Rapid Change in Bacterial Genome

Ben -   Also this week, researchers at the Wellcome Trust Sanger Institute who have been looking at changes in the genome of a troublesome pathogen, Streptococcus pneumoniae.  This bacteria is responsible for a broad range of human diseases including pneumonia, ear infections, and bacterial meningitis.  Fortunately, because this bacteria has been infecting humans for so long, we have samples from all over the world, going back many, many years, and this means, we can compare them genetically to see what has changed in response to modern antibiotic's and to vaccines.  We're joined now by Dr. Stephen Bentley from the Wellcome Trust Sanger Institute.  Thank you very much for joining us, Stephen.

Stephen -   My pleasure.

Ben -   Now the idea that bacteria change genetically in response to drugs or vaccines isn't new, we've known about this for a while.  What's novel about your work?

Stephen -   As you say, we've known about these phenomena occurring over time since antibiotics have been in use, but the kind of breakthrough that we have in the projects that we're now able to do has been driven by our ability to sequence large number of genomes.  Ten years ago, when we tried, when we sequenced the genomes of things like TB, Mycobacterium tuberculosis, one project to sequence the genome of one isolate, which was chosen to represent species, would have taken a couple of years and cost 1 or 2 million Pounds.  The new sequencing technologies allow us to sequence hundreds or thousands of isolates at a cost of around 100 Pounds per isolate, and the turnaround of generating sequence data is down to a matter of weeks.  That allows us now to exploit the collections of isolates that you mentioned earlier to really drill down onto the evolution of a population.

Ben -   So, what samples have you actually been looking at?  I understand they are from all over the world and actually going back quite a long time.

Stephen -   The idea of project was to really use the whole genomes to analyse the evolution over the period since the introduction of antibiotics.  So, when we tried to collect the samples we were looking for as much geographic and temporal coverage as possible. So, the collection we ended up with was 240 isolates, spanning the period of 1984 to 2007, and covering 22 different countries around the globe.

Ben -   What sort of scale of changes are you actually seeing in the genome here?

Stephen -   Substitution mutations which basically occur at random.  One of those will occur approximately every 15 weeks.  But there's another mechanism for variation which we call recombination, and that's basically where the bacteria are able to swap DNA with their neighbours.  On the BBC website, there's a guy, James Gallagher, who's come up with a really nice analogy that's like doing down to the shops and swapping eye colour with someone in the queue at the checkout.  But for the bacteria, that means they can generate enormous amounts of variation in their genomes and each variation that's generated then can be selected for possible advantage.  So in a situation where the organisms are being exposed to antibiotics on a regular basis, which is what's happened in all bacteria since the introduction of antibiotics in the '60s and '70s, if they acquire a variation in their sequence which gives them advantage over the rest, then that's going to give them a good chance of proliferating.

Ben -   Now obviously, the bacteria can't, as you say, meet down the shops and have a chat and say, "I've got these bright blue eyes.  They've been really useful for me for survival.  Why don't you have a copy?"  There can't be a process by which they know that if they translate these genes from another cell that'll be useful.  How does that process actually work?  Is it quite random?

Stephen -   It is entirely random.  So, what happens with Streptococcus pneumoniae, they live in the human nasopharynx and, as far as we know, that's their only natural niche, but there are other bacteria that live there, some of which will be members of the same species but maybe only slightly-related, and then there are other species as well. As cells grow and divide and go through a cycle where there's actually death, some cells will lyse.  Lysing the cells releases the [DNA] bacteria into the environment and that allows bacteria like Strep pneumo to potentially take-up that DNA.  So this ability to take-up DNA from the environment is not ubiquitous in bacteria, it only happens in certain species.  But Strep pneumo does that and is entirely random and some people have thought about whether it might actually be for nutritional needs and much as anything that they take up the DNA.  So, the randomness is there but then selection kicks in, so it's really - if the variation you introduce is disadvantageous, then that will very quickly die out.  If it happens to give you an advantage, then it will proliferate.

Ben -   With all these random changes, do you see changes evenly across the genome or are there regions that are very heavily conserved and regions that are very highly variable?

Stephen -   Yes.  So we see these variation hotspots in the genome and it's interesting that those variation hotspots tend to be associated with surface antigens.  So the major surface antigen in Strep pneumo is the surface polysaccharide and that's one region where we've seen frequent recombinations.

Ben -   Now these are the sugars or the proteins that are on the surface of the cell that it actually presents to, say, our immune system.

Stephen -   Exactly, yes.  And what's interesting from the data that we've seen is that that region is a hotspot, so it's swapping in and out with its neighbours and changing its surface polysaccharide.  But these changes were already in the population at fairly high levels, so that in around 2000, when we started to introduce a vaccine which targeted those surface polysaccharides, we could see that, in the population, there were already variants generated which we're going to be able to evade that vaccine because they didn't have the vaccine target.

Ben -   So they essentially come preloaded to avoid our vaccine attempts.

Stephen -   The population does, yes.

Ben -   How can this help us to develop better vaccines or better antibiotics?

Stephen -   Already, since the seven-valent vaccine was introduced, because these capsules have been spotted, new generations of vaccine have been generated where they now target 10 or 13 of those types.  So, going forward, we would hope to be able to continue to monitor the population in very, very high resolution, using whole genome sequencing, so that we'll be able to understand better how the pathogens are responding to the clinical practices; and then maybe we can adjust those clinical practices to make them more efficient.