PodcastCell Biology

Science Signaling Podcast: 23 June 2009

See allHide authors and affiliations

Science Signaling  23 Jun 2009:
Vol. 2, Issue 76, pp. pc11
DOI: 10.1126/scisignal.276pc11


This is a conversation with Andre Levchenko, author of a Protocol published in the 16 June 2009 issue of Science Signaling. He discusses a method for using a microfluidic device to quantify cell signaling events within individual cells.

(Length: 17 min; file size: 8.2 MB; file format: mp3; location: http://podcasts.aaas.org/science_signaling/ScienceSignaling_090623.mp3)

Technical Details

Length: 17 min

File size: 8.2 MB

File Format: mp3

RSS Feed: http://stke.sciencemag.org/rss/podcast.xml

Download Podcast: http://podcasts.aaas.org/science_signaling/ScienceSignaling_090623.mp3

Educational Details

Learning Resource Type: Audio

Context: High school upper division 11-12, undergraduate lower division 13-14, undergraduate upper division 15-16, graduate, professional, general public and informal education

Intended Users: Teacher, learner

Intended Educational Use: Learn, teach

Discipline: Biochemistry, Biotechnology, Cell Biology, Pharmacology, Immunology

Keywords: Science Signaling, engineering, microfabrication, microfluidic device, protocol


Host – Annalisa VanHookWelcome to the Science Signaling Podcast for June 23rd, 2009. I’m Annalisa VanHook. In this episode we’ll talk about using a microfluidic device to measure signaling events in individual cells with Andre Levchenko, corresponding author of a Protocol published in the June 16th issue of Science Signaling (1). Dr. Levchenko spoke with me from his office at the Johns Hopkins University in Baltimore.

Interviewer – Annalisa VanHookDr. Levchenko, hello.

Interviewee – Andre LevchenkoHello.

Interviewer – Annalisa VanHookYou’ve recently published a Protocol in Science Signaling about a method for quantifying signaling events at the cellular level. Why is quantitative information on the scale of individual cells particularly valuable?

Interviewee – Andre LevchenkoThere are two reasons for that. One reason is that if you try to determine, with whatever method is available to you, try to determine the level of signaling that takes place in individual cells, it is frequently very different from the level of signaling that you would have if you averaged across the cell population. And many, many phenomena, in our body, many processes in our body are dependent on the behavior of individual cells. For example, it is thought that tumors are formed by individual tumor cells that venture out of the tumors and through the bloodstream to the new host tissues. So, under many circumstances it is of interest how the behavior of single cells—individual cells—is different from the behavior of the community of the cells.

Interviewer – Annalisa VanHookHow do people generally quantify signaling in single cells, and how does that differ from the way one quantifies signaling within a cell population?

Interviewee – Andre LevchenkoIt is still predominant that so-called biochemical assays are used. For example, the Western blot or any type of blot or other types of assays where you require thousands, if not millions, of cells to acquire enough material to do some sort of biochemical study—a study where you have extraction of components and then determination of level—let's say phosphorylated proteins, or proteins that are altered in some ways. So, this gives you an averageD information about the behavior of the community. There are methods for single cell analysis, but they also frequently have important drawbacks. For example, it’s very popular to label proteins within cells with green fluorescent protein tag, or its variants. And one of the issues with that is that in mammalian cells, in particular, these studies imply that the protein of interest that is being tagged is overexpressed. And, if you’re interested in the dynamics of signaling, or in the quantitative nature of signaling events, usually it’s a big problem. So, it is important to find alternatives to this methodology, and an alternative that we’re using is a high throughput immunostaining of the cells so that they are exposed to a variety of different conditions or exposed to those conditions for different amounts of time, so we can reconstruct the signaling events—both in time and in terms of combinatorial response to different inputs.

Interviewer – Annalisa VanHookCould you explain just a little bit about the technique, and how the experiment is done, how it’s set up?

Interviewee – Andre LevchenkoSo, this is an experiment that is fully run within one small microfluidic device. The miniature nature of this means that single devices, similar to a quarter in terms of how large it is—it is compatible with the usual microscope slides that one would frequently use for imaging—but on this very small scale you have the ability to run dozens of very small, tiny experiments. So, if you take maybe ten thousands of cells—10,000 cells or 20,000 cells—you can distribute them within the device to different areas and run dozens of separate experiments in this defined different areas of the device. The way the liquid is handled on this micro scale on a scale of microns, gives you the ability to control the flow and control the amounts of molecules, for example, of interest, in the flow very precisely. And so, what you can do is you run these very different small experiments; you can assay different conditions—for example, with different exposure times or different concentrations of drugs of interest or growth factors of interest.

So, all of this can occur in one experiment at the same time, after which you can fix the cells in the device. You can stain the cells using antibodies against proteins of interest or other components of the cells, for example, the cytoskeleton or nucleus. And then you can visualize the results of the staining so that you get images of thousands of cells coming from just one experiment. Now, if you run dozens of different experiments in different parts of the device, you can keep track of what you do, and you can decide that, for example, 200 cells would be sufficient to get enough statistics for single one small experiment, so that if you introduce ten thousands of cells, you can distribute them in fractions of, let’s say 200 cells, into different parts and run these different experiments. So, in the end basically what you have is the results coming from thousands of cells and enough statistics to look at the different conditions; so that let’s say you have 200 cells telling you how the cells responds to one condition, and another group of 200 cells telling you how the cells to respond to a different set of conditions. So, it’s incredibly powerful in terms of statistics, in terms of the throughput, and in terms of how many experiments—small experiments—you can run simultaneously in one device.

So, to summarize, it’s a well-controlled experiment, where you control the conditions very precisely, that gives you very high throughputs and gives you the ability to look at multiple different types of conditions that the cells can be exposed to and respond to in different ways. I think the fact that we can handle liquids and change the solutions—for example, to expose the cells while there’s still alive, to fix the cells and then to stain the cells—all of this is done in one device, in part because we can handle the liquids precisely, we can exchange the liquids so precisely. So, this is extremely convenient. All of this is computerized. In fact, the experiment itself [is] done in such a way that you just press a button on a computer and the experiment basically runs itself. And that’s another advantage of using microfluidic chips—they’re very precisely fabricated, so that you know very precisely the structure of the tube, and you can combine this with microscopy in such a way that all of this can be automated very easily.

Interviewer – Annalisa VanHookAnd you can very rapidly change the conditions to which the cells are exposed.

Interviewee – Andre LevchenkoYes.

Interviewer – Annalisa VanHookSo, you can introduce a drug or a ligand or something, for example, and then fix them milliseconds later.

Interviewee – Andre LevchenkoAbsolutely. And it can do very complex things, also. For example, if you’re interested in how a drug affects tumor cells—in the body, frequently you’ll see that the drug is not present at the same level, it may enter the bloodstream and then be cleared out of the bloodstream, so it’s a very complicated curve, a very complicated change in the level of the drug. And we can reproduce in time how the cells are exposed to the drug, so that we can reproduce what is called the pharmacokinetics of the drug—the level of the drug in the bloodstream—very precisely. So, the ability to change things in time, and do it very precisely, can allow us to mimic the conditions that are more relevant to what is going on in patients, for example, or just create very interesting nonlinear inputs so that the amount of the drug can change in various ways. So, the ability to control the microenvironment of the cells is a very important advantage of the microfluidic chip experiment.

Interviewer – Annalisa VanHookTo understand how the microfluidic device itself works, you can put a large amount of cells into the device, and then you can divide these cells up into sectors, so all the cells have the same initial starting state, and then you can introduce a drug or something, and then, for example, fix cells at different point in their response to the drug—so cells that are in one sector you could fix half a second after their exposure, and cells in another sector you could fix to study at one second after the exposure?

Interviewee – Andre LevchenkoThat’s true. So, you can do it this way, or you can, for example, time the exposure in such a way that all the different exposures end at the same time, so that you can start exposing the cell to the drug at different time points, and then you can end your experiments in different sectors of the device at the same time and then follow this by fixation of the cells. So, there is a considerable amount of flexibility of how you would do such experiments. And the same device can allow you to play with the time course, or exposure times, or let’s say the very same device can allow you to do very detailed dose-response. So, if you’re not interested in time course you can also design experiments so that different sectors will receive different amounts of a interesting chemical or maybe different combinations of different chemicals. You can also vary, for example, the particular antibodies that you would use so that different sectors will then be stained in different ways. So, that allows you to have assays of different signaling events, for example. So, this flexibility is also a very nice feature of this device. The flexibility is very powerful, and that was one of the important factors for us, when we thought about the actual design of the device—it should be flexible enough to allow all those different types of experiments.

Interviewer – Annalisa VanHookWe know that you can use this device for a wide range of conditions, of environmental conditions. Can the device be used for a lot of different cell types—can it accommodate cells of very different sizes, or cells that require very different culture conditions?

Interviewee – Andre LevchenkoAbsolutely. So, we’ve tested almost every single line in the so-called NCI60 collection. This is a collection developed by National Cancer Institute—kind of a standard collection of tumor cell lines for research on different cancers. We have tested it for a variety of adult and embryonic stem cells, obviously for the usual model cell lines that people have used in the lab, including fibroblasts, for instance, and a variety of other model cell lines.

Interviewer – Annalisa VanHookCan you give us an example of the type of information you’ve been able to get about cell signaling using this device?

Interviewee – Andre LevchenkoAbsolutely. In fact, the first paper that used this technology, but didn’t describe it very well, was published a couple of years ago in Proceedings of [the] National Academy of Sciences, where we, in collaboration with Andy Feinberg here, an oncologist, studied the effects of epigenetic control of the expression of a growth factor in people that may have predisposition to colon cancer based on their epigenetic gene regulation (2). And what we found, using this device were the details of how a set of signaling pathways are activated in cells that are either epigenetically normal or something that you would probably call normal—something that doesn’t give you a predisposition to cancer—versus cells that can ultimately be more predisposed to form tumors. So, the detailed analysis we could do with thousands of cells and a few different signaling pathways revealed to us very important differences in how these pathways were activated over time and allowed us to conclude that the cells, due to this epigenetic control, become addicted to a specific growth factor, so to speak. And that, furthermore, allowed us to predict what perturbation of the signaling pathways might be very effective in taking advantage of this addiction.

So, very interestingly, following this publication, this—not this particular method, but the results and the ensuing methodology of treatment for prevention of cancer—were licensed by a pharmaceutical company. So, it provided something that was also more or less immediately useful for people who have clinical interest or pharmaceutical development interest in this problem. So, another interesting aspect of that that we analyzed was the characteristics of differential response, or distribution of responses, in the population of cells in the context of the immune response, or innate immune response, where the pathway of interest is the NF-κB pathway and other pathways—and that’s something we describe more in detail in the Protocol that is published in Science Signaling, and the details can be found there. But, the overall result was also very exciting in the sense that we could find that the cells separate naturally into populations that are distinguishable in terms of their responses. And you'll have some populations that are engaged in this innate response—or at least how we model this innate immune response—versus the cells that are not engaging in it. And so, we are using this information now to try to understand the details of the signaling events and potential implications for the immune response.

Interviewer – Annalisa VanHookThe use of this microfluidic device, then, can presumably be expanded to include any number of signaling pathways or conditions that researchers could come up with.

Interviewee – Andre LevchenkoOh, absolutely. And, it’s designed to be scaleable so that, for example, if experiments call for use of even more cells, say the 10- or 20,000 that one would use with the device as described, might not be sufficient, or one wants to look at more different signaling pathways or different conditions, one can easily scale this up so that the device itself may increase in number of, let’s say sectors of cells that one would use or a number of cells that it can accommodate. So, yes, it’s true, but also what we anticipate this device might be useful for is diagnostic use. So, one of the important features is that even though it’s thousands of cells—if you compare this to the number of cells you can get in a biopsy—even if it’s a so-called fine needle biopsy, which is minimally invasive—the number of cells is very small compared to what you can get from a patient. So, what we anticipate might be possible to do is to run this very detailed analysis based on a relatively small number of cells that may come from individual patients, and report, in a patient-specific way, what properties those particular cells have in terms of their responses to drugs or to growth factors or other conditions that may be present in the tumor during treatment or before treatment or after treatment. So, we anticipate that it might be useful even clinically and we’ve been discussing this with physicians here at Hopkins and have received a lot of interest so far in this possible use.

Interviewer – Annalisa VanHookPutting it within the realm of personalized medicine…

Interviewee – Andre LevchenkoYes.

Interviewer – Annalisa VanHook…in determining what particular drug regimen might be the best to treat a specific individual’s cancer.

Interviewee – Andre LevchenkoYes. And this would be very gratifying if that proves to be the case. But we are very hopeful—so far we haven’t found anything that might prevent this potential use. So, we are trying to see if this can be done, too. So we anticipate it might be useful to very different communities—from basic scientists interested in details of signaling to maybe pharmaceutical companies interested in running very high-throughput drug screening experiments to clinicians who might be interested in the applications, in terms of diagnostic or prognostic use of these devices.

Interviewer – Annalisa VanHookThank you for talking with me, Dr. Levchenko.

Interviewee – Andre LevchenkoThank you, thank you very much.

Host – Annalisa VanHookThat was Andre Levchenko talking about a Protocol from his group published in the June 16th issue of Science Signaling (1). The article is by Cheong and colleagues and is titled “Using a Microfluidic Device for High-Content Analysis of Cell Signaling.”

For more on how microfluidic devices are changing the way researchers study cell signaling, you can read a Perspective by Timothy Elston from October 2008 titled “Probing Pathways Periodically” (3). Also, in the Science Signaling archives you'll find a Perspective from Navratil and colleagues titled “Microfluidic Devices for the Analysis of Single Cells: Leaving No Protein Uncounted” (4).


That wraps up this Science Signaling Podcast. If you have any questions or suggestions, please write to us at sciencesignalingeditors{at}aaas.org. This show is a production of Science Signaling and of AAAS—Advancing Science, Serving Society. I'm Annalisa VanHook, and on behalf of Science Signaling and its publisher, the American Association for the Advancement of Science, thanks for listening.


View Abstract

Stay Connected to Science Signaling

Navigate This Article