Synchrotron microtomography reveals fish fossil's heart

Innovative use of advanced technology assisted research on the anatomy of Rhacolepis buccalis, an extinct fish that lived over 100 million years ago in waters covering Chapada do Araripe, Ceará State, Brazil.

By Peter Moon  |  Agência FAPESP – The innovative use of X-ray synchrotron microtomography reveals hitherto imperceptible aspects of specimens and samples, creating opportunities for entirely novel types of scientific research.

A fossil of Rhacolepis buccalis, a fish that swam over 100 million years ago in the waters that covered Chapada do Araripe, a site now in Ceará State, Brazil, offers a good example.

The fossil was studied by a research group led by José Xavier-Neto at Brazil’s National Bioscience Laboratory (LNBio) in Campinas, São Paulo State, and Vincent Fernandez, a French scientist at the European Synchrotron Radiation Facility (ESRF). The specialist in fish anatomy was Brazilian ichthyologist Murilo Carvalho, also affiliated with LNBio. Carvalho’s research was supported by FAPESP via the project entitled “Molecular evolution of regulatory regions of HOX genes associated with the morphology of fish fins, with special emphasis on Chondrichthyes”.

The findings have been published by eLife in the article entitled “Heart fossilization is possible and informs the evolution of cardiac outflow tract in vertebrates”.

The story of how 62 Brazilian fish fossils from Araripe were sent to France to be scanned using microtomography begins with Xavier-Neto’s interest in the evolution and biology of the multichambered heart.

All vertebrates have a multichambered heart. In mammals and birds, the heart has four chambers – two atria and two ventricles – through which blood is pumped. Amphibians and reptiles have hearts with three chambers, while fish have two-chambered hearts.

Invertebrates have much simpler cardiac structures, without chambers but with peristaltic pumps: the blood is pumped by alternating compression and relaxation, similar to the kneading movements of the intestine.

“No living vertebrate has a heart that can be considered a transition between the peristaltic pump and the multichambered heart,” Xavier-Neto said. “But it’s possible that an animal with this transitional profile once existed and became extinct. That’s why we decided to investigate the fossil record to see if we could find fossilized hearts. We found none. No one even knew if it was possible for a heart to fossilize.”

Vertebrates consist of hard parts, such as the skeleton, and soft parts, such as the internal organs, muscles, cartilage, fat, skin, fur or hair, feathers, and scales. The vast majority of animal fossils only preserve the hard parts. Exceptional circumstances are needed for soft parts to fossilize. Sites with the right conditions for this to happen are very rare. One such site is Chapada do Araripe, a high tabular plateau with a massive escarpment on the eastern side, located on the border between Ceará and Pernambuco States.

Between 119 million and 113 million years ago, in the Cretaceous, Araripe was the bottom of a shallow lagoon. Its ancient tropical waters teemed with fish. Dinosaurs, crocodiles, turtles and an infinity of insects inhabited the nearby coastal plain. And through the air roamed pterosaurs, extinct flying reptiles that have made Chapada do Araripe world famous.

Although pterosaurs are the stars of Araripe, the geological formation was discovered in the mid-nineteenth century thanks to the abundance and excellence of its fish fossils, which are among the most complete and perfectly formed of their kind. Xavier-Neto sought out these fossils in annual trips to the northeast of Brazil.

With government permission (from DNPM, the National Department of Mineral Production), he began taking dozens of fish fossils to Campinas, choosing the most complete and least deformed. His preference for research purposes was Rhacolepis buccalis, a relatively small fish spanning only 10 cm in length.

At LNBio, Xavier-Neto dissolved some fossils in acid, layer by layer, to investigate the internal structures of the species. “In the third fossil, I found a structure that caught my attention,” he said. “It was cone-shaped. This is the characteristic shape of the heart in fish.”

Computed tomography imaging was not successful, so he turned to the National Synchrotron Light Laboratory (LNLS) in Campinas. A synchrotron is an extremely bright light source used by scientists to determine the structural and chemical properties of materials at the molecular level.

When submitted to synchrotron microtomography, the fish fossils revealed their inner structures, but the images were blurred and indistinct. It was impossible to determine whether the third specimen did indeed contain a fossilized heart, let alone how many chambers it had. The X-ray beam produced by the Brazilian synchrotron, which has been in service since 1997, only penetrates a few micrometers into any material.

The solution was to turn to a more powerful synchrotron. The fossils were sent to the ESRF in Grenoble, France. The facility’s mighty beamline produced images with a resolution of 6 micrometers, a thousand times the resolution of medical tomography.

Achieving success at last, the French synchrotron revealed a wealth of details with “magnificent” image quality, according to Xavier-Neto. “We were able to see all the internal structures very distinctly,” he said. “We could even determine the species of shrimp the fish had eaten.” It was a shrimp species from Araripe that had already been described.

“When the images arrived from France, we saw the anatomy of a 113-million-year-old fish. And we obtained proof that the conical structure we had glimpsed was indeed the fish’s heart,” Carvalho continued.

“But, it wasn’t just any fish heart. It didn’t have the single outflow valve seen in most fish today. To our utter astonishment, this conical structure in the Rhacolepis specimen from the age of the dinosaur had five valves. There are no records of any other animal, living or extinct, with a heart like that,” Carvalho said.

Evolutionary transition

Xavier-Neto’s hypothesis is that an evolutionary transition occurred early on in the evolution of fish species, from an original heart with tens of valves to a heart with only five valves, as in the Rhacolepis fossil. The number of valves then decreased until most species had a single valve, as in modern fish. “The number of valves must have decreased gradually,” Carvalho said.

To determine whether there are indeed hearts with tens of valves, researchers need to study other fossils of different ages and belonging to other groups.

Identifying such fossils is no easy task, as they must have specific features, such as three-dimensional conservation of soft tissue. Another difficulty is access to fossils. Most of the material in these conditions is outside Brazil.

“It’s important to do this research because now we know we can use fossils to study the evolution of the heart. Moreover, the knowledge that synchrotron microtomography can literally perform a virtual ‘autopsy’ on a fish fossil opens up the possibility of investigating the internal structures of all the fossils that have a conserved record of their soft parts,” Xavier-Neto said.

“There are thousands of fossils in these conditions from Araripe alone, not to mention the material from fossil-bearing deposits that conserve magnificent remains of dinosaurs and birds in China, mammals in Germany, and even fish in Australia.”

For now, Xavier-Neto and Carvalho will have to continue sending their fossils out for analysis in France. That will change in 2019, when LNLS is due to start operating Sirius, Brazil’s fourth-generation synchrotron, currently under construction at Campinas with FAPESP’s support. “Sirius will be the most powerful synchrotron in the world when it comes on stream,” Carvalho said.

The article entitled “Heart fossilization is possible and informs the evolution of cardiac outflow tract in vertebrates” (doi: 10.7554/eLife.14698) by José Xavier-Neto et al., published in eLife, can be read at elifesciences.org/content/5/e14698v1.