Credit: Courtesy of Eric Olmon and Pamela Sontz. DNA strand modeled using model.it (http://hydra.icgeb.trieste.it/dna/). XPD modeled bound to DNA by Li Fan. AFM image generated with WSxM software (Horcas I, et al. 2007). Scene generated with PyMol and POV-Ray.
Caltech Study Supports Model for Locating Genetic Damage Through DNA Charge Transport
PASADENA, Calif.—Our genetic information is under constant attack—not only from outside sources such as UV radiation and environmental toxins, but also from oxidative stress, the production of highly reactive forms of oxygen, within our bodies. Luckily, repair proteins are typically hard at work, locating and fixing damaged DNA. Over the past decade, Caltech chemist Jacqueline Barton has been exploring a model that describes how repair proteins might work together in this scouting mission to efficiently home in on lesions or mismatches within the DNA.
Essentially, the model suggests that two DNA-bound repair proteins can use DNA like a wire to shuttle electrons between themselves—a process called charge transport. When one protein receives an electron from the other, its affinity for the DNA to which it clings decreases, causing the protein to fall off that strand. If instead, a lesion—a structural defect in the DNA—prevents that electron from being transferred between the proteins, both members of the pair remain bound to the DNA and begin inching toward the problem area. Since this signaling can be achieved over long molecular distances, the model could explain how it is that the relatively few repair proteins in our cells are able to scour so much genetic information to efficiently locate problems.
Providing support for that model, Barton's team has recently shown that two different repair proteins that are part of two different genome-maintenance pathways preferentially reposition themselves onto damaged strands of DNA. Publishing in the journal Proceedings of the National Academy of Sciences (PNAS), the group recently reported that repair proteins relocate in this way only if they have the capability to participate in charge transport. In the study, proteins with mutations that destroy their charge-transport capabilities could not zero in on DNA lesions. Versions of those mutated proteins found in humans are associated with diseases such as colorectal cancer, Cockayne syndrome, and xeroderma pigmentosum, suggesting that charge transport may indeed be a necessary part of genome maintenance.
"These findings are consistent with and provide data to support the model of facilitated search for lesions through DNA-mediated signaling between proteins," says Barton, the Arthur and Marian Hanisch Memorial Professor of Chemistry and chair of the Division of Chemistry and Chemical Engineering at Caltech. "This provides another piece of the puzzle."
In their study, the researchers investigated XPD, a protein involved in both DNA repair and replication. First, the scientists attached very short strands of DNA to a gold electrode, added the XPD, and used the electrode to measure the protein's electrical potential, or its ability to send or receive electrons. Separately, the chemists made a solution of the protein along with both regularly matched strands of DNA and longer strands that included a mismatched pair of nucleotides, which are the individual chemical units that make up DNA. Then they used microscopy to visualize and count the number of proteins that bound themselves to the different types of DNA. They found that only the proteins that were able to send and receive electrons through the DNA repositioned themselves in the vicinity of the mismatched nucleotides.
"We believe that the redistribution comes from two proteins using charge transport to communicate with one another, and falling off of the strands that don't have a lesion and attaching to the strands that do," says Pam Sontz, lead author on the study and a graduate student at Caltech.
In a previous study, the Barton lab had conducted similar experiments with Endonuclease III (EndoIII), a repair protein that removes damaged bases. They found that like XPD, EndoIII redistributes itself so that it can home in on mismatches in DNA, and that mutant forms of EndoIII that cannot participate in charge transfer do not relocalize.
In the new study, the researchers were also able to show that mixtures of EndoIII and XPD were able to coordinate in order to relocate onto the mismatched strands of DNA. The team was working with EndoIII from the bacteria Escherichia coli and XPD from a microorganism called Sulfolobus acidocaldarius.
"Our findings suggest that these two proteins are able to signal one another in order to zero in on a lesion," Sontz says. "They're from different DNA-repair pathways and from totally separate organisms, which is really neat. That's what's really opening the door for a lot of future studies. If EndoIII and XPD can do this, there are probably many other proteins from a variety of organisms that are also able to send or receive charge through DNA."
Coauthor and graduate student Tim Mui adds that this kind of cooperation between tested proteins that are normally completely isolated from one another could indicate that the mechanism has been conserved by organisms across evolutionary history. "It's kind of remarkable that these proteins, which should never ever be in contact with one another, can actually coordinate to do this," he says.
The key to this coordination seems to be that the two types of proteins have comparable electrical potentials when bound to DNA, meaning that they are similar in their likelihood to gain or lose electrons. "We're finding that one of the requirements for a protein to potentially participate in this process is that as it binds to DNA, it has to have a similar potential to other proteins, so that it can release an electron that can shuttle through the DNA to another protein," Sontz says. "If the two proteins are not at similar potentials, you're not going to get accurate cooperation."
The proteins have this ability to send and receive electrons because they contain what are called redox-active iron-sulfur clusters. Both XPD and EndoIII contain four-iron, four-sulfur clusters that play no clear structural role in the proteins but contain metals that can easily gain or lose an electron as they bind DNA. "It's really difficult for the cell to build these clusters," Mui says. "So there is a thought that they must play a significant role in something else, which could be this mechanism to locate lesions within the DNA."
Written by Kimm Fesenmaier