Near the top of your wish list might be a way to selectively turn on or off individual wires in the black box, because it would allow you to falsify causal linkages, associating observable changes in robotic behavior with your manipulations.
Near the top of your wish list might be a way to selectively turn on or off individual wires in the black box, because it would allow you to falsify causal linkages, associating observable changes in robotic behavior with your manipulations. Ultimately, you would create a circuit map that relates structure to function-the quickest, most robust way to accomplish your assignment.
The desire for circuitry control is familiar to brain biologists but its implementation has been quite elusive. The presence of a trusty electrode implanted in neural tissue has been helpful, but it is ultimately a blunt instrument. Electrodes generate local electrical fields over large areas, effectively triggering responses of all neurons within their range, eroding specificity and spatial control. Whether alien robot or flesh-and-blood brain, greater resolution-even on/off control of single cells in a neural circuit-would be a handy technology.
While it may sound like the stuff of science fiction, this goal is about to be realized. In this column, I examine an extraordinary series of experiments that led to the molecular engineering of on/off switches in nerves. These experiments involved manipulating wavelengths of light. First, I review some properties about protein chemistry and light, then move to a brief review of ion channels and, finally, describe the technique itself.
Neural on and off switches
As you remember from school days, neurons are studded with ion channels, which are complex aggregates of specialized proteins. When opened, these channel complexes permit selective movement of ions across the cell membrane, allowing the neuron to control the generation of action potentials (rapid, all-or-none electrical signals propagated along the length of the neuron that allow long-distance neural signaling). Other channel proteins, when opened, reverse the process, inhibiting electrical activity. The activation and inhibition of these complexes serve as the natural electrochemical on/off switch of the neuron.
The experiments described here involved creating unnatural regulatory switches. This was accomplished using photosensitive ion channels. Light has been interacting with biological systems for millions of years and ion channels are found in many of these systems. For example, plants take their place as the most important part of the food chain because of their photosynthetic ability to turn light into energy.You can read the words on this page because of the ability of proteins in your eyes to change their shape in response to light.
One of the most intensively studied light-sensitive proteins is rhodopsin. When light hits this protein in the retina, it causes the protein to change its 3-dimensional conformation. The cell that carries the protein registers this shape change and a chemical cascade is initiated, which eventually results in neural activity.
Although we often think of ion channels as functioning only in the complex world of mammalian neurophysiology, such proteins exist in many living things, including algae and bacteria, some of which are light sensitive. It was the clever use of these admittedly simpler photoreactive proteins that ultimately led to the ability to turn neurons both on and off with light frequencies (Figure) [Figure restricted. Please see print edition for content].
One such protein was discovered in Chlamydomonas reinhardtii, a green alga studied in molecular biology laboratories throughout the world. The protein is called Channelrhodopsin-2 (ChR-2) and is a cation channel that opens on exposure to blue light. The algae use this sensitivity in photo-orientation behaviors ("feeding," as it were), attempting to find the best environmental conditions for photosynthesis.
The gene for ChR-2 was eventually isolated and sequenced. As part of its general characterization, the gene was inserted into a neuron and its function was assessed (mammalian neurons do not possess this gene). When blue light was shone on the cell, the channel opened up, as it would in an algal background. To the researchers' delight, the neuron immediately depolarized! With selective exposure of this cell to a particular wavelength of light, an action potential was triggered. The effect was found to be both fast and reversible. Thus, an on switch was created.
Now to the off switch, which required researchers to turn to another simple ion channel protein. This was found in an archaeum, which normally resides in a dry salt lake in the middle of the Sahara Desert. The creature is called Natronomonas pharaonis and has a channel protein called Natronomonas pharaonis halorhodopsin (NpHR). This protein is used by the archaeum as a pump, allowing it to become engorged with almost ridiculous amounts of chloride ion. It uses this influx to generate an electrochemical gradient that is useful in energy production (adenosine triphosphate synthesis).
As part of its general characterization studies, NpHR was also inserted into a neuron and its function was assessed (mammalian neurons also do not possess this protein). When blue light was shone on the cell, nothing happened. The same was observed with red and green light. But, when yellow light was shone on the cell, the channel opened up. NpHR then behaved as if it were still in the desert, allowing copious amounts of chloride to enter the neuron (eventually filling the cell with almost 50 times the amount normally found in a neuron). This rapid influx of chloride had the effect of hyperpolarizing the cell, which prevented it from generating action potentials. The effect was rapid and just as reversible as the ChR-2 protein. With selective exposure of this cell to a particular wavelength of light, an action potential could be inhibited. Thus, an off switch was created.
For persons interested in selective activation, these proteins are research heaven. You could insert ChR-2 into a neuron and turn it on at will with a wavelength of light. You could drop NpHR into another neuron and turn it off at will with a wavelength of light. You could even drop both genes into a single neuron-or a group of neurons-and then turn them on and off by selectively exposing them to blue or yellow light. Not only do you have an on/off switch, you have a remote control!
Does it work in real life?
These manipulations worked as expected with cultured neural cells. The manipulations did not alter the cells' electrical properties or other normal functions-they just turned them on or off. In NpHR, the degree of inhibition could be manipulated like a dimmer switch; it was shown to be dependent on the intensity and duration of exposure to yellow light. The effect was so robust that short pulses of light could knock out single action potentials.
However good the work was in dishes, the results said little about real world biology. What about whole tissues? Whole organisms? A number of experiments investigated neural activity within increasingly sophisticated neurological backgrounds.
The first experiments involved transfecting neural tissues with specially engineered lentiviruses that carry these ion channel genes. The viruses are exposed to neural tissue, where they set up a productive infection cycle. This includes the insertion of the gene into the resident chromosomes and expression of their proteins. One such experiment infected the brains of mouse pups and the hippocampi of adult mice with viruses loaded with both ion channel genes. Subsequent assessment of the acute activity of cortical and hippocampal slices showed that neurological activity could be manipulated by simply flashing blue light (excitatory) or yellow light (inhibitory). The tissues turned on or off, depending on the wavelength deployed. This was the first evidence that complex neurological tissues could be regulated remotely with great precision and in a bidirectional manner.
The next experiments involved examining intact animals. The organism used was Caenorhabditis elegans, a tiny roundworm that is the subject of a great deal of genetic research (it has been particularly useful in elucidating the so-called genes of aging). It is possible to insert genes into specific tissues of this organism and assess changes in behavior. Such work was done on the roundworm using these ion channel genes. The creatures were specifically engineered to express the proteins in the body wall muscle cells that mediate the wiggling, swimming movement of the worm, or in the cholinergic neurons that control these muscles. The results were stunning. By flashing blue or yellow light onto the creature, they could control the wriggling, swimming behavior of the worm!
This is quite a technical achievement with plenty of research and clinical applications. One of the greatest potential uses of this technology is the elucidation of neuronal pathways involved in typical behaviors. Plans are under way to trace circuits involved in tactile sensations. Using animal models of human disease, other researchers are working to identify circuits relevant to depression, Parkinson disease, and epilepsy.
There are many potentially powerful applications of this technology for humans. It is possible that these switches could be used to coax neurons in the retinas of blind patients with no functioning rods or cones to fire in the presence of blue or yellow light, perhaps restoring aspects of vision.
There may even be a psychiatric application. The suitability of deep brain stimulation for patients with severe depression has been studied by researchers such as Helen S. Mayberg and colleagues for years.1 In this experimental protocol, electrodes are inserted deep in the brain in an attempt to stimulate certain neural circuits. Working something like localized electroconvulsive therapy, the hope is that activation of these circuits could successfully treat these patients. It is very possible that such photosensitive proteins might some day replace electrodes currently in use for deep brain stimulation investigations. This might allow the clinician to target only those neurons relevant to the disease. There is a large barrier to overcome before any of this could be done in humans, of course, including issues such as the insertion of these proteins into a living brain.
Still, what biologists have in hand is no less than the wish list for the alien brain mentioned at the beginning of this column. For the first time, there is the ability to turn neural circuits on and off at will and to do so from a remote location. It is an extraordinary collaboration between the behavioral, physiological, and molecular sciences. No doubt your boss at the electronics laboratory would be proud.
Mayberg HS, Lozano AM, Voon V, et al. Deep brain stimulation for treatment-resistant depression.