While high altitude mountain climbers know that the human brain can eventually adjust to low oxygen levels, sadly the same is not true in the clinical setting. Within minutes of a stroke or heart failure, delicate neural tissue is already beginning to shut down as the processes that initially protected the brain begin to fail and irreversible damage ensues. But not all creatures suffer such drastic consequences as tissue oxygen levels fall;several species of turtle and fish can survive lengthy periods of oxygen deprivation, emerging unscathed. The mechanisms that protect these hypoxia tolerant creatures don't fail, and Peter Lutz is optimistic that understanding how protective mechanisms differ between vulnerable and tolerant animals could eventually help clinicians design new therapies based on naturally tolerant organisms. Teaming up with Bob Boutilier, Lutz realised that the time was ripe to bring together a collection of review articles discussing the hypoxic brain and mechanisms of tolerance from an integrative perspective, ranging from creatures that survive truly anoxic environments and the adaptations that protect them, through to the hypoxic brain in a clinical context.

Hunkered down in an icy lake over winter, the chances for trapped fish and other animals must seem slim as oxygen levels in the water dwindle and even become anoxic. But come spring, when the captives are released, many of these creatures emerge unscathed. Despite the unfavourable conditions, they have adapted to survive low oxygen levels where most would perish. Curious to know how the brains of these, and other hypoxia tolerant creatures, survive the otherwise fatal conditions, scientists from across the planet have focused on animals ranging from insects, through to hibernating mammals, in an effort to understand the origin of each creature's strategy for success.

Although Drosophila can tolerate several hours of anoxia, one of the real champions of anoxia tolerance is the crucian carp, routinely surviving lengthy periods of anoxia entombed beneath ice during the long northern winter. As Göran Nilsson explains, the fish perform this featby rapidly and reversibly reprogramming their metabolism to adjust glycolysis and ATP consumption in a highly coordinated manner'(p. 3131). Outlining the crucian carp's survival strategy, Nilsson describes how the fish become almost inactive during periods of anoxia, downregulating their brain's metabolism with the neuroinhibitor GABA, to the extent that they become deaf and blind. Having reduced their metabolic demands, the fish switch to glycolysis to meet their meagre ATP requirements, and as Nilsson points out,the only factor that eventually limits [the crucian carp's] survival time is the total exhaustion of its glycogen store'. However, instead of producing the toxic byproduct, lactate, when switching to glycolysis, the fish generate and excrete ethanol instead.

While the crucian carp's anoxia resilience is almost legendry,overwintering turtles have also evolved a suite of coping strategies to protect their brains from damage while trapped in icy lakes. Fascinated by the reptile's resilience, Peter Lutz and Sarah Milton outline how the freshwater turtle's brain has relatively low aerobic requirements at the low temperatures experienced in a freezing lake(p. 3141). They add that the turtle's brain is well stocked with glycogen, should aerobic respiration fail to sustain the animal's metabolic needs, and that the animals conserve energy by shutting down costly ion-channels and producing inhibitory neurotransmitters to protect the neural tissue from the dangers of depolarization. As soon as the lake's oxygen level becomes dangerously low,the reptile drastically suppresses metabolism to the point of becoming comatose, cutting the brain's ATP demands and switching to glycolysis. However, in contrast to the crucian carp, which excretes ethanol, the turtles protect themselves from the effects of lactate by storing the acid in their shells. Lutz also outlines other energy saving molecular mechanisms that protect the reptile's dormant brain from damage, in readiness for it to resume activity as soon as the thaw begins.

Although turtles are always prepared to survive periods of anoxia,mammalian foetuses also seem well adapted to low oxygen levels throughout the course of their development. Outlining the factors that protect newborn mammals' brains, Susan Vannucci from Columbia University, New York, discusses how the relatively low metabolic rate of young mammals' brains protect them from hypoxia, as well as their ability to metabolise alternative energy sources such as ketone bodies to produce ATP(p. 3149). She also explains that the composition of young mammal's NMDA receptors, which mediate brain development, and that the receptor's regulation differs from adult NMDA receptors, altering the brains susceptibility to high levels of toxic glutamate released from neurones in response to an hypoxic insult. Despite the neonatal brain's resistance to some aspects of hypoxic stress, Vanucci warns that young mammals can also be susceptible to hypoxia due to the low levels of antioxidants in young brains, which leaves them vulnerable to damage by hydrogen peroxide produced when oxygen re-enters the system.

Neonatal brains are also vulnerable to hypoxia-induced apoptosis through several distinct apoptotic pathways, although their relative importance is still unclear.

Neonates aren't the only mammals that have developed tolerance to hypoxia;mammals that hibernate the winter away also benefit from hypoxia tolerance well beyond the early stages of life, thanks to a suite of neuroprotective mechanisms. Kelly Drew points out that this tolerance is hypothesized to stem from adaptations necessary for successful hibernation' even though most hibernating animals never experience hypoxia during their long hibernation(p. 3155). Drew describes how these dormant animals dramatically drop their body temperature and reduce their metabolism to conserve oxygen. They also suppress the immune system and activate a variety of cellular stress pathways to protect their brains from damage.

Although many resourceful creatures have developed a remarkable degree of tolerance to hypoxia, some better known mammals are also teaching us a great deal about mammalian responses to hypoxic conditions over short and long periods.

One major question that is only now being addressed is how do mammalian brains know when they are becoming hypoxic? What are the cellular signalling systems that alert the brain to an imminent disaster, and how does the tissue respond to protect itself? Outlining various molecular mechanisms that are invoked to protect the hypoxic brains, Helmut and Till Acker describe how the tissue conserves energy by inhibiting protein synthesis(p. 3171). The brain also inhibits neural depolarisation, which is not only energetically costly,but can also cause fatal damage. Many of the coping strategies employed by the brain are regulated by a key protein transcription factor called Hypoxia Inducible Factor-1 (HIF-1), which activates many of the key cellular process that initially afford protection to an oxygen-starved brain. Acker also describes several other signalling proteins that could be involved in detecting the initial drop in oxygen levels as hypoxia sets in, including NADPH, a key electron transport protein in the mitochondrial chain, and other haem proteins.

While understanding the key molecular signals that trigger the brain's hypoxia response is crucial if we are ever to develop a full understanding of the effects of hypoxia, Joe LaManna is focusing on the effects of long-term hypoxic exposure on rats (p. 3163). It turns out that rats have a mechanism for tolerating hypoxia that involves increasing blood delivery to the brain by increasing the tissue's vasculature during the first weeks of exposure; the brain's blood supply doubles in three weeks. LaManna suspects that a key component of the hypoxia signalling system, HIF-1, activates vascular endothelial growth factor to increase the capillary density, as well as activating transcription of a glucose transporter to increase the brain's glucose supply and supplement aerobic respiration. And the effect is reversible too. Once returned to a normoxic environment, the capillaries thin back to their normal density by apoptosis.

Despite the many coping mechanisms that have evolved to combat the effects of hypoxia, once the tissue has been damaged it is at risk of triggering apoptosis, and while the brain is particularly susceptible, the heart is no less vulnerable. However, while investigating the effects of hypoxia on rat neonate cardiac cells, Keith Webster and his colleagues in Miami, Florida made a startling discovery; in their hands, the cells appeared to tolerate hypoxia(p. 3189)! Puzzled,the team began testing the cardiac cells' resistance, and discovered that the cells only became vulnerable to hypoxia at acidic pH, while they retained function at neutral pH when supplied with glucose. Comparing the gene expression profiles of hypoxic cells at neutral and acidic pH, Webster found that a cell death factor known as BNIP3 became activated when the cells were hypoxic, but seemed to have no affect until the cell's pH fell. Only then could the deadly protein move across the cell's mitochondrial cell wall, and deliver its lethal message. So apoptosis is not always a natural consequence of an hypoxic insult, and certainly depends on physiological factors other than oxygen depletion. While several reviews in this issue describe some of the molecular mechanisms that allow brain tissue to sense the onset of hypoxia, Klaus Ballanyi discusses one aspect of the tissue's response to the approaching threat (p. 3201). Outlining many of the protective roles of the ATP-sensitive K channel (KATP) in brain hypoxia, Ballanyi explains how the channel detects falling levels of ATP and hyperpolarizes neural membranes, by increasing their permeability to potassium, to conserve energy. Unfortunately,this only offers a short-term solution to oxygen deprivation, as the increased permeability eventually leads to potassium losses that cannot be sustained by the brain's depressed metabolism. Ballanyi also describes how the channel inhibits calcium uptake, preventing neuronal calcium levels from reaching a critical level and triggering cell death. In the case of neurones involved in the regulation of respiration, KATP channels seem to play a significant role in respiratory regulation, and may in some cases also function as a glucose sensor.

While many cell factors outlined by Ballanyi and other contributors to this collection are relatively well known, Hugo Marti outlines the neuroprotective role of a glycoprotein better known for its role in erythrogenesis:erythropoietin (EPO). Marti explains how EPO may function as an oxygen sensor whose levels change in response to hypoxia and stimulate red blood cell synthesis in bone marrow (p. 3233). However, the discovery of EPO in the brain has triggered interest in its many neurological roles, including foetal brain development and enhanced neural survival, and EPO may also play a role in neurogenesis after hypoxia. Marti outlines how EPO may protect cells from damage by limiting the release of glutamate during a bout of hypoxia, thus preventing neurones from embarking on apoptosis, while also stimulating antiapoptotic agents through a variety of cell signalling pathways. Knowing that EPO seems to protect brain tissue from the effects of hypoxia, Marti points out thatEPO might have a beneficial effect for the treatment of stroke patients' andcould be of interest in a large variety of brain injuries'.

But a cell's survival doesn't depend solely on its ability to sense impending danger, it must also be able to repair damage that has already occurred, and Rona Giffard discusses how a family of stress proteins, known as the heat shock proteins, seem to repair proteins damaged during hypoxia(p. 3213). Heat shock proteins are a ubiquitous group of proteins that prevent the aggregation of damaged proteins, and in some cases assist unfolded proteins to refold. Giffard explains that regulation of the state of protein folding and protein association is... severely perturbed by ischemia and reperfusion'. Noticing that neurones that produce Hsp70 survived hypoxia and suffered a reduction in the incidence of protein aggregation, Giffard and her team also identified a co-chaperone, called HDJ-2, which prevents protein aggregation and seems to inhibit cell death. Giffard adds that identifying those actions of the chaperones which are most important for blocking injury will most likely lead to the development of novel approaches to reduce damage from both chronic and acute neurodegeneration'.

Although oxygen deprivation can be catastrophic for brain tissue, the damage caused when oxygen returns to the system can be equally disastrous. As David Warner points out, oxidative damage can occur through many possible mechanisms. Warner outlines how identifying key components of oxidative damage pathways could produce new targets for post-ischemic therapeutic intervention' (p. 3221), and suggests some possible ways of blocking oxidative damage by inhibiting lipid peroxidation, upregulating superoxide dismutase and other free radical scavengers, as well as upregulating glutathione peroxidase to remove toxic hydrogen peroxide.

Of course one of the main thrusts behind many of the studies outlined in this collection of reviews is the need to find new therapies offering protection to the millions of victims who suffer heart attacks and strokes every year; as Phil Bickler from the department of Anesthesia at University of California at San Francisco points out, an effective treatment for brain ischemia is a pressing medical need'(p. 3243). Unfortunately, despite the enormous amount of funding and research energy that have been poured into this effort, no effective therapies have emerged so far,and according to Bickler, only one study that re-vascularized blocked arteries in early stroke victims, has showed any promise. However, Bickler remains optimistic, pointing out how all neurones have some capacity to adapt to changing oxygen tension', and that clinical lessons can be learned from these intrinsic abilities. Focusing on turtle neurones as an example of mature neurones that are capable of tolerating hypoxia, Bickler discusses the possibility of transferring the lessons learned from turtle neurones,hypoxia-tolerant neonates and our own innate tolerance at high altitudes, to the development of neuroprotective therapies in mammalian brains.

Having drawn together such a stimulating collection of review articles,Lutz is optimistic that it will encourage much productive dialogue between traditional and comparative physiologists, in the hope of developing novel clinical approaches to treating the catastrophic neural damage caused by ischemia as we progress through the first decades of the 21st century.