Rediscovering Biology: Molecular to Global Perspectives
Genomics Expert Interview Transcript: James Carrington, Ph.D.
Director, Center for Gene Research and Biotechnology
Director of the Oregon State University’s Center for Gene Research and Biotechnology. Carrington’s lab conducts research on how viruses and host plants interact, using the model organism Arabidopsis. He uses genetic, genomic, and proteomic strategies to understand RNA silencing pathways, virus recognition events, and cellular targets for various RNA silencing suppressors. Carrington’s research in RNA silencing has been included in the 2002 scientific “Breakthrough of the Year” in the journal Science. The magazine cited a body of work being done by several research groups across the nation on small RNA molecules, calling them “electrifying discoveries, which are prompting biologists to overhaul their vision of the cell and its evolution.”
James Carrington, PhD, is the director of the Oregon State University’s Center for Gene Research and Biotechnology. Carrington’s lab conducts research on how viruses and host plants interact, using the model organism Arabidopsis. He uses genetic, genomic, and proteomic strategies to understand RNA silencing pathways, virus recognition events, and cellular targets for various RNA silencing suppressors. Carrington’s research in RNA silencing has been included in the 2002 scientific “Breakthrough of the Year” in the journal Science. The magazine cited a body of work being done by several research groups across the nation on small RNA molecules, calling them “electrifying discoveries, which are prompting biologists to overhaul their vision of the cell and its evolution.”
What is the big question that your research is trying to answer?
The question that my lab is trying to answer is: How are viruses recognized as foreign invasive unwanted entities that can cause damage and disease? How are they recognized and how do plants respond to those pathogens in ways that elicit a defense response or prevent future infection by that virus or related viruses?
[We want to know] how plants have evolved to recognize pathogens. There are a couple of major categories of pathogens: viruses, bacteria, fungi, nematodes, and a few other things. Plants have actually devoted a large chunk of their genome capacity to defending themselves against all of these different types of pathogens, including viruses.
What we’re really interested in is how plants utilize all of this genetic information to defend themselves against viruses. [That] involves recognizing viruses, eliciting defense responses, and in some cases, customizing defense responses against specific viruses and then remembering that they were infected by those viruses. [The plant has the capacity to make sure] that it can defend itself if that virus ever comes along again.
Why did you choose to use Arabidopsis in your research?
Arabidopsis is a wonderful organism. It’s a small plant that’s in the mustard family. You might see this plant if you’re living in a northern latitude growing in the cracks of sidewalks in the spring or as a little weed in your flower bed.
[What is important] for people like me, is it has a very small genome and it [has] a very rapid generation time. It can be analyzed genetically very easily. It crosses with other Arabidopsis and it also self-fertilizes very efficiently. And an important feature of Arabidopsis worldwide is that there are lots of so-called ecotypes with natural variation all around the world.
We have this worldwide diversity of Arabidopsis that we can go out and analyze and bring back to the lab and answer basic questions of how plants grow and develop, how they respond to and defend themselves against pathogens, and in many cases how eukaryotes grow and develop. Arabidopsis is a wonderful model organism.
Are a lot of people working with this same plant?
Arabidopsis has been adopted by the plant research community as its flagship organism. People around the country — and in fact around the world — have settled on Arabidopsis as their main, front-and-center model organism to study.
The Arabidopsis genome was the first plant genome to be sequenced. There are basic repositories and culture collections and seed banks that are very well organized that researchers deposit their materials and you can draw from those. What this has created is this worldwide network of researchers who all know what everybody else is doing and who all share their resources. That’s the great thing about working with a model organism: you’re not alone, you’re part of a much larger group. When you have these networks of researchers is progress goes much faster.
You mentioned that Arabidopsis has a small genome: How many genes does it have?
This was a surprise, not in the context of the roughly 25,000 that were discovered, but in the context of looking at other organisms. Arabidopsis has many more genes than Drosophila. Some people were initially very surprised at that because Drosophila can do things like think and fly and move from Point A to Point B very quickly. They can see things. They have brains even though they’re tiny brains.
Arabidopsis has 10,000 or so more genes than the fruit fly-Arabidopsis has almost as many genes as a human. The spectrum of genes that are in the Arabidopsis genome are very similar to the spectrum of genes that you find in organisms like fruit flies and humans with some unique features. There are, for example, a number of genes required to make cell walls, which are unique in plants.
Can you describe some specific aspects of your research?
One of the ways that we study plant defense [against viruses] is using quite old technology. We take a virus in one hand, we take a plant in another hand, and we either rub the plant with a suspension or solution containing the virus or we put the virus in this high-pressure artist airbrush device and literally blast a plant with a solution of the virus. These are very crude but effective ways of infecting plants with viruses like the tomato mosaic virus. We then let the natural course of infection occur and this can occur out in the greenhouse or in a controlled growth chamber.
The leaves initially become infected by the virus and then the virus moves one cell at a time in radial concentric patterns until it reaches the vasculature. Then the virus moves up the plant, through the equivalent of the plant bloodstream, the phloem. All of that takes literally just a few days and we analyze the infection at all of those points from the initial perception stage to the replication stage to the movement stage to the long-distance transport stage.
Many of the tools that we have to study virus infection of plants are facilitated by our ability to add genes that allow us to track the virus. So, for example, we can add genes to the virus that turn the virus fluorescent green. [We can] study how the virus replicates at the single-cell level and then invades different parts of the plant.
How has knowing the Arabidopsis genome helped your research?
It’s revolutionized it. If we wanted to analyze genes in plants and their role during virus infection in the old days (which was literally just a few years ago) we did it one gene at a time. Now, with the availability of the Arabidopsis genome sequence, we can build tools that allow us to analyze all of the genes in Arabidopsis simultaneously. One way we can do this is with what’s called a “microarray experiment.”
Why is it important to see all of the genes at once?
It’s revolutionized the type of information that it gives you because now we can not only in an experiment measure gene activity of genes that we’re interested in, but we can also measure the activity of genes that we don’t even know [we should] be interested in yet. The way we know we should be interested in them is how those genes behave during, for example, the course of a virus infection.
So if we have a resource like a microarray chip that gives us a readout or a measure of the activity of all of the genes in Arabidopsis in parallel, we can look at how they all respond during the course of a virus infection. It might turn out that the genes that we were initially interested in don’t do what we’re expecting them to do. However, there may be other genes that we have no concept of yet because we just haven’t thought about them.
In a microarray experiment, we’ll see [the genes] affecting one another or behaving as a part of a gene network.
So how exactly does a microarray experiment work?
First, we ask the following question: What are all of the genes in Arabidopsis that are either up-regulated or down-regulated during the course of a virus infection?
The microarray [itself] is a solid platform like a piece of glass on to which we’ve stuck or have had manufactured pieces of DNA-called features or probes. The benefit of a microarray is we can put those features at a very high density on a chip. In a square inch or a square centimeter, we can pack hundreds of thousands of DNA probes onto the surface of the glass.
The Arabidopsis chips have pieces of DNA that correspond to all of the known genes-and furthermore they have many different features corresponding to each gene. There are over ten features corresponding to each gene on the Arabidopsis microarray chip. And what that gives you is over ten individual readouts for gene activity for every gene in Arabidopsis.
In the experiment, we inoculate a set of plants with a virus and [another set of plants] receive no virus. We [then] extract the RNA from those two sets of plants [in order to] compare the gene expression profiles in Arabidopsis in the two sets of samples.
[After we] take the mRNA from a given plant, we fluorescently label it so that we can detect it. We [then] take our fluorescently labeled solution of mRNA and hybridize the microarray, meaning that we take that solution and allow Watson and Crick base pairing to occur between the labeled mRNA that we extract and the features on the chip.
So let’s say, for example, you have Gene A1 and you have several features on the chip that contain pieces of the A1 gene. If the A1 gene is expressed in Arabidopsis that means an mRNA will be produced and we extract that mRNA as part of the global pool of mRNA. We label the global pool with fluorescent tags and then hybridize it, we let it anile. We know that the mRNA pieces that are fluorescently labeled will stick tightly to the corresponding gene features but they won’t stick tightly anywhere else. So we rinse the chip off put it in a laser scanner to electronically capture the data.
The amount of the fluorescence is directly proportional to the amount of mRNA in the global pool from Arabidopsis, which is a direct reflection of the gene activity that was going on in Arabidopsis at the time you harvested the samples.
One of the things that a microarray experiment gives you is too much data to analyze-we could not possibly analyze it manually or by inspecting the fluorescence output from the chip.
When we scan to collect data from all of the hundreds of thousands of features on the chip, we can visualize each feature as a fluorescence intensity on a computer screen and we can even compare the corresponding spots on the infected and the non-infected chip samples. But our eye tends not to be good enough to discriminate differences in intensity. So, the computer provides us with a number that is a reflection of the fluorescence intensity, which is a reflection of the gene activity. The computer then compares those numbers for us and does some statistics through a whole series of steps that allow us to get a relative comparison of gene activity between the two samples.
And so in a microarray experiment, we quickly go from handling plants and mRNA to handling digital pieces of information. In fact, most of the work in a microarray experiment is sitting in front of a computer just trying to figure out this vast amount of data that the computer has provided you.
How do you compile all of these data?
One of the first things that we do when we analyze a microarray experiment is a graphic analysis of gene activity in our infected and non-infected plants. And since we have a number that corresponds to the gene activity for every gene that is on the chip, we take that number and simply do a scatter plot. A scatter plot simply plots the value of signal intensity from a given gene from the infected vs. the non-infected. On the Y-axis we have the infected, on the X-axis we have the non-infected.
Now, if all of the genes were expressed at the same level in the two different samples and you plotted the intensity of every gene on the graph, you would come out with a perfectly straight line at a 45-degree angle. However, what really happens when you look at the data is most of the genes are on that diagonal and there are a few that are off the diagonal. It’s those ones that are off the diagonal that tell you that the gene is differentially expressed between the two samples. Those are the ones that we’re really interested in. One of the great things about bioinformatics is we can rapidly identify all of those genes that are off the diagonal or that are induced or suppressed and we can put them all in a database and we can start asking questions about those genes: Are these genes known to be involved in defense against viruses? Or are these genes involved in other processes that previously we had no concept was related to virus infection?
A class of genes that we’re very interested in are those genes that are affected by the virus because the virus goes in and shuts certain things down. If you think of virus infection in plants or in any organism, for that matter, like an army invading a country, there are certain logical things that an army or an invading force would want to do. One of the first things that the invasion force wants to do is shut down the communication channels and activate the defenses. So in the army analogy, the first thing that happens is the radar towers are hit and the missile launchers are hit by the invading force. And that’s so that whatever’s being invaded can’t defend.
Viruses do exactly the same thing. One of the first things that they do is go in and shut down certain key host defense processes. One of the ways that that happens is certain genes are altered by the virus in terms of their expression levels. Other genes are affected by the virus not because of the expression level but because how the messenger RNA is processed and how it’s decoded by the normal process of translation. Viruses go in and alter certain gene’s expression programs and messenger RNA translation programs that normally occur and many of those things we can detect on a microarray experiment.
Now that you know which genes might be of interest to you, what is your next step?
The bioinformatics tools that we have to analyze a microarray experiment are fabulous experimental tools because they allow interactive analysis with the computer. [This is] because we have large databases that have information about all of the Arabidopsis genes, and furthermore these databases are all linked with other computers and through the Internet with other networks of information.
When we display on these scatter plots all of the genes in their relative expression in the infected and the non-infected, all we need to do is click on a position and windows will appear from all over the Internet that give us all the information we want to know to start doing experiments on those genes. When we click on a gene we can learn if there’s a known function for that gene, or if the gene looks like other genes for which there’s a known function.
More frequently however what the microarray experiment gives us is a large number of genes for which we were not expecting to be up-regulated or down-regulated, but these database resources that are available to us allow us to quickly get information about any gene that we want to study in more detail.
[When we are researching genes on the Internet], we start with learning some very basic information. What’s the name of the gene, wherein the Arabidopsis genome does that gene come from-there are five chromosomes that contain all of the nuclear genes in Arabidopsis-and because the complete genome sequence is known, we know precisely the locus or the site where that gene is located. And furthermore, we can learn what’s around that gene. If we’re lucky it’s a gene that has been studied well by others or perhaps we’ve studied it in the past and we might know something about how it’s normally regulated.
What particular genes have you investigated?
We’ve become very interested in genes that are differentially affected by viruses, meaning we’re interested in all those genes that the virus, by virtue of making viral proteins or replicating or doing something else, is causing to be misexpressed or misregulated.
One class of genes that we’ve become very interested in is called the scarecrow-like family of genes. This was one of these categories of genes that comes under the “surprise” category because we were not interested in scarecrow-like genes before. We [recently] learned that viruses are effecting how the mRNAs for the scarecrow-like genes are processed during the normal course of gene expression.
What really caught our attention is what these scarecrow genes or scarecrow-like genes are normally doing. These are normally genes that are involved in controlling some of the basic processes of cellular division in differentiation in plants. Some of these scarecrow-like genes also control how signals are produced and transported in plants and how hormones are perceived.
It turns out that what viruses are doing is interrupting certain processes that are normally required for defense. But these defense-like processes, like RNA silencing and RNA interference, are also involved in normal growth and development.
Why are these scarecrow-like genes important?
There’s about 30 or so of these [genes] and most of these that have functions assigned to them, are involved in controlling the activities of other genes. They code for proteins that are called “transcription factors” that either turn on or turn off other genes.
The scarecrow-like gene products or transcription factors are involved in very early cellular differentiation and development. In other words, what we think is happening are these scarecrow-like genes are activating or suppressing other groups of genes that are normally required for growth and development.
What further research is required after you have learned about the genes from these Internet databases?
When we do a microarray experiment or we’re interested in any given gene, we now have large numbers of resources available to us to study those genes and one of the most invaluable resources that we have in the Arabidopsis research community are the so-called “knockout lines.”
These are individual lines of Arabidopsis that have been experimentally manipulated such that single genes have disruptions. The labs that are making these knockout lines are doing an excellent job of cataloging and defining precisely which specific genes are disrupted in all of the given lines. For example, our colleague Joe Ecker at the Salk Institute is building a large collection of insertion knockout lines, and he’s mapping precisely which genes are disrupted. [He then posts] that information on the Internet for labs like mine to log on to and determine whether or not any gene of interest to us has a knockout line available.
After a microarray experiment, we might see a gene that’s induced, and we might ask the following question: Is induction of that gene involved in defense? One of the hypotheses that we would make if it is involved in defense is that if we knockout that gene we ought to create a plant that is more susceptible to the virus. Conversely, if a gene is required by the virus, we would predict that if we knock out that gene, we ought to make a plant that’s more resistant to the virus.
So we log on to the websites that have database information of all of the different knockout lines that are publicly available, we make a couple of clicks, and a couple of weeks later those seeds for those knockout lines arrives in the mail. We take those lines out to the greenhouse, plant them, and in several weeks when they’re old enough we’ll inoculate with the virus and ask: Are these plants more susceptible, less susceptible, or as susceptible as the parental wild type plant?
How does your research with Arabidopsis apply to farmers whose crops are infected with viruses?
One of the great things about working with Arabidopsis [plant] is it’s a model organism- [it] is essentially the stand-in model organism for all the different crop plants out there, most of which are very difficult to study.
Take cherry trees, for example, or take bananas growing in Central America-these are all essential crop and agricultural resources, but they’re very difficult to study for lots of reasons. They tend to have very long generation times. They tend to have other relatively undesirable features for doing rapid experiments. Arabidopsis allows us to identify those big questions that we want to ask in general for plants, be they cherry trees or corn or peppers growing in Southern California.
The whole community of plant biologists chose Arabidopsis [to be a model organism] because it has a very rapid generation time. It has a very small genome as far as plants go. So the genome size of Arabidopsis makes it a very amenable species, because you can go in and sequence the entire genome in a relatively short period of time.
[Even though our research is on Arabidopsis], it turns out that the underlying mechanisms of virus replication, virus spread through plants, and plant responses to the virus, are essentially the same, whether you’re talking about Arabidopsis or a cherry tree. And they’re the same because Arabidopsis and cherry trees are relatively evolutionarily close. In fact, some of the underlying basic mechanisms that you can [study in] Arabidopsis are even conserved beyond plants. They’re conserved in fruit flies and worms and even humans.
The long-term value of this research is, if we can understand the basic underlying mechanisms of how plants respond to pathogens, like viruses, we may be able to treat virus infections in [other crops] or prevent infections much better than we can do it today.
You mentioned that some of the defense mechanisms in Arabidopsis are conserved throughout evolution: Can you expand on that?
A question that we’re frequently asked is: Since we study viruses of plants does that have any bearing on viruses of animals, like humans? And the answer is, it absolutely does. It turns out that the genome revolution that we’ve gone through over the past ten years is really the second wave of genomes to be sequenced. The first wave of genomes to be sequenced happened back in the ’70s and ’80s and they were virus genomes. So there were hundreds of virus genomes that were sequenced 20 years ago. One of the really striking revelations from all that work was how closely the viruses of plants were related to viruses of animals like humans.
For instance, we study a group of viruses that are very similar to the polio virus and the common cold virus and a number of other animal viruses. It turns out [that] a lot of the underlying mechanisms of how they replicate [and] how they express their genetic information are exactly the same [as in] those viruses of animals. There are even some conserved host responses to both plant viruses and animal viruses. One of these responses that seem to be a universal response of infection of eukaryotes to viruses is this process called “RNA silencing” or “RNA interference.” This is a very old evolutionarily ancient defense response that plants and animals have to defend themselves against viruses and other invasive genomes.
Plants are wonderful organisms [with which] to study things like RNA silencing. We can go in and manipulate plants at the experimental level-we can grind them up without any thought of sacrificing animals or doing any kind of human experimentation. A lot of the things that you can do in plants at the experimental level are directly applicable to animals and humans in the context of virus infection.
So will this research help humans eventually?
I have no trouble thinking that the research that the Arabidopsis research community is doing will have a direct benefit to human biology and [the] treatment of human diseases. Again, there are examples of basic cellular processes that are conserved at the virus-host interaction level between plants and humans.
We can study these processes very well in plants. We can apply that information in humans to generate-and we’re talking over the long term-better antiviral therapies, better treatments. Furthermore, a lot of these basic cellular processes that come into play in plants during virus infection or RNA silencing or RNA interference are also playing essential roles in human and animal development.
What are some future questions in Arabidopsis research?
The next wave of questions is how are all of the genes expressed from the Arabidopsis genome? What are all of the proteins doing that are coded by all of those Arabidopsis genes? How are those proteins interacting and doing their jobs? How are proteins interacting with genes? How are enzymes produced that make things like cell walls and fats and carbohydrates? How are all of these things controlled and how are they all interacting in networks? And even further out than that-how are all of those networks that we see in a whole plant evolving over time to make new plants?
1.1 Online Textbook and Video
The online textbook chapters support and extend the content of each video. The Web version can be viewed as a full chapter or as individual sub-sections, and includes links to glossary terms and other related material.