In our cartoon-watching pasts, whenever an animated character—let’s say Wile E. Coyote—got a brilliant idea, we would know not simply because of the sly curl of his lip and the St. Louis-worthy arching of the eyebrow, but also because of the enormous light bulb that would appear instantly over his head. (If it was an amazing idea, the light bulb might even be joined by exclamation marks!!) In short, the brainwave triggered a light source.
But what if this process could be reversed? What if the illumination of a light source could be used to trigger a neurological impulse? One perhaps that did not help someone devise an over-elaborate trebuchet that was doomed to physics-defying failure, but instead maybe just initiated the impulse to run or collapse into an epileptic seizure.
No “what-ifs” about it. That technology is called optogenetics.
Flip the switch
“One of the main challenges in understanding neuroscience right now is understanding how all those different cell types form networks,” explains University of Minnesota neuroscientist Esther Krook-Magnuson, “and how all those networks interact to provide for the physiological functions that we need to be able to navigate our space, to understand what someone’s telling us and all that.”
And likewise, what happens when things go wrong, and in the case of Krook-Magnuson’s research, what happens when an apparently otherwise healthy neurological network produces an epileptic seizure.
“Optogenetics allows you to go in and tweak just particular players and at just particular times to see how that affects things,” she continues. “It allows you to test causality in a way that wasn’t previously possible.”
For epilepsy research in particular, she says, research generated a lot of information about physiological changes that happen in the brain and how epileptic tissue would be different than control tissue. But it was difficult to know which of those changes were simply compensatory changes, which ones caused the seizures and needed to be fixed and which ones were purely incidental, downstream changes that didn’t really contribute to the epilepsy phenotype or help correct the situation.
“With optogenetics, you can go in and ask 'what happens if I take this cell type out of the equation?'” she enthuses.
“The brain is incredibly heterogeneous in terms of cell type, and densely packed,” offers Christian Wentz, co-founder of Kendall Research Systems. “Optogenetics allows cell-type-specific control at millisecond timescales, so one can precisely tease apart one neural circuit's function from another literally microns away.”
He offers the contrast with electrical brain stimulation where thousands of neurons may be recruited simultaneously, providing little or no spatial resolution, or another recent intervention method known as DREADDs (designer receptors exclusively activated by designer drugs), where temporal resolution is much more coarse.
Optogenetics centers on a group of light-modulated or light-gated proteins known as opsins, which are found in a variety of organisms including algae, bacteria and fungi. Comprised of an ion channel and light-sensitive cofactor, these proteins depolarize or hyperpolarize cells in response to specific wavelengths of light, providing the switching effect described earlier.
In 2011, Karl Deisseroth and colleagues published an extensive review in Cell of the microbial opsin family, and since then, work has continued on multiple fronts to expand the repertoire of opsin proteins, in terms of their wavelength sensitivity, molecular activity and cellular biology.
A related group of opsins are chimeric proteins comprised of the light-responsive transmembrane protein and G-protein coupled receptors. Thus, rather than altering the flux of intracellular ions, these proteins trigger cell signaling cascades upon activation with the appropriate wavelength.
“You can generate mice which have specific neurons that express these light-gated ion channels,” explains Lynne Chang, senior application specialist for Nikon Instruments. That expression can be mediated using viral vectors (e.g., adeno-associated virus, or AAV), generating transgenic mouse lines or, in some cases, direct electroporation of the cells of interest.
“Let’s say they are the neurons involved in the mouse walking in circles,” Chang continues. “You open up the head of the mouse and shine light into the brain, and only the neurons that are expressing the specific ion channel send an electrical signal, and you can actually cause the mouse to move in a circle.”
Picking your brain
As suggested, the specificity of introducing opsins to particular cell types within the brain affords researchers the opportunity to identify precisely what cell types are involved in the pathogenesis of different neurological disease states and, in some cases, to actually induce the disease or a model of the disease within the test subjects. Once the opsins are in place, they can then be triggered using a fiberoptic cable emitting the appropriate wavelength of light, the cable typically penetrating directly through the skull and into the brain itself.
Over the past decade, optogenetic explorations have expanded to a wide range of neurological conditions, providing information about normal functioning and pathology in areas such as stroke, epilepsy, movement disorders and even addiction.
For example, Michelle Cheng and colleagues from Stanford University recently presented their work on stroke recovery at the Society for Neuroscience (SfN) 2014 conference in Washington, D.C. In this study, the researchers stimulated neurons in the deep cerebellar nucleus in a rodent stroke model to examine its impact on recovery and to better understand the underlying mechanisms following stroke.
They noted that mice that were stimulated from days five to 14 after stroke induction recovered quickly and with significant improvement in motor skills compared with non-stimulated mice, and that the improvement persisted for at least two weeks after stimulation ceased.
At the same meeting, Peter Kalivas and colleagues at the Medical University of South Carolina described their exploration of drug-seeking behavior in rats addicted to cocaine self-administration. Using optogenetics to turn off critical regions of the reward circuitry in the rats’ brains, the researchers noted a significant reduction in the rats’ drug-seeking efforts that was paralleled by a decrease in the cellular morphology changes that often accompany drug addiction.
Because of the sudden onset of seizures in epilepsy, Krook-Magnuson had to go one step further and design a system that would give her precise temporal control of opsin illumination.
“We developed the online seizure detection software that detects the seizure early on, before it physically appears,” she explains. “Then we can deliver light at just those times to try and stop the seizures.”
By monitoring electrical activity in the mouse brain, her group could identify when a seizure was about to happen and use light stimulation of different parts of the brain to prevent the seizure. As they quickly discovered, the earlier they intervened, the more effective the intervention.
The work also allowed them to virtually dissect which tissues or cell types within the brain offered the most significant effect on seizure development, work that would have been almost impossible with more systemic interventions, whether pharmacological or surgical.
“For example, with the cerebellar work that I talked about [at SfN], by doing it in this on-demand fashion, we could see that there was a difference between doing it in the midline cerebellum (the vermus) or the lateral cerebellum,” she offers. “And that effect would have been lost if we were doing it in a more chronic, constant manner without consideration of the underlying physiology of when the actual seizures were occurring.”
For Krook-Magnuson, optogenetics is about more than simply elucidating and understanding the underlying physiology of the brain and its diseases, though she does not minimize the importance of this—she also sees it as an opportunity to better fine-tune potential targets for intervention.
“We just had a paper come out that says if you just target the granule cells of the hippocampus, you get essentially the same inhibition of seizures as you get if you target a whole bunch of cells there,” she recounts. “So, if you could develop a drug that would just target those cells, then you could get that clinical benefit without putting optogenetics into people.”
By the same token, she also sees an exploratory benefit from the ability to not only inhibit a neurological activity but also induce it.
“You can also use optogenetics to induce seizures and develop new models that then drugs could be tested on,” she adds, although she is quick to point out that epilepsy—or perhaps more accurately epilepsies—are not just one condition, and that one must always view models with caution.
“Even though the healthy brain can have seizures—you can induce acute seizures in those—the biology can be very different than spontaneously occurring seizures where there have been modifications to the network that promote seizures,” she explains. “And so, studying the acute seizures may not translate to the system that is spontaneously generating seizures.”
Move to wireless
On the hardware side, a big area of development is the move toward wireless systems designed to free the test subjects and conceivably allow them to behave more normally.
“Big areas of research in basic and clinically translational neuroscience surround issues of social interaction, fear/anxiety, depression, etc.,” says Wentz, whose company has taken a significant lead in wireless optogenetics.
These behavioral models of psychiatric indications, he suggests, are quite sensitive to external influences, such as those presented by a physical tether to external equipment. This challenge prompted the company’s interest in pushing the use of optogenetics beyond a tethered paradigm.
“We were also interested in developing hardware that allowed us to look at chronic studies and studies of multiple animals interacting, both of which were virtually impossible with an optical fiber tethering to a benchtop laser/LED source,” he continues.
Such systems were largely lacking, however, and so Wentz says it quickly became obvious that to do such work, they would need to engineer their own solutions (see sidebar The accidental neurist below).
Given that one of the appeals of an untethered system is the ability to study behavioral models, however, miniaturization of the wireless units becomes a big issue. It does little good to unleash the mouse from its fiberoptic tether only to burden it with a hat that weighs a significant percentage of its own head.
“When you have a tethered system, all you need is a 200-micron fiber going into the brain, so it could be very light and all the heavy equipment is elsewhere,” explains Krook-Magnuson. “But if you put it all on the mouse in a wireless way, then it all has to be there.”
She then factors in the challenges of trying to do the same sort of experiments her lab is currently doing with its online monitoring system.
“Especially if you want to be able to control it in real time, then you have to have a system for recording and EEG processing and then delivering the light. So for a mouse, which is really small, that can be more difficult.”
From Wentz’s perspective, however, such miniaturization issues are more about financing rather than engineering, suggesting that systems can always be made smaller with sufficient capital.
Perhaps a more important issue for Wentz was the use of wireless power rather than batteries to maintain the portable systems.
“And within wireless power, having robust energy transfer is critical,” he adds. “We worked on the proprietary power transmitters and receivers with our partner company for quite some time, rather than just having two coils of wire inductively coupled.”
So far, Wentz suggests, the small animal models seem to tolerate the hardware itself very well.
“Our specific system design scales very well, so increasing throughput on delicate behavioral experiments has been a big advantage, as well as integration with existing tools via hardware and software APIs that we've released,” he notes.
From models to people
But what if rather than simply using optogenetics to improve our understanding of brain pathophysiology or build models against which to screen possible therapeutic interventions, optogenetics could be used as a therapeutic intervention itself?
“Drug-refractory epilepsy is a major issue,” explains Krook-Magnuson. “By some estimates, it is as high as 40 percent of people who don’t have their seizures adequately controlled through pharmacology.”
For some of those patients, the surgical removal of the damaged tissue may be an option, but even here, she warns, those patients tend to regress over time.
“And for some patients, it is not an option because you can’t remove the tissue without there being other effects, or you don’t know what tissue to take out,” she adds.
In a recent Nature Neuroscience review, she describes these interventions as the “hammer” approach to epilepsy, which “produce a variety of unwanted and sometimes debilitating side effects.”
Could optogenetics be used in a clinical setting much as it is used in the research setting?
The first challenge would be getting the opsins into the human brain (see the sidebar Optogenetic mind control below).
“The way you would probably end up doing optogenetics in people is to use a viral vector approach,” Krook-Magnuson explains. “I like to use transgenic mice because for experimental reasons, they’re nicer. But for people, you would put in a viral vector.”
The idea of using viral vectors within the brain is not new, she suggests, pointing to several clinical trials in Parkinson’s disease—unrelated to optogenetics—that have demonstrated their safe use for the expression of exogenous proteins.
“We need to show that these optogenetic proteins—the opsins—can be expressed safely in human tissue,” she adds. “And then we have to figure out how much expression we need, which cells to target and all of that.”
Wentz describes the opportunities for optogenetic interventions in humans as “very promising,” but like many, sees the challenge in safely delivering viral vectors into humans that provide efficacy without side effects. Outside of the biology, however, he is more confident.
“I believe that we have the hardware to pull this off clinically,” he enthuses.
Making such interventions more palatable is the growing understanding that you don’t necessarily need to involve the entire brain to see an effect of treatment.
In a recent Nature Communications paper, Krook-Magnuson and colleagues looked at the optogenetic control of spontaneous temporal lobe seizures in mice and found that an exceptionally small percentage of cells within a brain region may need to be impacted for an effect to be noted.
“Despite the fact that less than 5 percent of the illuminated neuronal population was directly affected by light intervention, significant seizure control was still achieved, stopping electrographic seizures and reducing the frequency of behavioral seizures,” the authors wrote.
That being said, Krook-Magnuson quickly cautions that the work to date has focused largely on rodents, which have small brains. Thus, when light is delivered to the brain, it impacts a significant proportion of the tissue in the organ, which in and of itself is not a lot of tissue.
“When you scale up to humans, that becomes an issue,” she warns. The physical amount of tissue that will require illumination is likely to be significantly higher, although what that amount is remains unknown and will depend to some extent on the particular technique employed.
This may be where the expanded development of red-shifted opsins—proteins that respond to red-spectrum light rather than blues and yellows—will be required, as red light passes through human tissues more deeply than the other wavelengths. Alternatively, she suggests, a strip of light rather than point illumination may allow for the broader illumination.
Whatever the solution, the growing list of publications related to optogenetics suggests that light bulbs are popping up everywhere.
The accidental neurist
In some respects, the launch of Kendall Research Systems (KRS) and the commercialization of its wireless optogenetics platform was a bit of an accident.
“This was meant to be only a research demonstration and for internal lab use,” says KRS co-founder Christian Wentz. “I had already co-founded a separate clinical neuromodulation venture—Cerenova, Inc.—with Massachusetts General Hospital Neurosurgery and thought that would be where my efforts would be solely focused after my M.Eng.”
But all that changed when the company presented the first demonstration device at the Society for Neuroscience conference, and they were immediately mobbed by academic and industry researchers asking when and how they could buy the technology.
“Initially I offered to build a one-off copy of our prototype for a research group in Germany,” he continues. “They became our (extremely patient) beta tester No. 1 and are still using our tools today.”
Within months, the group decided the technology was worth commercializing within the neuroscience space, but without funding, they needed to be creative.
“I set up KRS in a coworking space in Kendall Square area, pulled together some software and mechanical engineering friends willing to work on an interesting project for heavily discounted rates and served as hardware engineer,” he recounts. “We worked at night and on weekends, mostly outside of our day jobs.”
He reserves particular kudos for the “heroic efforts” of the KRS firmware engineer who had been previously involved in developing complex systems at Motorola.
With designs in hand, Wentz approached a handful of the most enthusiastic researchers who had clamoured to their original demo with an offer:
“We are in the process of development. If you're willing to put in a deposit, we will provide this technology to you as a beta tester at a heavy discount from target market price, and we'll continue to provide hardware/software release updates until we've met the design spec.”
Working out arrangements with each of the institutions and putting any fabrication cost shortfalls on Wentz’s personal AmEx card, KRS used the seed money to build a beta product and ultimately secured additional SBIR funding with its partner company Ferro Solutions.
They launched the first version of their platform commercially last November and are now scaling the team and product line beyond optogenetics.
“The goal of KRS has always been to take top engineering talent,” Wentz says, touting the resources of MIT and greater Boston, “and leverage technology gains from larger industries capable of supporting innovation at scale and applying it to neurotechnology.”
Optogenetic mind control
Turning the optogenetics paradigm described elsewhere in this article onto its head, Martin Fussenegger and colleagues in France and Switzerland recently described their efforts to use the human brain to control optogenetic expression of a transgene in mice with the help of a device that looks like a small USB key but of cells rather than memory.
Reporting in Nature Communications, the researchers described their efforts to use a near-infrared (NIR) light-activated protein in transgenically modified cells. When illuminated, the protein generated a secondary messenger that triggered a cell signalling cascade, which activated the final transgene.
The researchers initially tested the system in tissue culture, showing that exposure to NIR light led to the signalling cascade and transgene expression.
They then moved the experiments in vivo by implanting hollow-fiber microcontainers encapsulating the transgenic cells—the “USB key”—under the skin of the mice. Upon illuminating the mice transdermally, the researchers were able to detect elevated transgene product in the bloodstreams of the treated animals.
Taking the experiment into a mode reminiscent of the film “The Matrix,” the researchers then wirelessly linked the NIR light on the microcontainers to an EEG headset, creating a brain-computer interface. Using this interface, they could tune the illumination power and duration to three states of the headset wearer: self-trained biofeedback (holding an observed meditation-meter within a specific range), intense concentration (computer gaming) and meditation (relaxation).
They found that not only did the human subject control NIR illumination in the mice, but also the mind control intervention resulted in modified transgene expression levels as shown previously.
“By combining cybernetics with optogenetics, we now provide the missing link enabling mental states such as biofeedback, concentration and meditation to directly control the transgene expression in living cells and mammals,” the authors wrote.
They then pushed the envelope on what this could mean for patients.
“Far into the future, patients may either learn to generate specific mental states (for example, pain relief), locked-in syndrome programming or having disease-related brain activities (for example, epilepsy, neurodegenerative disorders) close-loop control, therapeutic implants producing corresponding doses of protein pharmaceuticals in real time.”