Thursday, November 20, 2014

K13 and the spread (or simultaneous emergence) of drug resistance in malaria parasites

We’ve mentioned this before, but the malaria and evolution story is complicated by multiple evolutionary tales:

  • Humans adapt to parasites
  • Parasites adapt to humans
  • Mosquitoes adapt to both

Parasites may adapt to mosquitoes too – and humans have adaptations to mosquitoes…

Malaria parasites bursting from red blood cells.  From National Geographic, June 1986 - scanned and shared online by Centuron:
This post is a story about parasites developing responses to some of the things we do to get rid of them.  Malaria parasites appear to have a real knack for survival, or at least the ones that survive and spread do.  Time and time again they have developed immune responses to our antimalarials.  Sometimes it happens quickly, sometimes it seems to take decades, but each time a new antimalarial is used, parasite strains emerge that are resistant to that antimalarial drug.

Southeast Asia appears to be a “special” place with regard to the evolution of antimalarial resistance.  For whatever reason, parasites that are resistant to new antimalarials always seem to be first documented here and then sometimes appear to subsequently spread globally.  (See Klein 2013 for a nice review of some theory around this problem (2)).  For example, chloroquine resistance in falciparum malaria seems to have independently emerged in both South America and Southeast Asia, but then seems to have spread globally from Southeast Asia (3).

Plasmodium falciparum parasites are almost globally resistant now to all antimalarials except for artemisinin.  In an attempt to keep these drugs effective, there has been a huge push to only use them in combination with other antimalarials.  The mechanisms of action of most antimalarials aren’t well understood, but the hope has been that by using different drugs, with different half-lifes, and probably different modes of action, then it will be much more difficult for parasites to develop resistance when compared to monotherapy (only using a single drug).

However, despite these efforts, artemisinin resistance has emerged in Southeast Asia (4).  It is not normally complete treatment failure at this point, but rather increased clearance times.  For example, while it would once take at most two days for parasites to be cleared from a patient’s blood stream after taking a dose of artemisinin, it now can take five.  Occasionally the treatment doesn’t work at all.   This is even occurring with artemisinin combination therapy.  Strains of parasites with “reduced sensitivity” have been found in Cambodia, in part of Vietnam, and along both sides of the Thailand-Myanmar border.

Some work has attempted to understand the genetics behind artemisinin resistance but many results, including a few I’ve been a part of, have contradicted each other.  However, one region on the parasite’s chromosome 13 keeps popping up in analyses.   Earlier this year, mutations in a particular gene (Kelch 13 (K13)-propeller) were identified as being potentially important in artemisinin resistance.  The function of this gene in the parasites isn’t well understood, but it is related to protein interactions.  And it isn’t a single point mutation that seems to confer resistance.  It appears that a wide variety of mutations, any of which are occurring in this gene, lead to parasites that are less sensitive to artemisinins – and this has now been confirmed both in vitro and in vivo.

The in vitro portion of this work began with a lab strain of falciparum malaria (3d7) which was intermittently exposed to artemisinins over a period of five years(5).  Doses of the drug were applied, then removed, then applied at higher proportions over this period of time.  Parasites from each dose cycle were sequenced so that the origin of mutations could be documented and so that mutations could be compared between case and control strains.  Ultimately the researchers narrowed their search down to a mutation in a single gene that corresponded to a point in time where some of the lab parasites seemed to no longer have strong, negative reactions to the antimalarial.

[It is important, I think, to remember that drug resistance isn’t usually an all or nothing type of trait, it is much more a trait of degree.  Even in situations where an antimalarial no longer works, it is likely that by increasing the dose of that antimalarial, there will be a point at which the parasites are still sensitive.  The problem is that it also becomes toxic to the human at some point.]

Next the researchers began looking at field isolates, across space and time, in Southeast Asia.   While they didn’t always find the same point mutations, they did find mutations in the same gene, in geographic areas where parasites are known to be less sensitive to artemisinins.  In areas where parasites still appear to be sensitive to the drug, they did not find mutations in this gene.  Furthermore, the prevalence of these mutations appears to have increased in certain regions (the ones that now have artemisinin resistance) over time.  

These findings are interesting I think for several reasons.
Here we have a gene in which mutations are somehow related to artemisinin resistance in malaria parasites.  But there isn’t a single mutation that leads to this resistance phenotype – rather it seems that just about any mutation(s) in this “gene” leads to resistance.  Does that make this a gene for resistance?

Another major finding, this time from a paper that came out in September 2014 (6), is that these mutations may not be spreading in the same way that other resistant strains (like chloroquine resistant falciparum malaria, for example) seem to have.  By analyzing the flanking regions of the K13 gene, analyzing patterns of linkage disequilibrium, the authors noted that several mutations in the K13 gene appear to have emerged independently and almost simultaneously both in Cambodia and along the Thailand-Myanmar border.

Once again the implications are quite interesting, if also scary.
One is that the evolutionary response seems less rare and unique if it can happen independently and simultaneously in different regions.  Does this mean that combination therapy is not working the way we hoped it would?

Another is more directly related to public health.  Right now there are several small scale elimination attempts occurring throughout Southeast Asia.  In fact, I’m working with one of the teams doing this (briefly discussed here).  Our hope is that we can wipe out resistant strains before they spread (via mosquitoes or humans) to other regions – perhaps especially Africa.  If resistance is likely to evolve anywhere that artemisinins are being used, we may not be able to halt this spread.  I would argue that our intentions to eliminate malaria in targeted subregions are worthwhile regardless.  But, it is a bit scary nevertheless.  

*** My opinions are my own!  This post and my opinions do not necessarily reflect those of Shoklo Malaria Research Unit, Mahidol Oxford Tropical Medicine Research Unit, or the Wellcome Trust.  

1. Network MGE. Reappraisal of known malaria resistance loci in a large multicenter study. Nat Genet. 2014;46(11):1197–205.

2. Klein EY. Antimalarial drug resistance: a review of the biology and strategies to delay emergence and spread. Int J Antimicrob Agents [Internet]. Elsevier B.V.; 2013 Feb 7 [cited 2013 Mar 8];41(4):311–7. Available from:

3. Payne D. Spread of chloroquine resistance in Plasmodium falciparum. Parasitol Today [Internet]. 1987 Aug;3(8):241–6. Available from:

4. Dondorp A, Nosten F, Yi P. Artemisinin resistance in Plasmodium falciparum malaria. New Engl J Med J … [Internet]. 2009 [cited 2013 Nov 17];455–67. Available from:

5. Ariey F, Witkowski B, Amaratunga C, Beghain J, Langlois A-C, Khim N, et al. A molecular marker of artemisinin- resistant Plasmodium falciparum malaria. Nature. 2014;505(7481):50–5.

6. Takala-harrison S, Jacob CG, Arze C, Cummings MP, Silva JC, Khanthavong M, et al. Independent Emergence of Artemisinin Resistance Mutations Among Plasmodium falciparum in Southeast Asia. J Infect Dis. 2014;491:1–10.


Anonymous said...

Calling K13 a artemisinin resistance gene is a bit like calling CCR5 a HIV resistance gene. Just about any inactivating mutation will confer this resistance. Perhaps K13 allows artemisinin to enter the cell and disrupt some process, but the malaria parasite can survive without K13.

Ken Weiss said...

The question here would be what sorts of mutations have been found. Are they all inactivating mutations? Or are some changes of other types, like regulatory. Or of unknown effect (at least unknown from informatic analysis)? Most 'disease' genes are found to have many different mutations in different patients, and only a few are clearly understood in terms of what they do to the protein. So knowing what's been found so far would be important and probably many people are working on this?

Ken Weiss said...

If I correctly read the Ariey et al paper in Nature last January, that Dan cited, there are several mutations in the 'propeller' part of the K13 gene, that were separately tested for artemisinin resistance. I think each case includes just one such mutation, not multiple ones in the same copy of the gene.

If my reading is correct, what is found is that several different codon-changing mutations are associated with some quantitative differences in parasite clearance. The differences vary by mutation, and they overlap normal to some extent. The different mutations do not (apparently) inactivate the gene as "anonymous" assumes, but in differing ways destabilize the protein. This seems consistent with Dan's description of the data.

The quantitative aspects probably aren't the same as inactivating mutations, but whether that's a major difference, a point that quantitative vs qualitative effects could be useful in applying treatment or prevention approaches etc. is something I have no idea about.

That the authors show evidence of selection certainly suggests that this is an important part of the protein and that the evolutionary reconstruction makes sense.

Multiple independent, regionally different genetic pathways to something strongly selected is what we should expect, is what the malaria story has previously shown (globins, G6PD, etc), and seems to show how challenging this all still is.

So this is a very interesting post!

Daniel M Parker said...

Hi Anonymous,
I mostly agree - my question about whether K13 is a resistance gene was a rhetorical question.

Ken Weiss said...

Doesn't the evidence that the variants seem (at least in the Nature paper, which listed several criteria) to be related to rapid adaptation, and malaria is a highly plausible cause given the pattern, make the phrase 'resistance gene' at least sensible (in the same sense that one calls BRCA a 'breast cancer' gene)? In this case, what is selected is K13's de facto resistance functions.

Daniel M Parker said...

Hi Ken,
Put that way, yes. But typically when people are talking about a gene "for" something, I think there is an assumption that the genes function is directly related to it. That is, there tends to be a perception that the gene is intended to do something specific - probably what the gene is normally associated with. In this situation at least its not that way at all. For whatever reason, when there is a mutation in this gene, it happens to help the parasite out when it is exposed to a specific drug. So perhaps what I mean, but didn't really flesh out, is that the concept of a gene being "for something" can be a bit strange given that a gene can be related to multiple outcomes. But perhaps even the concept of a gene as some discrete, easily definable entity is also a bit strange.

Anne Buchanan said...

I agree, Dan. I always go back to genes 'for' spermatozoa... expressed all over the developing embryo. That is, given pleiotropy, perhaps the best one can say is, 'this variant of this gene does this in this context." Whether or not the full context (genomic background/environment) can ever be known is another question.

Ken Weiss said...

Yes, I completely agree. I didn't mean to suggest that the gene evolved because of malaria. However, from what you've written and what I've seen elsewhere, the gene is functionally related to the biology of malaria, even if there's no reason to assume it evolved because of malaria. It's sort of what Gould and Lewontin called a 'spandrel' in the sense that, if I understand things, it exists for some biological function in the 'normal' organism but took on a new function in the presence of the parasite.

Anne Buchanan said...

Oh, but look! There's a 'happy gene'!