Researchers have traced the most comprehensive wiring diagram of the mammalian brain so far. The mouse brain “connectome,” charted by a large team of scientists led by Hongkui Zeng at the Allen Institute for Brain Science in Seattle, tracks the thousands of neuronal circuits that wire the brain. The researchers described the map and initial findings in the April 2 Nature, and the connectome is freely available online for Google Earth-style exploration. Researchers can zoom in to the cellular level and quantify the strength of connections across the brain, a feat that until now had only been achieved in the roundworm C. elegans. The work provides a “quantitative baseline” of mouse-brain circuitry, a tool that will aid researchers in understanding how connectivity goes haywire in brain disorders such as Alzheimer’s disease, Zeng told Alzforum.
“This work is a beautiful example of what neuroscience ought to be striving for: making data widely available, well-organized, and accessible,” said Marcus Raichle of Washington University in St. Louis, who applauded the effort along with other researchers. “It’s a very important contribution, and not just a flash in the pan.”
The connectome will likely make important contributions to the study of neurodegenerative disease, said Ashish Raj of Weill Cornell Medical College in New York. “All of these diseases, including Alzheimer’s and dementia, happen in neural networks,” not in isolation, Raj said. “Considering the connectome is one of the most important things people can do in neurology.”
The mouse connectome makes its debut in a climate of diverse efforts to map the anatomy, circuitry, and genetic architecture of the brain (see Feb 2013 news story). “It fits into the current quest that’s going on in many different places to understand brain connectivity patterns,” said Olaf Sporns of Indiana University in Bloomington. One example of this is the ongoing Human Connectome Project, funded by the National Institutes of Health, which uses live imaging to map the connectomes of 1,200 adult brains (see Oct 2012 conference coverage). The advantage of noninvasive imaging is that it can be conducted in humans and used to compare differences in brain wiring between healthy people and those with brain disorders. However, such maps are limited to resolving only large conduits of axons that flow between brain regions. In such connectomes, “you don’t really see how the axons branch out when they go into gray matter, and you have no idea about the strength of connections between neurons,” Zeng said. Being able to see where axonal projections start and end, and to quantify the strength of those connections, will be crucial to identifying early malfunctions that may precede the onset of disease, she added.
To construct a quantitative map of neural circuitry throughout the entire mouse brain, co-first authors Seung Wook Oh, Julie Harris, Lydia Ng, and colleagues injected animals with adeno-associated virus (AAV) expressing green fluorescent protein (GFP). The infected neurons near each injection site then expressed GFP in the cytoplasm, which allowed the researchers to trace neuronal projections. The researchers injected 469 mice, each in a different, carefully chosen brain coordinate. Together, the coordinates spanned 295 regions of the brain. The researchers then removed the brains and subjected them to serial slicing coupled with two-photon tomography, a technique in which 140 sections of the brain were individually imaged just prior to being sliced off (see Apr 2012 news story on Ragan et al., 2012). The process took nearly 19 hours for each brain, Zeng said. “We used the imaging system all the time, day and night, six days a week,” she said. “It generated beautiful, high-resolution, and highly coaligned sections that allowed us to coregister every image into one common three-D space.”
The researchers entered the results from each mouse into a template brain structure they generated by averaging images from 1,231 additional mouse brains. The result? A comprehensive map of neuronal projections originating from regions throughout the right side of the brain and extending outward to next-door neighbors as well as to the outer reaches of the left hemisphere. The researchers then split the three-dimensional map into thousands of voxels and measured the fluorescence in each one. Using this technique, they quantified the projection strength—a function of the number of axons—linking one region of the brain to another.
The quantification allowed Oh and colleagues to make several striking observations. The first one speaks to the power of the quantification itself: Connection strengths spanned five orders of magnitude throughout the brain. Notably, the majority fell on the weaker side of the spectrum, with relatively few strong connections. “To me, that is one of the core findings of this article,” said Sporns. “It raises the interesting question as to the relative functional contributions of the stronger versus the weaker connections.”
The map also revealed a surprising number of bilateral connections, crossing from the right hemisphere of the brain to the left, Zeng said, although the average strength of the same-side connections was more than four times greater than that of the connections venturing across the hemispheric divide. “That there are many bilateral connections throughout the brain indicates that there is so much coordination between the two sides,” she said. The bilateral connections also tended to match same-side connections. For example, the projections emanating from the right cortex tended to travel to similar areas on both the right and left sides of the brain. The importance of such bilateral connections, Zeng said, is poorly understood.
An analysis of the network architecture in the connectome revealed both clusters (cliquey connections between nearby neurons) and hubs (neuronal centers that connect to many areas of the brain). Interestingly, Zeng said, the combination of clusters and hubs did not conform entirely to existing network theories, which are based primarily on networks within the cortex rather than the whole brain (see Nov 2011 news story). “Subcortical-to-cortical connections are more heterogenous, divided into vastly different systems,” Zeng said. “It’s actually going to be a bigger challenge to understand hubs and subnetworks in the subcortical system.”
The mouse connectome provides a data-rich baseline model that researchers can use to understand healthy brain function, but Zeng said researchers at the Allen Institute already have their sights set on mapping connectomes in mouse models of developmental and neurodegenerative disease. The group is in the process of securing funding to build brain connectomes from mouse models of Alzheimer’s disease, she said: “We’d like to look for connectional changes that may occur even before plaques appear.” Zeng added that such changes are likely to occur in specific pathways or circuits, and to be quantitative in nature, making the researchers’ techniques uniquely applicable.
The researchers have already started rolling out a second wave of data on their website that targets specific types of neurons. Using different transgenic mice that switch on the virally delivered GFP in subsets of neurons, the researchers are teasing out the relative contributions of different cell types in the connectome.—Jessica Shugart
- Proposed Brain Activity Map Seeks to Crack Neuronal Code
- The Human Connectome Project: Deciphering the Brain’s Wiring
- Modern Microscopy Plumbs the Depths of Brain Tissue
- Cortical Hubs Form "Rich Club" in Human Brain
- Ragan T, Kadiri LR, Venkataraju KU, Bahlmann K, Sutin J, Taranda J, Arganda-Carreras I, Kim Y, Seung HS, Osten P. Serial two-photon tomography for automated ex vivo mouse brain imaging. Nat Methods. 2012 Mar;9(3):255-8. PubMed.
No Available Further Reading
- Oh SW, Harris JA, Ng L, Winslow B, Cain N, Mihalas S, Wang Q, Lau C, Kuan L, Henry AM, Mortrud MT, Ouellette B, Nguyen TN, Sorensen SA, Slaughterbeck CR, Wakeman W, Li Y, Feng D, Ho A, Nicholas E, Hirokawa KE, Bohn P, Joines KM, Peng H, Hawrylycz MJ, Phillips JW, Hohmann JG, Wohnoutka P, Gerfen CR, Koch C, Bernard A, Dang C, Jones AR, Zeng H. A mesoscale connectome of the mouse brain. Nature. 2014 Apr 10;508(7495):207-14. Epub 2014 Apr 2 PubMed.