Dinosaur Ancestor's Vision Possibly Nocturnal; Researchers Recreate 240-Million Year Old Protein In Test Tube

The pigment, rhodopsin, was recreated based on the scientists "inferring" its protein sequence.

Their findings, reported in the September issue of Molecular Biology and Evolution, offer the first look at a protein that has not been seen in 240 million years, and pave the way for scientists to study how the structure and function of vision pigments -- and ultimately other biologically important molecules -- have changed over the course of evolutionary time.

"Visual pigments trigger the critical first step in the biochemical cascade of vision in humans and other animals and obviously were present in now extinct species," says senior author Thomas P. Sakmar, M.D., head of Rockefeller University's Laboratory of Molecular Biology and Biochemistry.

"Recreating the inferred visual pigments of the archosaur ancestors in the laboratory should be a first step toward a better understanding what they could see -- and not see," adds Sakmar, an HHMI associate investigator.

In their journal paper, Sakmar and his colleagues report that archosaurs may have had a class of visual pigments that would support dim-light vision. "This is consistent with the intriguing though controversial possibility that nocturnal, not diurnal, life histories may have been the ancestral state in amniotes, which are birds, reptiles and mammals whose embryos are protected with a fluid-filled sac," says Belinda S.W. Chang, Ph.D., first author and research assistant professor at Rockefeller. "We are doing further biochemical studies on this recreated pigment to clarify this issue."

Chang turned to existing databases and employed sophisticated statistical methods to infer the most likely DNA sequences that the ancestral archosaur would have had for its rhodopsin.

"From the databases, we pulled rhodopsin gene sequences for such animals as dogs, rats, cows, birds, teleost fish, eels and amphibians. Then we aligned them," says Chang. "Using our knowledge of how these vertebrates are related to each other, the sequence alignment and a model of how often certain types of genetic changes occur over time, we calculated the most likely gene sequence."

In her calculations, she used maximum likelihood phylogenetic statistical methods.

Chang and her colleagues next took the inferred DNA sequence for archosaur rhodopsin and reconstructed a gene, which they then inserted into mammalian tissue cell cultures -- a standard method for producing rhodopsin in the lab. As expected, the gene instructed the cells to generate rhodopsin in the mammalian tissue.

But, did the protein structure and biological function of the artificially produced archosaur rhodopsin resemble "natural" rhodopsin? To answer this question, the researchers showed that it binds to a molecule called 11-cis--retinal, gives a characteristic absorption spectrum in the visible range and activates in response to light.

In order to see color, humans have three types of visual receptors, sensitive to red, green and blue light, but the primary molecule involved in all of these is a form of vitamin A called 11-cis-retinal.

"We found it does bind 11-cis-retinal and produces a very beautiful absorption spectrum with a maximal sensitivity at slightly red-shifted wavelengths when compared with our control in the laboratory, which is bovine rhodopsin," says Chang. "Although we don't know why the archosaur rhodopsin is shifted toward the red end of the spectrum, it is closest to the spectrum measured for bird rhodopsins."

The final piece of evidence that the researchers had produced a functional rhodopsin was that the activated form of rhodopsin triggered the rest of the signal transduction cascade in the photoreceptor cell; in other words, it interacted with the "second messenger," the G protein transducin.

"Using a technique that measures the increase of fluorescence in the G protein transducin upon activation, we found, strikingly, that the archosaur rhodopsin activated transducin in a similar time frame as bovine rhodopsin, which is very good at activating transducin," says Chang.

"Characteristics of rhodopsin determine characteristics of vision directly, so from this we can infer things about how archosaurs actually saw at night and under dim light conditions," adds Chang. "We can infer that their night vision was, at least on the level of their rhodopsin and its activation of G protein, basically as good as mammalian rhodopsin, which is surprising since mammals went through a nocturnal phase."

Sakmar's and Chang's co-authors are Manija A. Kazmi and Karolina Jönsson at Rockefeller and Michael J. Donoghue at Yale University.

This work was supported in part by the Howard Hughes Medical Institute, the Allene Reuss Memorial Trust, an NSF/Sloan Fellowship in Molecular Evolution and the Ellison Medical Foundation.