My friend Andrew sent me a nice Reddit thread with the charmingly long title “It’s hypothesized that warm-bloodedness evolved in mammals and birds because it provided defense against fungal infections. Very few fungi can survive the body temperatures of warm-blooded animals. By comparison, insects, reptiles, and amphibians are plagued by fungal infections.” The title is, it turns out, a quote from the Wikipedia page “Warm Blooded,” quoting in its entirety, the subsection “Defense against fungi.” This deep cut subsection cites not one, but four papers by one Arturo Casadeval of Johns Hopkins School of Public Health. One wonders if this subsection is brought to us by Arturo himself - he does, after all, have a voluminous biographical Wikipedia page.

So far as I understand it, the fungal-mammalian hypothesis goes like this (I am paraphrasing from Arturo’s 2012 PLoS Pathogens paper here): there is a geological signature called the K-T or K-Pg boundary where mammals abruptly replace reptiles in the fossil record. This event occurred about 66 million years ago and is usually hypothesized (with strong evidence as I understand) to be due to the impact of a very large asteroid that excited the dinosaurs - the so-called “Alvarez impact hypothesis” (Wikipedia, NYer). Anyway, Arturo finds this hypothesis troubling. He wonders: why didn’t the dinosaurs re-establish themselves after the asteroid dust settled. So he offers another hypothesis: mammals are more resistant to fungi than reptiles are. Most fungal pathogens of mammals (e.g. Candida albicans) are what we call “opportunistic pathogens” - pathogens that infect when the host is otherwise weakened. Casadeval gives a particular explanation for mammalian resistance to fungi that I’d like to take issue with. A even deeper cut, if you will.

Casadeval argues that part of the reason that fungi are so much less effective at infecting mammals is that mammals regulate their body temperature while reptiles do not. More specifically - He draws on literature about contempary fungi that infect birds and bats to argue temperature regulation acts synergistically with the mammalian immune system in some fashion to specifically enhance resistance to fungi. The problem is that there are plenty of fungi that can survive at 37 C or above. 37 C is, by the way, 98 F AKA human body temperature. Here’s a nice review of thermophillic fungi that grow happily at 45 C and above. Candida albicans is a famous human pathogen that is a fungus (a yeast) and brewers yeast grow fine at 37 C (we usually grow it at 30 C in the lab).

Three of my friends in Rachel Brem’s lab, Carly, Jeremy & Rylee, just published a great paper dissecting the divergence in thermotolerance between different yeast species. Brewers yeast (S. cervisiae) grows decently at 39 C and its close relative S. paradoxus does not. They showed that this difference was attributable to the loci (coding sequence + promoter region) of a very small number of genes that could be swapped into S. paradoxus and enable it to grow nicely at 39 C. So temperature-dependendence of growth does not explain any particular advantage that mammals might have over reptiles in defending against fungi - the fungi can adapt to mammalian body temps. Now you might argue (paraphrasing Jeremy here) that there are 5 million years between S. cerevisiae and S. paradoxus, so it might take 5 million years to adapt to mammalian body temperatures. But 5 million years is not that long in comparison to the 66 million years between us and the dinosaurs and the acquisition of thermotolerance is not the only thing that happened over that 5 million yr period - S. paradoxus differs from S. cerevisiae at many more than the 6-8 genomic loci identified in the Carly-Jeremy-Rylee paper. Moreover, laboratory experiments have been pretty successful at improving yeast thermotolerance over about 1 week and, Jeremy points out, clinical isolates of S. cerevisiae are pretty common and come from multiple genetic backgrounds, suggesting that opportunistic pathogenesis evolves pretty quickly in some fungi.

The Alvarez impact hypothesis posits that an asteroid struck the Earth and wiped out all the largest animals due to ecosystem collapse. My understanding is that the evidence for a large asteroid is quite good and the ecologists have good mathematical models predicting that apex predators would die in that situation. I should read that literature someday. Arturo wonders why the reptiles didn’t take over again after the Earth ecosystem re-stabilized, but we know from the pioneering work of Richard Lenski (and from intuition and mathematical models) that population evolution exhibits strong historical contingency - random events early in the evolutionary trajectory (e.g. small mammals occupying a particular niche) can have very big effects on the later trajectory (e.g. mammalian takeover of the Earth ecosystem). Not that we don’t deserve Jurassic park, but that we shouldn’t necessarily expect any particular outcome from a natural evolutionary process operating on mixed populations. Which is not to say that mammals have no advantage against fungi, or that fungi had nothing to do with mammalian dominance of Paleogene period, but that we should be skeptical of just-so stories about evolution and the temperature-based explanation in particular.

Which reminds me of another just-so story about Earth history and evolution that I learned about recently. Apparently there is a very large glut of coal in the geological record starting about 300 million years ago. This marks the beginning of the so-called “carboniferous” period for the obvious reason. Various studies have also inferred that the atmospheric CO2 concentration dipped around that same time. For a long time people hypothesized that the onset of coal accumulation was due to the evolution of lignin - the very stable polymer that makes up woody plant material. The idea was that lignin evolved but the fungal enzymes that today degrade lignin did not evolve for a few million years after that. So in that gap of ~100 million years a lot of coal was buried which then would explain why we find all this coal in those strata today. There are several problems with this story though: (1) many coal-forming plants of the carboniferous period did not make lignin (2) lignin biosynthesis is inferred to pre-exists the carboniferous period (3) there are several present-day lignin degrading enzymes, some of which have been inferred to predate the carboniferous as well. These issues are detailed very nicely in this paper (Nelsen et al. PNAS 2016) and a very clear commentary on that same paper by Isabel Montañez. Anyway, as usual, neat ideas for predicting evolution are often not so accurate, especially at the planetary scale.