Spreading depolarization is the sudden, near-complete but potentially reversible breakdown of the physiological ion concentration gradients across neuronal cell membranes. It occurs under pathological conditions such as intoxication with chemicals (e.g. potassium), pharmacological inhibition of the sodium-potassium pump (Na/K-ATPase) or the respiratory chain, or during hypoxia, hypoglycemia, brain trauma or stroke. A complete breakdown of the neuronal ion gradients does not exist in the living brain because neuronal membranes will lyse and the neurons will die beforehand.
Spreading depolarization occurs when neuronal cation outflux by the ATP-dependent Na/K-ATPase pumps locally fails to compensate for cation influx of sodium and calcium. Cells experience an abrupt sodium and calcium overload and a loss of potassium. Water follows sodium and calcium ions entering the intracellular space because these outnumber the potassium ions leaving the cells. The near-complete ionic breakdown and water movement into the cells inevitably lead to cytotoxic edema with swelling of the cellular soma and distortion of the dendritic spines (see Fig. 1: Mechanism of spreading depolarization).
Signaling overload of the system in relation to the normal membrane Na/K-ATPase pump activity can cause a temporary, relatively harmless spreading depolarization. In this case, additional pump activity is recruited to re-establish the normal steady state before any tissue damage occurs. This recruitment will only take about one minute, but even then, tissue ATP will fall transiently by about 50%.
In the event of a breakdown of the Na/K-ATPase function, however, the additional recruitment of pumps cannot re-establish the normal steady state. This situation is critical: the ionic breakdown will persist and will interfere in many ways with the structural metabolism of the neurons. If the cells remain in this non-physiological state for too long, they will eventually die. In animals, neurons can die within this state, or they can show a transient electrical recovery after prolonged metabolic compromise and then die later on. How long neurons tolerate this pathological condition presumably depends on the type of neuron, the developmental state and the initial damage that led to the ionic breakdown.
A Under physiological conditions, low intracellular sodium concentration and high intracellular levels of potassium maintain a membrane potential of -70mV. While there is a constant flux of ions across the cell membrane through ion-specific open non-gated channels, the Na/K-ATPase helps maintain the cells resting potential by hydrolysing one molecule of ATP to move 3 sodium ions out of the cell with each cycle, and 2 potassium ions into the cell with each cycle.
Physiological ion concentrations
|Intracellular concentration (mM)||Extracellular concentration (mM)|
|calcium||0.06 x 10-3||1.2-1.3|
B The core process of spreading depolarization is the failure of the Na/K-ATPase to provide sufficient outward currents to balance the persistent inward currents. This results in loss of electrochemical activity, almost passive ion distribution across the cellular membranes (breakdown of ion concentration gradients), intracellular hyperosmolality with cellular swelling and distortion of dendritic spines (cytotoxic edema), and extracellular hyposmolality with shrinkage of the extracellular space from 20% to 5%. Neuronal membrane potential during spreading depolarization approaches zero.
Ion concentrations during spreading depolarization
|Intracellular concentration (mM)||Extracellular concentration (mM)|
|calcium||25 x 10-3||0.08|
In extracellular recordings, we observe spreading depolarization as a large negative change of the slow electrical field potential that is often referred to as the direct current (DC) potential. Spreading depolarization induces a silencing or suppression of spontaneous brain electrical activity, observed in the alternating current (AC) range of electrocorticographic (ECoG) recordings as amplitude reductions that run between adjacent electrodes. This is termed spreading depression of activity (see Fig. 2: Spreading depolarization and spreading depression in the electrocorticogram).
This animation explains the difference between spreading depolarization and spreading depression of activity in the electrocorticogram. In order to simplify this, we look at the recordings from a single electrode, rather than several adjacent electrodes. This means that we cannot see the actual propagation of both phenomena. Therefore, we refer to them as "spreading" depolarization and depression, respectively.
Spreading depolarizations have been found in various animals, from locusts and cockroaches to monkeys. Therefore, it is not surprising that they also occur in the human brain: they have been electrophysiologically assessed in abundance in patients with brain trauma, intracerebral and subarachnoid hemorrhage and malignant hemispheric stroke. Similar to observations in animals, invasive full-band DC recordings proved that a spectrum from short-lasting to very prolonged spreading depolarizations exists in the injured human brain. Indirect evidence by measurements of regional cerebral blood flow (CBF) or its surrogates suggested that spreading depolarization also occurs during the debilitating (yet relatively harmless) condition of migraine aura.
Spreading depolarizations are restricted to the gray matter of the brain. They do not occur in the white matter. Locally, they last for at least 30 seconds and propagate in the tissue at a rate of about 3 mm/min, and the spreading depolarization-induced spreading depression of activity propagates at a similar rate. Spreading depolarizations can also arise in so-called “isoelectric” tissue, where the spontaneous activity has already been depressed e.g. by previous spreading depolarizations. In this case, spreading depolarization cannot induce further spreading depression.
Traces 1-3: Spreading depolarization in the near-DC/AC-ECoG (bandpass 0.01-45 Hz)
Consecutive onset of a polyphasic slow potential change at neighboring ECoG electrodes (monopolar array).
Traces 4-5: Spreading depression of spontaneous activity in the AC-ECoG (bandpass 0.5-45 Hz)
In electrically active tissue, spreading depolarization typically causes spreading depression of spontaneous activity. Neighboring electrodes show a rapidly developing amplitude reduction in the AC-ECoG (bandpass 0.5-45 Hz). These two traces were recorded in bipolar array but with the same electrodes, and simultaneously with the upper three traces.
Traces 1-3: Isoelectric spreading depolarization in the near-DC/AC-ECoG (bandpass 0.01-45 Hz)
Consecutive onset of a polyphasic slow potential change at neighboring ECoG electrodes (monopolar array).
Traces 4-5: Isoelectric spreading depolarization in the AC-ECoG (bandpass 0.5-45 Hz)
If the depression from a previous spreading depolarization is long-lasting, further spreading depolarizations can occur in this isoelectric tissue, but there is no activity that could be suppressed. Spreading depolarizations in electrically inactive tissue are called isoelectric spreading depolarizations.
An increase in neuronal activity requires vasodilation, and accordingly an increase in CBF to meet the increased energy demand. This process is called neurovascular coupling. The principles underlying the neurovascular response to spreading depolarization may be very similar to those underlying the neurovascular response to physiological neuronal activation. This applies to the release of glutamate and vasoeffectors (like nitric oxide (NO) and arachidonic acid metabolites), ion flux directions, and increase in metabolism and energy demand. During spreading depolarization, however, these changes are much more pronounced.
Under physiological conditions, CBF increases in response to spreading depolarization by more than 100%. This CBF increase typically propagates together with the depolarization wave in the tissue. It is therefore termed spreading hyperemia. Spreading hyperemia seems to serve several purposes: increased delivery of energy substrates (glucose, oxygen) to the tissue, and increased clearance of metabolites from the extracellular space. It outlasts spreading depolarization and only ends after about two minutes. Following spreading hyperemia, CBF shows a mild decrease for up to two hours. This type of mild CBF reduction in response to spreading depolarization is called spreading oligemia.
Under certain pathological conditions, spreading depolarization induces prolonged hypoperfusion due to severe arteriolar constriction. This occurs when there is local dysfunction of the microvasculature, and instead of vasodilation, microarterial spasm is coupled to the neuronal depolarization. The result is a spreading perfusion deficit that leads to a prolonged depolarization phase since there is no energy for neuronal repolarization. The prolongation of the neuronal depolarization and near-complete ionic breakdown are observed as a prolonged negative DC shift. Inverse neurovascular response poses a particular challenge to the neurons, as energy demand increases while energy delivery is decreased. Prolonged depolarization, near-complete ionic breakdown and cytotoxic edema are therefore more likely to cause lasting neuronal damage. In other words, the inverse neurovascular response to spreading depolarization can convert a relatively harmless short-lasting spreading depolarization into a harmful intermediate or even terminal spreading depolarization. In contrast to spreading depolarization under physiological conditions to which hyperemia is coupled, the inverse neurovascular response can thus cause widespread cortical necrosis.
To distinguish it from the harmless physiological spreading oligemia (see above), hypoperfusion as a consequence of the inverse CBF response to spreading depolarization is called spreading ischemia. Current experimental findings suggest that spreading ischemia is perpetuated due to disturbed microvascular reactivity. The classic condition that causes this vicious cycle is a decrease of cortical NO availability combined with an increase in baseline extracellular potassium.
This movie shows a microscopic view of of how pial arteries virtually disappear and a paleness propagates across the surface of a rat's brain in the wake of a typical spreading depolarization. The wave propagates from the left side to the right in this particular area of the cortex. The slow potential change (recorded in the middle panel at the bottom of the screen) starts somewhat earlier than the fall of regional cerebral blood flow (left and right panel). This indicates that the spreading depolarization precedes the perfusion deficit, in contrast to the sequence of events during a non-spreading ischemia (e.g. as a result of middle cerebral artery occlusion or cardiac arrest) in which the spreading depolarization follows the perfusion deficit after some minutes.
Severe focal cerebral ischemia triggers spreading depolarization at one or more points in the tissue of the ischemic core, within 2-5 minutes after the onset of ischemia. Neurons can survive this initial “anoxic” spreading depolarization if the tissue is reperfused, and recovers from it before the so-called commitment point (i.e. the point at which neurons will start to die). If, however, the commitment point is reached, the cytotoxic changes associated with spreading depolarization will initiate the cascades leading to neuronal cell death – even if there is subsequent reperfusion, repolarization, and some recovery of spontaneous activity. There is no general definition of the commitment point, it depends on local levels of perfusion and is different between the various types of neurons.
Anoxic spreading depolarization propagates in similar fashion to spreading depolarization in non-ischemic tissue. Starting from the focal ischemic core, it spreads against the gradients of oxygen, glucose, and perfusion into the adequately supplied surrounding tissue, changing its features in response to the local conditions of the tissue during the course of its propagation. By the time it has reached the surrounding non-ischemic tissue, the initially anoxic spreading depolarization will have become a short-lasting non-ischemic spreading depolarization. In other words, the full continuum of spreading depolarization is observed in this single initial wave, as well as in subsequent spontaneous depolarization waves (see Fig. 4: The spreading depolarization continuum in focal ischemia and in the ischemic penumbra). It is also important to realize that the pharmacological sensitivity of spreading depolarizations to spreading depolarization-inhibiting drugs increases along the continuum.
The duration of the negative DC shift of spreading depolarization indicates the duration of the depolarization and near-complete breakdown of the ion homeostasis. Repolarization requires activation of the energy-dependent Na/K-ATPase. Short-lasting DC shifts therefore indicate sufficient ATP supplies for repolarization at the recording site. Thus, the negative DC shift is a powerful indicator of both tissue energy status and the risk of injury at the recording site.
Many of the spreading depolarizations observed in patients have “intermediate” characteristics somewhat in between those of spreading depolarizations in severely ischemic tissue, and that of normal tissue. Spreading depolarizations with “intermediate” characteristics and duration are measured in the so-called ischemic penumbra. In general, this describes ischemic but still viable - and potentially rescuable - tissue. The original definition of the ischemic penumbra focused on the electrocorticographic features of the tissue: Astrup and colleagues defined the penumbra as a region where neurons are not yet terminally depolarized but where at the same time, CBF levels are so low that spontaneous brain electrical activity has already ceased and thus, spreading depolarizations cannot cause further depression of activity. Later on, Hossmann defined the ischemic penumbra as a region of constrained blood supply and disturbed protein synthesis where energy metabolism is still preserved.
This animation shows the pathophysiology of focal cerebral ischemia. The experimental literature roughly defines the ischemic core, the inner penumbra, and the outer penumbra by perfusion levels below 15ml/100g/min, below 20ml/100g/min, and below 55ml/100g/min, respectively.
In the first part of this animation, we see how minutes after the onset of focal ischemia, an initial spreading depolarization emerges from the ischemic core and spreads against the perfusion gradients into the surrounding tissue.
In the second part of the animation, we take a closer look at the depolarization wave as it propagates against the perfusion gradients. During the course of its propagation – and depending on the increasing levels of perfusion – the wave’s electrophysiological features change. This is indicated by the change of the negative DC shift. In the ischemic core where the tissue is electrically inactive (isoelectricity), the negative DC shift is persistent. With increasing distance to the ischemic core, the negative DC shift becomes shorter and shorter, just as the non-spreading depression is also less and less complete. Subsequently, the spreading depolarization induces long-lasting spreading depression of activity in the zones of impaired tissue perfusion, and shorter-lasting spreading depression in normally perfused tissue.
In the final part of the animation, we see how subsequent spreading depolarizations originate in the ischemic core and gradually cause an expansion of the zone of electrically inactive tissue, beyond the penumbra and into the surrounding, adequately perfused tissue.
One way to study the role of spreading depolarizations for cell death in the ischemic penumbra is the model of middle cerebral artery occlusion in animals. Here, it was shown that the cumulative duration of spreading depolarizations correlated with infarct size and growth. The classic experimental approach to demonstrate their deleterious effect takes advantage of their propagating nature: a harmless, rapidly reversible spreading depolarization is chemically provoked outside of the ischemic penumbra. The depolarization wave then propagates into the penumbra where it changes its biophysical features, becoming a spreading depolarization of intermediate character. With each incoming depolarization wave, the ischemic core now enlarges (see Fig. 4).
Direct histological evidence of the deleterious effect of spreading depolarizations in low-flow regions has been provided in an endothelin-1 (ET-1) model in rats. Topical application of ET-1 to the brain induced a low-flow region that gave rise to spontaneous spreading depolarizations in half of the animals. These animals later showed focal necrosis in areas exposed to ET-1 (as assessed by histological analysis). In the remaining half of the animals, the same ET-1 concentration led to a low-flow region but neither depolarization waves nor focal necrosis occurred (“non-responders”). However, when a depolarization wave was then chemically triggered in non-responders and propagated into the ET-1-induced low-flow region, those animals also developed focal necrosis in the ET-1 exposed cortex. The findings suggested that the depolarization waves initiated the cellular damage in the low-flow region, independently of whether the depolarization wave started in the healthy surrounding cortex or in the low-flow region.
It is quite astonishing how spreading depolarization is the common mechanism of two pathological conditions with very different symptoms and medical histories: migraine aura and stroke. Migraine is a relatively harmless disease but in stroke, spreading depolarizations initiate and facilitate neuronal injury. It was the Brazilian neurophysiologist Aristides Leão who first provided a plausible explanation for this paradox: the energy state of the tissue determines both the duration of the spreading depolarization and the type of depression of spontaneous activity that accompanies the depolarization.
Sudden and simultaneous neurological deficits of transitory ischemic attacks, non-migrainous stroke and cardiac arrest are associated with non-spreading depression of activity, whereas spreading depression of activity seems to be the correlate of the creeping neurological deficits of migraine aura and migrainous stroke. In other words: in migraine, a single spreading depolarization causes spreading depression of spontaneous brain electrical activity, which in turn is responsible for the transitory aura symptoms such as visual disturbances (scintillating scotoma), hemiparesis, or paresthesia. The spreading depolarization is only short-lasting and therefore harmless. By contrast, ischemia causes non-spreading depression of activity. This leads to complete electrical failure and a sudden onset of symptoms such as hemiparesis or aphasia. Only after this electrical failure does spreading depolarization arise. This type of spreading depolarization cannot trigger spreading depression, because the electrical activity has already been suppressed. The depolarization persists and initiates the cascades leading to cellular death and irreversible damage, unless there is timely tissue reperfusion. The spreading depolarization process – depending on its duration – thus initiates the countdown to neuronal death, whereas the depression pattern determines the clinical symptoms.
TBI is the leading cause of death and disability in children and young adults, and the incidence in elderly patients is also increasing. Causes include falls, vehicle and road traffic accidents, and violence. The pathophysiology of human TBI is heterogeneous and complex. The primary injury may consist of any combination of parenchymal contusion, intracerebral hemorrhage (ICH), subarachnoid hemorrhage (SAH), extraparenchymal hematoma, and diffuse axonal injury. Following the initial impact, a whole array of secondary complications such as hypotension, hypoxia, fever, or brain edema with elevated intracranial pressure can lead to long-term disability and neurological complications such as post-traumatic epilepsy.
It is assumed that similarly to the outer zone of focal cerebral ischemia, a hypoperfused, pericontusional or traumatic penumbra with low concentrations of brain glucose surrounds the zone of primary injury with irreversible neuronal loss. In addition to the state of hypoperfusion, another commonality between ischemic and traumatic penumbra is the occurrence of spreading depolarizations. Spreading depolarizations can be recorded in about 56% of patients with severe TBI. These spreading depolarizations can lead to further neuronal injury through spreading ischemia and prolonged ionic imbalance. In penumbral areas, the onset of irreversible cell damage (the commitment point) is reached later than in the inner zone. This gives treating physicians more time to salvage this tissue.
There is, however, one significant difference between the ischemic and the traumatic penumbra: penumbral tissue in ischemia shows a pathologically elevated oxygen extraction fraction (OEF), whereas in the traumatic penumbra OEF is low. This discrepancy suggests mechanistic differences between these two types of penumbra. There is a need for further investigations of the biochemical processes of neuronal injury after TBI, and the mechanisms that differentiate traumatic from ischemic injury.
Development of secondary brain damage following traumatic injury. The first image shows a post-surgery CT scan from the day of the injury with bilateral frontal bone fractures, hemorrhagic contusions, scattered subarachnoid hemorrhage, and diffuse cerebral edema. The second image shows the CT scan after the second surgery (bifrontal decompression) on day 9 after the injury, demonstrating severe injury progression.
aSAH is the second most common type of hemorrhagic stroke, and results from the rupture of a basal cerebral artery aneurysm. The blood leaves the cerebral circulation and accumulates in the subarachnoid space covering the brain surface. Mortality and disability rates of individuals reaching medical care are 30% and 30%, respectively. Among survivors, only 60% are able to resume their previous lifestyles. Delayed cerebral ischemia (DCI) is the most prominent in-hospital complication after aSAH and it is presumably caused by the breakdown of erythrocyte products in the subarachnoid space.
The clinical signs of DCI are global and focal neurological deficits that lead to significant deterioration in functional outcome. In ECoG recordings, clusters of recurrent spreading depolarizations with prolonged depression of the spontaneous activity indicate the occurrence of DCI. Spreading depolarizations can be recorded in approximately 70-80% of patients with poor-grade aSAH. They show an early peak associated with the initial brain damage and a late peak around day 7 associated with DCI.
Development of delayed ischemic infarcts after aneurysmal subarachnoid hemorrhage. The first MRI, taken on day 3 after the hemorrhage, shows an early infarct in the posterior territory of the left middle cerebral artery. On day 7 (second image), we can see a newly developed delayed infarct in the left anterior middle cerebral artery territory (including in the vicinity of the electrode on the left frontal cortex), and an additional small infarct in the left anterior cerebral artery territory.
In addition to DCI, malignant hemispheric stroke (MHS) is the other type of ischemic stroke in which spreading depolarizations can be monitored in patients. It occurs in around 10% of patients with middle cerebral artery infarcts. These patients have massive infarctions that require a decompressive hemicraniectomy to avoid life-threatening brain edema (and increased intracranial pressure). A craniectomy allows implantation of a subdural electrode strip for ECoG, and thus neuromonitoring of spreading depolarizations. Spreading depolarizations occur in practically 100% of patients with MHS.
Infarct progression after malignant hemispheric stroke. The first image shows a CT scan taken 24 hours after symptom onset, the second image another scan after 96 hours. The images show the course of malignant middle cerebral artery infarction: a massive space-occupying edema with subsequent brain tissue shift to the contralateral, compression of formerly healthy brain structures and finally, transtentorial herniation.