Most proteins are thought to fold into a single active shape. But the human immune protein XCL1 is a rare breed that can switch back and forth between two different structures, each with its own function. A recent study in Science elucidates the evolutionary history of XCL1, concluding that it evolved from an ancestor with a single active conformation about 350 million years ago. The authors used a combination of biochemistry and nuclear magnetic resonance (NMR) to synthesize XCL1’s ancestral proteins in the lab, and to determine when the protein’s ancestors became metamorphic.
The finding sheds light on mysterious “shapeshifting” protein behavior, which was first discovered about a decade ago, says lead author Acacia Dishman, a structural biologist and MD-PhD student at the Medical College of Wisconsin in Milwaukee. The vast majority of known proteins have just one thermodynamically stable conformation, the product of the natural folding of their amino acid sequence into secondary, tertiary, and even quaternary structures. The idea that every protein has just one stable structure is a pillar of protein design, a field that’s exploded in the last few decades. Shapeshifting designs could have wide-ranging pharmaceutical and biomedical applications in coming years, including protein-based therapeutics or switchable drug delivery vehicles, Dishman says. “You might want a pharmaceutical with two functions, or that can switch from on to off. They provide fast and reversible switching,” she says.
In 2008, XCL1 was discovered to be among the first proteins with two thermodynamically stable structures. That discovery raised an interesting question, Dishman says: Why do metamorphic proteins exist? Some researchers hypothesized that the metamorphic proteins were “caught in the act of evolving,” she says—in other words, they were anomalous “snapshots of an evolutionary intermediate.”
This latest study contradicts that hypothesis. Dishman and her doctoral advisor, senior author, structural biologist and professor of biochemistry Brian Volkman, set out to trace the evolutionary history of XCL1. Together with coauthors, they reconstructed the protein’s most likely family tree using phylogenetic software and about 450 amino acid sequences from XCL1’s close modern cousins in the chemokine protein family. Once the researchers predicted the amino acid sequences of XCL1’s ancestors, they then synthesized those ancient proteins in the lab. The authors used NMR spectroscopy to scan each protein sample, revealing a set of peaks on the spectra which correspond to the positions of amino acids in each protein. Proteins with one active conformation had one set of peaks, while two sets of peaks indicated two active conformations, meaning a metamorphic protein. The taller set of peaks indicated the more abundant protein conformation in each metamorphic protein sample.
XCL1’s earliest ancestor 350 million years ago was not a shapeshifter; it only had one active conformation according to the NMR. The first metamorphic ancestor appeared about 150 million years ago. It had a similar set of NMR peaks to the modern XCL1 protein, suggesting it shapeshifted between a similar set of conformations. Volkman compares the shape of the more ancient conformation to a single unpaired sock, which binds a receptor on white blood cells and traffics them to sites of infection. The second, more recently evolved conformation is a dimer, which looks like a pair of socks folded together, Volkman says, and this different structure encodes a different function: the protein breaks and disrupts bacterial and fungal membranes, making it antibacterial and antifungal.
When the first metamorphic ancestor appeared, it mostly existed in the single sock form, with a small concentration in the pair-of-folded-socks form, according to the NMR analysis. However, several million years later, another ancestor appeared, called “Ancestor 4” in the study. NMR spectra of samples of Ancestor 4 turned up a much higher concentration of the folded socks conformation. Finally, NMR analysis of XCL1 showed that it naturally exists at about 50% single-sock and 50% folded-socks forms.
The pattern of protein evolution does not support the hypothesis that XCL1 is just a passing anomaly. Rather, the findings suggest that the shapeshifting is “a phenotype that’s being selected for,” Dishman says. A broader implication, she adds, is that “there might be a lot more shapeshifters out there in the world than we ever imagined before.” Roughly 100 such proteins have been discovered to date.
Searching for many more new fold-switching proteins is a clear next step, says Lauren Porter, a structural bioinformaticist at the National Library of Medicine in Bethesda, MD, who was not involved in the new work. She says researchers probably haven’t found more metamorphic proteins because existing methods to predict protein structure assume there is only one thermodynamically stable conformation.
As studies like this one reveal more about the evolution and biology of metamorphic proteins, biologists will also have more fodder to design fold-switching proteins in the lab. Researchers could theoretically design such proteins to change their conformation and function in response to environmental cues, Porter says—an elegant and efficient solution to accomplishing two tasks with a single protein.
“So far, protein design has focused on proteins that adopt one folded state,” says biophysicist David Baker at the University of Washington in Seattle, who was not involved in the recent work. His lab is now collaborating with Volkman’s group to design such proteins.
Other recent papers recommended by Journal Club panelists: