Brave new brain science
Image: Professor Ed Boyden's unconventional imaging tools have changed the face of neuroscience.
From switching neurons on and off with light to swelling brains for imaging with nanoscale precision, Professor Ed Boyden's unconventional breakthroughs have changed the face of neuroscience.
Now heading up the Synthetic Neurobiology Group at the MIT Media Lab, the researcher is applying his eclectic brand of brain science to build tools to revolutionise biology and medicine. His methods are already being used worldwide to repair retinas, develop neural-control prosthetics and map parts of the human brain. But the road to radical research has been far from straightforward.
Like his research, Boyden's early education didn't follow convention. At age 14, the young student went to the University of North Texas to study chemistry.
As he tells Microscopy and Analysis: "I was a philosophical student and was working with a research group that was trying to create the building blocks of life from scratch."
"Our goal was to create molecules such as DNA," he says. "It was fun, I learned a lot but of course it didn't work or you'd have heard about it by now."
Two years later, in 1995, Boyden transferred to MIT to study Physics as well as Electrical Engineering and Computer Science. By 1999, he had graduated with degrees in both, as well as a Masters in Electrical Engineering and Computer Science working on quantum computation, and crucially had also discovered a passion for building tools.
In optogenetics, light-sensitive ion channels and pumps, known as microbial opsins, are genetically targeted to specific cells, so that upon light delivery, just those cells will be electrically activated or silenced. [Ed Boyden/MIT]
As part of his graduate research, the young Boyden had worked with Professor Neil Gershenfeld from MIT, as well as Professors Michale Fee and Sebastian Seung from Bell Labs, all renowned for a 'free-range' approach to research.
As such, Boyden developed computational models of how birds sing, engineered control software for an autonomous submarine, programmed machine-learning tools for reconstructing digital violin dynamics, and more.
Given this, as well as his interdisciplinary degrees, Boyden figured he was ready to study the brain. "I felt perfectly poised to bring these points of view into neuroscience and to really think like an inventor," he says. So he did.
Road to Optogenetics
In 1999, Boyden embarked on a PhD at Stanford University. Working for neurobiologists, Professors Jennifer Raymond and Richard Tsien, he studied how neural circuits use plasticity to store memories.
However, at the same time, he also started to wonder how technologies could be used to control the electrical activity of different neurons in the brain. A few late night conversations with his colleague, Dr Karl Deisseroth, led to the idea that light could be harnessed to control a single brain cell.
"We thought about magnetic fields, mechanical force, light and all sorts of ways to deliver energy into the brain," he explains. "But in the end we picked light as it is fast and can be focused."
At the time, other researchers were experimenting with light, using laser pulses for example, to activate many neurons. However, Boyden and Deisseroth wanted to switch specific neurons on or off as part of a grand plan to determine which type of neuron was responsible for a certain character trait, brain disorder or even disease. But of course, the first, critical question was how?
Help first came in 2002, when Professor Gero Miesenbock and colleagues from Sloan-Kettering Cancer Center in New York City, showed that genetically-sensitised Drosophila neurons could be driven by light.
Then, in 2003, Professor Georg Nagel, University of Würzburg, Germany, discovered a light-sensitive protein from green algae - Channelrhodopsin-2 - and showed that the protein could be used to depolarise mammal cells.
Excited by these developments, Deisseroth and Boyden obtained a clone of the protein from Nagel, and engineered cultured neurons that could express Channelrhodopsin-2. Then, late one night in 2004, Boyden placed a dish of the cultured neurons under an epifluorescent microscope, patch clamped one of the glowing neurons and pulsed blue light at the samples.
Immediately the patched neuron fired electrical signals, Boyden had demonstrated neurons could be activated with light, and suddenly, this fledgling field of neuroscience - now known as optogenetics - was ready to explode.
Blue light activates a single neuron. [Ed Boyden/MIT]
Since this time, Boyden and colleagues have developed strategies to 'silence' the cells and also used different proteins to control neurons in live mammal models. Key breakthroughs, primarily achieved working alongside experts on specific diseases,, have included restoring vision in mice and suppressing beta amyloid plaque production - the hallmark of Alzheimer’s disease - in mice brains.
Crucially, along the way, Boyden has also shared his methods with thousands of other research groups in academia and industry, and today, several start-ups are also pursuing optogenetics research in humans.
"The field is now mainstream, and there must be hundreds of papers published in which, say, researchers turn-off certain neurons in a mouse brain to try to shut down epileptic seizures or work out how emotional feelings arise," he says.
"A Caltech group used our tools to photo-activate neurons deep in the brains of mice, so the mice would become aggressive," he adds. "Such research is helping people realise the power of these tools in finding brain circuits that implement complex functions that you probably couldn't probe in any other way."
But while optogenetics progress has been rapid, for Boyden, it is just one pioneering method in his growing neuroscience toolbox. Completing his PhD in late 2005, the researcher moved to MIT Media Lab in 2006, to set up his own neuroengineering research laboratory, the Synthetic Neurobiology Group.
As he puts it: "The MIT Media Lab is MIT's home for misfits and has given us time to grow on our own outside the confines of traditional disciplines."
And unconventional imaging tools have followed. For example, in the Summer of 2014, Boyden and collaborator Alipasha Vaziri revealed a system based on light-field microscopy to generate 3D videos of the entire brains of C. Elegans nematodes and zebrafish larva at millisecond timescales.
A microlens array is inserted into the optical train of an epifluorescence microscope so that sensor pixels capture and recombine the 2D location and 2D angle of incident light to recreate the 3D structure.
However, in January 2015, Boyden revealed what is arguably his biggest breakthrough since Channelrhodopsin-2-based optogenetics.
So-called expansion microscopy uses an expandable polymer to swell tissue to around four and a half times its usual size, so that nanoscale structures appear within focus on an ordinary confocal microscope.
3D super-resolution microscopy of mouse brain tissue: volume rendering of a portion of hippocampus showing neurons (green) and synapses. [Fei Chen, Paul Tillberg, and Ed Boyden, MIT, Research published in Science January 2015, Vol 347, Issue 6221]
According to Boyden, his team first explored concepts similar to expansion microscopy in 2007 as a means to separate and more easily tag proteins in cells. However, they quickly shelved the idea as PALM, STORM and other super-resolution microscopy methods were emerging around the same time.
Still, come 2012, Boyden's graduate students Fei Chen and Paul Tillberg were struggling to use STORM as well as electron microscopy to map the brain. The old idea was unearthed.
"Expansion microscopy in 2012 emerged as a bit of a joke borne out of frustration as we asked ourselves, 'why can’t we just make the darn thing bigger?'," explains Boyden. "Performing multiplexed biomolecular mapping to look at how biomolecules are organized was appearing really difficult using STORM and electron microscopy, so we thought, why don't we just expand these samples."
They went onto infuse sodium acrylate into chemically fixed brain tissue, adding polymerisation agents to form a tissue-polymer network composite within the sample.
Fluorescent labels were covalently anchored to the polymer at biomolecular sites within the tissue-polymer sample, which was then treated with protease to homogenise its mechanical properties and immersed in water to trigger a 4.5 times linear expansion.
According to Boyden, labels spaced closer than the optical diffraction limit were separated and he and colleagues used the method to image human embryonic kidney cells and mouse brain tissue slices with 70 nm lateral effective resolution
"Embedding tissues in polymers has a very long history... but with our expansion method we can begin to image in 3D with nanoscale precision across large areas," says Boyden. "My hope is that we can use this tool to get full maps of the brain with molecular information, and maybe someday be able to simulate what is happening when a brain circuit generates a thought or a feeling."
Expansion microscopy of Brainbow Hippocampus: simple nanoscale resolution imaging on fast diffraction-limited microscopes. [Yosuke Bando, Fei Chen, Dawen Cai, Young Gyu Yoon, and Ed Boyden, Nature Biotechnology]
Since its development, the researchers have honed the method to rely on off-the-shelf chemicals, rather than the original custom-designed chemical tags. What's more, in the last year, and in a similar vein to their optogenetics tools, they have shared the method with more than 100 research groups.
"Researchers are using this tool to figure out where pathogens are hiding in tissues, as well as how cancers differ from normal tissues," highlights Boyden.
"We have no idea what's going on in so many diseases," he adds. "But if we can use tools such as optogenetics and expansion microscopy to figure out what's important, perhaps we can develop better therapies."
Boyden and colleagues now intend to combine optogenetic tools with expansion microscopy; the results look set to be mind-boggling.
As Boyden points out, he and colleagues could use optogenetics to activate certain neurons and, in turn, observe how these neurons control other neurons, then map the brain circuit using expansion microscopy.
"We could create computational models of brain circuit," he says. "We might be able to use these to predict what a brain circuit will do during a certain behaviour or disease."
So as research continues apace, what advice would Boyden give to the aspiring researchers of the future? In his words: "Every time somebody gives you advice, consider doing the exact opposite."