How evolution helps us manage antibiotic resistance

Invited author blog by Rees Kassen

It will come as no sur­prise to most read­ers of this blog that anti­mi­cro­bi­al res­ist­ance, or AMR for short, is a major glob­al health chal­lenge. Every year since 2013 there have been glob­al warn­ings – from the World Eco­nom­ic For­um, the World Health Organ­iz­a­tion, and the United Nations Gen­er­al Assembly – that the world is at risk of enter­ing a post-anti­bi­ot­ic era. When the drugs don’t work we get sick­er more often and stay sick­er longer. Some­thing has to be done.

Ideally, we would reduce the glob­al bur­den of AMR by decreas­ing the quant­ity of drugs we use. The logic here is straight­for­ward. The wide­spread use of anti­bi­ot­ics in medi­cine and agri­cul­ture means that drug res­ist­ant types enjoy a strong select­ive advant­age and can quickly spread. So, by redu­cing the use of drugs we reduce the selec­tion pres­sure on res­ist­ance and slow the rate at which AMR evolves and spreads. Simple.

Except it’s not. How, pre­cisely, should we reduce the volume of drugs we use? We can’t just intro­duce a blanket ban on pre­scrip­tions and stop using drugs alto­geth­er. If my loved one is sick and there is a drug that can make them bet­ter, I want that drug. But maybe we could deliv­er our drugs in ways that would make it harder for res­ist­ance to evolve and spread, and so pro­long the amount of time a drug remains effect­ive? This was the ques­tion that launched our research.

One sug­ges­tion is to make use of drug sanc­tu­ar­ies – drug-free envir­on­ments – to effect­ively reduce the strength of selec­tion for res­ist­ance. If drug sens­it­ive types have a growth advant­age over res­ist­ant types when no drug is around, then sens­it­ive gen­o­types would pre­dom­in­ate in drug-free refuges and help keep res­ist­ant ones from tak­ing over the pop­u­la­tion. Drug sanc­tu­ar­ies could be used in hos­pit­als, for example, if dif­fer­ent wards restrict the use of dif­fer­ent drugs (gen­er­at­ing a form of spa­tial sanc­tu­ary) or by altern­at­ing the use of a drug on the same ward over time (a form of tem­por­al sanctuary).

But which of these forms of sanc­tu­ary, space or time, is more effect­ive at slow­ing the spread of res­ist­ance? It turns out that mod­els for how genes under selec­tion move through pop­u­la­tions when the envir­on­ment var­ies in time and space can guide us here (see Levene 1953, Demp­ster 1955, Fel­sen­stein 1976). Vari­ation in time is like pay­ing taxes: you can’t avoid it. The type that evolves in a tem­por­ally vary­ing envir­on­ment is the one that does best across all con­di­tions exper­i­enced. Spa­tial vari­ation, on the oth­er hand, provides more options because the dif­fer­ent patches can act as refuges for spe­cial­ized types that do well on that patch. Because those patches are always avail­able, then spe­cial­ists can coex­ist indef­in­itely. At least in theory.

What does the data say? Our recent work in Evol­u­tion Let­ters, which involved allow­ing pop­u­la­tions of the oppor­tun­ist­ic patho­gen Pseudo­mo­nas aer­u­ginosa to evolve in the pres­ence or absence of the com­monly used anti­bi­ot­ic, cipro­floxa­cin, and track­ing the emer­gence and fate of res­ist­ant gen­o­types along the way, was very clear: drug sanc­tu­ar­ies in space, but not time, can slow the spread of res­ist­ance, just as the the­ory pre­dicts. In space, res­ist­ant and sens­it­ive gen­o­types coex­is­ted because of a trade-off between res­ist­ance and growth in the absence of anti­bi­ot­ic: res­ist­ant types could not grow as well as sens­it­ive types in the absence of drug, and the reverse was the case when the drug was present. In time, a single, gen­er­al­ist gen­o­type evolved that was both res­ist­ant and grew well in the absence of drug. These res­ults are exactly what is expec­ted from the­ory and they tell us that drug sanc­tu­ar­ies in space could be effect­ive at man­aging drug resistance.

Above: Pseudo­mo­nas aer­u­ginosa cul­tures in Luria-Ber­trani media dur­ing our selec­tion exper­i­ment. Photo: Alanna Leale

But we also noticed some­thing else that was not expec­ted. Over time, the trade-offs that ensured coex­ist­ence between res­ist­ant and sens­it­ive types broke down. In fact, the res­ist­ant strains star­ted to gain muta­tions that improved their abil­ity to grow in the absence of drug. The res­ult was that, by the end of the exper­i­ment, most pop­u­la­tions in the spa­tial sanc­tu­ary were dom­in­ated by res­ist­ant strains. The take home mes­sage here is that drug sanc­tu­ar­ies in space may be effect­ive at slow­ing the rate at which res­ist­ant gen­o­types spread, they can­not pre­vent res­ist­ant strains from tak­ing over the entire pop­u­la­tion eventually.

evl343-5
P. aer­u­ginosa colon­ies. Photo: Alanna Leale

The pro­gnos­is for using drug sanc­tu­ar­ies as a form of evol­u­tion­ary man­age­ment of res­ist­ance is there­fore mixed. Drug sanc­tu­ar­ies in space do help: while they can­not pre­vent res­ist­ance from evolving in the first place, they can slow the rate at which it spreads through a pop­u­la­tion, at least rel­at­ive to drug sanc­tu­ar­ies in time. In the long run, how­ever, when selec­tion has had more time to do its work, the trade-offs that pre­vent res­ist­ant strains from tak­ing over can be broken down by muta­tion and lead to the evol­u­tion of a gen­er­al­ist type that is both res­ist­ant and grows well in the absence of drug.

Of course, our exper­i­ments are highly con­trived and restric­ted to one strain of one patho­gen evolving in response to one drug in the defined and con­trolled con­di­tions of a labor­at­ory. Wheth­er our con­clu­sions hold for oth­er bugs and oth­er drugs, or more com­plic­ated net­works of trans­mis­sion char­ac­ter­ist­ic of hos­pit­als and the com­munit­ies they are embed­ded in, remains to be seen. What our work allows us to do is get dir­ectly at the eco­lo­gic­al and genet­ic mech­an­isms respons­ible for the emer­gence, coex­ist­ence, and even­tu­al demise of diversity. Using these evol­u­tion­ary prin­ciples in more com­plex, real-world situ­ations to man­age our exist­ing arsen­al of drugs for as long as pos­sible, will be a major chal­lenge for evol­u­tion­ary bio­logy in the years to come.

 

The full study, by Alanna Leale & Rees Kassen, is freely avail­able to read and down­load from Evol­u­tion Let­ters.

Leave a Reply