Professor of biological sciences
University of Texas
at El Paso
At A Glance
Name: Renato Aguilera
Position: Professor of biological sciences, University of Texas at El Paso, since 2002
Background: Assistant/associate professor, department of molecular, cell, and developmental biology, University of California, Los Angeles, 1989-2002; PhD and Postdoc, immunology, University of California at Berkeley, 1982-1989; MS, biology, UTEP, 1982.
Renato Aguilera and colleagues at the University of Texas at El Paso have developed what they believe is a novel approach to screening compounds in cells. The assay is based on the detection of GFP in a both eukaryotic cells, such as human HeLa cells, and prokaryotic cells, such as gram-negative and gram-positive bacteria. Using a simple GFP-based detection method, the researchers are screening a variety of chemical compounds for their potential cytotoxic and bacteriocidal properties. Aguilera, who has helped author two recent papers on the subject, discussed the assay last week with CBA News.
How did you develop your interest in call-based assays?
This came about from a simple conversation I had with a colleague in the chemistry department by the name of Luis Martinez. He has been developing potential anti-cancer compounds, or cytotoxic compounds that one would hope could eventually have anti-cancer properties. These are based on the general structure of plumbagin, which is from the black walnut and has some anti-cancer properties. It is basically a naphthoquinone that has been known for probably a couple of decades and has potent cytotoxic activities. He had been deriving something like 15 variants of the compound, and was struggling with the screening — and eventually he was going to want to ramp up to thousands of compounds. So I started thinking about that and came up with an idea that was actually quite simple, and that is to test them on cells that have the GFP gene expressed. These cells that we use were obtained from a colleague in San Diego and they actually express GFP in the nucleus, so it's a very clean background, and you can see them disintegrate when you add the compound using simple microscopy.
So what was the impetus for the tandem screening of eukaryotic and prokaryotic cells?
That was actually the second idea. There were two papers — one was published in 2004, which described the original concept of screening using simple fluorescent microscopy, which is actually quite tedious because you have to count the number of cells that are dead, visually. This was work that was done by an undergraduate in my laboratory. So the next step was to make this higher throughput to go to a 96-well format. So that was another simple idea that we thought of: Why not make things easy on ourselves? Put the compounds in overnight, kill the cell, and if they're dead, there won't be a signal from the GFP. And that worked beautifully, just as simple as it sounded. We'd put a certain number of cells in a well, in 96-well plates, and use 40-some-odd compounds in duplicate and triplicate, and we got very reproducible results on the cancer cells. But then we upped the ante, and decided to ask the next question. We could actually screen the same compounds simultaneously on other cell types, thus cutting down the screening process. If you're going to screen the same compounds anyway, on the same mycobacteria, then you can do it at the same time, and then compare the results simultaneously. We ended up with three different plates — one was cancer cells, one was mycobacteria, and the other was E. coli, all of them GFP-tagged. Surprisingly it worked very nicely. Doing it in three different organisms is not particularly trivial, but it was doable.
Is the thought, then, that you can find which compounds are toxic to one but not to another cell type?
That's the whole premise of it, exactly. And we found two sets of compounds from this new screen — compounds that were mainly toxic to everything, and compounds that were actually mostly toxic to the mammalian cells, but not to the bacteria. We would like to find the opposite — compounds that are toxic to the mycobacteria, but not toxic to the mammalian cells.
So the thought is that these compounds could be potential anti-cancer and anti-bacterial?
That's right. As far as I know, it's the first time that anybody has done tests on three different organisms simultaneously in these types of relatively high-throughput assays.
Are you thinking about doing this in other cell lines now?
Yes, we're trying to generate different cancer lines so that we can screen ovarian cancer, prostate cancer, liver cancer, et cetera, simultaneously. Right now we have an ovarian cancer line, and we want to go to other lines — basically any cell line that is amenable. It has to have certain characteristics for us to be able to manipulate. It's easier if it's an adherent cell, although we can probably work with both because eventually they'll all settle to the bottom of the plate. We're just at the proof of concept stage right now, though. We haven't gotten to do any of those experiments yet.
In a way, the cells are being used as a reagent…
Right, as biosensors.
Would you like to now follow up and find out why a compound is toxic to one cell line and not to another?
Yes, actually we've already done a little bit about that when we went to characterize the mechanism of toxicity. What is this particular compound doing to the mammalian cells? Most of the compounds that we tested induced both apoptosis and necrosis in mammalian cells — mostly through apoptotic pathways. But in the bacteria, if we find one that is only toxic to bacteria, then the prediction will be that it will be a fairly novel way of killing [the cells]. What we've seen is that if it kills the bacteria, it kills the mammalian cell. And it would be nice to eventually find the drugs that kill only mycobacteria. There are very few drugs, actually, against mycobacteria that are effective. There haven't been good drugs for at least a decade.
The tandem screening approach — do you think it is novel for the drug-discovery industry as well?
As far as we know — there is nothing in the literature. I have a review coming out next year called "Green fluorescent protein as a biosensor for toxic compounds," to be published in Annual Reviews of Fluorescence that actually scours the literature for the use of GFP as a biomarker.
What type of detection technology did you use for this?
For this particular experiment we used a commercial instrument called the FluoroScan, which is a 96-well reader that can basically do luminescence and fluorescence. We get results in seconds. We're used to taking a whole day to count cells, but now it takes seconds. And then we reproduce the results with a microscopy-based assay. One is the quick and dirty screen, and then we move to the other assay. We're now even using fluorescence-activated cell sorting to look at activation of aopototic markers.
Have you considered using any automated microscopy platforms for this?
That's exactly where we're going. We just purchased an instrument — basically an imaging cytometer. I want to keep it a bit under wraps, because we don't think anybody has done this, and we are thinking about the possibility of a patent. My colleague is going to develop thousands of compounds, and there are also compound libraries elsewhere at our disposal, such as those from the National Cancer Institute. So we want to go to the next format. Obviously we're not there yet, but… we have the machine sitting in loading dock, let me put it to you that way.
So this is from a commercial vendor?
Yes. We want to get to the point where we can use the same kind of approach in the 96-well format, but for 1,000 compounds at a time, so we can literally screen thousands of compounds in a period of days, as opposed to weeks. It's very doable. We'll have to work out the technical glitches, which we'll encounter for sure, like we did with this paper. But the idea would be higher throughput assays. There has been a lot of interest when I've given talks and at poster presentations at conferences, especially in these very simplistic types of assays that don't take a lot of manual labor. If we are going to go with the thousand-plate assays, then we'll have to use robotics. So that's another technical challenge we'll have to work out, and is probably another whole conversation.
So you are looking to eventually patent the assay methods?
The tandem analysis coupled with high-throuhgput screening. We already, just using 96-well plate technology — if your reads are only 10 seconds — we can easily do a thousand a day. But we're looking at 1,000 per plate, or whatever, depending on the plate format.
Have you received interest from pharma about this?
I'm just a basic academic scientist. My colleague, who actually works in the department of chemistry, actually worked in the biotech sector for a while, so I know he has interest in the applicability of this to the pharmaceutical industry.