Deoxyribonucleic acid (DNA) is susceptible to DNA damage and one form of DNA damage is caused by reactive oxygen and nitrogen species (RONS). RONS can oxidize guanine (G) into 8-oxo-7,8-dihydro-2’-deoxyguanosine (OG), which on its Hoogsteen face mimics thymine. During replication, adenine (A) can be placed across from OG resulting in a transversion mutation. To prevent the accumulation of mutations, the GO repair pathway has base excision repair enzymes to either cleave out OG across from cytosine (C) or A from OG:A base pairs. The enzyme responsible for the removal of A from OG:A base pairs is MutY in bacteria and Homo sapiens MutY, MUTYH, in humans. MUTYH-associated polyposis (MAP) is a genetic disease that results in early onset of colorectal cancer (CRC) due to inheritance of MUTYH variants. The decreased activity and function of MUTYH variants leads to overabundance of somatic mutations and inactivation of tumor suppressor genes, leading to CRC.This thesis reviews two different MutY enzymes, the first is Geobacillus stearothermophilus (Gs) MutY to determine the effects of cancer variant R241Q and the second is Lactobacillus brevis MutY (LbY) to begin to understand the nature of this “clusterless” MutY. Both enzymes will be reviewed mechanistically and structurally. MutY has several motifs of interest including the iron-sulfur (FeS) cluster and the FSH (Phe-Ser-His) loop. The FeS cluster is coordinated by four highly conserved cysteine (Cys) residues. Neighboring these four Cys are four arginine residues (Arg) in which previous research proposes that these Arg hydrogen bond with the conserved Cys coordinating the FeS cluster. To understand if alterations to one of these Arg could affect enzymatic activity and binding, cancer databases were searched for related mutations for all four of the Arg. Several mutations were found for each corresponding Arg. The variant studied in this thesis was found in a patient with MAP and breast cancer, R241Q. Preliminary unpublished A glycosylase assays demonstrate binding of R241Q but no activity. Enzymatic studies of other MAP variants have been done previously in Gs MutY, so the corresponding mutation, R149Q, was tested and was found to have activity. Previous research demonstrates the importance of the FeS cluster in DNA repair and recognition, however LbY does not contain the FeS cluster coining it as a “clusterless” MutY. Also, LbY does not have a serine (Ser) in the FSH loop but rather a threonine (Thr). In the FSH loop of Gs MutY, Ser308 hydrogen bonds with OG, but the structural differences between the hydroxymethyl of Ser and the secondary hydroxyl group of Thr could mean that LbY recognizes OG differently than Gs MutY. The goal of this thesis is separated into two parts: first, to determine the impact of the Arg149 to Gln149 mutation in Gs MutY to the hydrogen bond network, the active site, and the FeS cluster and second, to optimize the purification of LbY for structural studies as well as beginning to understand the role of Thr in the FTH loop.
In this thesis, R149Q Gs MutY was successfully crystallized with DNA duplexes containing OG opposite of tetrahydrofuran (THF), a product analog, and purine (P), a substrate analog. The biochemical results demonstrated decreased activity in R149Q compared to wild-type (WT) Gs MutY and even lower activity with OG:P. This suggested the potential for a product structure with R149Q OG:P and the product structure was captured. The best fit sugar pucker conformation of the product was 3’-exo-beta, which is found to be the most stable conformer of the oxocarbenium ion intermediate during A cleavage. Analysis of both crystal structures found that the effects of Arg mutation to Gln are dramatic: there are alternate conformations of the FeS cluster and the supporting Cys, a calcium ion is found coordinating with Asn146, and Gln149 can no longer coordinate the DNA backbone. Due to the shorter side chain of Gln149, a Ca2+ ion fills that space and coordinates Asn146. Asn146 forms hydrogen bonds with Asp144, so to maintain these connections Asp144 was found further away from the substrate in both R149Q-OG:AP, the product structure, and R149Q-OG:THF compared to the Transition State Analog Complex structure. These results illustrate the potential supportive role of these Arg residues near the coordinating Cys of the FeS cluster.
LbY was successfully purified in this thesis, but crystals could not be obtained. Further crystal conditions will need to be tested for LbY. Previous research has demonstrated the importance of the FSH loop by amino acid substitution studies, which resulted in decreased MutY activity. Previous research found that Ser hydrogen bonds with N7 and O8 of OG, however due to the different side chains of Ser and Thr, Thr may not be able to maintain these interactions with OG. To begin to understand OG recognition in LbY MutY, biochemical experiments were conducted with OG analogs, 8-thioguanine (8SG), 8-bromoguanine (8BrG), and 7-methyl-8-oxo-7,8-dihydroguanosine (7MOG), to see how MutY activity is affected by changes to N7 and O8 positions of OG. The OG analogs, G, 8SG, and 8BrG have modifications to the O8 position, but 8SG had the fastest rate of glycosylase bond cleavage of these three. G and 8BrG had similar rates of glycosylase bond cleavage which could be due to the modifications of the O8 position. 7MOG had the lowest rate in activity which could be due to the bulkiness of the methyl group and potential steric interactions with Thr316. The size of 8SG and 8BrG could also affect substrate recognition due to the size of the substrate and the side chain of Thr316.
This thesis furthers our understanding of the FeS cluster, expanding it beyond the coordinating Cys to the potential Arg near Cys, and illustrates the potential differences in OG recognition due to the FTH loop. Further research could investigate the other Arg residues in this area to see if they have similar effects to the active site and the FeS cluster. Additionally, continued attempts to crystalize LbY for structural analysis will increase the understanding of the FeS cluster area and the FTH/FSH loop comparison.