Remote Ischemic Preconditioning: A Novel Protective Method From Ischemia Reperfusion Injury—A Review

Background

Restoration of blood supply to an organ after a critical period of ischemia results in parenchymal injury and dysfunction of the organ referred to as reperfusion injury. Ischemia reperfusion injury is often seen in organ transplants, major organ resections and in shock. Ischemic preconditioning (IPC) is an adaptational response of briefly ischemic tissues which serves to protect against subsequent prolonged ischemic insults and reperfusion injury. Ischemic preconditioning can be mechanical or pharmacological. Direct mechanical preconditioning in which the target organ is exposed to brief ischemia prior to prolonged ischemia has the benefit of reducing ischemia-reperfusion injury (IRI) but its main disadvantage is trauma to major vessels and stress to the target organ. Remote (inter organ) preconditioning is a recent observation in which brief ischemia of one organ has been shown to confer protection on distant organs without direct stress to the organ. Aim: To discuss the evidence for remote IPC (RIPC), underlying mechanisms and possible clinical applications of RIPC.

Methods of search

A Pubmed search with the keywords “ischemic preconditioning,” “remote preconditioning,” “remote ischemic preconditioning,” and “ischemia reperfusion” was done. All articles on remote preconditioning up to September 2006 have been reviewed. Relevant reference articles from within these have been selected for further discussion.

Results

Experimental studies have demonstrated that the heart, liver, lung, intestine, brain, kidney and limbs are capable of producing remote preconditioning when subjected to brief IR. Remote intra-organ preconditioning was first described in the heart where brief ischemia in one territory led to protection in other areas. Translation of RIPC to clinical application has been demonstrated by the use of brief forearm ischemia in preconditioning the heart prior to coronary bypass and in reducing endothelial dysfunction of the contra lateral limb. Recently protection of the heart has been demonstrated by remote hind limb preconditioning in children who underwent surgery on cardiopulmonary bypass for congenital heart disease. The RIPC stimulus presumably induces release of biochemical messengers which act either by the bloodstream or by the neurogenic pathway resulting in reduced oxidative stress and preservation of mitochondrial function. Studies have demonstrated endothelial NO, Free radicals, Kinases, Opioids, Catecholamines and K_ATP channels as the candidate mechanism in remote preconditioning. Experiments have shown suppression of proinflammatory genes, expression of antioxidant genes and modulation of gene expression by RIPC as a novel method of IRI injury prevention.

Conclusion

There is strong evidence to support RIPC. The underlying mechanisms and pathways need further clarification. The effective use of RIPC needs to be investigated in clinical settings.

Introduction

The restoration of blood supply to organs after a certain period of no flow ischemia results in parenchymal damage referred to as ischemia-reperfusion injury (IRI). The critical ischemia period is dependent on the organ and is 15–20 min [1] in the liver and kidney, 2.5 h in skeletal muscle [2, 3, 4], whereas in the brain ischemia for more than 5 min, leads to considerable neuronal death and infarction. Reperfusion following periods exceeding the critical ischemia period results in endothelial and parenchymal injury. The liver is resilient to hypoxic injury. Low-flow ischemia found in hemorrhagic shock (mean arterial pressure [MAP]-40 mmHg for 120 min) followed by restoration of normal flow does not lead to activation of Kupffer cells, generation of free radicals and associated IRI in the initial resuscitation period [1]. This is because most Kupffer cells are located in the periportal region and hemorrhagic shock is characterized by ischemia in the pericentral regions with sinusoidal perfusion failure in the periportal region as a result of which Kupffer cells are not affected by pericentral hypoxia.

Following a period of ischemia, tissues adapt to anaerobic metabolism [5]. Restoration of blood supply results in oxygen supply in excess of the requirements that lead to activation of macrophages in the vasculature and consequently generation of super oxide radicals, also referred to as reactive oxygen species (ROS), causing oxidative stress. The key event in the initial phase of reperfusion injury is activation of macrophages that are the primary source of extracellular ROS. ROS are the key initiators of reperfusion injury, which leads to endothelial injury and further release of pro-inflammatory cytokines. Thus neutrophils are the key cells in the late phase of IRI.

IRI often happens following transplantation of organs, major organ resections, and trauma. IRI following transplantation can lead to primary nonfunction of the implanted organ (<5%), primary dysfunction (10–30%), and also multiple organ dysfunction syndrome, resulting in morbidity and mortality in adult and pediatric transplants [6, 7, 8]. In hypervolemic shock resuscitation leads to IRI in brain, gut, and pancreas [9], which results in more extensive tissue infarction. In the liver shock resuscitation reduces its tolerance to subsequent warm ischemia [10]. In major organ resections inflow occlusion [11] and subsequent restoration of blood supply causes reperfusion injury. This can lead to postoperative organ insufficiency. Ischemia reperfusion injury is also associated with chronic rejection due to arteriosclerosis [12] caused by IRI. IRI results in adipocutaneous and musculoskeletal flap necrosis as well as nonfunctioning microvascular flaps.

Protective strategies have been developed for protection of organs from ischemia reperfusion injury, which are referred to as organ preconditioning. IPC is only a method by which the target organ is conditioned prior to the ischemic insult to reduce the extent of injury. IPC could be mechanical or pharmacological. Mechanical preconditioning is either direct or remote (indirect). In IPC a brief direct ischemic insult to the target organ followed by reperfusion results in tolerance to subsequent insults of ischemia. Ischemic tolerance is induced by regulation of endothelial function, blood flow, and decreased macrophage as well as neutrophilic activity. This results in decreased endothelial injury and eventually decreased parenchymal injury. Direct IPC has been investigated as a surgical tool for many years [13]. Although direct IPC does reduce reperfusion injury [14, 15, 16] as well as its systemic consequences [17, 18], its main disadvantage is direct stress to the target organ and mechanical trauma to major vascular structures, which have limited its clinical application.

Remote ischemic preconditioning (RIPC) is a novel method where ischemia followed by reperfusion of one organ is believed to protect remote organs either due to release of biochemical messengers in the circulation or activation of nerve pathways, resulting in release of messengers that have a protective effect. This protects target tissue without direct stress. RIPC was first demonstrated in myocardium [19] by McClanahan in 1993. He found that ischemia in the kidney followed by reperfusion protected myocardium from ischemia and reduced infarct size. In animal models brief ischemia reperfusion of the limb, gut, mesenteric, or kidney reduces myocardial infarct size. In humans skeletal preconditioning has been used for myocardial protection with the beneficial effect being attributed to regulation of endothelial protection [20].

The aim of this article was to discuss the evidence for existence of remote preconditioning, the underlying mechanisms and pathways in different methods of remote preconditioning, the mediators of preconditioning, and the possible clinical applications of remote preconditioning.

A PubMed search with the keywords “ischemic preconditioning,” “remote preconditioning,” and “remote ischemic preconditioning” and “ischemia reperfusion” was done. All articles on remote preconditioning up to September 2006 have been reviewed. Relevant reference articles from within these have been selected for further discussion.

The very first evidence of RIPC can be traced to 1993 when Przyklenk conducted regional myocardial preconditioning in an experimental dog model [21]. In this model the circumflex coronary artery was occluded four times for 5 min followed by 5 min reperfusion prior to 1 h of sustained left anterior descending coronary artery occlusion. A significant reduction in myocardial infarct size was seen as compared to non-preconditioned groups. Nakano et al. questioned the existence of RIPC as in an experimental Langendorff model they showed two cycles of 5 min occlusion of a branch of coronary artery followed by reperfusion (5 min) prior to global cardiac ischemia protected the myocardium supplied by the arterial branch but not the rest of the myocardium against sustained ischemia [21, 22]. However, their experiments were in a rabbit model and they used a different preconditioning protocol. Moreover, their argument is not supported by any other studies. Hence evidence for nonexistence of RIPC in other species cannot be concluded from their study.

RIPC is not confined to within an organ and can be transferred from one organ to another.

Spontaneous ischemic events in the brain have been shown to induce adaptation of the heart to ischemia [23] and Tokuno et al. were the first to demonstrate in a mouse model decreased myocardial infarction and improved myocardial function in mice by prior brain ischemia (bilateral internal carotid ligation) with no evidence of brain damage except for transient neurological patterns [23]. Importantly, this study demonstrated that ischemia rather than reperfusion stimulated signaling pathways, which led to remote protection.

Five studies showed that brief renal ischemia reduced the size of infarct resulting from myocardial ischemia [24, 25, 26, 27, 28]. The first evidence of RIPC dates back to 1993 when McClanahan carried out experiments on a rabbit model and showed that a 10 min period of renal ischemia was as effective as a 5 min period of coronary artery occlusion in preconditioning the heart [19]. In 1996, Gho et al. [26] showed that brief periods of mesenteric artery occlusion (MAO) (15 min) and reperfusion prior to 60 min coronary artery occlusion (CAO) was as effective as direct CAO and 15 min renal artery occlusion (RAO) prior to 60 min CAO reduced infarct size under hypothermic conditions but not normothermic conditions, demonstrating the unmasking of effects of RAO by hypothermia. The onset of myocyte injury is associated with adenosine triphosphate (ATP) depletion and breaks in sarcolemmas, which are linked to intracellular acidosis and raised phosphocreatinine levels. Takoaka et al. showed by nuclear magnetic resonance spectroscopy that RAO (10 min) followed by reperfusion (20 min) in rabbits led to a decrease in myocardial infarct size on histology, attenuated depletion of ATP, preserved myocyte pH, and improved recovery of myocardial phosphocreatinine as well as ATP levels during subsequent ischemia-reperfusion (IR) [25]. Two studies, one in a rabbit model [27] and the other in a rat model [28], showed RIPC by RAO reduced myocardial infarct size prior to CAO (30 min) followed by 2 h reperfusion. One study showed that RIPC from kidney was as effective as direct myocardial IPC in reducing infarct size [25]. These studies collectively demonstrate that brief periods of RAO and reperfusion are necessary for RIPC. RIPC was not seen in permanent arterial occlusion, suggesting that infarct size limitation was due to washout of a protective substance during reperfusion. Both rat and rabbit models demonstrated the benefit of RIPC. The extent of myocardial protection by RAO was more in Takoaka's study as compared to McClanahan's, although both used rabbit models, possibly due to differences in core temperatures of the animal models. This remained unclear as McClanahan did not report core body temperature in his study. Further studies in different animal models and in clinical settings are needed to test the efficacy of RIPC by RAO and to define the indications for its application.

Ten studies demonstrated that RIPC from mesentery reduced myocardial infarct size on histology in a rat model. Seven studies [26, 29, 30, 31, 32, 33] showed that RIPC by single cycle of MAO (15 min) and subsequent reperfusion preconditioned the intestine and reduced myocardial infarct size. In contrast, two studies showed that RIPC, by multiple cycles of MAO [34, 35], reperfusion preconditioned the intestine and reduced myocardial infarct size, while one study demonstrated myocardial protection even after induction of myocardial ischemia 24 h later [36]. These studies collectively support the evidence for RIPC by MAO and reperfusion. All studies demonstrated an effective preconditioning protocol with brief MAO and only one study [33] demonstrated that a single cycle of RIPC was more effective than multiple cycles. Only one study (Vlasov [37]) showed that direct IPC is more effective than RIPC and one study (Wang et al. [36]) demonstrated the existence of late phase mesenteric RIPC. Histology is a common endpoint in all studies but myeloperoxidase activity (MPO) activity or myocardial creatinine kinase levels were not evaluated in all studies. One study showed that RIPC reduced myocardial MPO, which is a marker of systemic inflammatory response [36]. Two studies found significant hemodynamic alterations [30, 36] and showed that MAP increased on mesenteric occlusion and decreased on reperfusion with a gradual return to baseline. None of the studies measured ventricular function. Future studies aimed at clarifying the functional importance of mesenteric RIPC, comparisons of preconditioning protocols to establish the ideal protocol, and clinical applications are needed.

Twelve studies have shown the beneficial effects of brief limb ischemia in preconditioning the heart. Oxman et al. showed reduced cardiac tachyarrythmias following brief limb ischemia [38], while Birnbaum et al. demonstrated that gastrocnemius electrical stimulation coupled with brief limb ischemia reduced myocardial infarct size [39]. Konstantinov et al. [40] and Kharbanda et al. [20] by multiple cycles of limb preconditioning and Weinbrenner et al. by a single cycle of preconditioning [41] demonstrated that RIPC reduced myocardial infarct size in animal models. Kharbanda showed that limb preconditioning reduced endothelial injury in humans [20]. IR of the upper limb (ischemia 20 min, 200 mm Hg cuff) blunted its vasodilatory response to acetylcholine; however, RIPC of the contralateral arm (3 × 5 min ischemia) attenuated neutrophil activation following ipsilateral limb IR. Gunaydin et al. showed that RIPC by two cycles of 3 min upper limb ischemia (tourniquet pressure 300 mm Hg) separated by 2 min of reperfusion in patients undergoing coronary bypass showed an equal rise in creatinine phosphokinase-MB in both preconditioned and control groups; however, there was a significant increase in lactate dehydrogenase (LDH) in the preconditioned group with levels of myocardial lactate and lactate efflux twice as high compared to the control group, providing biochemical evidence for maintenance of myocardial anaerobic glycolysis in the preconditioned group [42]. From a clinical perspective RIPC (Kharbanda model, [20]) may be useful in patients undergoing coronary angioplasty for reducing myocyte and endothelial injury and in patients undergoing cardiopulmonary bypass (Gunaydin model, [42]) where RIPC by transient upper limb ischemia can be used to protect the myocardium against subsequent IR when the aorta is clamped. Patients with intermittent claudication may have a better tolerance to myocardial ischemia due to preconditioning by brief limb ischemia and consequently have a longer window period for thrombolytic therapy to salvage ischemic myocardium. Vlasov [43] showed that both brief limb ischemia and cardiac ischemia reduced myocardial infarct size. A significant reduction in nicotine amide dehydrogenase (NADH) diaphorase and LDH activity was seen in the ischemic zone in the IR group, which was attenuated by both RIPC and direct IPC. In addition an increase in NADH diaphorase activity was seen in intact cardiomyocytes of the preconditioned groups as compared to IR group animals. This provided evidence that both RIPC and direct IPC induced some form of metabolic activity and ischemic adaptation in both ischemic and intact cardiac myocytes. This is supported by Gunaydin et al. [42], who showed attenuated LDH activity due to anaerobic glycolysis in preconditioned myocardium. Li et al. [44] in a mouse model demonstrated that RIPC significantly reduced myocardial infarct size. Expression of nuclear factor-kappa B (NF-κB) proteins from both limb skeletal muscle after RIPC and myocardium suggested the induction of protective signals in the limb being transferred to the heart and leading to ischemic adaptation. Two studies demonstrated the role of RIPC in a rat cardiac transplant model. Kristiansen et al. preconditioned the donor (rat model, 4 × 5 min IR cycles, hindlimb ischemia), which significantly reduced coronary IR in donor hearts on implantation in the recipient [45]. In contrast, Konstantinov et al. preconditioned the recipient (rat model, 4 × 5 min IR cycles, hindlimb), which reduced myocardial infarct size in the donor heart on implantation [46].

Eight studies have demonstrated the beneficial effect of skeletal preconditioning on adipocutaneous, muscle, and cremasteric muscle flaps as well as remote skeletal tissue [47, 48, 49, 50, 51, 52, 53, 54]. Moses et al. [47] demonstrated reduced LD muscle flap infarct size in a pig model following RIPC (3 × 10 min). Contralateral hindlimb RIPC [48, 50, 51, 52, 53] prior to ipsilateral cremasteric flap ischemia (2 h) and reperfusion in a rat model improved RBC flow and reduced neutrophil adhesion. Similarly Liauw et al. in a rat model showed that RIPC (ipsilateral Gracilis) reduced muscle necrosis of contralateral muscle by 60% as compared to non-preconditioned [54].

RIPC reduced flap necrosis, which is of particular benefit in patients with irradiated tissues, smokers, and obese patients. This protective effect is more pronounced in the late phase of preconditioning. These studies showed remote ischemic preconditioning to be associated with better microcirculation, decreased leukocyte endothelial sticking, and endothelial dysfunction as well as better capillary blood flow with terminal arteriolar dilation [55].

The following four studies have shown hindlimb ischemic preconditioning to be beneficial in protection of the lung and the brain [56, 57, 58]. (1) RIPC by hindlimb ischemia (3 × 5 min), 5 min reperfusion prior to 2 h bilateral hindlimb ischemia, and 2.5 h reperfusion in a porcine model [58] protected against lung dysfunction secondary to limb IRI. RIPC reduced plasma cytokine levels, ameliorated impaired gas exchange and oxygen transport, reduced the elevation in pulmonary arterial pressure and vascular resistance, reduced pulmonary edema, and decreased lung tissue myeloperoxidase activity. During surgery of the lower limbs, there is release of cytokines after prolonged limb ischemia leading to acute respiratory distress syndrome and limb preconditioning could protect against lung dysfunction found with acute respiratory distress syndrome. (2) RIPC by left femoral artery occlusion (3 × 5 min) prior to lung IRI (90 min ischemia, 5 h reperfusion) [59] in a porcine model reversed the detrimental effects of lung IRI on lung function and attenuated pulmonary hypertension and impaired gas exchange. Thus, both models of RIPC were effective in protecting the lung. Harkin et al. demonstrated that IPC of the hindlimb prior to limb IR injury reduced cytokines (IL-1 and IL-1β), and activated neutrophils, systemic inflammatory response syndrome, and lung dysfunction [58]. In contrast, Waldow's experimental model demonstrated that RIPC (hindlimb ischemia) protects the lung from acute IR injury but does not modulate all indices of systemic inflammation (IL-6, ROS, and activated granulocytes were not modulated by RIPC) [59]. (3) Xia et al. [57] showed that RIPC by three episodes of 5 min occlusion and 5 min reperfusion of the iliac artery preserved lung function and prevented a rise in pulmonary vascular resistance as well as arterial pressure following myocardial reperfusion injury. This has important clinical applications in cardiac surgery as coronary IR has deleterious effects on the lung, which is a major cause of mortality in these patients. In beating heart surgery it is not possible to directly clamp the aorta and therefore direct cardiac preconditioning is not possible. Under such circumstances preconditioning from a remote organ such as the limb may be useful as shown by Xia et al. [57]. (4) RIPC by femoral artery occlusion (30 min), reperfusion (15 min) and (30 min) occlusion, 48 h (reperfusion) [56] prior to IR (30 min brain ischemia by carotid artery occlusion and 48 h reperfusion) reduced brain edema and circulating endotheliocytes and improved blood flow, demonstrating ischemic adaptation in the early and late phase of IRI.

Most models of limb preconditioning supported the effectiveness of multiple brief cycles of limb ischemia followed by reperfusion; conversely, Weinbrenner et al. showed a single cycle to be more effective than multiple cycles and prolongation of length of the single cycle led to more effective preconditioning [41].

Four studies showed the protective effects of brief myocardial and hepatic ischemia on remote organs [17, 24, 60, 61]. Both brief hepatic and myocardial ischemia have protective effects on the stomach [60, 61]. In an experimental rat model, Brzozowski et al. showed that two 5 min episodes of hepatic/myocardial ischemia followed by 10 min of reperfusion each was as effective as direct gastric preconditioning in reducing gastric erosions as well as increasing gastric blood flow following sustained gastric IR. Ates et al. demonstrated in a rat model that brief hepatic ischemia (10 min) prior to 45 min of ischemia in the left kidney was associated with better creatinine clearance as well as improved sodium fractional excretion 24 h after preconditioning. They showed reduced mitochondrial swelling, basement membrane detachment on electron microscopy, reduced renal tubular swelling, necrosis, tumor necrosis factor-alpha (TNF-α) levels, decreased lipid peroxidation (TBARS levels) [24], and a relatively rapid decline in LDH levels in the remote preconditioned group as compared to IRI group. This study provides evidence of preservation of ultrastructural, histopathological, and biochemical renal function in the RIPC group and supports data for beneficial effects of brief hepatic ischemia on remote tissues.

In an experimental rat model Liem et al. demonstrated that two cycles of CAO of 15 min ischemia followed by 15 min reperfusion preconditioned the myocardium prior to 60 min CAO and reduced infarct size, which was abrogated by 8-sulfophenyl theophylline (adenosine receptor blockade) prior to IPC, suggesting the role of adenosine in IPC. However, four cycles of 15 min CAO followed by 15 min reperfusion rendered the myocardium tolerant to IPC and subsequent preconditioning was ineffective in reducing myocardial infarct size [62] due to depletion of cardiac interstitial adenosine. However, RIPC with two cycles of MAO 15 min in tolerant myocardium was effective in reducing myocardial infarct size, suggesting alternate signaling pathways. This data suggests that repetitive brief ischemia of same duration may render tissue tolerant to preconditioning; however, in the clinical setting patients with unstable angina are unlikely to develop tolerance to adenosine since angina is of varying severity and duration. However, exogenous adenosine is effective in ameliorating IRI even in those who develop tolerance, as shown by Liem et al., and RIPC may be useful in situations where direct preconditioning is ineffective due to adenosine tolerance.

The role of remote trauma in preconditioning has been addressed in a recent study by Ren et al. [63]. They demonstrated that carotid artery catheterization (remote nonischemic vascular surgical trauma) aggravates myocardial ischemia; however, abdominal incision (remote nonischemic nonvascular surgical trauma) reduces myocardial infarct size following cardiac IR. This effect is more in the early phase of remote preconditioning (80% reduction in infarct size) and less in the late phase of preconditioning (40% reduction in infarct size). Remote preconditioning of trauma (RPCT), unlike ischemic preconditioning or remote ischemic preconditioning, does not involve ischemic insults to initiate preconditioning in remote organs. There is both an early and a late preconditioning phase in RPCT. The underlying mechanism of RPCT is unclear. RPCT further reduces myocardial infarct size in TNF-α knockouts, supporting the argument that TNF-α does not mediate remote preconditioning of trauma [63]. The role of adenosine in RIPC of the heart and noradrenaline in IPC has been shown. Therefore adenosine activity, sympathetic neuronal activity, and catecholamines are potential mediators of RPCT. Since norepinephrine is involved in cross-signaling with protein kinase C (PKC) [64], nitric oxide (NO) [65] modulates release of norepinephrine from skeletal muscle in ischemia and both PKC and NO play a role in remote preconditioning. One may speculate the role of catecholamines, PKC, and NO pathways in RPCT. Further studies in animal models using sympathetic blockade, knockouts of NO and PKC, and ganglion blockade prior to RPCT are needed to pinpoint the candidate mechanism and elucidate the pathway of signal transmission.

Ischemic preconditioning of the heart followed by transfer of coronary effluent from the preconditioned heart to the recipient heart proved beneficial in protecting the recipient heart from ischemia reperfusion injury [66, 67, 68, 69]. The mechanism of transferred preconditioning was via opioids such as metencephalins and not due to epinephrine or adenosine release into the effluent. Naloxone blocked opioid receptors and abrogated the beneficial effect of transfer of coronary effluent from the preconditioned heart.

Coronary effluent from preconditioned hearts is also effective in preconditioning segments of jejunum via opioid receptors and K_ATP channels [66]. Mesenteric ischemic tolerance induced by pretreatment of small bowel segments with coronary effluent leads to quicker recovery of contractile function of small bowel following reoxygenation and reperfusion.

Adenosine, NO, TNF-α, opioids, bradykinins, PKC, calcitonin gene related peptide (CGRP), cyclo-oxygenase, K_ATP channels, capsaicin, heat shock proteins, and norepinephrine are all involved in the mechanism of remote ischemic preconditioning. These substances are released as a response to stress and act via the neuronal or humoral pathway to produce organ protection. The pathways involved are different in response to different ischemic stimuli and often overlap. The evidence for the role of each of these compounds and the mechanisms by which they act and the pathways involved are reviewed.

In direct IPC there is evidence for different phases of ischemic adaptation and protection [70]. These phases were first identified in the heart and subsequently in other organs. In the heart the early phase protects against infarction but not against myocardial dyskinesia [70], whereas the late phase protects against both. The early phase begins soon after reperfusion and lasts for up to 3 h in ischemic preconditioning, whereas the late phase starts 12–24 h later [71]. The early phase is independent of protein synthesis and is due to release of endogenous substances, which stimulate posttranslational modifications in proteins, whereas the late phase is stimulated by release of endogenous substances, which lead to synthesis of new proteins and altered gene expression. This is referred to as the second window of protection [72]. The effects of the acute phase are short lived, lasting for 3–4 h, whereas the effects of the delayed phase are longer, lasting for 48–96 h or sometimes for weeks [25, 27, 28].

The protective effect of the delayed phase of preconditioning is less compared to the early phase and mechanistically different [70].

Eight studies have shown two phases of protection (ischemic adaptation) in RIPC [27, 29, 30, 36, 38]. These studies showed the existence of both phases in all organ systems in animal models and in recent human studies. (1) Two studies [25, 27, 28] showed brief MAO-induced early and delayed preconditioning in the heart. (2) Li et al. showed a pronounced reduction in myocardial infarct size 24 h after hindlimb RIPC [44]. (3) Four studies demonstrated late RIPC in remote skeletal muscle following skeletal muscle preconditioning [51, 52, 53]. (4) One study showed ischemic adaptation of the brain to hindlimb ischemia after 24 h of preconditioning [56]. (5) One study showed late phase of RIPC in the heart and liver due to heat shock protein (HSP) expression. (6) The first clinical application of RIPC in children undergoing cardiac surgery on bypass showed that RIPC was most effective in myocardial protection if applied 24 h prior to the coronary ischemic insult [73].

Four studies [25, 26, 27, 28] in renal preconditioning and four studies [30, 32, 33, 55] in mesenteric preconditioning showed that remote organ protection was associated with brief periods of reperfusion of the remote organ; however, persistent ischemia without reperfusion did not protect remote organs. Liem et al. [32] showed that brief MAO and reperfusion conferred myocardial protection demonstrating that reperfusion of virgin intestine was necessary for activation of the neurogenic pathway. Subsequent permanent MAO reduced infarct size significantly more than brief MAO alone, suggesting that brief periods of MAO are cardioprotective but not optimal as their effect was enhanced by further permanent occlusion. Four studies demonstrated the role of the neurogenic pathway in transduction of RIPC from the limb. In an animal model, Oxman et al. [38] showed rise in plasma catecholamine levels on preconditioning, which was abolished by autonomic nerve blockade with reserpine. Loukogeorgakis et al. showed the role of the autonomic nervous system in transduction of the protective signal. However, they were unable to define the specific component of the autonomic system [74]. Birnbaum et al. demonstrated RIPC by electrical stimulation of gastroenemius muscle in conjunction with brief limb ischemia in a rabbit model [39] and Kharbanda showed RIPC from the limb was abolished by sympathetic nerve blockade with reserpine [20]. In contrast six studies demonstrated rises in plasma levels of nitrates, opioids, free radicals, and catecholamines [36, 38, 41, 49, 75, 76, 77] following RIPC, providing evidence in support of a humoral pathway. In Weinbrenner's study [41], the protective effect was only seen in groups who had a period of reperfusion after ischemia in comparison to those who had no reperfusion after ischemia [25, 78] and simultaneous aortic occlusion along with coronary occlusion did not confer protection, indicating that preconditioning had to be prior to allow for the substance released to reach the heart. These findings demonstrate the release of protective substances into the circulation. Autonomic nerve blockade with hexamethonium did not abolish RIPC [75]. Two studies showed cardioprotection in a rat cardiac transplant model (denervated heart), suggesting the role of blood-borne factors in preconditioning [45, 46]. Whether the neurogenic pathway was activated locally in the mesenteric, renal, or skeletal beds or by release of mediators into the circulation (humoral pathway) and subsequent stimulation of sensory afferent fibers is unclear from these studies. The need for reperfusion to confer remote protection and rise in plasma levels of catecholamines, adenosine, neuropeptides, cytokines, or free radicals suggests that these substances may activate neuronal pathways after release into circulation. It seems that both the neurogenic pathway and the humoral pathway have some element of overlap and are not mutually exclusive. Measurement of interstitial levels of mediators and sensory afferent nerve activity prior to and after preconditioning and specific blockade of synthesis and release of the mediators would help clarify their role in activation of neuronal pathways. Blockade of neuronal pathways and end receptors would help define the predominant pathway for transduction of preconditioning in different organ systems.

Studies in direct IPC have demonstrated the protective effect of NO on microcirculation [14, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90]. NO is a free radical produced from l-arginine by the enzyme NO synthase, which has three forms: endogenous NOS (eNOS), neuronal NOS (nNOS), and inducible NOS (iNOS) [82, 83]. Of these, eNOS and nNOS are constitutively expressed, while iNOS is produced in response to cytokines and NF-κB. eNOS has a protective effect on microcirculation and always produces NO in small amounts, which predominates in the circulation, producing a protective effect on the microcirculation.

Six studies have demonstrated the role of NO [48, 50, 51, 52, 53, 56] in remote preconditioning of skeletal muscle, intestine, brain, and heart. These studies showed hindlimb RIPC induced protection of muscle flaps in the early and late phase of RIPC [36, 51, 52], induction of myocardial iNOS following mesenteric RIPC, and reduction in myocardial infarct size as well as MPO activity (marker of neutrophilic activation). Blockade of NO activity abrogated the decrease in myocardial infarct size and MPO activity, suggesting that NO inhibits neutrophil infiltration. Tokuno et al. showed that spontaneous brain ischemic events 24–48 h prior to cardiac IR injury reduced myocardial infarct size [23, 49] with loss of RIPC in iNOS knock out mice. Vlasov et al. [56] showed recovery of brain blood flow at 120 min after reperfusion and reduced cerebral edema in the late phase of RIPC (hindlimb) due to iNOS-induced cerebral ischemic adaptation. A recent study has shown expression of NF-κB due to brief limb ischemia and NF-κB-induced iNOS 24 h later, which reduced myocardial infarct size [44] with loss of protection in iNOS knockout mice. This observation is supported from studies of direct IPC that have demonstrated that cytokines induce iNOS production. Conversely Petrischev showed that NO is not involved in RIPC [43] since the nonspecific NO blocker l-NNA did not abolish increase in nicotinamide adenine dinucleotide phosphate activity in intact cardiomyocytes following limb/mesenteric remote ischemia [43] or block attenuation of decrease in nicotinamide adenine dinucleotide phosphate activity in ischemic tissue by preconditioning. These studies demonstrate the role of NO in acute RIPC of the heart and muscle but its role in acute preconditioning of the intestine and brain remains unproven. IR injury impairs endothelial function in the initial phase primarily by impairment of formation and bioavailability of NO, which may explain the lack of effect of the early phase of RIPC in the brain and intestine. This may also be explained by different protocols of preconditioning used for remote preconditioning in different organs and different criteria for assessment of preconditioning in the heart and muscle (cytoprotective effects) as against endothelial function in the intestine and brain. Therefore similar outcome measures are needed to investigate the role of NO.

In a mouse model increased NF-κB expression in the limb following RIPC induced NF-κB and iNOS in the heart, suggesting the role of cytokines in the signaling pathway for induction of NO [44]. In a rat model, brief hepatic ischemia increased gastric CGRP, increased mucosal flow [34], and reduced gastric mucosal erosions and this effect was abolished by l-NAME (NO blocker), suggesting that CGRP induces NO release and subsequent vasodilatory effects and increased flow. Evidence suggests release of ROS, cytokines, and NO into the venous effluent following intestinal ischemia reperfusion [91] activates NF-κB and PKC [92]. These signaling pathways induce iNOS [93] in the target organ. Peralta has shown adenosine increases NO formation and this effect is blocked by adenosine antagonists [94].

The data from the studies discussed show that brief periods of ischemia reperfusion in RIPC induce NO pathways. Li et al. showed that RIPC by hindlimb ischemia increased NF-κB in skeletal muscle and heart and induced ischemic adaptation in the heart by iNOS formation [44]. In view of the short half-life of NO (5 s), it is unlikely for NO to be produced in the remote organ and reach the target organ to confer protection by the blood stream. Chen et al. [75] showed that RIPC of the limb reduced myocardial infarct size through NO production. Since hexamethonium (autonomic ganglion blocker) did not abrogate reduction in myocardial infarct size, it seems that NO pathways act through the bloodstream. In an animal model of warm hepatic IRI, Kanoria's group showed increased hepatic venous plasma nitrates/nitrites [77] and amelioration of hepatic IR following hindlimb RIPC. This provides further support for the argument in favor of the bloodstream being the NO pathway. Clearly, further clarification of this is needed by measuring systemic venous plasma nitrate/nitrite levels and plasma arginine levels (NO precursors) after RIPC. Also, knockout models of NOS would be specific for investigating the role of NO pathways in preconditioning.

Tokuno et al. demonstrated reduced myocardial infarct size following induced brief brain ischemia without an increase in cardiac iNOS; however; the protection was abolished in NOS knockouts, suggesting the role of NO as a trigger [23]. Blockade of NO prior to RIPC by l-NAME abolished myocardial protection, supporting the role of NO as a trigger [75]. Increased iNOS production 24 h after RIPC would suggest its role as a mediator and future studies by blocking NO after RIPC are needed to demonstrate its role as a mediator.

IPC studies have demonstrated that NO modulates microvascular perfusion through its vasodilatory effect [95] and through its anti-inflammatory actions, including inhibition of stellate cell activation [96], neutrophil adhesion [88, 90, 97], and platelet aggregation. NO plays a key role in initiating and maintaining preconditioning. The early phase of preconditioning mediated by eNOS is through generation of cGMP [98], as shown by Lochner et al. in the myocardium [99] and subsequently inhibition of cAMP levels as well as reduction in energy demands. The late phase is protein synthesis dependent and is through activation of PKC, NF-κB, and transcription of iNOS [100]. In the liver NO mediates preconditioning by inhibitory actions on endothelin [14], activation of adenosine A2 receptors, and subsequent NO formation [94]. NO also has been shown to confer protection against cold ischemia [15] of liver. NO inhibits apoptosis of cells by inhibition of caspase activity, TNF-α, and up-regulation of Bcl-2 [101, 102, 103]. NO has a protective effect on intestinal microcirculation such as scavenging of oxygen free radicals, maintenance of normal vascular permeability, inhibition of smooth muscle proliferation, reduction of leukocyte adherence to the mesenteric endothelium, prevention of mast cell activation, and platelet aggregation [84, 86, 88, 104]. Kubes et al. [83] have demonstrated in the intestine that eNOS has a protective effect on intestinal mucosa, while increased iNOS has been shown to increase mucosal apoptosis by generation of free radicals such as peroxynitrate [105, 106]. However recent studies have shown that the production of peroxynitrate may be associated with loss of eNOS rather than increased iNOS production [107]. This observation suggests a dichotomous role for NO in IR injury with small quantities of NO produced by eNOS, reducing IRI, while excessive NO due to iNOS, causing deleterious effects.

Adenosine is an extracellular molecule that is both a trigger and a mediator of IPC as demonstrated from past studies [108]. Adenosine is a hormone widely distributed in human tissues. Adenosine production occurs in myocytes, endothelial cells, and vascular cells. During ischemia of the heart, brain, and the kidney, the imbalance between oxygen supply and demand results in net breakdown of adenosine triphosphate (ATP) and release of adenosine, which can increase up to 50 fold.

Three studies have shown the role of adenosine in RIPC of the heart by RAO and MAO. Two studies in a rabbit model [25, 27] showed RIPC by RAO (10 min); reperfusion (10 min) prior to CAO reduced myocardial infarct size, and 8-SPT (adenosine blocker) abolished the protective effects, while RIPC by single cycle of MAO (15 min) and reperfusion prior to CAO in a rat model increased plasma adenosine levels and reduced myocardial infarct size.

The following two studies demonstrated the role of adenosine in RIPC of skeletal muscle. Adenosine release has been shown to be an effector molecule in skeletal muscle IPC [109]. (1) RIPC increased adenosine plasma levels and the protective effect was partially blocked with reserpine. (2) Prior adenosine blockade [49] did not completely abolish latissmus dorsi flap protection by RIPC (limb ischemia, 3 × 10 min cycles); however, adenosine blockade (8-SPT) and free radical scavenger mercaptopropionyl glycine (MPG) completely abolished the RIPC effect, suggesting that adenosine plays a partial role in RIPC.

In a rabbit model adenosine blockade prior to RIPC abolished cardioprotection, demonstrating its role as a trigger of RIPC [27]. Adenosine blockade after RIPC before reperfusion abolished cardioprotection, suggesting its role as a mediator of RIPC also.

Following ischemia-induced ATP breakdown, adenosine crosses the cell membrane and enters into the interstitial space by simple diffusion. From the interstitial space it escapes into the intravascular space by paracellular washout (slow, 10% under physiological conditions). It would seem that in RIPC adenosine produced in the remote organ would reach its target organ by the bloodstream; however, it has a half-life of 0.6–1.5 s as it is rapidly taken up by endothelial cells, RBCs, and pericytes, which contain nucleoside transporters and are responsible for rapid degradation of adenosine [110], making it unlikely for adenosine to reach its target by the circulation. There are no studies in RIPC that demonstrate modulation of adenosine degradation or activation of nucleoside transport inhibitors. Two studies showed a rise in plasma adenosine levels following preconditioning, suggesting that reperfusion and release into circulation is required for adenosine-induced RIPC [25, 27]. One study [32] supported a neurogenic pathway for adenosine in RIPC. They demonstrated abrogation of preconditioning of the heart by brief cycles of MAO and reperfusion due to prior ganglion blockade. To further clarify the pathway, they demonstrated that intramesenteric infusion with adenosine mimicked the effects of brief MAO and prior ganglion blockade abolished the protective effects of RIPC. Ganglion blockade after reperfusion did not abrogate the protective effects but adenosine receptor blockade abolished myocardial protection. These observations suggest that adenosine acts locally to stimulate afferent nerves in the mesenteric bed, which in turn activate adenosine receptors in the heart. This argument is supported by data demonstrating lack of preconditioning following intraportal or intracaval infusion of adenosine and effectively excludes effects of adenosine spillover during mesenteric infusion, which could potentially stimulate adenosine receptors in the liver and contribute to myocardial protection.

8-SPT (adenosine receptor blocker) abolished RIPC by MAO. 8-SPT was given after reocclusion of mesenteric artery [32], preventing adenosine access to the mesenteric bed. This suggested the presence of adenosine receptors on the heart, the effect of which was blocked by 8-SPT.

In RIPC two studies (RAO reduced myocardial infarct size, rabbit model) [25, 27, 28] demonstrated that selective mitochondrial K_ATP channels blockers (5-HD) abolish the protective effect. The action potential of sarcolemmal K_ATP channels was unaffected. Modulation of K_ATP channels reduced ATP depletion, preserved intracellular pH, and enhanced recovery of ATP and phosphocreatinine levels during reperfusion [25]. Decreased acidosis reduced intracellular Ca accumulation and myocardial infarct size. These observations are supported by IPC studies [109] and suggest that KATP channels serve as end effectors in adenosine pathways.

IPC studies have shown that adenosine mediates both the early and the late phases of preconditioning via different end-organ receptors. In the heart adenosine acts via the A1 receptor. Adenosine receptors mediate anti-adrenergic effects indirectly by reducing C-AMP levels. The A1 receptor-mediated effects involve activation of PKC, tyrosine kinases [111], heat shock protein, and MAPK. Kinases modulate end effector mitochondrial KATP channels. A2 receptors act in the hepatic microcirculation as shown by Peralta et al. [94]. A2 receptors are linked to vasodilatation and antiplugging effects of adenosine. Activation of these receptors and their effects are due to endothelium dependent and independent mechanisms, indirectly through release of NO and through direct relaxation of vascular smooth muscles. A3 receptors are found in myocytes.

The effects of adenosine include vasodilatation [110], inhibition of leukocyte adhesion, neutrophil and platelet function [112], and free radical production. These have to be clearly demonstrated in future studies in RIPC.

Li et al. in a mouse model showed RIPC (six cycles brief hind limb ischemia) prior to myocardial IR reduced infarct size. IRI activated NF-κB in hearts but prior RIPC attenuated activation of NF-κB in IR [44]. However RIPC in sham animals demonstrated NF-κB activation in both the limb and hearts. This study demonstrates a dual role for NF-κB. While excessive NF-κB activation in IR injury has deleterious effects and increases infarct size, activation of NF-κB following limb preconditioning led to an adaptive response in the heart, increased iκβ (inhibitory KB) expression, which attenuated NF-κB activation following sustained IR injury and reduced myocardial infarct size, attenuated decrease in left ventricular developed pressures (LVDF) and increases in left ventricular end-diastolic pressures (LVEDP) on reperfusion. In NF-κB and iNOS knockout models, the decrease in LVDF was attenuated but preconditioning did not confer any additional benefit. NF-κB and iNOS knockouts had less severe increase in LVEDP but no additional attenuation was conferred by preconditioning. NF-κB is induced by ROS and subsequently acts through activation of kinases in the preconditioning response. Preconditioning may down-regulate the inflammatory response during reperfusion as NF-κB activation increases its own inhibitor Iκβ (Tahepold) or it may act through a mediator such as iNOS, as suggested by data from this study.

Increases in TNF-α levels following hepatic IR injury led to remote organ injury (lung and kidney) and IPC of the liver reduced both TNF-α levels and remote organ injury as demonstrated by Peralta et al.[113]. This study suggested that TNF-α may have a role in remote preconditioning. In RIPC two studies have investigated the role of TNF-α. Ates et al. showed raised TNF-α levels following renal IR, which were reduced by preconditioning with brief hepatic ischemia and associated with improved renal function as compared to IR groups [24]. RIPC by brief hepatic ischemia prior to gastric IR (30 min) attenuated plasma TNF-α levels in a rat model [60]. Ren et al. showed in a rat model of cardiac IR reduced myocardial infarct size following IPC and in TNF-α knockout mice. The protective effect of the early phase of IPC was not abolished in TNF-α knockouts but that of the late phase of IPC was abolished in knockouts, suggesting that late IPC is TNF-α dependent. Both the early and the late phase of RPCT (abdominal incision prior to cardiac IR) in TNF-α knockouts further reduced myocardial infarct size, suggesting that remote preconditioning of trauma was mechanistically different from TNF-α ablation [63]. Clinical efficacy of cardioprotective strategies maybe maximized by using a combination of RPCT and TNF-α ablation. This study also showed that blockade of NF-κB and TNF-α was additive, suggesting NF-κB to be involved in the signaling pathway of TNF-α. The mechanism of TNF-α-induced RIPC and IR is unclear and future studies are needed to study the mechanism and signaling pathways involved.

In lung IR injury secondary to remote limb ischemia reperfusion, blockade of interorgan inflammatory mediators such as cytokine IL-6 and levels of primed neutrophils by prior limb preconditioning confers protection with no significant difference in levels of TNF-α. Conversely lung protection against local IRI by remote brief limb ischemia is mechanistically different and entails incomplete blockade of systemic inflammatory response syndrome mediators such as IL-6 and circulating primed neutrophils and complete suppression of IL-1β, which is an early mediator of reperfusion injury. Therefore, despite incomplete blockade, it is likely that remote preconditioning reduces expression of adhesion molecules and neutrophilic infiltration in the lung.

RIPC has been shown to confer protection on the remote organ in lung dysfunction due to limb IR and local IR by blockade of IL-1β. The role of IL-1β was demonstrated by Harkin et al., who clearly showed an increase in IL-1β in limb IR, which led to lung injury and deterioration of lung function [58]. In sustained limb IR, limb IPC protected lung function from remote IR injury by blocking the cytokine IL-1β. A similar rise in IL-1β was shown by Waldow in IRI of the lung. RIPC of the limb ameliorated the rise in IL-1β and conferred protection on the lung. Thus IL-1β has been shown to have a role as an interorgan mediator of IR injury and as a mediator of local IRI [59].

Wolfrum et al. showed that PKC was responsible for reduction in myocardial infarct size after mesenteric ischemia reperfusion and blockade of PKC by a highly selective inhibitor chelerythrine prevented reduction in infarct size [30]. They showed an increase in myocardial PKC following RIPC and blockade of PKC abolished the protective effect. Weinbrenner et al. showed RIPC by infrarenal aortic occlusion for 15 min followed by reperfusion reduced myocardial infarct size, which was abolished by chelerythrine [78].

There is evidence that myocardial PKC undergoes activation following ischemic stimuli in IPC [114, 115, 116]. This leads to conversion of the cytosolic PKC to particulate PKC fraction, thereby increasing the ratio of the particulate to cytosolic fraction. Subsequent activation of mitochondrial PKC receptors results in activation of tyrosine kinases and MAP kinases, which lead to opening up of mitochondrial potassium receptor K_ATP channels. Mitochondrial K_ATP channels serve as end effectors in modulation of mitochondrial energy flow and preservation of mitochondrial membrane integrity. Both humoral and neuronal pathways have an important role in PKC-mediated RIPC. Wolfrum et al. demonstrated increased plasma bradykinin levels following mesenteric ischemic preconditioning, which activated myocardial PKC and blockade of bradykinin receptors with bradykinin antagonist HOE 140 abolished myocardial protection [30]. Ganglion blockade with hexamethonium did not alter the cytosolic to particulate ratio of PKC but prevented activation of PKC. These observations suggest that bradykinin-induced PKC activation is a prerequisite for the cardioprotective effect of RIPC, that activation of PKC is a decisive step in conferring cardioprotection, and that both the bradykinin-dependent humoral pathway as well as the neuronal pathway are essential for PKC activation.

Shoemaker et al. [29] were the first to demonstrate that mesenteric preconditioning induced increased endogenous bradykinin levels, which had a remote preconditioning effect on the heart and reduced infarct size following coronary infarction. Blockade of bradykinin receptors was associated with increased myocardial infarct size. However, bradykinin blockade in non-preconditioned animals did not influence infarct size and in the absence of preconditioning there was no change in basal bradykinin levels [117].

A combined sensory neurogenic and humoral pathway is strongly suggested in the bradykinin-mediated RIPC [29]. Following mesenteric ischemia reperfusion, there is local release of bradykinins, which stimulates the sensory afferent nerves projecting on efferent nerves to the heart, which in turn precondition the heart. Bradykinin receptors B2 are involved in sensory nerve stimulation and bradykinin receptor antagonists HOE-140 (Hoechst-140) abolish the protective effect. This study also showed that bradykinin receptor blockade led to loss of protection in both direct and remote IPC but ganglion blockade abolished protection only in remote preconditioning. Thus direct preconditioning is associated with blood-borne kininogens in contrast to the complementary effect of humoral and neurogenic pathways in RIPC. Bradykinins activate intracellular transduction of PKC as demonstrated by Wolfrum et al. [117]. However, the pathway downstream of kinases remains unclear. Although modulation of K_ATP channels is suggested, this needs to be clarified in future studies.

Toombs et al. showed IPC by 5 min of CAO, 10 min reperfusion in rabbits prior to 30 min CAO, and 120 min reperfusion reduced infarct size [118]. In reserpinized rabbits subsequent IPC failed to reduce infarct size, suggesting the role of noradrenaline release in IPC. In RIPC two studies provided evidence to support the role of catecholamines. Oxman et al. [38] showed an increase in cardiac norepinephrine release and increased baseline plasma levels after limb preconditioning due to a systemic stress response that was partially abolished by reserpine due to depletion of catecholamine stores. In reserpinized animals the antiarrhythmic effect of RIPC was blocked. Kharbanda et al. showed that sympathetic blockade abolished RIPC resulting in increased myocardial infarct size [20]. de Zeeuw et al. demonstrated that increase in cardiac interstitial norepinephrine levels following ischemia was cardioprotective but RIPC by cerebral ischemia was ineffective due to inadequate increase in norepinephrine levels [119]. The role of catecholamines in RPCT is only speculative and future studies are needed to clarify this.

In direct IPC endogenous catecholamines are known to act upon cardiac adenoreceptors, stimulating myocardial protein kinase eventually leading to preconditioning of the heart [120]. In RIPC the likely pathway seems to be the sympathetic nerve pathway as reserpine abolished protection. It is also possible that catecholamines released into the circulation may act on sympathetic nerve endings in target organs; however, this needs to be further clarified by measurement of plasma catecholamine levels and use of adrenergic receptor blockers. The postreceptor mechanisms are unclear and the role of catecholamines in activation of kinases as well as modulation of K_ATP channels as end effectors needs to be clearly demonstrated in future experiments.

IPC of the intestine (three cycles of 8 min ischemia, 10 min reperfusion) prior to intestinal IRI (30 min MAO, 2 h reperfusion) reduced intestinal injury, edema, LDH, and malionaldehyde levels (markers of oxidative stress) [121]. Pretreatment with morphine mimicked the effects of IPC and naloxone abolished the protective effects of both IPC and morphine [121]. The intestine and colon are rich in opioid receptors and contain opioid peptides [121]. This study demonstrated an increase in endogenous opioid peptides in the effluent collected after IPC and suggested that opioids are released in response to oxidative stress to confer a protective effect against stress.

Opioid release from the stress of brief IR could result in a remote protective mechanism. Five studies have demonstrated the role of opioids in RIPC in intestinal, skeletal muscle, and heart tissue. (1) On the basis of the theory that opioids are released as a natural response to stress, are released irrespective of whichever organ is stressed, and act widely and ubiquitously on remote organs, Patel et al. [33] demonstrated in a rat model that RIPC (MAO, 15 min) prior to CAO reduced myocardial infarct size, which was abolished by naloxone. (2) Hindlimb RIPC (3 × 10 min IR cycles) prior to 4 h of muscle flap ischemia and 48 h reperfusion reduced latissmus dorsi, rectus abdominis, and gracilis flap infarct size. This study demonstrated the remote protection of all skeletal tissue by RIPC. RIPC (3 × 5 min IR cycles) prior to 30 min CAO and 120 min reperfusion in a rat model reduced myocardial infarct size and plasma LDH levels [122] and naloxone abolished these effects. Brief infrarenal aortic occlusion (IOA) 15 min prior to prolonged infrarenal aortic occlusion (30 min) protected myocardium [78]. Dickson et al. demonstrated a role for opioids in transferred preconditioning. Met and leu-enkephalins were liberated from the preconditioned donor rabbit heart into the coronary effluent, which subsequently elicited protection when given to virgin acceptor hearts. Addition of naloxone to the coronary effluent abolished opioid-induced protection. Based on evidence that opioids induce ischemic tolerance in the intestine in direct IPC and the presence of abundant opioid receptors in the intestine, Dickson et al. treated ischemic gut with coronary effluent from preconditioned hearts and showed that the recovery of maximal contractile force of gut after ischemia was enhanced by opioids [66, 67, 68, 69].

Opioid receptors are known to be present in neuromuscular regions, especially delta receptors [123], and previous studies have described the humoral action of opioid receptors in skeletal muscle by demonstrating [124] that B-endorphins released in the circulation stimulate glucose uptake in muscle. In RIPC (hindlimb) autonomic ganglion blockade by hexamethonium did not abolish the effects of opioids [49], suggesting the blood circulation to be the likely pathway of transmission of protection in opioid-induced RIPC. The need for reperfusion following IOA for cardioprotection and lack of protection in occlusion without reperfusion [78] and the evidence for transferred preconditioning in virgin acceptor hearts by opioids in coronary effluent from preconditioned hearts are observations that support a humoral pathway.

Opioids act on receptors in the target organ. This has been shown in heart, skeletal muscle, and intestinal tissue [33, 49, 122]. Skeletal muscle flap protection by RIPC was abolished by selective δ1 opioid receptor antagonists and myocardial protection was abolished by selective κ1 receptor antagonists. δ1 receptors have been shown in all species and human cardiomyocytes [78]. Activated δ1 opioid receptors induce effects which mimic RIPC. In transferred preconditioning leu-enkephalins activate δ1 opioid receptors to induce ischemic tolerance in myocardium or gut [66].

Opioid-induced RIPC prior to skeletal muscle IRI reduced ATP depletion, lactate accumulation, neutrophil infiltration, and myeloperoxidase activity in preconditioned skeletal muscle [49], which was abolished by opioid receptor antagonists. These observations suggest that the energy-sparing effect coupled with attenuation of lactate accumulation during early reperfusion is triggered by activation of opioid receptors. This is supported by IPC studies ([125].

One study has shown the role of K_ATP channels in the end effector energy sparing effect [66, 122]. Coadministration of Glibenclamide to coronary effluent from preconditioned hearts abolished transferred preconditioning in ischemic gut [66]. RIPC by κ opioid receptors in the heart modulates mitochondrial pore mobility in the post receptor mechanism [122]. IPC studies have shown opioid receptors to act via inhibitory G-protein, activation of multiple kinases with modulation of mitochondrial and sarcolemmal K_ATP channels serving as final end effectors [122]. Future studies are needed to clarify postreceptor mechanisms in RIPC and resolve receptor subtypes.

IPC studies have shown that free radicals can directly activate kinases leading to transcription of protective proteins [92, 126], and free radical scavenger MPG abrogates the protective effects of direct preconditioning. Weinbrenner et al. were the first to show that free radicals are key candidate molecules in RIPC [41]. A single cycle of 15 min of infrarenal aortic occlusion followed by 10 min reperfusion reduced myocardial infarct size significantly. MPG blocked the effects of both RIPC and a single cycle of direct preconditioning but failed to block multiple cycles of direct preconditioning. This suggested that preconditioning is a graded phenomenon with multiple cycles producing a more robust preconditioning stimulus with free radicals being only partially involved in the mechanism of protection. This study did not demonstrate the source of free radicals. Recently, Patwell et al. demonstrated the appearance of hydroxyl free radicals in the circulation following IR of the limb [127], and the argument that free radicals may be produced from the ischemic limb in RIPC was supported by Chen et al. [128], who showed that myocardial infarction was reduced by four cycles of 10 min femoral artery occlusion-reperfusion associated with an elevation in whole blood free radical counts up to 2 h following RIPC. Since ROS levels were significantly low compared to the IRI group, this data suggest the role of low-dose ROS in inducing preconditioning. In addition, a period of reperfusion for generation and action of free radicals is needed. These findings collectively suggest that free radicals reach the remote organ via the bloodstream to induce a preconditioning effect. Conversely, in an animal model, occlusion of the coronary artery before preconditioning supported a nerve pathway of transmission of the preconditioning effect [128].

Direct IPC studies have shown that ROS led to the release of triggers such as NO, catecholamines, adenosine, and bradykinin. ROS activates intracellular kinases directly [126] and induces synthesis of protective proteins. Also ROS activates cytokine NF-κB, which induces iNOS mRNA transcription 24 h later to confer delayed protection in the target organ. None of these pathways have been clearly demonstrated in RIPC. Chen et al. [76] showed that ROS induced elevation of heat shock protein and mitochondrial antioxidant enzymatic activity, which helped maintain mitochondrial function and reduce apoptosis. However, pretreatment with MPG abrogated HSP and antioxidant activity. Further clarification of the end effector mechanism of ROS in RIPC, the source of ROS, pathway, and the role of low-dose ROS in maintenance of mitochondrial membrane permeability as well as prevention of apoptosis is needed. Also quantification of the levels of ROS, which induce protective mechanisms need to be determined since high levels of ROS are the key elements in the initial cascade of IR injury.

Moses et al. [47] showed that RIPC (three cycles hindlimb occlusion) reduced latissmus dorsi infarct size, which was abolished by nonselective blockers (glibenclamide) and selective mitochondrial channel blockers (5-HD). Kristiansen et al. showed RIPC of the donor heart by hindlimb IR (4 × 5 min cycles) was abolished by blockade of nonselective and mitochondrial K_ATP channels and increased infarct size in donor heart on implantation. Administration of diazoxide (selective mitochondrial k channel activator) prior to explant conferred protective effects similar to RIPC, suggesting that the effects of RIPC are memorized in the explanted hearts and are critically dependent on modulation of mitochondrial K_ATP channels [45]. Konstantinov et al. showed that blockade of K_ATP channels following RIPC in the recipient increased myocardial infarct size in the implanted heart [46]. Dickson et al. showed blockade of K_ATP channels abolished transferred preconditioning in virgin hearts conferred by coronary effluent from preconditioned hearts [66]. Mabanta et al. showed that remote microvascular preconditioning is mediated by K_ATP channels and blockade of K_ATP channels abolished preconditioning [129]. Pell et al. showed renal IR (10 min cycles) reduced myocardial infarct size and cardioprotection was abolished by K_ATP blockers (5-HD) [27]. Moses et al. showed remote skeletal protection by limb RIPC (3 × 10 min ischemia) was abolished by 5-HD (selective mitochondrial channel blocker) [47].

Two studies have demonstrated the modulation of IRI in denervated hearts in a transplant model; this suggests that the humoral pathway is involved in conferring protection. The role of opioids in transferred preconditioning supports the humoral pathway.

Moses showed blockade of K_ATP channels prior to RIPC abolished protection, suggesting their role in the trigger mechanism of RIPC [47]. Blockade of K_ATP channels after RIPC abolished protection; however, blockade 10 min prior to ischemia did not influence infarct size, indicating a role of K_ATP channels in the mediation of signal transduction. This study demonstrated that the critical time period for K_ATP channels to remain open after triggering was 10 min further, supporting a role in mediation of signal transduction. One study has shown the opening of mitochondrial K_ATP channels and ionic fluxes across mitochondrial K_ATP channels in the opioid pathway of RIPC [122] and one study showed ionic fluxes across K_ATP channels in RIPC by the adenosine pathway [27]. These observations suggest that K_ATP channels serve as end effectors in the pathway of RIPC and their role needs to be further clarified as end effectors for other candidate molecules.

All studies discussed above clearly demonstrate the role of mitochondrial K_ATP channels in RIPC as blockade with 5-HD (selective mitochondrial channel blocker) abolished RIPC. Pell et al. [27] showed that RIPC was abolished by 5-HD without affecting the action potential of sarcolemmal fibers or vasodilator effects of sarcolemmal K_ATP channels. The muscle used in the Moses model was surgically denervated, not associated with changes in action potential and muscle contractility, suggesting the unlikely role of sarcolemmal K_ATP channels in the energy-sparing effect. Two studies showed K flux across mitochondrial K_ATP channels after RIPC [27, 122].

Pell et al. [27] showed that K_ATP channels reduce ATP depletion and maintain intracellular pH and phosphocreatinine levels in heart muscle after RIPC. Moses et al. showed decreased neutrophilic infiltration, myeloperoxidase activity, and ATP-sparing effect in preconditioned skeletal muscle [47]. These data suggest that K_ATP channel opening reduces the rate of ATP hydrolysis or mitochondrial ATPase activity, thereby decreasing the rate of ATP depletion. Also, opening of mitochondrial K_ATP channels decreases mitochondrial calcium load, which preserves mitochondrial integrity. Mitochondrial volume is regulated by K_ATP channels and volume changes modify energy flow through the electron system, thereby influencing energy transfer between mitochondria and cellular ATPases. These studies showed that RIPC applied in vivo exerts protection upon skeletal muscle and heart after explantation from the body and has potential beneficial effects in relation to heart transplantation, cardiopulmonary bypass, and autologous skeletal muscle transplantation.

CGRP is a neuropeptide and principal neurotransmitter found in capsaicin-sensitive sensory nerves. CGRP receptor antagonists abolished reduction in infarct size by IPC [130] in a rat model, suggesting the role of CGRP in mediation of IPC of the heart.

Five studies demonstrated the role of CGRP in RIPC by brief MAO. Tang, Xiao, Wolfrum et al. showed that brief MAO increased plasma levels of CGRP following RIPC and reduced myocardial infarct size and blockade of CGRP receptors, and ganglion blockade abolished RIPC [31, 34, 35]. Ganglion blockade did not affect CGRP release. Brzozowski et al. showed decreased mucosal CGRP and blood flow in gastric IR and RIPC by hepatic/coronary ischemia (2 cycles × 5 min IR) restored mucosal CGRP levels in gastric IR and gastric flow and prior treatment with CGRP receptor antagonist and capsaicin-induced sensory nerve deactivation abolished RIPC by MAO [61]. Administration of exogenous CGRP in denervated animals restored RIPC effects.

An increase in plasma levels of CGRP as shown in three studies [31, 34, 35] supports a humoral pathway. However, ganglion blockade abolished myocardial protection from brief MAO, demonstrating a neuronal pathway for CGRP in RIPC by MAO.

Capsaicin selectively depletes neurotransmitters in sensory nerves and is known to cause deactivation of sensory afferent neurons. Two studies in RIPC by MAO have shown that treatment with capsaicin prior to RIPC by MAO abolished both reduction in myocardial infarct size and increase in CGRP levels [34, 35]. Two studies in RIPC by brief hepatic artery occlusion and CAO [60, 61] showed that prior treatment with capsaicin abolished the increase in gastric mucosal CGRP, increase in gastric blood flow, and reduction in gastric erosions following RIPC. These observations strongly support the role of sensory neurons in RIPC. In addition Brzozowski et al. showed that vagotomy abolished RIPC from brief hepatic ischemia or CAO and reduced plasma levels of CGRP. This observation suggests that vagal efferents form a part of the effector pathway in CGRP-mediated RIPC and demonstrates the role of the brain gut axis in RIPC. Finally, l-name (NO blocker) abolished reduction of myocardial infarct size by CGRP-mediated RIPC, suggesting that CGRP acts via the NO pathway [34].

One study showed that in RIPC by MAO, CGRP activated myocardial PKC and reduced infarct size [31]. PKC activation and myocardial protection was abolished by both ganglion blockade and CGRP receptor antagonists. These data suggest that in RIPC, CGRP acts at the cellular level by activation of kinases. Another study showed the NO pathway as discussed above and NO may be the end effector responsible for CGRP effects. This argument is supported by past studies providing evidence for CGRP-related NO release [34]. One study showed the inhibition of cytokines (TNF-α and IL-1β) release in the target organ [60]. Further clarification of the pathway downstream is needed in future studies.

Brzozowski et al. showed that indomethacin (nonselective cyclooxygenase blockers), SC-560 (selective cox-1), and rofecoxib (selective cox-2) blockers abolished the gastric protective effects of IPC by two cycles of 5 min celiac artery occlusion, RIPC by 2 cycles of 5 min hepatic ischemia, or CAO [61]. Concurrent treatment with exogenous PGE₂ counteracted the effects of Cox blockers and restored the hyperemic mucosal effects of RIPC. Brzozowksi et al. also showed increased gastric mucosal generation of PGE₂ in IPC and RIPC. These observations suggest that increased endogenous activity of prostaglandins particularly of the COX-2 variety could be involved in preconditioning and it is known from past studies that they have a protective effect in gastric ulcers.

Brozozowski et al. showed that functional ablation of sensory nerves by pretreatment with capsaicin abolished the protective effects in all preconditioned groups. This observation suggests the involvement of sensory afferent neurons in the pathway. Vagotomy abolished RIPC from brief hepatic/CAO, suggesting the role of vagal efferents and the brain gut axis in mediation of RIPC. RIPC was associated with increased CGRP and decreased TNF-α and IL-1β expression and release, suggesting amelioration of gastric IRI by prostaglandins through modulation of cytokine release and increased mucosal flow by vasodilatory effects of CGRP through NO release. These effects were attenuated by pretreatment with capsaicin and ablation of sensory nerves known to release NO and CGRP.

HSP28, 40, 60, 70, 90, 104 have an important physiological and pathological role. One of the most important HSPs is HSP70, which has a cytoprotective role. The constitutive form of HSP is HSP73, which all cells express to some extent. The inducible form is HSP72.

HSP expression is more in myocytes compared to microvascular cells. Enhanced HSP expression may serve to protect the heart against ischemia reperfusion, hypoxia, and chemicals in addition to hyperthermia. HSP express themselves as early as 3 h post reperfusion in the myocardium, begin to decay from 42 h post reperfusion, and persist up to 72 h as detected by immunohistochemistry.

In three studies in RIPC, Tanaka et al. showed in a rabbit model RIPC by four cycles of 5 min CAO and 5 min reperfusion [21, 131] increased cardiac tolerance due to increased HSP expression in the ischemic and remote myocardium; Konstantinov showed RIPC by six cycles of 4 min femoral artery occlusion and 4 min reperfusion increased expression of HSP73 genes (anti-inflammatory) in the heart of preconditioned mice as compared to sham [40] and Chen et al. in a rat model showed increased myocardial HSP expression (HSP70) following hindlimb RIPC [76].

HSP is induced by ischemia [131] and ROS [76]. Tanaka showed myocytes with immunoreactivity for HSP as early as 3 h and persistence of this immunoreactivity up to 72 h. This suggests that HSP expression may be associated with a delayed phase of protection. Mechanisms by which ischemia may induce HSP expression include enhanced tissue levels of catecholamines and angiotensins, which induce HSP expression, myocardial dyskinesia, which induces compensatory hypercontraction of non-preconditioned myocardium, increased workload and a modest increase in HSP in the nonischemic myocardium, and elevated end-diastolic pressure, which leads to increased workload and HSP expression. However these need to be clearly demonstrated in experimental settings.

One study [40] showed that reperfusion stimulates HSP gene expression in myocardium following limb RIPC in a rat model. Chen et al. demonstrated that reperfusion induced myocardial HSP expression following limb RIPC and this was abrogated by free radical scavengers (MPG) [76]. These studies suggest that circulatory mediators stimulate HSP expression in remote tissue and free radicals are one of these mediators. The effector mechanism may involve scavenging of ROS and modulation of K-channels but this needs to be clearly demonstrated.

HO-1 is a form of inducible HSP. It is an enzyme that mediates the rate-limiting step of breakdown of haem to biliverdin, CO, and iron. In this process it scavenges ROS and mitigates IRI. Previous studies have shown hemoxygenase to reduce hepatic IR and lung IR [132, 133, 134, 135].

One study showed RIPC by limb ischemia [136] (four cycles of 10 min ischemia) reduced hepatic IRI and improved liver function. Hemoxygenase expression at 3 h of reperfusion suggested its role in amelioration of IR. Zinc protoporphyrin (HO-1 inhibitor) 1 h prior to IR abolished RIPC effects.

Thus far no studies have demonstrated HO-1 activity in serum following RIPC. It is likely that following RIPC peripheral breakdown of haem may lead to increased CO levels in the blood and CO has both vasodilatory and antiplugging effects in hepatic sinusoids. However no study has measured post RIPC serum carboxyhemoglobin levels to demonstrate this. Lai et al. [136] demonstrated no increase in HO-1 in peripheral macrophages following RIPC and therefore the role of peripheral macrophages as a pathway for HO-1 induction in the liver is unlikely. RIPC induces a low-grade oxidative stress in the limb that may lead to release of ROS and cytokines into the circulation, low-grade oxidative stress in the liver, and increased hepatic HO-1 expression. This is hypothetical and needs to be proved by using ROS scavengers. HO-1 pathways scavenge ROS following IR, stabilize mitochondrial membrane permeability, and reduce apoptosis by decreased cytochrome C release as shown by previous studies. HO-1 also down-regulates iNOS and inhibits NF-κB expression. Future studies are needed to define the pathway and effector mechanism of HO-1.

The first study to look at the effect of remote preconditioning on inflammatory gene expression modification in a human model showed that brief forearm ischemia led to suppression of proinflammatory genes [137] encoding proteins in circulating leukocytes, reduction in neutrophil chemotaxis, adhesion, and exocytosis. Studies by Huda et al. (RIPC MAO model) [140] showed expression of genes associated with organ protection following RIPC.

Kharbanda et al. conducted the first clinical trial in humans [20] and showed that forearm ischemic preconditioning is associated with diminished ischemia reperfusion injury in the contralateral arm as well as diminished endothelial injury. The mechanism was not clearly identified, although it was thought to be due to a substance released in the circulation. In a recent trial in children undergoing cardiac surgery for congenital heart defects on cardiopulmonary bypass, Cheung et al. demonstrated that four cycles of right hindlimb ischemia (5 min) followed by reperfusion (5 min) prior to cardiopulmonary bypass preconditioned the heart and reduced infarct size [73]. Thereafter Kristiansen et al. studied the effect of forearm ischemia on cardiac ischemia in cardiopulmonary bypass procedures and found it to be protective. This protective effect was attributed to KATP channels [45].

In humans, Konstaninov et al. [137] showed diminished expression of inflammatory genes in neutrophils following limb RIPC. Since neutrophils are one of the key mediators of the late phase of IRI responsible for oxidative stress and injury to organs, modulation of neutrophil activation would be of prime importance in reducing IRI.