Deep brain stimulation in Parkinson's disease: Uncovering the mechanism

Parkinson's disease (PD) belongs to a group of diseases that are referred to as neurodegenerative because they involve the degeneration and death of neurons. In PD a group of structures called the basal ganglia, which play a role in facilitating movement, are predominantly affected. The substantia nigra, one of the basal ganglia nuclei as well as one of the most dopamine-rich areas in the brain, is severely impacted; by the end stages of the disease patients have often lost 50-70% of the dopamine neurons in this region. This excessive loss of dopamine neurons and the accompanying depletion of dopamine levels in the basal ganglia are associated with increasingly debilitating movement-related symptoms, such as rigidity, tremor, bradykinesia (slow movement), and postural impairment.

The most common method of treating PD involves the administration of L-DOPA. L-DOPA is a precursor to dopamine that the brain can use to synthesize more of the neurotransmitter; thus, it works to increase the dopamine levels that are being continuously reduced by the disease. PD, however, is progressive, meaning that neurodegeneration will continue once it has begun. L-DOPA isn't capable of halting neurodegeneration, and eventually the dopamine synthesized from L-DOPA is not enough to replace all that has been lost due to the disorder; with time L-DOPA begins to lose its effectiveness. Especially in the later stages of PD, L-DOPA provides diminishing returns, and the side effects of chronic L-DOPA treatment start to make its continued use more detrimental than beneficial.

Therefore, we continue to seek out treatments for PD that will be more effective in the advanced stages of the disease (while maintaining a manageable side effect profile). In the early 1990s, it was observed in non-human primates that lesions of the subthalamic nucleus (STN) appeared to effectively eradicate Parkinsonian symptoms. Although the reason for this was not fully understood, a hypothesis was formulated based on the understanding that one of the functions of the STN seems to be to inhibit unwanted movements. Normally, this suppression of movement should only occur when a movement is not desired, and thus the interference should be removed with the attempt to initiate movement. In PD, decreased dopamine levels may prevent another structure in the basal ganglia, the globus pallidus, from moderating activity in the STN. This can lead to excessive STN activity, which serves to overly-inhibit movements and may cause the difficulty in making movements that characterizes PD. Based on this rationale and related experimental evidence, the STN was identified as a potential therapeutic target in PD. At that point, though, the only way to reduce activity in the STN was through a surgical procedure that irreversibly destroyed the nucleus.

However, not long after the STN was identified as playing a role in PD symptoms, a new method of influencing activity in the STN (and other brain areas) was developed: deep brain stimulation (DBS). This method was tested in patients with PD for the first time in the mid-1990s. The results were encouraging, as in some cases symptoms improved drastically and the patients were able to reduce their dose of L-DOPA and related drugs significantly. An example of the marked improvements that can occur after DBS is initiated can be seen in the video to the right. Since the first experimental DBS procedures, the method has been used with thousands of patients, making it an established therapeutic approach for the treatment of advanced PD.

Deep brain stimulation procedure

DBS involves the insertion of an electrode into the brain. Thus, it requires an invasive surgical procedure that necessitates making one or two holes in the skull. An electrode is placed in the desired region of the brain (in the case of PD usually the STN but also sometimes the globus pallidus); the electrode is connected to a wire that runs under the skin to a device called a pulse generator, which is usually implanted under the collar bone.

When the pulse generator is turned on, it emits electrical impulses that seem to disrupt neural functioning. This can be used to cause changes in brain activity that resemble what happens when a lesion has been created. Thus, implanting an electrode near the STN and turning on the pulse generator reduces excessive activity in the STN; the abatement of STN activity is associated with an improvement of symptoms.

Although the DBS procedure has seen some success in alleviating symptoms in patients with advanced PD, the mechanism by which it achieves those effects is still unclear. DBS in the STN does reduce STN activity in patients with PD, but it is uncertain why stimulation of a brain region would have effects similar to the ablation of that brain region. Several hypotheses have been put forth to explain the mechanism of DBS, ranging from the assertion that DBS causes changes in neurotransmitter and hormone levels to the proposition that DBS disrupts abnormal neural oscillations in the brains of PD patients. This latter hypothesis has perhaps received the most research attention as a mechanism of DBS, and is considered by some to be the most viable explanation.

Neural oscillations and phase-amplitude coupling

The term neural oscillations describes rhythmic changes in the electrical activity of neurons and can involve fluctuations in the membrane potential of an individual neuron (i.e. action potential) or a small population of neurons (i.e. local field potential). These neural oscillations in certain areas of the brain tend to exhibit patterns of synchronization, which means that the activity of different neural populations becomes regulated on a similar timescale. In other words, synchronized neural populations may (on average) fire action potentials at the same time, then be at rest at the same time. It is thought that these synchronized patterns of neural activity are used to facilitate communication and integrate activity among groups of neurons from different parts of the brain, and thus normal oscillatory behavior may be essential for diverse functions ranging from sensory perception to motor movements

There are several different rhythms of oscillatory activity that can be detected throughout the brain; they range from low frequency delta oscillations (1-4 Hz) to high frequency gamma oscillations (>30 Hz). What makes understanding the effects of neural oscillations even more complicated is that these different frequencies of oscillations can be linked, or coupled, together in such a way that different areas of the brain with different patterns of oscillatory activity seem to work in concert with one another by coordinating their disparate oscillatory behavior. For example, a peak in the activity in one region might coincide with a valley in the activity of another. This mechanism, known as phase-amplitude coupling (PAC) may allow for the syncing of activity across a variety brain regions in a dynamic manner, and is becoming recognized as a key feature of healthy cognition.

Deep brain stimulation as a correction to abnormal oscillatory activity

Patients with PD display abnormally increased oscillatory activity in the STN in the beta frequency (13-30 Hz), which has been hypothesized to disrupt the normal functioning of the basal ganglia in such a way as to impair movement. And, some studies have found that the reduction of this oscillatory activity may be one mechanism by which DBS alleviates the symptoms of PD. However, the signal for voluntary movement originates in the motor areas of the cerebral cortex, and it remains unclear how abnormal beta oscillations in the basal ganglia might influence the motor cortex in such a way as to produce the movement-related symptoms of PD. Thus, it is also unknown how the stimulation provided by DBS might affect the motor cortex to alleviate those symptoms.

In a recent study published in Nature Neuroscience, de Hemptinne et al. (2015) explored the hypothesis that DBS helps to improve the symptoms of PD by reducing excessive coupling of neural oscillations in the motor cortex. In non-PD patients, PAC between high- and low-frequency oscillations in motor areas of the brain occurs when at rest and is reduced when movements are made. It has been suggested that this coupling may inhibit neural activity until movement is initiated; at that point the coupling is diminished so movement can occur. de Hemptinne et al. hypothesized that in PD patients, the PAC is exaggerated and continues to inhibit movement even when a movement is desired. DBS, however, may act to reduce PAC and increase the possibility of movement execution. To test this hypothesis, they used a procedure called electrocorticography in PD patients before, during, and after DBS stimulation of the STN.

Electrocorticography (ECoG), also sometimes called intracranial electroencephalogram (iEEG), involves the placement of electrodes directly on the surface of the brain to record the electrical activity of neurons. Although this is an invasive surgical procedure, the patients in the study by de Hemptinne et al. were already undergoing surgery for the placement of the DBS electrode and thus an additional surgical procedure wasn't required. In this study, the electrodes for ECoG were placed directly on the sensorimotor cortex.

As the authors hypothesized, the ECoG recording before the DBS device was turned on showed excessive beta frequency activity in the STN as well as exaggerated coupling of beta activity with gamma frequency oscillations in the motor cortex. Patients at this point displayed characteristic PD symptoms like rigidity, tremor, and bradykinesia. When the DBS device was turned on, however, the abnormal PAC in the motor cortex was reduced and symptoms were alleviated. Additionally, the degree to which PAC was reduced was associated with the degree to which the patients' symptom severity was mitigated. Thus, it appears plausible that DBS reduces PD symptoms in part by reducing PAC in the motor cortex.

DBS is still considered a last-resort option for most PD patients, as it does involve invasive surgery and all the associated risks, and it is not successful for everyone. However, if we can come to fully understand the mechanism by which it works, we may be able to refine the method and improve its success rate. For example, if PAC in the motor cortex is to blame for the severity of some of the symptoms of PD, future DBS devices could incorporate real-time monitoring of PAC and the automatic adjustment of stimulation to most effectively reduce it.

As we uncover more information about the mechanism of DBS, it may become a treatment for PD that can eventually replace L-DOPA which, despite its therapeutic value, remains a temporary solution to the problem. Regardless of the improvements we make to DBS, however, it still does not seem it will be capable of permanently arresting the neurodegeneration that occurs in PD. Thus, the search will continue for a therapeutic approach to treating the symptoms of PD that can at the same time either slow or stop the relentless loss of basal ganglia neurons that defines the disease.

de Hemptinne, C., Swann, N., Ostrem, J., Ryapolova-Webb, E., San Luciano, M., Galifianakis, N., & Starr, P. (2015). Therapeutic deep brain stimulation reduces cortical phase-amplitude coupling in Parkinson's disease Nature Neuroscience, 18 (5), 779-786 DOI: 10.1038/nn.3997

New approaches to epilepsy treatment: optogenetics and DREADDs

Epilepsy refers to a group of disorders that are characterized by recurrent seizures. It is a relatively common neurological condition, and is considered the most common serious (implying that there is a risk of mortality) brain disorder, affecting around 2.2 million Americans.

Watch this 2-Minute Neuroscience video to learn more about epilepsy.

The seizures associated with epilepsy are not homogenous; they can have a drastically different presentation depending on the patient, the part of the brain the seizure originates in, and how much of the brain the seizure affects. For example, seizures can involve clonic activity (i.e. jerking movements), tonic activity (i.e. rigid contraction of muscles), atonia (i.e. loss of muscle activity), or any combination of motor movements and/or loss of motor activity. On the other hand, they can simply be associated with a brief and subtle loss of consciousness, as in the case of an absence seizure.

One attribute that all seizures have in common, however, is excessive neural activity. Seizures are generally characterized by an increased rate of firing in a population of neurons and/or synchronous firing (i.e. neurons that are normally activated at disparate times are all firing together, leading to large spikes in neural activity) by a neuronal population. Because seizures involve an excessive level of neural activity, ictogenesis (i.e. the generation of seizures) has commonly been considered to involve either the direct excitation of neurons or the failure of a mechanism that inhibits the excitation of neurons.

Pharmacological treatments for epilepsy have been designed from this perspective, and have generally involved drugs that either decrease neural activation or increase neural inhibition. For example, drugs like carbamazepine and lamotrigine treat epilepsy by reducing activity at sodium channels in neurons, which makes neurons less likely to fire action potentials and leads to less overall neuronal activity. Other drugs, like phenobarbital and lorazepam, actively promote neural inhibition by increasing the stimulation of gamma-aminobutyric acid (GABA) receptors. GABA receptor activation typically makes neurons less likely to fire, which also reduces overall neural activity.

Pharmacological treatments for epilepsy, however, leave much to be desired. The side effects associated with them range from minor (e.g. fatigue) to serious (e.g. liver failure), and about 30% of epilepsy cases don't even respond to current pharmacological treatments. Surgical options (e.g. removing an area of brain tissue where seizures originate) can be considered in severe cases. Clearly, however, this is an irreversible treatment and also one that lacks specificity, which means that some potentially harmless (but important) brain tissue is likely to be removed along with areas from where seizures are emerging. Although surgical procedures can allow about 60-70% of patients to be seizure free within a year after the procedure, after 10 years more than half of patients begin to experience seizures again.

One of the major limitations to the current approaches to treating epilepsy is that they lack specificity. For, even if seizure activity can be traced back to excess neural excitation or deficiencies in neural inhibition, it is clear that these problems are not occurring all of the time because--except in rare cases--seizures are intermittent and represent only a small percentage of overall brain activity. Drugs that increase inhibition or reduce excitation, however, are having these effects continually (as is surgery, of course). Thus, the treatment of epilepsy is a rather crude approach that involves exerting a constant effect on the brain in the hopes of preventing a relatively rare event.

Because of this, efforts at designing new treatments for epilepsy have focused on more selective techniques. Although these approaches--which hypothetically involve targeting only neurons involved in ictogenesis--are still a long way from being used in humans, there is some promise associated with them. One method, optogenetics, targets seizure activity by incorporating light-sensitive proteins into neurons and then controlling their activity with the application of light. Another approach, designer receptors exclusively activated by designer drugs (DREADDs), focuses on ictogenesis by incorporating genetically engineered receptors that only respond to a specific ligand into neurons and then controlling neuronal activity through the administration of that ligand.

Optogenetics for the treatment of epilepsy

Optogenetics is a field that combines insights from optics and genetics to manipulate the activity of neurons. It generally involves the use of gene therapy techniques to promote the expression of genes that encode for light sensitive proteins called opsins. In most cases, genes for opsins are carried into an organism after being incorporated into a virus' DNA (the virus in this case is known as a viral vector) or an animal is genetically engineered to express opsin genes in certain neurons from birth. Opsin expression can be targeted to specific cell types and the proteins can be used to create receptors or ion channels that are sensitive to light. When light is delivered to these neurons--either via an optical fiber inserted into the brain or with newer technologies that allow wireless external delivery of light--the opsins are activated. This allows exposure to light to act like an on-off switch for the neurons in which opsins are expressed, making them suitable for experimental or therapeutic manipulation.

One way optogenetics can be used as a treatment for epilepsy is by promoting the expression of a light-sensitive ion channel that, when activated, allows a flow of negatively charged chloride ions into the cell. This hyperpolarizes the neuron and makes it less likely to fire an action potential. Alternatively, opsin ion channels can be expressed on GABA neurons that, when activated, cause increased GABA activity and thus promote general neuronal inhibition. Both of these approaches lead to reduced activity of neuronal populations, potentially decreasing the excessive activity associated with seizures.

When linked to some method of seizure detection, optogenetics can be used to inhibit seizure activity at the first indication of its occurrence. This has already been achieved in experimental animals. For example, Krook-Magnuson et al. (2012) promoted either the expression of inhibitory channel opsins or opsins that activate GABA neurons in different groups of mice, then monitored seizure activity using electroencephalography (EEG) after the injection of a substance that promotes seizures. When seizure activity was detected on the EEG, it automatically triggered the application of light to activate the opsins. In both groups (inhibitory channel opsins and excitatory opsins on GABA neurons), light application rapidly stopped seizures.

Thus, when combined with seizure activity monitoring, optogenetics provides a way to selectively control neural excitation, cutting seizures off as soon as they begin. Optogenetics is still a relatively new field, however, and the work in this area has not yet translated into clinical approaches with humans. There are some significant hurdles to overcome before that can happen. One involves the need for a device that can non-invasively and effectively deliver light, another concerns the need for a non-stationary device that can monitor seizure activity. Advances in wireless light delivery, however, have already been made, and an implantable device to monitor seizure activity in humans was recently tested for the first time. Therefore, while this technology is not ready to be applied to epilepsy treatment in humans yet, its use is feasible in the not-so-distant future.

DREADDs for the treatment of epilepsy

Designer receptors exclusively activated by designer drugs, or DREADDs, are another approach that addresses the desire for specificity in epilepsy treatment. The use of DREADDs involves the manipulation of genes that encode for neurotransmitter receptors, then the forced expression of those mutated genes in an experimental animal. Receptors can be engineered so they no longer respond to their natural ligand, but instead only respond to a synthetic, exogenously administered drug. DREADD expression can be targeted to specific cell populations and, like the optogenetic methods discussed above, the receptors can be used to activate inhibitory neurons or inhibit excitatory neurons.

For example, Katzel et al. (2014) modified an inhibitory muscarinic acetylcholine receptor to no longer respond to acetylcholine but instead only to a synthetic ligand called clozapine-N-oxide (CNO); they then promoted the expression of this receptor in the motor cortices of rats. They administered a seizure-causing substance, then administered CNO, and found that CNO administration significantly reduced seizure activity.

Therefore, it seems that DREADDs also have potential for the targeted treatment of epilepsy. Because activation of DREADDs only requires taking a pill, it is considered less invasive than current optogenetic approaches. However, optogenetics possesses greater temporal specificity in that it can be activated immediately upon the onset of seizure activity and terminated just as quickly. Synthetic ligands for DREADDs, on the other hand, must be administered in advance of ictogenesis to ensure the drug is available to inhibit seizure activity when it begins, and will remain active in a patient's system until the drug is metabolized by the body.

Just as with optogenetics, though, there are some hurdles that must be overcome for DREADD use to translate into the clinical treatment of epilepsy. For example, individuals tend to vary considerably in how quickly they metabolize drugs. Thus, there might be some variation in the time span of protection offered by administration of a DREADD ligand, which in the case of potentially severe seizures could be dangerous. Also, although the ligands used for DREADD activation are chosen based on their selectivity for the designer receptor, a metabolite of CNO is clozapine, a commonly-used antipsychotic drug that also activates other receptors. In rodents, this did not translate into side effects, but the potential for metabolites of synthetic ligands to be biologically active must be considered when attempting to apply the technology to human populations. 

Optogenetics and DREADDs both represent intriguing approaches to treating epilepsy in the future. The intrigue stems primarily from their ability to only target certain cells, which is likely to reduce the occurrence of side effects. Regardless, even if these technologies aren't able to be used to treat humans for a long time--or ever--they still have a place in epilepsy research. For, the use of these tools also allows us more control over seizures in experimental animals, which makes a more thorough dissection of the seizure process possible. At the very least, this should provide more insight into a dangerous, yet relatively common, neurological disorder.

Krook-Magnuson, E., & Soltesz, I. (2015). Beyond the hammer and the scalpel: selective circuit control for the epilepsies Nature Neuroscience, 18 (3), 331-338 DOI: 10.1038/nn.3943

Optogenetics, memories, and mind control

Optics

A few years ago (2010), the journal Nature Methods chose optogenetics as its "method of the year." The fact that optogenetics, in 2010, was already considered a viable approach to studying the brain is impressive in and of itself, considering that all of the seminal work with optogenetics has been done since the year 2000. Because the method is still a relatively recent development, however, it is probably true that the most intriguing work with optogenetics has yet to be done.

What is optogenetics?

Optogenetics incorporates methodology from the fields of optics and genetics in attempting to understand the activity of neurons. Specifically, optogenetic methods can be used to selectively activate individual neurons. This allows researchers to gain a better understanding of the function of these neurons by observing the effects of their activation.

There have been a few different approaches developed to activate neurons; one of the more common approaches was realized with the help of green algae. Green algae possess an ion channel that opens in response to light. When the channel is exposed to light it opens, allowing ions to rush into the cell and potentially causing an action potential to occur. The channel is called channelrhodopsin-2 (ChR2), and algae use its light sensitivity to grow towards sources of light.

Researchers in the early 2000s realized if they could get neurons to express light-sensitive ion channels like ChR2, then they could potentially control the activation of those neurons using pulses of light. So, they packaged the genes that encode for ChR2 into a viral vector and used it to infect neurons. The viral vector carries ChR2 genes into susceptible cells and "infects" them, causing the target cells to express the genes.

Once the genes for a light-sensitive ion channel become incorporated into a neuron, researchers can use light to activate that neuron. They can do this by inserting optical fibers into the brains of animals and using lasers or light emitting diodes (LEDs) to expose neurons to light. You can see this in action in the video below, which shows a mouse that (after being injected with a viral vector containing ChR2 genes) expresses ChR2 in its motor cortex. When researchers apply a burst of blue light to the mouse's brain, this causes a distinct pattern of movement.

What is optogenetics used for?

Optogenetics provides neuroscientists with a method to turn on specific neurons and then observe the effects. This gives researchers a way to make a strong connection between the activity of individual neurons and behavior. In other words, if researchers stimulate a particular area of the motor cortex (as seen in the video above) and this causes a mouse to move counterclockwise in circles, then we can hypothesize that the region stimulated plays a large role in that type of movement. Understanding the role of individual neuronal populations is crucial to understanding behavior and disease.

A study published this month in Nature provides a good example of influential optogenetics research. In the study, researchers (Nabavi et al.) used optogenetic methods to examine the behavior of neurons involved in conditioning fear responses. Normally, when you take a rodent and play a specific tone right before it receives an uncomfortable electric shock, it will begin to associate the tone with the shock; quickly it will come to fear the tone itself. In other words, it forms a memory of what normally follows the tone (a shock) and begins to anticipate it immediately upon hearing the tone.

It is thought that the mechanism of fear conditioning involves the amygdala, a region of the brain that plays an important role in fear processing of all kinds. In the case of associating fear with an auditory signal, the lateral amygdala receives auditory information from the thalamus. When there is an aversive stimulus associated with that auditory information, some neurons in the amygdala may undergo a process known as long-term potentiation, which is a term for the enhancement of synaptic communication thought to underlie memory formation (for more on this process, see this article). This enhancement allows those neurons in the amygdala to "remember" that the auditory information was followed by something aversive, and promotes avoidance behavior upon simply hearing the tone in the future.

Nabavi et al. injected a virus that expressed a variant of ChR2 into the brains of rats, and then waited until it was expressed in neurons of the lateral amygdala. Then, instead of pairing the shock with an auditory tone, they paired it with a burst of light that would hypothetically activate the same neurons in the amygdala that would be activated by a tone. This created a fear response that was similar to what was seen in rats who had the shock paired with a tone. So, the investigators essentially created a memory for an auditory stimulus in these rats, even though there was no auditory stimulus present.

The researchers went on to demonstrate that the memory formation was likely due to long-term potentiation. One way they did this was to use a method of optical stimulation thought to induce long-term depression, which is in some ways the opposite of LTP, as it acts to weaken the connection between synapses. By doing this, Nabavi et al. were able to abolish the memory. Amazingly, they were then able to reactivate it simply by re-stimulating the amygdala with light (foot shock was not needed again).

This experiment demonstrates how optogenetics can be used to activate specific neurons to help us to understand their role in behavior. In this case, researchers activated neurons in the amygdala to show how they are involved in fear conditioning, and expanded upon this by verifying that LTP is important to the fear conditioning process.

Optogenetics and mind control

While optogenetics gives us the ability to explore the functions of individual neurons, at the same time it provides us with the ability to modify the activity of those neurons. In this way, we can influence behavior, as can be seen in the video above when the mouse is prompted to move in a specific direction after optogenetic stimulation. Although tethering an animal to a cable for experimentation would seem to limit the possibilities that could be explored in terms of behavior, wireless optogenetic methods have already been introduced; their use will remove some of these limitations.

But how far can this technology go? Would it be possible to express ion channels sensitive to different wavelengths of light in different areas of the brain, thus giving scientists the capability of controlling a whole panoply of behavior? Indeed, work has already started to move in this direction. Will optogenetic technology one day be able to be applied to humans to influence things like addictive behavior or to treat disorders like depression, essentially modifying peoples' thought patterns in the process? Although this is not right around the corner, it is conceivable, and it is not a far stretch from approaches like deep-brain stimulation that are already being explored for these purposes. Thus, using optogenetics to exert some sort of control over the mind, albeit not of the devious sort that the phrase "mind control" seems to imply (hopefully), may be a distinct possibility at some point in the future.

Nabavi, S., Fox, R., Proulx, C., Lin, J., Tsien, R., & Malinow, R. (2014). Engineering a memory with LTD and LTP Nature DOI: 10.1038/nature13294