In Depth: Using Rabies Virus to Study Neural Circuits

Two papers lead up to the most recent publication by Weible and friends. The first (by Wikersham and friends) introduces a modified rabies virus that is only capable of initial infection and reproducing, but can’t move transsynptically from the initially infected cell. The authors accomplished this by deleting the rabies viral coat glycoprotein from its genome and replacing it with EGFP. Since the glycoprotein is necessary for infection, its deletion results in an inert virus. However, when raised in complementing cells that express the glycoprotein, the rabies incorporates it into its coat, resulting in active virus. The first paper stops here and shows that, since mammals don’t produce the glycoprotein, the virus can only infect individual cells and will not spread. Since the virus infects through axon terminals though, you can use it to trace inputs to a specific area by injecting it into the area of interest. There is no cellular specificity here, but Wickersham and friends tackle that issue in their next paper, published 3 months later. In this iteration we get a virus with 2 new features:

1.)   It still has the glycoprotein gene replaced with egfp, but along with it is a construct coding for a glycoprotein/EnvA fusion. The glycoprotein portion allows the fusion protein to incorporate into the rabies coat as if it were the endogenous protein, and the EnvA portion is presented. Using EnvA they were able to target the infection only to cells expressing its receptor, TVA. This method is doubly specific because mammals don’t express TVA, but also due to the “exquisite specificity” of the receptor ligand pair. So by targeting TVA expression to a specific neuronal subtype, you target initial infection very specifically.

2.)   Secondly, TVA is coexpressed with the rabies coat glycoprotein in the neuronal population, meaning that following targeted infection, the replicating rabies virus incorporates the glycoprotein into its shell, producing infection-competent virus. The resulting virus, with its glycoprotein supplied by this transcomplimentation process, will infect transynaptically from the initially infected, or first order neuron. However, unless the transsynaptically infected cell also expresses the TVA/glycoprotein construct (which will happen in a neuronal subtype that is interconnected) the virus labels the secondary neuron and stops there since the glycoprotein is not expressed and no transcomplimentation occurs.

This second point, while making the system very powerful for mapping connections onto specific neuron types, is also an important confound. As I alluded to above, unless the initial infection is targeted to a neuronal population whose cells are not interconnected at all, the virus will spread amongst the targeted cell type and potentially result in a tangled red (or green) mess.

In the third and most recent paper in the series, Weible and friends present the same system again, adding to the mix 8 transgenic mouse lines that express the TVA/glycoprotein construct under control of a tetracycline dependant promoter. They then crossed these lines to mice expressing tetracycline transactivator in specific neuronal subtypes. The result is tetracycline-mediated, neuronal subtype-specific expression of TVA and the rabies coat glycoprotein. In other words, they have cell type specific monosynaptic retrograde tracing capabilities. Weible et al. claim that their method “for the first time overcomes [the following] three constraints” of classic retrograde tracers:

1.)   Lack of cellular specificity of uptake of label.
2.)   Difficulties in determining whether a given neuron is monosynaptically or polysynaptically connected to the initially labeled neurons.
3.)   Highly variable intensity of labeling.

I agree that points 1 and 3 are solved. Despite some small issues with leaky tTA expression and a very low basal level of infection of cells not presenting TVA, their cellular specificity is pretty good. As for the intensity of the label, as I mentioned earlier, the rabies is still replication-competent, so theoretically, infection by a single virion will be amplified greatly, leading to strong label even in minor inputs. All three papers confirm this. However, I take issue with point 2. Despite the specificity of initial infection, if that original neuron is connected to others of the same cell type, those cells also express the rabies glycoprotein, meaning that the virus will spread beyond them, tracing polysynaptic connections. The authors even allude to this problem: “We chose to concentrate our efforts on CA1 pyramidal neurons since […] they are one of the few cell types known to have vanishingly few interconnections.” It’s a shame that such a beautifully worded sentence exposes the shortcoming. I may seem nit-picky here, but I want to see monosynaptic connections onto a single cell! I want to be able to adjust the titre of the virus to get single neuron initial infections without the infection muddying my images by spreading to other neurons in my densely interconnected cell type. Since rabies apparently takes 24 hours to cross a synapse, this issue could conceivably be circumvented by limiting infection time to something safely less than 48 hours – but can we get high enough signal amplification in the secondary neurons in this time span from something as little as a single virion? I don’t know the answer to that.

This system comes tantalizingly close to being able to map connections onto a single neuron but still needs a certain je ne sais quoi to get there. However, Weible and friends certainly have a good tool for determining monosynaptic inputs onto a specific population of neurons, which will no doubt be informative. And given that the virus can potentially make use of any neuron-specific Tet On or Off, it’s going to produce some results. This is definitely a technique to watch out for.

Bonus: With Ian Wickersham (arguably the key author in these three papers) postdocing with Sebastien Seung, we could be in for some impressive reconstruction of monosynaptic circuitry of single neurons in the not too distant future.


1. Wickersham IR, Finke S, Conzelmann K-K, Callaway EM. 2007. Retrograde neuronal tracing with a deletion-mutant rabies virus. Nat Meth 4:47-9

2. Wickersham IR, Lyon DC, Barnard RJO, Mori T, Finke S, et al. 2007. Monosynaptic Restriction of Transsynaptic Tracing from Single, Genetically Targeted Neurons. Neuron 53:639-47

3. Weible AP, Schwarcz L, Wickersham IR, DeBlander L, Wu H, et al. 2010. Transgenic Targeting of Recombinant Rabies Virus Reveals Monosynaptic Connectivity of Specific Neurons. J. Neurosci. 30:16509-13s.

This entry was posted in Home, In Depth. Bookmark the permalink.

One Response to In Depth: Using Rabies Virus to Study Neural Circuits

  1. Pingback: Using Rabies Virus to Study Neural Circuits |

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s