Free to roam: implantable optogenetic devices

Technical advancements in neuroscience are hard to keep track of. We've talked about optogenetics, which gives us the ability to activate or inactivate neurons in the brain using light and channelrhodopsin or halorhodopsin. Taking this a step further, we looked at small wires that can be implanted in the brain and are capable of optogenetic stimulation, neuronal recording, and drug delivery all in an awake and behaving mouse. As incredible as that is, and it most certainly is amazing, having a large wire implanted into a mouse brain and attached to a machine is a rather unnatural setup. It can limit the movement of the mouse, and alter its behavior. In order to fix this, scientists have been working on wireless optogenetic devices so can control neuronal activity in mice that are free to move around without having a wire attached to their head. The early instruments that scientists put onto the heads of mice were rather bulky, and weighed around 1-3 grams. This isn't ideal considering a mouse's head weighs 2 grams. One example of this is shown in the picture to the left. Imagine having a bowling ball taped to your head while you go about your everyday life, it would probably change the way you act considerably. Large devices can prevent normal movement and behavior in mice, not allowing them to perform a host of experimental tasks. Bulky machines also aren't ideal for use on other parts of the body besides the brain, like the spinal cord or limbs. Knowing this, the next logical step would be to make wireless optogenetic devices smaller - much smaller. Recently, two labs have made huge strides in slimming down these optogenetic machines. In two rather different ways, they managed to make devices that can be implanted into the mouse in several places on the body, allow for normal movement, and still have the capability to optogenetically activate neurons.&&

In the first article, Montgomery & Yeh et al. made small, implantable devices that weigh about 35mg and are about 20mm³. They are shown with a penny to scale in the picture to the right. To bring us back to the bowling ball analogy, this instead would be like having a AA battery on your head - far less intrusive! Part of the device includes a micro-LED light that can be used to activate channelrhodopsin expressed in neurons. To avoid the use of batteries, these small devices are powered wirelessly. Designing the wireless power mechanism involved some rather ingenious craftsmanship that uses radiofrequencies emitted from the base of a chamber to create electromagnetic energy in the mouse which is used to charge small coils (around 2mm) in the implanted devices. The manner in which this works is complicated and, honestly, a bit over this biologist's head. Suffice to say, it provides a means of constantly charging the device irrelevant to the location of the mouse within the cage. If you would like to learn more, the paper that describes how this works is here. The researchers were able to implant this device into the skull, the spine, and the leg of the mouse without limiting the movement of the animal. When they implanted it into the skull, they pointed the LED towards the premotor cortex which they had expressing channelrhodopsin. Activation of this area would make the mouse start moving considerably. Upon turning on the light, that mouse started running in circles around their cage. This is shown in the figure to the left. In mice expressing channelrhodopsin (ChR2), when the light was on, there was a drastic increase in running, especially compared to mice that didn't express channelrhodopsin. In another experiment, they put the device in the mouse's leg, with the light pointing towards channelrhodopsin expressing sensory neurons in the foot. Activation of these sensory neurons would induce slight pain in the mouse, something it would want to avoid. To show this, they allowed the mice to choose between two different chambers, one that activated the device, and another that wouldn't. When they had the activating chamber turned off, the mouse would spend equal time in both chambers. Once they turned on the radiofrequencies to turn on the LED light -thus activating the sensory neurons - the mouse would consistently move towards the chamber that wouldn't activate their sensory neurons and stay there. Both of these experiments show that the implantable devices are doing their job in activating neurons in both the central and peripheral nervous systems.

The second article that came out introduces us to a slightly different type of implant. In Park & Brenner et al., they developed a soft optogenetic implant that has the ability to move with the body of the mouse. It is shown in the image to the right. Their device has two parts: a base that, again, absorbs radiofrequencies to power itself and a long tail that can emit light to activate optogenetic proteins localized to neurons. This is shown in the picture to the right as well. A soft implant can be worn without impacting the mouse's life or behavior and is also extremely durable. After being implanted, it still functioned reliably for up to 6 months. Because it's so flexible, the researchers worried that bending it would affect how well it works, but they showed that there is only a small decrease in efficiently, even when bending beyond what would happen during normal movements. Devices like these can be used for long term experiments because body heat and consistent use don't affect their efficiency. To show that these devices work, they did a similar experiment to what we talked about above. They implanted them in the back leg of a mouse (this is shown in the lower left part of the diagram to the right), and had the tail of the device go over the sciatic nerve. The sciatic nerve in these mice expressed channelrhodopsin, so when it is hit with the LED light, it would fire causing slight pain. They then let the mouse choose between two cages, one where the implant would be activated, and one where it wouldn't. Again, the mice overwhelmingly chose the cage where the device wasn't turned on, showing that the implant was working as expected.

Both of these devices represent huge advancements in the field. They provide wireless optogenetic control that don't hinder animal behavior and movements, and also allow for implantation in the spinal cord and peripheral nerves. However, they do come with a couple caveats. The ways in which these devices are charged pose somewhat of a problem. The first implant we talked about only works within the confines of their wireless charging chamber, which drastically limits the amount of behavioral tasks they can accomplish. You can get an idea of the size of the chamber as the mouse in the second set of images above is shown inside of the cage. Many behavioral tests need to be performed in specific types of cages and setups, which this may not allow. The second implant is limited by about an 8 inch working distance from their radio-frequency generator. This gives them a little more freedom with what behavioral tests they can use, but it is still a limitation. Implant one shows a slight increase in the local temperature of the mouse by about 1 °C, which could cause some brain damage after long exposure. It also increases the body temperature of the mouse, although not above the normal range. That being said, this can be limited by controlling the amount and strength of use. Another limitation of device two could be its utility in brain implants. Because it's extremely flexible, it may not be ideal for use in the brain. That being said, it was developed as a way to study the spinal cord and peripheral nervous systems: parts of the body that weren't approachable with the last generation of wireless devices. Even with these limitations, both of these advances give us a way to test behavior using optogenetics in a more natural way than we ever could before. I'm sure in just a short time, they will become far more efficient than they already are. At that rate that this field is evolving, it won't surprise me if there are wireless implantable optogenetic devices that can also record neuronal signals and deliver drugs at the same time. In the future, devices similar to these could be used to optimize deep brain stimulation in humans, a technique used to treat Parkinson's Disease. It's exciting to see this field evolve and I look forward to seeing whats in store for us in the future.

References:

Anikeeva, P. Optogenetics unleashed. Nature Biotechnology 34, 43-44 (2016).

McCall, J.G., Kim, T., Shin, G., et al. Fabrication an dapplicaiton of flexible multimodal light-emitting devices for wireless optogenetics. Nature Protocols 8, 2413-2428 (2013)

Montgomery, K.L., Yeh, A.J., et al. Wirelessly powered, fully internal optogenetics for brain, spinal and peripheral circuits in mice. Nature Methods 12, 969 - 974 (2015).

Park, S.I., Brenner, D.S., Shin, G., Morgan, C.D., et al. Soft, stretchable, fully implantable miniaturized optoelectronic systems for wireless optogenetics. Nature Biotechnology 33, 1280 - 1286 (2015).

Image credits:

The first image comes from the McCall et al. paper cited above.

The second set of images are augmented from the Montgomery et al. paper cited above

The third set of images are augmented from the Park et al. paper cited above.