At A Glance
Name: Pilar Blancafort
Position: Senior research associate, Scripps Research Institute
Background: Research associate, Scripps Research Institute — 1999-2004; PhD and MSc, Université de Montreal — 1994-1999; BSc, Universitat de Barcelona — 1988-1993
Pilar Blancafort has been using transcription factor libraries as tools for gene expression analysis for several years under the tutelage of Scripps Research Institute scientist Carlos Barbas. In 2005, Blancafort will continue her research at the University of North Carolina, where she will serve as an assistant professor in the pharmacology department. Blancafort took some time last week to discuss her work — and how it integrates cell-based screening — with Inside Bioassays.
How did you develop an interest in your current field of study?
I became interested in the engineering of proteins for therapeutic reasons. I think that transcription factors are very interesting for correcting disease genes — generally speaking, molecular therapeutics. This means that if you have a gene that is somehow damaged, or is over-expressed, or not working properly, you can use a transcriptional de novo design of proteins to correct those genes and try and make the [expression] levels of those genes like a normal cell. So I have been interested in how we can use those proteins and how we can design them so they can be operational in cells, specifically for diseases such as cancer. So we can use a tool that exists naturally, because we build off of transcription factors. There are thousands of proteins that I’m working on as we speak. But how can we use these as tools to correct diseases?
It seems as if this is an alternative to methods for modifying and studying gene expression, such as RNAi. Can you explain how this differs?
It’s different basically because transcription factors can be used for two things. [They] can be used for upregulating genes, and can also be used to knock down gene expression or repress specific genes. That makes it a unique tool because within the same protein you have the ability to modulate a gene in both directions, which is pretty difficult to do, so far, with other tools. With a small interfering RNA technology, you can only knock down the gene function. Say you want to study the function of a gene — you can study what happens when you suppress that information, when you knock down RNA levels. But with the same transcription factor you have the possibility to go up or down with the gene. Depending on the cell line we work with, certain genes are more expressed or less expressed, so particularly in cancer cell lines, it will be very interesting to modulate in both directions.
Another thing that I want to point out is that although it’s just a different way to regulate the gene — it’s transcriptional. It operates via DNA-binding protein that can either activate or repress transcription. I see it as a complementary tool to current methodologies. It will be very powerful to use those transcription factors in cooperation with other methods such as RNAi, so you have a way to regulate the gene at two levels: one is transcriptional, the other is post-transcriptional. That’s the way we look at it: how we can regulate genes at multiple levels.
Another thing is that with siRNA, you are regulating a gene that is already expressing. You have an RNA population, and you use this siRNA to knock down that transcript. With a zinc finger, you go down to the first level, which is transcription, so that’s why they’re complementary.
Has this tool been explored at all as a way to modify phenotypes for drug screening purposes? It seems as if it would be a powerful tool to do so.
Yes, we actually do phenotypic screens in the lab using cellular populations. So you basically have a population of mammalian cells — usually we work with cancer cell lines. Then you deliver this library of transcription factors, so you have millions of different transcription factors with very unique DNA-binding domains combined together. So now you can look at the cell population and ask the question: Does any cell in the population have the phenotype of interest? That can be from a change in that cell, like a change in migration or the way they divide or look, but they can also be more complicated, such as the cellular resistance to particular drugs. Then you can isolate these cells from the population, and isolate the transcription factor that is most likely responsible for the observed phenotype, and then isolate the genes that are involved in that. You can definitely do a genome-wide screen, and we are currently doing that in the lab. There are many different ways to screen for phenotypes. One thing that is particularly interesting in our lab is to screen for cells that resist viruses. Or in another example — cells that are more metastatic, or cells that migrate more. So we are doing a variety of phenotypic screens right now.
What type of methods do use for these screens? Flow cytometry?
In the original Nature Biotechnology paper, what we did there was isolate from these cell populations members that upregulate or downregulate certain cell surface markers. And for that, we used cell sorting — an antibody that binds specifically that protein that is either up- or downregulated.
But right now, the ways to screen will be more functional assays, such as multilog resistance — cells that survive a high dose of therapeutic agents such as Taxol. So it’s basically asking the question: Can some of these proteins upregulate or downregulate the gene that is involved in some aspect of this resistance? You basically select for surviving cells. You can also do screens on the ability of these cells to be metastatic — and there are a variety of assays for that. We are now doing screens in whole organisms, as well. So we’re putting these cell populations in mice and seeing whether these cells can be more metastatic or migratory — looking at cells that move out of the primary tumor.
How do these cell-based and whole-organism screens fit in with biochemical screens?
It’s important to look at cell populations so you can take into account all the factors that are involved in tissue organization. For example, in a tumor, you take into account that it’s not just this particular cell, but you have surrounding tissues that are strongly in contact with this population. So the role of context is very important, especially in cancer biology. We always try to think that way — try to make the screening as physiologically [relevant] as possible. That would be the reason we are trying to move toward living tissue.
Are you planning to continue this research in your new laboratory?
More specifically I am interested in the mechanism of signal transduction, so this goes more into detail about how genes interact in a given pathway. So you use this type of approach to understand when genes are involved in a particular phenotype and trying to map them in a given pathway. The way that we used to study gene function in traditional genetics was by knocking down that gene or overexpressing that gene one at a time. But it runs out that the genes are interacting in the context of a genetic network, which is very interesting. And everything changes there, because you cannot really view the function of a gene as being isolated, but instead strongly influenced by many others, because these genes are interacting in a physiological situation. So I think that a transcription factor is a more compact cDNA library strategy where you can have a way to ask the question: What are the multiple genes involved in that particular change? So you have a way to understand genetic networks. I’m going to be in the pharmacology department, so these are the kinds of questions I’d like to answer in that context.