Some people fear them, others revere them, but there is no denying their place among nature’s iconic predators. Despite our collective fascination with snakes, their evolutionary origins and relationships to the other lizard families remain hotly debated in herpetology and palaeontology circles. Part of the challenge in untangling the evolutionary origins of snakes among the legged lizards has to do with their anatomy: snake skulls are highly modified relative to other lizards and their bodies are long and limbless.

However, as new discoveries of more ancient fossil snakes push their origins into the age of the dinosaurs, there may come a time where the lines between “lizard” and “snake” are so blurred, that we need different anatomical evidence to identify potential transitional fossils. What palaeontologists need are anatomical markers that are unique to snakes, conserved across living and extinct species, and that can be identified in fossils. One promising candidate that has interested me has nothing to do with their bodies or skulls, but is the way they replace their old teeth.

Snakes replace their teeth with new ones constantly, like most other reptiles, but where they differ is in the way they shed the old ones. If you picked up a snake skull in a museum collection, you wouldn’t know which teeth were new and which ones were about to be shed. Snake teeth show no signs of how old teeth are removed from the jaw. We know they do it all the time, but no one has seen how.

By comparison, lizards, crocodiles, and mammals have a replacement pit that forms when a new tooth begins to form. These replacement pits grow in response to the development of the new replacement tooth, eventually eating away at the old tooth and helping it to be shed from the mouth. This is why our baby teeth become loose when we are kids; the replacement tooth underneath causes the root of our baby tooth to dissolve away.

Snakes have many replacement teeth floating under their gums, but snakes lack these resorption pits. We wanted to identify the mechanism that allows them to shed their old teeth. If we could find the tooth shedding mechanism in extant snakes, maybe we could provide a new line of evidence, independent of the anatomy of the skull and body, to support the identification of fossil jaws as belonging to early snakes. This is the premise of our recent paper in Nature Communications, but the story really began with a chance discovery.

As a palaeontologist with an interest in teeth, I spend a lot of time looking at histological thin sections of teeth and jaws under the microscope. One day while working with a colleague on a project looking into the evolution of snakes’ venom fangs, I came across a strange-looking snake tooth in one of our slides.

This section showed that this snake’s tooth was being eaten away from the inside and that this might be the mechanism that allows snake teeth to be shed, a feature that has been hidden from view.

We then took a cells-to-fossils approach to snake tooth replacement. We first used a stain for labelling cells that remove tooth tissues (called odontoclasts) to track their positions in and around the teeth in thin sections of a corn snake. Because snakes continually replace their teeth, it’s possible to reconstruct the entire tooth replacement cycle from a single jaw. Catching teeth at various stages of this cycle helped us piece together the activity of the odontoclasts and reveal snakes’ secret for shedding their old teeth.

At some point in the life of each snake tooth, it undergoes a dramatic transformation: the inner pulp of each tooth becomes filled with large cells that begin to eat away at it from the inside. This process eventually weakens the tooth base enough to break it away from the jaw, allowing the new tooth to jostle into position and replace its predecessor.

We then compared tooth shedding in other lizards with that in snakes. It turns out that snake-type tooth replacement has no equivalent in other lizards, or any other reptile for that matter. But for this to be of any use for palaeontologists, it would need to be preserved inside the teeth of fossil snakes. To test this, we first used non-destructive CT scanning to look inside skeletons of extant species, not for seeing the cells, but the “bite marks” in the tooth tissues left behind by the odontoclasts that would have been resorbing the inside of the tooth.  

Amazingly, these “bite marks” are visible inside the teeth in CT scans. And, combined with some more histology data from other snake species, we were able to show that this unique form of tooth replacement is found all across the snake evolutionary tree. This also showed us what to look for in CT scans of fossil snake teeth, because these “bite marks” could preserve in fossils. We then looked inside the teeth of a primitive fossil snake from Australia, Yurlunggur…and we found them.

We then decided to take this a step further back in time and look inside the teeth of one of the world’s oldest snake fossils: a partial lower jaw of a poorly known snake from the Jurassic period of Portugal called Portugalophis. Together with our Portuguese colleagues, we were able to scan this fossil that was so old, it probably pre-dated the complete loss of limbs in snakes (though without complete body fossils of Portugalophis, we can’t be sure yet). Once again, we found good evidence for the same “bite marks” inside the teeth of this ancient fossil, revealing that this secret inside of snake teeth may be an ancient one, a feature that could help us identify putative snake fossils in the future, even from isolated jaws.

This was an ambitious project, but one of my favourites to work on. Not only did this start with a serendipitous peek down a microscope eyepiece (a stereotypical scientific discovery if I ever heard of one), but it eventually snowballed into a larger question around the origins of snakes that brought together talents from disparate fields, something that is becoming more common in 21^st century science.