Advancing Basic Science for Humanity
Creating a Perfect Brain Storm
WITH SOME 86 BILLION NEURONS forming hundreds of trillions of synaptic connections between them, the human brain is often said to be the most complex object in the known universe. Understanding how it develops, how it generates thoughts and complex behaviors, and what happens when it goes wrong is, therefore, humankind’s greatest intellectual challenge.
In recent years, there have been unprecedented advances in brain research, with the development of sophisticated new techniques for analyzing brain structure and function. This has led to an ever-growing mountain of information about the brain, yet there is still no unifying theory to tie these data together, and fundamental questions, such as how many different kinds of brain cells there are, remain unanswered.
The past few years have seen the launch of several large-scale, big budget collaborative projects around the world, which aim to accelerate our understanding of the brain through the development of new technologies. With deeper knowledge of the brain comes the promise of alleviating the huge personal, social and economic burdens associated with brain diseases, such as dementia, and mental health disorders, such as schizophrenia and depression.
In 2013, Europe and the United States spearheaded these efforts with the launch of the Human Brain Project and the U.S. Brain Research through Advancing Innovative Neurotechnologies (BRAIN) Initiative. Similar projects have been launched or are on the horizon in other regions of the world, most notably in East Asia, where a Japanese initiative is under way, and China and South Korea are set to launch their own.
All of these large-scale neuroscience initiatives are planned to last a decade or more. They share the common goal of advancing our understanding of the healthy and diseased brain, and adopt overlapping approaches to achieve that. But they also differ—and hence complement each other—in important ways. Recognizing this, their leaders—from Washington, D.C. to Seoul—are keen to foster collaborations that could ultimately lead to the emergence of a concerted global initiative to break through this final frontier of human knowledge.
The brain is a highly dynamic structure, which continuously reorganizes itself in response to everyday experiences. Learning and memory, for example, are widely believed to involve the modification of synaptic connections in networks of neurons that are sparsely distributed throughout the brain.
The U.S. BRAIN Initiative aims to map the dynamic brain in action, with a five-year plan focused on the development of novel tools and methods that will help researchers determine how individual neurons interact within complex circuits over time, followed by five years of applying the new tools to examining the brains of various species, from fruit flies to humans. This should provide fresh insights into how the brain records, stores and retrieves huge quantities of information, and may eventually lead to new ways of treating brain disorders.
The BRAIN Initiative is funded by federal agencies such as the National Institutes of Health (NIH) and private organizations such as the Howard Hughes Medical Institute and The Kavli Foundation. What distinguishes it even more is the mix of scientists who are involved. “One of the most impressive things about [the BRAIN Initiative] is that many of the people who have received grants have never applied to the NIH before,” says Richard Huganir, who directs the Kavli Neuroscience Discovery Institute at Johns Hopkins University and serves on a multi-council working group that oversees the BRAIN Initiative. “A lot of them are engineers, physicists and chemists, so we drew into neuroscience a whole new population of scientists, and their achievements have already been amazing.”
Among these achievements are the development of new techniques for classifying different types of neurons; automated methods for creating wiring diagrams of the connections between cells in the mammalian brain; new ways of recording and manipulating the electrical activity of hundreds, even thousands, of cells simultaneously; and, advanced microscopes that allow researchers to visualize the activity of all the cells in the brains of small animals.
BRAIN Initiative researchers and their partners are also developing new ways of visualizing data that makes the information more easily accessible. For instance, researchers at the Allen Institute for Brain Science, launched by Microsoft cofounder and philanthropist Paul Allen in 2003, and a BRAIN Initiative partner, are generating vast data sets and using them to compile comprehensive atlases of cell types, connectivity and gene expression patterns in the brains of mice and humans. Furthermore, they are making their data and analytical tools freely available for anyone to use. In July 2016, they released the Allen Brain Observatory, which will eventually survey the properties of more than 18,000 neurons in the visual region of the mouse brain. The ultimate aim is to use this information to explore how the networks of neurons process sensory information.
Whereas the U.S. BRAIN Initiative is focused on developing innovative tools for brain research, the Human Brain Project (HBP) aims instead to develop next-generation supercomputers to further our understanding of the brain. The HBP is one of two European Commission's Future and Emerging Technologies flagship programs, which were launched to promote interdisciplinary research into information and communications technology. Its original aim was to develop neuromorphic—or 'brain-inspired'—supercomputers that could run simulations of the human brain, using diverse data sets collected by HBP researchers and others.
In collaboration with IBM, HBP scientists have developed neuromorphic chips that are more energy-efficient than those used in existing computing systems, and have greater storage capacity because they are designed to mimic plasticity, the process by which synaptic connections are modified over time in response to neuronal activity and experience. The researchers have used these chips to create one of the most comprehensive brain simulations to date—a digital reconstruction of some 31,000 neurons, and 37 million synaptic connections, in a sand grain-sized chunk of the rat brain.
As new information is incorporated, such simulations should become more accurate and could be used to run virtual experiments and test predictions about brain function. Ultimately, the hope is that scaling up the simulations toward the whole-brain level will provide fresh insights into the workings of this mysterious organ and, possibly, lead to something like a grand unified theory of brain function.
Like other large-scale neuroscience initiatives, the HBP places great emphasis on collaboration, and released the 'Collaboratory,' a public portal to the various tools developed by HBP researchers, earlier this year. “We would like to take advantage of the expertise and data from different labs in order to integrate it and provide a platform to share and exchange data and tools,” says Katrin Amunts, Scientific Research Director of the HBP and a neuroscientist at the University of Düsseldorf.
In 2014, Japan launched a 10-year initiative called Brain Mapping by Integrated Neurotechnologies for Disease Studies (Brain/MINDS), with funding from the Japan Agency for Medical Research and Development. The project is unique in that it is largely focused on mapping the brain of a small New World monkey called the common marmoset. Marmosets are not native to Japan, but over the past decade scientists there have established a large research colony. The marmoset brain is a lot like the human brain in the way it develops and functions. Marmosets also exhibit social and maternal behaviors that are similar to those of humans; and, their vocal communication and learning abilities could provide clues into the evolution of language and how the human brain produces speech.
“Mapping the brain of the common marmoset would represent an important step toward understanding the human brain,” says Hideyuki Okano of Keio University, who leads a Brain/MINDS sub-project devoted to mapping the marmoset brain “and toward developing knowledge-based strategies for the diagnosis and treatment of human psychiatric and neurological disorders.”
Because marmosets are more closely related to humans than mice and other organisms more commonly used in research, they can also be used to generate more realistic models of human brain diseases. Working toward this goal, Okano and his colleagues created the world’s first genetically engineered non-human primates. “We can use transgenic marmosets to visualize particular neuronal pathways, and we have generated a model of Parkinson's Disease,” says Okano. These monkeys exhibit brain changes similar to those seen in humans with Parkinson’s, as well as some of the behavioral symptoms of the disease.
Okano’s group has also succeeded in using a state-of-the-art technique called genome editing in marmosets, which could be used to generate 'knock-out' monkeys, in which specific genes are deleted or inactivated in specified brain regions, and at a particular stage of brain development.
Other researchers involved in the Brain/MINDS project have produced a three-dimensional atlas of the marmoset brain, and a database of the expression patterns of more than 400 genes within it. This information could provide insights into the workings of both the marmoset and the human brain. It is available for other researchers to use. On the technological side, another research group has developed a new way to make biological samples, including brain tissue, transparent while preserving their integrity, which allows for detailed visualization of the structures within them.
Several other East Asian countries recently announced that they too are set to launch their own national large-scale neuroscience initiatives.
Compared with mice, the brains of non-human primates such as marmosets and macaque monkeys are far more similar to the human brain, making them a valuable research tool. (Credit: Martijn van den Heuvel/Dutch Connectome Lab, Utrecht University)
Last year, following two years of negotiations with the country’s leading researchers about what direction its scientific research should take, the Chinese government announced the China Brain Project. Though the details are still to be formalized, Mu-ming Poo, director of the Institute of Neuroscience at the Shanghai Institutes for Biological Sciences and one of the project leaders, says, “The China Brain Project is designed as one body with two arms, with the body being made up of basic scientific research.”
One arm of the project will focus on translating basic research into treatments for disease; the other, which will begin in a later phase, will involve parallel initiatives to develop brain-machine interfaces, brain-inspired computers and intelligent robots. “The next generation of intelligent technologies will depend on brain science, and we want to use artificial intelligence to develop new algorithms, computing methods and neuromorphic chips,” says Poo.
Both arms will eventually converge into a single project to create reliable diagnostic tools by 2030 that can detect brain diseases early, he says.
The Chinese initiative will make two unique contributions to neuroscience. It will use non-human primates to develop models of human diseases. But whereas the Japanese Brain/MINDS project is focused on studying the marmoset brain, the China Brain Project will concentrate instead on the rhesus macaque, an Old World monkey species, which resembles humans even more closely.
Chinese scientists are now establishing macaque colonies for research purposes, and earlier this year, they reported the creation of genetically engineered macaques carrying mutations associated with a rare developmental disorder called Rett Syndrome. These animals recapitulate most of the cognitive and physical problems that arise in children with the disorder, and so may help researchers learn more about its biology and how to treat it.
The other unique contribution rests on the country’s enormous population. With more than 1.35 billion people, China is the most populous country in the world, and as such it also has the world’s largest population of patients with neurological and psychiatric diseases. Combining genetic information and medical records could yield an enormous amount of useful information about the genetic risk factors for these diseases. A similar but dramatically smaller effort in Iceland has already led to the identification of a number of gene variants, including one that reduces the risk of Alzheimer’s Disease. “The China Brain Project hopes to introduce a national team to take advantage of this patient data,” Poo explains. “But these data are still very fragmented, so we really will need a national effort with standards for data collection and sharing.”
In May 2016, South Korea announced plans to launch its own large-scale neuroscience initiative. The project is currently undergoing feasibility studies and technological assessment; if these are successful, the South Korean government will launch the 10-year initiative in 2018.
The Korea Brain Initiative will be two-pronged: It will simultaneously initiate research and development projects and reinforce the neuroscience community both nationally and internationally, by offering expanded multidisciplinary degrees, co-operative training programs and collaborations between high-profile researchers.
“We hope to gain a comprehensive understanding of the connectivity and synaptic plasticity mechanisms underlying higher brain functions,” says neuroscientist Kyungjin Kim, president of the Korea Brain Research Institute. “We also aim to develop innovative neurotools and devices and personalized treatments for neurological disorders, and to create a sustainable neuroscience infrastructure through global cooperation.”
One of the proposed research projects aims to construct brain maps at multiple scales of organization—from the connections between individual cells to local neuronal circuits to long-range tracts from one brain area to another—and will also optimize existing brain mapping techniques and develop new ones. The researchers hope that this will result in two specialized brain maps by 2023—one for the healthy brain, another charting changes to the organ as it ages.
A second project is geared toward applying a wide range of different technologies, particularly brain-machine interfaces for controlling robotic prostheses, ‘mini-brains’ grown from stem cells, and artificial intelligence and machine learning strategies, to the treatment of brain disorders and injuries.
Other countries are following suit. For example, Brain Canada is a national non-profit organization that supports collaborative multidisciplinary brain research, with a focus on brain diseases, disorders and injuries; another non-profit organization, Israel Brain Technologies, aims to turn Israel into a global neurotechnology research hub by commercializing products that emerge from the country’s basic and clinical research. The Australian Brain Alliance, formed earlier this year, is a network comprising more than 20 research institutions and professional organizations advocating for a national brain research initiative. And New Zealand has also announced that it will launch a national brain research initiative.
Global Brain Projects: By the numbers
Human Brain Project
Computational techniques are helping neuroscientists search for meaningful patterns in neuroscience data. (Credit: Jeremy Freeman, Nikita Vladimirov, Takashi Kawashima, Yu Mu, Nicholas Sofroniew, Davis Bennett, Joshua Rosen, Chao-Tsung Yang, Loren Looger, Philipp Keller, Misha Ahrens/HHMI)
Technological advances have led to a dramatic increase in the rate of data production, and all of these brain research projects face the enormous challenge of how to deal with the vast amounts of data they are generating. For example, experiments involving sophisticated whole-brain imaging techniques in small animals can generate up to 10 terabytes per hour, and researchers at the Allen Institute for Brain Science, who released an atlas of all 21,000 genes in the mouse brain, produced more than 1 petabyte of data in 2015.
In terms of brain simulation, the challenges are even bigger: The world’s fourth most powerful supercomputer took 40 minutes to model one second of brain activity in a simulated network containing 1.73 billion neurons and over a trillion synapses, yet this represents just one per cent of neuronal networks in the human brain. Even if a simulation of the whole human brain were possible today, there simply is not enough computing power in the world to run it.
What’s more, different labs use different methods for collecting data, and there is no standardized format for storing the information, or for naming neural structures and components.
This massive data overload has led to the field of neuroinformatics, which grew out of the necessity to find better ways of storing, organizing and analyzing these ever-growing mountains of information, and to integrate different kinds of information about the brain into something coherent.
“This is a big problem for brain science in general,” says Huganir of the U.S. BRAIN Initiative. “We’re getting so much data, we’ve got to store it all somewhere, as well as analyze it, so there’s a great push to share data and put these in databases in the cloud.”
Neuroinformatics and data sharing are, therefore, at the center of all of these large-scale neuroscience initiatives. Some, most notably the HBP, are also dedicated to developing high-performance computing platforms to help deal with the problems of data storage and analysis.
What’s lacking, however, are standardized procedures for storing and organizing these large data sets, to ensure that they are compatible with each other and can be shared and used as widely as possible. Several organizations have been attempting to address this.
The International Neuroinformatics Coordinating Facility (INCF) brings together leading scientists from around the world in order to coordinate their activities and foster data integration. The Neuroscience Information Network (NIF) is an inventory of online neuroscience data, materials, tools and other resources developed to advance brain research by making research data and tools publicly accessible on an open-source platform.
More recently, The Kavli Foundation launched Neurodata Without Borders, a consortium of researchers who share the common goal of standardizing neuroscience on an international scale. The project has established a unified data format for cellular neurophysiology and will soon be expanding to include other types of data.
Data standardization would be a prerequisite for a successful global neuroscience initiative. International collaborations could contribute to fixing this problem by encouraging the sharing of standardized data between researchers in different countries and regions of the world.
An international neuroscience initiative
Given the individual strengths and complementary aspects of these large-scale projects, their leaders are keen to foster such collaborations. Indeed, promoting global cooperation and international consortia of neuroscientists is one of the core aims of the Korea Brain Project, and the leaders of the South Korean, Chinese and Japanese initiatives are already discussing a trilateral East Asian brain research consortium.
“Macaques and marmosets have different behavioral characteristics, and each one has its advantages and disadvantages,” says Okano of Japan’s Brain/MINDS. “It's important for a worldwide brain initiative to work on different non-human primate models. In that sense, Japan and China are using complementary approaches, but we can share technical developments.”
Poo concurs. “The China Brain Project will have a lot of interactions with Japan because of the primate research,” he says. “We have already visited their marmoset group, and they’ve visited us, and we’ve exchanged a lot of our technologies.”
Poo foresees East Asia becoming a world hub for primate neuroscience in years to come. “I believe that rodents will have exhausted their usefulness for basic brain research in the next 20 years, and people will have no choice but to turn to primates to understand higher brain function,” he says. “In the long run, when the primate centers are well established, international collaborations will flourish, because researchers from non-Asian countries will want to come and use our facilities.”
In fact, researchers from the U.S. are already collaborating with Chinese institutions to make use of their macaque colonies, and the leaders of the Brain/MINDS project are working with researchers at the Allen Institute for Brain Science to try to solve the big data problems associated with brain mapping. “We are going to make our data open for use by researchers across the world,” says Okano, “and we want to make a standard format for data from all species.”
The leaders of all the projects also agree that the grand challenge of understanding the brain will require a concerted global effort. “We have to bring together all the data about different aspects of brain organization from different species,” says Amunts, of the HBP, “but that’s a huge challenge that is far beyond the capacities of a single lab, or even of a single country.”
—Moheb Costandi, Summer 2016