Discovering new antibiotics

Using genetic techniques, scientists are screening for new drug molecules to which bacteria struggle to develop resistance...
21 October 2013

Interview with 

David Williams, CEO, Discuva


New antibiotics, which can tackle specific bacterial infections, are urgently needed to overcome rising rates of resistance. But how are these Enterobacter cloacae on agarnew drugs developed? One Cambridge-based company, Discuva, are using genetic techniques to identify new drug compounds to which bacteria struggle to develop resistance. David Williams, CEO of Discuva, took Ginny Smith to his lab to demonstrate how they go about finding potential new drugs...

Ginny -   So, you're looking for a compound that might work as a drug to kill bacteria.  Where do you even start?  There must be so many compounds out there.

David -   There's about 10 million compounds you can acquire.  So, what we do is we look at those computationally to look for chemical qualities that we think will allow them to penetrate a bacterial cell.  We screen them against different bacteria, that are pathogens in our hospital today.

Ginny -   And can you show me how you do that screening?

David -   Yeah, sure.  We have here a microtitre plate and this is essentially loads of test tubes joined together and we put our compounds, our chemicals, in each of these individual test tubes, and we put bacteria on top of them, and we grow them over time.  The bacteria that aren't affected by the compound grow and become cloudy in solution and for the ones that are prevented from growing, there's a clear solution.

Ginny -   How many wells are there in each of those?

David -   In each plate, there's 386 wells.

Ginny -   Wow!  So you can look at 386 different possible drugs at the same time.

David -   That's correct.

Ginny -   Can you show me the computer system that actually tells you which ones have worked?

David -   Yeah, I can take you over there now.

Ginny -   Okay, so it's a big grey cupboard basically.

David -   We have a reader inside here - a plate reader - which will scan through 50 plates at once.

Ginny -   And then there's a computer next to it.  It's got lots of little circles on it and they're all different colours.  What does that mean?

David -   Each of the coloured spots represents an individual well in the plate.

Ginny -   Okay, so there's a load of blue spots down one side.  What do they mean?

David -   Well, that's where we don't have any bacteria in the well at all.  They just have the growth medium in them.  So, the blue means there's been no bacterial growth, so no cloudiness of the solution.

Ginny -   So, what you're looking for is one of the other spots that has bacteria in it, but has a different medium to be blue as well because that would mean the bacteria hadn't grown.

David -   That's correct.  At the moment, they're green in colour which means there's a lot of growth.

Ginny -   So, none of these would be possible candidates for a drug.

David -   This particular plate, no, not at the moment.  This is a bacterium called Pseudomonas.  This particular bacteria is a really difficult pathogen to treat.  It causes various blood infections and respiratory infections.

Ginny -   Although there aren't any here that might be candidates, you do often get them coming up.  How do you figure out if this is going to be a really effective antibacterial treatment or if it's just going to work only in some circumstances and not for very long?

David -   Well, we've got a very neat technology that we've developed.  We had to take some of these pathogens that are problems and engineer them.  We've engineered them in such a way that we can use them as mini computers to tell us what's happening to a compound when they're treated with it. 

A bacterium has a piece of DNA, its chromosome, which is 8 million base pairs long.  We put individual tags in different bacteria at every position in that genome and those tags are pieces of DNA that can disrupt genes or activate genes in the position that the tag has been placed. 

We incubate all of those bacteria with our compounds and without our compound.  After we've grown those bacteria, we sequence the bacteria that've gone into the experiment and the bacteria that've come out of the experiment. It's the difference between the mutants that we see that gives us the information on what the compounds are doing. 

So, it can tell us exactly what component inside the cell the compound is binding to and affecting, and it can tell us what all the potential resistance mechanisms could be for that compound.  That's really important because we don't want to develop new antibiotics that are going to develop resistance quickly in the clinic. 

We want to be working on compounds that can last a long time once they're on the market and in patients.

Ginny -   So, you're essentially trying to speed up what would happen in nature once people started being treated with this new compound because bacteria are very fast at evolving and can develop resistance quite quickly.  So, is that what your technique is doing, sort of trying to speed that up and see what would happen in real life?

David -   Yes, absolutely right.  We can do what nature does in 10 years in a week.

Ginny -   Brilliant!  So then you can pick out the ones that it's hardest for the bacteria to evolve resistance towards and focus on those for the future.

David -   That's exactly what we do and all of our compounds that we're working on for new drugs have those qualities in built.

Ginny -   You're talking about 8 million base pairs, 8 million different types of bacteria and then you're sequencing the genome for each of those.  There must be huge amounts of data.  Does this take days and days to do and to process?

David -   Three or four years ago, it was impossible to do this.  It's really the rapid advances in next generation sequencing technology that have enabled us to be able to do this.  This particular instrument can give us sequence information back from 200 million different sequencing reactions in just a few hours.  The information that we generate is huge.  I mean, we generate in a week, more information than has been generated in the whole London Stock Exchange in its history.

Ginny -   Wow!  Have any of your drugs reached clinical trials yet?

David -   No, we're relatively new company.  We've only been in the laboratory for just under 2 years.  For us to pursue any of our drugs to clinic, it is going to be a minimum time of 3 years...


Add a comment