At A Glance
Name: Rich Jorgensen
Position: Associate professor, plant sciences, University of Arizona
Education: PhD, biochemistry, University of Wisconsin -- 1978
BS, engineering, Northwestern University -- 1973
While he is perhaps best known for stumbling upon the phenomenon of sense co-suppression when he was attempting to intensify the purple color of petunias, Rich Jorgensen has continued to work in the plant gene-silencing field. Currently, when he is not preoccupied with his role as editor-in-chief of The Plant Cell, he is working on a National Science Foundation-sponsored project to develop a new tool for functional genomics using RNAi.
Following up on an interview from last year (see RNAi News, 2/6/2004), Jorgensen spoke with RNAi News this week about his latest efforts.
I came across [an] NSF grant [you recently received] regarding sense RNAi. Could you explain what this is and what the grant entails?
Sense RNAi, as we've been calling it, is the same thing as what we originally called co-suppression and what we and others have called sense co-suppression over the years. [We use the name] to distinguish it from antisense silencing.
The idea is that if co-suppression of sense RNAi is efficient enough, then we ought to be able to take a population of cDNAs that would represent the genes that are expressed in a particular tissue or physiological response or developmental process, transfer them into an appropriate sense co-suppression vector, and produce a population of transgenic plants that have effectively been mutagenized in a targeted manner, that is, by silencing only genes represented in the cDNA population.
By doing this, we would be using RNAi for forward genetics rather than reverse genetics. What people normally do with RNAi is reverse genetics, where they take a sequence for whatever genes interest them and make standard RNAi constructs to introduce to plants or animals, and then try to understand the functions of the underlying genes. In practice, that's not always very satisfying because many times we don't see any phenotype whatsoever. The lesson from reverse genetics is unless you have some clear sense of what process the gene is involved in, you may not know what phenotype to look for.
I think it's been interesting that most of the advances, at least in our field of chromatin genetics in plants, have come from forward genetics by people who are not interested initially in chromatin but who stumble onto chromatin by identifying a gene, underlying a phenotype in the process they're interested in.
So it's been the forward genetic approaches that have led to the discovery of function underlying genes encoding chromatin proteins in plants by an order of magnitude, I think, over the genetic approaches -- even though people have made mutants of all the chromatin genes in plants. I think the problem is that the number of cell types and biological processes in a plant is huge and no one person can be expert enough in them all to identify what's changed in a given mutant.
So we thought we ought to really think more about using RNAi as a forward genetic approach, and that immediately brought us back to our original approach of making a sense over-expression construct to try to reduce the expression of a homologous gene. A complication with that is, of course, that a sense construct designed to over-expression a protein can also obviously over-express the protein, perhaps in different cell types or different tissues. So that complicates the experiment a lot. But there was an observation made at DNA Plant Technology Corp. that was published a few years ago showing that inclusion of a 3' inverted repeat of the 3' UTR would increase the silencing efficiency of a sense construct. It's probably an RNA-dependent RNA polymerase-dependent process. If the 3' inverted repeat were extended through the sense-coding sequence, using the 3' inverted repeat as a primer or by some other mechanism, you'd have double-stranded RNA.
In our initial experiments, we see that on average 50 percent or more of transformants are in fact silenced with these kinds of constructs. So it does seem that it will be feasible to put a population of cDNA molecules into an appropriate vector, with a strong promoter followed by a 3' UTR inverted repeat, where you'd introduce the cDNA population in between the two. We would produce was a large mixed population of agrobacterium carrying a variety of different constructs from the cDNA population, and then transform plants with that population of agrobacterium to produce populations of independent transformants, each with a different gene potentially knocked out.
This approach would be available to any biologist who would do a genetic screen just like any other genetic screen but in this case they would use a population that's targeted to the cell type or tissue type they're interested in.
One area that we're beginning in, to illustrate how this would work and to find out how well it would work, is trichomes -- the hairs on leaves. That's a single cell and we're collaborating with Mike Scanlon at the University of Georgia who can do laser-assisted microdissection to isolate those cells from which you can extract RNA and make a cDNA population to put into our vector. In the population for transformants we would be looking for mutations that affect trichome development.
Assuming that all goes well there, what's the next step after that? To optimize the whole process?
In parallel with that, we will be optimizing the process using photosynthetic genes because there're so many of those that give rise to what're called yellow-green phenotypes in leaves that it would be very efficient for optimizing the process. Three hundred to 500 genes in Arabidopsis apparently code for that phenotype, so depending on the efficiency of silencing, and depending on how diluted those RNAs would be in leaf mesophyll cells, anywhere from .1 percent to 10 percent of all transformants might have this phenotype.
Once you have the phenotype, by using PCR to amplify the inserted sequence in your transgene, and then sequencing that sequence, you can determine what gene the underlying probable effect lies in. We would then recapitulate that by making a standard RNAi construct, then re-transforming the plant to validate that result.
Are there other projects underway that have started since we last spoke about a year ago?
This has really kind of taken over the lab at the moment. The work we were trying to do on the mechanism by which the sense constructs lead to silencing we'll be returning to, and this more applied work may well produce a lot of materials that will be useful for exploring that. But for the moment everybody has gotten sidetracked onto this.
How long is the grant for?
Three years is what we wrote it for, and it's [worth] approximately $1.6 million split between Arizona and Georgia.
You also mentioned that you were interviewed for a planned Nova special on RNAi?
They're saying they're going to run it in July. I'm not sure what to say about it.
Were they mostly asking about the petunia work?
Yeah, and they wanted to spend time in the greenhouse and see how plants were transformed. They wanted to know how we stumbled upon this and what we found and why that was surprising to us. That kind of thing. I think what they want is to try to get the story across to the public that you can be doing fundamental research or applied research in whatever field, and important observations can come out of nowhere. You don't really go into the lab expecting to develop a particular. In this field, it's particularly striking because so many things came out of so many different labs and it all eventually came together.