Metalloproteins' Ranks Swell

Metalloproteins may be more numerous and diverse than previously suspected, researchers report (Nature, DOI: 10.1038/
nature09265).

Proteins often use metals as cofactors and catalysts, but predicting which proteins incorporate which metals is difficult because of the variety of metal-binding sites found in proteins.

“There is no routine method to analyze a complement of metalloproteins,” says Nigel J. Robinson of Newcastle University, in England, who was not involved with this study. “The challenge is an important one because nearly half of the structurally characterized enzymes in the Protein Data Bank need metals, yet it is not possible to predict with certainty which proteins will use which metals.”

Indeed, most metals associated with proteins are discovered “after the fact” as people study particular proteins, says study leader Michael W. W. Adams of the University of Georgia, Athens. “We used the reverse approach,” he says. “Rather than purifying proteins and seeing what metals they contain, we purified metal peaks and then tried to see what proteins were associated with the metal.”

The work is part of a current Department of Energy-funded collaborative project by researchers at Lawrence Berkeley National Laboratory, Scripps Research Institute, and the University of Georgia to investigate the nature of microbial macromolecular complexes that can’t be predicted from genome sequences. In the current study, Adams and coworkers used inductively coupled plasma mass spectrometry (ICP/MS) to detect metals associated with proteins from the microorganism Pyrococcus furiosus. They purified the metals and their associated proteins through multiple chromatography steps, using ICP/MS to detect the metals and tandem mass spectrometry to identify the proteins.

In their analysis, a single peak corresponded to one metal but could contain more than one protein. “We took as liberal an approach as we could in trying to assign a protein to a particular metal peak,” Adams says. For example, in analyzing a nickel peak, if any protein in that peak was in the same family as another known nickel-binding protein, the team assumed that that protein bound the nickel.

Even with such generous assignments, almost half (158 out of 343) of the metal peaks could not be assigned to known metalloproteins. The researchers selected eight of the peaks for further purification, two each of molybdenum, nickel, uranium, and lead. They were able to identify new molybdenum- and nickel-binding proteins. The uranium and lead were present at substoichiometric amounts and most likely represent misincorporation of those metals, Adams says.

Even such misincorporations could reveal something about how the microbes process metals. “The detection of interactions between proteins and elements such as uranium, which are unlikely to have a role in the physiology of organisms, is also important in understanding how cells manage unwanted metals,” comments Ivano Bertini of the University of Florence, Italy. “This observation may be crucial to shed light on the mechanisms of metal toxicity and on the cellular strategies for detoxification.”

Just as surprising as what they found was what they didn’t find. “Some elements are notably absent,” Robinson says. “There has been a debate about the recruitment of copper by soluble cytoplasmic proteins in prokaryotes. The absence of any copper proteins here might support a view that it is rare.”