Genes suggest that humans and all other mammals had insect-eating ancestors

Paleontologists have long noted how mammal fossils seem to change dramatically in form from one era to the next. In fact, few things in the fossil record seem to contrast more than the mammals that lived alongside the dinosaurs and those that survived after most dinosaurs went extinct some 66 million years ago.

One of the most obvious changes is that the Mesozoic (dinosaur era) mammals were generally very small and had teeth with sharp, pointed ends, both of which are characteristic of modern insect-eating mammals. Post-Mesozoic mammals, however, attained much larger body sizes and had teeth that were clearly used for grinding up plant material or slicing meat.

Why did mammals change? The typical thought among scientists is that when dinosaurs were the dominant land animals, they occupied nearly all of the herbivore and carnivore niches, restricting mammals to insect-eating and therefore small body sizes. Then when these dinosaurs went extinct, mammals were able to evolve and adapt to occupy these newly vacant ‘jobs’ within their respective ecosystems.

The last dinosaur
The extinction of most dinosaurs is thought to have allowed mammals to adapt and diversify

Together, this implies that all modern herbivores and carnivores, ranging from rhinos and cows to tigers and dolphins, descended from tiny, insect-eating ancestors. At some point, these ancestors began foregoing their insect prey in favor of plants and/or meat, and their bodies changed in form and function along with these newfound diets. To help you imagine the scope of what this means, picture the small insectivorous Mesozoic mammal below (top left), being the greatest of grandmothers to the herbivorous mammals in the images alongside it.

My colleagues and I wondered if there might be any evidence of this pattern, implied by the fossil record, in the genomes of modern day mammals. We looked at a gene encoding an enzyme called acidic mammalian chitinase (CHIA), which is produced in the stomach and other digestive organs and can break down the chitin-rich exoskeletons of insects [1]. We reasoned that if modern day herbivores and carnivores evolved from insect-eating ancestors, then they might have remnants (pseudogenes) of this gene in their genome.

Interestingly enough, we found that there are actually five chitinase genes, not one, and our analyses suggest that all five are common to the largest group of mammals (i.e., placental mammals). This implies that there were five in the common ancestor of these mammals, which we think lived in the Mesozoic alongside dinosaurs. Interestingly, modern mammals that have all five chitinase genes are all strongly insectivorous, suggesting that this ancestor was primarily an insect-eater too, consistent with the fossil evidence.

Chitinase gene family
Phylogeny showing the similarity between the different chitinase genes (closed circles) and pseudogenes (open circles). Colors and silhouettes correspond to different groups of mammals.

Furthermore, we found that in modern-day meat-eaters and plant-eaters, every species we examined has pseudogene remnants of one or more chitinase genes. So herbivores, like sloths, elephants, manatees, fruit bats, horses, rhinos, camels and rabbits, and carnivores, such as tigers, polar bears, walruses and dolphins, have remnants of genes that were likely once used to digest the prey of their ancestors: insects!

Chitinase pseudogenes
Chitinase gene remnants in herbivores and carnivores. Arrows indicate disabling mutations shared by two or more species.

What’s more is that many of these mammals that are thought to be related, based on DNA and anatomical similarity, share the exact same mutations in some or all of the chitinase genes. For example, horses and rhinos look quite different from one another, yet have been considered by anatomists, paleontologists and geneticists to be related for over a century and a half. They are both herbivores, which implies their diet was inherited from a common ancestor, and indeed the earliest fossils that resemble them had teeth and jaws that appeared to be particularly good at eating plants. Such adaptations would have rendered insect-digesting genes relatively useless. Indeed, horses and rhinos have pseudogene remnants for four of the five chitinases, and they share at least one disabling mutation in each gene, suggesting they inherited defunct copies from a plant-eating ancestor.

Humans also possess three chitinase pseudogenes, alongside a single functional chitinase gene. Two of the pseudogenes share the same inactivating mutations with monkeys and apes, and a third shares a mutation only with apes. Interestingly, many of the earliest primate fossils appear to have been insect-eaters, but as monkeys and apes appeared, plants, as well as meats, became more important for their diets, so insect-digesting genes were likely less useful.

Together, these data suggest that in our very own genomes, we retain ‘molecular fossils’, that hearken back to a time when our distant ancestors were not the top of the food chain, but rather scurried along amidst dinosaurs, eating insects.

Questions for Creationists

Why would the Creator design humans, rhinos, tigers and other mammals that never or almost never eat insects with remnants of insect-digesting genes? Is it just a coincidence that the earliest mammal fossils, found alongside dinosaurs, appear to have been insect eaters, and modern herbivores and carnivores have remnants of insect-digesting genes? For those that believe that all animals were plant-eaters in the Garden of Eden, why do so many herbivores have remnants of insect-eating genes? If mammals were created in the last 10,000–6,000 years, how could they evolve from insect-eaters to herbivores and carnivores so quickly, modifying their teeth, jaws, intestinal tracts, etc. to be optimized for their new diets?

References

1. Emerling, C. A., Delsuc, F., & Nachman, M. W. (2018). Chitinase genes (CHIAs) provide genomic footprints of a post-Cretaceous dietary radiation in placental mammals. Science advances, 4(5), eaar6478.

11 thoughts on “Genes suggest that humans and all other mammals had insect-eating ancestors

  1. g

    hi christopher. a question: how its possible that after so many years (say 50my) their sequence is almost identical?

    2) if many of these gene were lost by convergent loss isnt it a good evidence against a single common descent?

    3) is it possible that these gene had other functions?

  2. Hello again, g! Numbered answers bellow correspond to your numbered questions:

    (1) I assume you’re referring to the second figure? They really aren’t almost identical, except for species that we expect to be very closely related (e.g., chimps and humans). The figure is just showing a very small portion of the gene, but if you saw more of it, you’d see how different these sequences can be, due to (scientists believe) evolution.

    (2) Could you please clarify your question?

    (3) Yes, most definitely possible, though there’s only one function I can think of that’s very plausible. Functional versions of all five genes share a very similar gene structure (e.g., 11 exons, all exons same length except last one) and encode a protein that has a chitinolytic domain (a motif that indicates a capability to break chitin) and a chitin-binding domain. As such, we think these are all true chitinases, though this has yet to be verified in vitro/vivo (with the exception of the originally described chitinase, which we call CHIA5). Based on our analyses, we also now know that all five can be expressed in digestive tissues (e.g., stomach, pancreas, salivary glands). Plus there’s a positive correlation between the number of chitinase genes and the degree of insects/invertebrates in the diet. As such, we think that these were/are likely largely, if not primarily, involved in digestive functions.

    But, we do know that at least one of these chitinases (CHIA5) can be expressed in other tissues, such as the lung in humans and mice. What seems likely is that its turned on there to help protect the body from pathogenic fungi, since fungi have chitin in their cell walls.

    1. g

      hi. as for 1 or 3- ok. thanks.

      as for 2) if i got it right this gene get lost several times in different species. am i right?

  3. Hi Chris
    Are all 5 genes functionally equivalent? Do monotremes have any copies? (i think the common ancestor with placentals also ate insects) For animals that have just one copy could it be the case they only express it in the lungs and not the digestive system? Are there any animals that dont have any functional copies? Dolphins? ( whales of course would have it.)

    1. Hi Rod, thanks for commenting!:

      1. Functionally equivalent meaning they do the exact same thing?

      2. We kept our study limited to placental mammals, but a quick look showed that at least some marsupials have five chitinase genes, and the platypus has at least one. The platypus genome assembly is pretty shoddy, so I would wait for a better assembly or one of an echidna before looking too much into it.

      3. So far, gene expression analyses have been pretty darn limited, both in terms of # of species and # of tissues. Every mammal (that I can recall) that has had its stomach gene expression checked for the chitinase gene shows that its presence, except the cow. The problem is that (a) the cow has a complicated stomach, so I’m not sure which part they looked at, and (b) I don’t believe they looked at other digestive tissues. The gene is expressed in the liver of the cow though.

      4. Yes, there are several mammals we checked that apparently have zero functional copies: African elephant, West Indian manatee, three species of fruit bats, tiger, African wild dog, polar bear, Weddell seal, walrus, ferret, rhino, donkey, okapi, flying lemur, aye-aye, blue-eyed black lemur, Coquerel’s sifaka, Angola colobus monkey, pika, rabbit, 13-lined ground squirrel, two African mole rats, guinea pig, chinchilla, degu, beaver, jerboa, and a blind mole-rat. All four whales we looked at, including a bottlenose dolphin, had 1 intact gene.

      1. Hi Chris,
        Thanks for the reply!

        1. Yes, meaning they do exactly the same thing. If they do it would be hard to explain why the Intelligent Designer would make 5 when 1 would suffice. If there is some slight different substrate specificity that would take detailed biochemical analysis to detect but I suppose there might be patterns to which copies are lost in different groups depending on which copy they needed for their diet. If all copies are equivalent there would be no pattern. Are there any extra domains or motifs in any versions?
        4. Does that mean you can detect pseudogenes for all copies? I saw that there are only 4 copies total for humans. I assume a pseudogene or functional copy was deleted. Can we use synteny to see where the deletion occurred? Where other genes lost in the deletion?
        Since we share 2 pseudos with monkeys and apes the third must have occurred independently and its a coin toss which functional we have. Do we have the same one as monkeys and apes?

      2. I hate to admit it but I’m having trouble reading the phylogeny. Am I correct that humans are missing Chia1 and that our active copy is Chia5? It seems to me that apes also have functional Chia4 but you say thats not the case in the article. Is it monkeys that have Chia4?

        1. Hi Rod:

          1. Yes, meaning they do exactly the same thing. If they do it would be hard to explain why the Intelligent Designer would make 5 when 1 would suffice. If there is some slight different substrate specificity that would take detailed biochemical analysis to detect but I suppose there might be patterns to which copies are lost in different groups depending on which copy they needed for their diet. If all copies are equivalent there would be no pattern. Are there any extra domains or motifs in any versions?

          There are some differences for some of them, though we’re not yet sure what the functional implications are yet (this will probably be up to some biochemists to check out; we might collaborate with someone to do it, but just in the talking stage now). The main difference that we were able to detect is a difference in something called a hinge region. This is a sort of flexible region that allows for some independence between the chitinolytic domain and the chitin-binding domain. This length is variable, and there seems to be a tendency for similar paralogs to have similar lengths/motifs. For instance, CHIA1 does not appear to have a hinge region at all, whereas the other four do, some tend to be longer than others, etc.

          I will say that it is possible that an Intelligent Designer created five (or they evolved such that they had five) if having five allows you to produce more chitinase, and therefore digest insects more efficiently. My guess is that some are more specialized for different tissues, because different parts of the gastrointestinal tract have different pHs, different enzymes that they’re interacting with (e.g., different proteases that may interacts with chitinases), etc. We did find some evidence of tissue specialization, but we didn’t look at nearly enough tissues or species to feel confident about this. Much more work needs to be done!

          “4. Does that mean you can detect pseudogenes for all copies? I saw that there are only 4 copies total for humans. I assume a pseudogene or functional copy was deleted. Can we use synteny to see where the deletion occurred? Where other genes lost in the deletion?”

          No, not necessarily for all copies. Depends on the species. Rabbits, for instances, have pseudogenes for all five copies, whereas at the other extreme, for some rodents (e.g., beaver) you can only find one. We were able to use synteny to demonstrate that in quite a few cases whole gene deletion was likely. No evidence that we could see other genes were lost in the deletion.

          “Since we share 2 pseudos with monkeys and apes the third must have occurred independently and its a coin toss which functional we have. Do we have the same one as monkeys and apes?”

          Some New world monkeys have both CHIA4 and CHIA5. Old world monkeys tend to have CHIA4 intact, whereas apes have CHIA5 intact. Another study that came out right before ours looked at more Primates and found that there was some variation in Old World monkeys (note they use a different naming system that we did): https://academic.oup.com/mbe/article-abstract/35/3/607/4693806

          “I hate to admit it but I’m having trouble reading the phylogeny. Am I correct that humans are missing Chia1 and that our active copy is Chia5? It seems to me that apes also have functional Chia4 but you say thats not the case in the article. Is it monkeys that have Chia4?”

          No worries! The figure is mostly supposed to demonstrate the overall clade patterns, not necessarily to get into the nitty gritty. If you find the supplementary materials for the paper (follow the link of this page) you’ll see a more detailed figure with the species names (Figure S3).

          But yes, humans possess a functional CHIA5, and CHIA1 was deleted based on synteny data.

  4. Ronald Myers

    People can eat insects and it is probable that some subsistence cultures have done so. I also see my cats chase insects and presumably eat them. It would also be useful for a herbivore to be able to digest insects which are eaten plants. So some insect digestion may be useful even for non insectivores.

    1. Thanks for your comment, Ronald! We think it’s likely that the chitinases help free more tissues for digestion, but by pure mechanical digestion almost certainly a hungry cat can gain access to some non-chitinous tissues. For the record, humans appear to have one functional chitinase remaining, so we still have that capability!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s