By Dr. Lise Johnson (CSNE Education Manager)
Bill Shain is a neural engineer now, but that isn’t what he started out to be. For one thing, when Shain was deciding what he wanted to be when he grew up, neural engineering didn’t exist. There were people doing some proto-neural engineering, but Shain started out on a different track entirely. As an undergraduate at Amherst College he developed an interest in embryology, which as you may have guessed, is the science of how an embryo develops. That’s what took him to Temple University where he earned a Ph.D. in Biology. It wasn’t until he took a position as a postdoctoral researcher at the National Institutes of Health that he developed an interest in brains. As part of one of his projects Shain tried to culture neurons, which means he tried to grow them in a petri dish (or, if you want to sound scientific you can say he was growing them in vitro, which is Latin for “in glass”). By his own account, he was not very successful in this endeavor; the neurons were not easily coaxed into growing outside of the brain. However, while Shain and his colleagues were trying and failing to culture neurons, they succeeded in culturing a different type of cell. They were not initially all that interested in these cells, called astrocytes. Astrocytes are not neurons, but they are brain cells. They belong to a class of cells called the glia, and they’re called astrocytes because they’re shaped like stars (sort of). There are many more glial cells in the brain than neurons, and astrocytes are the most abundant of the glia, so if you take a sample of brain tissue, you’re fairly sure to get some astrocytes as part of the bargain. Glia do a lot of very nice things for neurons like feed them nutrients and oxygen, hold them in place, and clean up after them. As Shain discovered, they also grow pretty well in culture. And, because Shain had a lot of them, he started to get kind of interested in them, and that’s how he ended up an expert on astrocytes. He still was not a neural engineer, but he was getting closer, because as it turns out, there is another really important thing to know about glia – they protect the brain from foreign intruders.
Neurons have a lot of very valuable qualities, but they aren’t killing machines. Total annihilation is just not part of their repertoire. As I said in my last post, if the brain finds a foreign object in its midst, it will do its level best to destroy it, or at least to surround it. Neurons are not equipped for this sort of work, but the glia are. If you look carefully (under a microscope) at the insulating tissue that forms around implanted electrodes you’ll find a lot of glia. Consequently, scientists call it a gliosis or a glial sheath. If the implantation of the electrode causes significant injury, the gliosis will also contain scar tissue, in which case it is called a glial scar. There are lots of different kinds of glia, but there are basically two types that are involved in gliosis. First, there are the microglia. These guys are highly specialized killers and destroying invasive materials is their primary job. Their other job is cleaning up the mess after cell injury or death. So, it’s not surprising that they make up a large number of the cells in a gliosis. Interestingly enough, most of the rest of the cells are astrocytes. So, if you want to stop a gliosis from forming (and if you want your electrodes to function in the long term, you do) you need to know about microglia and astrocytes. Who do we know who knows about astrocytes? Bill Shain. And that’s how Bill Shain became a neural engineer.
To understand how Shain proposes to stop the glial sheath from forming around implanted electrodes, you have to know a little bit about how it gets there in the first place. The first thing to know is that in a normal situation, when there are no electrodes around, astrocytes are busily engaged in all of those support functions that I mentioned before. They will go on cheerfully taking care of domestic business until you give them a reason to do something else. This reason generally comes in the form of an injury, the sort of injury that you might cause when you implant an electrode, for example. When they are injured, astrocytes undergo a bit of an Incredible Hulk style transformation. They enter what is called the “reactive” state, and when they are in this state they start to change. They get bigger, they multiply, and they start to move around - specifically, they migrate towards the injury site. The astrocytes also call for help, and when they do, the microglia come running. The business of the microglia, as I said, is to kill things. When they’re not actively called upon to kill things (or mop up dead stuff), they exist in a resting state. But, when they’re needed, they’re activated, and they go to work. Like the astrocytes, the activated microglia mobilize and multiply. They set out on a hunt for bacteria and debris, and when they find them, they swallow them, and destroy them. I’m not exaggerating; they actually consume and digest intruders. It’s very dramatic. So, what happens when you implant an electrode is that the microglia and the astrocytes become activated, they rush over to the injury site, and there they find the electrode. They quite correctly deduce that the electrode does not belong there, and so they gang up on it and surround it. They literally cover the surface of the electrodes with their bodies, and this effectively isolates the electrode from the healthy tissue (including the neurons) around it. This is bad news, because the reason we put electrodes in the brain is to record the electrical activity of those neurons.
If you want to stop the glial sheath from forming, you have to stop the reactive glial response. To do this, you need to convince the glial cells that the electrode is not a stranger or an enemy and that they don’t need to attack it. But, sitting down and having a nice reasonable conversation with glia is not an option, so how do you do that? This is what I asked Shain. He told me that, before we even begin, we need to recognize that there are actually two different issues that arise whenever we implant electrodes in the brain: 1) the initial injury (which is effectively a stab wound) and 2) the continued presence of a foreign body. You need to solve both problems if you want to stop the gliosis.
It makes sense that you want to minimize the damage that you do when you put the electrodes in, and people have spent considerable time figuring out how to do this. According to Shain, here is the right way to do it:
- Do not put the electrodes in by hand, use a machine. Why is this important? Because you need to put the electrodes in with a constant speed (preferably a high speed) and you want to be able to stop them in exactly the right spot. No matter how dexterous you are, you can’t do this with your hands, so don’t try.
- Don’t put in a whole grid of electrodes at once; put them in one at a time. Electrodes are often manufactured in grids, or arrays, because this particular configuration has a lot of practical advantages. You can make them all at once, you can put them in all at once, they’re all evenly spaced, and you know where they are with respect to each other. The 96 electrodes implanted in Cathy’s brain, if you recall from my last post, were arranged in a grid. The one major disadvantage to the grid configuration is that it doesn’t take into account the topography of the brain. Everyone’s brain is different, and importantly, everyone has a different pattern of brain blood vessels. If you stick a grid in and it blocks one of the little blood vessels feeding that part of the brain, you’ve just killed the neurons you want to record from. So don’t do that.
- Put the electrodes in absolutely perpendicular to the surface of the brain. If they go in at an angle, they make a bigger wound, and we want as small a wound as possible.
- After you’re finished implanting the electrodes, let go of them very carefully, so that you don’t yank on them. This may seem obvious, but it’s easier said than done. However, it’s important.
If you follow these guidelines when you put in the recording electrodes, you should cause very minimal damage, and your gliosis won’t contain any scar tissue. That’s good, but not sufficient to stop a glial sheath from forming, so the problem remains. This remaining part of the problem is called the sustained response, and it is a thorny problem indeed. Clever people have been thinking about this for a while, and here are some of the approaches that they have come up with.
First, let’s remember that brains are distinctly squishy and electrodes are generally stiff and pointy. This would not be a problem if the brain never moved after you put the electrodes in. But, unfortunately, brains move around a lot. In fact, your brain jiggles around a little bit every time you breathe, walk around or drive on a dirt road. This means that if a person has electrodes implanted in their brain, every time they take a breath that stiff electrode rubs against their squishy brain and causes a little bit of damage. What we need are electrodes that are flexible and move with the brain instead of rubbing up against it. Believe it or not, engineers can make this kind of electrode. But, think about it, how are you supposed to put a bendy electrode into a squishy brain? Imagine trying to push a piece of cooked spaghetti into a bowl of Jello. If you push on it, it just folds up. What you want is something that is stiff when you implant it, and flexible after it is implanted, and that’s rather a lot to ask. Nevertheless, some progress has been made by taking the soft electrodes and coating them in a stiff material. The stiff shell dissolves after it is implanted, leaving only the soft electrode behind. Bingo. But unfortunately, while this helps, it doesn’t entirely solve the problem.
The next clever approach has been to coat the electrode with something to change the way cells stick to it. Remember, the real issue is that the astrocytes and microglia physically surround the electrode and barricade it off by clinging to it. Certain chemicals will make it harder for them to stick, so why don’t we just cover the electrode in one (or some) of those chemicals? This technique is called changing the surface chemistry because it has to do with the chemical reactions that happen at the interface between the cells and the electrode. I’m not going to spend a lot of time talking about it, because frankly, while it was a good idea, it doesn’t work all that well.
The next good idea: drugs. I didn’t mention this before, but the reactive glial response is similar in many ways to inflammation. Inflammation is how the body’s immune system responds to infection in the peripheral parts of your body. For example, if you get a splinter in your finger, the tissue around it swells, turns red, feels warm, and hurts. These are the classic signs of inflammation. An electrode is very much like a splinter in the brain, and even though the tissue doesn’t react in the same way, there are many similarities at the cellular level. This being the case, it makes sense that anti-inflammatory drugs might keep this reactive response down. It turns out that coating the electrodes with these drugs before implantation is, in fact, fairly effective at stopping the gliosis. As the outer coating slowly dissolves the drug is released around the implantation site and the glial response is kept at bay. You may have already spotted the flaw in this technique, and that is that eventually the drug coating completely dissolves, and then you just have a bare electrode again. So it works, but not forever, not even close to forever, actually, and that means it doesn’t work for long enough.
And now we come to the new and exciting part, the part that Bill Shain is working on right now. Instead of making the electrodes merely inoffensive to the glia, Shain is actually trying to make them invisible. It’s not that glia don’t mind that the electrode is there, they don’t even know that it’s there in the first place. Obviously I’m being highly metaphorical here because glia are just cells, they don’t really “know” anything. What I’m trying to say is that Shain has found a way to make the glia behave in exactly the same way as they would in the absence of an electrode. If that is what you want, then, according to Shain, what you need to do is use an electrode that has holes in it. Specifically, it should have holes that are spaced about 30 microns (that’s 30 millionths of a meter) apart; none of the solid parts of the electrode should be wider than 30 microns. Why is that important? Because of the astrocytes. Unlike neurons, astrocytes are physically connected to each other by special connectors called gap junctions. Lots and lots of them are connected together. Because of these many connections, astrocytes can share all of their resources and distribute all of their problems. This teamwork is one of the reasons they can take care of the neurons so effectively, and they need to be connected to each other to do their jobs properly. When you put an electrode, or anything else, into the brain, you sever the connections between the astrocytes, and they don’t like that one bit. But, they can and will heal, and over time they reconnect to each other and carry on with their business. At least, that’s what they’ll do if there isn’t anything in the way. And that turns out to be the key. Remember that even though electrodes are small, they’re huge compared to the size of a single cell. For astrocytes they are equivalent to an unscalable, impenetrable dividing wall, and so they can’t reconnect, and they can’t do their jobs properly, and that is what causes them to persist in their reactive state. Recall that the microglia respond when the astrocytes are in distress, and so as long as the astrocytes are separated by this wall, the microglia will also be in their activated state. Ideally you could make an electrode so small that the astrocytes could just reach around it, but practically this isn’t possible. As an alternative, you just open up some windows in the wall. You have to space the windows so that the astrocytes on both sides can find their partners on the other side, and 30 microns just happens to be the right spacing for this to be the case. And then… you’re done. The astrocytes recover from the initial implantation injury and grow through the electrode as if it wasn’t there. They reconnect with their friends, turn off their distress signals, and get back to work as usual. The microglia switch back into sleeper mode, and no gliosis forms. You can plug into your electrode and continue to listen to the surrounding neurons as long as you want. Or at least that’s the idea. Shain’s initial results have been very positive and he is continuing to test this new kind of electrode in his lab. If they work as well as expected, these new electrodes could make long-term implanted brain-computer interfaces a reality for people who need them.
Bill Shain started out to be an embryologist and ended up a neural engineer. In my next post we’ll talk about someone who got to neural engineering through a completely different door.