Name: Brian Harfe
Position: assistant professor, University of Florida
Background: Postdoc, Harvard Medical School — 2000-2003
Postdoc, Emory University — 1997-2000
PhD, muscle development, Johns Hopkins University — 1997
BS, molecular biology, University of Glasgow, Scotland — 1993
Having worked with Andy Fire during his graduate studies at Johns Hopkins, Brian Harfe has been interested in RNAi since the gene-silencing technology was first being uncovered. Since then, he has kept at least one foot in the field, and he currently researches the developmental role of miRNAs.
Additionally, he continues to collaborate with University of California, San Francisco, researcher Michael McManus on the creation of a collection of miRNA knockout mice.
This week, Harfe spoke with RNAi News about his research.
Let’s start with an overview of your lab.
My lab is a developmental biology lab so we study early vertebrate development, and concentrate on the limb a lot. We started to move into other areas, so we have a project going on in intervertebral disk development. We have also started a project on looking at microRNAs in the germ line.
I like to keep less than 50 percent of my lab working on microRNAs, but it seems like everybody that comes to the lab wants to work on these things — they’ve all heard about them. Probably about half my lab works on the microRNA-related projects and the other half is working on more classical developmental biology pathways … in the context of limb development and intervertebral disk development.
Did your interest in microRNAs start before you came to Florida?
Oh yeah. I was interested way back when Andy Fire came into the lab meeting one day and said, ‘Hey, you guys should be injecting double-stranded RNA to knock out your favorite genes in C. elegans.’ It was like, ‘Look what I figured out,’ and at the time we all thought it was a cool C. elegans thing where we could knock out our favorite genes instead of spending two years doing a classical reverse-genetics screen.
So even when I was in the yeast lab for my first postdoc we were doing some RNAi stuff. Then, when I moved [to Harvard], I did it as a side project, initially. Then Mike [McManus] and I joined forces to make [a conditional knockout allele of Dicer] and we went from there.
In your lab, the microRNAs you’re looking at are involved in limb development?
Yeah. Unlike what Mike’s doing, we’re not so much interested in how [miRNAs] work, we’re interested in what the microRNAs are doing — their developmental function. In the limb, we’ve spent a lot of time. We’ve also recently moved into the germ line, and we also have some nice stuff on what they’re doing in muscle development.
Can you talk about some recent findings?
In the limb, one of the initial publications from Mike and I using the Dicer knockout described what these microRNAs are doing in limb development. To our really big surprise, when you get rid of Dicer — so you get rid of all microRNAs in the limb — [mice] make a pretty normal limb. The limb is a lot smaller than normal, but they make all the bones that are found in a normal limb, they make all the muscles, they make tendons, they make ligaments, there are blood cells in there, there’s skin in there. That was a big surprise because we thought when you got rid of microRNAs you’d have no limb. That doesn’t appear to be the case. It seems like these microRNAs are more modulating how the limb forms and not specifying structures in the limb.
What about in the germ line?
None of this stuff is published yet, but in the germ line the only stuff that’s been done with microRNAs has been done by Alex Schier’s lab [at Harvard] in zebrafish. They can do a genetic trick in zebrafish where they can get rid of both maternal and zygotic Dicer.
If you just make a straightforward Dicer knockout, you can’t study germ line because its maternal-protein loaded and you can’t knock out Dicer in the mother because you’d have a dead mouse. So we’ve recently used our conditional allele … to knock out Dicer only in the female germ line or only in the male sperm cells so the rest of the animal is perfectly normal. We can mate these guys together to get embryos — and they do form embryos — that are completely Dicer-minus. Then we can ask what happens to those embryos. We can also ask what happens when sperm don’t have Dicer in them or oocytes don’t have Dicer in them. What are the developmental consequences of the loss of those things?
Is this the kind of thing where you go into it with some theories of what’s going to happen?
In this case, yeah. We had some theories because of the things Alex Schier has done in zebrafish. We had some idea that we’d probably have an interesting phenotype coming out.
Can you talk about what you expect might happen?
In the zebrafish system, when you have a Dicer-minus sperm fertilizing a Dicer-minus oocyte, they get an arrest that’s much earlier than the null phenotype because now they’re getting rid of the maternal. The embryo develops and goes through gastrulation and early development, then it kind of poops out. The most recent work has shown [that] the reason there are a lot of defects is because it looks like Dicer is important for clearing maternal messages from the oocyte. We haven’t proven that yet in the mouse system, but that’s something we’re investigating.
As far as what the sperm look like, the germ lines are not as perturbed as you would assume. They do make sperm, and we’re trying to figure out what the defects are in the sperm that lack Dicer and presumably lack microRNAs.
Do you have a sense of the broader implications of this work?
People always ask that: ‘Five years or 10 years from now, what is the field going to look like? Where is it going?’ My opinion is that it’s really going to move to the targets of microRNAs. MicroRNAs are just like any other gene — they’re like a Hox gene or homeodomain gene. There are lots of targets, and the question is figuring out what those targets do. So I think you’re going to see people moving away from, ‘Let’s knock out Dicer in this tissue and knock out hundreds of microRNAs at once and thousands of targets,’ to ‘Let’s knock out a single microRNA or family of microRNAs and see what the phenotype is.’
Are you still working with Mike [McManus] on the mouse knockout consortium (see RNAi News, 2/23/2006)?
Oh, yeah. He came up with this idea of knocking out all microRNAs, and this really means all of the ones that are in the database now — 350 or something like that. There are probably more but we’re starting with those, creating knockouts, and making them freely available to the scientific community to study them.
My part of the project is making the mice. Once the consortium has individual microRNAs knocked out in embryonic stem cells, those will be sent my way, and the University of Florida, with me coordinating, will make all the mouse lines. We’ll house them here at the University of Florida and post them immediately on a website. Anybody that wants one can then just order it, and we’ll send them out.
People ask, ‘Aren’t you going to do some initial characterization of them?’ [My response is that] if I’m dealing with hundreds of mouse lines, I’m making them, and as soon as I know they went germ line and they’re the right thing and its targeted correctly, you guys figure out what the phenotype is.
And the University of Florida has given us money to do that, to fund the work to make all these lines.
Has that work started?
That work has started. I know some of the targeting constructs are already done … [but] we’re at the early stages. But it is going forward.
Have you already started to hear rumblings of interest in this?
Oh, yeah. I gave a talk out in Virginia about a month ago, and there were very interested people. There’s also interest from companies that want this reagent to go study whatever they study.
Are they interested in terms of using it themselves, or have you gotten interest from companies that want to market it?
There is definitely interest from companies that want to help us fund the project so they get made more quickly, but Mike has been dealing with all that.