https://orcid.org/0000-0002-4519-6276
This is an uncorrected proof.
Figures
Abstract
Herpes simplex virus 1 (HSV-1) UL12 encodes a highly conserved 5′ → 3′ alkaline exonuclease that is essential for the production of infectious virus. Together with the viral single-stranded DNA-binding/annealing protein ICP8, UL12 functions as a two-component recombinase that mediates recombination-dependent viral DNA replication. Here, we present the crystal structure of the catalytic domain of the HSV alkaline nuclease (UL12), which provides the first view of an α-herpesvirus alkaline nuclease. Using this structure, we optimized a series of small-molecule viral nuclease inhibitors (VNIs) that target the UL12 active site and potently inhibit UL12 exonuclease activity in vitro. We have thus established a robust platform for structure-based docking, SAR analysis and rational inhibitor design. Because UL12 orthologs are conserved across all human herpesviruses, we examined the activity of these compounds against the β- and γ-herpesvirus alkaline nucleases UL98 and SOX and found that they inhibit all three enzymes. The VNIs also exhibit antiviral activity against HSV-1 and HCMV in cell culture. EC50 and IC50 values were in the nanomolar to low micromolar range. Together, these findings establish herpesvirus alkaline nucleases as conserved, druggable antiviral targets and provide a foundation for the development of broad-spectrum anti-herpesvirus therapeutics, either as standalone agents or in combination with existing nucleoside analogs.
Author summary
Human herpesviruses are widespread pathogens that establish lifelong infections and can cause severe disease, particularly in immunocompromised individuals, newborns, and older adults. Current antiviral treatments are limited and often target a narrow range of viruses, highlighting the need for new therapeutic strategies. In this study, we focus on UL12, a viral enzyme that is essential for herpesvirus replication and is conserved across all human herpesviruses. We determined the first crystal structure of the HSV-1 UL12 nuclease, providing structural insight into the catalytic mechanism and conservation of herpesvirus alkaline nucleases. Using this structural information, we evaluated small molecules that inhibit UL12 activity and found that they block viral replication in cell culture. Importantly, these inhibitors also act on related enzymes from other herpesviruses, supporting the idea that UL12 represents a shared vulnerability. Our findings establish a foundation for developing broadly active antiviral therapies targeting herpesvirus nucleases.
Citation: Sharma N, Xie X, Szczepaniak R, Rani C, Khosro SK, Krucinska J, et al. (2026) Viral Nuclease Inhibitors: Small molecule disruptors of the UL12 alkaline nuclease display broad anti-herpes virus activity. PLoS Pathog 22(8): e1014531. https://doi.org/10.1371/journal.ppat.1014531
Editor: Clinton Jones, Oklahoma State University College of Osteopathic Medicine, UNITED STATES OF AMERICA
Received: June 16, 2026; Accepted: August 6, 2026; Published: August 17, 2026
Copyright: © 2026 Sharma et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files. The atomic coordinates and structure factors for the HSV‑1 UL12.5 structure have been deposited in the Protein Data Bank (PDB) under accession number 10HO.
Funding: SW received AI069136, and SW and DW received R21 AI143272, R56 AI173955 and 5R01AI021747-35 grants from National Institute of Health/National Institute of Allergy and Infectious Diseases NIH/NIAID (https://www.niaid.nih.gov). The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
Competing interests: I have read the journal’s policy and the authors of this manuscript have the following competing interests: L. W., D. W. and S. W. are co-founders of Quercus Molecular Design, an early-stage biotech company focused on infectious disease. The remaining authors have declared that no competing interests exist.
Introduction
Human herpesviruses (HHVs) are large DNA viruses with a profound global clinical impact. They establish lifelong latent infections characterized by episodes of reactivation. Clinical manifestations span a wide spectrum, from neonatal infections and cancer to encephalitis and severe opportunistic disease in immunocompromised individuals. Among the most prevalent HHVs are the α-herpesviruses (HSV-1 and HSV-2), which are responsible for orolabial and genital herpes. HSV-1 affects an estimated 3.8 billion people under the age of 50 (64% worldwide) [1], while HSV-2 infects about 520 million individuals aged 15–49 (13%) [2,3].
Drug discovery for HHVs has centered largely on nucleoside analogs that inhibit the viral DNA polymerase. For HSV, the mainstay therapy is acyclovir (ACV) and its prodrug valaciclovir (VCV), which effectively reduces the severity and frequency of genital and orolabial lesions [4,5]. However, ACV, approved in 1982 [6] and VCV, approved in 1995 [7], do not prevent viral reactivation, viral shedding or partner transmission, nor are they effective against the β- (HCMV, HHV6A/B and HHV7) or γ- (EBV, KSHV) HHVs. Ganciclovir (licensed in 1989 [8]), and its oral prodrug valganciclovir (approved in 2001 [9]), are the first-line antivirals for HCMV infections, although their use is limited by off-target toxicities such as myelosuppression [10,11]. The only newly approved agent targeting HHVs since 2001 has been letermovir [12], which inhibits the terminase. However, letermovir has major limitations including narrow spectrum (HCMV only), lack of approval for treatment of active disease, rapid emergence of resistance and clinically significant drug-drug interactions [13–16]. These limitations highlight the ongoing need to develop anti-HHV therapeutics that act on novel viral targets.
Since 1978 it has been recognized that HSV-1 encodes a 5’ → 3’ alkaline deoxyribonuclease (UL12) [17]. All HHVs encode a well-conserved ortholog of UL12 [18–21], several of which (HCMV UL98, EBV BGLF5 and KSHV SOX) have been shown to be essential for viral growth [22–26]. HSV-1 UL12 is the best-characterized alkaline nuclease (AN), and we have shown that it plays a critical role in the synthesis of HSV genomes that can be packaged into infectious virions [26]. Interestingly, UL12 collaborates with the HSV-1 single-strand DNA binding protein/annealase, ICP8, to mediate recombination-dependent replication of viral DNA [27–29]. UL12 has also been shown to promote the maintenance of replication fork progression [30]. Interestingly, HSV-1 also expresses a truncated form of the AN known as UL12.5 that lacks the disordered tail of the full-length protein but retains the core catalytic domain and exonuclease activity [31]. Unlike the full-length protein, UL12.5 localizes to the mitochondria during infection [32]. From a drug discovery perspective, UL12 is an attractive target due to its essential role in viral DNA replication.
UL12 and its HHV AN orthologs are members of a large family of two metal ion-dependent (TMID) enzymes that share a dependency on magnesium ions for catalytic activity [33–35]. The presence of a well-defined active site also underscores its potential for the development of novel antiviral agents. We previously identified small-molecule inhibitors of HSV-1 UL12 by exploiting its two-metal ion catalytic center [26,33]. We have now established a platform to target the highly conserved HHV ANs which involves: 1) screening a proprietary library of flexible, metal-directed inhibitors; 2) optimizing substitution patterns of initial hits; 3) enhancing molecular rigidity; and 4) conducting off-target profiling and synthetic refinement to mitigate liabilities. Drug discovery efforts targeting TMID enzymes in HIV and influenza have successfully yielded selective and safe inhibitors in the past [36–39], demonstrating that enzymes of this class can serve as viable drug targets.
Results
We previously reported inhibitors of HSV-1 UL12, including a series of natural and synthetic α-hydroxytropolones that demonstrated inhibition of in vitro nuclease activity [26]. From an in-house compound library, we also identified a number of 8-hydroxyquinolines (8-HQ), exemplified by AK-157 and AK-166, that inhibited UL12 in the micromolar range and showed antiviral activity against both HSV-1 and HCMV [33]; however, these compounds demonstrated some cytotoxicity, leading to reduced selectivity indexes (SI) (Fig 1). Building on this work, we have identified a new scaffold and produced a novel series of amido substituted hydroxypyridinones (HPs) which exhibit excellent antiviral activity and desirable selectivity indexes.
Determination of the crystal structure of UL12.5, the catalytic domain of HSV-1 UL12
Although crystal structures of γ-herpesvirus alkaline nucleases, such as BGLF5 in its apo form [40] and SOX in both apo and DNA/RNA-bound forms [41–43] have been previously solved, no α- or β-HHV AN structures have been reported. In this study, we present the first crystal structure of HSV-1 AN (UL12 residues 126–626) from anα-HHV. Efforts to crystallize full-length UL12 were hindered by the presence of a disordered N-terminal region comprising 126 amino acids. Therefore, the truncated protein (UL12 residues 126–626; also referred as UL12.5), was expressed in E. coli system with an N-terminal 12x His-tag and a TEV protease cleavage site to enable affinity purification. Despite its limited solubility, we successfully purified the protein using immobilized metal affinity and size exclusion chromatography, yielding approximately 1 mg of protein from a 1.5 L culture. This preparation of UL12.5 retained exonuclease activity at levels comparable to the full-length protein (S1 Fig). Subsequent crystallization trials yielded diffraction-quality crystals in the apo form, with the structure determined at 2.46 Å resolution via molecular replacement (S1 Table).
Initial attempts to solve the UL12.5 structure by molecular replacement using available homologous structures, including SOX [43] and BGLF5 [40], were unsuccessful, likely due to limited sequence identity and structural divergence. Molecular replacement was ultimately successful using an AlphaFold2‑predicted model of UL12.5 as the search template [44] which yielded a clear solution and well‑defined electron density (Fig 2A and 2B). The final model includes residues 1–73, 83–137, 144–166, 196–308, 317–418, 429–477 with the partial tag on the N-terminus (residues -8–0) and contains one molecule in the asymmetric unit. The overall architecture closely resembles that of the γ-herpesvirus nucleases SOX and BGLF5, adopting a canonical nuclease fold with distinct N- and C-terminal domains (Fig 2A) [40,43]. The structural alignment of UL12.5 with SOX (PDB:3FHD) and BGLF5 (PDB:2W45) yielded RMSD values of 1.54 Å (205 atoms) and 1.63 Å (253 atoms), respectively (Fig 2C). The catalytic triad, comprising Glu154, Asp 214 and Glu238 (residues Glu280, Asp340, and Glu364 respectively in full length-UL12) with clear electron density, is positioned within a crevice formed at the central core. However, no electron density corresponding to the bound Mg2+ ions was observed, despite their expected role in catalysis. Interestingly, while one Mg2+ is clearly resolved in the structure of SOX (3FHD), it is similarly absent in the apo structure of BGLF5 (2W45), indicating that metal ion binding may be substrate-dependent or influenced by subtle differences in structural dynamics among herpesvirus nucleases. Notably, the side chains of two residues of the catalytic triad, Glu154 and Asp214 in UL12.5 are oriented away from the metal-binding site in comparison to the SOX structure (Fig 2D).
(A) Overall structure of UL12.5, with the N-terminal domain shown in wheat and the core C-terminal domain in light blue. The conserved catalytic triad is depicted as green sticks. (B) Representative 2Fo-Fc electron density map contoured around residues 233-238. (C) Structural superposition of UL12.5 (light blue) with the Epstein-Barr virus alkaline nuclease BGLF5 (2W45: brick red) and the Kaposi’s sarcoma-associated herpesvirus alkaline nuclease SOX (3FHD: cyan). (D) The catalytic triads of UL12.5, BGLF5, and SOX are depicted as light blue, brick red, and cyan sticks, respectively, highlighting their relative positions and conformations. The Mg2+ ion coordinated in the SOX structure is represented as a green sphere. (E) Electrostatic surface representation of UL12.5, colored according to surface charge. Double-stranded DNA from structurally aligned DNA-bound complexes of KSHV SOX (green) and bacteriophage λ-exonuclease Redα (yellow) is shown to indicate the predicted DNA-binding path along the positively charged surface of UL12.5.
Comparative structural analysis shows that HSV‑1 UL12 preserves the conserved 5′‑phosphate-binding architecture characteristic of EBV BGLF5, KSHV SOX, and bacteriophage λ‑exonuclease (Redα) [40,43,45,46]. In Redα, the DNA‑bound structure reveals anchoring of the 5′‑terminal phosphate within a deeply buried, positively charged pocket that is essential for substrate positioning and processive catalysis [47]. An analogous anion‑binding pocket is present in EBV BGLF5 and KSHV SOX, where sulfate, phosphate, or formate ions occupy the same site in apo and DNA‑bound structures. Whereas SOX and BGLF5 contain a conserved Ser-Ser-Ser stretch within motif I, UL12 instead features a Thr-Ala-Ser sequence, a composition that more closely resembles Redα and highlights sequence variation within an otherwise conserved structural framework (S2 Fig). Consistent with this conservation, UL12 displays clear electron density within the phosphate‑binding pocket modeled as a formate ion, paralleling the anion‑binding modes observed across the family. Together, these structures establish that UL12 retains the core framework for 5′‑phosphate recognition shared with λ‑exonuclease and other γ‑herpesvirus nucleases, while subtle differences in residue substitutions may modulate substrate engagement and catalytic efficiency. Electrostatic surface analysis further supports this model, revealing a positively charged groove that is predicted to accommodate DNA binding (Fig 2E).
Our initial molecular docking efforts to design inhibitors relied on homology models of UL12 derived from the KSHV SOX structure. However, these models failed to accurately recapitulate key structure-activity relationships in the enzyme active site. The newly determined UL12.5 crystal structure from this work provided us a far more robust foundation for structure-based drug design.
Molecular modeling with the UL12.5 structure
In order to conduct docking studies with our newly solved UL12.5 structure, it was necessary to model the active site divalent magnesium ions and bound waters. Structural modeling of UL12.5 was guided by the crystal structure of Redα bound to DNA (PDB: 3SM4) [47]. Although a nucleic acid-bound structure is available for the orthologous KSHV SOX protein [41], we used λ Redα(47) due to its closer functional correspondence with UL12.5, as SOX also degrades mRNA – reflecting distinct active-site features. Proteins were aligned by superimposing conserved metal-coordinating residues (Fig 3A). Following superposition, Redα protein and all non-bound waters were removed, leaving only active site waters and magnesium ions. Dihedral angles of the active site residues were adjusted as needed and the UL12.5 protein was prepared for docking using ProteinPrep in the Schrodinger Maestro suite [48–51].
(A) Top panel: Structural superposition of conserved active-site residues from UL12.5 (blue) and λ Redα (yellow), shown as sticks. Residue 131 in λ Redα was mutated from Lys to Ala to prevent DNA cleavage; Bottom panel: Alignment of conserved active-site residues from λ Redα, full-length UL12, and UL12.5; residue numbering for UL12.5 is offset by 126 residues relative to full-length UL12. (B) Docking model of a trinucleotide containing a free 5′ phosphate bound to UL12.5, illustrating extensive interactions between the nucleotide bases and active‑site residues, either directly or mediated by Mg2+ ions and water molecules. (C) Predicted binding orientation of the first-generation 8-hydroxyquinoline inhibitor AK-157 (shown in yellow sticks) within the UL12.5 active site. Purple stars indicate water molecules displaced upon ligand binding. (D) Two-dimensional ligand interaction diagram showing the predicted key contacts between AK‑157 and UL12.5 active-site residues. The colored contour surrounding the ligand represents the local binding-pocket environment, and interaction types are indicated according to the embedded legend. In panels B and C, UL12/UL12.5 active-site residues are shown as dark gray sticks, DNA bases as magenta sticks with the backbone in light gray, inhibitors as yellow sticks, Mg2+ ions and water molecules as cyan and red spheres respectively.
To evaluate the model, a trinucleotide (AGC) bearing a free 5’ - phosphate was prepared with Schrodinger LigPrep (pH = 7 + /-2) and docked to the protein using the Schrodinger Maestro Glide flexible docking (Fig 3B) [51]. Notable high affinity interactions include: (1) binding of the 5’- terminal phosphodiester bond to both magnesium ions, positioning it adjacent to the K240/magnesium A-bound nucleophilic water which performs the bond cleavage, (2) binding of the 5’- terminal phosphodiester oxygen to magnesium B, (3) a network of H-bonds between the 5’- free phosphate and residues W105, R109, S116 (via water), and S212 [52]. Additionally, W105 is at an appropriate distance to engage the 5’ terminal nucleobase through hydrophobic π interactions.
Docking of first generation hydroxyquinoline inhibitors
As previously reported, we identified a series of 8-hydroxyquinolines (HQs) that inhibited UL12 enzymatic function and demonstrated good antiviral activity against HSV-1 in cell culture [33]. Extensive docking was performed with HQ inhibitors (i.e., AK-157, Fig 1) to help elucidate the mode of binding and orientation of the molecule in the active site. Iterative modeling suggested that binding of the inhibitor (Fig 3C) would promote the displacement of two bound water molecules from the active site. AK-157 interacts primarily through binding of the heteroatom triad with the magnesium ions, while the p-bromophenyl amide side chain sits in a hydrophobic region created by the hydrocarbon portion of the sidechains of residues R151, E154, and T197 (Figs 3C and 3D).
Implications of sequence homology for Pan-HHV activity
The HHV alkaline nucleases share significant homology (Figs 3A and S3A), including identical active site catalytic arrays (UL12 PD-(D/E)XK motif: L339, D340, E364, V365, K366 and E280; denoted by blue stars in S3A Fig) as well as the 5’-phosphate binding pocket (WRSS denoted by green stars in S3A Fig). We rationalized that potency of these compounds could be enhanced by engaging the unique 5’ terminal phosphate binding pocket (WRSS), which is not found in eukaryotic or prokaryotic exonucleases [47]. Given the planar mode of binding of the HQs (S3B Fig), we determined that a different scaffold would be required to access the terminal phosphate binding pocket. We selected a hydroxypyridinone (HP) chemotype, which would be more suitable for targeted design as the substitution on the nitrogen of the 4-pyridinone could be readily exchanged using well-established chemistry. Modeling indicated that aromatic groups in this region could potentially form contacts with the WRSS region and, thus, we prepared analogs bearing N-alkyl, aryl and benzyl substitutions on the pyridinone ring. Although this first iteration of HPs was inferior to the most potent HQs, the cellular CC50s were clearly improved (S3D Fig). Molecular docking of representative compounds (i.e., DD-I-38) indicated that the pyridinone substitution was insufficiently flexible for accessing the WRSS pocket (S3C Fig). Iterative molecular modeling determined that an azacycle spacer could effectively position the pendent aromatic groups proximal to the WRSS site. Subsequently, we prepared a series of HP analogs, including 4, 5, and 6-membered rings using the route shown in Fig 7.
New compounds were screened against the UL12 enzyme, followed by antiviral evaluation (HSV-1) in cell culture. HPs that exhibited strong antiviral activity (EC50 < 500nM) were prioritized for additional evaluation (Fig 4).
The sulfonamide design improved both target level activity and SI. Once baseline activities for the new series of compounds had been established, induced-fit molecular docking was performed to aid in developing a refined SAR model. Fig 5 shows representative sulfonamide class inhibitor VNI-5168 docked to the UL12.5 structure. The binding of the hydroxypyridinone core is dominated by interactions between the planar oxygen triad and the two divalent magnesium ions; however, the non-planar distortion of the aminomethyl piperidine group allows the aromatic sulfonamide to access the WRSS pocket, unlike the earlier HP inhibitors (S3C Fig). Molecular docking predicted that the aromatic sulfonamides would engage this region through a combination of H-bonding to the sulfonamide oxygen atoms, π-interactions with the W105-R109 cation-π complex and H-X hydrogen bonding (Fig 5). To capture a diverse range of substitution patterns among representative highly active sulfonamide inhibitors, VNI‑5162, VNI‑5163, VNI‑5167, and VNI‑5168 were selected for more detailed evaluation.
(A) Docking model of the sulfonamide‑class inhibitor VNI‑5168 bound to the UL12.5 active site. The active‑site pocket is shown in surface, with catalytic and ligand‑interacting residues depicted as dark gray sticks. VNI‑5168 is shown as yellow sticks, highlighting extensive interactions with UL12.5 residues and engagement of the WRSS pocket. Mg2+ ions and coordinating water molecules are depicted as cyan and red spheres respectively. (B) Two-dimensional ligand interaction diagram showing the predicted interactions between VNI-5168 and UL12.5 active-site residues. The colored contour surrounding the ligand represents the local binding-pocket environment, and interaction types are indicated according to the embedded legend.
Testing for breadth of coverage
As described above, there is a high level of active site homology across all seven HHV exonucleases. In principle, similar architectures could allow broad spectrum inhibition from a single compound. To better evaluate this potential, we compared the degree of inhibition of enzyme activity for the four select compounds between HSV-1 UL12 and two additional ANs, HCMV UL98, a β-herpesvirus and KSHV SOX, the γ-family homolog (Fig 6A). Although all four analogs at a single fixed 5 µM concentration demonstrated complete inhibition of both UL12 and UL98, the activity against KSHV SOX was more variable. VNI-5162 and VNI-5168, which bear halogen substitutions on the pendent sulfonamide, exhibited strong inhibitory activity. In contrast, inhibition of SOX by the methyl (VNI-5163) and methoxy (VNI-5167) analogs was significantly lower. The observed disparity between the γ- (SOX) and α- (UL12)/β- (UL98) ANs is potentially related to an additional RNA endonuclease activity of the γ orthologs.
(A) Percentage inhibition of HSV-1 UL12, HCMV UL98, and KSHV SOX, representative alkaline nucleases from the alpha-, beta-, and gamma-herpesvirus subfamilies, respectively. Nuclease activity was measured using a PicoGreen-based assay in the presence of 5 µM inhibitor. Percent inhibition was calculated relative to the uninhibited DMSO control. Data represent the mean of three independent experiments, with each experiment consisting of 2–3 technical replicates. (B-F) Dose-response analyses for enzymatic inhibition, binding and antiviral activity. IC50 values for full‑length (FL)-UL12 (B) and UL12.5 (C) were determined using the PicoGreen‑based nuclease assay. (D) Binding affinities between UL12 and the indicated inhibitors were measured by microscale thermophoresis (MST), with normalized fluorescence plotted as a function of compound concentration. Data were fitted to a Kd binding model using MO.Affinity Analysis software (v3.0.5; NanoTemper Technologies) and plotted using GraphPad Prism. Data shown are from three independent experiments, each performed in technical duplicate. EC50 values for HSV-1 (E) and HCMV (F) were determined in HFF cells infected with fluorescently tagged viruses in the presence of increasing concentration of the compound and antiviral activity was evaluated using high-content image-based assay. Percentage inhibition in panels B, C, E and F was calculated relative to the uninhibited DMSO control. IC50 and EC50 values were obtained by nonlinear regression analysis in GraphPad Prism using a four-parameter logistic model (log[inhibitor] versus response, variable slope). Data in panels B, C, E and F represent the mean of three independent experiments, with each experiment comprising 2-3 technical replicates.
Detailed characterization of lead sulfonamides
The four compounds were resynthesized and the inhibition of UL12 was re-confirmed, yielding IC50 values ranging from 0.59-2.1 µM (Table 1, column 1) with similar inhibitory profile observed against UL12.5 (Table 1, column 2 and Fig 6B, 6C). Additionally, a Km value for UL12 was determined in the absence of inhibitor (S4 Fig; see S1 Text for experimental details) and used together with the measured IC50 values to estimate Ki values for each compound using the Cheng–Prusoff equation [53]. The resulting sub-micromolar Ki values reflect robust inhibition of this nuclease (Table 1, column 3). However, it is important to note that UL12 is a processive enzyme, and interpretation of IC50/Ki values can be more complicated than those observed in more traditional distributive enzymes that follow classical Michaelis-Menten kinetics. In contrast, processive exonucleases employ a multi-turnover mechanism whereby multiple nucleotide excisions occur per each DNA-binding event. This often results in an under- or over-estimation of the true Ki. To provide an independent assessment of target-engagement, dissociation constants (Kd) for two lead inhibitors, VNI-5168 and VNI-5162, were determined using microscale thermophoresis (MST) [54,55]. As shown in Table 1 (column 4), direct binding was observed between UL12 and these two compounds. Titration of the labeled protein with increasing concentrations of the inhibitors yielded Kd values of 0.145 [0.069-0.304] µM for VNI-5168, and 0.249 [0.114-0.545] µM for VNI-5162 (Fig 6D). These high binding affinity values align closely with our extrapolated Ki values (Table 1, column 3), further validating potent interactions between the sulfonamide inhibitors and UL12.
Antiviral activity of lead sulfonamides
The lead compounds described above were further evaluated against HSV‑1 and HCMV using a fluorescence‑based assay. HFF cells were infected with recombinant viruses expressing fluorescently tagged late gene products. Because late gene expression depends on ongoing viral DNA replication, this readout serves as a direct indicator of active replication. HFF cells were infected with an HSV-1 recombinant virus, eYFP-ICP0/gC-mCherry, generously provided by Dr. Colin Crump (Cambridge, UK) [56]. A multiplicity of infection (MOI) of 1 PFU/cell was used to ensure that majority of the cells were infected, and expression of mCherry-tagged glycoprotein gC, encoded by the UL44 gene was measured using Cytation5 high-content microscopy. To assess inhibition of HCMV, HFF cells were infected with TB40/E-mCherry-UL99eGFP, generously provided by Eain A. Murphy (Upstate Medical University, Syracuse, NY) [57]. Expression of GFP-tagged pp28 encoded by the late gene, UL99, was measured.
The lead compounds exhibited antiviral activity against HSV‑1, with EC50 values comparable to the Kᵢ values observed for UL12 inhibition (Table 1, column 5 and Fig 6E). These compounds also showed significant inhibition of HCMV, consistent with their potency against the UL98 enzyme (Table 1, column 6 and Fig 6F). We also tested antiviral activity against HSV-1 using a viral yield assay in which cells were infected at a low MOI (0.05 PFU/cell) for 65 h to test the ability of the virus to spread. In this assay, compounds produced a 2–3 fold stronger effect compared to the fluorescence assay (Table 1, column 7, S5A Fig). This is likely due to the ability of the virus to undergo multiple rounds of infection at the low MOI used in this experiment. No significant cytotoxicity was observed at 24 hrs post-treatment, while some increase was evident at 65 hrs (Table 1, column 8, S5B Fig). However, for most compounds, even at the longer time-points, the CC50 for each compound was not reached at the highest compound concentration tested (100 µM). This resulted in selectivity indexes (SI) between 19–470 fold for fluorescence assays and even higher for yield assays (Table 1, column 9 and 10).
Exploration of SAR in the initial HQ series suggested a strong preference for halo aromatic units on the amido sidechain frequently seen in other nuclease inhibitors [58,59]. However, modeling suggested that alternative hydrophobic moieties could also be well tolerated in the pocket. To evaluate alternative substitutions, two additional analogs were prepared bearing a dibenzothiophene (VNI-5173) and dibenzofuran (VNI-5174) sidechain. These extended analogs exhibited strong inhibition of the recombinant UL12 enzyme and correspondingly strong antiviral activity with limited toxicity, suggesting additional freedom-to-operate in the region of the inhibitor structure (Table 1, S6 Fig).
Discussion
Extensive genetic and biochemical studies support a model in which HSV DNA synthesis proceeds through a recombination-dependent single-strand annealing pathway mediated by a two-component viral recombinase consisting of the alkaline nuclease UL12 and the SSAP ICP8 [28,60,61]. The HSV-1 Exo/SSAP recombinase is reminiscent of related two-component recombination systems that are widely conserved among DNA viruses infecting bacteria, protozoa, plants, mammals and insects [62,63]. The preservation of viral Exo/SSAP mechanism suggests that they play a significant role in the evolutionary success of these viruses, supporting the choice of this complex as a novel target for drug discovery [26,33].
Analysis of UL12 mutants lacking exonuclease activity, demonstrated a significant defect in the production of infectious viral progeny [26], suggesting that pharmacological inhibition of nuclease activity could be antiviral. To test this hypothesis, we leveraged our high-resolution structure of the catalytic domain of UL12 to design potent active site inhibitors. The resulting compounds show strong antiviral activity against HSV-1 with favorable selectivity (Table 1). Our structure provides the first view of an α-herpesvirus alkaline nuclease and establishes a robust platform for structure-based docking, SAR analysis and rational inhibitor design. Leveraging the high conservation across herpesvirus alkaline nucleases facilitated the design of broadly active antiviral inhibitors. Compounds that inhibited HSV-1 UL12 nuclease activity and viral replication also potently inhibited HCMV. Although inhibition of the KSHV SOX nuclease is more variable, several analogs effectively target all three enzymes, thus establishing the feasibility of developing pan-herpesvirus antivirals. While this study focuses on human herpesviruses, the findings may have broad veterinary applications, extending to feline, bovine, and equine herpesviruses.
In this work, we identified a highly flexible scaffold and numerous lead compounds which exhibit strong potency, high selectivity and are broadly active against multiple HHVs. Although the observed antiviral potency is significant, it will be necessary to further increase antiviral potency in order to develop a preclinical candidate. This will require enhancement of target potency against the viral alkaline nucleases. At present, it is unclear whether the active site homology between the different HHV ANs presented herein would lead to a broadly active, low nM inhibitor. Alternatively, highly potent derivatives may have a more restrictive spectrum of enzyme coverage. This would likely be a significant determinant for whether further development strategies focus on a broadly active HHV antiviral or compounds with more specific clinical indications.
In addition to these considerations on potency and spectrum, other key optimization parameters would be toxicity and pharmacokinetic properties. The derivatives evaluated thus far exhibit high selectivity with respect to human cytotoxicity, suggesting minimal off-target effects. However, further lead optimization may yield compounds with increased activity against human nucleases. Should this occur, subsequent optimization efforts would focus on identifying the structural determinants that govern selectivity between viral nucleases and their cellular counterparts. We also recognize that the metal-binding functionality of these inhibitors may represent a metabolic liability, potentially necessitating a prodrug strategy similar to that employed for baloxavir marboxil, the prodrug of the influenza antiviral baloxavir.
Materials and methods
Bacterial constructs
The pETM6T1‑SOX vector was generously provided by Dr. Tracey E. Barrett. Codon‑optimized UL12 and UL98 genes were synthesized by GenScript and subcloned into the pETM6T1‑SOX backbone by replacing the SOX coding sequence with either UL12 or UL98. The NusA tag was subsequently removed and pETM6T2-UL12.5 construct was created by deleting 126 amino acids (378 bp) from the N-terminal region using NEB’s Q5 Site-Directed Mutagenesis Kit (NEB #E0554).
Protein expression
All proteins bearing an N-terminal His tag were expressed in Escherichia coli BL21. For protein expression, a starter culture was prepared by inoculating cells into 100 mL of LB/2YT broth (Invitrogen) supplemented with kanamycin (25 µg/mL) and incubated overnight at 37°C with shaking. The following day, a 1:100 dilution of the overnight culture was transferred into 1.5-2L of fresh media and grown at 37°C until the optical density at 600 nm (A600) reached 0.6-0.8. Protein expression was induced by the addition of 1 mM isopropyl β-D-1-thiogalactopyranoside (IPTG), and cultures were incubated overnight at 18°C. Cells were harvested by centrifugation and stored at -80°C until further use.
Protein purification
HSV-1 UL12.5: Cell pellets were resuspended in buffer A (25 mM Tris-HCl, pH 8.5; 300 mM NaCl; 5 mM β-mercaptoethanol) supplemented with 10 mM imidazole and 1 mM PMSF, followed by sonication on ice. Lysates were clarified by centrifugation at >26,000 × g for 1 hour, and the supernatant was filtered through a 0.45 µm syringe filter before incubation with pre-equilibrated Ni-NTA resin (50% slurry) at 4°C for 1 hour with gentle shaking. Bound proteins were eluted using buffer A supplemented with 300 mM imidazole. Fractions containing the highest purity proteins were pooled, concentrated, and subjected to size exclusion chromatography using a Superdex 200 column pre-equilibrated with UL12 storage buffer (25 mM Tris-HCl, pH 8.5; 300 mM NaCl; 10% glycerol; 5 mM DTT). Fractions containing pure protein were verified by SDS-PAGE gel, pooled, concentrated, and stored at -70 °C until further use.
HCMV UL98: Cell pellets from 1.5 L of culture expressing UL98 were resuspended in buffer B (25 mM Tris‑HCl, pH 8.5; 200 mM NaCl) supplemented with cOmplete protease inhibitor, DNase I, 1 mM PMSF, and 5 mM β‑mercaptoethanol. Cells were lysed by sonication on ice, and the lysate was clarified by centrifugation at >26,000 × g for 1 h. The supernatant was filtered through a 0.45 µm syringe filter and incubated with pre‑equilibrated Ni‑NTA resin at 4 °C for 1 h with gentle agitation. Bound proteins were eluted with buffer B supplemented with 500 mM imidazole. The affinity tag was removed by addition of TEV protease during overnight dialysis against buffer B containing 2 mM DTT. The dialyzed sample was applied to a HiTrap Q FF column equilibrated in buffer B with 2 mM DTT and eluted using a linear NaCl gradient from 50 mM to 1 M. Fractions containing UL98 were pooled, concentrated, and further purified by size‑exclusion chromatography on a Superdex 200 column equilibrated in storage buffer (25 mM Tris‑HCl, pH 8.5; 300 mM NaCl; 10% glycerol; 5 mM DTT). Fractions containing pure protein were verified by SDS-PAGE gel, pooled, concentrated, and stored at -70 °C until further use.
KSHV SOX: KSHV SOX was purified as described by Bagneris et al. [41]. Briefly, cell lysates were prepared in buffer C (25 mM Tris-HCl, pH 8.5, 200 mM NaCl) supplemented with DNase I, PMSF, and cOmplete protease inhibitor. Clarified lysates were subjected to Ni-affinity chromatography, washed with buffer C containing 50 mM imidazole, and eluted with buffer C supplemented with 500 mM imidazole. The fusion protein was further purified by anion-exchange chromatography, followed by TEV protease-mediated tag removal during overnight dialysis in 25 mM Tris-HCl (pH 8.5), 200 mM NaCl, and 1 mM DTT. A second anion-exchange chromatography step followed by size-exclusion chromatography in 25 mM Tris-HCl (pH 8.5), 300 mM NaCl, and 10% glycerol were used to obtain purified SOX. Fractions containing pure protein were pooled, concentrated, and stored at -70°C until use.
Protein crystallization
Initial crystallization trials with 192 conditions from NeXtal JCSG Core Suite (II and III) were conducted at a protein concentration of 5 mg/mL using the sitting drop vapor diffusion method at 4 °C. Crystallization screening was automated with a Gryphon crystallization robot (Art Robbins Instruments), where 200 nL of protein solution was mixed with 200 nL of crystallization solution and equilibrated against 80 µL of reservoir solution. Initial crystal hits were obtained from the Core Suite II, conditions G8 containing 0.16 M ammonium sulfate, 0.08 M sodium acetate (pH 4.6), 20% (w/v) PEG 4000, 20% (v/v) glycerol. To improve crystal quality, the above conditions were further optimized using the hanging drop vapor diffusion method with 2 µL each of protein and reservoir solutions. Diffraction-quality crystals grew within a few days from the well solution comprising 0.12 M ammonium sulfate, 0.08 M sodium acetate (pH 5.0), 15% (w/v) PEG 4000, 10% (v/v) glycerol, and 4% DMSO. Current efforts are focused on generating co-crystals of the UL12 with the most potent inhibitors.
Data collection and structure solution
X-ray diffraction data were collected at the National Synchrotron Light Source II, using the 17-ID-1 (AMX) beamline at Brookhaven National Laboratory. Diffraction data were processed with XDS for indexing and integration [64], followed by space‑group determination with Pointless [65], which identified P 2₁ 2₁ 2₁ as the correct space group. Scaling and merging were performed in Aimless [66]. The structure of UL12 was determined by molecular replacement using Phaser [67] as part of the CCP4i2 suite of programs [68] employing an AlphaFold-predicted model as the initial search template [69,70]. Iterative rounds of model building and refinement were conducted using Coot [71] and Refmac [72]. Comprehensive data collection parameters, refinement statistics, and final structural metrics are summarized in S1 Table.
PicoGreen alkaline nuclease assay
Alkaline exonuclease (UL12) activity was assessed using the PicoGreen fluorescence assay, as previously described [33]. In brief, 25-µL reaction mixtures containing 20 mM Tris-HCl (pH 8.2), 40 mM NaCl, 1 mM MgCl2, 1 mM dithiothreitol (DTT), 10 nM UL12, and either DMSO or varying concentrations of inhibitors (with final DMSO concentration of 1.6%) were preincubated at room temperature for 10 minutes. The nuclease reaction was initiated by the addition of 1 nM linearized pUC119 DNA, followed by a 10-minute incubation at room temperature. Reactions were quenched with 10 µL of 125 mM EDTA (pH 8.0). For fluorescence detection, 150 µL of PreciseGreen dsDNA quantification reagent (Lumiprobe; diluted 1:500 in Tris-EDTA buffer, pH 7.5) was added to each quenched reaction. A total of 170 µL of the reaction mixture was transferred to a 96-well Flurotrac 200 black plate (Greiner Bio-One), and fluorescence was measured using a SpectraMax 3 plate reader with excitation and emission wavelengths of 480 nm and 520 nm, respectively.
The exonuclease activities of SOX and UL98 were assessed using similar assay conditions with a few variations. For SOX, reactions were carried out in a buffer containing 20 mM Tris-HCl (pH 8.8), 50 mM NaCl, 10 mM MgCl2, and 1 mM DTT. For UL98, the reaction buffer consisted of 50 mM Tris‑HCl (pH 8.8), 2 mM MgCl2, and 1 mM DTT. The final protein concentrations were 50 nM for SOX and 20 nM for UL98. All reactions were incubated at 37 °C for 10 minutes.
PicoGreen fluorescence is proportional to the amount of intact double-stranded DNA present in the reaction. Therefore, degradation of DNA by exonuclease results in reduced fluorescence, whereas inhibition of nuclease activity preserves DNA and yields increased fluorescence relative to the uninhibited enzyme control. Percent inhibition was calculated relative to the DMSO- control, and dose-response curves were generated by plotting percent inhibition as a function of inhibitor concentration. IC50 values were determined by nonlinear regression using a four-parameter logistic model (log[inhibitor] versus response, variable slope) in GraphPad Prism version 10.6.1 for Mac OS X, GraphPad Software, Boston, Massachusetts USA, www.graphpad.com. The raw fluorescence measurements and calculated inhibition values used for IC50 determination are provided in S1 Data.
Microscale Thermophoresis (MST)
Binding affinities of two most potent compounds, VNI-5168 and VNI-5162, with UL12 were determined using MST [54,55]. The protein was diluted to 200 nM in PBS-T buffer (137 mM NaCl, 2.5 mM KCl, 10 mM Na2HPO4, 2 mM KH2PO4, pH 7.4; 0.05% Tween-20) supplemented with 2.5 mM MgCl2 and 2.5 mM DTT and was labeled with RED-tris-NTA 2nd generation dye in the 1:2 ratio (100 nM dye, 200 nM protein) for 30 minutes at room temperature. Stock solutions of both compounds (20 mM in DMSO) were diluted to 5–10 μM in PBS-T buffer containing 3% DMSO, followed by 16 points, 1:1 serial dilution. Each ligand dilution was then mixed with an equal volume of labeled protein to yield final concentrations of 50 nM UL12 and 1.5% DMSO. After 20 minutes incubation time, all samples were centrifuged for 10 min at 12000 g in 4 °C and loaded into premium capillaries (Nanotemper Technologies). The measurements were performed on a NanoTemper MonolithNT.115 pico instrument. The initial fluorescence was uniform among the sixteen capillaries, and the thermographs showed no signs of aggregation or molecule adsorption to the capillary. Data was collected at 22°C, at 40% LED power and medium MST power. Three independent experiments were performed with technical duplicate. MO.Affinity Analysis software (v3.0.5, Nano Temper Technologies) was used to fit the data and to calculate the apparent Kd values with their respective 68.3% confidence intervals [CI] using a 1:1 binding model (see S1 Data for underlying measurements used to calculate Kd values).
Cells and viruses
African green monkey kidney epithelial cells (Vero) were obtained from ATCC and propagated in Dulbecco’s Modified Eagle Medium (DMEM) (Invitrogen, Carlsbad CA) supplemented with 5% fetal bovine serum (FBS, Gemini Bio-Products, Woodland CA). Human foreskin fibroblast (HFF) cells were purchased from ATCC and cultured in Eagle’s Minimal Essential Medium (MEM) (Invitrogen, Carlsbad, CA) supplemented with 10% FBS. Fluorescently labeled, recombinant eYFP-ICP0/gC-mCherry virus, a gift from Colin Crump (Cambridge, UK), expressing YFP tagged versions of ICP0 and mCherry tagged UL44 (gC) late gene was used in HSV-1 inhibition assays [57]. The TB40/E-mCherry-UL99eGFP, a recombinant HCMV TB40/E strain generously provided by Eain A. Murphy (Upstate Medical University, Syracuse, NY), expressing mCherry under SV40 promoter and GFP-tagged version of UL99 late gene was used in HCMV inhibition assay [57].
Antiviral experiments
HFF cells were seeded in 96-well clear plates in 5% FBS-MEM. At 24 h post seeding, cells were infected with eYFP-ICP0/gC-mCherry HSV-1 virus at a multiplicity of infection (MOI) of 1 PFU/cell for antiviral high-content image-based compound screen (fluorescent-based assay) or with MOI of 0.05 PFU/cell for yield-based assay and incubated for 20 h or 65 h, respectively, in the presence of various concentrations of the compounds or DMSO alone in 100 μL MEM at final concentrations of 2% FBS and 0.5% DMSO. For HCMV antiviral high-content image-based assays, cells were infected at MOI of 2 PFU/cell with fluorescently labeled TB40/E-mCherry-UL99eGFP and incubated for 65 h under the same experimental setup as for HSV-1 virus.
For HSV-1 inhibition assay, the entire plate was subjected to 2 freeze-thaw cycles, and viral yields were assessed by plaque assay on Vero cells as described previously [73]. For HSV-1 and HCMV antiviral high-content image-based assays at 20 h or 65 h post infection, respectively, plates were fixed with 2% Paraformaldehyde for 15 min at room temperature (RT) in the dark. Plates were washed twice with phosphate buffered saline (PBS) and wells were overlayed with PBS. Images were acquired with Agilent BioTek Cytation5 high-content microscope using 4X air objective. For HSV-1 images were collected using Channel 1: YFP (excitation 500nm, emission 542nm), Channel 2: Texas Red (excitation 586nm, emission 647nm) and for HCMV Channel 1: GFP (excitation 469nm, emission 525nm), Channel 2: Texas Red (excitation 586nm, emission 647nm). Images from both channels were acquired. For every image, mean intensity of the Texas Red (HSV-1) or GFP (HCMV) fluorescent representing true late gene expression in each virus was used for calculation corrected for the background from untreated cells. Signal obtained from DMSO treated infected cells was normalized to 100% infection (0% inhibition) and used for calculation of percent viral inhibition in infected cells treated with increasing concentrations of compound. The raw fluorescence intensity measurement and plaque-forming unit (PFU) counts, underlying the antiviral analyses are provided in S1 Data.
Cytotoxicity assay
HFF cells were seeded in 96-well white plates in 5% FBS-MEM. At 24-h post seeding, cells were incubated in the presence of various concentrations of the compounds or DMSO alone in 100 μL of MEM at final concentrations of 2% FBS and 0.5% DMSO. At 24 h and 65 h post-treatment, cytotoxicity was assessed using CellTiter Glo (Promega, Madison WI) luciferase assay according to the manufacturer’s instructions (Raw cytotoxicity measurement are provided in S1 Data).
Software and programs used
Molecular modelling and docking were performed using ProteinPrep, LigPrep and Glide in Schrodinger Maestro suite [48–51]. Multiple sequence alignment was performed using Clustal Omega [74,75]. Final structural figures were generated using PyMol Version 3.1.7.2 (Schrödinger, LLC). IC50 and EC50 values were obtained by nonlinear regression analysis using a four-parameter logistic model (log[inhibitor] versus response, variable slope) in GraphPad Prism version 10.6.1 for Mac OS X, GraphPad Software, Boston, Massachusetts USA, www.graphpad.com.
General synthetic scheme
Compounds were synthesized as shown in Fig 7. Commercially available pyrone 1 underwent aminolysis and ring transformation with the corresponding Boc-protected aminomethyl piperidine to afford protected hydroxypyridinone intermediate 2. Ester hydrolysis provided acid 3, which was coupled with the appropriate aniline to furnish amide intermediates 4. Subsequent Boc deprotection afforded free piperidine intermediates 5, which were functionalized with the corresponding sulfonyl chlorides to provide sulfonamide intermediates 6. Final debenzylation afforded the target hydroxypyridinone analogs. Detailed synthetic procedures and characterization data are provided in S1 Text and S1 File, respectively.
(a) EtOH, HOAc, 2-amino-2-phenylethan-1-ol; (b) NaOH, EtOH, 80°C; (c) DIPEA, Pybop, DMSO, DMAP, 4-bromoaniline; (d) DCM, TFA; (e) DMF, DIPEA, X-Cl; (f) TFA 90°C. X denotes the variable sulfonyl group used in the N-sulfonylation step to generate the corresponding sulfonamide analogs.
Supporting information
S1 Fig. UL12.5 retains nuclease activity comparable to full-length UL12.
The nuclease activities of purified full-length UL12 (FL‑UL12) and UL12.5 were evaluated using a PicoGreen-based assay with 1 nM linearized dsDNA substrate. Reactions (25 μL) contained 10 nM enzyme and were performed under standard assay conditions as described in the Methods. Fluorescence (RFU) is inversely proportional to nuclease activity, with lower fluorescence indicating increased DNA degradation. Data are shown as the mean ± SD of three technical replicates. Under the assay conditions used, UL12.5 exhibited nuclease activity comparable to that of full-length UL12.
https://doi.org/10.1371/journal.ppat.1014531.s001
(TIF)
S2 Fig. Sequence alignment of UL12, BGLF5, SOX, and Redα showing the conserved motif Thr‑Ala‑Ser in motif I of UL12 and λ exonuclease Redα and the Ser‑Ser‑Ser motif in the γ‑herpesvirus nucleases BGLF5 and SOX (black box).
https://doi.org/10.1371/journal.ppat.1014531.s002
(TIF)
S3 Fig. Conserved UL12 active‑site features and predicted inhibitor binding modes.
(A) Multiple sequence alignment highlighting conserved, identical, and essential catalytic active‑site residues (blue stars) and the 5′‑phosphate binding pocket defined by the WRSS motif (green stars). (B) Predicted binding site of the HQ inhibitor AK‑157 within the UL12 active site, showing its spatial separation from the WRSS pocket. (C) Predicted binding site of the HP inhibitor DD‑I‑38 within the UL12 active site, similarly positioned distal to the WRSS pocket. In (B) and (C), inhibitors are shown as yellow sticks, UL12 active‑site residues as dark gray sticks, coordinating water molecules as red spheres, and Mg2+ ions as cyan spheres. (D) Summary of the biological and biochemical activities of the HQ inhibitor AK‑157 and the HP inhibitor DD‑I‑38 shown in panels B and C respectively. Shown are the IC50 values determined in the UL12 nuclease inhibition assay, EC50 values determined in HSV‑1 (yield) and HCMV antiviral assays, CC50 values determined in HFF cells, and the corresponding selectivity indices (SI) calculated from these measurements.
https://doi.org/10.1371/journal.ppat.1014531.s003
(TIF)
S4 Fig. Michaelis–Menten plot showing initial reaction velocity (v₀) as a function of substrate concentration for UL12.
Experiments were performed four times, and data points represent the mean ± SEM from two independent experiments. The solid line represents a nonlinear least‑squares fit to the Michaelis–Menten equation using GraphPad Prism, from which the apparent Kₘ value was determined. The Km value and corresponding IC50 values were used to extrapolate the Ki using the Cheng–Prusoff equation.
https://doi.org/10.1371/journal.ppat.1014531.s004
(TIF)
S5 Fig. Dose response curves used to determine A) EC50 values for HSV-1 using yield assay in HFF cells infected with fluorescently tagged HSV-1 virus.
B) cytotoxicity on HFF cells at 65 h post-treatment was measured using Promega Cell Titer-Glo.
https://doi.org/10.1371/journal.ppat.1014531.s005
(TIF)
S6 Fig. Biochemical and antiviral characterization of analogs bearing a dibenzothiophene (VNI-5173) and dibenzofuran (VNI-5174) sidechain.
Dose-response analyses for antiviral activity to determine EC50 values for HSV-1 (A and B) and HCMV (C) in HFF cells infected with respective viruses in the presence of increasing concentration of the compound. Antiviral activity was evaluated using either high-content image-based assay or yield assay. (D) Dose-response analyses for enzymatic inhibition to determine IC50 values for full‑length (FL)-UL12 using the PicoGreen‑based nuclease assay. Percentage inhibition was calculated relative to the uninhibited DMSO control. IC50 and EC50 values were obtained by nonlinear regression analysis in GraphPad Prism using a four-parameter logistic model (log[inhibitor] versus response, variable slope). Data represent the mean of three independent experiments, with each experiment comprising 2–3 technical replicates.
https://doi.org/10.1371/journal.ppat.1014531.s006
(TIF)
S1 Table. Data collection, refinement, and validation statistics for HSV-1 UL12.5 (PDB ID: 10HO).
Highest-resolution shell values are shown in parentheses.
https://doi.org/10.1371/journal.ppat.1014531.s007
(PDF)
S1 Data. Primary and processed data underlying all figures and tables.
This file contains the raw experimental measurements used in this study, including PicoGreen fluorescence values, antiviral assay fluorescence intensities, viral yield measurements, and CellTiter-Glo luminescence values, together with the corresponding calculated percent inhibition, percent cytotoxicity and other raw values used to determine IC50, EC50, CC50, Kd, and Km parameters.
https://doi.org/10.1371/journal.ppat.1014531.s008
(XLSX)
S1 Text. Additional methods for steady-state kinetics and compound synthesis.
Detailed protocols for kinetic analyses, compound synthesis, purification, and characterization of viral nuclease inhibitors.
https://doi.org/10.1371/journal.ppat.1014531.s009
(PDF)
S1 File. Analytical characterization of synthesized viral nuclease inhibitors.
Mass spectra, HPLC chromatograms, and ^1H/^13C NMR spectra confirming compound identity and purity.
https://doi.org/10.1371/journal.ppat.1014531.s010
(PDF)
Acknowledgments
This research used resources at beamlines 17-ID-1 of the National Synchrotron Light Source II, a U.S. Department of Energy (DOE) Office of Science User Facility operated for the DOE Office of Science by Brookhaven National Laboratory under Contract No. DE-SC0012704. We thank Dr. Chathura Abeywickrama for collecting HPLC data for the compounds. AI‑assisted tools (ChatGPT and Microsoft Copilot) were used for language refinement of the manuscript.
References
- 1. World Health Organization WHO. Over 1 in 5 adults worldwide has a genital herpes infection – WHO. https://www.who.int/news-room/fact-sheets/detail/herpes-simplex-virus 2024. Accessed 2025 January 1.
- 2. James C, Harfouche M, Welton NJ, Turner KME, Abu-Raddad LJ, Gottlieb SL, et al. Herpes simplex virus: global infection prevalence and incidence estimates, 2016. Bull World Health Organ. 2020;98(5):315–29. pmid:32514197
- 3. Harfouche M, AlMukdad S, Alareeki A, Osman AMM, Gottlieb S, Rowley J, et al. Estimated global and regional incidence and prevalence of herpes simplex virus infections and genital ulcer disease in 2020: mathematical modelling analyses. Sex Transm Infect. 2025;101(4):214–23. pmid:39658199
- 4. Elion GB. Mechanism of action and selectivity of acyclovir. Am J Med. 1982;73(1A):7–13. pmid:6285736
- 5. Elion GB. The biochemistry and mechanism of action of acyclovir. J Antimicrob Chemother. 1983;12 Suppl B:9–17. pmid:6313600
- 6.
U.S. Food and Drug Administration. Zovirax (acyclovir) – injectable; NDA original approval. Silver Spring, MD: U.S. Food and Drug Administration. 1982.
- 7.
U.S. Food and Drug Administration. Drug approval package: Valtrex (valacyclovir hydrochloride) NDA 020487. Silver Spring, MD: U.S. Food and Drug Administration. 1995.
- 8.
U.S. Food and Drug Administration. Ganciclovir Injection – Original NDA Approval. Silver Spring, MD. 1989.
- 9.
U.S. Food and Drug Administration. Valcyte (valganciclovir hydrochloride) – original NDA approval. Silver Spring, MD. 2001.
- 10.
Mcintosh M, Hauschild B, Miller V. Human cytomegalovirus and transplantation: drug development and regulatory issues. 2016.
- 11.
Van Der Wekken-Pas LC, Totté J, Lunel FV, Van Zuilen A, Van Luin M. Therapeutic drug monitoring of ganciclovir in cytomegalovirus-infected patients with solid organ transplants and its correlation to efficacy and toxicity. 2023.
- 12.
U.S. Food and Drug Administration. Prevymis (letermovir) – original NDA approval. Silver Spring, MD: U.S. Food and Drug Administration. 2017.
- 13. Malnoë D, Bories M, Marchand T, Le Corre P. Drug-drug interaction between letermovir and ciclosporin in allogeneic haematopoietic cell transplantation recipients. J Antimicrob Chemother. 2025;80(5):1269–73. pmid:40036739
- 14. James SH. Letermovir Resistance in Hematopoietic Stem Cell Transplant Recipients: The Risks Associated with Cytomegalovirus Prophylaxis. J Infect Dis. 2020;221(7):1036–8. pmid:31784759
- 15. Chou S. Rapid In Vitro Evolution of Human Cytomegalovirus UL56 Mutations That Confer Letermovir Resistance. Antimicrob Agents Chemother. 2015;59(10):6588–93. pmid:26259791
- 16. Douglas CM, Barnard R, Holder D, Leavitt R, Levitan D, Maguire M, et al. Letermovir Resistance Analysis in a Clinical Trial of Cytomegalovirus Prophylaxis for Hematopoietic Stem Cell Transplant Recipients. J Infect Dis. 2020;221(7):1117–26. pmid:31781762
- 17. Hoffmann PJ, Cheng YC. The deoxyribonuclease induced after infection of KB cells by herpes simplex virus type 1 or type 2. I. Purification and characterization of the enzyme. J Biol Chem. 1978;253(10):3557–62. pmid:206546
- 18. Baylis SA, Purifoy DJ, Littler E. The characterization of the EBV alkaline deoxyribonuclease cloned and expressed in E. coli. Nucleic Acids Res. 1989;17(19):7609–22. pmid:2552412
- 19. Lin SF, Hsu TY, Liu MY, Lin LS, Yang HL, Chen JY, et al. Characterization of Epstein-Barr virus DNase and its interaction with the major DNA binding protein. Virology. 1995;208(2):712–22. pmid:7747443
- 20. Stolzenberg MC, Ooka T. Purification and properties of Epstein-Barr virus DNase expressed in Escherichia coli. J Virol. 1990;64(1):96–104. pmid:2152838
- 21. Sheaffer AK, Weinheimer SP, Tenney DJ. The human cytomegalovirus UL98 gene encodes the conserved herpesvirus alkaline nuclease. J Gen Virol. 1997;78 (Pt 11):2953–61. pmid:9367383
- 22. Gao M, Robertson BJ, McCann PJ, O’Boyle DR, Weller SK, Newcomb WW, et al. Functional conservations of the alkaline nuclease of herpes simplex type 1 and human cytomegalovirus. Virology. 1998;249(2):460–70. pmid:9791036
- 23. Yu D, Silva MC, Shenk T. Functional map of human cytomegalovirus AD169 defined by global mutational analysis. Proc Natl Acad Sci. 2003;100(21):12396–401.
- 24. Feederle R, Bannert H, Lips H, Müller-Lantzsch N, Delecluse H-J. The Epstein-Barr virus alkaline exonuclease BGLF5 serves pleiotropic functions in virus replication. J Virol. 2009;83(10):4952–62. pmid:19264771
- 25. Hartenian E, Mendez AS, Didychuk AL, Khosla S, Glaunsinger BA. DNA processing by the Kaposi’s sarcoma-associated herpesvirus alkaline exonuclease SOX contributes to viral gene expression and infectious virion production. Nucleic Acids Res. 2023;51(1):182–97. pmid:36537232
- 26. Grady LM, Szczepaniak R, Murelli RP, Masaoka T, Le Grice SFJ, Wright DL, et al. The Exonuclease Activity of Herpes Simplex Virus 1 UL12 Is Required for Production of Viral DNA That Can Be Packaged To Produce Infectious Virus. J Virol. 2017;91(23):e01380-17. pmid:28956767
- 27. Weller SK, Sawitzke JA. Recombination promoted by DNA viruses: phage λ to herpes simplex virus. Annu Rev Microbiol. 2014;68:237–58. pmid:25002096
- 28. Reuven N, Staire AE, Myers R, Weller SK. The herpes simplex virus type 1 alkaline nuclease and single-stranded DNA binding protein mediate strand exchange in vitro. Journal of Virology. 2003;77(13):7425–33.
- 29. Reuven NB, Willcox S, Griffith JD, Weller SK. Catalysis of strand exchange by the HSV-1 UL12 and ICP8 proteins: potent ICP8 recombinase activity is revealed upon resection of dsDNA substrate by nuclease. J Mol Biol. 2004;342(1):57–71. pmid:15313607
- 30. Mullon PJ, Maldonado-Luevano E, Mehta KPM, Mohni KN. The herpes simplex virus alkaline nuclease is required to maintain replication fork progression. J Virol. 2024;98(12). pmid:39508568
- 31. Martinez R, Sarisky RT, Weber PC, Weller SK. Herpes simplex virus type 1 alkaline nuclease is required for efficient processing of viral DNA replication intermediates. J Virol. 1996;70(4):2075–85. pmid:8642627
- 32. Corcoran JA, Saffran HA, Duguay BA, Smiley JR. Herpes simplex virus UL12.5 targets mitochondria through a mitochondrial localization sequence proximal to the N terminus. J Virol. 2009;83(6):2601–10. pmid:19129438
- 33. DiScipio KA, Weerasooriya S, Szczepaniak R, Hazeen A, Wright LR, Wright DL, et al. Two-Metal Ion-Dependent Enzymes as Potential Antiviral Targets in Human Herpesviruses. mBio. 2022;13(1):e0322621. pmid:35073739
- 34. Steitz TA, Steitz JA. A general two-metal-ion mechanism for catalytic RNA. Proc Natl Acad Sci U S A. 1993;90(14):6498–502. pmid:8341661
- 35. Yang W. Nucleases: diversity of structure, function and mechanism. Q Rev Biophys. 2011;44(1):1–93.
- 36. Esposito F, Tramontano E. Past and future. Current drugs targeting HIV-1 integrase and reverse transcriptase-associated ribonuclease H activity: single and dual active site inhibitors. Antivir Chem Chemother. 2014;23(4):129–44. pmid:24150519
- 37. Semenova EA, Johnson AA, Marchand C, Davis DA, Yarchoan R, Pommier Y. Preferential inhibition of the magnesium-dependent strand transfer reaction of HIV-1 integrase by alpha-hydroxytropolones. Mol Pharmacol. 2006;69(4):1454–60. pmid:16418335
- 38. Mouscadet J-F, Tchertanov L. Raltegravir: molecular basis of its mechanism of action. Eur J Med Res. 2009;14 Suppl 3(Suppl 3):5–16. pmid:19959411
- 39. McColl DJ, Chen X. Strand transfer inhibitors of HIV-1 integrase: bringing IN a new era of antiretroviral therapy. Antiviral Res. 2010;85(1):101–18. pmid:19925830
- 40. Buisson M, Géoui T, Flot D, Tarbouriech N, Ressing ME, Wiertz EJ, et al. A bridge crosses the active-site canyon of the Epstein-Barr virus nuclease with DNase and RNase activities. J Mol Biol. 2009;391(4):717–28. pmid:19538972
- 41. Bagnéris C, Briggs LC, Savva R, Ebrahimi B, Barrett TE. Crystal structure of a KSHV-SOX-DNA complex: insights into the molecular mechanisms underlying DNase activity and host shutoff. Nucleic Acids Res. 2011;39(13):5744–56. pmid:21421561
- 42. Lee H, Patschull AOM, Bagnéris C, Ryan H, Sanderson CM, Ebrahimi B, et al. KSHV SOX mediated host shutoff: the molecular mechanism underlying mRNA transcript processing. Nucleic Acids Res. 2017;45(8):4756–67. pmid:28132029
- 43. Dahlroth S-L, Gurmu D, Schmitzberger F, Engman H, Haas J, Erlandsen H, et al. Crystal structure of the shutoff and exonuclease protein from the oncogenic Kaposi’s sarcoma-associated herpesvirus. FEBS J. 2009;276(22):6636–45. pmid:19843164
- 44. Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O, et al. Highly accurate protein structure prediction with AlphaFold. Nature. 2021;596(7873):583–9. pmid:34265844
- 45. Kovall R, Matthews BW. Toroidal structure of lambda-exonuclease. Science. 1997;277(5333):1824–7. pmid:9295273
- 46. Kovall RA, Matthews BW. Structural, functional, and evolutionary relationships between lambda-exonuclease and the type II restriction endonucleases. Biochemistry. 1998;95:7893–7.
- 47. Zhang J, McCabe KA, Bell CE. Crystal structures of lambda exonuclease in complex with DNA suggest an electrostatic ratchet mechanism for processivity. Proc Natl Acad Sci U S A. 2011;108(29):11872–7. pmid:21730170
- 48. Maiti M, Lakshminarayanan A, Sastry S. Characterization of void space in polydisperse sphere packings: Applications to hard-sphere packings and to protein structure analysis. Eur Phys J E Soft Matter. 2013;36(1):5. pmid:23355091
- 49. Lu C, Wu C, Ghoreishi D, Chen W, Wang L, Damm W, et al. OPLS4: Improving Force Field Accuracy on Challenging Regimes of Chemical Space. J Chem Theory Comput. 2021;17(7):4291–300. pmid:34096718
- 50. Halgren TA. Identifying and characterizing binding sites and assessing druggability. J Chem Inf Model. 2009;49(2):377–89. pmid:19434839
- 51. Friesner RA, Murphy RB, Repasky MP, Frye LL, Greenwood JR, Halgren TA, et al. Extra precision glide: docking and scoring incorporating a model of hydrophobic enclosure for protein-ligand complexes. J Med Chem. 2006;49(21):6177–96. pmid:17034125
- 52. Fujii H, Kato A, Mugitani M, Kashima Y, Oyama M, Kozuka-Hata H, et al. The UL12 protein of herpes simplex virus 1 is regulated by tyrosine phosphorylation. J Virol. 2014;88(18):10624–34. pmid:24991005
- 53. Cheng Y, Prusoff WH. Relationship between the inhibition constant (K1) and the concentration of inhibitor which causes 50 per cent inhibition (I50) of an enzymatic reaction. Biochem Pharmacol. 1973;22(23):3099–108. pmid:4202581
- 54. Jerabek-Willemsen M, André T, Wanner R, Roth HM, Duhr S, Baaske P. MicroScale Thermophoresis: Interaction analysis and beyond. J Mol Struct. 2014;1077:101–13.
- 55. Jerabek-Willemsen M, Wienken CJ, Braun D, Baaske P, Duhr S. Molecular interaction studies using microscale thermophoresis. Assay Drug Dev Technol. 2011;9(4):342–53. pmid:21812660
- 56. Scherer KM, Manton JD, Soh TK, Mascheroni L, Connor V, Crump CM, et al. A fluorescent reporter system enables spatiotemporal analysis of host cell modification during herpes simplex virus-1 replication. J Biol Chem. 2021;296:100236. pmid:33380421
- 57. O’Connor CM, Nukui M, Gurova KV, Murphy EA. Inhibition of the FACT Complex Reduces Transcription from the Human Cytomegalovirus Major Immediate Early Promoter in Models of Lytic and Latent Replication. J Virol. 2016;90(8):4249–53. pmid:26865717
- 58. Smith SJ, Zhao XZ, Passos DO, Lyumkis D, Burke TR Jr, Hughes SH. Integrase Strand Transfer Inhibitors Are Effective Anti-HIV Drugs. Viruses. 2021;13(2):205. pmid:33572956
- 59. Wilcken R, Zimmermann MO, Lange A, Joerger AC, Boeckler FM. Principles and applications of halogen bonding in medicinal chemistry and chemical biology. J Med Chem. 2013;56(4):1363–88. pmid:23145854
- 60. Reuven NB, Antoku S, Weller S. The UL12.5 gene product of herpes simplex virus type 1 exhibits nuclease and strand exchange activities but does not localize to the nucleus. J of Virology. 2004;78(9):4599–608. pmid:15078942
- 61. Schumacher AJ, Mohni KN, Kan Y, Hendrickson EA, Stark JM, Weller SK. The HSV-1 exonuclease, UL12, stimulates recombination by a single strand annealing mechanism. PLoS Pathog. 2012;8(8):e1002862. pmid:22912580
- 62.
Myers R, Rudd KE. Mining DNA sequences for molecular enzymology: The red alpha superfamily defines a set of recombination nucleases. In: Proceedings of the 1998 Miami Nature Biotechnology Winter Symposium, 1998. 49–50.
- 63. Iyer LM, Koonin EV, Aravind L. Classification and evolutionary history of the single-strand annealing proteins, RecT, Redbeta, ERF and RAD52. BMC Genomics. 2002;3:8. pmid:11914131
- 64. Kabsch W. XDS. Acta Crystallogr D Biol Crystallogr. 2010;66(2):125–32.
- 65. Evans P. Scaling and assessment of data quality. Acta Crystallogr D Biol Crystallogr. 2006;62(Pt 1):72–82. pmid:16369096
- 66. Evans PR, Murshudov GN. How good are my data and what is the resolution? Acta Crystallographica Section D. 2013;69(7):1204–14.
- 67. McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. Phaser crystallographic software. J Appl Crystallogr. 2007;40(Pt 4):658–74. pmid:19461840
- 68. Agirre J, Atanasova M, Bagdonas H, Ballard CB, Baslé A, Beilsten-Edmands J, et al. The CCP4 suite: integrative software for macromolecular crystallography. Acta Crystallogr D Struct Biol. 2023;79(Pt 6):449–61. pmid:37259835
- 69. Soh TK, Ognibene S, Sanders S, Schäper R, Kaufer BB, Bosse JB. A proteome-wide structural systems approach reveals insights into protein families of all human herpesviruses. Nat Commun. 2024;15(1):10230. pmid:39592652
- 70. McCoy AJ, Sammito MD, Read RJ. Implications of AlphaFold2 for crystallographic phasing by molecular replacement. Acta Crystallogr D Struct Biol. 2022;78(Pt 1):1–13. pmid:34981757
- 71. Emsley P, Lohkamp B, Scott WG, Cowtan K. Features and development of Coot. 2010;:486–501.
- 72. Murshudov GN, Vagin AA, Dodson EJ. Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr D Biol Crystallogr. 1997;53(Pt 3):240–55. pmid:15299926
- 73. Livingston CM, Ifrim MF, Cowan AE, Weller S. Virus-Induced Chaperone-Enriched (VICE) Domains Function as Nuclear Protein Quality Control Centers during HSV-1 Infection. PLoS Pathog. 2009;5(10):1000619.
- 74. Madeira F, Madhusoodanan N, Lee J, Eusebi A, Niewielska A, Tivey ARN, et al. The EMBL-EBI Job Dispatcher sequence analysis tools framework in 2024. Nucleic Acids Res. 2024;52(W1):W521–5. pmid:38597606
- 75. Gouet P, Robert X, Courcelle E. ESPript/ENDscript: Extracting and rendering sequence and 3D information from atomic structures of proteins. Nucleic Acids Res. 2003;31(13):3320–3. pmid:12824317