DNA repair of mutagenic lesions is imperfect, allowing mutations to accumulate unevenly across the genome. In base excision repair, glycosylases must locate rare damaged bases embedded in diverse sequence contexts, yet how these contexts shape recognition and mutational outcomes remains unresolved. Here, we introduce a high-throughput approach that quantifies glycosylase binding across thousands of lesion-containing sequences. Focusing on the cytosine deamination pathway, we map the recognition landscapes of human UDG, TDG, and MBD4. Binding depends strongly on sequence context, extending several bases beyond the lesion and including non-additive interactions between neighboring positions. Structural analyses and molecular dynamics simulations implicate DNA-shape features, including minor groove width, as determinants of recognition. Nearest-neighbor preferences resemble deamination-related cancer mutational signatures, whereas broader-context preferences track variation in cytosine-thymine balance across matched human genomic contexts. Together, these findings establish a versatile and generalizable platform for decoding glycosylase recognition and linking repair specificity to mutational patterns.

Proteins contain small pockets that form due to imperfections in residue packing or the rotational and conformational movement of amino acids. In this work, we used the fluorescent probe 1-aminoanthracene (AMA) to detect a small ligand pocket in the hydrophobic core of human proliferating cell nuclear antigen (PCNA), which is a critical protein for DNA replication and repair. Fluorescence measurements of AMA reported that the core of PCNA had a dielectric constant (ε) of 4, which was very apolar and similar to cyclohexane (ε = 2). Protein mutagenesis, photoaffinity labeling, and molecular dynamics simulations localized the binding site for AMA next to PCNA residues L90 and L101, which also interacted with general anesthetics (sevoflurane and propofol). The ligand binding site was cryptic, i.e., it formed transiently and was only detectable in certain structural states of PCNA. Ligand binding to the cryptic site on PCNA structurally stabilized the trimeric protein and reduced its ability to disassemble and reassemble its subunits. Thus, the cryptic site in PCNA’s core serves to destabilize the assembled protein and promotes structural and oligomeric flexibility. Finally, the hydrophobic site is widely conserved among homologous β clamp proteins with a similar fold as PCNA. This work highlights how small fluorescent probes can reveal ligand sites within proteins, defines the chemical features of protein hydrophobic cores, and introduces a novel approach to modulate the oligomeric stability of PCNA.

Because of their ubiquitous presence, ions interact with numerous macromolecules in the cell and affect critical biological processes. Here, we discuss how cations including Mg2+ alter the enzymatic activity of a DNA glycosylase by tuning its affinity for DNA. The response of uracil DNA glycosylase (UNG2) to Mg2+ ions in solution is biphasic and paradoxical, where low concentrations of the ion stimulate the enzyme, but high concentrations inhibit the enzyme. We analyzed this phenomenon by modeling experimental data with a statistical framework that we empirically derived to understand molecular systems that display biphasic behaviors. Parameters from our statistical model indicate that DNA substrates are nearly saturated with cations under ideal conditions for UNG2 activity. However, the enzyme slows rather abruptly when the ionic content becomes too low or too high due to changes in the electrostatic environment that alter protein affinity for DNA. We discuss how ion occupancy on DNA is dependent on DNA length; thus, the sensitivity of UNG2 to cations is also dependent on DNA length. Finally, we found that Mg2+-induced changes in DNA base stacking and dynamics have minimal effects on UNG2, as these outcomes occur at ion concentrations that are much lower than is required for efficient enzyme activity. Altogether, our work demonstrates how cation-DNA interactions, which are likely common in the nucleus, are a key determinant of uracil base excision repair mediated by UNG2.

The compounds pemetrexed and 5-fluorodeoxyuridine (FdU) are widely used for cancer therapies and disrupt cell proliferation by inducing DNA damage and stressing DNA replication. The drugs disrupt pyrimidine nucleotide metabolism and promote the accumulation of uracil bases in genomic DNA, which are repaired by uracil DNA glycosylase (UNG2) and downstream base excision repair proteins. UNG2 interacts with Proliferating Cell Nuclear Antigen (PCNA) and Replication Protein A (RPA), which localize to the replication fork during DNA damage responses to orchestrate DNA repair. In this work, we tested whether UNG2 requires interaction with PCNA and RPA to repair DNA damage in a colorectal cancer model during treatment with pemetrexed or FdU. We genetically knocked out UNG2 in HT29 cells and engineered the cells to express UNG2 variants that cannot bind to PCNA or RPA. We found that eliminating UNG2 activity or disrupting its interaction with PCNA or RPA sensitized the cells to the DNA-damaging effects of pemetrexed and FdU. The ability of UNG2 to localize to stalled replication forks was impaired when the enzyme could not interact with PCNA or RPA. Finally, disrupting the interaction of UNG2 with PCNA or RPA sensitized the cells to the cytotoxicity of the drugs. We concluded that certain cancers may be sensitized to pemetrexed and FdU by directly inhibiting the enzymatic activity of UNG2, by depleting UNG2 levels in the cell, or by impairing UNG2 function by inhibiting its protein-protein interactions.

Proliferating cell nuclear antigen (PCNA) is a conserved eukaryotic DNA sliding clamp that is essential to DNA metabolism. PCNA interacts with hundreds of DNA replication and repair proteins, and inhibition of clamp function compromises genomic integrity and cell survival. Here, we discovered a novel cryptic binding site on PCNA that interacts with general anesthetics and their fluorescent analog, 1-aminoanthracene, which has environment-dependent properties. The fluorescence of 1-aminoanthracene blue-shifted 80 nm and increased 200-fold upon PCNA binding. 1-aminoanthracene was competed off PCNA by propofol and sevoflurane, indicating the site binds more than one class of general anesthetic. LC-MS/MS identified two residues on PCNA (L90, L101) that interacted with a photoactive propofol analog. Molecular dynamics simulations (>1 μs) confirmed that a cryptic site of suitable volume for the ligands (213 Å3) was formed adjacent to L90, L101 and neighboring residues, confirming the mass spectrometry results. Additionally, mutagenesis of L90 or L101 to polar residues red-shifted the fluorescence of 1-aminoanthracene relative to its emission when bound to wild-type PCNA. Characterization of the cryptic site properties, and experiments to determine the functional relevance of ligand binding, are ongoing.

Sirtuin-2 (SIRT2) is an NAD+-dependent deacylase that removes acetyl and fatty acyl modifications from lysine residues. Here, we used chemical crosslinking, size exclusion chromatography, and crystallographic analyses to demonstrate that SIRT2 efficiently dimerizes in solution at concentrations greater than ∼100 nM. We also used a split-GFP system to show that SIRT2 dimerizes in cells. SIRT2 undergoes a dimer to monomer transition when bound to long fatty acyl groups such as myristoyl-lysine, but SIRT2 remains a dimer in the presence of acetyl-lysine.

Human sirtuin-2 (SIRT2) has emerged as an attractive drug target for a variety of diseases. The enzyme is a deacylase that can remove chemically different acyl modifications from protein lysine residues. Here, we developed a high-throughput screen based on a homogeneous time-resolved fluorescence (HTRF) binding assay to identify inhibitors of SIRT2’s demyristoylase activity, which is uncommon among many ligands that only affect its deacetylase activity. From a test screen of 9600 compounds, we identified a small molecule that inhibited SIRT2’s deacetylase activity (IC50 = 7 μM) as well as its demyristoylase activity (IC50 = 37 μM). The inhibitor was composed of two small fragments that independently inhibited SIRT2: a halogenated phenol fragment inhibited its deacetylase activity, and a tricyclic thiazolobenzimidazole fragment inhibited its demyristoylase activity. The high-throughput screen also detected multiple deacetylase-specific SIRT2 inhibitors.

Human sirtuin isoform 2 (SIRT2) is an NAD+-dependent enzyme that functions as a lysine deacetylase and defatty-acylase. Here, we report that SIRT2 readily dimerizes in solution and in cells and that dimerization affects its ability to remove different acyl modifications from substrates. Dimerization of recombinant SIRT2 was revealed with analytical size exclusion chromatography and chemical cross-linking. Dimerized SIRT2 dissociates into monomers upon binding long fatty acylated substrates (decanoyl-, dodecanoyl-, and myristoyl-lysine). However, we did not observe dissociation of dimeric SIRT2 in the presence of acetyl-lysine. Analysis of X-ray crystal structures led us to discover a SIRT2 double mutant (Q142A/E340A) that is impaired in its ability to dimerize, which was confirmed with chemical cross-linking and in cells with a split-GFP approach. In enzyme assays, the SIRT2(Q142A/E340A) mutant had normal defatty-acylase activity and impaired deacetylase activity compared with the wild-type protein. These results indicate that dimerization is essential for optimal SIRT2 function as a deacetylase. Moreover, we show that SIRT2 dimers can be dissociated by a deacetylase and defatty-acylase inhibitor, ascorbyl palmitate. Our finding that its oligomeric state can affect the acyl substrate selectivity of SIRT2 is a novel mode of activity regulation by the enzyme that can be altered genetically or pharmacologically.

Human uracil DNA glycosylase (UNG2) is an enzyme whose primary function is to remove uracil bases from genomic DNA. UNG2 activity is critical when uracil bases are elevated in DNA during class switch recombination and somatic hypermutation, and additionally, UNG2 affects the efficacy of thymidylate synthase inhibitors that increase genomic uracil levels. Here, we summarize the enzymatic properties of UNG2 and its mitochondrial analog UNG1. To facilitate studies on the activity of these highly conserved proteins, we discuss three fluorescence-based enzyme assays that have informed much of our understanding on UNG2 function. The assays use synthetic DNA oligonucleotide substrates with uracil bases incorporated in the DNA, and the substrates can be single-stranded, double-stranded, or form other structures such as DNA hairpins or junctions. The fluorescence signal reporting uracil base excision by UNG2 is detected in different ways: (1) Excision of uracil from end-labeled oligonucleotides is measured by visualizing UNG2 reaction products with denaturing PAGE; (2) Uracil excision from dsDNA substrates is detected in solution by base pairing uracil with 2-aminopurine, whose intrinsic fluorescence is enhanced upon uracil excision; or (3) UNG2 excision of uracil from a hairpin molecular beacon substrate changes the structure of the substrate and turns on fluorescence by relieving a fluorescence quench. In addition to their utility in characterizing UNG2 properties, these assays are being adapted to discover inhibitors of the enzyme and to determine how protein-protein interactions affect UNG2 function.

DNA repair proteins participate in extensive protein-protein interactions that promote the formation of DNA repair complexes. To understand how complex formation affects protein function during base excision repair, we used SpyCatcher/SpyTag ligation to produce a covalent complex between human uracil DNA glycosylase (UNG2) and replication protein A (RPA). Our covalent "RPA-Spy-UNG2" complex could identify and excise uracil bases in duplex areas next to ssDNA-dsDNA junctions slightly faster than the wild-type proteins, but this was highly dependent on DNA structure, as the turnover of the RPA-Spy-UNG2 complex slowed at DNA junctions where RPA tightly engaged long ssDNA sections. Conversely, the enzymes preferred uracil sites in ssDNA where RPA strongly enhanced uracil excision by UNG2 regardless of ssDNA length. Finally, RPA was found to promote UNG2 excision of two uracil sites positioned across a ssDNA-dsDNA junction, and dissociation of UNG2 from RPA enhanced this process. Our approach of ligating together RPA and UNG2 to reveal how complex formation affects enzyme function could be applied to examine other assemblies of DNA repair proteins.

Biphasic, non-sigmoidal dose-response relationships are frequently observed in biochemistry and pharmacology, but they are not always analyzed with appropriate statistical methods. Here, we examine curve fitting methods for "hormetic" dose-response relationships where low and high doses of an effector produce opposite responses. We provide the full dataset used for modeling, and we provide the code for analyzing the dataset in SAS using two established mathematical models of hormesis, the Brain-Cousens model and the Cedergreen model. We show how to obtain and interpret curve parameters such as the ED50 that arise from modeling, and we discuss how curve parameters might change in a predictable manner when the conditions of the dose-response assay are altered. In addition to modeling the raw dataset that we provide, we also model the dataset after applying common normalization techniques, and we indicate how this affects the parameters that are associated with the fit curves. The Brain-Cousens and Cedergreen models that we used for curve fitting were similarly effective at capturing quantitative information about the biphasic dose-response relationships.

Human sirtuins are a family of nicotinamide adenine dinucleotide (NAD+)‐dependent enzymes that are responsible for removing acyl modifications from lysine residues. The deacylase activity of sirtuin isoform‐2 (SIRT2) is involved in the formation and proliferation of cancers and is thought to regulate the progression of neurodegenerative diseases. The acyl substrates of SIRT2 are chemically diverse and range from small acetyl modifications to much larger, fatty acyl groups such as myristoyl. Here, we show that recombinant SIRT2 self‐associates to form dimers in a substrate‐dependent manner that regulates its deacylase activities. SIRT2’s Kd for self‐association was determined to be 98 nM, which is within range of its cellular concentration. SIRT2 alone readily forms dimers, but the enzyme undergoes a transition to monomer when bound to myristoyl substrate. SIRT2’s oligomeric transition from dimer to monomer upon myristoyl substrate binding slows its demyristoylase reaction by adding an additional step to its reaction mechanism. In contrast, SIRT2 appears to remain dimerized when performing its deacetylase reaction which enhances its activity. We propose that SIRT2 abundance in cells may regulate its oligomeric state and deacylation kinetics, and additionally, its oligomeric states may be pharmacologically targeted to modulate acyl substrate selectivity. Finally, sirtuin isoforms 3 and 6 (SIRT3 and SIRT6) do not oligomerize at concentrations relevant to cellular activity, indicating that this mode of regulation is not inherent to all human sirtuins.

Proliferating Cell Nuclear Antigen (PCNA) is a homotrimeric protein involved in DNA replication and repair. It contains three equivalent binding sites that are known to interact with dozens of replication and repair proteins. In theory, PCNA could bind multiple proteins at the same time, but steric hindrance prevents this for some protein combinations. Here, we developed an approach to detect the formation of multi‐protein complexes containing PCNA and base excision repair proteins using fluorescence anisotropy and hormetic modeling. Initially, we measured binding of fluorescent‐labeled pogo‐ligase peptide (PL) to PCNA and determined a Kdof 118 nM. This assay measured an increase in the fluorescence anisotropy of 50 nM PL upon PCNA binding because the fluorescent peptide becomes part of a larger macromolecular complex that slows its movement in solution (anisotropy min/max: 0.043/0.121). Next, we competed 50 nM PL from 0.25 μM PCNA using uracil DNA glycosylase (UNG2), which reduced the anisotropy from 0.086 to 0.057, and we determined an IC50 of 4 μM with a conventional sigmoidal curve. In contrast, displacement of PL from 3 μM PCNA using UNG2 resulted in a hormetic dose response that was fit using the Brain‐Cousens equation for hormesis. In these experiments, the anisotropy of 50 nM PL increased from 0.095 in the presence of 3 μM PCNA to 0.127 with the addition of 5 μM UNG2, and the anisotropy then reduced to 0.086 with 50 μM UNG2. The hormetic response occurred because a ternary PCNA‐PL‐UNG2 complex formed at UNG2 concentrations that were sufficient to bind open PCNA sites, but were insufficient to displace PL from PCNA, and the anisotropy of PL bound to the ternary complex was higher than its anisotropy bound to PCNA alone. Additional values derived from the Brain‐Cousens curve included a significant hormesis parameter f (0.024), a maximum stimulatory response of 134% of control at a dose of 4 μM UNG2, and a limited dose for stimulation at 31 μM UNG2. We also report additional datasets where UNG2 displaced 50 nM PL from other PCNA concentrations to train our hormetic modeling and optimize assay signal/noise. Continuing experiments using fluorescent‐labeled UNG2 and fluorescent‐labeled DNA polymerase β (POLB), in addition to DNA Ligase 1 (LIG1), will explore the simultaneous and/or sequential interactions of base excision repair proteins with PCNA.

Hormesis refers to dose-response phenomena where low dose treatments elicit a response that is opposite the response observed at higher doses. Hormetic dose-response relationships have been observed throughout all of biology, but the underlying determinants of many reported hormetic dose-responses have not been identified. In this report, we describe a conserved mechanism for hormesis on the molecular level where low dose treatments enhance a response that becomes reduced at higher doses. The hormetic mechanism relies on the ability of protein homo-multimers to simultaneously interact with a substrate and a competitor on different subunits at low doses of competitor. In this case, hormesis can be observed if simultaneous binding of substrate and competitor enhances a response of the homo-multimer. We characterized this mechanism of hormesis in binding experiments that analyzed the interaction of homotrimeric proliferating cell nuclear antigen (PCNA) with uracil DNA glycosylase (UNG2) and a fluorescein-labeled peptide. Additionally, the basic features of this molecular mechanism appear to be conserved with at least two enzymes that are stimulated by low doses of inhibitor: dimeric BRAF and octameric glutamine synthetase 2 (GS2). Identifying such molecular mechanisms of hormesis may help explain specific hormetic responses of cells and organisms treated with exogenous compounds.

Sirtuins are a class of proteins belonging to the Sir2 (Silencing information regulator 2) family of NAD+ dependent protein lysine deacylases. Different Isoforms (SIRT1-SIRT7) differ in their specific deacylase activity and cellular location. They have roles in DNA repair, glucose metabolism, and cellular proliferation which make them highly desirable targets for carcinoma therapeutics. We previously used 1-aminoanthracene’s (AMA) fluorescent properties when bound with SIRT2 (Kd of 37 μM) to develop a high-throughput screen to identify novel ligands that inhibit SIRT2’s enzymatic activities. We hope to reveal other potential probes for future high-throughput screening with all the sirtuin isotopes. 1-AMA’s fluorescence along with fluorescent labeled peptides “Cy3-PEG4-H4K16(myr)” and “FAM-PEG4-H4K16(myr)” were used in binding assays to determine their affinities with SIRT2, SIRT3, and SIRT6. Further, we determined 1-AMA’s ability to bind sirtuin isoforms when they were equilibrated with 100 μM of various acyl-peptides. 1-AMA displays weak binding to SIRT3 and SIRT6 when compared to SIRT2. FAM-PEG4-H4K16(myr) binds SIRT2 with a Kd of 7nM which is much higher than its interaction with SIRT3 and SIRT6 (Kd of 6 μM and 2 μM, respectively). Cy3-PEG4-H4K16(myr) binds as expected SIRT2,3,6 although its affinity for SIRT6 changes minimally with the addition of ADP-ribose, which suggests Cy3 may facilitate binding in the absence of SIRT6’s co-factor. Future work will test additional probes with the other sirtuin proteins and establish their competency to be utilized for high-throughput screening.

Human uracil DNA glycosylase (UNG2) is responsible for removal of uracil bases from DNA and initiates base excision repair pathways. Accumulation of uracil or its fluorinated analogs in DNA is one of the killing mechanisms of thymidylate synthase (TS) inhibitors in cancer cells, and depletion of UNG2 often enhances the toxicity of these anticancer drugs. We used CRISPR to knockout UNG2 from HT29 colon cancer cells and confirmed the absence of protein by western blot and uracil excision assays. We tested the effect of UNG2 KO on the efficacy of multiple TS inhibitors (5‐fluorouracil, fluorodeoxyuridine, pemetrexed, and raltitrexed), and we determined that only fluorodeoxyuridine and raltitrexed were significantly more potent in UNG2 KO cells compared to wild‐type HT29 cells (fluorodeoxyuridine IC50: 2 mM (wt) vs. 3 nM (KO); raltitrexed IC50: 14 nM (wt) vs. 2 nM (KO)). Interestingly, UNG2 protein levels can also be depleted by the HDAC inhibitors SAHA and MS275, providing a pharmacologic strategy to reduce UNG2 activity in cells. Unexpectedly, the HDAC inhibitors synergized with 5‐fluorouracil, but not fluorodeoxyuridine, in both wild‐type and UNG2‐knockout cells. This suggested that HDAC inhibitors sensitized cells to 5‐fluorouracil through an UNG2‐independent mechanism. Moreover, cell death pathways activated by fluorodeoxyuridine and regulated by UNG2 activity are not sensitized by HDAC inhibitors. Our combined genetic and pharmacologic strategies targeting UNG2 activity in cells are defining cell death mechanisms for combination therapies of TS inhibitors and HDAC inhibitors. Future work will examine these drug combinations in additional cell lines to understand optimal therapeutic combinations and to further refine mechanisms of cell death.

Small-molecule inhibitors of the human sirtuin SIRT2 are being developed because of their therapeutic potential in a variety of diseases. Here, we developed a high-throughput screen to identify novel SIRT2 inhibitors using a fluorescent SIRT2 probe, 1-aminoanthracene (AMA). AMA has high fluorescence when bound to SIRT2, and its fluorescence reduces >10-fold when it is displaced from SIRT2 by other ligands. We used this property of AMA to screen a library of known bioactive compounds for SIRT2 binding and discovered two known pharmaceutical compounds that bind SIRT2 with Kd values in the low μM range, ascorbyl palmitate and pictilisib. Both compounds inhibit the deacetylase and defatty-acylase activities of SIRT2. While pictilisib has selectivity for SIRT2, ascorbyl palmitate also inhibits the enzymatic activities of SIRT1 and SIRT6. Finally, we show that ascorbyl palmitate inhibits SIRT2 deacetylase and defatty-acylase activities in cells, and SIRT2 inhibition by ascorbyl palmitate contributes to the cytotoxicity of the compound. Our work discovered novel SIRT2 deacylase inhibitors and presents a screening approach that can be applied on a larger scale.

It is well established that thymidylate synthase inhibitors can cause cellular toxicity through uracil DNA glycosylase (UNG2)-dependent pathways. Additionally, thymidylate synthase inhibitors and HDAC inhibitors are known to act synergistically in a variety of cancer types. A recent article from J. Transl. Med. links these together by demonstrating widespread depletion of UNG2 levels across a variety of cell lines treated with HDAC inhibitors. Recent findings suggest that UNG2 depletion by HDAC inhibitors would likely be an effective method to sensitize cells to thymidylate synthase inhibitors. This is particularly important for cancer types that are typically resistant to thymidylate synthase inhibitors, such as cells that are deficient in p53 activity.

Sirtuin isoform 2 (SIRT2) is an enzyme that catalyzes the removal of acyl groups from lysine residues. SIRT2's catalytic domain has a hydrophobic tunnel where its substrate acyl groups bind. Here, we report that the fluorescent probe 1-aminoanthracene (AMA) binds within SIRT2's hydrophobic tunnel in a substrate-dependent manner. AMA's interaction with SIRT2 was characterized by its enhanced fluorescence upon protein binding (>10-fold). AMA interacted weakly with SIRT2 alone in solution (Kd = 37 μM). However, when SIRT2 was equilibrated with a decanoylated peptide substrate, AMA's affinity for SIRT2 was enhanced ∼10-fold (Kd = 4 μM). The peptide's decanoyl chain and AMA co-occupied SIRT2's hydrophobic tunnel when bound to the protein. In contrast, binding of AMA to SIRT2 was competitive with a myristoylated substrate whose longer acyl chain occluded the entire tunnel. AMA competitively inhibited SIRT2 demyristoylase activity with an IC50 of 21 μM, which was significantly more potent than its inhibition of other deacylase activities. Finally, binding and structural analysis suggests that the AMA binding site in SIRT2's hydrophobic tunnel was structurally stabilized when SIRT2 interacted with a decanoylated or 4-oxononanoylated substrate, but AMA's binding site was less stable when SIRT2 was bound to an acetylated substrate. Our use of AMA to explore changes in SIRT2's hydrophobic tunnel that are induced by interactions with specific acylated substrates has implications for developing ligands that modulate SIRT2's substrate specificity.

Replication Protein A (RPA) is a single-stranded DNA binding protein that interacts with DNA repair proteins including Uracil DNA Glycosylase (UNG2). Here, I report DNA binding and activity assays using purified recombinant RPA and UNG2. Using synthetic DNA substrates, RPA was found to promote UNG2's interaction with ssDNA-dsDNA junctions regardless of the DNA strand polarity surrounding the junction. RPA stimulated UNG2's removal of uracil bases paired with adenine or guanine in DNA as much as 17-fold when the uracil was positioned 21 bps from ssDNA-dsDNA junctions, and the largest degree of UNG2 stimulation occurred when RPA was in molar excess compared to DNA. I found that RPA becomes sequestered on ssDNA regions surrounding junctions which promotes its spatial targeting of UNG2 near the junction. However, when RPA concentration exceeds free ssDNA, RPA promotes UNG2's activity without spatial constraints in dsDNA regions. These effects of RPA on UNG2 were found to be mediated primarily by interactions between RPA's winged-helix domain and UNG2's N-terminal domain, but when the winged-helix domain is unavailable, a secondary interaction between UNG2's N-terminal domain and RPA can occur. This work supports a widespread role for RPA in stimulating uracil base excision repair.