Director, Center for Systems Biology, Duke University
• Professor/chair, biology, Duke University — 2002-2007
• Professor, biology, New York University — 2001-2002
• Co-director, NYU Center for Comparative Functional Genomics — 2000-2002
• PhD, cell/developmental biology, Harvard University — 1981-1986
• BSc, biochemistry, University of Paris VI — 1981
An international team of researchers led by investigators from the Duke Institute for Genome Sciences and Policy this week published a report in Nature describing how microRNAs travel between cells in the model plant Arabidopsis.
Also contributing to the work were groups at the University of Helsinki, Uppsala University, and Cornell University.
While investigating plant root development, the investigators discovered that a transcription factor called Short-root travels from the plant's vascular cylinder to the surrounding endodermis, where it activates another transcription factor called Scarecrow. These two then activate two miRNAs that travel back to the vascular cells, degrading target mRNAs encoding certain transcription factors and thereby determining cell types in a dose-dependent manner.
RNAi News spoke with the study's senior author, Duke's Philip Benfey, about the findings.
Let's start with background on the research focus of your lab.
We work on plant development in the larger scope of developmental biology. We have a strong interest in a number of areas, but they primarily center around the trajectory from stem cell to differentiated tissue.
In that context, we work primarily on the root of Arabidopsis because it's a great model with some simplifying aspects. That is, the different tissues are arranged essentially as concentric cylinders around the vascular tissue at the center of the root. Also, new cells are made from a set of stem cells at the tip of the root, and then … all the progeny are constrained in files behind them, so you always know the age of every cell and the identity of every cell in, essentially, a two-dimensional matrix — on the radial orientation you have the cell type, and on the longitudinal orientation you have the developmental stage of the cell.
Given all the different things you investigate in the course of your research, how much of a focus are microRNAs?
This was a bit of serendipity, as much of biology really is. It came about from overlapping work from three different labs … two of which [are run] by former postdocs of mine — Yka Helariutta at the University of Helsinki [and] Ji-Young Lee, who's at the Boyce Thompson Institute [for Plant Research] at Cornell — and one [of which is run by one] of Helariutta's former postdocs, Annelie Carlsbecker, who is at Uppsala University. So this is sort of like working with your children and grandchildren.
[We had observed a] change in the cell specification in different mutant backgrounds, and … knew there was a transcription factor moving in one direction, [so how this] … could influence cell identity where it didn't seem to be acting was the great conundrum.
Previous work had shown that microRNAs were involved in targeting the genes that were involved in [this phenomenon], so [by] putting together a lot of different parts of this story, we came upon microRNAs as the central acting element.
Can you give an overview of the research and your findings?
What we found was in the two mutants there was a change of identify of the vascular tissue of the water-conducting tissue where, normally, you have [a cell type] called metaxylem and, exterior to it, [a cell type] called protoxylem. In these two mutants, there was only metaxylem [including in places where protoxylem should be].
A normal place to start to understand how this works would be to think that one of the two genes [that control xylem patterning,] called Short-root, [which activates another transcription factor called Scarecrow] was actually expressed in the tissue in which the protoxylem was normally forming.
To test whether the expression [of these genes] in these tissues was cell autonomous, we made a form of Short-root, which is [normally] able to translocate from cell to cell, that couldn't move. We found that when it couldn't move and we placed it where it was normally expressed, it had no effect — it could not turn on the correct identity of the cell type.
We then made a form of it that couldn't move and put it in the cells to which it normally moves, and found in that case that it could actually rescue the mutant phenotype; that is, we could get this protoxylem to form.
That said that this transcription factor had to move from where it is normally made to another cell in which it has activity. And then, somehow, it has to send a signal back to the tissue from where it's made to change the identity of a different cell type.
What led us to understanding that there is are microRNAs involved is [the fact] that a set of transcription factors known to be involved in the cell identity of xylem tissue in the aerial part of the plant were expressed in a way that strongly suggested that they might be involved in this case. And they were known targets of a microRNA family.
We started to look at that family and, indeed, two members, [miR-165a and miR-166b,] were expressed precisely where we would expect them to be expressed. We showed they were direct targets of Short-root and Scarecrow. We then used a sensor construct for microRNAs to show that although the microRNA was normally only expressed in a single cell type, it actually was active in more than that cell type.
I should point out that that did not fit the dogma when we were in the early stages of working on this. At that time, there were some very specific statements in the literature that microRNAs don't move — siRNAs move, tasiRNAs move, but microRNAs do not.
We spent a lot of effort to be absolutely certain that this was actually a moving process — that the activity was microRNA activity and that it was non-cell autonomous. By expressing microRNAs in different mutant backgrounds and by using different sensor constructs, by in situ hybridization, we tried to address this issue and showed that there was actually what appeared to be a gradient of non-cell-autonomous activity from that microRNA spreading out in both directions.
At this point, do you have any understand of how these microRNAs are moving?
We don't. The only thing we can say is that plants have special conduits between cells called plasmodesmata, and there is a reasonable amount of evidence that transcription factors like Short-root actually move through these plasmodesmata. So it would be a reasonable thing to believe that microRNAs also move through them.
We think it's unlikely, but we have no evidence for or against, that it is moving through a process similar to diffusion; we think it is probably an active process of some sort. But that is pure speculation.
If the movement is occurring through plasmodesmata, then would it be a phenomenon limited to plants?
Yes, but there are these means of moving small RNAs around C. elegans, for example, that Craig Hunter [at Harvard University] discovered quite a few years back. So there are means of moving small RNAs in other organisms. The fact that this process is particular to plants, in my mind, does not rule out the possibility that there are similar gradients formed in other organisms.
Is there work underway to follow up on these findings?
There are efforts in collaborators' labs to understand the actual mechanism of movement. There are also efforts ongoing to understand how the precise mechanism by which [Short-root] specifies tissue [type], and [whatever else] it may be doing — from everything we can see, it is not just moving directionally inward from where it is made, but moves outward, as well. So it may be interacting with other mRNAs in other cells to do other things.
Most of that is occurring outside my lab in former postdocs' labs.