At A Glance
Name: James Patton
Position: Professor, biological sciences, Vanderbilt University
Background: Postdoc, Harvard University — 1988-1992
PhD, molecular biology, Mayo Clinic — 1988
BA, chemistry, University of St. Thomas, Minnesota
With a background in alternative splicing, James Patton's migration to microRNAs and siRNAs seems a natural one. Now at Vanderbilt University, he is studying the function of vertebrate microRNAs during zebrafish development, and pursing the development of what could become an RNAi-based treatment for growth hormone deficiency.
Recently, he spoke with RNAi News about his work.
Can you give a little bit of your background and what you do in you lab?
I have a long-term research interest in the role of small RNAs in regulating gene expression. Mostly that's been splicing, but in the last couple of years the discovery of this gene family called microRNAs and the silencing RNA pathway RNAi has been exciting so I jumped into it. Our newest focus then is using zebrafish as a model system to study the role of microRNAs and how they regulate gene expression during early development.
That's the primary focus of the lab?
It's sort of half and half; half is splicing and half is microRNAs and development.
Can you touch on the microRNA work and what you're looking at?
Sure. We created a microarray to determine the temporal expression patterns of probably around 350 to 400 microRNAs in zebrafish. We have their patterns from fertilized egg to about five days old, and we're doing these arrays so we'll know the expression pattern of most microRNAs in zebrafish. Then the next step is to also take mutant lines of zebrafish in specific signaling pathways, do the same array analysis, and eventually have these patterns where microRNA expression is up or down in wild-type versus mutant. Ultimately, [we'll] use collaborations with developmental biologists to look and see from the patterns what the likely targets [of the microRNAs] are.
So we'll go to the computer and come up with a list of targets. But they have to be validated, experimentally determined, so we use developmental defects, either by over-expression or under-expression of a microRNA, and its temporal expression pattern to then narrow down the list of targets. Then [we] actually go ahead and do the experiments to find out what the targets are.
Are you focusing on particular developmental processes?
My two closest collaborators [are] Lilianna Solnica-Krezel — she studies gastrulation — [and] Bruce Appel, [who] studies neurogenesis. So we kind of, just by default, tend to look at those.
Are they are at Vanderbilt as well?
Yeah. And they're the ones who brought zebrafish to Vanderbilt, as well. I've studied mammalian splicing for all my years, so zebrafish are new to me.
This microarray that you mentioned, can you discuss how you developed it?
Now there are actually several published studies using microarrays to study microRNA expression. We'd actually gotten underway before a lot of these were published, and we haven't published ours yet. I'd had fairly good expertise at isolating small RNAs and fluorescently labeling them and doing a variety of things so we think we have a better array — I think it's better; other people might argue with me. We just isolate RNA, put a fluorescent tag on in, and go to do our microarray.
What are the advantages of this approach that make you think it's better than others?
A lot of microarrays will label up their RNA probes with some step of needing reverse transcription to make a radioactive DNA. We wanted to avoid those steps because you automatically lose some species [that] can be reverse transcribed well. We didn't want to have an amplification step because that would skew the quantitation. So we think that by the way we're doing it taking straight off an RNA prep, labeling it, and going to the array … that this gives us a pretty accurate measure of expression levels and timing of different microRNAs.
Are there plans to eventually publish the details?
I'm not going to divulge everything, but we're about two weeks from what I think will be a strong publication. We've identified a target in a signaling pathway in zebrafish using this array, so we'll probably publish the array as figure 1 in this paper that describes the new target.
What about long-term implications of your work? Where do you see this leading you?
There's a debate and a couple of different papers that say there are anywhere from 300 to 1,000 microRNA genes in a vertebrate cell. I don't know where that real number is going to fall, but either way it's a large number. If you take the known targets right now, there is just a handful. So it's going to take a while, and I think even the best computer algorithms that are going to start leading us toward targets are never going to be sufficient to nail it down. You have to do an experiment; you have to go to the bench.
I wanted to do [my research] in zebrafish where we had developmental possibilities. I didn't want to do it in cell lines, and doing it in mice was inordinately expensive. I think just from that first family [of microRNAs in zebrafish], if we can be the leader in identifying targets, I'll be thrilled.
Longer term, maybe this is just the first family of microRNAs and there're going to be several. Our genomes are full of what's been called junk DNA, and maybe that junk DNA encodes functional small RNAs that work in a variety of processes.
Have you worked with these microRNA prediction algorithms?
I'm not a computer software guy. We certainly have used them, so we will take a given microRNA, apply those algorithms, and come up with putative targets. That's how we do our work in zebrafish. The putative targets will be anywhere from 25 to 40 microRNA targets. We go by what we see from developmental defects and what genes seem to make more sense and try to narrow the search down.
I haven't worked [on] them. I know several people who do, and they're busy with the anchor sequences and whether the 5' end pairing is more important than the 3' end pairing. I guess maybe I have a bias because I'm more of a bench scientist. Ultimately, [the algorithms] all have to be proved.
Do you anticipate ever expanding your work into mammals?
That's probably one of the great things about using zebrafish. The vertebrate microRNAs that we're looking at are almost completely conserved to the human genes. So we think the zebrafish model basically [translates] to looking at human microRNAs.
Do you have any ideas or thoughts on what this [microRNA] research could mean for humans? Do you anticipate seeing these as targets for drugs or drugs themselves?
Well, I'll back up and tell you about a project in my lab that I haven't mentioned. One of the splicing stories I work on is the human growth hormone gene. There are a variety of people out there who have mutations in the human growth hormone gene that causes them to skip one of the exons in growth hormone. If they skip exon 3, it makes a dominant negative form of growth hormone that then prevents secretion of the wild-type hormone from the other allele. So these patients are short in stature.
We've recreated the condition in mice, and have a growth hormone-deficient mouse strain that involves an aberrant splicing event. We can introduce siRNAs against the bad transcript, destroy it, and now the wild-type allele can be secreted and growth hormone levels can come back.
And the mice start to grow?
We've only done it in cells. We're doing it in mice, [but] we haven't tried it from a therapy standpoint. We've made a transgenic mouse that expresses the silencing RNA, we're going to mate it to our growth hormone-deficient mouse and find out if the progeny are wild-type sized or at least increased in size. So can we rescue a defect? That's gene therapy.
We're doing it genetically because that will prove whether it will work. Then the key, if you're going to extend this down to humans, is [whether] you can deliver silencing RNAs as a form of gene therapy. This growth hormone system will be a good system in which to test this because your measure is weight — all you're going to do is weigh the mice.
Hopefully, somebody will pick that up from us and want to do more therapy in terms of injecting siRNAs or — you can imagine all kinds of experiments about how to deliver them, whether it's encased in lipids or however you're going to do it. That's the crux of gene therapy — how to deliver it and get it to the right spot. SiRNAs aren't going to be any different in having that delivery problem, but they're a little easier than having to fix the genome.
So, the more we understand about microRNAs and silencing RNAs, I do think, long term, this could be a viable gene therapy.
When you [talk about] having somebody else come along and run with [the growth hormone work], are you thinking a company?
Have you actually talked to anybody at this point?
I pursued one of those SBIR grants for the growth hormone thing, but everyone is holding off until we see if the genetic approach will work. If that works, we'll see.
We have to deliver these small RNAs to the anterior pituitary, and most of the stuff that's been published [involves delivery] to the liver. It's a whole different kettle of fish to deliver to the pituitary versus the liver.
Do you have a sense when you might know?
Well, the mice are made, and we have the growth hormone-deficient strain. We know the sizes they are, and as soon as we have enough mice to do a statistically meaningful experiment we're going to mate them. So it will be this year.