First author Aaron Rozelle and coauthors in the Lee lab have found a new role for oxidative stress in creating interstrand cross links (ICLs), a form of DNA damage that binds opposite DNA strands together . While oxidized forms of adenine and guanine (oxoA and oxoG) are prevalent in cells, whether or not they could form ICLs was unknown. Previously, only thymine radicals were known to crosslink under oxidative conditions. Aaron’s work, out now in Nature Communications, shows that oxoA, but not the more prevalent oxoG, can form ICLs, filling in an important gap about how oxidative damage can affect DNA integrity.
Oxidative stress impacts cellular health by creating reactive oxygen species (ROS) that can react with everything in the cell, including DNA bases. Normal metabolic activity, as well as disease states like cancer and inflammation all produce ROS. These attack DNA bases, favoring purines due to their lower redox potential. Adenine and guanine bases become oxoA and oxoG which in turn have even lower redox potential than their unmodified precursors. This leads to further reactions and damage to the bases themselves. OxoG has taken the limelight due to it’s higher occurrence, leaving oxoA understudied. While it’s known that oxoG tends to accumulate further damage and is capable of protein-DNA crosslinking, no ICLs have been isolated from either oxoA or oxoG.
Because oxopurines occur a lot in cells, and the base position on oxoA where an ICL could form was shown to react with other nucleophiles, Aaron set out to see if oxoA could cause DNA damage via ICLs. ICLs are highly toxic — the crosslinked DNA becomes a problem when strands need to separate during replication and transcription. ICLs are very toxic to cells, so they’re mainly studied in vitro. Plus, trying to oxidize DNA in cells causes a ton of other damage (reviewer 2, if you’re reading this, stop asking Aaron to induce ICLs in cells, he’s tried). We do know that they happen in nature, though, because ICL repair pathways are conserved and cancer cells can evolve resistance to ICLs. ICLs can form in response to a few types of cellular stress, but only one pathway was previously known for oxidative damage causing ICLs, via thymine radicals. Could oxopurines also cause ICLs?
Aaron tackled the oxidative ICL project with the pragmatic approach that was driven hard as an engineering undergrad at North Carolina State. From the beginning of his project, he mocked up a problem tree of experiments, results and alternative approaches to follow depending on each result. He starts by informally outlining with pen and paper (he had just cleaned up his stacks of paper notes related to this project when we met) and then makes a more polished digital version with complete hypotheses and diagrams that he can use in slides or send to his PI. Aaron cites the planning routine as a therapeutic and grounding practice. This is the biggest way he’s grown over the last five years as a grad student.
“I don’t really think I’m any better or worse at science—like lab work—than I was in the beginning,” Aaron said, “but I’ve gotten much better about the process of outlining things and kind of working through problems.”
Science doesn’t always follow plans though. Aaron learned this during his second year, when his original dissertation project—trying to make new synthetic crosslinking reagents—hit a dead end. None of them yielded significant amounts of ICLs. Aaron realized he needed to find a more promising project, so he started paying careful attention to lab meetings and thinking more deeply about his labmates’ work. He read more and noticed where there were opportunities to branch off of others’ work with a new perspective and learn something new.
He pivoted to the oxopurine ICL project, picking up on preliminary work from the second author of the paper. There was promise that oxoA could do cool chemistry—a previous paper showed that oxoA could take on modifications at the same position that ICLs potentially could form, but those authors never showed it was capable of crosslinking. Despite this possibility, hopes of synthesizing oxoA cross links in the lab were low because previous attempts generated tiny yields. Aaron decided to tweak the existing DNA sequence and try again, and changing one base did the trick. He generated a high yield of ICLs on the first try with the new sequence, demonstrating that oxoA, but not the more prevalent oxoG, can crosslink with multiple different bases to create many different ICLs.
Aaron noticed that the ICLs at oxoA could form quickly, which helped make a case for biological relevance even though pulling oxoA ICLs out of living cells wasn’t feasible. The timescale he saw matches up with the idea that oxoA ICLs can form within the timeframe DNA is exposed to the environment, such as during replication. Together, the known prevalence of oxoA and the ease with which it forms ICLs in high yields makes a case for oxoA ICLs to occur in vivo.
Aaron’s outlining his next few months now, mapping out what he needs to get done before he’s out of here. The acceptance of this paper that was years in the making, plus the finish line of graduation being almost in reach has put Aaron in a reflective state of mind. That, and he recently celebrated his 30th birthday. He’s noticed a shift in himself: less ego driven and more comfort with himself and his contributions to science.
“I’ve stopped viewing myself as better than people and just kind of different than people, and part of that is being happy with what I have done,” Aaron said.
He’s grappled a bit with his source of motivation as he’s matured and relies less on comparison and competitiveness to drive his work. He’s landed on a more altruistic motivation strategy.
“It sounds kind of cheesy, but the good that something will do…I care more about the impact of something than I used to,” he said.
(1) Rozelle, A.L., Cheun, Y., Vilas, C.K. et al. DNA interstrand cross-links induced by the major oxidative adenine lesion 7,8-dihydro-8-oxoadenine. Nat Commun 12, 1897 (2021). https://doi.org/10.1038/s41467-021-22273-2
Leave a Reply