Epilepsy clinical + neurobio

What is the difference between normal and epileptic (seizure) neuronal activity?

NORMAL:

• low-voltage, mixed frequencies, carries complex signals necessary for complex behavior

  • think ordered, like workers in cubicles at an office

SEIZURE:

• high-voltage, rhythmic signal (where the complexity of information is degraded)

  • the order is lost (think an office party)

• this tracks with the excess excitation we see for seizure patients, OR failure of normal inhibitory mechanisms

  • the balance of excitation and inhibition has been thrown off, in favor for increased excitation

During a seizure, the person’s behavior is driven/unresponsive to the environment.

Do we need medical interventions to end seizures, or do they terminate/end spontaneously?

Spontaneously

What are the types of seizure that we can observe?

partial (where only a small part of the brain is involve - the seizure is localized)

• during this type of seizure, the person remains conscious (memory/awareness is working)

• except for the focus area of the seizure, they are able to move (so motor control is largely preserved)

complex seizures are like partial seizures, but the person’s memory and awareness are affected

generalized (bilateral - affects both sides of the brain) —> we see HUGE spikes in brain activity

• has four stages:

  • absence of motor control (petit mal)

  • atonic (drop) —> sudden loss of muscle tone, person collapses/DROPS

  • tonic-clonic (grand-mal) —> muscle stiffening + jerking movements

  • myoclonic (jerks)

secondary seizures start as partial and then become generalized (due to a spread through the brain)

What are the criteria needed to be diagnosed with epilepsy?

• history of 2 or more unprovoked seizures (with no injury, illness, or drug withdrawal involved

What problem does epilepsy reflect?

hyperexcitability of the CNS

• idiopathic (disease causing itself): arises from GENETICALLY-BASED issues with either the functionality of GABAergic neurons, or ion channels

  • we can’t identify a structural cause, so we assume it’s genetic

  • problem is on the cell level

• cryptogenic: abnormal development of the NETWORK associated with lesions (ex. tumors, migrational errors of cells during development, issues with hippocampus)

  • problem is on the NETWORK level

• symptomatic: seizures are secondary to a known brain injury (ex. stroke, infection, traumatic brain injury) —> the seizures are a SYMPTOM of something else

How do anti-epilepsy drugs work?

• by blocking ion channels to bring about hyperpolarization + slow firing rates

• enhancing GABA activity

These drugs can have a lot of side effects though!

What is the incidence by age?

kids are more likely to have seizures than adults, but we see a BIGGER peak/rise between the ages of 60-80

• this tracks because brain injuries can increase for the elderly (and lead to symptomatic seizures) via falls or stroke

Talk about the progressive development of epilepsy. What happens?

• we progress from a healthy brain to an epileptic one gradually

• there is usually an initial insult (event that triggers the development of epilepsy, such as neurodevelopmental issues due to some stressor before or during birth)

• then we go to the latent/sub-threshold period, where changes occur that lower the epileptic threshold

  • ex. inflammation, loss of inhibitory GABAergic neurons, changes in ion channel expression

• then we have the second hit (ex. traumatic brain injury, sleep deprivation)

• then we cross the threshold briefly for the first seizure (and dip back down until the next seizure takes place)

An epileptic brain usually doesn’t go back significantly below threshold and remain just under it - why?

Neurons that fire together wire together - once you have the first BIG burst of SYNCHRONIZED excitability, chronic seizures become much more likely (i.e. it becomes much more likely you will have another seizure)

What does the circuitry look like for each stage?

• early stressor (convulsion/maternal loss) —> leads to development of a “bad brain”

  • leads to abnormal development of the hippocampus

    • the hippocampus is a really important negative feedback regulator of the HPA axis, since glucocorticoids bind to GRs there and (indirectly) inhibit their own release

    • so we end up seeing dysfunction of the HPA axis, which drives up cortisol levels and lowers BDNF

So cell death/decreased neurogenesis happens in the hippocampus (which is ALSO driven by the balance of excitation/inhibition being thrown off overall, leading to glutaminergic excitotoxicity)

This is what eventually leads to seizures (the abnormal phenotype) - and having the first seizure makes it easier for you to have more seizures

What is the effect of stress on epilepsy?

So not only can stress drive relapse (via negative reinforcement) for drug use, and instigate/trigger depressive episodes….

….but it can also exacerbate seizures (shown by the exacerbation of seizures of those who were in DC on 9/11, not seen in those who were not there)

• 50% increase in both self-rated stress and seizure exacerbation

What kinds of animal models do we use in the study of seizures? Genetic or environmental?

Genetic animal models aren’t used very often - instead, we induce seizures in the animals (chemically or electrically) to the point that they will have spontaneous/recurrent seizures on their own

What is the difference between the kindling and status epilepticus models, both of which are used to study epileptogenesis?

Kindling model: the process by which we progressively increase electrical or chemical stimulations to lead to progressively STRONGER/MORE severe seizures

• a pro of this is that this mimics the progressive nature of epilepsy development + most animals don’t die from it

• a con of this is that it is very slow (takes days to weeks)

Status epilepticus model: we induce continual seizures over several hours, JUST once, and have a latent/silent period before we begin seeing spontaneous seizures

• this provides a really good model for temporal lobe epilepsy (rather than the development of epilepsy)

• however, because we see so much excitotoxicity (which also leads to too much Ca2+/free radicals) and complications in other parts of the body, this leads to HIGH mortality

  • the ones who are already most susceptible to epilepsy/seizures will be gone, leaving behind only the most resilient animals

  • in real life, it’s those who already had a vulnerability (via initial insult) that will end up developing seizures —> so the rats don’t reflect the real-life population we want to study

What is the chemical animal model used to study epilepsy? What is the function of kainic acid?

Since epilepsy results from hyperexcitability, it makes sense that we would use an agonist of glutamate receptors in order to model it in animals. This is what kainic acid does

• control animals have regular neuronal activity (low amplitude, low voltage)

• after administration, we see rhythmic BURSTS in excitatory activity, with a much higher voltage/amplitude

Talk about the role of potassium channels in epilepsy - what happens? What happens when we administer an indirect treatment?

K+ channels are critical for repolarizing/hyperpolarizing neurons

• help return the membrane potential to its resting state

• limit neuronal excitability/the firing frequency

• act like a BRAKE on excessive activity

loss of function mutations can lead to seizure activity

However, we can treat this indirectly via adding/transplanting GABA neurons, which leads to seizures being FEWER and SHORTER in duration

What are the subunits of the GABA receptor?

The GABA(a) receptor has 5 subunits

• it’s a ligand-gated ion channel that works by letting in Cl- ions in response to the binding of GABA, which hyperpolarizes the cell

• loss of GABA(a) receptors can lead to hyperexcitability

So, drugs that open GABA(a) channels (agonists) and work to reduce neuronal excitation can be effective anti-seizure medications

What is a functional consequence of atypical GABA subunit distribution in epileptic hippocampal neurons?

we see a decrease in the alpha-1 subunit (which promotes the FAST inhibitory function of the GABA(a) receptor and makes it more responsive to GABA agonists, like benzodiazepines) —> fast, powerful inhibition (phasic)

• this means that we see slower or weaker inhibitory postsynaptic currents, which weakens the brain’s ability to “clamp down” on excitation quickly

we also see an increase in alpha-4 (which are less responsive to benzodiazepines/GABA agonists) —> tonic (slow/”chronic” inhibition)

The changes in function of the GABA(a) receptor, which lets in Cl-/hyperpolarizing current, are apparent even before the actual emergence of seizure activity

• we see a difference even in people who are latent for epilepsy

Talk about astrocyte domains in epilepsy - how are they relevant? Is it healthy to have overlapping or non-overlapping domains?

It is best for astrocytes (glial cells) to have non-overlapping domains - that is what is considered healthy. In epilepsy we see overlapping astrocyte domains, which is unhealthy + CENTRAL to seizure biology. If we correct the overlap, we could correct seizures.

• this allows astrocytes to PRECISELY modulate neuronal activity (ex. via local control of glutamate clearance with EAAT2 glutamate reuptake transporters; synaptic pruning, which is relevant for schizophrenia since excess pruning = decreased connectivity)

• we can stain for these boundaries, and the locations of astrocytes themselves, using GFAPs

  • the astrocyte boundaries can be treated using anti-epileptic treatment (and we see a reduction or rescue in the epileptic phenotype)

What happens when excitatory neurons are hyperactive? How do astrocytes come into play, and how do free fatty acids relate to excess glutamate levels in the synapse/excitotoxicity?

• hyperactive (excitatory) neurons release not only excess glutamate (into the synapse), but also free fatty acids

  • this is the result of breakdown of cell membrane phsopholipids

  • free fatty acids can be neurotoxic at high levels, so they have to be cleared away quickly

    • this is where astrocytes come into play: they produce apolipoprotein E/ApoE to shuttle FFAs AWAY from neurons, which they are toxic to in high levels, and bring them to astrocytes to be taken up

    • so basically ApoE acts like a chaperone protein for them

However, the astrocytes can get overwhelmed by having to store so many free fatty acids. They no longer have enough space for glutamate storage and damage to the cell increases, which can contribute to less expression of EAAT2 (the glutamate reuptake transporter) AND leads to glutamate BUILDUP in the synapse

• more excitotoxicity

This astrocyte is called the lipid-accumulated reactive astrocyte (LARA)

Do we see too little (hypo) or too much (hypermyelination) in animals with seizures? How might this unhealthy myelination contribute to seizure progression, and what knockout model can rescue the phenotype?

we see hypermyelination, which can lead to FASTER transmission (signals will spread faster since axon potentials can summate more easily)

• to counteract the effects of hypermyelination, we can turn to oligodendrocytes (which are responsible for myelinating cells)

• we obviously can’t knock out oligodendrocytes before adulthood, so that normal development can progress. But we can knock out oligodendrocytes (via knocking out progenitor cells) in adulthood through HDAC inhibitors, for example

• this leads to limiting excess myelination and blocking seizures

!! Myelination is activity-dependent - the more that a synapse is “going off” (we constantly see release of neurotransmitters there), the more likely it is to be myelinated because it is electrically active. The oligodendrocytes will SELECTIVELY myelinate axons based on their activity level.

• the inhibition works because myelination is activity-dependent. Epilepsy = too much activity/synchronized firing, leading to hypermyelination.

What are current treatments for epilepsy?

• benzodiazepines (GABA(a) receptor agonists, most effective for the receptors with the FAST/phasic subunit, alpha-1)

• Na+ and Ca+ channel blockers (CArmazapine) —> if we reduce cation influx we can reduce neuronal excitability

• briviact - allows for a protective role of a synaptic vesicle protein (which regulates neurotransmitter release)

• ketogenic diet (since if the brain is constantly using fat for energy, not carbs, we will see more free fatty acids being broken down, and consequently less overwhelm of the astrocytes)

• we can also perform surgery on the temporal lobe, which reduces temporal lobe epilepsy)

How does cooling therapy work to treat patients with epilepsy?

Cooling reduces neuronal activity, which slows the opening of ion channels (which can lead to excess cation influx —> hyperexcitability)

It also leads to a reduction in glutamate levels (thus affecting excitability) without affecting glucose (which is a precursor for glutamate)