et al., 1999). Indeed, use of pan-caspase inhibitors protected rat liver SECs and hepatocytes against I/R injury after prolonged periods of both cold and warm ischemia. On the other hand, other groups oppose the view that the majority of cells undergo apoptosis in response to either warm or cold I/R injury, believing that necrosis is the principle form of cell death (Massip-Salcedo et al., 2007). They believe that in a number of studies the proportion of cells undergoing apoptosis is not of significant magnitude and that the degree of caspase activation does not correlate with the number of SECs and hepatocytes supposedly undergoing apoptosis. Thus, a controversy has emerged over the past years as to whether

antiselectin therapy against intestinal reperfusion injury (Kubes et al., 2002).

**3.3 Cell death in liver transplantation** 

(Kukan & Haddad, 2001; Vajdova et al., 2002).

necrotic or apoptotic cell death accounts for the severe parenchymal injury observed during hepatic reperfusion. Although it has long been assumed that necrosis and apoptosis are different processes this may not actually be the case. First we will briefly review some basic background information on death cell signalling pathways in hepatocytes in order to understand the shared pathway that leads to both necrosis and apoptosis.

Apoptosis occurs through two main pathways. The first, referred to as the intrinsic (mitochondrial) pathway, is typically activated by a variety of stressors such as DNA damage, p53 activation, growth factor deprivation, and metabolic disturbances (Malhi et al., 2006). The second is the extrinsic pathway that is triggered through death receptors (Malhi et al., 2006). It is well known that one of the most important regulators of intrinsic pathway is the Bcl-2 family of proteins. The Bcl-2 family includes proapoptotic members such as Bax, Bak, Bad, Bid and antiapoptotic members such Bcl-2, Bcl-Xl and Bcl-W (Ghobrial et al., 2005). Following death signal, proapoptotic proteins undergo posttranslational modifications resulting in their activation and translocation to the mitochondria. Then, the outer mitochondrial membrane becomes permeable, leading to the release of Cyt c, which promotes caspase 9 activation, which then activates caspase 3 and the final stages of apoptosis (Ghobrial et al., 2005). In the extrinsic pathway, a variety of mediators, including tumor TNF-α, Fas ligand, and tumour necrosis factor-related apoptosis-inducing ligand (TRAIL) first bind to their respective death receptors, which cause receptor oligomerization and the association of various adapter proteins, including Fas-associated death domain, TNF-α receptor-associated death domain, and TNF-α receptor-associated factor. Fasassociated death domain and TNF-α receptor-associated death domain promote binding of procaspase 8 and its proteolytic activation to catalytic caspase 8. If sufficient amounts of caspase 8 are generated at the receptor, caspase 8 can directly activate procaspase 3. In hepatocytes the caspase 8 interacts with the intrinsic pathway and cleaves Bid, a BH3 only proapoptotic Bcl2 family member, to a truncated form, tBid. tBid translocates to mitochondria, causing mitochondrial permeabilization and release of mitochondrial effectors of apoptosis, such Cyt c (Yin, 2000) (see Fig. 2).

The mechanisms that induce the release of mitochondrial intermembrane proteins such as Cyt c remain controversial (Jaeschke & Lemasters, 2003). In hepatocytes TNF-α and Fas dependent signalling induce the onset of the mitochondrial permeability transition (MPT), which leads to large-amplitude mitochondrial swelling, rupture of the outer membrane, and release of Cyt c and other proteins from the intermembrane mitochondrial space (Jaeschke & Lemasters, 2003). In some models, tBid interaction with either Bax or Bak, forms channels in the mitochondrial outer membrane that release Cyt c and other proteins from the intermembrane space. If MPT onset occurs in relatively few mitochondria, the organelles become sequestered into autophagosomes for lysosomal digestion, a process that eliminates the damaged and potentially toxic mitochondria (Casillas et al., 2006; Jaeschke & Lemasters, 2003). When the MPT involves more mitochondria, mitochondrial swelling leads to outer membrane rupture and Cyt c realease. Provided that ATP is available from glycolysis and still-intact mitochondria, Cyt c activate downstream caspases and other executioner enzymes of apoptosis. When MPT onset is abrupt and involves most mitochondria, ATP becomes profoundly depleted, which blocks caspase activation. Instead, ATP depletion culminates with plasma membrane rupture and the onset of necrotic cell death (Jaeschke & Lemasters, 2003). Hence, the new term "necrapoptosis" has been coined to describe a process that begins with a common death signal and which culminates in either cell lysis

(necrotic cell death) or programmed cellular resorption (apoptosis), depending on factors such as the decline of cellular ATP levels (see Fig. 2).

Fig. 2. Scheme of possible cell death pathway in hepatic I/R. (Alfany et al., 2009; Ben Mosbah et al., 2010; Casillas et al., 2006; Fernández et al., 2004; Ghobrial et al., 2005; Jaeschke &Lemasters, 2003; Malhi et al., 2006; Massip-Salcedo et al., 2007; Selzner et al., 2000; Yin, 2000)
