AUSTRALIAN FRONTIERS OF SCIENCE, 2008

The Shine Dome, Canberra, 21-22 February

Emergence of a new disease as a result of interspecific virulence gene transfer
by Dr Peter Solomon

Peter Solomon has a degree and PhD from The University of Queensland on the role of molybdenum-containing enzymes in the photosynthetic bacterium Rhodobacter capsulatus. He held a postdoctoral position at the Carlsberg Laboratory in Denmark investigating the nutritional basis of the tomatoCladosporium fulvum interaction. In 2000, he moved to the Australian Centre for Necrotrophic Fungal Pathogens located at Murdoch University in Perth where he continues to investigate fungalplant interactions. Recently, Peter accepted a senior lecturing position at Murdoch University where he is continuing his studies on the molecular interaction of fungal pathogens and their hosts.

(Click on image for a larger version)

I have a very simple talk today, with a very simple message. I am going to put up some evidence and try and convince you that around 60 to 70 years ago a nasty little fungus, a favourite of mine, transferred a gene to another fungus, a harmless little fellow that sat on the surface of a wheat leaf, and caused it to become one of the most damaging pathogens that we have around parts of the world today one gene!

(Click on image for a larger version)

Before the serious stuff, I should remind us all about what marvellous creatures these little organisms are, these fungi that we work with. Peter Dodds has just outlined how elaborate, how clever they are, to get around resistance genes and everything else. That is really remarkable. And, importantly, it is remarkable what they are capable of.

We see here some absolutely beautiful examples of how spectacular they can be in nature. As we wander through the forests at the right time of year, we see them, we pick them hopefully, we pick the right ones!

(Click on image for a larger version)

But fungi also have their uses. We see here an example of the black truffle at $2000 a kilo, I am told they are beautiful. Blue cheeses are all due to our friends the fungi. There are a couple of other products which occasionally some of us find of use and are due to a non-filamentous fungus, yeast. And we can't forget things like penicillin as well, also due to our friends.

(Click on image for a larger version)

As Peter Dodds has told us, however, all is not well. Much of the food spoilage we see when the kids leave the bread out, and things like that, are due to fungi. They have also got a nasty habit of using biosynthetic pathways, much like what we do with penicillin, to produce things like aflatoxin, said to be the nastiest biologically produced chemical on the Earth horrible stuff.

Fungi can also affect humans, particularly immuno-compromised people. Aspergillus fumigatus, Candida and so on are fungal diseases that unfortunately we can cop from time to time.

(Click on image for a larger version)

As Peter outlined, one of the most annoying habits of fungi is their use of an ability to infect and colonise plants. They cost billions of dollars a year in damage to all sorts of plants. They can affect fruit, crop, leaves, roots, anything you name it, these things can get in anywhere. The list here is just a minute handful of the plant diseases that exist around the world.

But the interesting thing about this list is the column on the right-hand side. That shows the year in which these plant diseases were first observed. We are talking evolution here, but we're talking in the 1970s. Some of these diseases had never been seen before. How did they come about? Where did they come from? Things don't just happen.

(Click on image for a larger version)

I have got an interest in wheat, over in Western Australia a large wheat-growing part of the world. So I have got a keen interest in the first disease in this list. It results from interaction with wheat and a pathogen that is much different from what Peter has just described. (I will go into some of those differences in a minute in fact, they are opposites.) So let's look at the organism called Pyrenophora tritici-repentis (PTR).

(Click on image for a larger version)

PTR is known as tan spot or yellow spot, depending on whether you are from the northern or the southern hemisphere. It looks completely dull and boring, compared with rust; it just produces these little blotches on the leaves. It is a significant pathogen and you will find that we people in molecular plant pathology will always talk up how significant our plant pathogen is, particularly round ARC discovery time!

What is interesting is that PTR has been a pathogen only since about the 1940s. People knew about it before then. People could culture a fungus off a leaf prior to that point, but it was never a pathogen never reported to be one, anyway. It was reported throughout the 1940s in the upper east of the United States and in the '60s and '70s here in Australia.

One very well characterised aspect of this fungus PTR is its host-specific toxin, called ToxA. A host-specific toxin essentially defines the organism's host range, and typically is required for disease. So this toxin will enable that particular fungus to be infectious on that particular plant. To go further into this: ToxA will only work if the wheat carries what is known as a dominant susceptibility gene, Tsn1.

(Click on image for a larger version)

This is where, if you followed Peter Dodds through his great talk, you find something that is completely opposite. This slide concerns the dominant susceptibility gene-for-gene model. In Peter's case which is the classical gene-for-gene model, proposed many years ago if the fungus itself was carrying an avirulence protein and the host carried the resistance protein, there was no disease. You had what was called the hypersensitive response: that localised cell killed itself and prevented the fungus from going any further.

This is opposite. This is a necrotrophic fungus. This loves dead stuff. It secretes its ToxA, and the plant is actually carrying a gene that renders it susceptible. (Perhaps somebody can explain to me why it is doing that. We can't find a good reason. It may be a good point for discussion later.) So there is disease unbelievable, the plant is carrying a gene that makes the plant itself sick. If any of these combinations are missing, if the fungus doesn't produce ToxA, or Tsn1 is missing, in this case, there is no disease. So it is called the dominant susceptibility gene-for-gene model.

(Click on image for a larger version)

Roughly, the way this works is as follows. (Peter outlined this really well in his talk.)

The fungus produces a ToxA which interacts with a protein we know as Tsn1. We know nothing more about Tsn1. We do now have a very good idea of where it is, but we don't know what it is. We know that they interact, and they even get internalised through the process Peter was talking about before, where the fungal protein gets internalised within the host cell. They make their way to the chloroplast, where the only other thing that is known about this interaction is that they bind to something called ToxA binding protein, which results in cell death.

Now, unlike Peter's system, which is a biotrophic interaction where the fungus requires living tissue to feed, ours doesn't like living tissue. It just wants the thing dead. And it likes it dead for two good reasons, in our case. Our fungus feeds on dying cells. Cells are full of nutrients, so it just chomps away. It's happy. Secondly, dying cells can't fight back in this system. So it is in a win-win situation.

(Click on image for a larger version)

My favourite disease is Stagonospora nodorum, a very old wheat disease. It has got plenty of names shown there on the slide. Again the symptoms are dull by comparison with the rust spores flying through the air.

(Click on image for a larger version)

This disease's infection cycle looks something like this. These are wheat leaf cross-sections, and the fungus here is expressing a green fluorescent protein (GFP), so it is just green. We see that at three days post infection it has gone through an open stoma and is rapidly ramifying through the host tissue. By four days we see autofluorescence, which means that those cells that are alive are not happy about it that is typical of a defence response but the cellular structure is generally intact.

By six days, though, the cellular leaf structure has completely collapsed. The fungus has taken over. We can't see any green in the original position any more, because there is nothing left to feed on, and that GFP essentially is expressed in the presence of the fungus feeding. But what we can see is large asexual sporulation bodies called pycnidia, and inside the middle lower image here we see all the nice little pycnidia ready to jump up to the next leaf in infection. That is the way this fungus works: it jumps from leaf to leaf, until it gets to the head and the glume, and that is where the yield losses come in.

(Click on image for a larger version)

All fungi I work with cause significant economic losses 10 per cent, in this case. To put that into perspective: the average wheat yield in Western Australia is 10 million to 12 million tonnes, and this year we had $400 a tonne. If you lose 10 per cent of that, it is more than I get paid!

This was first described as a disease in 1850, and the resistance is quantitative. You can't take out a single gene and fight the disease; there are lots of little bits that just add up.

Interestingly, there is evidence of toxins that we came across with our collaborator Tim Friesen, in North Dakota. But a big asset of this particular system is the fact that we have had a genome sequence for a few years now. This was funded through the foresight of the Grains Research and Development Corporation, and, as I think you will see through the rest of this talk, it has paid for itself by now.

One thing you do when you get a genome sequence, if you are talking lots of data ours is about 37 million base pairs is to put it through a computer and compare that with databases of other known proteins. You try and find out what proteins are in your database that other people have already identified. So we did this.

(Click on image for a larger version)

What we came up with straight away was an unbelievably high hit, a protein of very high similarity to the Pyrenophora tritici-repentis (PTR) ToxA that I was talking about before. So our fungus, Stagonospora, has a gene which is very, very similar and I will come to the similarity later to ToxA. Similarity is one thing; it doesn't mean that they are the same protein. It doesn't mean they have the same activity, either. Does it interact with Tsn1, for example?

(Click on image for a larger version)

An advantage we also have with this fungus is that we can say to a gene, 'You're gone.' We can selectively remove any gene we want to from that fungus, which has an advantage over an obligate biotroph such as Peter Dodds just talked about. We grow this in a lab; with comparative ease we can knock a gene out. We did this with ToxA.

(Click on image for a larger version)

We then take the protein extracts from the fungus that contain the ToxA fraction, and we can simply infiltrate them into the leaf, as you can see in the image on this slide, with a 1 ml syringe. We infiltrated those extracts into leaves known to contain the Tsn1 gene that is required for ToxA activity.

The first of the four examples here is the wildtype fungus. We see the typical chlorotic/necrotic response; when ToxA and Tsr1 are in the same environment you see this reaction.

The fourth example, at the bottom, is our negative control: we should see the response here, and we do.

The examples in the middle are two independent knockouts, where ToxA has been removed from the fungus. Its protein extract now does not contain a protein that is able to interact with the Tsn1 the first piece of evidence that ToxA in Stagonospora nodorum is homologous to Ptr ToxA.

(Click on image for a larger version)

The second piece of evidence is that we can take a fungal strain of Stagonospora, known not to contain ToxA, avirulent when inoculated onto a Tsn1 background, and we can put the ToxA gene back into it. We can do that as well we can knock genes out, we can put them in.

The example at the top left is the fungal strain without the ToxA in it: nothing. But if we can put it back in, we get a good strong reaction with the Stagonospora ToxA. And we can also put the Ptr ToxA back into the Stagonospora and show we get exactly the same reaction. This is confirming functional identity between the Ptr ToxA and the Stagonospora ToxA. These are the same proteins, which have the same activity.

The examples at the lower right are protein extracts infiltrated, and above them are the actual sprays with the spores on the leaf, and we see exactly the same thing.

This is conclusive evidence that Sn ToxA is required for pathogenicity with an interaction with Tsn1. Host-specific toxins in Stagonospora had never been identified before, so this was very exciting for us.

(Click on image for a larger version)

But what I mentioned before was the fact that these genes were very, very similar. In fact, they are 99 per cent identical at the nucleotide level. Generally, this is unheard of, and given that I was the one who prepared the DNA for the genome sequence, I thought, 'Oh God, I've prepared the wrong sample!' I thought we'd just spent half a million dollars sequencing Pyrenophora. Fortunately, that was not the case.

When you see identity to that level you can put it in one of two ways: is this gene simply incredibly highly conserved between the two, or has it been subject to a horizontal gene transfer event? That is where one organism transfers its gene, its piece of DNA, to another, when that other is not its offspring it just shuffles across. It is talked about a lot these days but evidence is still lacking, particularly in fungal systems.

(Click on image for a larger version)

How far does this level of identity extend? We have talked about 99 per cent at the nucleotide level within the ToxA. At the top of this slide we have the 11 kb, or 11 kilobases, 11,000 bases, of the Stagonospora nodorum ToxA region the ToxA is there in yellow. Below that we have the PTR sequence.

The graph at the bottom of the slide depicts the percentage identity of the base pairs between those. We can see that, for all that 11 kb, they are almost identical. In fact, for 7.3 kb, I think it is, they are identical with the odd exception, for a total of about 99.83 per cent. It drops off to 80 to 90 per cent, which is still very high, towards the ends, but after that point identity is gone. There is nothing. It is just as if that piece of DNA has just landed there.

It also contains a transposase, a protein which, to put it very simply, helps a piece of DNA jump from one part of the genome to another also interesting.

(Click on image for a larger version)

We can also look at the sequence variation within the ToxA itself. Sequence variation is interesting because it gives us an idea of, essentially, the age of the gene, or how long that gene has been there. Has it had time to evolve?

Of 529 isolates of Stagonospora nodorum screened, only 29 per cent contained ToxA. And of those 29 per cent, we sequenced those and identified that there were 11 different versions. So it is quite polymorphic: there is a lot of variation out there for the Stagonospora ToxA gene. Thankfully, the strain of Stagonospora we sequenced was among that 29 per cent. (I wouldn't be standing here today if it didn't, I am sure of that.)

(Click on image for a larger version)

For Ptr ToxA a few less were screened we screened 59 isolates of the ToxA gene and there seemed to be much more coverage, 80 per cent, of the ToxA within it. And there was no sequence variation, none at all, which is interesting because if you start to think about age of genes and time of evolution, you realise that this thing has probably just not been there long enough to evolve.

(Click on image for a larger version)

I will just put forward to you now the evidence that we have collected, and try to convince you that a horizontal gene transfer has taken place and has created this pathogen which farmers sometimes just smack their hand on their head about.

It is highly conserved, with 99 per cent identity.

The typical markers that people look for, for conserved genes, is something called an ITS (intergenic transcribed spacer) region, or an actual enzyme called glyceraldehyde-3-phosphate dehydrogenase. These were at 80 to 83 per cent, as you would expect for fungi that are related to this degree. But not 99 per cent.

There is a very similar sequence at the 11 kb region, with this transposase-like gene which can help DNA jump.

ToxA has never been found in another Stagonospora or Pyrenophora species. This is it. It is just found in these two, in everything we have looked for thus far.

(Click on image for a larger version)

There is a lot of variation in the Stagonospora ToxA; there is no variation in the PTR.

And, importantly, Stagonospora nodorum has been a pathogen since its first description in 1889 and who knows how long before that we are talking domestication of wheat and everything else. PTR has only been described as a pathogen since 1941, very recently.

When you put all this together, you look at the similarity of the genes, the time that these things have been pathogens, the time that they have had to evolve the gene sequence only 60 years, in the case of PTR, is the time we think it's been there, so it is not surprising that we haven't seen much of a change. The Stagonospora ToxA gene has been there for a long time. So what has happened around some time prior to 1941, we think, is that Stagonospora nodorum and PTR, both sitting on a wheat leaf surface, have just come together and Stagonospora nodorum has passed on, essentially, its gene to Pyrenophora and made it a pathogen. One gene, and you ought to see the losses this thing can cause because of one gene.

(Click on image for a larger version)

The last experiment I will talk about is a very recent one. It was done on the Broadbalk collection at Rothamsted, just north of London. This is the world's oldest continuous experiment. Since about 1830 they have continually planted wheat on this one piece of field and have tested different nutrient regimes, different fertiliser regimes, and they have collected it since 1830. What a resource. Through wars this stuff has been collected.

It just sits in a shed there, carefully monitored, and a good scientist there by the name of Bart Fraaije extracted DNA. And we can analyse that. We have done that, looking at DNA as far back as 1862, just using PCR (polymerase chain reaction) molecular techniques. We know ToxA is present; we know Stagonospora nodorum is there; but there is no evidence of PTR. This further confirms the likelihood that ToxA was involved in a horizontal gene transfer from Stagonospora nodorum to Pyrenophora tritici-repentis.

(Click on image for a larger version)

A lot of work has gone into this, not only by myself but also by Richard Oliver, at Murdoch, Tim Friesen, in North Dakota, and Bruce McDonald, in Switzerland.