David Van Vactor
Professor, cell biology, Harvard Medical School
• Associate professor, cell biology, Harvard Medical School — 2001-2007
• Assistant professor, cell biology, Harvard Medical School — 1995-2001
• Postdoc, University of California, Berkeley — 1991-1995
• PhD, biological chemistry, University of California, Los Angeles — 1991
• BA, behavioral biology, Johns Hopkins University — 1985
Instructor, cell biology, Harvard Medical School
• Postdoc, Harvard Medical School — 2004-2007
• Postdoc, Salk Institute for Biological Studies — 2003-2004
• PhD, molecular biology, European Molecular Biology Laboratory — 2003
• BSc, genetics/molecular biology, University of Bucharest — 1997
Earlier this month, David Van Vactor and colleagues at Harvard Medical School, including instructor Tudor Fulga, reported in Nature Methods on the refinement of a microRNA-inhibition technology, dubbed microRNA sponges, for use in vivo.
Originally developed in the lab of Phillip Sharp at the Massachusetts Institute of Technology (see RNAi News, 8/31/2007), miRNA sponges are transcripts expressed from strong promoters that contain multiple tandem binding sites to an miRNA of interest.
Aiming to improve upon the work done in Sharp's lab, Van Vactor and Fulga developed an improved version of the sponges to enable their use in intact Drosophila. Last week, RNAi News spoke with the researchers about their efforts.
Let's start with a little background on the microRNA sponges. Are these based on the same technology developed at Phillip Sharp's lab [at the Massachusetts Institute of Technology]?
TF: Yes, it's a similar technology. It's based on the sponge technology from the Sharp lab … [but] the difference and the big step forward is moving the technology into a transgenic setup, which can be applied to living organisms.
DVV: The combination of the in vivo transgenic and the modular [and conditional] expression system … realizes the potential of the system that Margaret [Ebert] and Phil [Sharp] initially developed.
Can you give an overview of where you started and how you went about making [the sponge technology] applicable in whole organisms?
TF: In [Van Vactor's] lab we are interested in microRNAs in the context of the nervous system, and we faced the same problem that everybody was facing at the moment: how to efficiently inhibit endogenous [miRNA] function during development or the adult life [of an animal].
In trying to find a way to overcome this difficulty, we came across the sponge paper from the Sharp lab and immediately thought of how to apply this to transgenic animals. The way that we came up with was to use the very well-established and powerful Gal4-UAS targeted-expression system.
Initially, we collaborated with [Ebert] in the Sharp lab to understand the sponge system and improve the way they were made originally by changing the number of copies, the design of spacers, et cetera. But it is [based on] the same principle [as the original sponges] of repetitive microRNA complementary sequences specific to a given family of microRNAs that share the same seed sequence.
The big breakthrough was placing these [constructs, called miR-SP], under the control of UAS regulatory sequences. This gave us the opportunity to express these transgenic lines using a variety of Gal4 drivers. The Gal4-UAS technology has been extensively used for years now, and there are thousands and thousands of drivers for very specific tissues, cell types, or developmental times, allowing extreme versatility for conditional expression.
DVV: The nervous system is particularly challenging, and this is reflected by the investments of [the Howard Hughes Medical Institute's] Janelia Farm to create a very large collection of highly specific Gal4 drivers. It's a principle that applies to essentially all developmental processes, so the applicability [of the new miR-SP technology] is very broad.
This is reflected by a lot of enthusiasm in the field already, sparked by [Fulga's] first presentation of the technology at a [Cold Spring Harbor Laboratory] meeting — we've already been contacted by many people, and we're very excited about applying this technology to problems ranging from development to behavior.
[ pagebreak ]
Using the new and improved microRNA sponge, how did you go about giving it a test run and what did you find?
TF: Originally when we designed [the sponges], we had no idea if they were going to work. The major question that was left after the Sharp publication, which he discussed in [RNAi News], was whether or not scientists would be able to achieve sufficient expression and specificity to induce effective silencing of a microRNA in living organisms.
Once we designed [the sponges], we did a few proof-of-principle experiments.
There is a small number of well-characterized microRNA mutant phenotypes in Drosophila that have gross morphological expressivity like defects of the leg, wing, and such. What we ultimately wanted to have this construct do was completely mimic a mutant phenotype, and thus provide an alternate way to inhibit completely inhibit microRNA activity.
Interestingly enough, [using] various genetic combinations, we ultimately reached sufficient expression to completely mimic mutant microRNA phenotypes.
DVV: There were about three microRNAs for which there were published phenotypes when [Fulga] began designing this system. It turns out that each one of these served a slightly different purpose. For example, miR-7 was very useful because it showed that the system could be used to titrate the level of expression to create a sensitized background that would allow genetic interactions to be uncovered by the sponge system, which indicates that one could do large interactome screens, potentially to identify targets or other genetic partners in vivo.
[Additionally], miR-8, which was a gene we were interested in based on its synaptic phenotypes, was a beautiful demonstration of how one could use the tissue specificity [of the miRNA sponge system] to not only understand the biology of miR function — in this case, that muscle expression of this microRNA is essential for pre-synaptic development and promotion of synaptic growth — but also the tissue specificity of one of its major targets in the muscle, which turns out to be an actin-binding protein.
In the case of miR-9, again [we had] a very nice opportunity to show not only that [Fulga] could recapitulate the published phenotype, but also take it a step further and determine the minimal set of cells in the wing that could elicit that.
Fortuitously, each of these provided a very nice way to put the system through its paces to show how it could be utilized in a way that would extend capabilities of researchers beyond simple loss-of-function mutations. [And while these] can be made, the lack of conditional control for loss-of-function mutations in microRNAs limit the ability of the investigator to explore the complex functional dimensions.
TF: In other words, what came out through these proof-of-principle experiments, plus the novel function [of miR-8] we discovered, was that not only can we recapitulate a microRNA loss of function, but that [the system is enormously versatile] in dissecting a particular function from the perspective of the cells where a microRNA is required and the type of targets that are regulated differentially in that tissue.
[ pagebreak ]
Will there be further refinement made to the system? Are there areas that you think need tweaking?
TF: Certainly. Actually, the system that is published in the paper is already obsolete in a sense. It's still very effective, but we've already refined it for the next generation. The difference, specifically for Drosophila, is the way the constructs are made. If you look at the system we published, it was using 10 repetitive copies of the sponges and is employing the conventional type of transgene generation, by random insertion into the genome.
The next step, which is done in collaboration with [Harvard Medical School's] Norbert Perrimon, is building a library of transgenic miR-sponge constructs … for every microRNA in Drosophila.
This collection … which should be completed by the end of the year … is improved by the fact that the constructs carry a higher number of repetitive copies  and [are] using targeted insertion into the attP system that Perrimon perfected.
DVV: The targeted insertion will create better uniformity in expression for the various lines, and also provide the advantage that the insertions are mapped at defined loci. We'll have a collection by the end of the calendar year with insertions on two different chromosomes, giving us the ability to combine these sponge insertions with a variety of different genetic backgrounds.
In January, we plan to begin our first pilot screens to define novel microRNA functions at different stages of development and in a variety of tissues. This will be the first application of the system as a comprehensive screening tool.
TF: There is another interesting dimension we are thinking of exploring. Given that this technology is using the Gal4-UAS, bipartite, modular expression system, it isn't restricted to Drosophila. The Gal4-UAS has recently been applied to zebrafish, and other systems of conditional expression, like Cre-Lox, for example, work very well in mice.
We are now discussing with people from the mouse and zebrafish fields to determine if the same principle will work in higher organisms. The next immediate step is probably going to be in zebrafish, because the exact same Gal4-UAS system we used in flies appears to be effective in zebrafish.