Painting Neural Circuitry With a Viral Brush: Are the Neighbors Green?

November 1, 2008
John J. Medina, PhD

Psychiatric Times, Psychiatric Times Vol 25 No 13, Volume 25, Issue 13

In last month’s column (“Painting Neural Circuitry With a Viral Brush,” Psychiatric Times, October 2008, page 16), I used Michelangelo’s famous fresco, “Hand of God Giving Life to Adam” on the ceiling of the Sistine Chapel as a metaphor to introduce a series of technologies that have allowed researchers to map the complex interactions of neural connections in continuously functioning neural tissues.

In last month’s column (“Painting Neural Circuitry With a Viral Brush,” Psychiatric Times, October 2008, page 16), I used Michelangelo’s famous fresco, “Hand of God Giving Life to Adam” on the ceiling of the Sistine Chapel as a metaphor to introduce a series of technologies that have allowed researchers to map the complex interactions of neural connections in continuously functioning neural tissues. This technology promises to deliver accurate synaptic associations-one finger to another-at a very high level of resolution

These extraordinary cartographic techniques involve exploiting the natural ability of the rabies virus to set up productive infections in neural tissues. For simplicity’s sake, we are examining the genetic manipulation of hypothetical “Neuron A” and its reaction to a previously engineered rabies virus. Although the manipulations to the virus are complex, using the data obtained, researchers seek to answer a simple, seemingly innocuous question: Are the neighbors green?

In case you do not have last month’s column handy, let me briefly review the life cycle of the rabies virus and reexamine the reengineered virus and Neuron A. We can then turn directly to the data.

Rabies virus

As mentioned last month, the rabies virus has several biological aspects that make it an ideal delivery device for working with living neural tissues. Once inside a nerve cell, the virus sets up a manufacturing site to create more viruses, like any typical virus. At maturity, however, these progenies jump to neighboring neurons, which allows the virus to spread along specific neural routes. This life cycle is handy if you are interested in synaptic connections throughout the body. The infection can start in the peripheral nervous system and then jump the stout molecular border that separates it from the CNS. (That is why a bite anywhere on the body can result in a catastrophic brain infection.) If one could find a way to follow the virus, one could identify the routes by which it travels.

Aspects of this jumping ability were exploited in the circuit-mapping experiments we are about to review. Both virus and cell had to be genetically manipulated in 3 different ways for the experiment to work.

Glycoprotein gene deletion. First, a mutation was engineered that deleted a viral glycoprotein gene. This mutation rendered the virus incapable of spreading the infection beyond the first encountered cell, thereby stalling infection. If the virus were introduced to Neuron A, it could reproduce itself in Neuron A’s cytosol but would have no means of escaping Neuron A.

Artificial addition of EnvA gene. Second, a gene encoding the envelope protein EnvA from the avian sarcoma and leukosis virus (ASLV-A) was inserted into the manipulated viral genome. This allowed the virus, upon infection of Neuron A, to create the EnvA protein on its surface.

Why is this important? Putting the EnvA protein on the viral surface allows the virus to bind to and then enter any cell carrying EnvA’s natural receptor-a protein called TVA. Conveniently, human cells do not naturally possess TVA. Therefore, if you want this virus to infect a human cell, you are going to have to supply that cell with TVA artificially.

That change in function could be of great value to a neurobiologist. Suppose Neuron A, sporting its foreign EnvA, is embedded in a thicket of normal neurons that are not engineered in such fashion. If the hobbled virus were exposed to the entire thicket, the only cell that would become infected is Neuron A, not the neighbors.

With a one-two punch, a virus has been created that needs external help to set up a productive infection. It then needs further help if its progenies are going to infect the neighbors. The result? A “defanged” viral particle whose direction of infection can be manipulated. All that is needed is one more addition: an onboard tracking agent (such as a colorant) that would allow visual inspection of viral progress as the infection advanced.

Another artificial addition. Addition of just such a colorant was the third manipulation. The gene encoding a green fluorescent protein was also stitched into the rabies virus genome, which caused infected cells to glow green. Researchers could then detect the presence of infection simply by looking for the green protein.

Nerve cell modification. Next, Neuron A was engineered to interact with the rabies virus from the third manipulation. As mentioned last month, we are describing events in just Neuron A; however, the actual manipulation involved a large number of newborn rat cell hippocampal slices. In this experiment, the task was to engineer some of the neurons in such a fashion that the connec-tions between the modified neurons and their nonmanipulated neighbors could be discerned.

To begin, Neuron A was given the gene for the TVA receptor from ASLV-A. Any externally supplied viral particle that carried the EnvA protein could now bind to Neuron A and set up a productive infection.

Neuron A was then given the gene encoding the rabies glycoprotein that had previously been deleted from the rabies virus. As a result, the cell’s interior became flooded with the protein. Why is that important? Remember that the defanged virus needs the glycoprotein to mount a productive infection and to spread beyond Neuron A. Giving this neuron the missing glycoprotein means that only Neuron A can act as the nursery for viral progenies.

Finally, the gene that encoded a red fluorescent protein was also stitched into the neural genome. This maneuver allowed the researchers to identify Neuron A simply by looking for the red coloring.

The experiment

With both the engineered virus and the engineered neuron in mind, we are ready to tackle the experiment. (Feel free to consult the Figure if this gets a bit confusing.) The researchers’ goal was to identify only those unmanipulated cells that formed direct synaptic connections with Neuron A and not the thousands of nearby connections with which it did not directly interact.

The engineered virus, sporting its EnvA protein, was allowed to infect Neuron A, which was fluorescing red because of its previous engineering. Because Neuron A possessed the TVA receptor whereas the unmanipulated neighboring cells did not, the engineered virus only infected Neuron A.

The virus now set up a productive infection that normally would not occur (remember, its glycoprotein has been gutted). Because Neuron A already possessed the rabies glycoprotein, the viral life cycle was rescued. A change from red to green showed the presence of reproductive viral activity.

What happened next reveals the reason that the researchers went to all this trouble. The progeny viruses from Neuron A escaped their cellular womb and infected the nerve cells that were directly connected to Neuron A even though these neighboring cells had not been given the glycoprotein.

Remember that the progeny viruses are as defective as the parents. After they entered these new cells, the viruses could still make the next generation, but they would do so without the glycoprotein background available in Neuron A. This means the viruses cannot escape the neighbors’ neural interiors once infected. The neighbors turned green, which identified them as infected neighbors, but then everything came to a halt. No other cells would become infected.

The upshot? You now have a perfect wiring diagram of every neighboring cell that forms a direct synaptic connection with Neuron A-and nothing else. All you have to do is look for little glowing green cells.


This is quite an achievement! Although the manipulations are complex, the view that is attained is an unambiguous neural circuit diagram of every cell that forms a direct connection with Neuron A. Such a feat has never been possible at so fine a resolution.

There are many uses for such basic information. If you have ever looked at the almost ridiculous complexity of neural interactions at the cellular level, you will be reminded of 2 trees that have been uprooted, turned 90 degrees to face each other’s root systems, and then jammed together. If you were to add about 1000 other cells in the neighborhood, you could then fully grasp just how intricate this webbing is in the real neural cytological world.

These complexities can be very frustrating to view, especially because most other techniques (such as staining technologies) miss many synaptic connections. Without a fuller understanding of how the billions of brain neurons are wired together, we will never be able to completely understand how the neural architecture of the brain gives rise to its amazing functions.

Other benefits from these manipulations may accrue as well, including the design of CNS–specific drug delivery mechanisms that use modifications of these techniques.

Regardless of its ultimate utility, these results represent a quantum leap in our ability to directly view specific neural connections. It is a far cry from thinking about neurons in overly simplified metaphors, like Michelangelo’s fresco, beautiful as it is. I am, however, a bit hesitant to throw out the metaphor. Like so much of the work in the Sistine Chapel, a deeper appreciation for the complexities only reveals just how much genius truly exists. In the case of these data, the benefit lies in answering simply whether the neighbors are green or not. When the answer is yes, the masterpiece is revealed.

Note: This is the second installment of a two-part series describing the use of engineered rabies viruses in the elucidation of neural circuits.