At A Glance
Name: Rich Jorgensen
Position: Associate professor, plant sciences, University of Arizona
Background: Research scientist, University of California, Davis — 1990-1997; Researcher/director of flower color engineering program, DNA Plant Technology — 1983-1990; Postdoc, UC, Davis — 1981-1983; Postdoc, Stanford University — 1978-1981; PhD, biochemistry, University of Wisconsin — 1978; BS, engineering, Northwestern University — 1973
After his attempts to intensify the purple color of petunias resulted in a whitening of the flowers, Rich Jorgensen unknowingly and accidentally stumbled upon a phenomenon that would have a profound impact on the biological sciences — the petunias were in fact exhibiting sense co-suppression. RNAi News recently spoke with Jorgensen to get the story behind his discovery and where it has taken him.
I normally start off by asking how you got involved with RNAi, but perhaps it would be better to ask how you got involved with co-suppression.
[At DNA Plant Technology], we were trying to establish a program in the engineering of flower color — in cut flower crops, primarily. In order to raise money from venture capitalists, the idea was to do something in a model flower color species — petunia — that would allow us to put something on the table for them to see. In other words, showing them that we could genetically engineer chrysanthemums to, say, kanamycin resistance would be an important step but it wouldn’t look any different to the venture capitalists. So, our research director said we really needed to do something visible even though it’s not going to be of any commercial significance.
At the time, we thought that we should do an antisense construct for one of the genes responsible for flower color, and at the same time we decided to try to over-express the protein, as well, so that we could attempt to get higher pigmentation levels, deeper colors. So we made both types of constructs and in the end it turned out to be very important that we had engineered the construct for high protein expression.
The antisense worked to some extent, but surprisingly the sense construct was much more effective at silencing than the antisense construct. And it did so in a very different way — it gave us a variety of different flower color patterns that we didn’t see at all with the antisense construct.
At first, we figured that we had probably made a dominant negative protein by mistake, that there was a mutation in the construct. So we did a lot of analysis: sequencing the construct, et cetera, to try to rule that out. We also compared notes with collaborators in Holland at the Free University in Amsterdam who had published, while our work was going on, experiments with antisense constructs showing that they could silence [the gene] chalcone synthase in petunia. They reported, in one line in the paper, that the sense control, as they called it, didn’t have any effect.
In one of my trips to Holland to interact with the growers and the venture capitalists, I visited the university, which we were beginning to set up a collaboration with, to talk about our results and their results, and everybody was quite puzzled about why our sense construct did what it did and theirs didn’t do anything.
Eventually, the graduate student Ander van der Krol, who was doing that work, decided that if his antisense flowers were effected by conditions of light and temperature to give more or less silencing, perhaps he just needed to push the sense plants a little bit harder. So he put the growth chamber under more extreme conditions, and lo and behold, he saw some silencing. In that sense, he was able to confirm our results and both labs showed that it was the accumulation of the RNA from both the transgene and the endogenous gene that was associated with silencing. Initially, everybody seemed to think it was probably a transcriptional silencing, and it was ultimately shown by several labs that it was a post-transcriptional silencing event.
Then what? What did you do?
Well, I was with a company trying to develop products, so there wasn’t a lot of scope for pursuing this interesting observation. Also, when I decided that I would pursue it in academia, I had to do it without a position, without a lab. So I went up to UC, Davis where my wife had just gotten a faculty position and kind of worked on the side. I decided that what I would pursue in the absence of much funding was the epigenetic aspects of the phenomenon because what interested me in particular about it was all the complex changes that were going on in the plants somatically and biotically.
Clearly, this transgenic effect wasn’t stable and there was quite a lot of interesting epigenetic stuff that turned out to be much like the phenomenon of paramutation in maize, which is largely a transcriptional phenomenon. So in effect, we had a complex hierarchy of transcriptional and post-transcriptional silencing that led to the diversity of flower color patterns.
That’s the direction I pursued rather than trying to pursue the underlying mechanism. Then, after a little while, I started focusing more on how it was that the transgene constructs that we had produced were so efficient at producing silencing whereas most labs, including the lab in Holland, had very inefficient silencing. There was clearly something about the constructs, and what it turned out to be, after quite a bit of experimentation, was [that] there were two features of our constructs that were important, that were related to the fact that we were trying to produce a high level of protein. It turns out that you need both high transcription and you need the transcript to be fully translated.
People began realizing, by the mid-90s, that many cases of transgene silencing in plants were post-transcriptional but there were also some cases of transcriptional silencing. [They also realized] that it was duplications of the coding sequences, or the transcribed sequences, that were responsible for post-transcriptional silencing, and duplications of promoter sequences that were responsible transcriptional silencing.
The thing sort of fell out into two different phenomena, which turned out to be related to some extent by double-stranded RNA.
The next major event in the plant area was when Peter Waterhouse showed, in the same year that Andy Fire did, that double-stranded RNA was at the root of this. What [Waterhouse] did was to cross together the sense and antisense lines [finding] that they were much more effective at silencing when both sense and antisense transgenes were present. Then he made a construct that made double-stranded RNA directly as a fold-back RNA, a hairpin RNA, and that was very effective.
That was in 1998, the same year that Fire’s stuff came out — an underappreciated fact, I think.
What about now? What sort of projects do you have ongoing?
The epigenetic aspects of the problem got me quite interested in chromatin and transcriptional silencing — that’s one avenue. Another is using RNAi in plants for functional genomics. The third is still trying to understand the mechanism of sense co-suppression. It’s known from the work of others in Arabidopsis that that mode depends on RNA-dependent RNA polymerase that’s encoded by the plant to copy the transcript that produces double-stranded RNA as a template, presumably, for Dicer. It’s that recog- nition process that’s very puzzling — we’ve engineered transgenes to produce transcripts that are as close as we can make them to the normal transcript for chalcone synthase in petunia, and we still get the silencing just the same way we do with a chimeric transgene. The silencing is not dependent on whether there’s an intron present or not, or what the 5’ UTR is or what the 3’ UTR is. So, the real question is: How is it that a fairly normal transcript is recognized and other transcripts are not to enter this pathway? And: To what extent is that potential mode of control used in plants to control their own gene expression?
It would be a different post-transcriptional mode than microRNAs because it would relate to the recognition of the transcript by the RDRP rather than a Dicer-produced microRNA.
That’s where the mechanistic side of the lab is going. The other aspect of the lab is to use that same mode of silencing to do functional genomics, mostly in polyploids.
Are there any questions that may fall outside your focus but that you’d be interested in seeing answered?
The main one is what I’ve already talked about: anything that has to do with how transcripts are recognized or not — is that subcellular localization or what? There are a number of people in the field doing experiments, both biochemical and genetic, trying to isolate mutants that would maybe illuminate that process. That is the understudied part of the field, I think. There’s been a lot of focus on what happens once you have double-stranded RNA, but less focus on how you get to the point where you produce double-stranded RNA to initiate the process.
The other thing is that the field is looking more and more at transcriptional silencing via double-stranded RNA. I think that’s important in several respects: It could be that functional genomics will be more effective, possibly, with transcriptional silencing than with post-transcriptional silencing.
Then, interestingly, also, because you have the possibility of the genome being modified epigenetically, being methylated via RNA, you have this possibility of a flow or feedback of information about gene expression states back to the chromosomes. On top of that, in plants, you have the possibility that the RNA can move between cells and even throughout the plant via the phloem. So, there’s the possibility of an RNA-based signaling system, what we’ve called … the RNA information superhighway. There’s a lot of work that needs to be done to understand the RNA-based signaling system that seems to exist in plants. How important is that? To what extent does it imprint information on the genome?
Because in plants epigenetic information can even be transmitted between generations, we need to know to what extent information about gene expression states might be imprinted on the genome and then passed onto the progeny, so that the progeny will have some knowledge of perhaps the environment of the previous generation, for instance.
It’s sort of a pseudo-Lamarckian issue.
Any predictions on where the field is going, looking five years out?
I think we’re going to continue to see a lot more virus resistant plants produced this way, because it is a very effective way of modifying virus resistance. People are modifying the biochemical content of plants for both nutritional quality and to remove less desirable compounds — for instance, to improve the oil quality of some oil seed crops, like soy bean, so they don’t have to be hydrogenated, [or] to produce more olive oil-like, more mono-unsaturated cooking oils, which are more stable and reputed to be healthier.
I think there will be a lot of those kinds of things coming as transgenic crop lines in the next five years, for sure.