At A Glance
Name: Francis Ka-Ming Chan
Position: Assistant professor, pathology, University of Massachusetts Medical School — 2002-present
Background: Visiting/research fellow, National Institutes of Health — 1998-2002; PhD/Postdoc, University of California, Berkeley — 1992-1997; BA, University of California, San Diego — 1991
Francis Ka-Ming Chan, an assistant professor of pathology at the University of Massachusetts Medical School, is conducting cellular analysis of signal transduction pathways by combining a relatively young assay technique — FRET (fluorescence resonance energy transfer) — with a time-tested analysis method, flow cytometry. Earlier this week, he presented his research at Cambridge Healthtech Institute’s Fluorescent Proteins in Drug Development Conference in La Jolla, Calif. Chan also took a few minutes to discuss his work with Inside Bioassays.
Can you describe the type of assay that you’re conducting?
We focus on the use of GFP or GFP variants, and it’s a flow cytometric-based assay. The major breakthrough that we’ve made is with FRET — people have done it for a long time, and the use of GFP variants has also been used for many years — but to adapt it into more of a flow-cytometric assay, I think, we are one of the first groups, if not the first. That of course offers the versatility in terms of adapting it into high-throughput pharma application. That’s also something that eventually we’re interested in doing. But what we have done so far is to basically use it as a tool to examine specific interactions, and also to confirm that the biochemical interactions we have seen using standard biochemistry.
So this is FRET inside of cells. What are the specific challenges of conducting a FRET-based assay in flow format?
The biggest challenge is that no one has really done it before. People have done it using a microscopic method, and people have done it using fluorescence-tagged antibodies, which are basically restricted to looking at interactions outside the cell. The advantage of using GFP variants is that you can monitor interactions in a live cell. And potentially — in fact, what we have done with TNF (tumor necrosis factor) receptor-like molecules — to demonstrate ligand-dependent changes in the receptor complex using a FRET approach. For example, in our case, with the TNF receptors, the level of FRET signal was increased when we stimulated the cells with the ligand. That indicates that the receptor change within the complex is undergoing some kind of reorientation. That’s something that you wouldn’t be able to do as easily with fluorescence-tagged antibodies. First of all, you would need to have antibodies that can actually recognize the appropriate epitope, and at the same time, wouldn’t disrupt the interaction you want to study. So with GFP, to some extent, you can put the GFP moiety any way you want within the molecule to circumvent problems you might encounter otherwise with the antibody-based approach.
The idea of putting it into flow — is that an attempt to increase throughput of this type of assay?
Well, at that time, there were really two reasons that drove us to do that. One is that we wanted a more quantitative assay, and also to be able to look at — unlike the microscopic method, where you can examine a few cells at a time — actually examine a large population of cells to determine whether it’s a more global phenomenon. When you’re examining cells under a microscope, you’re looking at a few cells, and you may be knowingly or unknowingly looking at events that may or may not be representative of what’s going on. But with a flow-based assay, you’re talking about tens of thousands of events, and we thought it would really complement what we’re doing with the microscope. So the challenge, coming back to your earlier question, is obviously that no one has really done that. The GFP variants come with their own set of challenges because of some of the issues about how the molecules actually have a little bit of a tendency to dimerize — even though it is not so bad with cyan and yellow fluorescent proteins, it is a problem that you can occasionally encounter. So I think a lot of the work that went into this was basically trying to make the instrument optimal for the kind of assay we wanted to run.
We first started with the BD FACSVantage flow cytometry model. Not everything was optimal, so we had to play around a bit with the channels and stuff like that. But eventually we got it there, so I think that this is something that is very useful, and we’re still kind of using it to look at cell-surface interactions as well as intracellular interactions. When you think about the antibody-based assay, if you want to monitor biological processes in a living cell, you can’t do that with that approach, because the antibody would have to successfully get inside the cells, which normally requires fixing the cells. So with the GFP variant approach, the cells are still living, and are active.
Are you still using the BD FACSVantage platform?
We started with that, but we’ve moved on to another version — the BD LSR II. Unlike the high-end ones, the basic ones come with one laser, but clearly, especially if you want to be able to monitor fluorescent signals from both the cyan and yellow fluorescent proteins, [you need more]. With cyan protein, in some case when you have enough of it, you can excite it with a 488-nm laser, which is the standard laser for all flow cytometers. But this is not optimal. You may sometimes see a little bit, but the signal is very weak. So you need a second laser to see this cyan fluorescent protein. Now, in theory, you don’t need two lasers, but we always use two, because then we can really see and make sure that the proteins — the FRET pairs — are expressed.
Do you have special analysis software?
We just use standard software that comes with platform for flow analysis. There’s nothing unique about that. The only other thing that we’re still kind of working on is to be able to quantify the level of FRET. Unlike microscopic methods, which can give you some level of quantitation, the flow method we’re still kind of working on. There’s a group of investigators in Hungary, people like [Janos] Szollosi, that have been doing a lot of work on trying to derive an algorithm that will allow you to quantitatively interpret the data from flow cytometric FRET analysis. But we’re not using their method. … Maybe when we’re 100-percent sure that the method will work. But we’re working on it.
What have been the major findings of your research, and what’s next?
So we started off using this method to look at cell surface receptor interactions. We used this to demonstrate that these actually are preassociative complexes to begin with. The interaction, for a long time, was not detected because it was a very unstable interaction. For example, with a standard biochemical method that people use to study interactions, you basically dissolve the interaction. So FRET allows us to look at these relatively weak interactions in the living cell. What we’re doing now is to move from the surface into the cell using this same type of approach to study different signal transduction molecules within the TNF receptor signaling pathway — how they interact, and when they interact when the cells are stimulated.