Advances in Science come in equal parts from theoretical insights, novel data and technological breakthroughs. The power of technological advances lies in bringing to light phenomena that were previously unobservable, revealing just how much we were missing out on. Adam Packer did not invent a new technique. What he succeeded in doing, however, was no small feat: bringing together arguably the two most exciting technical developments in Neuroscience in the decade – Optogenetics and 2-photon calcium imaging.
Optogenetics encompasses a range of approaches enabling neuroscientists to control brain activity using light. Methods for stimulating neurons precisely are crucial for Neuroscience’s goal of charting brain connectivity because, to a large extent, this is done by stimulating a brain structure and recording downstream from it in search of a response. Previously, scientists relied on electrodes to stimulate neurons. However, in most brain structures very heterogeneous types of neuron co-exist side by side, differing in morphology, gene expression and connection targets. Electrical stimulation of a structure thus evokes activity in a variety of cells, and if a response is recorded downstream it can be challenging to trace its origin with precision.
Optogenetics provided a solution to this problem in the shape of Channelrhodopsins, a family of light-sensitive, pore-forming proteins. One such protein is called ChR2. When ChR2 absorbs blue light, it changes shape and temporarily opens a pore into the cell, enabling various ions to pass through. ChR2 is found naturally in unicellular green algae, which use it to detect and orient towards light. Microbiologists have long known about channelrhodopsins and other similar light-gated channels, but this entire field of research was unknown to most neuroscientists for decades. In the early 2000s, Ed Boyden, Karl Deisseroth, and Gero Miesenboeck, were creative enough to consider the far-fetched: ChR2s are permeable to the same ions that mediate neuronal action potentials; could we genetically engineer these algal light-sensitive ion channels into mammalian neurons and use light to stimulate them?
The answer turned out to be yes, and optogenetics took the world of neuroscience by storm. Because ChR2 is a naturally occurring protein whose instruction set is coded in DNA, scientists could now use a whole range of available genetic engineering tools to determine where and when to express it. Neurons can be targeted for ChR2 expression based on their physical location, genetic properties, output target, recent activity, or even a combination of these properties. This means we can now be exquisitely precise in defining which neurons to stimulate. This precision applies also to optical stimulation itself, and a key advance has been the application of two-photon excitation, a principle of optical physics that allows a laser to precisely excite a point in 3D space, avoiding positions above or below that point. Adam Packer worked on the application of this principle to the stimulation of single neurons during his PhD with Rafael Yuste, at Columbia University. Using the knowledge he gained there, he showed us how he and others are now able to use two-photon excitation to stimulate single ChR2-expressing neurons. Building on this principle, now working at Michael Hausser’s lab in UCL, Adam has been able to engineer sophisticated microscope optics involving holograms, which allow him to simultaneously stimulate up to 50 specific and discretely-located neurons, including cells located at different depths.
A method for stimulating neurons would be useless without a way to record the products of this activity, bringing us to the second half of Adam’s approach. For decades in neuroscience, electrical readings of activity using microelectrodes have been mainstay. The gold standard in resolution is the whole-cell patch-clamp technique, which allows researchers to record voltage or ionic current flow in single neurons, at sub-millisecond timescale. However, the patch-clamp recording technique can only be applied to one neuron at a time. At best, several electrodes can be used to record from 8-16 neurons simultaneously. Such numbers are well below the estimated 71 million neurons in the mouse brain. How then do we zoom out from the tree and look at the forest? Is it possible to record the individual activity of dozens or hundreds of neurons simultaneously?
In a sense, it is. Spiking activity in neurons is accompanied by large and reliable increases in intracellular calcium concentration. Neuroscientists now have at their disposal Genetically-Encoded Calcium Indicators (GECIs), proteins that change brightness whenever they bind calcium. During baseline conditions, fluorescence is low; but when a neuron spikes, intracellular calcium concentration rises 10-100 times, calcium binds to the GECI, and the GECI protein fluoresces and emits photons. Powerful detectors can catch these photons and provide actual images of neural activity, where spiking cells appear brighter than silent ones. Current methods allow the visualization of activity in dozens to hundreds of mouse neurons, almost in real-time.
By combining 2-photon optogenetics and calcium imaging, Adam Packer has achieved an experimental setup where he can individually stimulate several neurons of a particular type and monitor the response produced in a target population of cells. This approach will be key to understanding how coordinated groups of cells can work together to encode messages in the brain. Adam is particularly interested in perception. His aim in neuroscience is to figure out the minimal neuronal correlate of the perception of a sensory stimulus. In simpler terms, how many neurons does it take to form a percept?
Adam outlined an elegant experimental plan that he is currently undertaking in the lab of Michael Hausser, taking advantage of the mouse’s highly sensitive whisker system. Whenever a target whisker is touched, the mouse is trained to report this percept by licking a spout. If the report was correct and the whisker had indeed been touched, the mouse receives a reward. Concurrently, Adam will image the activity evoked in the barrel cortex, the area of the mouse brain where whisker sensation is processed, aiming to identify which neurons are reliably active whenever a particular whisker is stimulated. He will characterize in detail not only the identity of active cells, but also the pattern of activity: when they are active, and how active they are. This information will allow him to imitate the patterns of activity produced by sensory stimulation using optogenetics. In a sense, what Adam will do is place mice into The Matrix: he is attempting to bypass sensory stimulation entirely and instead use optogenetics to control activity in barrel cortex so precisely that it mimics the neuronal response produced by an actual sensory stimulus. Although no sensory stimulus is actually presented, the prediction is that the mouse will “feel” as though the whisker was touched and report this using a spout lick, in the way it was trained to do with actual whisker touches. By picking and stimulating decreasing numbers of neurons and observing the mouse’s reporting behavior, Adam will eventually find what is the smallest number of neurons necessary to stimulate for the mouse to interpret a “matrix” whisker touch as a “real” one.
This is the sort of experiment that could raise even more questions than it answers – one of the hallmarks of good science. Though its aim is to define the minimal neuronal correlate of a percept, it will also potentially shed light on the role of subcortical pathways in perception. The converse experiment – temporarily inhibiting cortical neurons responding to whisker touch – could also prove extremely informative, and Adam has the tools in place to perform such an experiment. These are exciting times to be a circuit neuroscientist, and I for one will be following updates from Adam Packer’s work closely. Throughout the talk, Adam was enthusiastic about the continuous state of development and improvement in 2-photon microscopy, repeating the mantra “Wider, Faster, Deeper, Stronger”. This reminded me of the lyrics to the Daft Punk song, “Harder, Better, Faster, Stronger”. Do you remember how that song’s lyrics end? I do:
“Our work is never over.”
Andre Marques Smith
To read the latest Adam Packer’s paper in Nature Methods, please click on the link below.