Cutting edge techniques that use light waves to visualize and stimulate the brain could offer hope for sufferers of Parkinson’s and Alzheimer’s disease, as well as pushing forward the frontiers of how we understand the human mind. Viviana Gradinaru, a World Economic Forum Young Scientist, discusses the technology.
Video footage from your research shows a mouse with a fiberoptic cable connected to its brain. When you shine a blue light down the cable, the mouse runs around in circles. When you switch off the light, the mouse stops. What’s happening?
This is an illustration of optogenetics. In simple terms, we have modified some neurons in the rodent brain so that they respond to light. When you shine a light, the neurons fire, and that expresses itself in a certain behaviour – in this case, the mouse running around in a circle. Turn off the light, those neurons stop firing, and the mouse returns to its normal behaviour – albeit looking a little confused.
We can also use the same technique to inhibit neurons from firing – that is, to cause a certain behaviour to stop, rather than start. It is even possible to cause different neuronal circuits to react to different wavelengths of light, and cause different behaviours by varying the colour of the light.
How do you make neurons fire, or not fire, in response to light?
We use substances called opsins, which are a type of protein that responds to photons, or light particles. We find opsins in nature, in unlikely places – they exist in various simple unicellular organisms such as bacteria, fungi, and algae in saltwater lakes. There is also a lot of work happening now to engineer opsins with better properties in the lab.
And how do you get the opsins into the rodent’s brain?
By using viral vectors, as are used in gene therapy. Basically, we wrap the genetic code for the opsin up in a specially-constructed virus, and the virus delivers and inserts it into brain cells. The genetic code is customized with primers, which act as zip codes, to make sure opsins are produced in cell types we want, and not in others.
What are the alternative ways of manipulating brain circuits, and what are the advantages of using light instead?
There are two other main ways to modulate brain function – electrical and chemical. The electrical approach, directly stimulating brain areas with electrodes, works very quickly, but it lacks specificity. When you directly stimulate a part of the brain, all the neurons in the vicinity will fire – not only the circuits you want to affect, but possibly some others, too. So there is a big risk of side-effects.
Using chemicals – that is, drugs – to modulate brain functioning is much more specific. You can tailor drugs much more easily to act on only certain neuronal circuits, and not the ones right next to them. But drugs are much, much slower than electrodes to take effect.
Optogenetics offers the best of both worlds. It can be as specific as drugs, and as fast-acting as electrical stimulation.
You’re not yet wiring fiberoptic cables into the brains of humans. What are the implications of optogenetics for human health?
The highest potential of optogenetics is in helping us to understand more about how the brain functions. Studying brain functioning in animal models, such as rodents, increases our understanding of how human brains work – and, with it, our ability to design better treatments for brain-related conditions.
In the near term, the benefits will come from improving the process of testing and developing drugs. In the longer term, optogenetics could make electrical stimulation of the brain more precise and effective by helping us to know where to better position the electrodes that are already being used in humans to ameliorate symptoms of, for example, Parkinson’s.
Or it could ultimately offer an alternative to electrical stimulation in the human brain, combining the best-of-both-world features of speed and specificity. However using optogenetics for fundamental brain research – whose results could lead to non-optogenetic therapies – is in my opinion the most likely contribution of this technology for better mental health.
What are examples of conditions for which optogenetics could potentially improve treatments?
Mood or motor disorders – for example, Parkinson’s disease, tremor, depression, anxiety, addiction, obsessive-compulsive disorder, epilepsy, Alzheimer’s disease, or control of chronic pain.
To start with the near term issue, how could optogenetics improve the process of developing drugs for these conditions?
One of the big problems in behavioural neuroscience is that when you do an experiment involving animal models, obviously you need an experimental cohort and a control, and there will inevitably be some variability between the two groups. Optogenetic treatments are fully reversible, so you have the advantage of being able to use the same animal both for the experiment and as a control. Another advantage is that by varying the intensity of light you can effectively vary the intensity of a condition, which helps to predict which drugs could be most effective in the early or later stages of a condition, all within the same experimental subject. That should make for more reliable statistics, less rodent use, and faster usable outcomes.
How does deep brain stimulation currently work, and how could optogenetics improve it?
The challenge with electrical stimulation is its lack of specificity – you implant an electrode into the brain, and when you switch it on it indiscriminately affects all the circuit elements in the immediate vicinity, even the blood vessels. Some of those may not be ones you want to stimulate, so there can be some serious side effects.
Controlling the tremors caused by Parkinson’s Disease is currently the most successful application of deep brain stimulation. About 100,000 patients have had electrodes implanted in the brain, with very few side-effects. It’s remarkable to see a patient whose hands are shaking uncontrollably instantly regaining control of their movements, literally with the press of a button that activates an electrode implanted in their brain.
We already know where in the brain to implant the electrode to ameliorate Parkinsonian symptoms, but we do not understand why it works and that knowledge (of both electrode placement and mechanisms of action) is even less clear for other disorders. For example in 2009 we published in Science the results of a study using optogenetics to understand how deep brain stimulation works to ameliorate Parkinsonian symptoms. Surprisingly, this study highlighted the importance of selectively controlling axons and not local cell bodies at the electrode tip in modulating animal behaviour, a principle that might play a generalized role across many deep brain stimulation paradigms for motor and mood disorders.
We also showed that a likely origin of these axonal fibers could be cortical areas: shallow rather than deep electrode placement could be beneficial in minimizing effects from haemorrhage, for example. Greater understanding from optogenetics could help guide where to place electrodes more safely, for greater efficacy and fewer side-effects in other disorders as well, such as in treatment-resistant depression.
Would patients need to have fiberoptic cables wired into their brains?
Not necessarily. There are a number of future possibilities. One is developing step function opsins, which would need to be triggered only intermittently rather than requiring a continuous stream of photons – so you would need only occasional photon recharge, analogous to taking a daily pill.
We are also putting a lot of effort into making opsins respond specifically to infrared light, which penetrates a larger area of tissue than other wavelengths. That offers the potential to control neuronal circuits in a wider part of the brain, and also to position your light source further away from the circuits you’re trying to affect, so it becomes less invasive. Other labs are also trying to make opsin-like controllers respond to things other than light, such as ultrasound. The end result would be the same as optogenetics but without the need for fiber insertion in the tissue.
What other directions are you working on?
Two main directions, both complementary to optogenetics: voltage sensing with opsins and anatomical mapping of intact brains by tissue clearing to make brains transparent for better imaging of neurons and their fine projections. Our long-term goal is to understand the long-term effects of deep brain stimulation on neuronal health, function, and ultimately behaviour.
Why are you developing voltage sensing and what is your approach?
In addition to optogenetic control of neuronal activity we need feedback on how exactly the tissue is responding to light modulation. Ideally you want a closed loop, so you can measure the effect that the light is having in real time, and turn it up or down accordingly – for example to stop seizures. We are now using opsins to detect membrane voltage in real-time in defined cell populations.
Opsins can be engineered for diverse properties, including increased radiance, the level of which tracks the membrane voltage changes with high temporal precision. We used directed evolution of opsins to make them better at reporting voltage. We hope to combine optogenetic sensors and actuators to modulate and read activity in defined neuronal populations.
Why do you need to make brains transparent and how do you do this?
Even with the power of optogenetics for control and readout of brain networks, a standing challenge is knowing which circuits to modulate for an intended therapeutic effect: we do not have detailed maps of connectivity across large brain volumes. This can be a serious problem, as our optogenetic study on deep brain stimulation discussed above showed. That electrical deep brain stimulation might act fundamentally through white matter away from the electrode site highlights the need for better brain maps.
It is, however, difficult to create such maps for fine axons that run in bundles throughout the brain when the common method to do this is sectioning the tissue in paper-thin slices, then imaging each one, and putting it all back together with imaging software: it is slow, tedious, costly and error prone. We invested in tissue clearing instead to remove the lipids, which obstruct the view, and created a new method known as CLARITY, which renders the tissue transparent for easy visualization and identification of cellular components and their molecular identity without slicing. This method complements optogenetics, in that it can reveal, with ease, circuit-wide effects of optogenetic manipulations and also aid in mapping novel circuits that need tuning in disease.
However, the mastery, improvement and implementation of both these methods require an ongoing, large-scale, cross-disciplinary effort. This is why Caltech provided resources for us to establish The Beckman Institute Optogenetics Neuroscience Initiative and CLARITY Center (BIONIC Center), a Caltech-wide research effort for the further development, application and dissemination of optogenetics and CLARITY. In an attempt to perfect the execution of CLARITY, we have recently reported in Cell the first case of whole-body clearing – transparent rodents that can be used to obtain detailed maps of both central and peripheral nervous systems.
How does one make transparent mice and what are the benefits of such work?
We’ve learned a great deal from naturally-transparent organisms before: think about all the research on the nematode worm, or the zebrafish, that allows for easy identification of gene mutations and their consequences on organ development and survival. Some of this work was recognized with a Nobel Prize in 2002.
Unfortunately, looking through mammalian tissue is not an equally easy task due to the lipids that are present throughout thick tissues, including the cell membrane, making organs opaque. This means that is difficult to detect normal events and pathological ones too, such as cancer cells spreading or viruses such as HIV infecting our bodies.
The method we developed uses gentle delivery of structural-supportive hydrogels and removal of light-obstructing lipids through the vasculature of intact post-mortem organisms. The hydrogel mesh itself is transparent and secures proteins and nucleic acids into place so we can later detect them with fluorescent labels under a microscope. We call the method Perfusion-Assisted Agent Release in Situ (PARS) and the applications are not limited to neuroscience. PARS clearing and labelling with phenotypic markers could be useful to visualize sparse stem cells, which are easy to miss with conventional methods, or to map cancerous cells or viral particles throughout the body and see where they are most resistant to our treatment attempts.
This method should result in easier, cheaper and faster practices for projects relying on immunohistochemistry and could also have the added benefit of decreased usage of rodents. Just as in all histology work on tissue from humans or animal models, all the steps in our method are done on dead tissue and no pain is inflicted at any time. The result, a transparent organism, means that fewer rodents will have to be used to design and test treatments that might work in humans.
A transparent whole-organism is particularly useful to map the distribution and molecular identity of peripheral nerves at their target organs throughout the body; such nerves could then be modulated with optogenetics in animal models of disease to understand what needs tuning to improve symptoms and the resulting knowledge could facilitate better therapies that rely on, for example, electrically stimulating nerves for better organ function or for decreasing chronic pain.
How quickly do you see progress happening?
The field of optogenetics started as recently as 2005, and has already progressed a long way in that time. The work that led to CLARITY we started in 2010 and we have made great improvements since then – including being able to make whole bodies, not only brains, transparent for applications across all organs.
We have a lot of work ahead, though. For PARS and CLARITY we need better probes for easy diffusion in large tissue volumes and microscopy with suitable hardware and image analysis software for large 3D reconstructions. Widely-available optical platforms such as confocal microscopy can be used to image cleared tissues but continuous developments could cut down on imaging time or increase the field of view – both of which would be very useful. With a standard confocal microscope it takes a full day to image just one small rodent brain. A good scientific question and strong image analysis capabilities can make the fullest use of the techniques we developed.
While it is difficult to predict how quickly research will progress to reach the stage of clinical trials, there is lots of momentum and enthusiasm. Pharmaceutical companies are becoming interested in creating optogenetics units and we also had interest in PARS clearing, which can even be used in human tissue. We have collaborators who are using a passive form of CLARITY (PACT) on human brain tissue from patients who died of Alzheimer’s disease or dementia. And we have also shown that it works to detect cancer cells in a biopsy from a human skin-cancer patient. Looking through tissues with cell and protein resolution can teach us a great deal about pathology progression – which is the first step towards better cures.
It’s a fast-changing – and exciting – area of biomedical research.
Dr. Viviana Gradinaru is an Assistant Professor of Biology and Biological Engineering at Caltech as well as the faculty director of the Beckman Institute for Optogenetics Neuroscience Initiative and CLARITY (BIONIC) Center. She is also a World Economic Forum Young Scientist.
Image: A 3-D visualization of fluorescently-labeled brain cells within an intact brain tissue. Credit: Bin Yang and Viviana Gradinaru