Researchers use rabies to draw brain circuit diagrams

The brain is an organ of staggering complexity, consisting of hundreds of billions of cells (and tens of thousands of different cell types) which form millions of specialized circuits that are organized into thousands of discrete areas. Neuroscientists have a number of methods for investigating brain circuitry and the connectivity of neurons within circuits. One of these involves exploiting the abilities of certain viruses, such as the herpes viruses, to target nerve cells; genetically manipulated viruses can be used to trace the synaptic connections between cells. This method has its limitations, however; the results obtained are ambiguous, because such viral tracers spread across strong connections more quickly than across weaker ones, and cannot be prevented from continuing to spread further than the cells being targeted.

Edward Callaway and his colleagues at the Salk Institute’s Systems Neurobiology Laboratory, together with collaborators from the Max von Pettenkofer Institute and Gene Center at Ludwig Maximilians University in Munich, Germany, have developed a transsynaptic tracer which enables them to examine neuronal connectivity far more accurately than existing methods. The tracer, which is based on the rabies virus, can be targeted to individual neurons, and spreads only to those cells which form direct synaptic connections with it. The work is reported in the current issue of Neuron.

Rabies was chosen because, like herpes viruses, it infects neurons in the peripheral nervous system and is then transported along the nerve fibres to enter the central nervous system. It is less damaging to neurons, but infects them far more efficiently, than herpes viruses, which also target peripheral neurons and which are commonly used for tracing neuronal connections. The intact rabies virus is non-specific in its infection of peripheral neurons, and expresses a surface glycoprotein that is required for it to spread from one cell to another.

In order to use it to infect specified neurons, Gallaway’s group constructed a modified version of the rabies virus. The construct contained a deletion mutation – the surface glycoprotein was removed, and replaced with the gene encoding green fluorescent protein (GFP). The gene encoding an envelope protein (EnvA) from the avian sarcoma and leukosis virus (ASLV-A) was also inserted into the construct. Target cells in slices of brain from newborn rats were transfected with three genes – the gene encoding an avian receptor protein called TVA (an interaction between EnvA and the TVA receptor occurs during ASLV-A infection of an avian cell), the gene encoding the glycoprotein normally used by the rabies virus to spread from one cell to another, and a gene encoding red fluorescent protein.

Thus, when the construct was introduced into the culture dishes containing the brain slices, it infected only the target cells – those expressing the TVA receptor, the EnvA protein and the red fluorescent protein. Infection with the modified virus caused them to emit a green fluorescence as well. Because the target cells contained the glycoprotein that was deleted from the rabies virus, the construct could spread from the target cells to those connected directly to them by synapses. And, because the modified virus contained GFP, the cells to which it spread began to emit a green fluorescence. But once inside these cells, the viral particles remained stranded there, because only the targeted cells, and not those to which the particles spread, contained the glycoprotein required for them to spread further.


Target cells are labelled red, target cells successfully transfected with the viral construct are labelled red/green and marked with a dashed line, and cells which form synapses with the latter, into which the construct has spread, are labelled green. (From Wickersham, et al, 2007; scale bar = 200 μm)

Upon examination, the brain slices were seen to contain large clusters of green fluorescent neurons surrounding individual red/ green fluorescent cells. This strongly suggested that the modified viruses had spread from the cells into which they had been transfected to cells connected to them. The researchers then used the patch-clamp technique to determine the specificity of this tracing, which was presumed to be transsynaptic. Microelectrodes were inserted into the presumed target cells, and the cells into which the virus was believed to have spread. A solution of positive ions was injected into the cells emitting green fluorescence, causing them to generate action potentials. Consequently, action potentials were recorded in the cells emitting red fluorescence, revealing that the tracer viruses had travelled retrogradely (backwards) and spread from the targeted cells to those connected to them presynaptically (that is, those which send action potentials to them).

This is the first time a transsynaptic tracer has been used to label all the cells connected to a target cell. But it remains unclear whether the viral particles spread to all the neurons which form synapses with the targeted cells. Given more time, the technique may have revealed even more connections that were not made apparent in these experiments. This, and similar techniques being developed by others, will enable researchers to gain a better understanding of neural circuitry. But even accurate tracing methods also have their limitations. Some brain circuits are so complex that one could not hope to accurately trace all the connections within them. In the cerebellum, for example, Purkinje cells form synapses with hundreds of thousands (or perhaps up to a million) parallel fibres. Attempting to visualize the connections of a Purkinje cell, even with the most accurate transsynaptic tracer, would result in an incomprehensible image consisting of a large green blur.


Wickersham, I. R., et al. (2007). Monosynaptic restriction of transsynaptic tracing from single, genetically targeted neurons. Neuron 53: 639-647. [Abstract]


3 thoughts on “Researchers use rabies to draw brain circuit diagrams

  1. Thanks for the succinct explanation of their methods, and the picture – it’s especially helpful for those of us without Neuron access. I wonder if they will ever be able to verify that this method does reach all the target neuron’s connections? I don’t see how they’d manage that. Regardless, a very useful addition to the neuroimaging toolkit.

  2. Thanks to MC for doing a nice job of describing our work.

    In response to Bioephemera, we do think that we will be able to test whether the method is capable of labeling all of the connections that are made onto a single neuron. The way we plan to do it is to express a protein which is transported to presynaptic nerve terminals (unlike the green fluorescent protein you see in the pictures) and can also be visualized with electron microscopy. We can then use the electron microscope to identify the synapses that are made onto the postsynaptic neuron (the red ones in the pictures). If every one of the neurons making connections to the red cell has been labeled then we should be able to see reaction product in all of the presynaptic terminals.

Comments are closed.