Purified herba leonuri and leonurine protect middle cerebral artery occluded rats from brain injury through antioxidative mechanism and mitochondrial protection 1

For ischemic stroke Figure 2-1b, which accounts for 80% of all stroke cases, the brain damage is caused by a reduction or complete blockage of blood flow, resulting in the deficiency of

2.1 Pathophysiology of stroke

Acute stroke can be divided into 2 categories: hemorrhagic stroke and ischemic stroke

Hemorrhagic stroke (Figure 2-1a) describes sudden rupture of the blood vessels within

the brain that causes the leakage of the blood into brain cavity, and therefore brain

damage results (Zemke et al, 2004) For ischemic stroke (Figure 2-1b), which accounts

for 80% of all stroke cases, the brain damage is caused by a reduction or complete

blockage of blood flow, resulting in the deficiency of glucose and oxygen supply to the

territory of the affected region (Zemke et al, 2004) Neurons are the most vulnerable cells

to hypoxia due to their strong dependence on the oxidative metabolism of glucose for

energy Since ischemia is the major problem among all the stroke patients, many

researches have been targeting on the treatment of ischemia This project focused also

ischemic stroke as a model to test on the therapeutic potential of studied drugs (see

below)

Figure 2-1: a) Ischemic stroke; and b) hemorrhagic stroke (arrow)

a) b)

2.1.1 Ischemic stroke

As mentioned, ischemic stroke results from a transient or permanent reduction in cerebral blood flow (CBF) by an embolus or a thrombus, leading to brain injury A critical reduction of CBF causes failure of cellular transport mechanisms and massive release of potentially toxic neurotransmitters, subsequently the formation of free radicals, inflammation, induction of immediate early genes and later, cell death by necrosis or apoptosis (Barber, 2008) Therefore, the extent of brain injury is dependent on level of CBF reduction, duration of ischemic insult, tissue temperature, blood glucose concentration and many other physiological variables

The two principle models of human stroke are global ischemia and focal ischemia, either permanent or transient Global ischemia occurs when CBF is reduced throughout most or all parts of the brain as a result of cardiac arrest or other causes of collapse of system circulation, and subsequently failure of brain perfusion The tissue injury is dominated by neurons, occurring especially in the most vulnerable region (CA1 region of hippocampus)

of the brain first and then proceeding to the less vulnerable region such as thalamus or caudate putamen (Miller, 1999; Canese et al, 1997) Neuronal death in global ischemia is always detected in the hippocampus, striatum, neocortex with most susceptible population lies within CA1 and CA4 area of hippocampus and layer 2 and 5 of the cerebral cortex (Taoufik and Probert, 2008)

As for focal ischemia, it is represented by a reduction of blood flow to a specific brain region, such as the occlusion of middle cerebral artery Although occlusion of vessel occurs during focal ischemia, there is rarely complete blockade of CBF to the area supplied by the occluded vessel because plethora of collateral vessels provides some flow

to the area (Horst and Korf, 1997) During focal ischemia, there will be a core infarct area which does not receive sufficient perfusion to sustain any of the neurons, glia, or even the vasculature The volume of core infarct is correlated to the severity of neurological deficit

On multi-tracer O Positron emission tomography (PET), ischemic core exhibits very low CBF, cerebral blood volume (CBV), and metabolic rates of oxygen and glucose (Marchal

et al, 1999) In this region, a CBF of <10ml/100g of brain tissue per minute severely impair the cellular function by depleting the energy metabolites and causing the failure of the cell membrane to maintain ion homeostasis This manifests as massive efflux of potassium and reciprocal influx of sodium, calcium and water In addition, the anaerobic respiration is initiated due to the impairment of mitochondrial oxidative phosphorylation Anaerobic respiration leads to the production lactic acid which could cause acidosis toxicity to the cell The cells in this area are destined to have irreversible damage and undergo necrosis, which means out of therapeutic rescue (Marchal et al, 1999)

In addition, within the necrotic core, the vasculature may also be severely damaged, exposing the risk of undergoing haemorrhagic transformation This is especially important for the case of extensive infarction and with the use of thrombolytics for the stroke treatment, as it might worsen the clinical condition Therefore, reduction the risk

of haemorrhagic transformation is one of the therapeutic goals as for instance, new

thrombolytic agents that do not interfere with endothelial function or induce matrix metalloproteinase dysregulation (Moustafa and Barron, 2008)

Surrounding this core will be hypoperfused tissue, namely ischemic penumbra that

receives collateral flow that sufficient to prevent cells from undergoing necrosis At reduced CBF of about 20ml/100g of brain tissue per minute, the cerebral metabolic rate

of oxygen (CMRO2) starts to fall, thereby impairing the normal blood flow autoregulatory mechanisms, and the neuronal electroencephalographic activity ceases Therefore, this threshold represents the threshold for loss of neuronal electrical function (Wise et al, 1983) Cells in this region are functionally silent but remain metabolically active and maintain a very low level of adenosine triphosphate (ATP) Tissue within this area is potentially salvageable Reports (Touzani et al, 1997; Heiss et al, 1998) showed that large volumes of tissue with penumbral level of CBF escape necrosis if arterial recanalization is achieved in time However, with prolonged ischemic insult, the cells in penumbra is continuously bombarded by waste products from the dead cells in ischemic core, the cell recovery will therefore decrease over time and cells in this area undergo

delayed (hours or days) cell death (Zemke et al, 2004)

Furthermore, as if ischemia persists, large slow voltage shifts occur at the borders of the core infarct and propagate as spreading depolarization waves that compromise the survival of surrounding tissue (Selman et al, 2004), together with spreading of the inflammation and excitotoxicity in the ischemic core, tissue in ischemic penumbra will gradually transform into the core The course of events varies from patients to patients,

most exhibit substantial volumes of penumbra for many hours, with exceptionally days after stroke onset (Moustafa and Baron, 2008) Using multi-tracer PET, substantial volumes of cortical penumbra have been reported to decline over time, that being present

in over 50% of the patients studied within 9 hours after stroke onset, and in about third of the patients studied between 5 to 18 hours (Mousfara and Baron, 2008) Thus, it emphasizes the urgency of acute stroke management The reduction or prevention of the cell death in the ischemic penumbra so to prevent the growth of the ischemic core lesion within the temporal therapeutic window is the main target of pharmacological intervention studies

one-2.1.2 Cell death in stroke

2.1.2.1 Ischemic cascade

At cellular level, ischemic cascade is resulted from a severe prolonged ischemic insult (Figure 2-1) It begins with progressive derangements in energy and substrate metabolism (Horst and Korf, 1997) Energy deficiency leads to interruption of ATP dependent process, such as sodium/potassium ATPase (Na+/K+ ATPase) which can subsequently causes the disruption of ion homeostasis as most of the ATP generated from mitochondrial oxidative phosphorylation is used for the stabilization of transmembrane ion concentration gradients of sodium, potassium and calcium which are important for neuronal impulse conduction and synaptic function Therefore, onset of ischemia results

in the movement of ions down their electrochemical gradients such that intracellular

calcium and sodium concentration and extracellular potassium concentration increase dramatically within one or two minutes after ischemia (Horst and Korf, 1997)

Increase in extracellular potassium triggers depolarization and reversal of direction of action of the amino acid (such as glutamate) transporters Under these conditions, Ca2+enters the cells via voltage-dependent channel and a massive release of excitatory amino acid such as glutamate out of the cells, resulting in excitotoxicity This initiates a positive feedback loop where excessive glutamate activates AMPA, kainate and N-methyl-D-aspartate (NMDA) receptor to consume more ATP and promote further release of glutamate Ionotropic NMDA receptor potentiates the efflux of K+ and influx of Na+

together with water, leading to the cell edema Together with the Ca2+ entry via voltage dependent channel, ionotropic NMDA receptor also promotes excessive Ca2+ influx, leading to intracellular Ca2+ overload A range of downstream nuclear and cytoplasmic lethal metabolic derangement will be resulted by Ca2+ overload These include the activation of phospholipases and proteases that could degrade membrane and proteins that are essential for cellular integrity Ca2+ overloaded mitochondria will be severely impaired and hence the inhibition of ATP production (Nakka et al, 2008) Augmented intracellular Ca2+ further promotes the release of glutamate and thus propagates the excitotoxicity Increase in Ca2+ level causes also the increase in free radicals production which will be discussed later

As mentioned, oxygen deficiency during ischemia results in anaerobic respiration due to the inability of mitochondria to perform oxidative phosphorylation However, energy

obtained from anaerobic respiration is not enough to compensate the energy needed for neurons since brain has limited amount of glycogen stores In addition, anaerobic respiration leads to accumulation of lactic acid which causes a local rise of lactate production and a fall in pH, leading to intra- and extra-cellular acidosis, reflecting a marked imbalance between energy use and production (Barber, 2008) Low oxygen level will also cause free radical generation by incomplete oxidative phosphorylation Free radicals are known to react with and damage whole range of organelles and plasma

membrane (Zemke et al, 2004) In conclusion, mechanisms that contribute to the

neuronal cell death predominantly occur via 3 major mediators: unregulated intracellular increase of Ca2, tissue acidosis, nitric oxide (NO·) and free-radical production (Barber, 2008)

Figure 2-2: A diagram illustrates the ischemic cascade (Adapted from Crack and Taylor, 2005)

2.1.2.2 Apoptosis

In the past, neuronal cell death after cerebral ischemia was considered to be exclusively necrotic Research over the past decade has revealed that a portion of cells in ischemic penumbra or periinfarct zone undergo programmed cell death (PCD), namely apoptosis, via caspase dependent or caspase independent pathways, after stroke Thus they are potentially recoverable after the onset of stroke via pharmacological intervention of PCD (Brad et al, 2009)

Necrosis is commonly resulted from the accumulation of deleterious changes that disrupt vital cell viability Necrosis is irreversible massive cell death characterized by shrunken cells with darkened nuclei, swelling of cytoplasms and organelles and loss of membrane integrity, resulting in cell lysis and release of the cellular content that in turn lead to local inflammation to surrounding tissue (Taoufik and Probert, 2008) In contrast to necrosis, apoptosis is orderly process of energy dependent programmed cell death characterized by morphological features as cell shrinkage, membrane blebbing, chromatin condensation, and DNA fragmentation (Nakka et al, 2008) Cell undergoing apoptosis will be recognized and removed in an organized way (phagocytosis) to avoid inflammation and minimize the damage and disruption of neighboring cells (Taylor et al, 2008) A more unique morphological characteristic of neuron undergoing apoptotsis is the neurite fragment (dendrites and axons) that occurs early during the cell death process (Taoufik and Probert, 2008) Mixed morphologies of apoptosis and necrosis observed during

ischemic insult could be result from the initiation of apoptosis that are then overtaken by the molecular event associated with necrosis (Roy and Sapolosky, 1999)

Evidences of involvement of apoptosis in stroke comes from a small number of studies showing that neuronal apoptosis is involved in human stroke (Guglielmo et al, 1998; Love S et al, 1998), as well as a large body of support from animal studies where apoptotic markers are co-localized at the ischemic affected regions (Taoufik and Probert, 2008) Apoptosis was showed to contribute to ischemic damage by TUNEL staining, which could detect the DNA fragmentation of cell death Through TUNEL staining, apoptosis was found scattered throughout the ischemic territory with more apparent at the perifocal tissue (Sims and Anderson, 2002)

Generally, apoptosis can be executed via two pathways: Extrinsic pathway and intrinsic pathway (Figure 2-3) Extrinsic pathway initiates apoptosis through the engagement of plasma membrane death receptors, therefore also referred as “death receptor pathway” (Ashe and Berry, 2003) Death receptors belong to the tumor necrosis factor receptor (TNFR) family They transmit the apoptotic signal through binding of death ligand Fas is one of the best characterized family members Its preferred ligand is (Fas ligand) FasL (Ashe and Berry, 2003) There were reports on Fas/FasL system that it is also involved in neuronal apoptosis following traumatic brain injury and cerebral ischemia (Beer et al, 2000; Martin-Villalba et al, 1999; Rosenbaum et al, 2000) Trimerization of Fas followed

by ligation of FasL promotes the recruitment of the cytosolic adaptor protein associated death domain protein (FADD) through complementary death domain (DD)

Fas-FADD contains also death effector domain (DED) which is responsible to bind with complementary DED in procaspase 8 and 10 This complex (FasL, Fas, FADD, procaspase 8 or 10) is referred as death-inducing signaling complex (DISC) DISC close positions the DED containing initiator caspase (procaspase 8) and therefore cause the activation of initiator caspase by their autolytic cleavage (Figure 2-3) Activation of these initiator caspases results in the execution of the apoptotic programme by cleavage of downstream targets (Ashe and Berry, 2003)

Generally, there are two types of Fas-mediated apoptosis Type 1 requires the activation

of caspase 8 that is closely followed by the activation of caspase 3 Apoptosis in Type 1 cells cannot be rescued by inhibitor of Bcl-2 family which plays a central role in intrinsic pathway of apoptosis Type II has limited activation of caspase 8 Caspase 8 in type II cells involves cleavage of the BH3-only protein, Bcl-2 interacting domain (BID), to release truncated BID (tBID) BID is a proapoptotic cytosolic member of Bcl-2 family that translocates to mitochondria when cell receives death signal (Sugawara et al, 2004) which is crucial for the release of cytochrome c and Smac/DIABLO from mitochondria Activation of Bid by caspase 8 results in an amplification loop by means of extrinsic apoptotic pathway recruits an intrinsic apoptotic pathway (Ashe and Berry, 2003) Regardless the type, caspase 8 is the apical caspase in DR signaling and its activity is detected after permanent middle cerebral artery occlusion (MCAO) (Taoufik and Probert, 2008)

The onset of stroke causes the cytotoxic intracellular accumulation of Ca2+ which triggers the activation of intrinsic pathway of apoptosis (Dirnagl et al, 1999) Activation of calpain by increased Ca2+ or stimulation of caspase-8 via extrinsic pathway results in the activation of BID to its truncated active form tBID (Brad et al, 2009) Recent studies have shown the involvement of BID in cerebral ischemia that Plesnila and colleagues (2002) found that deletion of BID gene in mice reduced ischemic infarct size tBID causes the conformational changes of other proapoptotic proteins situated on mitochondria, such as BAX and Bcl-xS so to execute the apoptotic signaling (Figure 2-3) These proapoptotic proteins can also heteromerize with antiapoptotic members of bcl-2 family situated on outer mitochondrial membrane, such as Bcl-xL, Bcl-2, so to counteract their antiapoptotic function (Saito et al, 2003) In addition, studies showed that BAX can form channel across the mitochondrial membrane that are large enough to allow the passage of cytochrome c (Kirkland et al, 2002)

After the disruption of mitochondria or the opening of mitochondrial permeability transition pore (MPTP), mitochondrial proapoptotic protein such as cytochrome c, Smac/DIABLO, serine protease HtrA2/Omi will be released into the cytoplasm Once released, these proteins will be involved in caspase-dependent apoptotic pathway Cytochrome c, a water soluble mitochondrial protein that is an essential component of mitochondrial respiratory chain, forms apoptosome by binding to Apaf-1, ATP and pro-caspase 9 Caspase 9 will then be activated and subsequently activate caspase 3 as a executor of apoptosis (Brad et al, 2009) Caspase 3 has been documented to be involved

in cerebral ischemia (Asahi et al, 1997) and it cleaves many substrates such as poly

(ADP-ribose) polymerase (PARP) For Smac/DIABLO, it lifts the inhibition of caspase 9 and caspase 3 via neutralizing the caspase-inhibitory properties of the IAP (inhibitor of apoptosis) family of proteins, particularly XIAP thereby allowing apoptosis to occur (Christophe and Nicholas, 2006)

Both intrinsic and extrinsic pathways of apoptosis lead to activation of caspase 3 Caspase 3 is the executioner caspase in the cascade Caspase 3 activation was observed in neurons 24 hours after MCAO Administration of caspase 3 inhibitor reduces the infarct size after focal ischemia Caspase 3 inhibition also protected mice from transient MCAO (Taoufik and Probert, 2008)

Increasing evidences showed the significance of caspase independent apoptotic pathway

is involved in ischemic stroke (Elmore, 2007) A group of proteins will be released out from the mitochondria during apoptosis such as apoptosis inducing factor (AIF), endonuclease G and Bcl-2/adenovirus E1B 19kDa-interacting protein (BNIP3) Studies have demonstrated that the involvement of AIF and endonuclease G in cerebral ischemia that both AIF and endonuclease G translocate from mitochondria to nucleus after cerebral ischemia (Culmsee et al, 2005; Lee et al, 2005) In particular, AIF causes large scale of DNA fragmentation and peripheral condensation of peripheral nuclear chromatin, which

is distinct from the global chromatin condensation and oligonucleosomal DNA fragmentation of caspase-dependent death (Cho and Toledo, 2008)

With the sudden increase of intracellular Ca2+ after excitotoxicty insult, calpains, cytoplasmic calcium sensitive cysteine proteases have been implicated in the pathogenesis of ischemic stroke (Lau and Tymianski, 2010) Previous report showed a modest neuroprotection in hippocampal cell cultures from NMDA insults by calpain inhibitors Faddis et al, 1997) Calpain proteolytic activity is necessary for the cleavage and release of AIF from mitochondria (Polster et al, 2005) Neuronal cultures subjected

to oxygen-glucose deprivation and calpain inhibitor treatment was shown to be prevented from undergoing neuronal death due to the inhibition of AIF translocation into nucleus (Cao et al, 2007)

Figure 2-3: A schematic diagram of apoptosis There are considerable cross talks between intrinsic and extrinsic pathway of apoptosis which could ultimately lead to cell death (Adapted from Nakka et al, 2008)

In addition to the necrosis and apoptosis, it was suggested that a third type of cell death, autophagy might be involved in the stroke pathology Autophagy is a fusion process which enables cells to dispose cytoplasms or organelles by fusion of vesicles containing these cellular compartments with lysosomes (Taoufik and Probert, 2008) However, a more detailed understanding is needed for this type of cell death

2.1.3 Oxidative stress of stroke

Under physiological condition, reactive oxygen species (ROS) including superoxide anion (O2˙¯), and nitric oxide (NO˙) is produced at low level and plays a role in cellular signaling such as regulation of blood flow and neurotransmission Intracellular sources of ROS include xanthine oxidase, mitochondrial electron transport chain, arachidonic acid and NADPH oxidase (Brad et al, 2009) Most importantly,ROS production is controlled

by endogenous antioxidants for instances superoxide dismutase (SOD) to dismutate O2˙¯, glutathione peroxidase and catalase to detoxify H2O2

Increased levels of ROS are the major cause of tissue injury after cerebral ischemia (Figure 2-4), in which there are overproduction of ROS, inactivation of antioxidant enzymes, consumption of antioxidants such that endogenous antioxidant defense mechanisms are failed to protect neurons from oxidative damage (Brad et al, 2009) Oxidative stress is the state of imbalance between the two opposing antagonistic forces, ROS and antioxidant, in which the effects of former predominate over the compensating action of latter (Fernández-Checa et al, 1997)

NO˙ and O2˙¯ are two major free radicals responsible in oxidative stress These two free radicals react with each other to produce powerful oxidant peroxynitrite (ONOO¯) Other ROS includes hydrogen peroxide (H2O2) and hydroxyl radical (OH˙) As reported by Zhu

et al, (2004), there are multiple possible mechanisms of free radical production In addition to the basal level of O2˙¯ generation by mitochondria, disruption of the mitochondria electron transport chain can result in autoxidation of flavoprotein and ubisemiquinone to form O2˙¯ Ischemia induced excessive release of glutamate results in increased intracellular Ca2+ which will in turn activates Ca2+ dependent nitric oxide synthase (NOS) and NO˙ production Metabolism of phospholipase A2 and subsequent release of arachidonic acid, prostaglandins, leukotrienes, thromboxanes, and platelet activating factor will be activated during ischemic cascade and produce free radicals as intermediates

In healthy tissue, xanthine oxidase exits as NAD reducing dehydrogenase (Lindsay et al, 1991) However, under ischemic condition, Ca2+ stimulated proteases irreversibly convert xanthine dehydrogenase to free-radical producing xanthine oxidase ATP hydrolysis during ischemic condition causes the accumulation of hypoxanthine Xanthine oxidase catalyzes the oxidation of hypoxanthine to xanthine which can be further oxidized by xanthine oxidase to produce uric acid, O2˙¯ and H2O2 (Parks and Granger, 1986; Warner

et al, 2004) It was shown that xanthine oxidase was increased significantly from 8% to

44% after 30 mins of global ischemia (Kinuta et al, 1989) Allopurinol, a competitive

inhibitor of xanthine oxidase, provides protection against ischemic injury in intestine, heart, kidney and brain (Parks and Granger, 1986; Isik et al, 2005)

In addition, NADPH oxidase is believed to be another major source of O2˙¯ during cerebral ischemia (Jackman et al, 2009; Abramov et al, 2007) NADPH oxidase is expressed in neurons, microglia and astrocytes constitutively (Bedard and Krause, 2007) According to Abramov and co-workers (2007), during ischemic condition, the ROS production in neurons is initiated from mitochondria, followed by the secondary phase of ROS generation associated by xanthine oxidase The third phase of ROS generation is associated with NADPH oxidase

In response to inflammatory response, leukocytes will generate large amounts of O2˙¯ and H2O2 In the extracellular compartment, autoxidation of catecholamines is another pathway for free radical production Endothelial cells also produce free radicals such as NO˙ which is a major component of endothelial-derived relaxing factor (Zhu et al, 2004)

Free radicals damage the membrane lipids, peroxidize the docosahexaenoic acid, a precursor of neuroprotective docosanoids proteins, cleave DNA during the hydroxylation

of guanine, and methylate the cytosine Free radicals block the mitochondrial respiration and facilitate the formation of mitochondrial transition pore permeability (MPTP), resulting in the initiation of apoptosis Free radicals also activate various cell signaling pathway and transcription factors such as nuclear factor-kappa B (NFkB) which regulates the cell death and survival (Nakka et al, 2008) While more intense oxidative stresses can

cause cell death, a moderate oxidative stress is a potent promoter for the apoptosis pathway The mechanism by which oxidative stress promotes the apoptosis is far from understood It has been reviewed that oxidative stress and redox state of neurons are implicated in the signaling pathway that involves phosphatidylinositol 3-kinase/Akt and downstream signaling, which is important for the cell survival (Yamamoto and Takahara, 2009) The possible mechanisms include increased expression of p53, a redox sensitive transcriptional activator of several proapoptotic genes and activation of mitochondrial permeability transition pore (MPTP) to release of cytochrome c from mitochondrial (Fiskum et al, 2004)

Figure 2-4: A flow chart showing the involvement of ROS in multiple ischemic cascades (adapted from Nakka et al, 2008)

Compared to other organs, brain is particularly vulnerable to oxidative stress due to the reason: (i) Brain depends almost exclusively on oxidative phosphorylation for energy production The neurons utilize 20% of the oxygen consumed by the body but constitute only 2% of the body weight, indicating the need of brain for high oxygen consumption and the potential generation of ROS during oxidative phosphorylation in brain; (ii) A high content of iron has been reported in some areas of brains, which can catalyze the formation of ROS; (iii) The brain is rich in poly-unsaturated fatty acids, the targets of ROS attack; (iv) Brain contains relatively low antioxidant defense mechanisms such as SOD, catalase, glutathione peroxidase; and (v) Loss of neurons cannot generally be compensated by regenerating new neurons (Ralf D, 2000)

Several lines of evidence indicate that oxidative stress is a primary mediator of neurologic injury during cerebral ischemia Most of the evidences that ROS participates

in neuronal ischemic injury comes from the use of antioxidants and free radical scavengers that prevent the infarct expansion and restore the neurological deficit function after ischemia (Braughler and Hall, 1989; Tagami et al, 1999) Cerebral protection was observed with mice with overproduction of free radical scavenging enzymes (Weisbrot-Leftkowitz et al, 1998) Furthermore, the extent of delayed neuronal death correlates well with prelethal markers of oxidative molecular alterations (Fiskum et al, 2004)

Neuroprotection was observed in vivo when the animals subjected to stroke insult were

treated with antioxidant or inhibitors of free-radicals producing enzymes Studies on genetic animal models demonstrated that neuroprotection could be observed where genes

encoding for free radical producing enzymes are knocked out or genes encoding for antioxidant enzymes are over-expressed (Fiskum et al, 2004) Therefore, it is believed that pharmacological modification of oxidative damage is one of the most promising avenues for stroke therapy

2.1.4 Rodent ischemic stroke models

Currently, there are two types of animal models of cerebral ischemia used in brain

ischemia studies: global ischemia and focal ischemia (Zemke et al, 2004)

Global ischemia affects the entire brain, which results most commonly from cardiac arrest or other causes of collapse of system circulation, and subsequently failure of brain perfusion The tissue injury is dominated by neurons, occurring especially in the most vulnerable region of the brain first and then proceeding to the least vulnerable region (Miller, 1999) Global ischemia can be imitated by the occlusion of both carotid arteries Two rodent models of global ischemia are routinely used: the 4-vessel occlusion (4-VO) transient severe forebrain ischemia model (Pulsinelli and Brierley, 1979) and the 2-VO plus hypotension model (Smith et al, 1984) 4-VO is caused by the permanent coagulation of the vertebral arteries and temporary ligation of two common carotid arteries while 2-VO is caused by the ligation of the two common carotid arteries with the reduction of blood pressure (Taoufik and Probert, 2008) Both of them have been used for the examination of selective hippocampal CA1 and neuronal death These models create a transient oligemia in the hippocampus, cortex and striatum during ischemia The

occlusion is followed by a complete restoration of energy by blood reperfusion Therefore, the ischemic insult is brief but severe Hypoxia from both cases is termed incomplete with residual 1-4% of blood flow However, complete hypoxia from global ischemia can be achieved by cardiac arrest or ligation of all arteries streaming from the heart In summary, global ischemia involves a short but very intense insults that results in the drastic reduction of ATP and delayed type of cell death to a portion of specific neuronal population, making this model is relatively simplified and less informative to stroke in humans (Taoufik and Probert, 2008)

In contrast, focal ischemia causes the damage only to a portion of brain The size and part

of affected area depends on which vessel is occluded (Zemke et al, 2004) In addition,

collateral flow contributes another major difference between global and focal ischemia (Horst and Korf, 1997) Regions of the brain with most severely impaired blood flow will

be rapidly and irreversibly injured This region is commonly termed as ischemic core Surrounding the ischemic core is hypoperfused region where cells receive moderate blood flow, referred as ischemic penumbra Cells within ischemic penumbra are functionally impaired but metabolically silent Most of the cells in ischemic penumbra undergo delayed cell death and therefore it is potentially salvageable (Brouns and De Deyn, 2009) Therefore, focal ischemia involves much more complicated ischemic cascade events as compared to global ischemia, and represents the closest model to stroke

in human; therefore it is most widely used model in stroke study In rodent models, focal ischemia can be mimicked by the occlusion of one of the major blood vessels that supply the brain, such as common carotid artery and the middle cerebral artery, permanent or

transient followed by reperfusion There are a few models of focal ischemia, for examples, intraluminal suture (Hata et al, 2000), a more distal extravascular clip (Buchan

et al, 1992) and clot embolic model (Kaplan et al, 1991) Intraluminal suture induces severe ischemia in striatum but mild ischemia to cortex while extravascular clip induces more severe cortical ischemia Clot embolic models (Kaplan et al, 1991) have the disadvantage of controlling the accurate timing for reperfusion

In order to resemble a massive and potentially fatal ischemic stroke in humans, permenant focal ischemic stroke by left middle cerebral artery occlusion (MCAO) was chosen as a model for studies in this research Most experiment and clinical research have focused much on MCAO as the infarct formed by MCAO is similar to the brain damage

of ischemic stroke in humans (Miller, 1999) MCA can be occluded close to its branching from internal carotid so that caudate putamen, most neocortical regions, the somatosensory and entorhinal cortex will be affected, and at the distal part where the flow to the basal ganglia will not be blocked and the damage spans the parietal cortex (Taoufik and Probert, 2008) In the case of left middle cerebral artery, it supplies the blood to left cortical areas and also corpus striatum Damage of these areas results in impairment of motor, speech and swallowing functions

Early metabolic responses following MCAO has been well characterized by Folbergrová

et al (1992, 1995), that little changes was observed in between 15 minutes and 2 hours after ischemia Impaired glucose delivery in the core infarct causes the decreased to glucose to the affected territory to 10-26% of non-ischemic values, glycogen was

essentially depleted to 5-12% During ischemia, major losses of ATP (18-32%) and phosphocreatine (16-28%) are resulted Lactate was greatly increased to 5-14 times in severely ischemic core regions For ischemic penumbra, ATP and phosphocreatine are moderately decreased, 53% and 70% of non-ischemic region, respectively However, the lactate accumulation was substantial

Interruption of blood flow by MCAO to the supplied basal ganglia, white matter and cortex causes a gradient of hypoperfusion to emerge, rather than a complete homogenous ischemia of the entire MCA territory (Figure 2-5) The striato-capsular and opercular/insular regions are often the earliest to exhibit irreversible damage Subsequently, as the penumbra is recruited into the core, the latter progressively expands

to other areas, including the cortical mantle The maximum extent of the core will become the final infarct volume (Moustafa and Baron, 2008)

Figure 2-5: The spatial pattern of cerebral blood flow (CBF) in MCAO The figure illustrate the CBF reduction following middle cerebral artery (MCA) occlusion in the baboon brain, demonstrating a gradient from ischemic core (red) through to penumbra and oligaemia (blue) to normally perfused cortex (grey) Values indicate approximate CBF in ml100g-1min-1 (adapted from Moustafa and Baron, 2008)

In this study, permanent left MCAO was induced by transcranial approach in rats Rats are currently the best species to perform MCAO, because it is relatively inexpensive, its cerebrovascular anatomy and physiology resemble that of higher species, and physiologic parameters can be easily monitored The transcranial approach requires a careful removal

of a section of the skull and the underlying dura in order to occlude the middle cerebral artery Tamura et al (1981) developed a subtemporal approach of proximal MCAO at the point near the origin of the lateral striate arteries, which produced infarction of both cortex and the caudate putamen The original technique, however, was very invasive and the rats survived only for a few hours Subsequent modifications including preserving the zygoma and the masseter muscle improved the postoperative survival and eventually the subtemporal approach becomes a standard technique of permanent focal ischemia in rats (Duverger et al, 1988; Nakayama et al, 1988; Menzies et al 1992)