In this study, we performed a detailed and complex investigation of changes in mitochondrial respiration, microcirculation, and histology during global cerebral ischemia in the pig. The use of appropriate animal models of cerebral ischemia is essential for furthering our understanding of the mechanisms of ischemic brain injury. Several models of cerebral ischemia have been developed in different species [9–11, 16–18]. Model of cerebral ischemia should mimic the pathophysiological changes found in human stroke, use procedures that are relatively simple and non-invasive, of low financial cost, and enable monitoring of physiologic parameters and analysis of brain tissue for outcome measures. Although the pig model has a high cost and is labor intensive, the larger animal size makes it easier perform physiological monitoring at multiple time points in the same animal. The technique of induced ventricular fibrillation for establishing global cerebral ischemia is commonly performed in large animals [17, 19], and is very similar to the physiological changes occur during human cardiac arrest. Although the use and development of primates and higher mammalian stroke models is an important future goal, small animal models remain widely used at the preclinical level [10, 20, 21].
The three models of cerebral ischemia used in the present study involved unilateral and bilateral carotid artery occlusion, and bilateral occlusion with hypotension. These surgical procedures produced different effects on microcirculation, with unilateral occlusion for 3 h resulting in a 50% reduction of flow through the small capillaries in the frontal cortex, while 3 h bilateral carotid artery occlusion (without or with hypotension) completely blocked perfusion of these capillaries. Perfusion of medium blood vessels in the frontal cortex was less affected in these models of cerebral ischemia, and flow through large vessels remained constant. Tissue sampling for mitochondrial and histological evaluation was performed in the same region of the cortex after direct visualization of the cortical surface. We chose a 3 h ischemic period because thrombolytic therapy, leading to restoration of blood flow to the ischemic area, could be performed during the first 3–4.5 h after onset of ischemia . Therefore, it is important to explore the processes occurring in the brain during that time.
The impairment of capillary blood flow after carotid occlusion is an expected finding, and may cause mitochondrial dysfunction despite the presence of the rete mirabile. Using the same technique, Perez-Barcena et al.  found significant microvascular blood flow alterations in small, medium, and large vessels in humans who had undergone decompressive surgery as a result of a space-occupying hemispheric infarction when compared with nonstroke control patients. In our study, the relatively high blood flow in medium and large vessels in the UCO group may represent better protective circulation in pigs.
We also observed a decrease in mitochondrial oxidative phosphorylation, but no neuronal death, after 3 h ischemia in all models used. This suggests that dysfunction of mitochondrial oxidative phosphorylation system, and damage to complex I of the respiratory chain in particular (reflected in the decreased respiration with pyruvate but not succinate), may be the primary target of an ischemic insult, and occurs before signs of neuronal death can be detected (by histological methods and TUNEL staining), in our pig model of global cerebral ischemia. Thus, we established that bilateral carotid occlusion together with systemic hypotension is a reliable method to cause cerebral ischemia in pigs without major disruption of brain tissue morphology, and that disruption of mitochondrial respiratory function may be an indicator of early brain tissue damage.
Morphological changes assessed by histology are traditionally used to identify dead or dying cells. Visualization of cell death after ischemic stroke by light microscopy can be used to visualize and measure the complete area of infarction or to analyze alterations at the single cell level. However, a major challenge of this technique is reliable detection of early ischemic changes. Nevertheless, numerous staining methods have been developed to overcome these limitations, including routine histology, silver staining, and fluorescence markers. It is important to note that each of the methods has advantages and disadvantages and there is no 'best’ solution. The most frequently used routine histological stains include hematoxylin and various modifications of the Nissl stain, which can be quickly and easily performed on tissue sections. Using these standard protocols, numerous studies have described the distinct temporal and spatial patterns of postischemic morphological alterations . However, because the progression of postischemic changes depends on the severity and duration of ischemia, the pathology observed can only be reliably interpreted under their specific standardized experimental conditions. In general, there is a loss of Nissl substance detectable approximately 2–3 h after experimental ischemia. Our study confirmed that such morphological changes are not pronounced and are difficult to evaluate objectively in the early period of experimental brain ischemia. These limitations of routine staining procedures can be overcome by the use of suppressed silver stains, fluorescence-based techniques, or evaluation of other parameters such as microcirculation and/or mitochondrial function.
It is now well documented that mitochondria play a central role in the pathophysiology of many neurodegenerative diseases, including stroke and ischemic brain injury. Several potentially deleterious mitochondrial responses including impaired ability to generate ATP , induction of free radical production and formation of a mitochondrial permeability transition pores (mPTPs) , and membrane permeabilization resulting in release of factors that promote apoptotic cell death [27, 28], have been detected during cerebral ischemia. Measurement of the respiratory activity of isolated mitochondria is a sensitive method to determine persistent alterations in mitochondrial function induced by ischemia or other insults. Using this method, we determined that the primary cerebral ischemia-induced injury to mitochondria involved a decrease in phosphorylating respiration (or OXPHOS capacity) with NADH-dependent substrates at 3 h of bilateral carotid artery occlusion, which developed further when bilateral artery occlusion was combined with hypotension. OXPHOS capacity with succinate as a substrate was inhibited only for bilateral artery occlusion with hypotension. These data suggest that suppression of complex I of the mitochondrial respiratory chain was likely the earliest event in the development of ischemic injury in cerebral mitochondria during ischemia. However, we cannot exclude ischemia-induced inactivation or degradation of the pyruvate dehydrogenase complex , which may also lead to inhibition of mitochondrial respiration with NAD-linked substrates. In support of these results, it was previously reported that brain mitochondrial respiration supported by either NAD-linked or FAD-linked substrates exhibited similar changes in rats , while the activity of the complexes I, II, and III of the mitochondrial respiratory chain were decreased after global ischemia in the rat brain [30–32]. Furthermore, brain ischemia does not affect complex IV activity, although a long period of reperfusion markedly inhibited its activity . In our pig models, ischemia-induced inhibition of respiration with both substrates was not induced in the presence of exogenous cytochrome c, suggesting that this inhibition was not caused by loss of cytochrome c from brain mitochondria. This finding is in contrast to observations that loss of cytochrome c in heart mitochondria was the earliest event during ischemic injury, and occurred well before inhibition of complex I activity . We also found that LEAK respiration (with pyruvate + malate) was not affected by ischemia in any experimental group, suggesting that proton permeability of the mitochondrial inner membrane was not changed. This also contrasts with results from heart mitochondria where ischemia has been shown to cause an increase in proton leak . Therefore, our study suggests that there are differences in the dynamics of ischemic injury development in mitochondria between pig and rat models of cerebral ischemia. There are also differences in comparison with ischemic damage to heart mitochondria. Furthermore, our data suggest that mitochondria play a key role in global cerebral ischemia, and may be a potential target for neuroprotection.