It's Tough to Make Predictions, Especially About the Future
As I noted here at ENV on Monday, a recent paper confirms a key inference I made in 2007 in The Edge of Evolution. Writing in PNAS, Summers et al. 2014 conclude that "the minimum requirement for (low) [chloroquine] transport activity ... is two mutations." This is the second of three posts on the matter.
Actually, with apologies to Yogi Berra, on some topics it's not all that hard to predict the future -- like on the need for multiple mutations to get some selectable biological function. Way back in 1970, when thinking about protein sequences was avant-garde, the late, eminent theoretical evolutionary biologist John Maynard Smith worried about how they might evolve. He cast the problem in terms of a little game:
The model of protein evolution I want to discuss is best understood by analogy with a popular word game. The object of the game is to pass from one word to another of the same length by changing one letter at a time, with the requirement that all the intermediate words are meaningful in the same language. Thus word can be converted into gene in the minimum number of steps as follows:
WORD WORE GORE GONE GENE
Of course mutations come in many different flavors besides the simple substitutions that Maynard Smith considered, as I discussed in detail in The Edge of Evolution. Nonetheless, those are the kind that are relevant to chloroquine resistance. What's more, Smith's general point about the need to proceed one step at a time applies to all mutations. For example, if a selectable effect requires a certain gene duplication plus the deletion of a distinct region, that combination will occur much less frequently than if only one of those two steps were needed.
That having been said, not all words can be reached in this way. Maybe not all proteins either. Later theoretician Allen Orr also acknowledged the general restriction of evolution to one change at a time:
Given realistically low mutation rates, double mutants will be so rare that adaptation is essentially constrained to surveying -- and substituting -- one-mutational step neighbors. Thus if a double-mutant sequence is favorable but all single amino acid mutants are deleterious, adaptation will generally not proceed.
Yes, adaptation generally won't proceed, but in special circumstances it can. For example, one way to get past the one-mutation-at-a-time hurdle would be to increase those low mutation rates, as HIV does. Another way is to have enormous numbers of organisms to increase the chances -- an astronomical population size. That's out of the question for larger animals. But for single-cell organisms such as the malarial parasite (Plasmodium falciparum), it's doable. The distinguished Oxford University malariologist Nicholas White noted about the world malaria population that "in any 2-day period ... ill people would contain [a total of] between 5 � 1016 and 5 � 1017 malaria parasites." And:
Resistance to chloroquine in P. falciparum has arisen spontaneously less than ten times in the past fifty years. This suggests that the per-parasite probability of developing resistance de novo is on the order of 1 in 1020 parasite multiplications.
If you do the arithmetic, that astounding number of parasites would be produced on Earth over the course of perhaps every few years!
Now, why does it take such an enormous number of organisms to adapt when it takes far, far fewer to counter other malarial drugs? What might be the hurdle to developing chloroquine resistance? A decade ago, Hayton and Su 2004 had a good idea: "Based on the mutant pfcrt haplotypes known so far, it is likely that simultaneous multipoint changes in pfcrt are necessary to confer [chloroquine resistance]." So, from the known mutant sequences, from the math that shows adaptation can't "generally" proceed if multiple mutations are needed, yet with the enormous population available to originate resistance, it's just not that hard to predict the future. In fact Hayton and Su did predict it: multiple mutations would be found necessary for malarial parasites to adapt to chloroquine. Yogi Berra would have been pleased.
But some other people weren't pleased at all, not one bit -- mostly Darwinist reviewers of The Edge of Evolution, where I quoted all the folks above. Not only were they not pleased, they went into denial -- either ignoring it, pooh-poohing it, or denying it altogether. So we got the spectacle of Sean Carroll lecturing in Science:
Multiple replacements can accumulate when each single amino acid replacement affects performance, however slightly, because selection can act on each replacement individually and the changes can be made sequentially.
Well, that's the theory. But what if a necessary single amino acid replacement doesn't affect performance at all (because it's neutral), or "affects performance" by making it worse? (See my response to Carroll's review here.) Tell your ideas about "sequential changes" to the malarial parasites dying in droves because they have only one of the two required mutations for chloroquine resistance. Tell it to the people saved over the course of decades because chloroquine was still effective, unlike drugs such as atovaquone that barely work at all because they become ineffective after just one mutation. Next Kenneth Miller harrumphed in Nature:
Behe waves away evidence suggesting that chloroquine resistance may be the result of sequential, not simultaneous, mutations, boosted by the so-called ARMD (accelerated resistance to multiple drugs) phenotype, which is itself drug induced.
But if anyone was obfuscating and waving away evidence, it was Miller. (See my response to Miller's review here.) Then we heard from yet another partisan of the rabidly Darwinist National Center for Science Education (Carroll and Miller are also affiliated with it), Nicholas Matzke, writing in Trends in Ecology and Evolution:
Behe's two mutations do not always co-occur. As a result, [chloroquine resistance] is both more complex and vastly more probable than Behe thinks.
He's right that the same two mutations that I discussed in my book didn't always occur together. (See my response to Matzke's review here.) But that didn't mean two mutations weren't required, just that their identities hadn't yet been pinned down with certainty. Now they have, and we see why chloroquine resistance is vastly less probable than Matzke thinks.
And in the New York Times Richard Dawkins himself warned us not to put too much faith in arithmetic:
If correct, Behe's calculations would at a stroke confound generations of mathematical geneticists, who have repeatedly shown that evolutionary rates are not limited by mutation.
But surely something was limiting the rate of development of chloroquine resistance. (As I pointed out in my response, the calculations in The Edge of Evolution don't contradict any "mathematical geneticists" at all, simply because no one has ever tried to model the complex systems I described in anything like sufficient detail.) Now what could that be? As Maynard Smith took for granted, as Allen Orr calculated, and as Hayton and Su 2004 easily deduced from the data, if two changes were required, the rate of getting those mutations would be greatly limited. They were all right. Dawkins was all wrong.
The need for multiple mutations for chloroquine resistance in malaria just wasn't that hard to see. So why do you think Darwinist reviewers of The Edge of Evolution missed it?