Associate professor of chemistry
Massachusetts Institute of Technology
Name: Alice Ting
Position: Associate professor of chemistry, Massachusetts Institute of Technology, 2007 to present
Background: Assistant professor of chemistry, MIT, 2002 to 2007; post-doctoral fellowship, University of California, San Diego, 2001 to 2002; PhD in chemistry, University of California, Berkeley, 1996 to 2000
To help them analyze protein-protein interactions, researchers predominantly use fluorescence resonance energy transfer and protein complementation assays. Each method, however, has shortcomings that muddy the results: FRET techniques have limited sensitivity, while PCAs, which include split-green-fluorescent proteins, can yield high false-positive and false-negative rates.
In an article published June 27 in the online edition of the Journal of the American Chemical Society, researchers from the Massachusetts Institute of Technology describe a probe they say can image thousands of interactions within a living cell, including protein-protein interactions.
The probes are derived from an enzyme and its peptide substrate: if the probe-linked proteins interact, the enzyme and substrate also interact, which can be easily detected. According to the authors, their probe is an improvement over existing PPI methods because it reduces false data while remaining highly sensitive.
This week, ProteoMonitor spoke with Alice Ting, senior author of the paper. Below is an edited version of the conversation.
What was the motivation behind the work described in the paper?
Existing live cell protein-interaction detection methods have problems. There’s the problem of a lot of false positives and false negatives.
How does the probe you’ve developed compare to GFP probes on false positives?
What we’ve done is engineered it so that the intrinsic affinity between our enzymes and our substrates is like 400 micromolars, so it’s very weak interaction that will not just happen spontaneously when the two proteins to which they’re fused do not interact.
That’s a consistent problem with a lot of these split-reporter systems … Even when you don’t have an interaction, the reporters themselves come together.
How did you decide to try to work around this using an enzyme and its peptide substrate?
Because I thought about the split-reporter system and I thought, ‘Well, if it’s just a binding interaction … if you rely on two reporter halves coming together to generate a signal, then you want the signal to be relatively stable and so you want the two halves to bind with relatively high affinity.’
If I just use an enzyme substrate interaction, I can still have a signal even after enzyme and substrate dissociate from one another. And so it removes some of the burden of having to have a high-affinity interaction.
Aside from lower false-positive rates, what other advantages does your probe have over other PPI methods?
Also lower false-negative rates because the substrate peptide is so small, so it’s not going to be blocking interactions as frequently as say GFP does.
Is this a technique that can be easily duplicated by other researchers?
Yes, Sue Lindquist’s lab [at MIT] is duplicating it right now and extending it to yeast, as is Peter Walter’s lab at the University of California, San Francisco.
Are there drawbacks to this method?
Right now, in order to read out the biotinylation signal, we have to fix the cell, but we’re going to develop versions of the technology where we don’t have to do that, where we can do live, real-time read-out from live cells.
Have you used this process to identify new PPIs?
That’s also the next step of what we’re focusing on now. I think it should be very well adapted for that.
Generally, what does this method mean for protein research? Does this open up new research areas or kinds of PPIs that can now be analyzed that couldn’t before?
In principle, it should be possible to detect interactions between transient protein-protein pairs … or weakly interacting pairs, whereas a lot of the other methodologies give false positives as soon as they start looking at weakly interacting proteins. So this opens the window of the possible protein interactions that can actually be imaged in live cells.
Is this method most applicable to transient proteins, or are there other kinds as well?
Perhaps proteins that are sensitive to fusion tags because the fusion tag is much smaller and thus it should be less invasive and proteins that are sensitive to that may perform better using our reporter.
Does this have any applicability to membrane proteins?
It should. It should be easily extensible to membrane proteins.
Would someone need to take special steps to apply this to membrane proteins?
No, it should be completely generalizable to membrane proteins.
Is this method not applicable for any kinds of proteins?
Right now, it’s not good for time-lapse imaging, so dissociation of protein-protein interactions.
Is there any way for a researcher to work around this himself or herself?
Probably not. They probably will have to rely on advances from our lab, but we’re working on it.
In your article, you say that the method overcomes some of the limitations of other existing methods. What are limitations that you have not been able to overcome?
Well, one thing is that it’s still a recombinant fusion technique. … This still requires fusing tags. It would be best to have just a completely tagless system. That doesn’t exist yet, so we’re not solving all the problems of live cell detection methods, but I think we have a method that is superior in many respects to existing methods.