AUSTRALIAN FRONTIERS OF SCIENCE, 2005
Walter and Eliza Hall Institute of Medical Research, Melbourne, 12-13 April
What the genetics of toxin production by blue-green algae tell us about drug design and discovery
Associate Professor Brett Neilan, School of Biotechnology and Biomolecular Sciences, University of New South Wales
I
want to talk a bit about how my research started in this field, over 10 years
ago, and essentially where it has got to at this stage. I was a human
geneticist and had an opportunity to move into environmental science, and now I
am back doing medical research. So it didn’t work but I still get to travel a
bit, which is good.
(Click on image for a larger version)
I want to introduce the concept of chemical ecology. I have found it hard to pigeonhole myself. They don’t call me a microbiologist where I work because I don’t know the different between Gram stains, so I have come up with this term ‘chemical ecology’. I read it in Science last year, and I think it fits pretty well. It is essentially how small molecules are made, why they are made, and potentially what use we can make of them in the pharmaceutical arena.
These are typically natural products, produced by bacteria and algae. They have various activities for the organisms that produce them, including chemical signalling and chelation of ions. We probably know them better as toxins and antibiotics.
(Click on image for a larger version)
This slide summarises about five people’s PhDs (but students are cheap!). This is an example of a secondary metabolite, a cyclic peptide somewhat similar in structure to the depsipeptides we have just been told about, and of how bacteria spend a lot of energy doing something that we have traditionally considered useless. So what I want to go through here is what is happening, how this is made into the structure, when it is made by the organisms that produce it, and why they are making it.
(Click on image for a larger version)
I guess this story all goes back to the early 1990s and problems in Queensland, New South Wales and South Australia, where we had massive proliferation of blue-green algae, scientifically known as cyanobacteria. The Murray-Darling river system was, essentially, infected or infested by this blue-green algae. The people who lived along that water system couldn’t drink the water. So if you lived in Bourke, you had to buy truckfuls of water. There was also an issue with crop irrigation and farming.
Australia invested a lot of resource 12 years ago, and that is when I went from medical to environmental science. When this issue was arising I could see a use for molecular genetics in, first of all, detecting low levels of these organisms when they are dormant in river sediments, and then working out how and when they are producing the toxins which is the real problem.
(Click on image for a larger version)
These organisms are not a problem normally. The oceans are full of what we call blue-green algae or plankton, really and these organisms produce around 75 per cent of the oxygen that is in our atmosphere. It is nice to have rainforests and the macrodiversity that rainforests harbour, but the rainforests of Indonesia and the Amazon aren’t really producing our oxygen. These are the organisms that are doing it, in the oceans of the world.
So you can’t just get rid of all blue-green algae, as in the engineering outlook. The water corporations of Australia thought, ‘We will just put copper sulphate in the rivers and kill all the algae,’ and that would be great. But also the fish die, oxygen is depleted, less nitrogen is fixed, and the whole cycle just erodes.
(Click on image for a larger version)
Also, blue-green algae is a food source for large populations of the world.
(Click on image for a larger version)
One of my pet scientific interests is how rocks are made by bacteria. The rocks shown on this slide are extremely hard rocks which are believed to be made it is debatable by cyanobacteria and the associated bacterial populations. This is at Shark Bay in Western Australia. (That’s a good reason to do environmental research. You can visit these places and then charge the government for it.)
We have here an extant example of a stromatolite. These are about 8000 years old and can be found today in Western Australia. Further inland in Western Australia, the Pilbara region is where you find fossils of these rocks, up to 3.5 billion or 4 billion years old. So these are the earliest examples of life on Earth, and this is an area of research now that people are referring to as astrobiology: the origins of life on Earth, and therefore an analogue, if we found these geological structures on Mars, to say that there was possibly life on that planet as well.
Looking at that rock, we are saying that that is a biogenic structure: it is actually a geological structure formed by a specific bacterial population.
(Click on image for a larger version)
I am getting a bit off the topic of drugs, but I will get back to them. These 3.8 billion year old microfossils are the first evidence of life on Earth, and about one billion years after that these stromatolithic reefs proliferated on Earth as much as the coral reefs do today. That is when our atmosphere became oxygenic and allowed the evolution of aerobic life.
So the cyanobacteria interest me in terms of their morphological and habitat diversity they live in absolutely every environment on Earth, from the most thermophilic to the most psychrophilic and halophilic environments. But their genetic diversity is extremely small, so that on a phylogenetic tree of cyanobacteria the branch lengths would be miniscule compared with those associated with other bacteria.
(Click on image for a larger version)
But back to the problem. You see here the problem when you get build-up of blue-green algae in this bloom, or this scum, type of situation, and then when that water is fed to people or livestock eat that scum.
(Click on image for a larger version)
The secondary metabolites which are the problem are the geosmin and isoborneol odour factors that is, if your water smells like earth, it is because of those compounds, which are produced by cyanobacteria and Streptomyces and depsipeptides, which you have been hearing about, cyclic peptides, which I will focus on now, and alkaloids, the first examples of bacterial alkaloid production. In the future, I predict, alkaloids will be a major source of pharmaceuticals derived from either wild-type bacteria or engineered bacteria. Scytonemin is another compound from cyanobacteria, a natural UV-blocking agent which is actually being researched in Germany now. That UV-blocking agent is what allows the communities in these stromatolites to survive in periods of very high UV irradiation and desiccation.
(Click on image for a larger version)
Cyclic peptide, the secondary metabolite shown at the top left of this slide as a seven-membered ring, is the subject of this talk. We have worked for several years on the genetics of its production.
After finding the genetics behind this compound, we were then able to find the gene cluster of the compound at the middle left of the slide and how that is synthesised, and then related compounds such as at the bottom left, toxins that are found more in tropical regions including Queensland and Brazil.
Lastly, we move on to analogues of nicotinic acid receptors, shown at the middle right of this slide, and saxitoxin, at the bottom right, which is a paralytic shellfish poison and is the topic of our most recent work. So this is the class containing essentially all the cyanobacterial toxins all the ones that affect water quality and all the ones that we have been working on over the last 10 years.
(Click on image for a larger version)
The main problem in Australian drinking water supplies is Microcystis aeruginosa, which is about a 5-micron diameter organism that is photosynthetic and produces oxygen. The main problem with the toxin from Microcystis occurs with acute doses. It is actually a protein-phosphotase inhibitor and it has its major effect on the liver. The toxin is actively transported by an anionic organic transporter a specific residue on the toxin molecule binds to that transporter from the bile to the liver, where it hyperphosphorylates hepatocytes. The sinusoidal membranes then expand, the liver haemorrhages and you die.
But what is turning out to be the major problem from the toxin from this organism is that it is a tumour promoter: although it doesn’t initiate tumours, it promotes them. And it is promoting tumours to a fairly high degree in countries where there is an initial incident or primary damage to the liver, typically by things like hepatitis. The reason why I have got the photo of China up here is that people drinking surface waters in China are particularly prone to tumour promotion by the toxins of this organism, because they drink the surface water that is not treated at all. We are very privileged in Australia to have treated water, particularly in the big cities.
So this is a future problem associated with the toxins of blue-green algae.
(Click on image for a larger version)
About 10 years ago we sat down and looked at the structure of the molecule of this cyclic peptide, and due to the fact that it has non-natural amino acids the D-form amino acids, some methylated amino acids, and some erythro-type amino acids or hydroxy acids we came to the conclusion that this peptide is not made on a ribosome as most proteins are, but is actually made on a thiotemplate. So it is a thiotemplate mechanism of protein synthesis. And it requires phosphopantetheine transferase as a co-factor. That activity was detected and we also showed that if the cells were challenged with chloramphenicol, which stops ribosomal protein synthesis, this peptide continued to be produced. So we had enough evidence to say it was non-ribosomal.
(Click on image for a larger version)
At the time when we were studying this work, if you mentioned non-ribosomal peptides to someone they would say, ‘What are you talking about? There’s no such thing.’ But even 10 years ago there were a couple of examples of proteins not made on a ribosome essentially, those from Bacillus that are involved in spore formation or the return to a vegetative state after spore formation: gramicidine and surfactin. You yourselves would probably know better examples, including cyclosporin, which is an 11-amino acid structure used as an immunosuppressant, various plant toxins, and enterochelin, from E. coli, which is a siderophore, so binds iron. And now the list of these non-ribosomal peptides has really burgeoned, to the extent where most toxins produced by bacteria are now identified as being from this class.
(Click on image for a larger version)
So we are not really looking at infectious microbiology here, if you are anthropocentric and say that infectious micro-organisms are only those that infect humans. As I see it, however, these types of pathogenic organisms are actually infectious to the planet. And they are real-time. They are moving from Australia to Europe, and by weird mechanisms including ocean currents and bird migration. These gene clusters that we are finding can rapidly move between organisms, so they are transferring either by phage or by transposition, and they are moving around this planet. So it is an infectious disease, in a new sense.
It was the work of several PhD students to clone and characterise this full gene cluster, which was at the time and maybe still is the largest gene cluster found in bacteria. It is about 65 kilobases from end to end, and it has a central biodirectional promoter, identified at the top of this model where the arrows change direction from left to right. It is essentially a whole bunch of peptide synthetases, which activate and then extend the amino acid or the peptide region of the molecule identified here at the left of the model, and a polyketide region of the molecule, which makes this side chain and this is the actual bioactivity of the molecule. This moiety at the top left of the molecule is the one that I said binds to a component of bile acid and transports the molecule to the liver. So one of the ideas that we are trying to get funded is to chop off this polyketide region at the far left of the model, and have this molecule at the top left of the main structure as a carrier for liver drugs. But it is all theoretical at the moment.
There are also a whole bunch of other ‘tailoring enzymes’, one or two of which I might talk of in a minute.
So it is a huge gene cluster, and 2 per cent of the genome of the organism produces it. It produces about 2 per cent of its dry weight of this compound, and we don’t know why it does it. But a biochemist will always tell you that is a secondary metabolite and therefore it is of no use.
(Click on image for a larger version)
We identified 45 of the 48 catalytic reactions necessary for production of a small heptapeptide, and listed here are the precursors. At the time, again, it was the first integrated example of a peptide and polyketide synthetase; it had two classes of secondary metabolite joining together.
(Click on image for a larger version)
The novel aspects were the tailoring enzymes the methyl transfer events, which also add to the bioactivity.
It is a non-linear process. Whereas most non-ribosomal production of proteins is traditionally linear or colinear, this one isn’t. It is from a central promoter. And there are numerous post-synthetic modifications, which I will mention briefly right at the end as being potentially useful for biotechnology.
(Click on image for a larger version)
This slide shows green fluorescent protein in a filamenta strain.
One of the major steps we had to overcome was the laboratory transformation of this organism. I am assuming that these organisms are easily transformable in the environment. They grow up to huge cell densities almost overnight, so it is a rapid proliferation, and then the next day they are gone, the water is clear. I am assuming that there is massive cell lysis and that water is full of DNA, and the potential for that DNA then to move on to other cells is very high.
Natural transformation of cyanobacteria is quite common in the environment. However, to do it in the laboratory is near impossible.
(Click on image for a larger version)
I would like to talk about one application of the molecular genetics, or discovering the gene cluster. We were then able to identify the promoter region and all transcript initiation points along this gene cluster, and to identify that the promoter itself actually has alternate transcription initiation points, based on the level of light that the organism is exposed to. This is not so strange, considering that the organism is photosynthetic and it is found in the environment all the time.
(Click on image for a larger version)
What was interesting to find out is that when you are exposing the organism to high light, especially during early and mid growth phase of the organism, you get a very large up-regulation of transcription. What you don’t see, however, is a concomitant increase in toxin within the cell.
(Click on image for a larger version)
So we looked at one of the tailoring enzymes that I was saying are in the gene cluster. It is an ABC transporter (an ATPase binding cassette transporter), indicated here with an arrow, and we are predicting that that transporter, because it is within the gene cluster, can actually bind and transport microcystin. There is enough phylogenetic evidence or structural motive evidence to say that that is the case. And now we have made antibodies and found that the toxin and the transporter are co-localised.
(Click on image for a larger version)
The transporter is also up-regulated in high light and the hypothesis now is that in high light there is more toxin exported out of the cell which again is a problem for water quality, because water quality works by filtering this organism out of the water, and if toxin is in the water we can’t stop that. But simply placing trees back on the edges of rivers and lakes shades the areas where the toxic organism is found, and doesn’t allow it to get to the levels of high light where toxin is exported. This is a very simple remediation of the problem, with a little bit of scientific evidence for people to sell it. So to sell this idea to a farmer, to put trees back on his river bank, stopping access by his cattle, has worked out to be a very good use of this scientific discovery.
(Click on image for a larger version)
So what is the toxin doing? We are saying it is there, it is activated by light, but light may not be the actual factor. It could be that sunlight itself is actually affecting the availability of iron in the water, and the toxin is a siderophore.
(Click on image for a larger version)
In all organisms that produce the toxin, we find the gene, and therefore we have a genetic test for that problem. And we have now found that the organisms that have the toxin gene have obtained it by lateral gene transfer events. We have shown that to be by transposition.
(Click on image for a larger version)
It has also been found in other genera of cyanobacteria, Nodularia being the main case.
(Click on image for a larger version)
Depicted here is the genetic test, showing how we now test water quality in water that looks as if it healthy and there is no problem with it. We can find single cells that may cause problems in the future, given the right conditions for bloom proliferation.
(Click on image for a larger version)
The next move has been towards elucidating the genetics behind this structure, cylindrospermopsin, a tropical cyanobacterial toxin.
(Click on image for a larger version)
To illustrate the latest work we are doing, this photograph is of a red tide, which accumulates in shellfish and is actually the worst of the algal or cyanobacterial toxins because we are the end users of shellfish. This is a bacterial alkaloid and we have recently found some indication of what genes are involved in the production of this compound. (I mentioned in connection with the secondary metabolites that alkaloids, which have analgesic or anaesthetic qualities, are a new class of compounds that are now being discovered in bacteria.)
