EBV BALF2 DNA annealing intermediate structure reveals the mechanism of annealing during recombination

EBV BALF2 DNA annealing intermediate structure reveals the mechanism of annealing during recombination | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article EBV BALF2 DNA annealing intermediate structure reveals the mechanism of annealing during recombination Gökhan Tolun, Jordan Nicholls, Jodi Brewster, Nehad El Salamouni, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6794668/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract Epstein-Barr virus is an oncogenic herpesvirus present in 95% of the global population. It encodes the highly conserved BALF2 protein as an essential member of its replisome. BALF2 is a multifunctional protein which acts as a general single-stranded DNA-binding protein during replication, and as an ATP-independent recombinase involved in the single-strand annealing homologous recombination pathway. Several lines of evidence suggest that homologous recombination is an integral feature of herpesvirus DNA replication, required for the generation of concatemeric replication intermediates, genomic maintenance, and as a major driver of genetic diversity. BALF2 and its homologues are therefore promising antiviral targets. Despite over half a century of research into the herpesvirus annealase proteins, a significant roadblock persists in our understanding of their binding and annealing mechanisms. Here, we present a structure of a BALF2 DNA annealing intermediate, determined to 2.2 Å resolution by cryogenic electron-microscopy (cryo-EM). This structure allowed for the identification and characterisation of an oligonucleotide-binding fold, a zinc-binding loop, an active site of ssDNA-annealing, and suggests a model for cooperative binding and oligomerisation. We also investigated BALF2 through biochemical assays, bioinformatic sequence analysis and molecular dynamics simulations to further characterise regions of the protein’s structure. These findings will strongly inform future studies on herpesvirus annealases and have great potential as a starting point for structure-based drug design. Biological sciences/Structural biology/Electron microscopy/Cryoelectron microscopy Biological sciences/Molecular biology/DNA recombination Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction Epstein-Barr Virus and BALF2 The Orthoherpesviridae family is a division of large spherical viruses containing relatively complex linear dsDNA genomes which vary between 125–230 kb in length (Davison et al., 2009 ). Orthoherpesviridae is divided into three subfamilies; α-herpesvirinae , β-herpesvirinae , and γ-herpesvirinae ; based on replication strategy, host-range, and evolutionary lineage (Sehrawat et al., 2018 ). Of the nine species that infect humans, the γ-herpesvirus human herpesvirus 4 (HHV-4), also known as Epstein-Barr virus (EBV), is among the most common. It is highly contagious and boasts a global lifetime prevalence of ~ 95% (Bakkalci et al., 2020 ). It is associated primarily with infectious mononucleosis but is also an oncovirus that has been implicated in the incidence of both lymphoma and carcinoma-type cancers. It is estimated ~ 85% of nasopharyngeal carcinoma incidence worldwide is EBV-linked (Wong et al., 2022 , Baumforth et al., 1999 ). EBV infection is characterised by asymptomatic periods of latency, during which viral DNA resides primarily as an episome within B-lymphocytes or epithelial cells and is replicated passively every cell cycle by host cell replication machinery (Serquiña and Ziegelbauer, 2017 , Hammerschmidt and Sugden, 2013 ). These latent periods are broken up by periods of lytic replication, where a virally derived replisome carries out amplification of viral DNA at specified areas of the nucleus known as replication compartments (Hammerschmidt and Sugden, 2013 , Fixman et al., 1992). EBV replication intermediates appear initially as long concatemeric molecules, before being packaged into mature virions (Hammerschmidt and Sugden, 1988 ). The essential proteins of the replisome include an expression factor and origin-binding protein (BZLF1); a viral polymerase (BALF5) and processivity factor (BMRF1); a heterotrimeric primase-helicase complex (BSLF1-BBLF4-BBLF2/3); and a multifunctional annealase/single-stranded DNA-binding protein (SSB), called BALF2 (Tsurumi, 2001 , Hammerschmidt and Sugden, 2013 ). This replisome is highly conserved in all human herpesviruses. BALF2 is a 123 kDa zinc metalloprotein essential for virion production (Angel et al., 1987 , Zhang et al., 1988 , Decaussin et al., 1995 , Mumtsidu et al., 2008 ). Though literature on BALF2 is relatively sparse, reference to the homologous ICP8 from α-herpesvirus human simplex virus-1 (HSV-1), which shares 30% identity with BALF2, and its closest homologue ORF6 from γ-herpesvirus Karposi’s sarcoma-associated herpesvirus (KSHV), which shares 41% identity, attributes several assumed functions to BALF2 (Quinn and McGeoch, 1985 , Nicholas et al., 1997 , Wu et al., 2001 , Ozgur et al., 2011 ). BALF2 was first identified through both genetic and biochemical homology to ICP8; and antigenic cross-reactivity with ICP8-antibody (Quinn and McGeoch, 1985 , Zhang et al., 1988 , Angel et al., 1987 ). It was defined as essential for viral production when Raji cells, which are non-producing BALF2-deficient carriers of the virus, successfully replicated viral DNA after transfection with the BALF2 gene (Decaussin et al., 1995 , Zhang et al., 1988 , Fixman et al., 1995 ). Through homology to ICP8, BALF2 is hypothesised to localise at punctate regions called pre-replicative sites where it recruits members of the replisome to facilitate the formation of replication compartments (Darwish et al., 2016 , Taylor et al., 2003 , Uprichard and Knipe, 2003 , de Bruyn Kops and Knipe, 1988 , Bush et al., 1991 ). As an SSB, it stabilises the replication fork to streamline polymerase activity and protects ssDNA from nuclease digestion (Hernandez and Lehman, 1990 , Tsurumi et al., 1996 , Tsurumi et al., 1998 , Decaussin et al., 1995 , O'Donnell et al., 1987 ). In addition, it also has several interactions with other members of the viral replisome, as well as the virally-encoded alkaline exonuclease BGLF5 (Zeng et al., 1997 , Lin et al., 1995 , Gao et al., 1998 , Fujii et al., 2000 , Calderwood et al., 2007 , Hara et al., 2022 ). BALF2 binding to ssDNA is preferential, non-specific, cooperative, and direction independent (Tsurumi et al., 1998 , Tsurumi et al., 1996 ). A BALF2 deletion mutant missing the C-terminal 60 residues (BALF2ΔC) was observed to bind to ssDNA as monomers forming complexes which appeared as irregular bead-like nucleoprotein filaments (Mumtsidu et al., 2008 ). The number of nucleotides covered by a single herpesvirus annealase monomer is approximately ~ 14–20 nt, though this has historically varied considerably (Mumtsidu et al., 2008 , Tsurumi et al., 1996 , Ozgur and Griffith, 2014 , Gourves et al., 2000 ). Affinity to ssDNA is comparable across ICP8 and ORF6 ( K d = ~ 0.1–0.5 µM) but has not been investigated for BALF2 (Weerasooriya et al., 2019 , Gourves et al., 2000 , Darwish et al., 2016 , Dudas and Ruyechan, 1998 , Ozgur and Griffith, 2014 ). For these homologues, binding to dsDNA or RNA is approximately equal and is 5-fold weaker than ssDNA-binding (Boehmer, 2004 , Lee and Knipe, 1985 , Ruyechan and Weir, 1984 ). There is also in vitro evidence that BALF2 can melt short sections of dsDNA, inferring an ability to melt secondary structure at the replication fork (Mumtsidu et al., 2008 , Tsurumi et al., 1998 ). This is a putative feature of this family of proteins (Boehmer and Lehman, 1993 , Wang and Hall, 1990 ). Exonuclease-Annealase Two-Component Recombinase (EATR) Systems dsDNA breaks (DSBs) are considered one of the most detrimental DNA mutations. Many organisms utilise aspects of homologous recombination (HR) for the repair of these breaks. HR relies on the manipulation of homologous sequences to use as substrates in DNA-annealing reactions to repair these DSBs through recombination events, often at the cost of genetic deletions. As a result, HR is also a major contributor to genome maintenance, efficient DNA replication, and is a powerful driver of genetic diversity (Fu et al., 2002 , Kowalczykowski et al., 1994 ). Exonuclease-annealase two-component recombinase (EATR) systems describe a highly conserved protein interaction which carries out HR through the single-strand annealing (SSA) pathway. Though ubiquitous, they are particularly common in the genomes of linear dsDNA viruses (Weller and Sawitzke, 2014 ). Canonically, SSA is carried out by a 5′-3′ exonuclease, which acts to expose ssDNA, and an annealase protein which facilitates annealing of two homologous single strands. They are differentiated from other HR pathways by an ATP-independent mechanism and a direct interaction between the two components. The model EATR system is the λ-Exonuclease/Redβ system from bacteriophage-λ, which is also referred to as the Red system (Signer and Weil, 1968 ). It is shown that the Red system may facilitate incorporation of Okazaki-like fragments into replication products through homologous pairing, in addition to its putative DSB repair function (Lin et al., 1984 , Mosberg et al., 2010 , Maresca et al., 2010 , Brewster and Tolun, 2020 , Newing et al., 2022 ). As a result, it has been utilised in the biotechnological process known as recombineering, a highly accurate method of in vivo genetic manipulation (Mosberg et al., 2010 , Weller and Sawitzke, 2014 , Czarniak and Hensel, 2015 , Valledor et al., 2018, Fels et al., 2020, Fitschen et al., 2024 ). In addition to its role as an SSB, BALF2 is reported to act as the annealase component in an EATR system, with the BGLF5 viral exonuclease, analogous to the λExo/Redβ complex (Weller and Sawitzke, 2014 , Valledor et al., 2018). Many lines of evidence support an intimate coupling of herpesvirus DNA replication with homologous recombination (Dutch et al., 1995 ). Inversion of the long and short regions of the herpes simplex genomes, flanked by homologous repeats, commonly occurs during lytic replication (Mocarski and Roizman, 1982 , Sheldrick and Berthelot, 1975 , Wagner and Summers, 1978 ). This process is so prevalent that chimeric viral strains can be produced this way (Amundsen and Parris, 1984 ). These terminal repeats in EBV facilitate circularisation of the viral genome, which allows replication to proceed initially via the θ-model (Zimmermann and Hammerschmidt, 1995 ). A switch to rolling-circle replication initiates the formation of long head-to-tail concatemeric replication intermediates; however, these are produced at logarithmic rates and may adopt branched conformations. This is inconsistent with a traditional rolling-circle mechanism but may be explained by HR events (Jacob and Roizman, 1977 , Severini et al., 1996). Additionally, herpesvirus DNA inherently contains many nicks and gaps (Smith et al., 2014 , Wilkie, 1973 , Jacob and Roizman, 1977 ). While these have been implicated in the recruitment of host-cell recombination machinery, they have also been suggested to act as dsDNA-breaking points during replication, which are inherently recombinogenic (Smith et al., 2014 , Wilkinson and Weller, 2003 ). The further processing of concatemeric products into unit-length viral DNA for packaging into capsids is believed to depend on exonuclease-dependent recombination events within the flanking terminal repeats of each unit (Zimmermann and Hammerschmidt, 1995 , Feederle et al., 2009 ). BGLF5 is annotated as an alkaline exonuclease (Zhang et al., 1987 , Baylis et al., 1989 ). Like λExo, it is a member of the D(D/E)xK family of nucleases, and they share sequence homology. Likewise, BGLF5 also possesses magnesium-dependent 5′-3′ exonuclease activity, and additionally can create nicks in dsDNA and breaks in ssDNA through endonuclease activity (Baylis et al., 1989 , Stolzenberg and Ooka, 1990 , Lin et al., 1994 ). dsDNA digestion proceeds processively, while ssDNA digestion occurs sporadically (Lin et al., 1994 ). While not explicitly essential, it is required for processing and packaging viral DNA into capsids, and thus greatly enhances viral progeny (Feederle et al., 2009 ). Though not well defined, several studies have provided evidence of a direct BGLF5-BALF2 interaction (Calderwood et al., 2007 , Hara et al., 2022 , Lin et al., 1995 ). In herpesviruses, the best characterised EATR is the HSV-1 UL12/ICP8 system (Weller and Sawitzke, 2014 , Reuven et al., 2003 , Valledor et al., 2018). BGLF5/BALF2 is evidenced to perform a very similar role, although it has not been studied as thoroughly (Tsurumi et al., 1996 , Tsurumi et al., 1998 , Lin et al., 1995 ). ICP8 is capable of ATP- and cofactor-independent melting of short regions of DNA, and can subsequently anneal homologous substrates to these regions in the presence of Mg 2+ (Boehmer and Lehman, 1993 , Dutch and Lehman, 1993 , Bortner et al., 1993 ). This annealing reaction is second order and relies first on the binding of ssDNA by annealase, followed by an interaction between coated substrates via a filamentous annealing complex which facilitates annealing (Dutch and Lehman, 1993 , Makhov and Griffith, 2006 , Weerasooriya et al., 2019 ). The interaction involves homology searching between the two strands, before the dsDNA is ejected from the complex. ICP8-facilitated annealing is direction-independent but limited to substrates with overhanging ssDNA, which can be generated by UL12 exonuclease activity; moreover, UL12 experiences an ICP8-specific increase to nuclease efficiency (Reuven et al., 2003 , Reuven et al., 2004, Reuven and Weller, 2005 ). When both components are present, significantly higher levels of SSA are detected in vivo , and other methods of recombination are downregulated (Schumacher et al., 2012). Like BGLF5, UL12 is also implicated in the processing of DNA concatemers for processing into viral capsids (Martinez et al., 1996, Goldstein and Weller, 1998 , Weller et al., 1990 , Grady et al., 2017 ). Recently, Valledor et al. (2018) demonstrated a proof of concept that this system could be used in controlled recombineering practices in eukaryotic cells, which could not be performed using the prokaryotic λ-Exonuclease/Redβ EATR. Though functional homologues of EATR systems are ubiquitous, resolving the phylogenetic relationships between annealases has been difficult due to low sequence similarity. There are three major families of DNA-annealing proteins: the Rad51-like family, which carry out HR through an ATP-dependent strand-invasion mechanism; and the gp-like and Rad52-like families which contain the annealases capable of ATP-independent SSA through EATR activity. The latter is composed of subfamilies Rad52, RecT/Redβ, Erf, and Sak3 (Iyer et al., 2002, Lopes et al., 2010 ). The gp-like family, defined by T7 bacteriophage gp2.5 protein, and the Rad52-like family, defined by human Rad52, are separated based on virulence of the source organism and specificity of their oligonucleotide-binding fold (OB-fold) (Lopes et al., 2010 , Cernooka et al., 2017 ). BALF2 does not formally belong to either annealase family, however, ICP8 has been suggested to share some structural motifs with an SSB from Enterobacter carcinogenesis phage Enc34, a putative member of the gp-like family, despite a marked size difference between the two proteins (128 kDa vs 26 kDa, respectively) (Cernooka et al., 2017 , Hernandez and Richardson, 2019 , Kazlauskas and Venclovas, 2012 ). BALF2 is reported to behave as a monomer in solution (Tsurumi et al., 1996 , Tsurumi et al., 1998 , Mumtsidu et al., 2008 ). It is also reported to form a concentration-dependent dimer (Mumtsidu et al., 2008 ). ICP8 is known to form tight left-handed bipolar filaments in the presence of Mg 2+ and absence of ssDNA (Makhov et al., 2009 , Makhov and Griffith, 2006 , Weerasooriya et al., 2019 , O'Donnell et al., 1987 , Darwish et al., 2016 ). During annealing, ICP8 appears as a super-helical complex (Makhov and Griffith, 2006 ). Similarly, the homologue ORF6 is also reported to form DNA-free left-handed bipolar filaments (Ozgur et al., 2011 , Ozgur and Griffith, 2014 ). In contrast, ORF6 filaments have a looser turn and are physically distinct from those formed by ICP8. They also can form in the absence of Mg 2+ but are dependent on a reducing environment. Both filaments have been suggested as scaffolds to which ssDNA may bind during annealing; each strand coated by each protofilament of the helix, consistent with the current annealing model (Makhov et al., 2009 , Ozgur and Griffith, 2014 , Tolun et al., 2013 ). An additional role of these filaments is thought to be in the formation of pre-replicative sites, as a scaffold and point of localisation for the replisome (Darwish et al., 2016 ). Removal of the C-terminal 60 residues (ΔC) abolishes the ability for these proteins to form filaments, in addition to decreasing cooperative binding (Makhov et al., 2009 , Ozgur and Griffith, 2014 , Mumtsidu et al., 2008 , Mapelli et al., 2000 ). No such filaments or high-order oligomers have been reported for BALF2, though the full-length protein has not been tested (Mumtsidu et al., 2008 ). BALF2 Structure Though very little is known about the structure of BALF2, the crystal structure of ICP8ΔC is available (Mapelli et al., 2005 ). It is composed of a large non-contiguous N-terminal domain (NTD) connected via a linker to a smaller C-terminal domain (CTD). This has been fit into low-resolution negative-staining electron microscopy maps of the ssDNA-free filament, as well as a toroidal annealing intermediate, which is thought to represent an initial turn of the filament (Tolun et al., 2013 , Makhov et al., 2009 , Makhov and Griffith, 2006 ). In the latter, annealing is proposed to occur at the interface of two monomers, which is consistent with the spatial arrangement of ICP8 in the protein-only filaments. Additionally, ICP8ΔC is known to coordinate zinc through a CCCH motif in the N-terminal domain (Mapelli et al., 2005 ). BALF2 is also proposed to bind zinc in a 1:1 ratio, though since only the cysteine residues are conserved it was unknown how this occurs (Mumtsidu et al., 2008 ). Similarly, the role of magnesium is not fully understood. Although it is required for filament formation and annealing by ICP8, it is not required for filament formation by ORF6 (Makhov et al., 2009 , Makhov and Griffith, 2006 , Weerasooriya et al., 2019 , O'Donnell et al., 1987 , Darwish et al., 2016 , Ozgur et al., 2011 , Ozgur and Griffith, 2014 ). Despite the clinical relevance of herpesviruses, our understanding of the homologous recombination mechanisms essential for herpesvirus DNA replication is limited. We therefore investigated DNA binding and annealing by BALF2 annealase, using cryogenic electron-microscopy (cryo-EM), biochemical assays, sequence analysis and molecular dynamics simulations. Here, we present the structure of full-length BALF2 captured as a novel filamentous annealing intermediate at a resolution of 2.2 Å. The structure of the monomer consists of a large non-contiguous NTD containing an OB-fold and zinc-finger loop; connected by a short linker to a small CTD which docks into a neighbouring NTD, driving filament formation through domain-swapping. The filament is a bipolar arrangement of dimeric asymmetric units that gives the ssDNA its correct directionality for annealing. Each monomer of the asymmetric unit holds ssDNA in a planar orientation to anneal 3 bases at a time at the dimeric interface. These findings provide key insights into the oligomerisation and cooperative ssDNA-binding and annealing mechanisms possessed by this class of proteins, and into DNA replication and repair by herpesviruses. Methods and Materials DNA Design DNA sequences used in this study are provided in Table 1 . pFASTBac1 plasmid containing BALF2 between restriction enzyme sites Eco RI and Hind III was ordered from Gene Universal (pFASTBac1- BALF2 ) (Delaware, USA). Primers for PCR amplification (P1-P4) and for sequencing (S1-S5) were ordered from Integrated DNA Technologies (IDT) (Coralville, Iowa). Phosphorylated oligonucleotides for binding assays (O1-O3) were also ordered from IDT. Table 1 Oligonucleotide sequences used for this study. Complementary regions in O2 for formation of a self-dimer are shown in bold. The overhanging region is underlined. Name Sequence (5′–3′) Description P1 GCTAGAATTCATGCAGGGTGCACAGACT Forward primer for amplification of BALF2 from pFASTBac1. Includes Eco RI cut site. P2 CTATAAGCTTCTACGCCTCTGGTTCGACCTCGAGTCCGGGGAG Reverse primer for amplification of BALF2 from pFASTBac1. Includes HindIII cut site and EPEA affinity tag (C-tag) sequence. P3 ATTTCAGGTGGCACTTTTCG Forward primer for amplification of ampicillin resistance gene from pFASTBac1. P4 CTGACAGTTACCAATGCTTAATCAG Forward primer for amplification of ampicillin resistance gene from pFASTBac1. O1 TAGCCGTATGTCATCCGCAAAAATCGAGCTATGCAGGGCGATTCTGCTCTAAGCCACAGT 60-mer oligonucleotide, 5′-labelled with Cy5 fluorophore. Complementary to O2. O2 ACTGTGGCTTAGAGCAG AATCG CCCTGCATAGCT CGATT TTTGCGGATGACATA CGGCTA 60-mer oligonucleotide, 3′-labelled with Cy3 fluorophore. Complementary to O1. O3 TCGACCACTAGCCATGCCATTGCCTCTTAGACACCCCGATACAGTGATTATGAAAGGTAT 60-mer oligonucleotide, 3′-labelled with Cy3 fluorophore. Non-complementary to O1. S1 CGCTCTACGACAAGGAG Internal BALF2 sequencing primer. S2 AAACTACGCTGTGGAGCAC Internal BALF2 sequencing primer. S3 CCCAGCTGTTTTACCGC Internal BALF2 sequencing primer. S4 CAACGTCATAGATGTGGTGC Internal BALF2 sequencing primer. S5 TGAGAACATCAGGGCTGG Internal BALF2 sequencing primer. Cloning The BALF2 sequence was PCR amplified using pFASTBac1- BALF2 as a template with forward primer (P1) to flank the gene with an Eco RI cut site and reverse primer (P2) to flank the gene with a C-tag affinity tag followed by a Hind III cut site. Strain AN1459 E. coli containing pFASTBac1, kindly provided by the Mace Lab, was grown in an overnight culture in the presence of 50 µg/ml ampicillin and DNA isolated by QIAprep® Spin Miniprep Kit (QIAGEN). Isolated pFASTBac1 and PCR-amplified BALF2 were double digested with Eco RI-HF and Hind III-HF (NEB) overnight and gel-purified using GIAEX®II Gel Extraction Kit (QIAGEN). The recombinant plasmid, pFASTBac1 containing BALF2 flanked by a C-tag sequence (pFASTBac1- BALF2 ), was then generated by ligation using an insert:vector ratio of 1:5. The newly constructed recombinant plasmid was confirmed by diagnostic digest using Eco RI-HF, Hind III-HF, and Not I-HF (NEB); and sequencing by the Garvan Institute (Sydney, Australia) using primers P1, P2, S1, S2, S3, S4, S5. Chemically competent strain DH10Bac E. coli was then transformed with pFASTBac1- BALF2 , which was confirmed by growth in 50 µl/ml kanamycin, 7 µl/ml gentamycin, and 10 µl/ml tetracycline; and blue/white screening using plates containing 50% X-gal. The Bac-to-Bac® system was then used to transfect 3 ml SF9 insect cells at ~ 1.0 x 10 6 cells/ml with recombinant baculovirus. After 5 days of incubation (27°C at 120 rpm) the supernatant (V1) was collected. Transfection of SF9 cells was then repeated, with 1 ml V1 added to 10 ml SF9 cells and harvested as above (V2). Transfection of SF9 cells was again repeated, with 1 ml V2 added to 100 ml SF9 cells and harvested as above (V3). V3 was then used for expression of BALF2 in 600 ml culture, for 72 hours at 27°C and 120 rpm. Expression and Purification of BALF2 600 ml SF9 insect cells were transfected with recombinant baculovirus (V3) and incubated for 72 hours at 27°C and 120 rpm. BALF2 expression culture was pelleted and resuspended in Lysis buffer (20 mM Tris-HCl (pH 7.5), 200 mM NaCl, 5 mM EDTA, 10% v/v glycerol, 3 mM β-mercaptoethanol (BME), 0.01 U/ml Benzonase® (Merck), 1X cOmplete™ Protease Inhibitor (Roche)). Cells were then lysed by sonication. The soluble portion was passed through a 0.22 µm filter and then applied to a 1 ml CaptureSelect™ C-tagXL pre-packed column (Thermo Fisher) using an ӒKTA pure™ system (Cytiva). The column was washed with 10 CV Wash buffer (20 mM Tris-HCl (pH 7.5), 500 mM NaCl, 10% v/v glycerol, 3 mM BME, 0.01 U/ml Benzonase®, 1X cOmplete™ Protease Inhibitor). Sample was eluted off the column with 10 CV Elution buffer (20 mM Tris-HCl (pH 7.5), 200 mM NaCl, 10% v/v glycerol, 3 mM BME, 2M MgCl 2 ). The pure sample was then buffer-exchanged into Storage buffer (20 mM Tris-HCl (pH 7.5), 200 mM NaCl, 10% v/v glycerol, 3 mM BME, 5 mM MgCl 2 ) by application to a HiPrep 26/10 Desalting column (Cytiva). During the purification process, aliquots were taken and analysed by SDS-PAGE. Pure (> 95%) BALF2 was frozen in liquid nitrogen and stored at -80°C. Negative-Staining Electron Microscopy BALF2 (50 µg/ml) in Imaging buffer (20 mM Tris-HCl (pH 7.5), 50 mM NaCl, 10% v/v glycerol) was applied to a Carbon Film 300 Mesh Copper grid (Electron Microscopy Sciences), prepared by glow-discharge for 3 minutes at 0.15 mA using a Denton Evaporator (Denton). After 1 minute, sample was blotted off with filter paper. The grid was then washed with ultrapure water, then stained with 2% w/v uranyl acetate for 30 seconds. The grid allowed to dry for 5 minutes. Imaging took place using a Tecnai T-12 (FEI) microscope equipped with a Gatan Rio™ 4 camera with an accelerating voltage of 120 kV. For the formation of protein-only irregular filaments, BALF2 (500 µg/ml) was incubated in Reducing buffer (20 mM Tris-HCl (pH 7.5), 50 mM NaCl, 10% v/v glycerol, 1 mM Dithiothreitol (DTT)) for 6 hours. For the formation of regular helical filaments, BALF2 (500 µg/ml) was incubated with O2 oligonucleotide in a 10:1 molar ratio in Filamentation buffer (20 mM Tris-HCl (pH 9.0), 50 mM NaCl, 10% v/v glycerol, 6 mM BME, 10 mM MgCl 2 ) for 30 minutes at 37°C before a further incubation for 20 hours at 4°C. Samples were diluted in their respective buffers to 50 µg/ml before being applied to a grid as above. For binding to M13 mp18 ssDNA, BALF2 at subsaturating (20 µg/ml) or saturating (50 µg/ml) concentrations was incubated with 1 ng/µl M13 mp18 ssDNA in Binding buffer (20 mM Tris-HCl (pH 7.5), 50 mM NaCl, 5 mM EDTA, 10% v/v glycerol) at 37°C for 30 minutes. This correlates with stoichiometric binding ratios of BALF2 to ssDNA of 0.8x and 2x, respectively, assuming a site-size of 15 nt. For binding to the short oligonucleotide O2, BALF2 (20 µg/ml) was incubated with 12.5 µg/ml O2 in Binding buffer at 37°C for 30 minutes. For the time-course assay of BALF2-facilitated annealing, the 1 kb bp gene for ampicillin resistance was amplified by PCR from pFASTBac1 using primers P3 and P4. This was then heat-denatured and incubated at 4 ng/µl with 500 µg/ml BALF2 in Annealing Buffer (20 mM Tris-HCl (pH 7.5), 50 mM NaCl, 10 mM MgCl 2 , 10% v/v glycerol) at 37°C. Aliquots were taken throughout the reaction and diluted to 50 µg/ml before being applied to a grid as above. Cryogenic Electron Microscopy BALF2 in Storage buffer was buffer exchanged using an Amicon 500 centrifugal filter (MWCO = 30 kDa) (Cytiva) into cryo-EM buffer (20 mM Tris-HCl (pH 9), 100 mM NaCl, 6 mM BME, 10 mM MgCl 2 ). 0.9 mg/ml Protein was then incubated at 37°C in the presence of 1.83 µM O2 for 30 minutes, before being incubated at 4°C for 20 hours. 3 µL sample was then applied to UltrAuFoil® R 1.2/1.3 gold foil 300 mesh grid (Quantifoil), with a blot force of 0 and 4 second blotting time, using a Thermo Fisher Mark IV Vitrobot. 6248 movies were collected as a series of 50 frames with a total dose of 65 e − /Å 2 at a pixel size of 0.84 Å/pixel at an average defocus of ~ 1 µm by a Titan Krios cryo-electron microscope (Thermo Fisher), operating at 300 kV and equipped with a Gatan K3 detector and BioQuantum LS 967 energy filter (Dataset 1). In addition, a dataset of 5571 movies was also collected on a similarly prepared and imaged R 1.2/1.3 Continuous Carbon 300 mesh grid (Quantifoil) (Dataset 2). Processing and Model Building Processing was performed remotely at the high-performance computing facility MASSIVE (Monash University, Australia) using the processing package CryoSPARC v4.0. Dataset 1 movies were first processed by patch-based motion correction and CTF estimation. 5469 total micrographs were selected for particle processing based on manual filtering by full-frame motion, CTF estimation, relative ice thickness, and defocus. Blob-based particle picking was used to generate templates with a box size of 320 pixels. Successive rounds of 2D classification and further template-based picking were used to generate suitable 2D class averages from which a multiclass ab initio and subsequent multi-class refinement jobs could be generated to further filter particles, resulting in a final particle count of 557,352. Refinement of higher order aberrations and positive Ewald sphere correction was then performed. Finally, a manually generated mask was used to isolate a single asymmetric unit and perform particle signal subtraction and local refinement, during which the enforcement of C2 symmetry was also used to increase the resolution. 3D variability analysis and further local refinement to isolate a monomeric subunit was then performed. Global sharpening and local resolution estimation of the final map was performed in CryoSPARC v4.0 and local sharpening was performed using deepemhancer (Sanchez-Garcia et al., 2021 ). An overview of the processing workflow is show in Supplementary Fig. 5. An AlphaFold2 structural prediction of BALF2 was used as an initial structure and was built into the density map of the monomer using the ChimeraX v1.5 implementation of ISOLDE v1.3 (Pettersen et al., 2021 , Croll, 2018 ). Refinement and validation of the structure was performed in ISOLDE v1.3 and the Windows installation of PHENIX v1.20.1, WinPHENIX (Liebschner et al., 2019 ). Successive rounds of building and refinement were used to iteratively finalise the structures of the asymmetric unit and of a single monomeric unit. In addition, a medium resolution (nominal 2.9 Å) tetrameric map was determined from particles in dataset 1 which the asymmetric unit structure could be rigid fitted into. This tetrameric structure was used as the starting structure for the molecular dynamics simulations. Further, a low resolution (nominal 7.8 Å) map of ~ 1 pitch was determined from particles in datasets 1 and 2, into which the asymmetric unit structure could be rigid fitted unambiguously, to generate an atomic model of the filament. Identification of conserved regions by sequence analysis and DALI structural comparison Amino acid sequences representing the entire database of herpesvirus annealase proteins from Pfam (PF00747) were downloaded and combined into a single FASTA file using Geneious Prime v.2023.2.1 (Biomatters). A multiple sequence alignment (MSA) was then prepared from the combined FASTA file using the EINSI algorithm within MAFFT v.7.52 (Katoh and Standley, 2013 ). An unrooted maximum likelihood phylogenetic analysis of the MSA was then completed in IQ-TREE v.1.6.12, using MFP + MERGE to simultaneously predict the best model for the data and complete the phylogenetic analysis. The best evolutionary model for the MSA was determined to be LG + F + R6 (Four-matrix model, with empirical AA frequencies and FreeRate heterogeneity across sites with six categories) by both corrected Akaike Information Criterion and Bayseian Information Criterion. Support for the resultant phylogenetic tree was estimated using both ultra-fast bootstrapping (-bb 10000) (Minh et al., 2013) and the SH-aLRT test (-alrt 10000) (Guindon et al., 2010 ) each with 10000 iterations. Maximum likelihood analysis was completed on the University of Technology Sydney eResearch High Performance Compute Facility. Three lists of accession numbers representing the taxa in each clade of the resultant tree (representing the α/β/γ subfamilies) were then extracted using the package ggtree v.3.10 (Yu et al., 2017 ) within Rstudio v.2023.12.0.396 ( https://www.posit.com ) using R v.4.3.2 ( https://www.R-project.org ). The sequences matching these accession numbers were then bulk downloaded from UniProt and aligned again using the EINSI algorithm within MAFFT v.7.52 resulting in three MSA representing each of the α/β/γ subfamilies respectively. The BALF2 monomeric subunit was then coloured in ChimeraX according to conservation in both the full MSA and the γ subfamily MSA using AL2CO with an averaging window of 3 residues, entropy-based conservation measure and unweighted frequency distribution method (Pei and Grishin, 2001 , Pettersen et al., 2021 ). The BALF2 monomeric unit containing both NTD and CTD was used to query the PDB using the DALI server with default settings (Holm et al., 2023 ). Results were then inspected for relevance manually. This search was then repeated using just the OB-fold of the BALF2 monomer model. Molecular Dynamics Simulations Molecular dynamics (MD) simulations were performed either on the monomeric or the tetrameric structures, and for the latter, two nucleotides were added to bridge the gap between resolved ssDNA strands to make the ssDNA continuous between asymmetric subunits, to simulate the filament more accurately. All unresolved residues were modelled using Modeller v10 (Sali and Blundell, 1993 ). Six systems of BALF2 were simulated: BALF2 monomer with and without a Zn 2+ bound at the zinc binding site, BALF2 monomer with 10 Mg 2+ bound to proposed magnesium binding sites, BALF2 tetramer alone with no bound DNA, BALF2 tetramer bound to one DNA strand and BALF2 tetramer bound to two strands of DNA. Systems were setup using CHARMM-GUI (Jo et al., 2008 , Lee et al., 2020 ). Protonation states of ionizable residues were predicted with PROPKA 3.0 (Olsson et al., 2011 ). The three cysteines (C453, C456 and C464) that coordinate the Zn 2+ at its binding site were deprotonated. The AMBER protein FF19SB force field (Maier et al., 2015 ) and nucleic acid BSC1 force field (Ivani et al., 2016 ) were applied for the protein and DNA, respectively. The TIP3P model was used for water (Jorgensen et al., 1983). MD simulations were conducted using the GROMACS 2022.3 simulation package (Abraham et al., 2015 ). Simulations were performed using periodic boundary conditions in a cubic box of ~ 125 Å × 125 Å × 125 Å for the monomeric simulations and ~ 210 Å × 210 Å × 210 Å for the tetrameric simulations that extended at least 10 Å from the solute surface. Na + and Cl − counter ions were added to neutralize the system and achieve a salt concentration of 150 mM. Simulations of BALF2 monomer to predict magnesium binding sites were performed in 1 mM of MgCl 2 instead. A constant temperature of 303.15 K was maintained using a Nosé-Hoover temperature coupling thermostats for solute and solvent separately (Nosé, 1984 , Hoover, 1985 ) with a time constant of 1 ps. An isotropic Parrinello − Rahman barostat (Parrinello and Rahman, 1981 ) with a time constant of 5 ps and a compressibility of 4.5 × 10 − 5 bar − 1 was used to maintain a pressure of 1 atm. A time step of 2 fs was used, where bonds involving hydrogen atoms were constrained using the LINCS algorithm (Hess et al., 1997 ). A cutoff distance of 0.9 nm was used for short-range van der Waals interactions. Particle mesh Ewald was used to treat electrostatic interactions with a real space cutoff of 0.9 nm. For all systems, 5000 steps of energy minimization were performed using the steepest descent algorithm followed by 125 ps of equilibration with positional restraints placed on all the protein and DNA heavy atoms (a force constant of 400 kJ/mol/nm 2 on the backbone atoms and 40 kJ/mol/nm 2 on the side chain atoms). This was followed by production runs of 1.5 µs in case of monomers and 200 ns for the tetramer systems. Five independent replicates for each system were simulated. Snapshots were saved every 100 ps. VMD 1.9.4 alpha 57 (Visual Molecular Dynamics) (Humphrey et al., 1996 ) and MDAnalysis (Michaud-Agrawal et al., 2011 ) were used to analyse the trajectories. Electromobility Shift Assays To assess binding to a single oligonucleotide, BALF2 (0-2000 nM) was titrated against 150 nM O2, then incubated for 30 minutes at 37°C in Binding buffer. Samples were then mixed with 10X Orange G Loading Dye (0.01% w/v Orange G, 10 mM Tris-HCl (pH 7.5), 5 mM EDTA) before loading on a 4–20% Mini-PROTEAN® TGX™ Precast Gel (Bio-Rad) and run for 150 minutes at 50 V in 2X TBE. Gels were imaged using an Amersham Typhoon™ (Cytiva). This was also repeated in Annealing buffer. Experiments were performed in triplicates. Densitometric analysis was conducted using ImageQuant (Cytiva). To assess ssDNA annealing, 1 µM BALF2 was incubated with 0.1 µM O1 at 37°C in Annealing buffer. After 30 minutes, 0.1 µM O2 or 0.1 µM O3 was added before further incubation for 5 minutes. Thermally annealed reactions without BALF2 were also included as controls. Samples were then mixed with 10X Orange G Loading Dye before loading on 1% Agarose gel and run for 150 minutes at 50 V in 2X TBE. Gel was imaged using an Amersham Typhoon™ (Cytiva). This was repeated with reactions run in Binding buffer. Results Cloning, Expression, Purification, and Initial Characterisation BALF2 was expressed in SF9 insect cells and lysed by sonication (Extended Data 1a). The soluble fraction of the lysate was filtered and purified by C-tag affinity chromatography to >95% purity (Extended Data 1b, c). BALF2 was flash-frozen in liquid nitrogen and stored in Storage Buffer at -80 °C. Initially, negative-staining electron microscopy (NS-EM) was used to assess the structural homogeneity of BALF2. BALF2 appeared as a field of irregularly shaped particles 10.4 ± 0.2 nm (n = 50) in diameter, consistent with monomers (Extended Data 2a) (Mumtsidu et al., 2008). Irregular protein filaments, appearing to consist of a single line of monomers, could be formed after incubation with reducing agent (Extended Data 2b). These formed best after 6 hours of incubation, after which they began to fall apart and were completely absent by 10 hours. Regular helical assemblies could be formed after incubation with the semi-self-complementary O2 oligonucleotide (Extended Data 2c,d). MgCl 2 was essential for the formation of these filaments, which were noticeably thicker than the thin protein-only irregular complexes described above. These could be very long (> 1 µm) and tended to break, branch, or loop back on themselves to form circular complexes. Under NS-EM conditions, their pitch (41.7 ± 0.4 nm; n = 114), diameter (13.9 ± 0.2 nm; n = 175), and helical striation (22 ± 1.1°; n = 11) could be estimated. The stability of these filaments was optimal in the presence of reducing agent, at a pH 9.0, and at 2.5x stoichiometric binding ratio (assuming a BALF2 site-size of 15 nt). To investigate binding to ssDNA, BALF2 was incubated with a long ssDNA substrate (mp 18 M13) for 30 minutes, and timepoints analysed by NS-EM. BALF2-coated M13 ssDNA appears as thin irregular complexes resembling beaded filaments, not dissimilar to the thin protein-only filaments formed under reducing conditions. At sub-saturating concentrations, BALF2 appears initially in short stretches on DNA, indicative of a cooperative binding mode (Extended Data 3a). These intermediate complexes were often tangled and could still be seen after 15 minutes of incubation. In contrast, at saturating concentrations of BALF2, this reaction is rapid as fully coated M13 DNA were immediately visible (Extended Data 3b). Given their widths (10.6 ± 0.2 nm; n = 31), we hypothesised that these were nucleoprotein complexes composed of a single strand of monomers coating a single strand of DNA. Under NS-EM conditions, the fully coated complexes were held in a slightly extended conformation of 3.6 ± 0.06 Å/nt compared to dsDNA. By measuring the distance between single BALF2 particles on the DNA (5.7 ± 0.1 nm), we were able to estimate an approximate site-size of 16.1 ± 0.2 nt (n = 245). This value is in good agreement with those for ICP8 and ORF6 (Mumtsidu et al., 2008, Tsurumi et al., 1996, Ozgur and Griffith, 2014, Gourves et al., 2000). Similar complexes, resembling short rods, were also observed when binding to a much shorter substrate O3 (Extended Data 3c). Considering the findings above, we hypothesised that the thick regular helical filaments which form upon incubation with the semi-self-complementary O2 may be BALF2 annealing intermediates. To assess the identity of these thick filaments, we ran a time-course assay of BALF2-facilitated annealing of heat-denatured DNA (Extended Data 4). Upon addition of the ssDNA, thick helical filaments resembling those produced in the presence of O2 began to form and generally became longer and more regular upon prolonged incubation. At points within these thicker filaments, they either broke apart, forked, or unravelled, revealing that these were composed of two thin nucleoprotein filaments which wound about each other. We refer to each thin filament as a protofilament. They also commonly looped back on themselves to form a hairpin complex showing a thin region within the loop. This is consistent with the current herpesvirus-annealase facilitate SSA model, which features two protofilaments, each coating ssDNA, winding about one another to facilitate annealing (Weerasooriya et al., 2019). After 24 hours, the complexes were still visible. Interestingly, this meant that the filaments observed in the presence of our 60-mer oligonucleotide would require further interaction between nucleoprotein complexes to reach the lengths observed. In our experiments, we commonly observed both short thick complexes, and longer beaded concatemeric BALF2-ssDNA complexes initially and after longer periods of incubation; we therefore cannot conclude as to the mechanism of filament growth, though it is noteworthy it can occur at all. Cryo-EM of BALF2 ssDNA-annealing Intermediate Since we were able to form stable annealing intermediates with a regular turn through the addition of a O2, we proceeded this sample to cryo-EM (Fig. 1). BALF2 was incubated with O2 for 30 minutes at 37 °C in the presence of BME and Mg 2+ , before further incubation for 20 hours overnight at 4 °C. The sample was then cryo-plunged and imaged using a Titan Krios equipped with a Gatan K3 detector (Fig. 1a). 6,248 movies were collected on UltrAUfoil grids and processed using CryoSPARC v4.0. Due to the inherent flexibility of the filaments as visualised by three-dimensional variability analysis (3D-VA), a more traditional single-particle reconstruction approach was used over helical processing, where one asymmetric unit was treated as a single particle (Supp. Movie 1). Through template-based picking and particle classification, a total of 557,352 asymmetric units were identified for further processing (Fig. 1b). By local refinement, the asymmetric unit was isolated and refined to 2.19 Å, using the gold-standard FSC cutoff 0.143 (Fig. 2c) (PDB ID: 9BYQ). Further movement was observed by 3D-VA so local refinement to isolate a single subunit was also performed to generate a 2.16 Å map (Fig. 1d, Supp. Movie 2) (PDB: 9BYP). Structures were built into these maps, revealing the asymmetric unit to consist of a dimeric BALF2 complex, with each subunit bound to ssDNA (Fig. 2e-g). Using an additional continuous carbon grid dataset, a low-resolution map of approximately one helical pitch was also generated and used to rigid fit the structure of the asymmetric unit to generate a composite model of the filament (PDB ID: 9BYR). An overview of the processing pipeline is included in Extended Figure 5. Unfortunately, due to the inherently flexible nature of the filaments, the helical parameters are not static and could not be used for further processing of the filament. Overall Structure The annealing intermediate of BALF2 is a highly flexible, left-handed bipolar filament consisting of two protofilaments of BALF2 monomers which dimerize at each asymmetric unit, related by C2 symmetry (Fig. 1h,i). Each protofilament contains a single strand of DNA, which runs opposite in each direction of each protofilament, providing the bidirectionality required for the formation of B-DNA. With reference to Fig. 1e, and following the established nomenclature, the monomeric structure consists of a large non-contiguous NTD (residues 9–981) consisting of head and shoulder regions which meet at a narrow neck. The head region is alpha helical and contains 8 α-helices and 1 3 10 -helix. The neck region is almost entirely composed of a β-barrel-like OB-fold which contributes to an electropositive cleft bound to ssDNA. In the asymmetric unit, each BALF2 monomer is positioned so the ssDNA is brought together at the dimeric interface to form dsDNA. The shoulder region sits below the neck and is by far the largest of the three regions, containing 36 helices (13 α-helices, 17 3 10 -helices, 5 mixed helices), and 15 β-strands. The strands are organised into 5 antiparallel β-sheets, the most prominent of which is a 6-stranded curved sheet on the sloped side of the shoulder. In addition, this region also contains a zinc finger with bound zinc atom. The NTD also contains two large unresolved loops which we refer to as loop 1 (residues 511–524) and loop 2 (residues 950–960) positioned on either side of each monomer at the interfaces between asymmetric units. The NTD is connected by a short linker to a small C-terminal domain (CTD) (residues 988–1090) which docks into the NTD of the neighbouring asymmetric unit in a domain-swapped configuration. The dimeric partner in the other protofilament also engages in CTD docking, however in the opposite direction. The CTD is composed of an α-helical bundle comprised of six helices (residues 988–1076), which docks into a cleft between the head and body regions, and a tail (residues 1083–1090) which interacts with a region in the head domain to form an antiparallel β-sheet. Residues at the beginning of the chain (residues 1–8) and end (residues 1091–1128) are unresolved and assumed to be disordered. DNA-Binding Mechanism Two strands of ssDNA consisting of 12 nucleotides could be modelled around the neck of each monomer in a continuous channel (Fig. 2a). We number these positions p1–12 in a 5′ – 3′ direction. The sequences chosen to be modelled were from the region of greatest self-complementarity of O2 since individual bases could not be identified, likely due to particle averaging. The ssDNA is held in an extended conformation of 6.3 ± 0.2 Å/nt. It consists of bases which face inwards to the monomeric unit (p1–5, p9–12; unpaired bases) and outwards to the dimerizing BALF2 monomer (p6–8; paired bases) to facilitate homology-pairing. The unpaired bases are bound primarily through π-π interactions (Y497, F509, W528, Y920 and Y937) or through the formation of hydrogen bonds directly with the bases themselves (S849 and N934), or phosphate backbone (N732) (Fig. 2b). We noted that these binding mechanisms create promiscuous architecture at each nucleotide-binding site, which we hypothesised allows BALF2 to accommodate any base at that position. To demonstrate this, each base was modelled into the same site (p3), supported by density, using molecular dynamics simulations implemented in ISOLDE (Croll, 2018) (Extended Data 6). In this example, the bases are accommodated for by the OH group of S849 which may act as a donor (for guanine and thymine) or acceptor (for cytosine or adenosine), and a water molecule which helps bridge the gap between the pyrimidines and the peptide backbone. Interestingly, we were also able to model in uracil. The ssDNA is bound by a β-barrel-like motif capped on top by a short α-helix and supported by a much longer kinked helix below (Fig. 2c). The architecture of this OB-fold is β-β-β-α-β-α-β-β-β (3β-αβα-3β) (Fig. 2d). Due to the low sequence similarity between annealases, a search using the DALI server for structural homologues of BALF2 was conducted (Extended Data 7a). Unsurprisingly, the best match was for ICP8 (PDB ID: 1URJ) which showed significant secondary structural similarity throughout the entire BALF2 chain. Otherwise, four relevant structures were identified: two undefined ‘phage-related’ proteins from Bacillus cereus (PDB ID: 4JG2) and Enterococcus faecalis (PDB ID: 4KLK), the family-defining gp2.5 annealase from bacteriophage T7 (PDB ID: 1JE5), and the Enc34 SSB (PDB ID: 5ODL), a phage of Enterobacter cancerogenus , bound to ssDNA (Extended Data 7b,c) . In these, β6/7 were combined into a single strand and α1 was not kinked. The active site is located at the dimeric interface, where nucleotides in p6–8 are paired with nucleotides p8–6 (respectively) from the opposite monomer to facilitate microhomology-based annealing (Fig. 2e). This site is stabilised through several interactions (Fig. 2f). First, hydrogen-bonds are maintained by β4 positively charged residues R718, K721, and K723 and the phosphate backbone. Second, a short loop which we term the D NA- S tabilisation L oop (DSL) between β5/6 (residues 920–930) frames the site on each end. Y920 forms hydrogen-bonds with the phosphate backbone and induces π-stacking with the unpaired p5 nucleotide; N924 and Q926 coordinate the paired bases at p6/8; and F930 engages the p6 nucleotide. Third, Q928 of this loop interacts with K670 on the opposite monomer. This network of bonds holds each strand so that the annealing interface is a planar conformation. Though these residues show varied levels of conservation in our MSA, Y920 and F930 universally occur as either phenylalanine or tyrosine, highlighting their importance in coordinating DNA. Oligomerisation The mechanism of filament formation appears to be driven primarily by a domain-swapping interaction between the CTD and the NTD of the neighbouring subunit of the same protofilament (Fig. 3a,b). This is a two-fold mechanism. Firstly, the α-helical bundle region of the CTD is highly negatively charged, which drives docking into the positive cleft between the head and shoulder domains. This interaction is strengthened by several salt bridges (Fig. 3c). Secondly, hydrophobic residues in the CTD tail (F1085, I1086 and V1088) interact with a shallow hydrophobic cleft in the head of the neighbouring subunit, forming an antiparallel β-sheet (Fig. 3d). In addition to this domain swap, Loop 1 (residues 511–524) is in the head domain and appears to interact with a region in the head of the neighbouring monomer. Loop 2 (residues 950–960) is in the neck region and is positioned nearby the 5′-end of the ssDNA, and may be involved in ssDNA-interactions as well as potentially oligomerisation . There are also relatively weaker protein-protein interactions at the dimeric interface. While the most dominant interaction seems to be via DNA-DNA interactions, dimerization also appears to be mediated by hydrogen-bonding through two sites. At the interface of the head domains, a single hydrogen bond is formed between N585 residues. In the other (Fig. 3e), what appears to be the major site of interaction features a kinked α-helix followed by a short loop (residues 648–670) which is positioned just below the active annealing site. Collectively, we term this region the Dimerization Helix (DH). The short loop forms hydrogen bonds with the DSL on the opposite monomer, through D669 and K670 which interact with N927 and Q928, respectively. This would presumably also aid to stabilise the active site. The lower helical region of the DH forms hydrogen bonds with the corresponding residues on the opposite monomer through Q657 and T658. Zinc Binding Site Zinc bound within a zinc finger loop, flanked by helical regions (residues 440–472), was identified on the opposite side of the protein to the active site (Fig. 4a). The position of the loop is stabilised by a conserved arginine (R471) which sits at the top of the second flanking helix and interacts with the backbone immediately preceding the first helical structure. The finger coordinates zinc tetrahedrally through a universally conserved C-C-C motif (C453, C456, and C464) and the participation of a water molecule. The water molecule is bound through interactions with the backbone; and by Y443 and T467 through another water molecule. The latter residues are also universally conserved, except Y443 which in β-herpesviruses is a histidine. The fold this site adopts, through zinc coordination, causes it pack between three different regions of the non-contiguous NTD. Overall, the site is very similar to that of ICP8, in its relative position to the active site (Mapelli et al., 2005). Sequence Analysis A multiple sequence alignment (MSA) of all herpesvirus annealases was first generated using all 652 amino acid sequences in the herpesvirus annealase family as compiled by pFam (PF0047) (Supp. Data 1). Phylogenetic analysis of this alignment resolved three distinct and well supported clades which correlated with the three subfamilies of orthoherpesviridae (full support for both ultrafast bootstrap and SH-aLRT for each clade) (Supp. Data 2). Within our phylogenetic tree human and animal herpesvirus annealases generally resolved together forming taxonomically broad clades which represented each of the orthoherpesviridae viral subfamilies. The only exception to this is ICP8 (from HSV-1 and HSV-2) and U41 (from HHV-6a, HHV-6b, and HHV-7) which each formed a distinct clade. Given the only moderate sequence similarity between subfamilies, indicated by our alignment and cited in literature (Mumtsidu et al., 2008, Mapelli et al., 2005), we also created subalignments for each of the γ, α, and β clades to enable assessment of conservation within and between families (Supp. Data 3, 4, 5). When conservation of these subalignments was mapped onto the BALF2 structure, we found that in both the global and γ alignments conservation was most apparent within the OB-fold and the core of the protein. Notably, in the γ alignment (the subfamily in which BALF2 is placed), we saw significantly greater conservation within these regions, as well as the DH region and within the head of the protein. A reduced alignment of the global MSA containing only seven representative sequences is included in Fig. 5c. Molecular Dynamics Simulations of BALF2 Metal3D (Dürr et al., 2023) predicted the presence of a Zn 2+ in the experimentally identified zinc binding site. To investigate the importance of the Zn 2+ on the structural integrity of the protein, we ran molecular dynamics (MD) simulations of BALF2 monomers in the presence and absence of a Zn 2+ (Fig. 6a,b). Both systems showed stable RMSD with increased flexibility at the zinc binding site region (residues 440-472) in absence of a bound Zn 2+ (Extended Data 8a,b). In the simulations that had a Zn 2+ , it remained bound at its binding site coordinated by the three cysteines (C453, C456, and C464) along the simulations. The two water molecules at the zinc binding site were maintained during the simulations as revealed by the average water density map calculated using the VolMap plugin in VMD (Humphrey et al., 1996) and the radial distribution function of water molecules surrounding the Zn 2+ (Extended Data 8c,d). We also conducted MD simulations of a system composed of two asymmetric units, which we call a tetramer without DNA, with one strand of ssDNA, or with both strands of ssDNA. These simulations of BALF2 tetramers show more stable average RMSD in presence of the two DNA strands (Fig. 6c-e and Extended Data 9a,b) and higher flexibility of loop 2 (residues 950–960) in absence of DNA (Fig. 6f-h and Extended Data 9c). In addition, loop 2 when at the interface of asymmetric units compared to a terminal end was also more constrained, suggesting it has roles both in oligomerisation and in ssDNA interaction. Metal3D (Dürr et al., 2023), BioMetAll (Sánchez-Aparicio et al., 2021) and MIB (Lin et al., 2016) were used to predict the magnesium binding site. Ten common binding sites were identified where Mg 2+ bound to E155, E159, D190, E238, E256, E684, D800, D899, D1032, D1033, and E1035 and were used as a starting structure for MD simulations of BALF2 monomers. All Mg 2+ remained bound during the simulations, and we didn’t observe any unbinding. This is likely due to tight binding between Mg 2+ and the negatively charged coordinating residues and much longer time scale required for unbinding comparing to the simulation time scale. Biochemical Characterisation of BALF2 In addition to the structural characterisation addressed by this work, we also investigated the biochemical properties of BALF2 by electromobility shift assay (EMSA). BALF2 was titrated against O2 in binding buffer before being run on 4–20% native PAGE in 2X TBE in triplicate (Fig. 7a,b). As the concentration of BALF2 was increased, bands representing multiple oligomeric weight species appeared, even though there was still unbound DNA in solution. This is indicative of a cooperative binding mode, agreeing with our negative-staining results. A dissociation constant ( K d ) of 467 ± 13 nM was calculated via densitometry and sigmoidal curve fitting (R 2 = 0.989). When this was repeated in the presence of MgCl 2 , the curves overlaid well and there was no significant difference (p > 0.05) in the dissociation constant (R 2 = 0.995; K d = 473 ± 1 nM). We therefore conclude that MgCl 2 does not contribute to binding to ssDNA, despite being required for filament formation. To analyse the annealing activity of BALF2, complexes were formed using fluorescently labelled oligonucleotides O1 (5′-labelled with Cy5), O2 (3′-labelled with Cy3), and O3 (3′-labelled with Cy3). Cy3 and Cy5 are a FRET (Förster resonance energy transfer) pair, where excitation of Cy3 causes emission at a wavelength capable of exciting Cy5, whose emission can then be measured (Fig. 7c). This only occurs over short (<10 nm) distances, and therefore only occurs when O1 is annealed to the complementary O2, but not when O1 is in the presence of the non-complementary O3. For each reaction we can therefore tell what DNA is bound, and if it is ssDNA or dsDNA. We found that BALF2 forms a stable complex with all three oligonucleotides even after 1 hour of incubation, however the addition of a complementary oligonucleotide after 30 minutes causes BALF2 to dissociate from the DNA, which we conclude is due to an annealing reaction and thus the formation of dsDNA. This reaction was rapid and was almost complete after only 5 minutes of incubation, indicated by the very faint white band when O2 is added to the reaction (lane 7 in the presence of MgCl 2 ). When this was repeated in the presence of EDTA, only very weak signal was visualised representing an annealed product and remaining visible DNA was present as BALF2-bound complexes. When the non-complementary oligonucleotide was added instead, no FRET signal could be visualised in either treatment; interestingly, unbound O1 was seen in the presence of MgCl 2 perhaps due order of addition of oligonucleotides. We therefore concluded that BALF2 can facilitate annealing of short oligonucleotides in a MgCl 2 -dependent manner. Discussion As a herpesvirus annealase, BALF2 represents a member of a long-studied group of proteins, first discovered over 50 years ago. Despite its potential as a drug target and as a biotechnological tool, its structure and mechanism have remained unknown, until now. In this study, we have determined the structure of BALF2 as a ssDNA annealing intermediate to 2.2 Å. This has allowed us to characterise the helical filament BALF2 forms, an OB-fold involved in DNA binding, the active site of ssDNA annealing, and a zinc-finger fold. We also performed MD simulations on monomeric and tetrameric structures and investigated BALF2-DNA interactions by EMSA biochemical assays. Our findings for the first time effectively corroborate and explain many of the observations recorded in the literature (Mumtsidu et al., 2008 , Weerasooriya et al., 2019 , Ozgur and Griffith, 2014 ), and suggest further avenues for investigation. It is well-established that BALF2 shows sequence-independent ssDNA-specific binding ability (Tsurumi et al., 1996 , Tsurumi et al., 1998 ). The ssDNA binding cleft was revealed to be around the neck of the protein, consistent with a number of proposed basic and aromatic residues from literature (Mapelli et al., 2005 , Wang and Hall, 1990 , Gao and Knipe, 1989 , Gao et al., 1988 , Leinbach and Heath, 1989 ). Promiscuous architecture at each nucleotide binding site appears to allow BALF2 to bind any ssDNA sequence, as we demonstrated at nucleotide binding site p3. This binding mechanism, which relies on both the exposed bases and the relaxed conformation inherent to ssDNA, provides convincing evidence for why herpesvirus annealases would show reduced affinity to B-DNA, but can non-specifically bind ssDNA (Ruyechan and Weir, 1984 , Lee and Knipe, 1985 ). Interestingly, a low RNA affinity has also been reported on the same scale as dsDNA-affinity (Ruyechan and Weir, 1984 , Boehmer, 2004 , Lee and Knipe, 1985 ). Interestingly, we were able to successfully model uracil into the site (Extended Data 6). This may suggest that the reduced RNA affinity may instead be due to steric hindrance by the additional hydroxyl group in the RNA ribose ring. F930, whose aromatic property is universally conserved (Fig. 5 c), is engaged in π-stacking interactions with the ribose ring of the nucleotide in p6, which may not be possible in RNA. The BALF2 OB-fold is 3β-αβα-3β (Fig. 2 ), has not, despite the structural homologues we identified, been described before as it appears in BALF2. The key differences between BALF2 and the gp2.5/Enc34 SSB are the β6/7 strands, which are is split in our structure but is part of a continuous strand in these homologues; and the α1 helix, which is longer and kinked in BALF2. Nonetheless, the identification of gp2.5 and Enc34 SSB as structural homologues was unexpected. Gp2.5 is the much better established of the two and is the family-defining annealase of the gp-like family (Hernandez and Richardson, 2019 , Lopes et al., 2010 ). In short, it has many similarities to our current understanding of BALF2. gp2.5 is produced with the immediate-early genes of the T7 replisome (Reuben and Gefter, 1973 , Reuben and Gefter, 1974 ). It preferentially binds ssDNA cooperatively and eliminates secondary structure in a direction-independent manner (Xu et al., 2023 , Reuben and Gefter, 1973 , Reuben and Gefter, 1974 , Zou et al., 2018 ). It is capable of strand-transfer and annealing reactions through a similar mechanism to BALF2: first by the coating of two ssDNA regions sharing homology, before they are brought together and annealed, potentially through gp2.5-gp2.5 interactions (Zou et al., 2018 , Hernandez and Richardson, 2019 , Makhov et al., 2009 , Kim and Richardson, 1993 , Hyland et al., 2003, Rezende et al., 2003, Kong et al., 1997). It binds to ssDNA as a monomer but can oligomerise in solution, as dimers, which has also been reported for BALF2 (Hollis et al., 2001 , Rezende et al., 2002 , Mumtsidu et al., 2008 ). Mutations that impact ssDNA-binding, DNA-annealing, or oligomerisation are lethal to the T7 phage (Hyland et al., 2003, Rezende et al., 2002 , Rezende et al., 2003). Though it consists almost solely of the OB-fold, it also has a long C-terminal tail, which is implicated in protein-protein interactions and oligomerisation, like the herpesvirus annealases (Ghosh et al., 2010 , Kim et al., 1992 , Kim and Richardson, 1994 ). Despite low (~ 15%) sequence identity to gp2.5, Enc34 SSB possesses some, if not all, of these traits (Cernooka et al., 2017 ). The Enc34 SSB structure is bound to ssDNA, which is positioned and oriented very similarly to our structure; this may also indicate a similar annealing or oligomerisation mechanism (Cernooka et al., 2017 ). The major biochemical difference between these proteins and BALF2 is an axillary feature of the C-terminal tail, which is thought to occupy the electrochemically complementary DNA-binding site in Enc34 SSB and gp2.5 in a regulatory role; its deletion increases ssDNA-affinity (Hollis et al., 2001 , Xu et al., 2023 , Marintcheva et al., 2008 , Cernooka et al., 2017 ). The opposite is true for BALF2, ICP8 and ORF6, which do not have an acidic tail, and experience a loss of ssDNA-affinity upon its removal (Mapelli et al., 2000 , Mumtsidu et al., 2008 ). Combined with a previous report that mutations or deletions of conserved residues within the CTD preclude filament formation (Darwish et al., 2016 ), our results suggest this is due to a loss in cooperativity and filament formation and/or stability since the CTD tail cannot dock into the neighbouring monomer, as is described below. Surprisingly, the OB-fold also strongly resembles the well-characterised OB-fold of the single-stranded binding proteins E. coli SSB and human RPA, which lack the α1 helix and are non-recombinogenic (Naufer et al., 2021 , Zou et al., 2006 ). Like Redβ, BALF2 also holds ssDNA in a planar orientation during the annealing process; therefore, we expected BALF2 to share a similar fold (Newing et al., 2022 ). In contrast, the Rad52-like superfamily (Redβ, RecT and Rad52) was not detected in our search, and upon manual inspection does not possess the 3β-αβα-3β fold (Caldwell et al., 2022 , Kinoshita et al., 2023 , Newing et al., 2022 , Lopes et al., 2010 ). Why the OB-fold of BALF2 seems to more closely resemble non-recombinogenic SSB proteins rather than annealases from the Rad52-like family is unclear, however it may reflect observed differences in ssDNA-binding. For example, like SSB proteins and gp2.5, the herpesvirus annealases preferentially bind to ssDNA, whereas the stability of a Redβ nucleoprotein complex is greatest as a double-stranded annealing intermediate (Newing et al., 2022 , Zakharova et al., 2021 , Naufer et al., 2021 , Zou et al., 2006 , Tsurumi et al., 1996 , Tsurumi et al., 1998 , Reuben and Gefter, 1973 , Reuben and Gefter, 1974 ). Overall, based on structural similarity, we suggest that the herpesvirus annealases likely constitute a group within the gp-like superfamily (Lopes et al., 2010 ). Whether this is due to divergent or convergent evolution is not clear. However, a shared lineage between herpesviruses and bacteriophage T7 is not a new avenue of thought (Baker et al., 2005 , Kazlauskas and Venclovas, 2012 ). Moreover, bacteriophage T7 and λ-phage, though both dsDNA phages, differ structurally and in terms of replication strategy (Blasche et al., 2013 ). BALF2 was observed to oligomerise into a bipolar filament composed of two protofilaments, each of which coats a strand of ssDNA. We could not find any previous reports of BALF2 forming a helical assembly, whose architecture closely resembles what has been shown in the literature for KSHV ORF6 and HSV-1 ICP8 (Weerasooriya et al., 2019 , Makhov and Griffith, 2006 , Makhov et al., 2009 , Tolun et al., 2013 , O'Donnell et al., 1987 , Darwish et al., 2016 , Mumtsidu et al., 2008 , Ozgur et al., 2011 , Ozgur and Griffith, 2014 ). The driving mechanism for the formation of these filaments is a domain-swap of the CTD into the NTD of the neighbouring subunit. A key interaction of this is an antiparallel β-sheet further stabilised by hydrophobic interactions formed between the CTD tail and NTD. As we noted above, it has been shown that mutation (or removal) of conserved residues within the tail or hydrophobic cleft eliminate filament formation (Darwish et al., 2016 ). This interaction therefore is likely essential to stabilise the domain-swap and the annealing complex. Further, since this interaction would also likely occur when BALF2 coats ssDNA as a single nucleoprotein filament, this domain-swap mechanism explains why cooperative binding is abolished in mutants where the C-terminal tail has been removed (Mumtsidu et al., 2008 , Mapelli et al., 2000 , Ozgur and Griffith, 2014 ). Domain-swapping, particularly of a C-terminal element, is a common approach utilised in several systems to modulate cooperative binding to DNA. For example, the C-terminal tail residues of both the adenovirus DNA-binding protein or the bacteriophage Pf1 SSB are proposed to interact with the globular NTD of neighbouring subunits when forming a nucleoprotein filament (Chang and Shenk, 1990 , Fox et al., 1999 ). More recently, Träger et al. ( 2025 ) found that the annealing protein P12 from bacteriophage PRD1 forms a helical filament in the same bipolar arrangement as BALF2, with cooperative binding dependent on a CTD helical motif that docks into the NTD; and homology pairing at the dimeric interface. In addition to the domain-swap, the position of Loop 2 and the results of our molecular dynamics simulations suggest that it may also play a role in oligomerisation between neighbouring subunits. From our structures, we can’t conclude as to why we see more stable filaments under reducing conditions. Two cysteine residues (C48 and C212) are within close enough proximity to each other to form an intramolecular disulfide bond, however the density present did not support this, likely due to the sample preparation containing BME which reduces disulfide bonds. In vivo , it is feasible that these residues may aid to stabilise the monomer during periods of inactivity. Overall, this filament structurally resembles the ORF6 filaments observed by Ozgur and colleagues ( 2011 , 2014). The ORF6 filaments presented were reported to form during the purification process; were DNA-free and could incorporate ssDNA; and required a reducing environment but not MgCl 2 . While we also observed longer more stable filaments under reducing conditions, these previous findings generally contradict our results. Yet, there are factors which may have contributed to this disparity: First, intracellular ssDNA may have been present within the ORF6 preparation. Likewise, the nuclease protection activity associated with this class of proteins would explain why the filaments persisted after a nuclease incubation. Second, cellular MgCl 2 already involved in filament formation was not accounted for when magnesium was removed and could have also resisted chelation. Third, the observation of thin and thick ORF6 filaments in the same sample was initially determined to be due to the presence and absence of ssDNA, however our results suggest that this was a mixture of thin ORF6-ssDNA complexes and thick double-helical annealing intermediates. Given that ORF6 is the closest homologue to BALF2, and that ORF6 filaments do have the ability to incorporate ssDNA, it is tempting to conclude that these are the same filament (Ozgur and Griffith, 2014 , Ozgur et al., 2011 ). In contrast, low-resolution NS-EM maps of ICP8 suggest it adopts a tighter conformation, though is still a domain-swapped bipolar arrangement composed of dimeric asymmetric units, which are thought to dimerize to anneal ssDNA at the same interface as we present in our structure, likely reflecting a shared mechanism (Tolun et al., 2013 , Makhov et al., 2009 ). Zinc coordination in proteins may serve either a catalytic or purely structural purpose. We were able to identify density corresponding to a zinc ion coordinated by a CCC zinc finger and the participation of a water molecule (CCCw). We feel confident defining this as a structural zinc site due to several factors: First, structural zinc atoms are often bound tetrahedrally, commonly by cysteine residues, which facilitate a higher charge transfer than other residues and thus form a stronger bond (Lee and Lim, 2008 , Nguyen et al., 1999 ). In contrast, catalytic zinc contacts usually consist of a mixture of cysteine, histidine, glutamate, or aspartate ligands and often take on coordination numbers of five or six; these interactions contribute to a more electronegative zinc ion which more readily acts as a Lewis acid and nucleophile (Lee and Lim, 2008 , Wolfe et al., 2000 ). For example, alcohol dehydrogenase features both a catalytic site, involving a variety of C/H/D/E/H 2 O contacts, and a structural site which is almost always four cysteine residues (Auld and Bergman, 2008 ). Despite the participation of a water molecule, the BALF2 zinc finger is therefore more characteristic of a structural site (Lee and Lim, 2008 ). Second, the ICP8 zinc finger (CCCH) is considered structural: mutation of the cysteine ligands results in a severe reduction in viral progeny in complementation assays, bound zinc resists extensive chelation which successfully removes zinc from alcohol dehydrogenase, and when zinc is removed ICP8 retains transient ssDNA-binding activity suggesting slow unfolding of the protein (Gao et al., 1988 , Gupte et al., 1991, Auld and Bergman, 2008 ). Third, the BALF2 zinc finger is positioned on the opposite side of the protein to the active site and is packed against multiple regions of the protein. Fourth, the most accurate Zn 2+ location predictor to date Metal3D (Dürr et al., 2023 ), predicted the presence of a Zn 2+ in the experimentally identified zinc binding site. As has been noted previously for ICP8, the occupation of the binding loop by Zn 2+ causes it to fold and, given the non-contiguous nature of the structure, allows this loop and associated helices to stabilise the full length of the protein (Mapelli et al., 2005 ). This is reflected in the increased flexibility of the zinc binding site region (residues 440–472) upon removal of Zn 2+ in our MD simulations. Considering that the ICP8 zinc finger is only conserved in α-herpesviruses, it would be unsurprising if ORF6 and members of the β-herpesvirus family also coordinated zinc tetrahedrally through a CCCw site, since the residues involved in water coordination are conserved. Interestingly, to our knowledge, a structural CCCw zinc site has not been described previously. Where waters are involved, the site architecture often consists of histidines and/or acidic residues unlike in BALF2 (Karpusas et al., 1997 , Bouckaert et al., 1996 ). Though magnesium was reported to be crucial for both filament formation and annealing activity, we were not able to identify density corresponding to Mg 2+ anywhere in our structure. A divalent cation binding site consisting of two universally conserved acidic residues (D1032 and D1033) identified by Bryant et al. ( 2012 ) is at the border of the CTD helical region and the NTD neck cleft. In theory, Mg 2+ coordination here may facilitate the domain-swap and filament formation. Interestingly, though not well-resolved, this site was also at the opening of a small pocket containing a network of water density which could also contain magnesium; but we did not build magnesium into this pocket since the density was ambiguous. Alternatively, magnesium could be located at the interface between asymmetric units, which would not be resolved due to our local refinement-based approach to reconstruction. Both D1032 and D1033 were also predicted as potential Mg 2+ binding sites in addition to E155, E159, D190, E238, E256, E684, D800, D899, and E1035. However, during the MD simulations, all Mg 2+ remained bound. At the dimeric interface, we identified two major stabilisation interactions: the hydrogen bonds formed by residues of the Dimerization Helix (DH) region, and the base-pairing hydrogen bonds created by each strand of ssDNA. Previously, mutations to Q706 and F707 in ICP8 (corresponding to BALF2 residues H655 and Y656) have eliminated annealing ability (Weerasooriya et al., 2019 ). The aromaticity of F707/Y656 is universally conserved and stabilises the helix in our structure by packing closely with aromatic hydrophobic residues (F693, F703 in ICP8; Y637 and W651 in BALF2). Mutation at this position may disorder the DH enough to disrupt crucial interactions involved in dimerization, eliminating annealing activity. A similar effect has been demonstrated in gp2.5, where mutation at an equivalent position relative to the OB-fold also removed annealing activity (Rezende et al., 2003). The DH-DH and ssDNA base-pairing interactions occur within proximity to each other. Using 3D-VA, we were able to demonstrate that this creates a hinge-like interaction, resulting in flexing, rotating, and sliding between the two monomers about the central planar ssDNA-ssDNA interaction (Supp. Movie 2). This is likely what gives the filament its flexibility. This movement involves the breaking and reformation of the inter-chain hydrogen bonds of the DH region, suggesting that the driver of dimerization is instead the ssDNA itself, which may only be a weak interaction given it only occurs across three base pairs in p6-8. Microhomology-based annealing has also been observed for DSB repair in human polymerase θ helicase. Though acting as a dimer only, polymerase θ anneals homologous short 3′ overhangs at the dimeric interface, then translocates away from the newly synthesised duplex which allows any gaps to be filled for repair of the DSB (Zerio et al., 2024 ). An analogous mechanism may occur for BALF2: after the ssDNA is coated and annealed, the current annealing model maintains that the herpesvirus annealase then dissociates from the newly synthesised ssDNA, consistent with our EMSA annealing assay (Weerasooriya et al., 2019 ). Considering this, we propose the following annealing mechanism: first, ssDNA is coated by BALF2, cooperatively driven by a domain-swapping interaction between the CTD into the NTD of the neighbouring monomer. Two sufficiently homologous BALF2-coated ssDNA protofilaments are then brought together to form a bipolar filament to facilitate microhomology-based searching at the dimeric interfaces of each asymmetric unit. Once paired correctly, this causes dsDNA to form, which takes on the canonical B-DNA helical twist. In this form, the DNA is more rigid and condensed, which places steric stress on both the annealing site and non-pairing bases. This, coupled with the relatively dynamic DH-DH interactions, causes the monomers to dissociate from the complex, completing the reaction. Contrary to the description of the last step (annealed product being released), we were still able to visualise annealing intermediates in our NS-EM annealing reaction after a prolonged incubation of 24 hours, which is inconsistent with both our suggested model and literature. We interpret this as an incomplete annealing reaction that could be the result of significant homology within the same ssDNA strand, or inadequate reaction conditions to facilitate efficient homology searching on a long template, despite complete reactions being demonstrated previously for ICP8 under similar conditions (Weerasooriya et al., 2019 ). While the findings presented explain several key concepts of annealase-facilitated ssDNA annealing, our structure poses further questions. Why is the oligomerisation in α- and γ-herpesviruses different? What are the detailed dynamics for homology searching, and is this linked at all to the movement observed in 3D-VA? What is the basis for the high degree of structural homology between the gp2.5-like family and BALF2? Outside the scope of this work, lines of inquiry still unaddressed include how exactly BGLF5 interacts with BALF2, and how, if at all, may this system be effectively used in recombineering or as a drug target. Nonetheless, this structure sheds significant light on the mechanisms involved in the essential process of SSA in human herpesviruses. Abbreviations 3D-VA: Three-dimensional variability analysis BALF2ΔC: BALF2 deletion mutant, missing the C-terminal 60 residues BME: β-mercaptoethanol CTD: C-terminal domain DH: Dimerization Helix dsDNA: double-stranded DNA DSB: double-stranded DNA break DSL: DNA-stabilisation loop EBV: Epstein-Barr virus EATR: Exonuclease-Annealase Two-component Recombinase FRET: Förster resonance energy transfer HHV-4: Human herpesvirus 4 HR: Homologous recombination HSV-1: Human herpes simplex virus 1 ICP8ΔC: ICP8 deletion mutant, missing the C-terminal 60 residues KSHV: Kaposi’s sarcoma-associated herpesvirus NS-EM: Negative-staining electron microscopy NTD: N-terminal domain OB: Oligonucleotide-binding ORF6ΔC: ORF6 deletion mutant, missing the C-terminal 60 residues SSA: single-strand annealing SSB: single-stranded DNA-binding protein ssDNA: single-stranded DNA Declarations Acknowledgements Parts of this research were supported by NHMRC (Ideas scheme, APP1184012 [GNT1184012]), ARC (Australian Research Council Centre of Excellence in Quantum Biotechnology project number CE230100021) and QUBIC Aspire Fellowship (awarded to N.E.S). This research was supported by resources provided by the National Computational Infrastructure (NCI) and the Pawsey Supercomputing Research Centre with funding from the Australian Government and the Government of Western Australia. References ABRAHAM, M. J., MURTOLA, T., SCHULZ, R., PÁLL, S., SMITH, J. C., HESS, B. & LINDAHL, E. 2015. GROMACS: High performance molecular simulations through multi-level parallelism from laptops to supercomputers. SoftwareX, 1-2 , 19-25. AMUNDSEN, S. K. & PARRIS, D. S. 1984. Detection of herpes simplex virus intertypic recombinant genomes in infected cell DNA. J Virol Methods, 8 , 19-25. ANGEL, F., DECAUSSIN, G., DAILLIE, J. & OOKA, T. 1987. Comparison of early polypeptides induced in EBV producer and non-producer lymphoid cell lines. Annales de l'Institut Pasteur / Virologie, 138 , 169-181. AULD, D. S. & BERGMAN, T. 2008. Medium- and short-chain dehydrogenase/reductase gene and protein families. Cellular and Molecular Life Sciences, 65 , 3961. BAKER, M. L., JIANG, W., RIXON, F. J. & CHIU, W. 2005. Common ancestry of herpesviruses and tailed DNA bacteriophages. J Virol, 79 , 14967-70. BAKKALCI, D., JIA, Y., WINTER, J. R., LEWIS, J. E., TAYLOR, G. S. & STAGG, H. R. 2020. Risk factors for Epstein Barr virus-associated cancers: a systematic review, critical appraisal, and mapping of the epidemiological evidence. J Glob Health, 10 , 010405. BAUMFORTH, K. R., YOUNG, L. S., FLAVELL, K. J., CONSTANDINOU, C. & MURRAY, P. G. 1999. The Epstein-Barr virus and its association with human cancers. Mol Pathol, 52 , 307-22. BAYLIS, S. A., PURIFOY, D. J. M. & LITTLER, E. 1989. The characterization of the EBV alkaline deoxyribonuclease cloned and expressed in E.coli. Nucleic Acids Research, 17 , 7609-7622. BLASCHE, S., WUCHTY, S., RAJAGOPALA, S. & UETZ, P. 2013. The Protein Interaction Network of Bacteriophage Lambda with Its Host, Escherichia coli. Journal of virology, 87. BOEHMER, P. E. 2004. RNA binding and R-loop formation by the herpes simplex virus type-1 single-stranded DNA-binding protein (ICP8). Nucleic Acids Res, 32 , 4576-84. BOEHMER, P. E. & LEHMAN, I. R. 1993. Herpes simplex virus type 1 ICP8: helix-destabilizing properties. J Virol, 67 , 711-5. BORTNER, C., HERNANDEZ, T. R., LEHMAN, I. R. & GRIFFITH, J. 1993. Herpes simplex virus 1 single-strand DNA-binding protein (ICP8) will promote homologous pairing and strand transfer. J Mol Biol, 231 , 241-50. BOUCKAERT, J., POORTMANS, F., WYNS, L. & LORIS, R. 1996. Sequential structural changes upon zinc and calcium binding to metal-free concanavalin A. J Biol Chem, 271 , 16144-50. BREWSTER, J. L. & TOLUN, G. 2020. Half a century of bacteriophage lambda recombinase: In vitro studies of lambda exonuclease and Red-beta annealase. IUBMB Life, 72 , 1622-1633. BRYANT, K. F., YAN, Z., DREYFUS, D. H. & KNIPE, D. M. 2012. Identification of a divalent metal cation binding site in herpes simplex virus 1 (HSV-1) ICP8 required for HSV replication. J Virol, 86 , 6825-34. BUSH, M., YAGER, D. R., GAO, M., WEISSHART, K., MARCY, A. I., COEN, D. M. & KNIPE, D. M. 1991. Correct intranuclear localization of herpes simplex virus DNA polymerase requires the viral ICP8 DNA-binding protein. J Virol, 65 , 1082-9. CALDERWOOD, M. A., VENKATESAN, K., XING, L., CHASE, M. R., VAZQUEZ, A., HOLTHAUS, A. M., EWENCE, A. E., LI, N., HIROZANE-KISHIKAWA, T., HILL, D. E., VIDAL, M., KIEFF, E. & JOHANNSEN, E. 2007. Epstein-Barr virus and virus human protein interaction maps. Proc Natl Acad Sci U S A, 104 , 7606-11. CALDWELL, B. J., NORRIS, A. S., KARBOWSKI, C. F., WIEGAND, A. M., WYSOCKI, V. H. & BELL, C. E. 2022. Structure of a RecT/Redβ family recombinase in complex with a duplex intermediate of DNA annealing. Nature Communications, 13 , 7855. CERNOOKA, E., RUMNIEKS, J., TARS, K. & KAZAKS, A. 2017. Structural Basis for DNA Recognition of a Single-stranded DNA-binding Protein from Enterobacter Phage Enc34. Scientific Reports, 7 , 15529. CHANG, L. S. & SHENK, T. 1990. The adenovirus DNA-binding protein stimulates the rate of transcription directed by adenovirus and adeno-associated virus promoters. J Virol, 64 , 2103-9. CROLL, T. I. 2018. ISOLDE: a physically realistic environment for model building into low-resolution electron-density maps. Acta Crystallographica Section D: Biological Crystallography, 74 , 519-530. CZARNIAK, F. & HENSEL, M. 2015. Red-Mediated Recombineering of Salmonella enterica Genomes. In: SCHATTEN, H. & EISENSTARK, A. (eds.) Salmonella: Methods and Protocols. New York, NY: Springer New York. DARWISH, A. S., GRADY, L. M., BAI, P. & WELLER, S. K. 2016. ICP8 Filament Formation Is Essential for Replication Compartment Formation during Herpes Simplex Virus Infection. Journal of Virology, 90 , 2561-2570. DAVISON, A. J., EBERLE, R., EHLERS, B., HAYWARD, G. S., MCGEOCH, D. J., MINSON, A. C., PELLETT, P. E., ROIZMAN, B., STUDDERT, M. J. & THIRY, E. 2009. The order Herpesvirales. Arch Virol, 154 , 171-7. DE BRUYN KOPS, A. & KNIPE, D. M. 1988. Formation of DNA replication structures in herpes virus-infected cells requires a viral DNA binding protein. Cell, 55 , 857-68. DECAUSSIN, G., LECLERC, V. & OOKA, T. 1995. The lytic cycle of Epstein-Barr virus in the nonproducer Raji line can be rescued by the expression of a 135-kilodalton protein encoded by the BALF2 open reading frame. J Virol, 69 , 7309-14. DUDAS, K. C. & RUYECHAN, W. T. 1998. Identification of a region of the herpes simplex virus single-stranded DNA-binding protein involved in cooperative binding. J Virol, 72 , 257-65. DÜRR, S. L., LEVY, A. & ROTHLISBERGER, U. 2023. Metal3D: a general deep learning framework for accurate metal ion location prediction in proteins. Nature Communications, 14 , 2713. DUTCH, R. E., BIANCHI, V. & LEHMAN, I. R. 1995. Herpes simplex virus type 1 DNA replication is specifically required for high-frequency homologous recombination between repeated sequences. J Virol, 69 , 3084-9. DUTCH, R. E. & LEHMAN, I. R. 1993. Renaturation of complementary DNA strands by herpes simplex virus type 1 ICP8. Journal of Virology, 67 , 6945-6949. FEEDERLE, R., BANNERT, H., LIPS, H., MÜLLER-LANTZSCH, N. & DELECLUSE, H. J. 2009. The Epstein-Barr virus alkaline exonuclease BGLF5 serves pleiotropic functions in virus replication. J Virol, 83 , 4952-62. FELS, U., GEVAERT, K. & VAN DAMME, P. 2020. Bacterial Genetic Engineering by Means of Recombineering for Reverse Genetics. Frontiers in Microbiology, 11. FITSCHEN, L. J., NEWING, T. P., JOHNSTON, N. P., BELL, C. E. & TOLUN, G. 2024. Half a century after their discovery: Structural insights into exonuclease and annealase proteins catalyzing recombineering. Engineering Microbiology, 4 , 100120. FIXMAN, E. D., HAYWARD, G. S. & HAYWARD, S. D. 1992. trans-acting requirements for replication of Epstein-Barr virus ori-Lyt. J Virol, 66 , 5030-9. FIXMAN, E. D., HAYWARD, G. S. & HAYWARD, S. D. 1995. Replication of Epstein-Barr virus oriLyt: lack of a dedicated virally encoded origin-binding protein and dependence on Zta in cotransfection assays. J Virol, 69 , 2998-3006. FOX, D. G., CARY, P. D. & KNEALE, G. G. 1999. Conformational studies of the C-terminal domain of bacteriophage Pf1 gene 5 protein. Biochimica et Biophysica Acta (BBA) - Protein Structure and Molecular Enzymology, 1435 , 138-146. FU, X., WANG, H. & ZHANG, X. 2002. High-frequency intermolecular homologous recombination during herpes simplex virus-mediated plasmid DNA replication. J Virol, 76 , 5866-74. FUJII, K., YOKOYAMA, N., KIYONO, T., KUZUSHIMA, K., HOMMA, M., NISHIYAMA, Y., FUJITA, M. & TSURUMI, T. 2000. The Epstein-Barr virus pol catalytic subunit physically interacts with the BBLF4-BSLF1-BBLF2/3 complex. J Virol, 74 , 2550-7. GAO, M., BOUCHEY, J., CURTIN, K. & KNIPE, D. M. 1988. Genetic identification of a portion of the herpes simplex virus ICP8 protein required for DNA-binding. Virology, 163 , 319-29. GAO, M. & KNIPE, D. M. 1989. Genetic evidence for multiple nuclear functions of the herpes simplex virus ICP8 DNA-binding protein. J Virol, 63 , 5258-67. GAO, Z., KRITHIVAS, A., FINAN, J. E., SEMMES, O. J., ZHOU, S., WANG, Y. & HAYWARD, S. D. 1998. The Epstein-Barr virus lytic transactivator Zta interacts with the helicase-primase replication proteins. J Virol, 72 , 8559-67. GHOSH, S., HAMDAN, S. M. & RICHARDSON, C. C. 2010. Two modes of interaction of the single-stranded DNA-binding protein of bacteriophage T7 with the DNA polymerase-thioredoxin complex. J Biol Chem, 285 , 18103-12. GOLDSTEIN, J. N. & WELLER, S. K. 1998. In vitro processing of herpes simplex virus type 1 DNA replication intermediates by the viral alkaline nuclease, UL12. J Virol, 72 , 8772-81. GOURVES, A.-S., LE GAC, N. T., VILLANI, G., BOEHMER, P. E. & JOHNSON, N. P. 2000. Equilibrium Binding of Single-stranded DNA with Herpes Simplex Virus Type I-coded Single-stranded DNA-binding Protein, ICP8*. Journal of Biological Chemistry, 275 , 10864-10869. GRADY, L. M., SZCZEPANIAK, R., MURELLI, R. P., MASAOKA, T., LE GRICE, S. F. J., WRIGHT, D. L. & WELLER, S. K. 2017. 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, 91. GUINDON, S., DUFAYARD, J.-F., LEFORT, V., ANISIMOVA, M., HORDIJK, W. & GASCUEL, O. 2010. New Algorithms and Methods to Estimate Maximum-Likelihood Phylogenies: Assessing the Performance of PhyML 3.0. Systematic Biology, 59 , 307-321. GUPTE, S. S., OLSON, J. W. & RUYECHAN, W. T. 1991. The major herpes simplex virus type-1 DNA-binding protein is a zinc metalloprotein. Journal of Biological Chemistry, 266 , 11413-11416. HAMMERSCHMIDT, W. & SUGDEN, B. 1988. Identification and characterization of oriLyt, a lytic origin of DNA replication of Epstein-Barr virus. Cell, 55 , 427-33. HAMMERSCHMIDT, W. & SUGDEN, B. 2013. Replication of Epstein-Barr viral DNA. Cold Spring Harb Perspect Biol, 5 , a013029. HARA, Y., WATANABE, T., YOSHIDA, M., MASUD, H. M. A. A., KATO, H., KONDO, T., SUZUKI, R., KUROSE, S., UDDIN, M. K., ARATA, M., MIYAGI, S., YANAGI, Y., SATO, Y., KIMURA, H. & MURATA, T. 2022. Comprehensive Analyses of Intraviral Epstein-Barr Virus Protein–Protein Interactions Hint Central Role of BLRF2 in the Tegument Network. Journal of Virology, 96 , e00518-22. HERNANDEZ, A. J. & RICHARDSON, C. C. 2019. Gp2.5, the multifunctional bacteriophage T7 single-stranded DNA binding protein. Semin Cell Dev Biol, 86 , 92-101. HERNANDEZ, T. R. & LEHMAN, I. R. 1990. Functional interaction between the herpes simplex-1 DNA polymerase and UL42 protein. J Biol Chem, 265 , 11227-32. HESS, B., BEKKER, H., BERENDSEN, H. J. C. & FRAAIJE, J. G. E. M. 1997. LINCS: A linear constraint solver for molecular simulations. Journal of Computational Chemistry, 18 , 1463-1472. HOLLIS, T., STATTEL, J. M., WALTHER, D. S., RICHARDSON, C. C. & ELLENBERGER, T. 2001. Structure of the gene 2.5 protein, a single-stranded DNA binding protein encoded by bacteriophage T7. Proc Natl Acad Sci U S A, 98 , 9557-62. HOLM, L., LAIHO, A., TÖRÖNEN, P. & SALGADO, M. 2023. DALI shines a light on remote homologs: One hundred discoveries. Protein Science, 32 , e4519. HOOVER, W. G. 1985. Canonical dynamics: Equilibrium phase-space distributions. Physical review A, 31 , 1695. HUMPHREY, W., DALKE, A. & SCHULTEN, K. 1996. VMD: visual molecular dynamics. Journal of molecular graphics, 14 , 33-38. HYLAND, E. M., REZENDE, L. F. & RICHARDSON, C. C. 2003. The DNA Binding Domain of the Gene 2.5 Single-stranded DNA-binding Protein of Bacteriophage T7*. Journal of Biological Chemistry, 278 , 7247-7256. IVANI, I., DANS, P. D., NOY, A., PÉREZ, A., FAUSTINO, I., HOSPITAL, A., WALTHER, J., ANDRIO, P., GOÑI, R., BALACEANU, A., PORTELLA, G., BATTISTINI, F., GELPÍ, J. L., GONZÁLEZ, C., VENDRUSCOLO, M., LAUGHTON, C. A., HARRIS, S. A., CASE, D. A. & OROZCO, M. 2016. Parmbsc1: a refined force field for DNA simulations. Nature Methods, 13 , 55-58. IYER, L. M., KOONIN, E. V. & ARAVIND, L. 2002. Classification and evolutionary history of the single-strand annealing proteins, RecT, Redbeta, ERF and RAD52. BMC Genomics, 3 , 8. JACOB, R. J. & ROIZMAN, B. 1977. Anatomy of herpes simplex virus DNA VIII. Properties of the replicating DNA. J Virol, 23 , 394-411. JO, S., KIM, T., IYER, V. G. & IM, W. 2008. CHARMM-GUI: A web-based graphical user interface for CHARMM. Journal of Computational Chemistry, 29 , 1859-1865. JORGENSEN, W. L., CHANDRASEKHAR, J., MADURA, J. D., IMPEY, R. W. & KLEIN, M. L. 1983. Comparison of simple potential functions for simulating liquid water. The Journal of chemical physics, 79 , 926-935. KARPUSAS, M., NOLTE, M., BENTON, C. B., MEIER, W., LIPSCOMB, W. N. & GOELZ, S. 1997. The crystal structure of human interferon β at 2.2-Å resolution. Proceedings of the National Academy of Sciences, 94 , 11813-11818. KATOH, K. & STANDLEY, D. M. 2013. MAFFT Multiple Sequence Alignment Software Version 7: Improvements in Performance and Usability. Molecular Biology and Evolution, 30 , 772-780. KAZLAUSKAS, D. & VENCLOVAS, C. 2012. Two distinct SSB protein families in nucleo-cytoplasmic large DNA viruses. Bioinformatics, 28 , 3186-90. KIM, Y. T. & RICHARDSON, C. C. 1993. Bacteriophage T7 gene 2.5 protein: an essential protein for DNA replication. Proc Natl Acad Sci U S A, 90 , 10173-7. KIM, Y. T. & RICHARDSON, C. C. 1994. Acidic carboxyl-terminal domain of gene 2.5 protein of bacteriophage T7 is essential for protein-protein interactions. J Biol Chem, 269 , 5270-8. KIM, Y. T., TABOR, S., CHURCHICH, J. E. & RICHARDSON, C. C. 1992. Interactions of gene 2.5 protein and DNA polymerase of bacteriophage T7. J Biol Chem, 267 , 15032-40. KINOSHITA, C., TAKIZAWA, Y., SAOTOME, M., OGINO, S., KURUMIZAKA, H. & KAGAWA, W. 2023. The cryo-EM structure of full-length RAD52 protein contains an undecameric ring. FEBS Open Bio, 13 , 408-418. KONG, D., NOSSAL, N. G. & RICHARDSON, C. C. 1997. Role of the bacteriophage T7 and T4 single-stranded DNA-binding proteins in the formation of joint molecules and DNA helicase-catalyzed polar branch migration. Journal of Biological Chemistry, 272 , 8380-8387. KOWALCZYKOWSKI, S. C., DIXON, D. A., EGGLESTON, A. K., LAUDER, S. D. & REHRAUER, W. M. 1994. Biochemistry of homologous recombination in Escherichia coli. Microbiol Rev, 58 , 401-65. LEE, C. K. & KNIPE, D. M. 1985. An immunoassay for the study of DNA-binding activities of herpes simplex virus protein ICP8. J Virol, 54 , 731-8. LEE, J., HITZENBERGER, M., RIEGER, M., KERN, N. R., ZACHARIAS, M. & IM, W. 2020. CHARMM-GUI supports the Amber force fields. The Journal of Chemical Physics, 153. LEE, Y.-M. & LIM, C. 2008. Physical Basis of Structural and Catalytic Zn-Binding Sites in Proteins. Journal of Molecular Biology, 379 , 545-553. LEINBACH, S. S. & HEATH, L. S. 1989. Characterization of the single-stranded DNA-binding domain of the herpes simplex virus protein ICP8. Biochimica et Biophysica Acta (BBA) - Gene Structure and Expression, 1008 , 281-286. LIEBSCHNER, D., AFONINE, P. V., BAKER, M. L., BUNKOCZI, G., CHEN, V. B., CROLL, T. I., HINTZE, B., HUNG, L.-W., JAIN, S., MCCOY, A. J., MORIARTY, N. W., OEFFNER, R. D., POON, B. K., PRISANT, M. G., READ, R. J., RICHARDSON, J. S., RICHARDSON, D. C., SAMMITO, M. D., SOBOLEV, O. V., STOCKWELL, D. H., TERWILLIGER, T. C., URZHUMTSEV, A. G., VIDEAU, L. L., WILLIAMS, C. J. & ADAMS, P. D. 2019. Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix. Acta Crystallographica Section D, 75 , 861-877. LIN, F. L., SPERLE, K. & STERNBERG, N. 1984. Model for homologous recombination during transfer of DNA into mouse L cells: role for DNA ends in the recombination process. Mol Cell Biol, 4 , 1020-34. LIN, S. F., HSU, T. Y., LIU, M. Y., LIN, L. S., YANG, H. L., CHEN, J. Y. & YANG, C. S. 1995. Characterization of Epstein-Barr virus DNase and its interaction with the major DNA binding protein. Virology, 208 , 712-22. LIN, S. F., LIN, S. W., HSU, T. Y., LIU, M. Y., CHEN, J. Y. & YANG, C. S. 1994. Functional analysis of the amino terminus of Epstein-Barr virus deoxyribonuclease. Virology, 199 , 223-7. LIN, Y.-F., CHENG, C.-W., SHIH, C.-S., HWANG, J.-K., YU, C.-S. & LU, C.-H. 2016. MIB: Metal Ion-Binding Site Prediction and Docking Server. Journal of Chemical Information and Modeling, 56 , 2287-2291. LOPES, A., AMARIR-BOUHRAM, J., FAURE, G., PETIT, M. A. & GUEROIS, R. 2010. Detection of novel recombinases in bacteriophage genomes unveils Rad52, Rad51 and Gp2.5 remote homologs. Nucleic Acids Res, 38 , 3952-62. MAIER, J. A., MARTINEZ, C., KASAVAJHALA, K., WICKSTROM, L., HAUSER, K. E. & SIMMERLING, C. 2015. ff14SB: Improving the Accuracy of Protein Side Chain and Backbone Parameters from ff99SB. Journal of Chemical Theory and Computation, 11 , 3696-3713. MAKHOV, A. M. & GRIFFITH, J. D. 2006. Visualization of the annealing of complementary single-stranded DNA catalyzed by the herpes simplex virus type 1 ICP8 SSB/recombinase. J Mol Biol, 355 , 911-22. MAKHOV, A. M., SEN, A., YU, X., SIMON, M. N., GRIFFITH, J. D. & EGELMAN, E. H. 2009. The bipolar filaments formed by herpes simplex virus type 1 SSB/recombination protein (ICP8) suggest a mechanism for DNA annealing. J Mol Biol, 386 , 273-9. MAPELLI, M., MÜHLEISEN, M., PERSICO, G., VAN DER ZANDT, H. & TUCKER, P. A. 2000. The 60-residue C-terminal region of the single-stranded DNA binding protein of herpes simplex virus type 1 is required for cooperative DNA binding. J Virol, 74 , 8812-22. MAPELLI, M., PANJIKAR, S. & TUCKER, P. A. 2005. The Crystal Structure of the Herpes Simplex Virus 1 ssDNA-binding Protein Suggests the Structural Basis for Flexible, Cooperative Single-stranded DNA Binding*. Journal of Biological Chemistry, 280 , 2990-2997. MARESCA, M., ERLER, A., FU, J., FRIEDRICH, A., ZHANG, Y. & STEWART, A. 2010. Single-stranded heteroduplex intermediates in λ Red homologous recombination. BMC Molecular Biology, 11 , 54. MARINTCHEVA, B., MARINTCHEV, A., WAGNER, G. & RICHARDSON, C. C. 2008. Acidic C-terminal tail of the ssDNA-binding protein of bacteriophage T7 and ssDNA compete for the same binding surface. Proc Natl Acad Sci U S A, 105 , 1855-60. MARTINEZ, R., SARISKY, R. T., WEBER, P. C. & WELLER, S. K. 1996. Herpes simplex virus type 1 alkaline nuclease is required for efficient processing of viral DNA replication intermediates. J Virol, 70 , 2075-85. MICHAUD‐AGRAWAL, N., DENNING, E. J., WOOLF, T. B. & BECKSTEIN, O. 2011. MDAnalysis: a toolkit for the analysis of molecular dynamics simulations. Journal of computational chemistry, 32 , 2319-2327. MINH, B. Q., NGUYEN, M. A. T. & VON HAESELER, A. 2013. Ultrafast Approximation for Phylogenetic Bootstrap. Molecular Biology and Evolution, 30 , 1188-1195. MOCARSKI, E. S. & ROIZMAN, B. 1982. Structure and role of the herpes simplex virus DNA termini in inversion, circularization and generation of virion DNA. Cell, 31 , 89-97. MOSBERG, J. A., LAJOIE, M. J. & CHURCH, G. M. 2010. Lambda Red Recombineering in Escherichia coli Occurs Through a Fully Single-Stranded Intermediate. Genetics, 186 , 791-799. MUMTSIDU, E., MAKHOV, A. M., KONAREV, P. V., SVERGUN, D. I., GRIFFITH, J. D. & TUCKER, P. A. 2008. Structural features of the single-stranded DNA-binding protein of Epstein–Barr virus. Journal of Structural Biology, 161 , 172-187. NAUFER, M. N., MORSE, M., MÖLLER, G. B., MCISAAC, J., ROUZINA, I., BEUNING, P. J. & WILLIAMS, M. C. 2021. Multiprotein E. coli SSB–ssDNA complex shows both stable binding and rapid dissociation due to interprotein interactions. Nucleic Acids Research, 49 , 1532-1549. NEWING, T. P., BREWSTER, J. L., FITSCHEN, L. J., BOUWER, J. C., JOHNSTON, N. P., YU, H. & TOLUN, G. 2022. Redβ177 annealase structure reveals details of oligomerization and λ Red-mediated homologous DNA recombination. Nature Communications, 13 , 5649. NGUYEN, L. T., LE, T. N., DE PROFT, F., CHANDRA, A. K., LANGENAEKER, W., NGUYEN, M. T. & GEERLINGS, P. 1999. Mechanism of [2+1] cycloadditions of hydrogen isocyanide to alkynes: Molecular orbital and density functional theory study. Journal of the American Chemical Society, 121 , 5992-6001. NICHOLAS, J., RUVOLO, V., ZONG, J., CIUFO, D., GUO, H. G., REITZ, M. S. & HAYWARD, G. S. 1997. A single 13-kilobase divergent locus in the Kaposi sarcoma-associated herpesvirus (human herpesvirus 8) genome contains nine open reading frames that are homologous to or related to cellular proteins. J Virol, 71 , 1963-74. NOSÉ, S. 1984. A molecular dynamics method for simulations in the canonical ensemble. Molecular physics, 52 , 255-268. O'DONNELL, M. E., ELIAS, P., FUNNELL, B. E. & LEHMAN, I. R. 1987. Interaction between the DNA polymerase and single-stranded DNA-binding protein (infected cell protein 8) of herpes simplex virus 1. J Biol Chem, 262 , 4260-6. OLSSON, M. H. M., SØNDERGAARD, C. R., ROSTKOWSKI, M. & JENSEN, J. H. 2011. PROPKA3: Consistent Treatment of Internal and Surface Residues in Empirical pKa Predictions. Journal of Chemical Theory and Computation, 7 , 525-537. OZGUR, S., DAMANIA, B. & GRIFFITH, J. 2011. The Kaposi’s sarcoma-associated herpesvirus ORF6 DNA binding protein forms long DNA-free helical protein filaments. Journal of Structural Biology, 174 , 37-43. OZGUR, S. & GRIFFITH, J. 2014. Interaction of Kaposi's sarcoma-associated herpesvirus ORF6 protein with single-stranded DNA. J Virol, 88 , 8687-95. PARRINELLO, M. & RAHMAN, A. 1981. Polymorphic transitions in single crystals: A new molecular dynamics method. Journal of Applied physics, 52 , 7182-7190. PEI, J. & GRISHIN, N. V. 2001. AL2CO: calculation of positional conservation in a protein sequence alignment. Bioinformatics, 17 , 700-12. PETTERSEN, E. F., GODDARD, T. D., HUANG, C. C., MENG, E. C., COUCH, G. S., CROLL, T. I., MORRIS, J. H. & FERRIN, T. E. 2021. UCSF ChimeraX: Structure visualization for researchers, educators, and developers. Protein Sci, 30 , 70-82. QUINN, J. P. & MCGEOCH, D. J. 1985. DNA sequence of the region in the genome of herpes simplex virus type 1 containing the genes for DNA polymerase and the major DNA binding protein. Nucleic Acids Res, 13 , 8143-63. REUBEN, R. C. & GEFTER, M. L. 1973. A DNA-binding protein induced by bacteriophage T7. Proc Natl Acad Sci U S A, 70 , 1846-50. REUBEN, R. C. & GEFTER, M. L. 1974. A Deoxyribonucleic Acid-Binding Protein Induced by Bacteriophage T7: PURIFICATION AND PROPERTIES OF THE PROTEIN. Journal of Biological Chemistry, 249 , 3843-3850. REUVEN, N. B., STAIRE, A. E., MYERS, R. S. & WELLER, S. K. 2003. The herpes simplex virus type 1 alkaline nuclease and single-stranded DNA binding protein mediate strand exchange in vitro. J Virol, 77 , 7425-33. REUVEN, N. B. & WELLER, S. K. 2005. Herpes simplex virus type 1 single-strand DNA binding protein ICP8 enhances the nuclease activity of the UL12 alkaline nuclease by increasing its processivity. J Virol, 79 , 9356-8. REUVEN, N. B., WILLCOX, S., GRIFFITH, J. D. & WELLER, S. K. 2004. 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, 342 , 57-71. REZENDE, L. F., HOLLIS, T., ELLENBERGER, T. & RICHARDSON, C. C. 2002. Essential amino acid residues in the single-stranded DNA-binding protein of bacteriophage T7. Identification of the dimer interface. J Biol Chem, 277 , 50643-53. REZENDE, L. F., WILLCOX, S., GRIFFITH, J. D. & RICHARDSON, C. C. 2003. A single-stranded DNA-binding protein of bacteriophage T7 defective in DNA annealing. J Biol Chem, 278 , 29098-105. RUYECHAN, W. T. & WEIR, A. C. 1984. Interaction with nucleic acids and stimulation of the viral DNA polymerase by the herpes simplex virus type 1 major DNA-binding protein. J Virol, 52 , 727-33. SALI, A. & BLUNDELL, T. L. 1993. Comparative protein modelling by satisfaction of spatial restraints. J Mol Biol, 234 , 779-815. SÁNCHEZ-APARICIO, J.-E., TIESSLER-SALA, L., VELASCO-CARNEROS, L., ROLDÁN-MARTÍN, L., SCIORTINO, G. & MARÉCHAL, J.-D. 2021. BioMetAll: Identifying Metal-Binding Sites in Proteins from Backbone Preorganization. Journal of Chemical Information and Modeling, 61 , 311-323. SANCHEZ-GARCIA, R., GOMEZ-BLANCO, J., CUERVO, A., CARAZO, J. M., SORZANO, C. O. S. & VARGAS, J. 2021. DeepEMhancer: a deep learning solution for cryo-EM volume post-processing. Communications Biology, 4 , 874. SCHUMACHER, A. J., MOHNI, K. N., KAN, Y., HENDRICKSON, E. A., STARK, J. M. & WELLER, S. K. 2012. The HSV-1 Exonuclease, UL12, Stimulates Recombination by a Single Strand Annealing Mechanism. PLOS Pathogens, 8 , e1002862. SEHRAWAT, S., KUMAR, D. & ROUSE, B. T. 2018. Herpesviruses: Harmonious Pathogens but Relevant Cofactors in Other Diseases? Frontiers in Cellular and Infection Microbiology, 8. SERQUIÑA, A. K. & ZIEGELBAUER, J. M. 2017. How herpesviruses pass on their genomes. J Cell Biol, 216 , 2611-2613. SEVERINI, A., SCRABA, D. G. & TYRRELL, D. L. 1996. Branched structures in the intracellular DNA of herpes simplex virus type 1. J Virol, 70 , 3169-75. SHELDRICK, P. & BERTHELOT, N. 1975. Inverted repetitions in the chromosome of herpes simplex virus. Cold Spring Harb Symp Quant Biol, 39 Pt 2 , 667-78. SIGNER, E. R. & WEIL, J. 1968. Recombination in bacteriophage lambda. I. Mutants deficient in general recombination. J Mol Biol, 34 , 261-71. SMITH, S., REUVEN, N., MOHNI, K. N., SCHUMACHER, A. J. & WELLER, S. K. 2014. Structure of the herpes simplex virus 1 genome: manipulation of nicks and gaps can abrogate infectivity and alter the cellular DNA damage response. J Virol, 88 , 10146-56. STOLZENBERG, M. C. & OOKA, T. 1990. Purification and properties of Epstein-Barr virus DNase expressed in Escherichia coli. J Virol, 64 , 96-104. TAYLOR, T. J., MCNAMEE, E. E., DAY, C. & KNIPE, D. M. 2003. Herpes simplex virus replication compartments can form by coalescence of smaller compartments. Virology, 309 , 232-247. TOLUN, G., MAKHOV, A. M., LUDTKE, S. J. & GRIFFITH, J. D. 2013. Details of ssDNA annealing revealed by an HSV-1 ICP8-ssDNA binary complex. Nucleic Acids Research, 41 , 5927-37. TRÄGER, L. K., DEGEN, M., PEREIRA, J., DURAIRAJ, J., TEIXEIRA, RAPHAEL D., HILLER, S. & HUGUENIN-DEZOT, N. 2025. Structural basis for cooperative ssDNA binding by bacteriophage protein filament P12. Nucleic Acids Research, 53. TSURUMI, T. 2001. EBV Replication Enzymes. In: TAKADA, K. (ed.) Epstein-Barr Virus and Human Cancer. Berlin, Heidelberg: Springer Berlin Heidelberg. TSURUMI, T., KISHORE, J., YOKOYAMA, N., FUJITA, M., DAIKOKU, T., YAMADA, H., YAMASHITA, Y. & NISHIYAMA, Y. 1998. Overexpression, purification and helix-destabilizing properties of Epstein-Barr virus ssDNA-binding protein. J Gen Virol, 79 ( Pt 5) , 1257-64. TSURUMI, T., KOBAYASHI, A., TAMAI, K., YAMADA, H., DAIKOKU, T., YAMASHITA, Y. & NISHIYAMA, Y. 1996. Epstein-Barr virus single-stranded DNA-binding protein: purification, characterization, and action on DNA synthesis by the viral DNA polymerase. Virology, 222 , 352-64. UPRICHARD, S. & KNIPE, D. 2003. Conformational Changes in the Herpes Simplex Virus ICP8 DNA-Binding Protein Coincident with Assembly in Viral Replication Structures. Journal of virology, 77 , 7467-76. VALLEDOR, M., MYERS, R. S. & SCHILLER, P. C. 2018. Herpes ICP8 protein stimulates homologous recombination in human cells. PLOS ONE, 13 , e0200955. WAGNER, M. J. & SUMMERS, W. C. 1978. Structure of the joint region and the termini of the DNA of herpes simplex virus type 1. J Virol, 27 , 374-87. WANG, Y. S. & HALL, J. D. 1990. Characterization of a major DNA-binding domain in the herpes simplex virus type 1 DNA-binding protein (ICP8). J Virol, 64 , 2082-9. WEERASOORIYA, S., DISCIPIO, K. A., DARWISH, A. S., BAI, P. & WELLER, S. K. 2019. Herpes simplex virus 1 ICP8 mutant lacking annealing activity is deficient for viral DNA replication. Proceedings of the National Academy of Sciences, 116 , 1033-1042. WELLER, S. K. & SAWITZKE, J. A. 2014. Recombination promoted by DNA viruses: phage λ to herpes simplex virus. Annu Rev Microbiol, 68 , 237-58. WELLER, S. K., SEGHATOLESLAMI, M. R., SHAO, L., ROWSE, D. & CARMICHAEL, E. P. 1990. The herpes simplex virus type 1 alkaline nuclease is not essential for viral DNA synthesis: isolation and characterization of a lacZ insertion mutant. J Gen Virol, 71 ( Pt 12) , 2941-52. WILKIE, N. M. 1973. The synthesis and substructure of herpesvirus DNA: the distribution of alkali-labile single strand interruptions in HSV-1 DNA. J Gen Virol, 21 , 453-67. WILKINSON, D. E. & WELLER, S. K. 2003. The role of DNA recombination in herpes simplex virus DNA replication. IUBMB Life, 55 , 451-8. WOLFE, S. A., AND, L. N. & PABO, C. O. 2000. DNA Recognition by Cys2His2 Zinc Finger Proteins. Annual Review of Biophysics and Biomolecular Structure, 29 , 183-212. WONG, Y., MEEHAN, M. T., BURROWS, S. R., DOOLAN, D. L. & MILES, J. J. 2022. Estimating the global burden of Epstein-Barr virus-related cancers. J Cancer Res Clin Oncol, 148 , 31-46. WU, F. Y., AHN, J. H., ALCENDOR, D. J., JANG, W. J., XIAO, J., HAYWARD, S. D. & HAYWARD, G. S. 2001. Origin-independent assembly of Kaposi's sarcoma-associated herpesvirus DNA replication compartments in transient cotransfection assays and association with the ORF-K8 protein and cellular PML. J Virol, 75 , 1487-506. XU, L., CABANAS-DANÉS, J., HALMA, M. T. J., HELLER, I., STRATMANN, SARAH A., VAN OIJEN, ANTOINE M., LEE, S.-J., PETERMAN, E. J. G. & WUITE, G. J. L. 2023. Regulation of T7 gp2.5 binding dynamics by its C-terminal tail, template conformation and sequence. Nucleic Acids Research . YU, G., SMITH, D. K., ZHU, H., GUAN, Y. & LAM, T. T.-Y. 2017. ggtree: an r package for visualization and annotation of phylogenetic trees with their covariates and other associated data. Methods in Ecology and Evolution, 8 , 28-36. ZAKHAROVA, K., CALDWELL, B. J., TA, S., WHEAT, C. T. & BELL, C. E. 2021. Mutational Analysis of Redβ Single Strand Annealing Protein: Roles of the 14 Lysine Residues in DNA Binding and Recombination In Vivo. Int J Mol Sci, 22. ZENG, Y., MIDDELDORP, J., MADJAR, J. J. & OOKA, T. 1997. A major DNA binding protein encoded by BALF2 open reading frame of Epstein-Barr virus (EBV) forms a complex with other EBV DNA-binding proteins: DNAase, EA-D, and DNA polymerase. Virology, 239 , 285-95. ZERIO, C. J., BAI, Y., SOSA-ALVARADO, B. A., GUZI, T. & LANDER, G. C. 2024. Human polymerase theta helicase positions DNA microhomologies for double-strand break repair. bioRxiv . ZHANG, C. X., DECAUSSIN, G., DAILLIE, J. & OOKA, T. 1988. Altered expression of two Epstein-Barr virus early genes localized in BamHI-A in nonproducer Raji cells. J Virol, 62 , 1862-9. ZHANG, C. X., DECAUSSIN, G., DE TURENNE TESSIER, M., DAILLIE, J. & OOKA, T. 1987. Identification of an Epstein-Barr virus-specific desoxyribonuclease gene using complementary DNA. Nucleic Acids Res, 15 , 2707-17. ZIMMERMANN, J. & HAMMERSCHMIDT, W. 1995. Structure and role of the terminal repeats of Epstein-Barr virus in processing and packaging of virion DNA. J Virol, 69 , 3147-55. ZOU, Y., LIU, Y., WU, X. & SHELL, S. M. 2006. Functions of human replication protein A (RPA): from DNA replication to DNA damage and stress responses. J Cell Physiol, 208 , 267-73. ZOU, Z., WU, S., XIONG, J., LI, H., JIANG, Y. & ZHANG, H. 2018. ssDNA hybridization facilitated by T7 ssDNA binding protein (gp2.5) rapidly initiates from the strand terminus or internally followed by a slow zippering step. Biochimie, 147 , 1-12. Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryMovie1Filament.mp4 Supplementary Movie 1 SupplementaryMovie2Dimer.mp4 Supplementary Movie 2 ExtendedData.docx Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6794668","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":467715521,"identity":"f8a306d2-ba3f-47d6-afb2-ccf47745ee18","order_by":0,"name":"Gökhan Tolun","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA1klEQVRIiWNgGAWjYFAC5gYQmcDGwHwASEvIENTAw8AI08KWANLCQ7wWINMAIkAI2LM3tkkX/GLI42Pv+fzqRo0FDwP74aMb8NrCc7BNemYfQzEbz9lt1jnHgA7jSUu7gVeLRGKbNG8PQ2KbRO424xw2oBYJHjNiteQ8M875R6wWnh9gLcyPc9uI0XLmYLM1b4ME0C/HzJhz+yR42Aj5hb29+eBtnj82efLtzY8/53yrk+NnP3wMrxYwYGyTAFFsEJKgcjD4AyaZPxCnehSMglEwCkYaAABB3T+PAQRqdQAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0001-6166-9451","institution":"University of Wollongong","correspondingAuthor":true,"prefix":"","firstName":"Gökhan","middleName":"","lastName":"Tolun","suffix":""},{"id":467715522,"identity":"ce09cd48-7b76-4423-88a6-945bbffb2f24","order_by":1,"name":"Jordan Nicholls","email":"","orcid":"","institution":"University of Wollongong","correspondingAuthor":false,"prefix":"","firstName":"Jordan","middleName":"","lastName":"Nicholls","suffix":""},{"id":467715523,"identity":"a3f91c22-d2a0-4310-ae8d-384c87110d62","order_by":2,"name":"Jodi Brewster","email":"","orcid":"","institution":"Teva Pharmaceuticals","correspondingAuthor":false,"prefix":"","firstName":"Jodi","middleName":"","lastName":"Brewster","suffix":""},{"id":467715524,"identity":"0facf25b-4d59-4ce6-9398-5c2c59204a48","order_by":3,"name":"Nehad El Salamouni","email":"","orcid":"","institution":"University of Wollongong","correspondingAuthor":false,"prefix":"","firstName":"Nehad","middleName":"El","lastName":"Salamouni","suffix":""},{"id":467715525,"identity":"9e6b1227-acad-46a4-afb9-92325b45da3d","order_by":4,"name":"Nikolas Johnston","email":"","orcid":"https://orcid.org/0000-0002-4208-4666","institution":"University of Wollongong","correspondingAuthor":false,"prefix":"","firstName":"Nikolas","middleName":"","lastName":"Johnston","suffix":""},{"id":467715526,"identity":"14460033-04b2-4418-96b5-daa4d3cb757d","order_by":5,"name":"Haibo Yu","email":"","orcid":"https://orcid.org/0000-0002-1099-2803","institution":"University of Wollongong","correspondingAuthor":false,"prefix":"","firstName":"Haibo","middleName":"","lastName":"Yu","suffix":""}],"badges":[],"createdAt":"2025-06-01 08:45:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6794668/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6794668/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":87293368,"identity":"cfd454f0-d008-4f4c-ae93-efa9416a960f","added_by":"auto","created_at":"2025-07-22 12:00:00","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":185485,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ea)\u003c/strong\u003eRepresentative motion-corrected micrograph, low pass filtered to 1 Å. In total, 6248 movies were collected. Scale Bar = 25 nm. \u003cstrong\u003eb)\u003c/strong\u003e A subset of representative 2D class averages used to generate a 3D volume of the dimeric asymmetric unit. \u003cstrong\u003ec\u003c/strong\u003e,\u003cstrong\u003ed)\u003c/strong\u003e Reconstructions of the BALF2 filament asymmetric unit and monomeric unit, respectively, coloured by local resolution. The overall global resolution as determined by FSC cutoff of 0.143 was 2.19 Å (PDB ID: 9BYQ) and 2.16 Å (PDB ID: 9BYP), respectively. Maps have been sharpened using deepemhancer (Sanchez-Garcia et al., 2021). \u003cstrong\u003ee)\u003c/strong\u003e Solved structure of the BALF2 asymmetric unit, composed of one monomeric unit, which is comprised of the NTD (blue) and CTD (green) from two different chains dimerizing with another monomeric unit (NTD: red; CTD; orange), overlaid with the cryo-EM map. The DNA annealing site is at the dimeric interface. \u003cstrong\u003ef)\u003c/strong\u003e Fit of specific residues and nucleotides. \u003cstrong\u003eg)\u003c/strong\u003eBALF2 monomeric unit, coloured by region (blue: shoulder; green: neck; orange: head; brown: Dimerization Helix; yellow: zinc-binding site) and including the helical region (red) and tail (purple) of the CTD from the neighbouring asymmetric unit. \u003cstrong\u003eh)\u003c/strong\u003e Low-resolution map of BALF2 annealing filament. \u003cstrong\u003ei)\u003c/strong\u003e BALF2 dimeric reconstruction fit into filament map, coloured as per \u003cstrong\u003ee\u003c/strong\u003e. The ssDNA (magenta and cyan) runs almost continuously through the filament in opposite directions.\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6794668/v1/8ec166de28acc4711f01e31b.jpg"},{"id":87292505,"identity":"a68bec3e-a1f0-43bd-9c8f-2dad805101a0","added_by":"auto","created_at":"2025-07-22 11:52:00","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":169003,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ea) \u003c/strong\u003eBinding groove of BALF2 coloured by electrostatic surface potential with ssDNA shown in ribbon conformation (non-pairing bases: purple; exposed bases for annealing at the active site: yellow). Each base position is numbered 1–12 from the 5′-end. \u003cstrong\u003eb) \u003c/strong\u003essDNA binding on either side of the active site is through π-stacking interactions and hydrogen-bonding with bases. \u003cstrong\u003ec) \u003c/strong\u003essDNA binding occurs through an oligonucleotide-binding fold composed of a β-barrel (orange) and supported from below by a kinked α-helix and topped by a short α-helix (blue). \u003cstrong\u003ed) \u003c/strong\u003eDiagram of the OB-fold (3β-αβα-3β), structurally conserved in the proteins gp2.5, Enc334 SSB, BALF2 and ICP8. Dotted lines indicate a discontinuity between structural elements. Purple arrow indicates the path of ssDNA around the fold in a 5′–3′ direction. \u003cstrong\u003ee) \u003c/strong\u003eTransparent representation of the asymmetric unit of the BALF2 annealing filament, showing DNA coloured as per a. Three bases are turned outwards to facilitate annealing with the corresponding bases bound by the opposite BALF2 monomeric unit. \u003cstrong\u003ef) \u003c/strong\u003eThe active site of dsDNA annealing. Both monomeric units (red and blue) stabilise the site through positively charged residues and the DNA Stabilisation Loop (DSL). The latter frames the site from both sides and interacts with the DNA through hydrogen bonding and π- π interactions; it is also involved in an interaction with the opposite monomeric unit.\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6794668/v1/ca55876050c17f23a7a5f0a6.jpg"},{"id":87293370,"identity":"89d7ec02-2043-4408-9b9e-f8b58992c426","added_by":"auto","created_at":"2025-07-22 12:00:00","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":184421,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ea)\u003c/strong\u003e Diagram of the domain swapping interaction driving filament formation of BALF2. Each asymmetric unit is composed of two N-terminal domains oriented in a bipolar conformation which are docked into by a CTD from monomers in the above or below neighbouring asymmetric units. \u003cstrong\u003eb) \u003c/strong\u003eThree asymmetric units of the BALF2 filament, positioned by rigid-fitting the dimeric model into the low-resolution filament map, coloured as per \u003cstrong\u003ea\u003c/strong\u003e. \u003cstrong\u003ec)\u003c/strong\u003e The C-terminal domain docking site in the N-terminal domain, with the C-terminal domain transformed laterally to expose the interface between the two, coloured by surface electrostatic potential. This is an interaction driven by high electro-complementarity between the highly negatively charged C-terminal domain and highly positively charged N-terminal cleft. Each residue shown is involved in forming a salt bridge. \u003cstrong\u003ed)\u003c/strong\u003e N-terminal binding strand shown in cartoon from the C-terminal domain docking into a hydrophobic cleft in the N-terminal domain, which is coloured by lipophilicity. Bold residues are located on the strand, italic residues make up the cleft. \u003cstrong\u003ee)\u003c/strong\u003e The oligomerisation region of the N-terminal domain, which mediates the dimeric interface. Both the upper (middle panel, located directly under the active site) and lower (left panel) dimerization sites are mediated by hydrogen bonding interactions.\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6794668/v1/e30ae730df0653f14d8a46a9.jpg"},{"id":87292509,"identity":"bad8c4af-e095-4ca5-8727-69c1c934af76","added_by":"auto","created_at":"2025-07-22 11:52:00","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":143041,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ea)\u003c/strong\u003e BALF2 N-terminal domain coloured rainbow from N-terminus (blue) to C-terminus (red). The zinc binding site (light green) is located on the opposite side of the model to the DNA Binding Cleft. It is packed against two sequentially distant regions (orange, dark blue). \u003cstrong\u003eb)\u003c/strong\u003eThe zinc binding site is a zinc-finger motif composed of a descending helix (Helix 1, green), loop (yellow) and ascending helix (Helix 2, pink). Helix 2 contains a universally conserved arginine which interacts with preceding residues. \u003cstrong\u003ec) \u003c/strong\u003eZinc bound tetrahedrally in BALF2 by a CCC motif and a water network stabilised by Y443 and T467. Metal coordination bonds: orange; hydrogen bonds: blue. \u0026nbsp;\u003cstrong\u003ed) \u003c/strong\u003eReduced version of our full MSA showing range of sequence variation within the zinc binding site. Sequences in descending order include BALF2 (EBV) and homologues from KSHV, Bovine Gammaherpesvirus 6 (BoGHV6), Rhinolophus Gammaherpesvirus 1 (RGHV-1), HSV-1, Varicella-Zoster Virus (VZV), Human Cytomegalovirus (HCMV). Conservation and logo consensus for each position is included under each position, numbered with reference to BALF2. The full sequence alignment can be viewed in Figure 5c.\u003c/p\u003e","description":"","filename":"4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6794668/v1/bc0e343e7f7720721c8fa970.jpg"},{"id":87294948,"identity":"01c5d121-d291-47b0-a805-3726045bee67","added_by":"auto","created_at":"2025-07-22 12:16:00","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":228066,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ea)\u003c/strong\u003e Maximum likelihood phylogenetic tree of the herpesvirus annealase family of proteins. Proteins are divided into three major clades which correlates with the subfamily of the source viruses. The location of each of the human herpesvirus annealases is indicated in the tree. *BALF2 homologues from HSV-1 and HSV-2 (called ICP8). ^BALF2 homologues from HHV-6a, HHV-6b, and HHV-7 (called U41). A black filled circle is included at major nodes with less than 100% bootstrap or SH-aLRT support. \u003cstrong\u003eb) \u003c/strong\u003eBALF2 subunit coloured by sequence conservation according to our MSA of all herpesviruses or γ-herpesviruses. \u003cstrong\u003ec) \u003c/strong\u003eReduced representation of the full alignment containing all herpesvirus annealases. Sequences in descending order are BALF2 from Epstein-Barr Virus (EBV); homologues from γ-herpesviruses from KSHV, Bovine Gammaherpesvirus 6 (BoGHV6), and Rhinolophus Gammaherpesvirus 1 (RGHV-1); and from human α-herpesviruses Human Simplex Virus (HSV-1) and Varicella-Zoster Virus (VZV), and β-herpesvirus Human Cytomegalovirus (HCMV). MSA is coloured according to conservation where white is not conserved, and dark blue is highly conserved.\u003c/p\u003e","description":"","filename":"5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6794668/v1/6f99a93a82f0890c587464a1.jpg"},{"id":87292512,"identity":"b5c547ca-fdf9-4d18-970c-40922a1a0f7b","added_by":"auto","created_at":"2025-07-22 11:52:00","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":173994,"visible":true,"origin":"","legend":"\u003cp\u003eMolecular dynamics simulations of BALF2.\u003cstrong\u003e \u003c/strong\u003eThe sampled ensemble of BALF2 monomer simulated in \u003cstrong\u003ea\u003c/strong\u003e absence and \u003cstrong\u003eb\u003c/strong\u003e presence of a Zn\u003csup\u003e2\u003c/sup\u003e (blue) bound to C453, C456, and C464 at the zinc binding site. Fifteen snapshots of 100 ns intervals from 1.5 μs simulations are shown. Structures were fitted to the monomer backbone atoms. Residues 440 to 472 which comprise the zinc binding site are circled. The sampled ensemble of BALF2 tetramer simulated \u003cstrong\u003ec \u003c/strong\u003ealone, \u003cstrong\u003ed\u003c/strong\u003e in presence of one strand of DNA and \u003cstrong\u003ee\u003c/strong\u003e two strands of DNA. Zoom-in at loop 2 (residues 950-960) away and at the interface of BALF2 tetramer simulated \u003cstrong\u003ef,i \u003c/strong\u003ealone, \u003cstrong\u003eg,j\u003c/strong\u003e in presence of one strand of DNA from protofilament containing the single strand and \u003cstrong\u003eh,k\u003c/strong\u003e two strands of DNA with Ca atoms of residues 950 and 960 shown as balls. Ten snapshots from 200 ns simulations of 20 ns intervals are shown. Structures were fitted to the backbone atoms of the protein N-terminal domains (residues 8-983). Blue, silver, and red colours correspond to early, mid, and late time intervals, respectively.\u003c/p\u003e","description":"","filename":"6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6794668/v1/a9850b59f6900bb7b9bc5672.jpg"},{"id":87293883,"identity":"1b8a35d5-99e8-47d7-8cbe-b6fb95a1bbcc","added_by":"auto","created_at":"2025-07-22 12:08:00","extension":"jpg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":69800,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ea)\u003c/strong\u003e Native PAGE of BALF2 binding to labelled O2. BALF2 (0-2000 nM) was incubated with 150 nM labelled O2 in the presence of either MgCl\u003csub\u003e2\u003c/sub\u003e or EDTA before loading on 4-20% polyacrylamide gel in 1X TBE in triplicate. The gel shown is a representative gel of the sample containing EDTA. \u003cstrong\u003eb)\u003c/strong\u003e Bands were quantified using ImageQuant™ and fitted to a sigmoidal curve. There was no significant difference in the dissociation constant calculated for both curves (p \u0026gt; 0.05). \u003cstrong\u003ec)\u003c/strong\u003e Agarose gel electrophoresis of BALF2-facilitated annealing of Cy5-labelled O1, Cy3-labelled O2 (complementary to O1), and Cy3-labelled O3 (non-complementary to O1). BALF2 was incubated with O1 in the presence of either MgCl\u003csub\u003e2\u003c/sub\u003e or EDTA before a further incubation with O2 or O3 for 5 minutes. Additionally, thermally annealed DNA was also included as a control (\u003cstrong\u003e*\u003c/strong\u003e). Scans for Cy5 (red), Cy3 (cyan), and Cy3 to Cy5 FRET (blue) were overlaid. Annealed DNA is indicated by purple (red Cy5 + blue FRET). Colocalised but not annealed DNA is indicated by white/grey (red Cy5 + cyan Cy3).\u003c/p\u003e","description":"","filename":"7.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6794668/v1/0995bf8a7945cccf66d78a61.jpg"},{"id":87296117,"identity":"61391d91-9c2e-4cbe-a334-f1e65a77520e","added_by":"auto","created_at":"2025-07-22 12:24:01","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2394967,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6794668/v1/b42792eb-fc77-4f4b-bee2-4f5fcd26ff3c.pdf"},{"id":87292514,"identity":"adf1efe8-6557-446a-acee-e21311d3f6e9","added_by":"auto","created_at":"2025-07-22 11:52:00","extension":"mp4","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":37266023,"visible":true,"origin":"","legend":"Supplementary Movie 1","description":"","filename":"SupplementaryMovie1Filament.mp4","url":"https://assets-eu.researchsquare.com/files/rs-6794668/v1/ddda1773186ff6edf6190ad3.mp4"},{"id":87292516,"identity":"0c006462-52c3-4db4-9ceb-dafb7dbba957","added_by":"auto","created_at":"2025-07-22 11:52:02","extension":"mp4","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":111600765,"visible":true,"origin":"","legend":"Supplementary Movie 2","description":"","filename":"SupplementaryMovie2Dimer.mp4","url":"https://assets-eu.researchsquare.com/files/rs-6794668/v1/7ff26569acab41b932568ee1.mp4"},{"id":87292513,"identity":"b34fe45b-2d6e-4cde-b8da-b0708a5f5a86","added_by":"auto","created_at":"2025-07-22 11:52:00","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":8121294,"visible":true,"origin":"","legend":"","description":"","filename":"ExtendedData.docx","url":"https://assets-eu.researchsquare.com/files/rs-6794668/v1/0198cf183c746ab214bfe619.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"EBV BALF2 DNA annealing intermediate structure reveals the mechanism of annealing during recombination","fulltext":[{"header":"Introduction","content":"\u003cdiv id=\"Sec2\" class=\"Section2\"\u003e\u003ch2\u003eEpstein-Barr Virus and BALF2\u003c/h2\u003e\u003cp\u003eThe \u003cem\u003eOrthoherpesviridae\u003c/em\u003e family is a division of large spherical viruses containing relatively complex linear dsDNA genomes which vary between 125\u0026ndash;230 kb in length (Davison et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). \u003cem\u003eOrthoherpesviridae\u003c/em\u003e is divided into three subfamilies; \u003cem\u003eα-herpesvirinae\u003c/em\u003e, \u003cem\u003eβ-herpesvirinae\u003c/em\u003e, and \u003cem\u003eγ-herpesvirinae\u003c/em\u003e; based on replication strategy, host-range, and evolutionary lineage (Sehrawat et al., \u003cspan citationid=\"CR121\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Of the nine species that infect humans, the γ-herpesvirus human herpesvirus 4 (HHV-4), also known as Epstein-Barr virus (EBV), is among the most common. It is highly contagious and boasts a global lifetime prevalence of ~\u0026thinsp;95% (Bakkalci et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). It is associated primarily with infectious mononucleosis but is also an oncovirus that has been implicated in the incidence of both lymphoma and carcinoma-type cancers. It is estimated\u0026thinsp;~\u0026thinsp;85% of nasopharyngeal carcinoma incidence worldwide is EBV-linked (Wong et al., \u003cspan citationid=\"CR144\" class=\"CitationRef\"\u003e2022\u003c/span\u003e, Baumforth et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e1999\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eEBV infection is characterised by asymptomatic periods of latency, during which viral DNA resides primarily as an episome within B-lymphocytes or epithelial cells and is replicated passively every cell cycle by host cell replication machinery (Serqui\u0026ntilde;a and Ziegelbauer, \u003cspan citationid=\"CR122\" class=\"CitationRef\"\u003e2017\u003c/span\u003e, Hammerschmidt and Sugden, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). These latent periods are broken up by periods of lytic replication, where a virally derived replisome carries out amplification of viral DNA at specified areas of the nucleus known as replication compartments (Hammerschmidt and Sugden, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2013\u003c/span\u003e, Fixman et al., 1992). EBV replication intermediates appear initially as long concatemeric molecules, before being packaged into mature virions (Hammerschmidt and Sugden, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e1988\u003c/span\u003e). The essential proteins of the replisome include an expression factor and origin-binding protein (BZLF1); a viral polymerase (BALF5) and processivity factor (BMRF1); a heterotrimeric primase-helicase complex (BSLF1-BBLF4-BBLF2/3); and a multifunctional annealase/single-stranded DNA-binding protein (SSB), called BALF2 (Tsurumi, \u003cspan citationid=\"CR131\" class=\"CitationRef\"\u003e2001\u003c/span\u003e, Hammerschmidt and Sugden, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). This replisome is highly conserved in all human herpesviruses.\u003c/p\u003e\u003cp\u003eBALF2 is a 123 kDa zinc metalloprotein essential for virion production (Angel et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e1987\u003c/span\u003e, Zhang et al., \u003cspan citationid=\"CR151\" class=\"CitationRef\"\u003e1988\u003c/span\u003e, Decaussin et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e1995\u003c/span\u003e, Mumtsidu et al., \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Though literature on BALF2 is relatively sparse, reference to the homologous ICP8 from α-herpesvirus human simplex virus-1 (HSV-1), which shares 30% identity with BALF2, and its closest homologue ORF6 from γ-herpesvirus Karposi\u0026rsquo;s sarcoma-associated herpesvirus (KSHV), which shares 41% identity, attributes several assumed functions to BALF2 (Quinn and McGeoch, \u003cspan citationid=\"CR108\" class=\"CitationRef\"\u003e1985\u003c/span\u003e, Nicholas et al., \u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e1997\u003c/span\u003e, Wu et al., \u003cspan citationid=\"CR145\" class=\"CitationRef\"\u003e2001\u003c/span\u003e, Ozgur et al., \u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). BALF2 was first identified through both genetic and biochemical homology to ICP8; and antigenic cross-reactivity with ICP8-antibody (Quinn and McGeoch, \u003cspan citationid=\"CR108\" class=\"CitationRef\"\u003e1985\u003c/span\u003e, Zhang et al., \u003cspan citationid=\"CR151\" class=\"CitationRef\"\u003e1988\u003c/span\u003e, Angel et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e1987\u003c/span\u003e). It was defined as essential for viral production when Raji cells, which are non-producing BALF2-deficient carriers of the virus, successfully replicated viral DNA after transfection with the BALF2 gene (Decaussin et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e1995\u003c/span\u003e, Zhang et al., \u003cspan citationid=\"CR151\" class=\"CitationRef\"\u003e1988\u003c/span\u003e, Fixman et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e1995\u003c/span\u003e). Through homology to ICP8, BALF2 is hypothesised to localise at punctate regions called pre-replicative sites where it recruits members of the replisome to facilitate the formation of replication compartments (Darwish et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2016\u003c/span\u003e, Taylor et al., \u003cspan citationid=\"CR128\" class=\"CitationRef\"\u003e2003\u003c/span\u003e, Uprichard and Knipe, \u003cspan citationid=\"CR134\" class=\"CitationRef\"\u003e2003\u003c/span\u003e, de Bruyn Kops and Knipe, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e1988\u003c/span\u003e, Bush et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e1991\u003c/span\u003e). As an SSB, it stabilises the replication fork to streamline polymerase activity and protects ssDNA from nuclease digestion (Hernandez and Lehman, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e1990\u003c/span\u003e, Tsurumi et al., \u003cspan citationid=\"CR133\" class=\"CitationRef\"\u003e1996\u003c/span\u003e, Tsurumi et al., \u003cspan citationid=\"CR132\" class=\"CitationRef\"\u003e1998\u003c/span\u003e, Decaussin et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e1995\u003c/span\u003e, O'Donnell et al., \u003cspan citationid=\"CR101\" class=\"CitationRef\"\u003e1987\u003c/span\u003e). In addition, it also has several interactions with other members of the viral replisome, as well as the virally-encoded alkaline exonuclease BGLF5 (Zeng et al., \u003cspan citationid=\"CR149\" class=\"CitationRef\"\u003e1997\u003c/span\u003e, Lin et al., \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e1995\u003c/span\u003e, Gao et al., \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e1998\u003c/span\u003e, Fujii et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2000\u003c/span\u003e, Calderwood et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2007\u003c/span\u003e, Hara et al., \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). BALF2 binding to ssDNA is preferential, non-specific, cooperative, and direction independent (Tsurumi et al., \u003cspan citationid=\"CR132\" class=\"CitationRef\"\u003e1998\u003c/span\u003e, Tsurumi et al., \u003cspan citationid=\"CR133\" class=\"CitationRef\"\u003e1996\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eA BALF2 deletion mutant missing the C-terminal 60 residues (BALF2ΔC) was observed to bind to ssDNA as monomers forming complexes which appeared as irregular bead-like nucleoprotein filaments (Mumtsidu et al., \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). The number of nucleotides covered by a single herpesvirus annealase monomer is approximately\u0026thinsp;~\u0026thinsp;14\u0026ndash;20 nt, though this has historically varied considerably (Mumtsidu et al., \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2008\u003c/span\u003e, Tsurumi et al., \u003cspan citationid=\"CR133\" class=\"CitationRef\"\u003e1996\u003c/span\u003e, Ozgur and Griffith, \u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e2014\u003c/span\u003e, Gourves et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2000\u003c/span\u003e). Affinity to ssDNA is comparable across ICP8 and ORF6 (\u003cem\u003eK\u003c/em\u003e\u003csub\u003e\u003cem\u003ed\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;~\u0026thinsp;0.1\u0026ndash;0.5 \u0026micro;M) but has not been investigated for BALF2 (Weerasooriya et al., \u003cspan citationid=\"CR138\" class=\"CitationRef\"\u003e2019\u003c/span\u003e, Gourves et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2000\u003c/span\u003e, Darwish et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2016\u003c/span\u003e, Dudas and Ruyechan, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e1998\u003c/span\u003e, Ozgur and Griffith, \u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). For these homologues, binding to dsDNA or RNA is approximately equal and is 5-fold weaker than ssDNA-binding (Boehmer, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2004\u003c/span\u003e, Lee and Knipe, \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e1985\u003c/span\u003e, Ruyechan and Weir, \u003cspan citationid=\"CR116\" class=\"CitationRef\"\u003e1984\u003c/span\u003e). There is also \u003cem\u003ein vitro\u003c/em\u003e evidence that BALF2 can melt short sections of dsDNA, inferring an ability to melt secondary structure at the replication fork (Mumtsidu et al., \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2008\u003c/span\u003e, Tsurumi et al., \u003cspan citationid=\"CR132\" class=\"CitationRef\"\u003e1998\u003c/span\u003e). This is a putative feature of this family of proteins (Boehmer and Lehman, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e1993\u003c/span\u003e, Wang and Hall, \u003cspan citationid=\"CR137\" class=\"CitationRef\"\u003e1990\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eExonuclease-Annealase Two-Component Recombinase (EATR) Systems\u003c/h2\u003e\u003cp\u003edsDNA breaks (DSBs) are considered one of the most detrimental DNA mutations. Many organisms utilise aspects of homologous recombination (HR) for the repair of these breaks. HR relies on the manipulation of homologous sequences to use as substrates in DNA-annealing reactions to repair these DSBs through recombination events, often at the cost of genetic deletions. As a result, HR is also a major contributor to genome maintenance, efficient DNA replication, and is a powerful driver of genetic diversity (Fu et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2002\u003c/span\u003e, Kowalczykowski et al., \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e1994\u003c/span\u003e). Exonuclease-annealase two-component recombinase (EATR) systems describe a highly conserved protein interaction which carries out HR through the single-strand annealing (SSA) pathway. Though ubiquitous, they are particularly common in the genomes of linear dsDNA viruses (Weller and Sawitzke, \u003cspan citationid=\"CR139\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). Canonically, SSA is carried out by a 5\u0026prime;-3\u0026prime; exonuclease, which acts to expose ssDNA, and an annealase protein which facilitates annealing of two homologous single strands. They are differentiated from other HR pathways by an ATP-independent mechanism and a direct interaction between the two components. The model EATR system is the λ-Exonuclease/Redβ system from bacteriophage-λ, which is also referred to as the Red system (Signer and Weil, \u003cspan citationid=\"CR125\" class=\"CitationRef\"\u003e1968\u003c/span\u003e). It is shown that the Red system may facilitate incorporation of Okazaki-like fragments into replication products through homologous pairing, in addition to its putative DSB repair function (Lin et al., \u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e1984\u003c/span\u003e, Mosberg et al., \u003cspan citationid=\"CR94\" class=\"CitationRef\"\u003e2010\u003c/span\u003e, Maresca et al., \u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e2010\u003c/span\u003e, Brewster and Tolun, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2020\u003c/span\u003e, Newing et al., \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). As a result, it has been utilised in the biotechnological process known as recombineering, a highly accurate method of \u003cem\u003ein vivo\u003c/em\u003e genetic manipulation (Mosberg et al., \u003cspan citationid=\"CR94\" class=\"CitationRef\"\u003e2010\u003c/span\u003e, Weller and Sawitzke, \u003cspan citationid=\"CR139\" class=\"CitationRef\"\u003e2014\u003c/span\u003e, Czarniak and Hensel, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2015\u003c/span\u003e, Valledor et al., 2018, Fels et al., 2020, Fitschen et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn addition to its role as an SSB, BALF2 is reported to act as the annealase component in an EATR system, with the BGLF5 viral exonuclease, analogous to the λExo/Redβ complex (Weller and Sawitzke, \u003cspan citationid=\"CR139\" class=\"CitationRef\"\u003e2014\u003c/span\u003e, Valledor et al., 2018). Many lines of evidence support an intimate coupling of herpesvirus DNA replication with homologous recombination (Dutch et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e1995\u003c/span\u003e). Inversion of the long and short regions of the herpes simplex genomes, flanked by homologous repeats, commonly occurs during lytic replication (Mocarski and Roizman, \u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e1982\u003c/span\u003e, Sheldrick and Berthelot, \u003cspan citationid=\"CR124\" class=\"CitationRef\"\u003e1975\u003c/span\u003e, Wagner and Summers, \u003cspan citationid=\"CR136\" class=\"CitationRef\"\u003e1978\u003c/span\u003e). This process is so prevalent that chimeric viral strains can be produced this way (Amundsen and Parris, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e1984\u003c/span\u003e). These terminal repeats in EBV facilitate circularisation of the viral genome, which allows replication to proceed initially via the θ-model (Zimmermann and Hammerschmidt, \u003cspan citationid=\"CR153\" class=\"CitationRef\"\u003e1995\u003c/span\u003e). A switch to rolling-circle replication initiates the formation of long head-to-tail concatemeric replication intermediates; however, these are produced at logarithmic rates and may adopt branched conformations. This is inconsistent with a traditional rolling-circle mechanism but may be explained by HR events (Jacob and Roizman, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e1977\u003c/span\u003e, Severini et al., 1996). Additionally, herpesvirus DNA inherently contains many nicks and gaps (Smith et al., \u003cspan citationid=\"CR126\" class=\"CitationRef\"\u003e2014\u003c/span\u003e, Wilkie, \u003cspan citationid=\"CR141\" class=\"CitationRef\"\u003e1973\u003c/span\u003e, Jacob and Roizman, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e1977\u003c/span\u003e). While these have been implicated in the recruitment of host-cell recombination machinery, they have also been suggested to act as dsDNA-breaking points during replication, which are inherently recombinogenic (Smith et al., \u003cspan citationid=\"CR126\" class=\"CitationRef\"\u003e2014\u003c/span\u003e, Wilkinson and Weller, \u003cspan citationid=\"CR142\" class=\"CitationRef\"\u003e2003\u003c/span\u003e). The further processing of concatemeric products into unit-length viral DNA for packaging into capsids is believed to depend on exonuclease-dependent recombination events within the flanking terminal repeats of each unit (Zimmermann and Hammerschmidt, \u003cspan citationid=\"CR153\" class=\"CitationRef\"\u003e1995\u003c/span\u003e, Feederle et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). BGLF5 is annotated as an alkaline exonuclease (Zhang et al., \u003cspan citationid=\"CR152\" class=\"CitationRef\"\u003e1987\u003c/span\u003e, Baylis et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e1989\u003c/span\u003e). Like λExo, it is a member of the D(D/E)xK family of nucleases, and they share sequence homology. Likewise, BGLF5 also possesses magnesium-dependent 5\u0026prime;-3\u0026prime; exonuclease activity, and additionally can create nicks in dsDNA and breaks in ssDNA through endonuclease activity (Baylis et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e1989\u003c/span\u003e, Stolzenberg and Ooka, \u003cspan citationid=\"CR127\" class=\"CitationRef\"\u003e1990\u003c/span\u003e, Lin et al., \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e1994\u003c/span\u003e). dsDNA digestion proceeds processively, while ssDNA digestion occurs sporadically (Lin et al., \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e1994\u003c/span\u003e). While not explicitly essential, it is required for processing and packaging viral DNA into capsids, and thus greatly enhances viral progeny (Feederle et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). Though not well defined, several studies have provided evidence of a direct BGLF5-BALF2 interaction (Calderwood et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2007\u003c/span\u003e, Hara et al., \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2022\u003c/span\u003e, Lin et al., \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e1995\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn herpesviruses, the best characterised EATR is the HSV-1 UL12/ICP8 system (Weller and Sawitzke, \u003cspan citationid=\"CR139\" class=\"CitationRef\"\u003e2014\u003c/span\u003e, Reuven et al., \u003cspan citationid=\"CR111\" class=\"CitationRef\"\u003e2003\u003c/span\u003e, Valledor et al., 2018). BGLF5/BALF2 is evidenced to perform a very similar role, although it has not been studied as thoroughly (Tsurumi et al., \u003cspan citationid=\"CR133\" class=\"CitationRef\"\u003e1996\u003c/span\u003e, Tsurumi et al., \u003cspan citationid=\"CR132\" class=\"CitationRef\"\u003e1998\u003c/span\u003e, Lin et al., \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e1995\u003c/span\u003e). ICP8 is capable of ATP- and cofactor-independent melting of short regions of DNA, and can subsequently anneal homologous substrates to these regions in the presence of Mg\u003csup\u003e2+\u003c/sup\u003e (Boehmer and Lehman, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e1993\u003c/span\u003e, Dutch and Lehman, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e1993\u003c/span\u003e, Bortner et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e1993\u003c/span\u003e). This annealing reaction is second order and relies first on the binding of ssDNA by annealase, followed by an interaction between coated substrates via a filamentous annealing complex which facilitates annealing (Dutch and Lehman, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e1993\u003c/span\u003e, Makhov and Griffith, \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e2006\u003c/span\u003e, Weerasooriya et al., \u003cspan citationid=\"CR138\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). The interaction involves homology searching between the two strands, before the dsDNA is ejected from the complex. ICP8-facilitated annealing is direction-independent but limited to substrates with overhanging ssDNA, which can be generated by UL12 exonuclease activity; moreover, UL12 experiences an ICP8-specific increase to nuclease efficiency (Reuven et al., \u003cspan citationid=\"CR111\" class=\"CitationRef\"\u003e2003\u003c/span\u003e, Reuven et al., 2004, Reuven and Weller, \u003cspan citationid=\"CR112\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). When both components are present, significantly higher levels of SSA are detected \u003cem\u003ein vivo\u003c/em\u003e, and other methods of recombination are downregulated (Schumacher et al., 2012). Like BGLF5, UL12 is also implicated in the processing of DNA concatemers for processing into viral capsids (Martinez et al., 1996, Goldstein and Weller, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e1998\u003c/span\u003e, Weller et al., \u003cspan citationid=\"CR140\" class=\"CitationRef\"\u003e1990\u003c/span\u003e, Grady et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Recently, Valledor et al. (2018) demonstrated a proof of concept that this system could be used in controlled recombineering practices in eukaryotic cells, which could not be performed using the prokaryotic λ-Exonuclease/Redβ EATR.\u003c/p\u003e\u003cp\u003eThough functional homologues of EATR systems are ubiquitous, resolving the phylogenetic relationships between annealases has been difficult due to low sequence similarity. There are three major families of DNA-annealing proteins: the Rad51-like family, which carry out HR through an ATP-dependent strand-invasion mechanism; and the gp-like and Rad52-like families which contain the annealases capable of ATP-independent SSA through EATR activity. The latter is composed of subfamilies Rad52, RecT/Redβ, Erf, and Sak3 (Iyer et al., 2002, Lopes et al., \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). The gp-like family, defined by T7 bacteriophage gp2.5 protein, and the Rad52-like family, defined by human Rad52, are separated based on virulence of the source organism and specificity of their oligonucleotide-binding fold (OB-fold) (Lopes et al., \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e2010\u003c/span\u003e, Cernooka et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). BALF2 does not formally belong to either annealase family, however, ICP8 has been suggested to share some structural motifs with an SSB from \u003cem\u003eEnterobacter carcinogenesis\u003c/em\u003e phage Enc34, a putative member of the gp-like family, despite a marked size difference between the two proteins (128 kDa vs 26 kDa, respectively) (Cernooka et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2017\u003c/span\u003e, Hernandez and Richardson, \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2019\u003c/span\u003e, Kazlauskas and Venclovas, \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2012\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eBALF2 is reported to behave as a monomer in solution (Tsurumi et al., \u003cspan citationid=\"CR133\" class=\"CitationRef\"\u003e1996\u003c/span\u003e, Tsurumi et al., \u003cspan citationid=\"CR132\" class=\"CitationRef\"\u003e1998\u003c/span\u003e, Mumtsidu et al., \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). It is also reported to form a concentration-dependent dimer (Mumtsidu et al., \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). ICP8 is known to form tight left-handed bipolar filaments in the presence of Mg\u003csup\u003e2+\u003c/sup\u003e and absence of ssDNA (Makhov et al., \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2009\u003c/span\u003e, Makhov and Griffith, \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e2006\u003c/span\u003e, Weerasooriya et al., \u003cspan citationid=\"CR138\" class=\"CitationRef\"\u003e2019\u003c/span\u003e, O'Donnell et al., \u003cspan citationid=\"CR101\" class=\"CitationRef\"\u003e1987\u003c/span\u003e, Darwish et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). During annealing, ICP8 appears as a super-helical complex (Makhov and Griffith, \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). Similarly, the homologue ORF6 is also reported to form DNA-free left-handed bipolar filaments (Ozgur et al., \u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e2011\u003c/span\u003e, Ozgur and Griffith, \u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). In contrast, ORF6 filaments have a looser turn and are physically distinct from those formed by ICP8. They also can form in the absence of Mg\u003csup\u003e2+\u003c/sup\u003e but are dependent on a reducing environment. Both filaments have been suggested as scaffolds to which ssDNA may bind during annealing; each strand coated by each protofilament of the helix, consistent with the current annealing model (Makhov et al., \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2009\u003c/span\u003e, Ozgur and Griffith, \u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e2014\u003c/span\u003e, Tolun et al., \u003cspan citationid=\"CR129\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). An additional role of these filaments is thought to be in the formation of pre-replicative sites, as a scaffold and point of localisation for the replisome (Darwish et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Removal of the C-terminal 60 residues (ΔC) abolishes the ability for these proteins to form filaments, in addition to decreasing cooperative binding (Makhov et al., \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2009\u003c/span\u003e, Ozgur and Griffith, \u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e2014\u003c/span\u003e, Mumtsidu et al., \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2008\u003c/span\u003e, Mapelli et al., \u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e2000\u003c/span\u003e). No such filaments or high-order oligomers have been reported for BALF2, though the full-length protein has not been tested (Mumtsidu et al., \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2008\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eBALF2 Structure\u003c/h3\u003e\n\u003cp\u003eThough very little is known about the structure of BALF2, the crystal structure of ICP8ΔC is available (Mapelli et al., \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). It is composed of a large non-contiguous N-terminal domain (NTD) connected via a linker to a smaller C-terminal domain (CTD). This has been fit into low-resolution negative-staining electron microscopy maps of the ssDNA-free filament, as well as a toroidal annealing intermediate, which is thought to represent an initial turn of the filament (Tolun et al., \u003cspan citationid=\"CR129\" class=\"CitationRef\"\u003e2013\u003c/span\u003e, Makhov et al., \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2009\u003c/span\u003e, Makhov and Griffith, \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). In the latter, annealing is proposed to occur at the interface of two monomers, which is consistent with the spatial arrangement of ICP8 in the protein-only filaments. Additionally, ICP8ΔC is known to coordinate zinc through a CCCH motif in the N-terminal domain (Mapelli et al., \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). BALF2 is also proposed to bind zinc in a 1:1 ratio, though since only the cysteine residues are conserved it was unknown how this occurs (Mumtsidu et al., \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Similarly, the role of magnesium is not fully understood. Although it is required for filament formation and annealing by ICP8, it is not required for filament formation by ORF6 (Makhov et al., \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2009\u003c/span\u003e, Makhov and Griffith, \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e2006\u003c/span\u003e, Weerasooriya et al., \u003cspan citationid=\"CR138\" class=\"CitationRef\"\u003e2019\u003c/span\u003e, O'Donnell et al., \u003cspan citationid=\"CR101\" class=\"CitationRef\"\u003e1987\u003c/span\u003e, Darwish et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2016\u003c/span\u003e, Ozgur et al., \u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e2011\u003c/span\u003e, Ozgur and Griffith, \u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eDespite the clinical relevance of herpesviruses, our understanding of the homologous recombination mechanisms essential for herpesvirus DNA replication is limited. We therefore investigated DNA binding and annealing by BALF2 annealase, using cryogenic electron-microscopy (cryo-EM), biochemical assays, sequence analysis and molecular dynamics simulations. Here, we present the structure of full-length BALF2 captured as a novel filamentous annealing intermediate at a resolution of 2.2 \u0026Aring;. The structure of the monomer consists of a large non-contiguous NTD containing an OB-fold and zinc-finger loop; connected by a short linker to a small CTD which docks into a neighbouring NTD, driving filament formation through domain-swapping. The filament is a bipolar arrangement of dimeric asymmetric units that gives the ssDNA its correct directionality for annealing. Each monomer of the asymmetric unit holds ssDNA in a planar orientation to anneal 3 bases at a time at the dimeric interface. These findings provide key insights into the oligomerisation and cooperative ssDNA-binding and annealing mechanisms possessed by this class of proteins, and into DNA replication and repair by herpesviruses.\u003c/p\u003e"},{"header":"Methods and Materials","content":"\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003eDNA Design\u003c/h2\u003e\u003cp\u003eDNA sequences used in this study are provided in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. pFASTBac1 plasmid containing \u003cem\u003eBALF2\u003c/em\u003e between restriction enzyme sites \u003cem\u003eEco\u003c/em\u003eRI and \u003cem\u003eHind\u003c/em\u003eIII was ordered from Gene Universal (pFASTBac1-\u003cem\u003eBALF2\u003c/em\u003e) (Delaware, USA). Primers for PCR amplification (P1-P4) and for sequencing (S1-S5) were ordered from Integrated DNA Technologies (IDT) (Coralville, Iowa). Phosphorylated oligonucleotides for binding assays (O1-O3) were also ordered from IDT.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eOligonucleotide sequences used for this study. Complementary regions in O2 for formation of a self-dimer are shown in bold. The overhanging region is underlined.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"3\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eName\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSequence (5\u0026prime;\u0026ndash;3\u0026prime;)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eDescription\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eP1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eGCTAGAATTCATGCAGGGTGCACAGACT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eForward primer for amplification of BALF2 from pFASTBac1. Includes \u003cem\u003eEco\u003c/em\u003eRI cut site.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eP2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCTATAAGCTTCTACGCCTCTGGTTCGACCTCGAGTCCGGGGAG\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eReverse primer for amplification of BALF2 from pFASTBac1. Includes HindIII cut site and EPEA affinity tag (C-tag) sequence.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eP3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eATTTCAGGTGGCACTTTTCG\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eForward primer for amplification of ampicillin resistance gene from pFASTBac1.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eP4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCTGACAGTTACCAATGCTTAATCAG\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eForward primer for amplification of ampicillin resistance gene from pFASTBac1.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eO1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eTAGCCGTATGTCATCCGCAAAAATCGAGCTATGCAGGGCGATTCTGCTCTAAGCCACAGT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e60-mer oligonucleotide, 5\u0026prime;-labelled with Cy5 fluorophore. Complementary to O2.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eO2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eACTGTGGCTTAGAGCAG\u003cb\u003eAATCG\u003c/b\u003eCCCTGCATAGCT\u003cb\u003eCGATT\u003c/b\u003eTTTGCGGATGACATA\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eCGGCTA\u003c/span\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e60-mer oligonucleotide, 3\u0026prime;-labelled with Cy3 fluorophore. Complementary to O1.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eO3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eTCGACCACTAGCCATGCCATTGCCTCTTAGACACCCCGATACAGTGATTATGAAAGGTAT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e60-mer oligonucleotide, 3\u0026prime;-labelled with Cy3 fluorophore. Non-complementary to O1.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eS1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCGCTCTACGACAAGGAG\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eInternal BALF2 sequencing primer.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eS2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAAACTACGCTGTGGAGCAC\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eInternal BALF2 sequencing primer.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eS3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCCCAGCTGTTTTACCGC\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eInternal BALF2 sequencing primer.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eS4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCAACGTCATAGATGTGGTGC\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eInternal BALF2 sequencing primer.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eS5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eTGAGAACATCAGGGCTGG\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eInternal BALF2 sequencing primer.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eCloning\u003c/h3\u003e\n\u003cp\u003eThe \u003cem\u003eBALF2\u003c/em\u003e sequence was PCR amplified using pFASTBac1-\u003cem\u003eBALF2\u003c/em\u003e as a template with forward primer (P1) to flank the gene with an \u003cem\u003eEco\u003c/em\u003eRI cut site and reverse primer (P2) to flank the gene with a C-tag affinity tag followed by a \u003cem\u003eHind\u003c/em\u003eIII cut site. Strain AN1459 \u003cem\u003eE. coli\u003c/em\u003e containing pFASTBac1, kindly provided by the Mace Lab, was grown in an overnight culture in the presence of 50 \u0026micro;g/ml ampicillin and DNA isolated by QIAprep\u0026reg; Spin Miniprep Kit (QIAGEN). Isolated pFASTBac1 and PCR-amplified \u003cem\u003eBALF2\u003c/em\u003e were double digested with \u003cem\u003eEco\u003c/em\u003eRI-HF and \u003cem\u003eHind\u003c/em\u003eIII-HF (NEB) overnight and gel-purified using GIAEX\u0026reg;II Gel Extraction Kit (QIAGEN). The recombinant plasmid, pFASTBac1 containing \u003cem\u003eBALF2\u003c/em\u003e flanked by a C-tag sequence (pFASTBac1-\u003cem\u003eBALF2\u003c/em\u003e), was then generated by ligation using an insert:vector ratio of 1:5. The newly constructed recombinant plasmid was confirmed by diagnostic digest using \u003cem\u003eEco\u003c/em\u003eRI-HF, \u003cem\u003eHind\u003c/em\u003eIII-HF, and \u003cem\u003eNot\u003c/em\u003eI-HF (NEB); and sequencing by the Garvan Institute (Sydney, Australia) using primers P1, P2, S1, S2, S3, S4, S5. Chemically competent strain DH10Bac \u003cem\u003eE. coli\u003c/em\u003e was then transformed with pFASTBac1-\u003cem\u003eBALF2\u003c/em\u003e, which was confirmed by growth in 50 \u0026micro;l/ml kanamycin, 7 \u0026micro;l/ml gentamycin, and 10 \u0026micro;l/ml tetracycline; and blue/white screening using plates containing 50% X-gal. The Bac-to-Bac\u0026reg; system was then used to transfect 3 ml SF9 insect cells at ~\u0026thinsp;1.0 x 10\u003csup\u003e6\u003c/sup\u003e cells/ml with recombinant baculovirus. After 5 days of incubation (27\u0026deg;C at 120 rpm) the supernatant (V1) was collected. Transfection of SF9 cells was then repeated, with 1 ml V1 added to 10 ml SF9 cells and harvested as above (V2). Transfection of SF9 cells was again repeated, with 1 ml V2 added to 100 ml SF9 cells and harvested as above (V3). V3 was then used for expression of BALF2 in 600 ml culture, for 72 hours at 27\u0026deg;C and 120 rpm.\u003c/p\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003eExpression and Purification of BALF2\u003c/h2\u003e\u003cp\u003e600 ml SF9 insect cells were transfected with recombinant baculovirus (V3) and incubated for 72 hours at 27\u0026deg;C and 120 rpm. BALF2 expression culture was pelleted and resuspended in Lysis buffer (20 mM Tris-HCl (pH 7.5), 200 mM NaCl, 5 mM EDTA, 10% \u003cem\u003ev/v\u003c/em\u003e glycerol, 3 mM β-mercaptoethanol (BME), 0.01 U/ml Benzonase\u0026reg; (Merck), 1X cOmplete\u0026trade; Protease Inhibitor (Roche)). Cells were then lysed by sonication. The soluble portion was passed through a 0.22 \u0026micro;m filter and then applied to a 1 ml CaptureSelect\u0026trade; C-tagXL pre-packed column (Thermo Fisher) using an ӒKTA pure\u0026trade; system (Cytiva). The column was washed with 10 CV Wash buffer (20 mM Tris-HCl (pH 7.5), 500 mM NaCl, 10% \u003cem\u003ev/v\u003c/em\u003e glycerol, 3 mM BME, 0.01 U/ml Benzonase\u0026reg;, 1X cOmplete\u0026trade; Protease Inhibitor). Sample was eluted off the column with 10 CV Elution buffer (20 mM Tris-HCl (pH 7.5), 200 mM NaCl, 10% \u003cem\u003ev/v\u003c/em\u003e glycerol, 3 mM BME, 2M MgCl\u003csub\u003e2\u003c/sub\u003e). The pure sample was then buffer-exchanged into Storage buffer (20 mM Tris-HCl (pH 7.5), 200 mM NaCl, 10% \u003cem\u003ev/v\u003c/em\u003e glycerol, 3 mM BME, 5 mM MgCl\u003csub\u003e2\u003c/sub\u003e) by application to a HiPrep 26/10 Desalting column (Cytiva). During the purification process, aliquots were taken and analysed by SDS-PAGE. Pure (\u0026gt;\u0026thinsp;95%) BALF2 was frozen in liquid nitrogen and stored at -80\u0026deg;C.\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eNegative-Staining Electron Microscopy\u003c/h3\u003e\n\u003cp\u003eBALF2 (50 \u0026micro;g/ml) in Imaging buffer (20 mM Tris-HCl (pH 7.5), 50 mM NaCl, 10% \u003cem\u003ev/v\u003c/em\u003e glycerol) was applied to a Carbon Film 300 Mesh Copper grid (Electron Microscopy Sciences), prepared by glow-discharge for 3 minutes at 0.15 mA using a Denton Evaporator (Denton). After 1 minute, sample was blotted off with filter paper. The grid was then washed with ultrapure water, then stained with 2% \u003cem\u003ew/v\u003c/em\u003e uranyl acetate for 30 seconds. The grid allowed to dry for 5 minutes. Imaging took place using a Tecnai T-12 (FEI) microscope equipped with a Gatan Rio\u0026trade; 4 camera with an accelerating voltage of 120 kV.\u003c/p\u003e\u003cp\u003eFor the formation of protein-only irregular filaments, BALF2 (500 \u0026micro;g/ml) was incubated in Reducing buffer (20 mM Tris-HCl (pH 7.5), 50 mM NaCl, 10% \u003cem\u003ev/v\u003c/em\u003e glycerol, 1 mM Dithiothreitol (DTT)) for 6 hours. For the formation of regular helical filaments, BALF2 (500 \u0026micro;g/ml) was incubated with O2 oligonucleotide in a 10:1 molar ratio in Filamentation buffer (20 mM Tris-HCl (pH 9.0), 50 mM NaCl, 10% \u003cem\u003ev/v\u003c/em\u003e glycerol, 6 mM BME, 10 mM MgCl\u003csub\u003e2\u003c/sub\u003e) for 30 minutes at 37\u0026deg;C before a further incubation for 20 hours at 4\u0026deg;C. Samples were diluted in their respective buffers to 50 \u0026micro;g/ml before being applied to a grid as above.\u003c/p\u003e\u003cp\u003eFor binding to M13 mp18 ssDNA, BALF2 at subsaturating (20 \u0026micro;g/ml) or saturating (50 \u0026micro;g/ml) concentrations was incubated with 1 ng/\u0026micro;l M13 mp18 ssDNA in Binding buffer (20 mM Tris-HCl (pH 7.5), 50 mM NaCl, 5 mM EDTA, 10% \u003cem\u003ev/v\u003c/em\u003e glycerol) at 37\u0026deg;C for 30 minutes. This correlates with stoichiometric binding ratios of BALF2 to ssDNA of 0.8x and 2x, respectively, assuming a site-size of 15 nt. For binding to the short oligonucleotide O2, BALF2 (20 \u0026micro;g/ml) was incubated with 12.5 \u0026micro;g/ml O2 in Binding buffer at 37\u0026deg;C for 30 minutes. For the time-course assay of BALF2-facilitated annealing, the 1 kb bp gene for ampicillin resistance was amplified by PCR from pFASTBac1 using primers P3 and P4. This was then heat-denatured and incubated at 4 ng/\u0026micro;l with 500 \u0026micro;g/ml BALF2 in Annealing Buffer (20 mM Tris-HCl (pH 7.5), 50 mM NaCl, 10 mM MgCl\u003csub\u003e2\u003c/sub\u003e, 10% \u003cem\u003ev/v\u003c/em\u003e glycerol) at 37\u0026deg;C. Aliquots were taken throughout the reaction and diluted to 50 \u0026micro;g/ml before being applied to a grid as above.\u003c/p\u003e\n\u003ch3\u003eCryogenic Electron Microscopy\u003c/h3\u003e\n\u003cp\u003eBALF2 in Storage buffer was buffer exchanged using an Amicon 500 centrifugal filter (MWCO\u0026thinsp;=\u0026thinsp;30 kDa) (Cytiva) into cryo-EM buffer (20 mM Tris-HCl (pH 9), 100 mM NaCl, 6 mM BME, 10 mM MgCl\u003csub\u003e2\u003c/sub\u003e). 0.9 mg/ml Protein was then incubated at 37\u0026deg;C in the presence of 1.83 \u0026micro;M O2 for 30 minutes, before being incubated at 4\u0026deg;C for 20 hours. 3 \u0026micro;L sample was then applied to UltrAuFoil\u0026reg; R 1.2/1.3 gold foil 300 mesh grid (Quantifoil), with a blot force of 0 and 4 second blotting time, using a Thermo Fisher Mark IV Vitrobot. 6248 movies were collected as a series of 50 frames with a total dose of 65 e\u003csup\u003e\u0026minus;\u003c/sup\u003e/\u0026Aring;\u003csup\u003e2\u003c/sup\u003e at a pixel size of 0.84 \u0026Aring;/pixel at an average defocus of ~\u0026thinsp;1 \u0026micro;m by a Titan Krios cryo-electron microscope (Thermo Fisher), operating at 300 kV and equipped with a Gatan K3 detector and BioQuantum LS 967 energy filter (Dataset 1). In addition, a dataset of 5571 movies was also collected on a similarly prepared and imaged R 1.2/1.3 Continuous Carbon 300 mesh grid (Quantifoil) (Dataset 2).\u003c/p\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003eProcessing and Model Building\u003c/h2\u003e\u003cp\u003eProcessing was performed remotely at the high-performance computing facility MASSIVE (Monash University, Australia) using the processing package CryoSPARC v4.0. Dataset 1 movies were first processed by patch-based motion correction and CTF estimation. 5469 total micrographs were selected for particle processing based on manual filtering by full-frame motion, CTF estimation, relative ice thickness, and defocus. Blob-based particle picking was used to generate templates with a box size of 320 pixels. Successive rounds of 2D classification and further template-based picking were used to generate suitable 2D class averages from which a multiclass \u003cem\u003eab initio\u003c/em\u003e and subsequent multi-class refinement jobs could be generated to further filter particles, resulting in a final particle count of 557,352. Refinement of higher order aberrations and positive Ewald sphere correction was then performed. Finally, a manually generated mask was used to isolate a single asymmetric unit and perform particle signal subtraction and local refinement, during which the enforcement of C2 symmetry was also used to increase the resolution. 3D variability analysis and further local refinement to isolate a monomeric subunit was then performed. Global sharpening and local resolution estimation of the final map was performed in CryoSPARC v4.0 and local sharpening was performed using deepemhancer (Sanchez-Garcia et al., \u003cspan citationid=\"CR119\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). An overview of the processing workflow is show in Supplementary Fig.\u0026nbsp;5.\u003c/p\u003e\u003cp\u003eAn AlphaFold2 structural prediction of BALF2 was used as an initial structure and was built into the density map of the monomer using the ChimeraX v1.5 implementation of ISOLDE v1.3 (Pettersen et al., \u003cspan citationid=\"CR107\" class=\"CitationRef\"\u003e2021\u003c/span\u003e, Croll, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Refinement and validation of the structure was performed in ISOLDE v1.3 and the Windows installation of PHENIX v1.20.1, WinPHENIX (Liebschner et al., \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Successive rounds of building and refinement were used to iteratively finalise the structures of the asymmetric unit and of a single monomeric unit.\u003c/p\u003e\u003cp\u003eIn addition, a medium resolution (nominal 2.9 \u0026Aring;) tetrameric map was determined from particles in dataset 1 which the asymmetric unit structure could be rigid fitted into. This tetrameric structure was used as the starting structure for the molecular dynamics simulations. Further, a low resolution (nominal 7.8 \u0026Aring;) map of ~\u0026thinsp;1 pitch was determined from particles in datasets 1 and 2, into which the asymmetric unit structure could be rigid fitted unambiguously, to generate an atomic model of the filament.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003eIdentification of conserved regions by sequence analysis and DALI structural comparison\u003c/h2\u003e\u003cp\u003eAmino acid sequences representing the entire database of herpesvirus annealase proteins from Pfam (PF00747) were downloaded and combined into a single FASTA file using Geneious Prime v.2023.2.1 (Biomatters). A multiple sequence alignment (MSA) was then prepared from the combined FASTA file using the EINSI algorithm within MAFFT v.7.52 (Katoh and Standley, \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). An unrooted maximum likelihood phylogenetic analysis of the MSA was then completed in IQ-TREE v.1.6.12, using MFP\u0026thinsp;+\u0026thinsp;MERGE to simultaneously predict the best model for the data and complete the phylogenetic analysis. The best evolutionary model for the MSA was determined to be LG\u0026thinsp;+\u0026thinsp;F\u0026thinsp;+\u0026thinsp;R6 (Four-matrix model, with empirical AA frequencies and FreeRate heterogeneity across sites with six categories) by both corrected Akaike Information Criterion and Bayseian Information Criterion. Support for the resultant phylogenetic tree was estimated using both ultra-fast bootstrapping (-bb 10000) (Minh et al., 2013) and the SH-aLRT test (-alrt 10000) (Guindon et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2010\u003c/span\u003e) each with 10000 iterations. Maximum likelihood analysis was completed on the University of Technology Sydney eResearch High Performance Compute Facility. Three lists of accession numbers representing the taxa in each clade of the resultant tree (representing the α/β/γ subfamilies) were then extracted using the package ggtree v.3.10 (Yu et al., \u003cspan citationid=\"CR147\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) within Rstudio v.2023.12.0.396 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.posit.com\u003c/span\u003e\u003cspan address=\"https://www.posit.com\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) using R v.4.3.2 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.R-project.org\u003c/span\u003e\u003cspan address=\"https://www.R-project.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The sequences matching these accession numbers were then bulk downloaded from UniProt and aligned again using the EINSI algorithm within MAFFT v.7.52 resulting in three MSA representing each of the α/β/γ subfamilies respectively. The BALF2 monomeric subunit was then coloured in ChimeraX according to conservation in both the full MSA and the γ subfamily MSA using AL2CO with an averaging window of 3 residues, entropy-based conservation measure and unweighted frequency distribution method (Pei and Grishin, \u003cspan citationid=\"CR106\" class=\"CitationRef\"\u003e2001\u003c/span\u003e, Pettersen et al., \u003cspan citationid=\"CR107\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe BALF2 monomeric unit containing both NTD and CTD was used to query the PDB using the DALI server with default settings (Holm et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Results were then inspected for relevance manually. This search was then repeated using just the OB-fold of the BALF2 monomer model.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003eMolecular Dynamics Simulations\u003c/h2\u003e\u003cp\u003eMolecular dynamics (MD) simulations were performed either on the monomeric or the tetrameric structures, and for the latter, two nucleotides were added to bridge the gap between resolved ssDNA strands to make the ssDNA continuous between asymmetric subunits, to simulate the filament more accurately. All unresolved residues were modelled using Modeller v10 (Sali and Blundell, \u003cspan citationid=\"CR117\" class=\"CitationRef\"\u003e1993\u003c/span\u003e). Six systems of BALF2 were simulated: BALF2 monomer with and without a Zn\u003csup\u003e2+\u003c/sup\u003e bound at the zinc binding site, BALF2 monomer with 10 Mg\u003csup\u003e2+\u003c/sup\u003e bound to proposed magnesium binding sites, BALF2 tetramer alone with no bound DNA, BALF2 tetramer bound to one DNA strand and BALF2 tetramer bound to two strands of DNA.\u003c/p\u003e\u003cp\u003eSystems were setup using CHARMM-GUI (Jo et al., \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2008\u003c/span\u003e, Lee et al., \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Protonation states of ionizable residues were predicted with PROPKA 3.0 (Olsson et al., \u003cspan citationid=\"CR102\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). The three cysteines (C453, C456 and C464) that coordinate the Zn\u003csup\u003e2+\u003c/sup\u003e at its binding site were deprotonated. The AMBER protein FF19SB force field (Maier et al., \u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) and nucleic acid BSC1 force field (Ivani et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) were applied for the protein and DNA, respectively. The TIP3P model was used for water (Jorgensen et al., 1983). MD simulations were conducted using the GROMACS 2022.3 simulation package (Abraham et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Simulations were performed using periodic boundary conditions in a cubic box of ~\u0026thinsp;125 \u0026Aring; \u0026times; 125 \u0026Aring; \u0026times; 125 \u0026Aring; for the monomeric simulations and ~\u0026thinsp;210 \u0026Aring; \u0026times; 210 \u0026Aring; \u0026times; 210 \u0026Aring; for the tetrameric simulations that extended at least 10 \u0026Aring; from the solute surface. Na\u003csup\u003e+\u003c/sup\u003e and Cl\u003csup\u003e\u0026minus;\u003c/sup\u003e counter ions were added to neutralize the system and achieve a salt concentration of 150 mM. Simulations of BALF2 monomer to predict magnesium binding sites were performed in 1 mM of MgCl\u003csub\u003e2\u003c/sub\u003e instead. A constant temperature of 303.15 K was maintained using a Nos\u0026eacute;-Hoover temperature coupling thermostats for solute and solvent separately (Nos\u0026eacute;, \u003cspan citationid=\"CR100\" class=\"CitationRef\"\u003e1984\u003c/span\u003e, Hoover, \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e1985\u003c/span\u003e) with a time constant of 1 ps. An isotropic Parrinello\u0026thinsp;\u0026minus;\u0026thinsp;Rahman barostat (Parrinello and Rahman, \u003cspan citationid=\"CR105\" class=\"CitationRef\"\u003e1981\u003c/span\u003e) with a time constant of 5 ps and a compressibility of 4.5 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;5\u003c/sup\u003e bar \u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e was used to maintain a pressure of 1 atm. A time step of 2 fs was used, where bonds involving hydrogen atoms were constrained using the LINCS algorithm (Hess et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e1997\u003c/span\u003e). A cutoff distance of 0.9 nm was used for short-range van der Waals interactions. Particle mesh Ewald was used to treat electrostatic interactions with a real space cutoff of 0.9 nm. For all systems, 5000 steps of energy minimization were performed using the steepest descent algorithm followed by 125 ps of equilibration with positional restraints placed on all the protein and DNA heavy atoms (a force constant of 400 kJ/mol/nm\u003csup\u003e2\u003c/sup\u003e on the backbone atoms and 40 kJ/mol/nm\u003csup\u003e2\u003c/sup\u003e on the side chain atoms). This was followed by production runs of 1.5 \u0026micro;s in case of monomers and 200 ns for the tetramer systems. Five independent replicates for each system were simulated. Snapshots were saved every 100 ps. VMD 1.9.4 alpha 57 (Visual Molecular Dynamics) (Humphrey et al., \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e1996\u003c/span\u003e) and MDAnalysis (Michaud-Agrawal et al., \u003cspan citationid=\"CR91\" class=\"CitationRef\"\u003e2011\u003c/span\u003e) were used to analyse the trajectories.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003eElectromobility Shift Assays\u003c/h2\u003e\u003cp\u003eTo assess binding to a single oligonucleotide, BALF2 (0-2000 nM) was titrated against 150 nM O2, then incubated for 30 minutes at 37\u0026deg;C in Binding buffer. Samples were then mixed with 10X Orange G Loading Dye (0.01% \u003cem\u003ew/v\u003c/em\u003e Orange G, 10 mM Tris-HCl (pH 7.5), 5 mM EDTA) before loading on a 4\u0026ndash;20% Mini-PROTEAN\u0026reg; TGX\u0026trade; Precast Gel (Bio-Rad) and run for 150 minutes at 50 V in 2X TBE. Gels were imaged using an Amersham Typhoon\u0026trade; (Cytiva). This was also repeated in Annealing buffer. Experiments were performed in triplicates. Densitometric analysis was conducted using ImageQuant (Cytiva). To assess ssDNA annealing, 1 \u0026micro;M BALF2 was incubated with 0.1 \u0026micro;M O1 at 37\u0026deg;C in Annealing buffer. After 30 minutes, 0.1 \u0026micro;M O2 or 0.1 \u0026micro;M O3 was added before further incubation for 5 minutes. Thermally annealed reactions without BALF2 were also included as controls. Samples were then mixed with 10X Orange G Loading Dye before loading on 1% Agarose gel and run for 150 minutes at 50 V in 2X TBE. Gel was imaged using an Amersham Typhoon\u0026trade; (Cytiva). This was repeated with reactions run in Binding buffer.\u003c/p\u003e\u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003eCloning, Expression, Purification, and Initial Characterisation\u003c/p\u003e\n\u003cp\u003eBALF2 was expressed in SF9 insect cells and lysed by sonication (Extended Data 1a). The soluble fraction of the lysate was filtered and purified by C-tag affinity chromatography to \u0026gt;95% purity (Extended Data 1b, c). BALF2 was flash-frozen in liquid nitrogen and stored in Storage Buffer at -80 \u0026deg;C.\u003c/p\u003e\n\u003cp\u003eInitially, negative-staining electron microscopy (NS-EM) was used to assess the structural homogeneity of BALF2. BALF2 appeared as a field of irregularly shaped particles 10.4\u0026nbsp;\u0026plusmn;\u0026nbsp;0.2 nm (n = 50) in diameter, consistent with monomers (Extended Data 2a) (Mumtsidu et al., 2008). Irregular protein filaments, appearing to consist of a single line of monomers, could be formed after incubation with reducing agent (Extended Data 2b). These formed best after 6 hours of incubation, after which they began to fall apart and were completely absent by 10 hours. Regular helical assemblies could be formed after incubation with the semi-self-complementary O2 oligonucleotide (Extended Data 2c,d). MgCl\u003csub\u003e2\u003c/sub\u003e was essential for the formation of these filaments, which were noticeably thicker than the thin protein-only irregular complexes described above. These could be very long (\u0026gt; 1 \u0026micro;m) and tended to break, branch, or loop back on themselves to form circular complexes. \u0026nbsp;Under NS-EM conditions, their pitch (41.7 \u0026plusmn; 0.4 nm; n = 114), diameter (13.9 \u0026plusmn; 0.2 nm; n = 175), and helical striation (22 \u0026plusmn; 1.1\u0026deg;; n = 11) could be estimated. The stability of these filaments was optimal in the presence of reducing agent, at a pH 9.0, and at 2.5x stoichiometric binding ratio (assuming a BALF2 site-size of 15 nt).\u003c/p\u003e\n\u003cp\u003eTo investigate binding to ssDNA, BALF2 was incubated with a long ssDNA substrate (mp 18 M13) for 30 minutes, and timepoints analysed by NS-EM. BALF2-coated M13 ssDNA appears as thin irregular complexes resembling beaded filaments, not dissimilar to the thin protein-only filaments formed under reducing conditions. At sub-saturating concentrations, BALF2 appears initially in short stretches on DNA, indicative of a cooperative binding mode (Extended Data 3a). These intermediate complexes were often tangled and could still be seen after 15 minutes of incubation. \u0026nbsp;In contrast, at saturating concentrations of BALF2, this reaction is rapid as fully coated M13 DNA were immediately visible (Extended Data 3b). Given their widths (10.6 \u0026plusmn; 0.2 nm; n = 31), we hypothesised that these were nucleoprotein complexes composed of a single strand of monomers coating a single strand of DNA. Under NS-EM conditions, the fully coated complexes were held in a slightly extended conformation of 3.6 \u0026plusmn; 0.06 \u0026Aring;/nt compared to dsDNA. By measuring the distance between single BALF2 particles on the DNA (5.7 \u0026plusmn; 0.1 nm), we were able to estimate an approximate site-size of 16.1 \u0026plusmn; 0.2 nt (n = 245). This value is in good agreement with those for ICP8 and ORF6\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e(Mumtsidu et al., 2008, Tsurumi et al., 1996, Ozgur and Griffith, 2014, Gourves et al., 2000). Similar complexes, resembling short rods, were also observed when binding to a much shorter substrate O3 (Extended Data 3c).\u003c/p\u003e\n\u003cp\u003eConsidering the findings above, we hypothesised that the thick regular helical filaments which form upon incubation with the semi-self-complementary O2 may be BALF2 annealing intermediates. To assess the identity of these thick filaments, we ran a time-course assay of BALF2-facilitated annealing of heat-denatured DNA (Extended Data 4). Upon addition of the ssDNA, thick helical filaments resembling those produced in the presence of O2 began to form and generally became longer and more regular upon prolonged incubation. At points within these thicker filaments, they either broke apart, forked, or unravelled, revealing that these were composed of two thin nucleoprotein filaments which wound about each other. We refer to each thin filament as a protofilament. They also commonly looped back on themselves to form a hairpin complex showing a thin region within the loop. This is consistent with the current herpesvirus-annealase facilitate SSA model, which features two protofilaments, each coating ssDNA, winding about one another to facilitate annealing (Weerasooriya et al., 2019). After 24 hours, the complexes were still visible. Interestingly, this meant that the filaments observed in the presence of our 60-mer oligonucleotide would require further interaction between nucleoprotein complexes to reach the lengths observed. In our experiments, we commonly observed both short thick complexes, and longer beaded concatemeric BALF2-ssDNA complexes initially and after longer periods of incubation; we therefore cannot conclude as to the mechanism of filament growth, though it is noteworthy it can occur at all.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eCryo-EM of BALF2 ssDNA-annealing Intermediate\u003c/p\u003e\n\u003cp\u003eSince we were able to form stable annealing intermediates with a regular turn through the addition of a O2, we proceeded this sample to cryo-EM (Fig. 1). BALF2 was incubated with O2 for 30 minutes at 37 \u0026deg;C in the presence of BME and Mg\u003csup\u003e2+\u003c/sup\u003e, before further incubation for 20 hours overnight at 4 \u0026deg;C. The sample was then cryo-plunged and imaged using a Titan Krios equipped with a Gatan K3 detector (Fig. 1a). 6,248 movies were collected on UltrAUfoil grids and processed using CryoSPARC v4.0. Due to the inherent flexibility of the filaments as visualised by three-dimensional variability analysis (3D-VA), a more traditional single-particle reconstruction approach was used over helical processing, where one asymmetric unit was treated as a single particle (Supp. Movie 1). Through template-based picking and particle classification, a total of 557,352 asymmetric units were identified for further processing (Fig. 1b). By local refinement, the asymmetric unit was isolated and refined to 2.19 \u0026Aring;, using the gold-standard FSC cutoff 0.143 (Fig. 2c) (PDB ID: 9BYQ). Further movement was observed by 3D-VA so local refinement to isolate a single subunit was also performed to generate a 2.16 \u0026Aring; map (Fig. 1d, Supp. Movie 2) (PDB: 9BYP). Structures were built into these maps, revealing the asymmetric unit to consist of a dimeric BALF2 complex, with each subunit bound to ssDNA (Fig. 2e-g). Using an additional continuous carbon grid dataset, a low-resolution map of approximately one helical pitch was also generated and used to rigid fit the structure of the asymmetric unit to generate a composite model of the filament (PDB ID: 9BYR). An overview of the processing pipeline is included in Extended Figure 5. Unfortunately, due to the inherently flexible nature of the filaments, the helical parameters are not static and could not be used for further processing of the filament.\u003c/p\u003e\n\u003cp\u003eOverall Structure\u003c/p\u003e\n\u003cp\u003eThe annealing intermediate of BALF2 is a highly flexible, left-handed bipolar filament consisting of two protofilaments of BALF2 monomers which dimerize at each asymmetric unit, related by C2 symmetry (Fig. 1h,i). Each protofilament contains a single strand of DNA, which runs opposite in each direction of each protofilament, providing the bidirectionality required for the formation of B-DNA. With reference to Fig. 1e, and following the established nomenclature, the monomeric structure consists of a large non-contiguous NTD (residues 9\u0026ndash;981) consisting of head and shoulder regions which meet at a narrow neck. The head region is alpha helical and contains 8\u0026nbsp;\u0026alpha;-helices and 1 3\u003csub\u003e10\u003c/sub\u003e-helix. The neck region is almost entirely composed of a\u0026nbsp;\u0026beta;-barrel-like OB-fold which contributes to an electropositive cleft bound to ssDNA. In the asymmetric unit, each BALF2 monomer is positioned so the ssDNA is brought together at the dimeric interface to form dsDNA. The shoulder region sits below the neck and is by far the largest of the three regions, containing 36 helices (13\u0026nbsp;\u0026alpha;-helices, 17 3\u003csub\u003e10\u003c/sub\u003e-helices, 5 mixed helices), and 15\u0026nbsp;\u0026beta;-strands. The strands are organised into 5 antiparallel\u0026nbsp;\u0026beta;-sheets, the most prominent of which is a 6-stranded curved sheet on the sloped side of the shoulder. In addition, this region also contains a zinc finger with bound zinc atom. The NTD also contains two large unresolved loops which we refer to as loop 1 (residues 511\u0026ndash;524) and loop 2 (residues 950\u0026ndash;960) positioned on either side of each monomer at the interfaces between asymmetric units. The NTD is connected by a short linker to a small C-terminal domain (CTD) (residues 988\u0026ndash;1090) which docks into the NTD of the neighbouring asymmetric unit in a domain-swapped configuration. The dimeric partner in the other protofilament also engages in CTD docking, however in the opposite direction. The CTD is composed of an\u0026nbsp;\u0026alpha;-helical bundle comprised of six helices (residues 988\u0026ndash;1076), which docks into a cleft between the head and body regions, and a tail (residues 1083\u0026ndash;1090) which interacts with a region in the head domain to form an antiparallel\u0026nbsp;\u0026beta;-sheet. Residues at the beginning of the chain (residues 1\u0026ndash;8) and end (residues 1091\u0026ndash;1128) are unresolved and assumed to be disordered.\u003c/p\u003e\n\u003cp\u003eDNA-Binding Mechanism\u003c/p\u003e\n\u003cp\u003eTwo strands of ssDNA consisting of 12 nucleotides could be modelled around the neck of each monomer in a continuous channel (Fig. 2a). We number these positions p1\u0026ndash;12 in a 5\u0026prime; \u0026ndash; 3\u0026prime; direction. The sequences chosen to be modelled were from the region of greatest self-complementarity of O2 since individual bases could not be identified, likely due to particle averaging. The ssDNA is held in an extended conformation of 6.3 \u0026plusmn; 0.2 \u0026Aring;/nt. It consists of bases which face inwards to the monomeric unit (p1\u0026ndash;5, p9\u0026ndash;12; unpaired bases) and outwards to the dimerizing BALF2 monomer (p6\u0026ndash;8; paired bases) to facilitate homology-pairing. The unpaired bases are bound primarily through \u0026pi;-\u0026pi; interactions (Y497, F509, W528, Y920 and Y937) or through the formation of hydrogen bonds directly with the bases themselves (S849 and N934), or phosphate backbone (N732) (Fig. 2b).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe noted that these binding mechanisms create promiscuous architecture at each nucleotide-binding site, which we hypothesised allows BALF2 to accommodate any base at that position. To demonstrate this, each base was modelled into the same site (p3), supported by density, using molecular dynamics simulations implemented in ISOLDE (Croll, 2018) (Extended Data 6). In this example, the bases are accommodated for by the OH group of S849 which may act as a donor (for guanine and thymine) or acceptor (for cytosine or adenosine), and a water molecule which helps bridge the gap between the pyrimidines and the peptide backbone. Interestingly, we were also able to model in uracil.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe ssDNA is bound by a \u0026beta;-barrel-like motif capped on top by a short \u0026alpha;-helix and supported by a much longer kinked helix below (Fig. 2c). The architecture of this OB-fold is \u0026beta;-\u0026beta;-\u0026beta;-\u0026alpha;-\u0026beta;-\u0026alpha;-\u0026beta;-\u0026beta;-\u0026beta; (3\u0026beta;-\u0026alpha;\u0026beta;\u0026alpha;-3\u0026beta;) (Fig. 2d). Due to the low sequence similarity between annealases, a search using the DALI server for structural homologues of BALF2 was conducted (Extended Data 7a). Unsurprisingly, the best match was for ICP8 (PDB ID: 1URJ) which showed significant secondary structural similarity throughout the entire BALF2 chain. Otherwise, four relevant structures were identified: two undefined \u0026lsquo;phage-related\u0026rsquo; proteins from \u003cem\u003eBacillus cereus\u003c/em\u003e (PDB ID: 4JG2) and \u003cem\u003eEnterococcus faecalis\u0026nbsp;\u003c/em\u003e(PDB ID: 4KLK), the family-defining gp2.5 annealase from bacteriophage T7 (PDB ID: 1JE5), and the Enc34 SSB (PDB ID: 5ODL), a phage of \u003cem\u003eEnterobacter cancerogenus\u003c/em\u003e, bound to ssDNA (Extended Data\u0026nbsp;7b,c)\u003cem\u003e.\u0026nbsp;\u003c/em\u003eIn these, \u0026beta;6/7 were combined into a single strand and \u0026alpha;1 was not kinked.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe active site is located at the dimeric interface, where nucleotides in p6\u0026ndash;8 are paired with nucleotides p8\u0026ndash;6 (respectively) from the opposite monomer to facilitate microhomology-based annealing (Fig. 2e). This site is stabilised through several interactions (Fig. 2f). First, hydrogen-bonds are maintained by\u0026nbsp;\u0026beta;4\u0026nbsp;positively charged residues R718, K721, and K723\u0026nbsp;and\u0026nbsp;the phosphate backbone. Second, a short loop which we term the \u003cstrong\u003eD\u003c/strong\u003eNA-\u003cstrong\u003eS\u003c/strong\u003etabilisation \u003cstrong\u003eL\u003c/strong\u003eoop (DSL) between\u0026nbsp;\u0026beta;5/6 (residues 920\u0026ndash;930) frames the site on each end. Y920 forms hydrogen-bonds with the phosphate backbone and induces \u0026pi;-stacking with the unpaired p5 nucleotide; N924 and Q926 coordinate the paired bases at p6/8; and F930 engages the p6 nucleotide. Third, Q928 of this loop interacts with K670 on the opposite monomer. This network of bonds holds each strand so that the annealing interface is a planar conformation. Though these residues show varied levels of conservation in our MSA, Y920 and F930 universally occur as either phenylalanine or tyrosine, highlighting their importance in coordinating DNA.\u003c/p\u003e\n\u003cp\u003eOligomerisation\u003c/p\u003e\n\u003cp\u003eThe mechanism of filament formation appears to be driven primarily by a domain-swapping interaction between the CTD and the NTD of the neighbouring subunit of the same protofilament (Fig. 3a,b). This is a two-fold mechanism. Firstly, the \u0026alpha;-helical bundle region of the CTD is highly negatively charged, which drives docking into the positive cleft between the head and shoulder domains. This interaction is strengthened by several salt bridges (Fig. 3c). Secondly, hydrophobic residues in the CTD tail (F1085, I1086 and V1088) interact with a shallow hydrophobic cleft in the head of the neighbouring subunit, forming an antiparallel \u0026beta;-sheet (Fig. 3d). In addition to this domain swap, Loop 1 (residues 511\u0026ndash;524) is in the head domain and appears to interact with a region in the head of the neighbouring monomer. Loop 2 (residues 950\u0026ndash;960) is in the neck region and is positioned nearby the 5\u0026prime;-end of the ssDNA, and may be involved in ssDNA-interactions as well as potentially oligomerisation\u003cstrong\u003e.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThere are also relatively weaker protein-protein interactions at the dimeric interface. While the most dominant interaction seems to be via DNA-DNA interactions, dimerization also appears to be mediated by hydrogen-bonding through two sites. At the interface of the head domains, a single hydrogen bond is formed between N585 residues. In the other (Fig. 3e), what appears to be the major site of interaction features a kinked\u0026nbsp;\u0026alpha;-helix followed by a short loop (residues 648\u0026ndash;670) which is positioned just below the active annealing site. Collectively, we term this region the Dimerization Helix (DH). The short loop forms hydrogen bonds with the DSL on the opposite monomer, through D669 and K670 which interact with N927 and Q928, respectively. This would presumably also aid to stabilise the active site. The lower helical region of the DH forms hydrogen bonds with the corresponding residues on the opposite monomer through Q657 and T658.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eZinc Binding Site\u003c/p\u003e\n\u003cp\u003eZinc bound within a zinc finger loop, flanked by helical regions (residues 440\u0026ndash;472), was identified on the opposite side of the protein to the active site (Fig. 4a). The position of the loop is stabilised by a conserved arginine (R471) which sits at the top of the second flanking helix and interacts with the backbone immediately preceding the first helical structure. The finger coordinates zinc tetrahedrally through a universally conserved C-C-C motif (C453, C456, and C464) and the participation of a water molecule. The water molecule is bound through interactions with the backbone; and by Y443 and T467 through another water molecule. The latter residues are also universally conserved, except Y443 which in \u0026beta;-herpesviruses is a histidine. The fold this site adopts, through zinc coordination, causes it pack between three different regions of the non-contiguous NTD. Overall, the site is very similar to that of ICP8, in its relative position to the active site (Mapelli et al., 2005).\u003c/p\u003e\n\u003cp\u003eSequence Analysis\u003c/p\u003e\n\u003cp\u003eA multiple sequence alignment (MSA) of all herpesvirus annealases was first generated using all 652 amino acid sequences in the herpesvirus annealase family as compiled by pFam (PF0047) (Supp. Data 1). Phylogenetic analysis of this alignment resolved three distinct and well supported clades which correlated with the three subfamilies of \u003cem\u003eorthoherpesviridae\u0026nbsp;\u003c/em\u003e(full support for both ultrafast bootstrap and SH-aLRT for each clade)\u003cem\u003e\u0026nbsp;\u003c/em\u003e(Supp. Data 2).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWithin our phylogenetic tree human and animal herpesvirus annealases generally resolved together forming taxonomically broad clades which represented each of the \u003cem\u003eorthoherpesviridae\u003c/em\u003e viral subfamilies. The only exception to this is ICP8 (from HSV-1 and HSV-2) and U41 (from HHV-6a, HHV-6b, and HHV-7) which each formed a distinct clade. Given the only moderate sequence similarity between subfamilies, indicated by our alignment and cited in literature (Mumtsidu et al., 2008, Mapelli et al., 2005), we also created subalignments for each of the \u0026gamma;, \u0026alpha;, and \u0026beta; clades to enable assessment of conservation within and between families (Supp. Data 3, 4, 5). When conservation of these subalignments was mapped onto the BALF2 structure, we found that in both the global and \u0026gamma; alignments conservation was most apparent within the OB-fold and the core of the protein. Notably, in the \u0026gamma; alignment (the subfamily in which BALF2 is placed), we saw significantly greater conservation within these regions, as well as the DH region and within the head of the protein. A reduced alignment of the global MSA containing only seven representative sequences is included in Fig. 5c.\u003c/p\u003e\n\u003cp\u003eMolecular Dynamics Simulations of BALF2\u003c/p\u003e\n\u003cp\u003eMetal3D (D\u0026uuml;rr et al., 2023) predicted the presence of a Zn\u003csup\u003e2+\u003c/sup\u003e in the experimentally identified zinc binding site. To investigate the importance of the Zn\u003csup\u003e2+\u0026nbsp;\u003c/sup\u003eon the structural integrity of the protein, we ran molecular dynamics (MD) simulations of BALF2 monomers in the presence and absence of a Zn\u003csup\u003e2+\u003c/sup\u003e (Fig. 6a,b). Both systems showed stable RMSD with increased flexibility at the zinc binding site region (residues 440-472) in absence of a bound Zn\u003csup\u003e2+\u003c/sup\u003e (Extended Data 8a,b). In the simulations that had a Zn\u003csup\u003e2+\u003c/sup\u003e, it remained bound at its binding site coordinated by the three cysteines (C453, C456, and C464) along the simulations. The two water molecules at the zinc binding site were maintained during the simulations as revealed by the\u0026nbsp;average water density map calculated using the VolMap plugin in VMD\u0026nbsp;(Humphrey et al., 1996)\u0026nbsp;and the radial distribution function of water molecules surrounding the\u0026nbsp;Zn\u003csup\u003e2+\u0026nbsp;\u003c/sup\u003e(Extended Data 8c,d).\u003c/p\u003e\n\u003cp\u003eWe also conducted MD simulations of a system composed of two asymmetric units, which we call a tetramer without DNA, with one strand of ssDNA, or with both strands of ssDNA. These simulations of BALF2 tetramers show more stable average RMSD in presence of the two DNA strands (Fig. 6c-e and Extended Data 9a,b) and higher flexibility of loop 2 (residues 950\u0026ndash;960) in absence of DNA (Fig. 6f-h and Extended Data 9c). In addition, loop 2 when at the interface of asymmetric units compared to a terminal end was also more constrained, suggesting it has roles both in oligomerisation and in ssDNA interaction.\u003c/p\u003e\n\u003cp\u003eMetal3D (D\u0026uuml;rr et al., 2023), BioMetAll (S\u0026aacute;nchez-Aparicio et al., 2021) and MIB (Lin et al., 2016) were used to predict the magnesium binding site. Ten common binding sites were identified where Mg\u003csup\u003e2+\u0026nbsp;\u003c/sup\u003ebound to\u003csup\u003e\u0026nbsp;\u003c/sup\u003eE155, E159, D190, E238, E256, E684, D800, D899, D1032, D1033, and E1035 and were used as a starting structure for MD simulations of BALF2 monomers. All Mg\u003csup\u003e2+\u003c/sup\u003e remained bound during the simulations, and we didn\u0026rsquo;t observe any unbinding. This is likely due to tight binding between Mg\u003csup\u003e2+\u0026nbsp;\u003c/sup\u003eand the negatively charged coordinating residues and much longer time scale required for unbinding comparing to the simulation time scale.\u003c/p\u003e\n\u003cp\u003eBiochemical Characterisation of BALF2\u003c/p\u003e\n\u003cp\u003eIn addition to the structural characterisation addressed by this work, we also investigated the biochemical properties of BALF2 by electromobility shift assay (EMSA). BALF2 was titrated against O2 in binding buffer before being run on 4\u0026ndash;20% native PAGE in 2X TBE in triplicate (Fig. 7a,b). As the concentration of BALF2 was increased, bands representing multiple oligomeric weight species appeared, even though there was still unbound DNA in solution. This is indicative of a cooperative binding mode, agreeing with our negative-staining results. A dissociation constant (\u003cem\u003eK\u003csub\u003ed\u003c/sub\u003e\u003c/em\u003e) of 467 \u0026plusmn; 13 nM was calculated via densitometry and sigmoidal curve fitting (R\u003csup\u003e2\u003c/sup\u003e = 0.989). When this was repeated in the presence of MgCl\u003csub\u003e2\u003c/sub\u003e, the curves overlaid well and there was no significant difference (p \u0026gt; 0.05) in the dissociation constant (R\u003csup\u003e2\u003c/sup\u003e = 0.995; \u003cem\u003eK\u003csub\u003ed\u003c/sub\u003e\u003c/em\u003e = 473 \u0026plusmn; 1 nM). We therefore conclude that MgCl\u003csub\u003e2\u003c/sub\u003e does not contribute to binding to ssDNA, despite being required for filament formation.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo analyse the annealing activity of BALF2, complexes were formed using fluorescently labelled oligonucleotides O1 (5\u0026prime;-labelled with Cy5), O2 (3\u0026prime;-labelled with Cy3), and O3 (3\u0026prime;-labelled with Cy3). Cy3 and Cy5 are a FRET (F\u0026ouml;rster resonance energy transfer) pair, where excitation of Cy3 causes emission at a wavelength capable of exciting Cy5, whose emission can then be measured (Fig. 7c). This only occurs over short (\u0026lt;10 nm) distances, and therefore only occurs when O1 is annealed to the complementary O2, but not when O1 is in the presence of the non-complementary O3. For each reaction we can therefore tell what DNA is bound, and if it is ssDNA or dsDNA. We found that BALF2 forms a stable complex with all three oligonucleotides even after 1 hour of incubation, however the addition of a complementary oligonucleotide after 30 minutes causes BALF2 to dissociate from the DNA, which we conclude is due to an annealing reaction and thus the formation of dsDNA. This reaction was rapid and was almost complete after only 5 minutes of incubation, indicated by the very faint white band when O2 is added to the reaction (lane 7 in the presence of MgCl\u003csub\u003e2\u003c/sub\u003e). When this was repeated in the presence of EDTA, only very weak signal was visualised representing an annealed product and remaining visible DNA was present as BALF2-bound complexes. When the non-complementary oligonucleotide was added instead, no FRET signal could be visualised in either treatment; interestingly, unbound O1 was seen in the presence of MgCl\u003csub\u003e2\u003c/sub\u003e perhaps due order of addition of oligonucleotides. We therefore concluded that BALF2 can facilitate annealing of short oligonucleotides in a MgCl\u003csub\u003e2\u003c/sub\u003e-dependent manner.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eAs a herpesvirus annealase, BALF2 represents a member of a long-studied group of proteins, first discovered over 50 years ago. Despite its potential as a drug target and as a biotechnological tool, its structure and mechanism have remained unknown, until now. In this study, we have determined the structure of BALF2 as a ssDNA annealing intermediate to 2.2 \u0026Aring;. This has allowed us to characterise the helical filament BALF2 forms, an OB-fold involved in DNA binding, the active site of ssDNA annealing, and a zinc-finger fold. We also performed MD simulations on monomeric and tetrameric structures and investigated BALF2-DNA interactions by EMSA biochemical assays. Our findings for the first time effectively corroborate and explain many of the observations recorded in the literature (Mumtsidu et al., \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2008\u003c/span\u003e, Weerasooriya et al., \u003cspan citationid=\"CR138\" class=\"CitationRef\"\u003e2019\u003c/span\u003e, Ozgur and Griffith, \u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), and suggest further avenues for investigation.\u003c/p\u003e\u003cp\u003eIt is well-established that BALF2 shows sequence-independent ssDNA-specific binding ability (Tsurumi et al., \u003cspan citationid=\"CR133\" class=\"CitationRef\"\u003e1996\u003c/span\u003e, Tsurumi et al., \u003cspan citationid=\"CR132\" class=\"CitationRef\"\u003e1998\u003c/span\u003e). The ssDNA binding cleft was revealed to be around the neck of the protein, consistent with a number of proposed basic and aromatic residues from literature (Mapelli et al., \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2005\u003c/span\u003e, Wang and Hall, \u003cspan citationid=\"CR137\" class=\"CitationRef\"\u003e1990\u003c/span\u003e, Gao and Knipe, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e1989\u003c/span\u003e, Gao et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e1988\u003c/span\u003e, Leinbach and Heath, \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e1989\u003c/span\u003e). Promiscuous architecture at each nucleotide binding site appears to allow BALF2 to bind any ssDNA sequence, as we demonstrated at nucleotide binding site p3. This binding mechanism, which relies on both the exposed bases and the relaxed conformation inherent to ssDNA, provides convincing evidence for why herpesvirus annealases would show reduced affinity to B-DNA, but can non-specifically bind ssDNA (Ruyechan and Weir, \u003cspan citationid=\"CR116\" class=\"CitationRef\"\u003e1984\u003c/span\u003e, Lee and Knipe, \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e1985\u003c/span\u003e). Interestingly, a low RNA affinity has also been reported on the same scale as dsDNA-affinity (Ruyechan and Weir, \u003cspan citationid=\"CR116\" class=\"CitationRef\"\u003e1984\u003c/span\u003e, Boehmer, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2004\u003c/span\u003e, Lee and Knipe, \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e1985\u003c/span\u003e). Interestingly, we were able to successfully model uracil into the site (Extended Data 6). This may suggest that the reduced RNA affinity may instead be due to steric hindrance by the additional hydroxyl group in the RNA ribose ring. F930, whose aromatic property is universally conserved (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec), is engaged in π-stacking interactions with the ribose ring of the nucleotide in p6, which may not be possible in RNA.\u003c/p\u003e\u003cp\u003eThe BALF2 OB-fold is 3β-αβα-3β (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), has not, despite the structural homologues we identified, been described before as it appears in BALF2. The key differences between BALF2 and the gp2.5/Enc34 SSB are the β6/7 strands, which are is split in our structure but is part of a continuous strand in these homologues; and the α1 helix, which is longer and kinked in BALF2. Nonetheless, the identification of gp2.5 and Enc34 SSB as structural homologues was unexpected. Gp2.5 is the much better established of the two and is the family-defining annealase of the gp-like family (Hernandez and Richardson, \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2019\u003c/span\u003e, Lopes et al., \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). In short, it has many similarities to our current understanding of BALF2. gp2.5 is produced with the immediate-early genes of the T7 replisome (Reuben and Gefter, \u003cspan citationid=\"CR109\" class=\"CitationRef\"\u003e1973\u003c/span\u003e, Reuben and Gefter, \u003cspan citationid=\"CR110\" class=\"CitationRef\"\u003e1974\u003c/span\u003e). It preferentially binds ssDNA cooperatively and eliminates secondary structure in a direction-independent manner (Xu et al., \u003cspan citationid=\"CR146\" class=\"CitationRef\"\u003e2023\u003c/span\u003e, Reuben and Gefter, \u003cspan citationid=\"CR109\" class=\"CitationRef\"\u003e1973\u003c/span\u003e, Reuben and Gefter, \u003cspan citationid=\"CR110\" class=\"CitationRef\"\u003e1974\u003c/span\u003e, Zou et al., \u003cspan citationid=\"CR155\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). It is capable of strand-transfer and annealing reactions through a similar mechanism to BALF2: first by the coating of two ssDNA regions sharing homology, before they are brought together and annealed, potentially through gp2.5-gp2.5 interactions (Zou et al., \u003cspan citationid=\"CR155\" class=\"CitationRef\"\u003e2018\u003c/span\u003e, Hernandez and Richardson, \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2019\u003c/span\u003e, Makhov et al., \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2009\u003c/span\u003e, Kim and Richardson, \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e1993\u003c/span\u003e, Hyland et al., 2003, Rezende et al., 2003, Kong et al., 1997). It binds to ssDNA as a monomer but can oligomerise in solution, as dimers, which has also been reported for BALF2 (Hollis et al., \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2001\u003c/span\u003e, Rezende et al., \u003cspan citationid=\"CR114\" class=\"CitationRef\"\u003e2002\u003c/span\u003e, Mumtsidu et al., \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Mutations that impact ssDNA-binding, DNA-annealing, or oligomerisation are lethal to the T7 phage (Hyland et al., 2003, Rezende et al., \u003cspan citationid=\"CR114\" class=\"CitationRef\"\u003e2002\u003c/span\u003e, Rezende et al., 2003). Though it consists almost solely of the OB-fold, it also has a long C-terminal tail, which is implicated in protein-protein interactions and oligomerisation, like the herpesvirus annealases (Ghosh et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2010\u003c/span\u003e, Kim et al., \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e1992\u003c/span\u003e, Kim and Richardson, \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e1994\u003c/span\u003e). Despite low (~\u0026thinsp;15%) sequence identity to gp2.5, Enc34 SSB possesses some, if not all, of these traits (Cernooka et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). The Enc34 SSB structure is bound to ssDNA, which is positioned and oriented very similarly to our structure; this may also indicate a similar annealing or oligomerisation mechanism (Cernooka et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). The major biochemical difference between these proteins and BALF2 is an axillary feature of the C-terminal tail, which is thought to occupy the electrochemically complementary DNA-binding site in Enc34 SSB and gp2.5 in a regulatory role; its deletion increases ssDNA-affinity (Hollis et al., \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2001\u003c/span\u003e, Xu et al., \u003cspan citationid=\"CR146\" class=\"CitationRef\"\u003e2023\u003c/span\u003e, Marintcheva et al., \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e2008\u003c/span\u003e, Cernooka et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). The opposite is true for BALF2, ICP8 and ORF6, which do not have an acidic tail, and experience a loss of ssDNA-affinity upon its removal (Mapelli et al., \u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e2000\u003c/span\u003e, Mumtsidu et al., \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Combined with a previous report that mutations or deletions of conserved residues within the CTD preclude filament formation (Darwish et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), our results suggest this is due to a loss in cooperativity and filament formation and/or stability since the CTD tail cannot dock into the neighbouring monomer, as is described below.\u003c/p\u003e\u003cp\u003eSurprisingly, the OB-fold also strongly resembles the well-characterised OB-fold of the single-stranded binding proteins \u003cem\u003eE. coli\u003c/em\u003e SSB and human RPA, which lack the α1 helix and are non-recombinogenic (Naufer et al., \u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e2021\u003c/span\u003e, Zou et al., \u003cspan citationid=\"CR154\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). Like Redβ, BALF2 also holds ssDNA in a planar orientation during the annealing process; therefore, we expected BALF2 to share a similar fold (Newing et al., \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). In contrast, the Rad52-like superfamily (Redβ, RecT and Rad52) was not detected in our search, and upon manual inspection does not possess the 3β-αβα-3β fold (Caldwell et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2022\u003c/span\u003e, Kinoshita et al., \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2023\u003c/span\u003e, Newing et al., \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e2022\u003c/span\u003e, Lopes et al., \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). Why the OB-fold of BALF2 seems to more closely resemble non-recombinogenic SSB proteins rather than annealases from the Rad52-like family is unclear, however it may reflect observed differences in ssDNA-binding. For example, like SSB proteins and gp2.5, the herpesvirus annealases preferentially bind to ssDNA, whereas the stability of a Redβ nucleoprotein complex is greatest as a double-stranded annealing intermediate (Newing et al., \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e2022\u003c/span\u003e, Zakharova et al., \u003cspan citationid=\"CR148\" class=\"CitationRef\"\u003e2021\u003c/span\u003e, Naufer et al., \u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e2021\u003c/span\u003e, Zou et al., \u003cspan citationid=\"CR154\" class=\"CitationRef\"\u003e2006\u003c/span\u003e, Tsurumi et al., \u003cspan citationid=\"CR133\" class=\"CitationRef\"\u003e1996\u003c/span\u003e, Tsurumi et al., \u003cspan citationid=\"CR132\" class=\"CitationRef\"\u003e1998\u003c/span\u003e, Reuben and Gefter, \u003cspan citationid=\"CR109\" class=\"CitationRef\"\u003e1973\u003c/span\u003e, Reuben and Gefter, \u003cspan citationid=\"CR110\" class=\"CitationRef\"\u003e1974\u003c/span\u003e). Overall, based on structural similarity, we suggest that the herpesvirus annealases likely constitute a group within the gp-like superfamily (Lopes et al., \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). Whether this is due to divergent or convergent evolution is not clear. However, a shared lineage between herpesviruses and bacteriophage T7 is not a new avenue of thought (Baker et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2005\u003c/span\u003e, Kazlauskas and Venclovas, \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Moreover, bacteriophage T7 and λ-phage, though both dsDNA phages, differ structurally and in terms of replication strategy (Blasche et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2013\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eBALF2 was observed to oligomerise into a bipolar filament composed of two protofilaments, each of which coats a strand of ssDNA. We could not find any previous reports of BALF2 forming a helical assembly, whose architecture closely resembles what has been shown in the literature for KSHV ORF6 and HSV-1 ICP8 (Weerasooriya et al., \u003cspan citationid=\"CR138\" class=\"CitationRef\"\u003e2019\u003c/span\u003e, Makhov and Griffith, \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e2006\u003c/span\u003e, Makhov et al., \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2009\u003c/span\u003e, Tolun et al., \u003cspan citationid=\"CR129\" class=\"CitationRef\"\u003e2013\u003c/span\u003e, O'Donnell et al., \u003cspan citationid=\"CR101\" class=\"CitationRef\"\u003e1987\u003c/span\u003e, Darwish et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2016\u003c/span\u003e, Mumtsidu et al., \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2008\u003c/span\u003e, Ozgur et al., \u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e2011\u003c/span\u003e, Ozgur and Griffith, \u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). The driving mechanism for the formation of these filaments is a domain-swap of the CTD into the NTD of the neighbouring subunit. A key interaction of this is an antiparallel β-sheet further stabilised by hydrophobic interactions formed between the CTD tail and NTD. As we noted above, it has been shown that mutation (or removal) of conserved residues within the tail or hydrophobic cleft eliminate filament formation (Darwish et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). This interaction therefore is likely essential to stabilise the domain-swap and the annealing complex. Further, since this interaction would also likely occur when BALF2 coats ssDNA as a single nucleoprotein filament, this domain-swap mechanism explains why cooperative binding is abolished in mutants where the C-terminal tail has been removed (Mumtsidu et al., \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2008\u003c/span\u003e, Mapelli et al., \u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e2000\u003c/span\u003e, Ozgur and Griffith, \u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). Domain-swapping, particularly of a C-terminal element, is a common approach utilised in several systems to modulate cooperative binding to DNA. For example, the C-terminal tail residues of both the adenovirus DNA-binding protein or the bacteriophage Pf1 SSB are proposed to interact with the globular NTD of neighbouring subunits when forming a nucleoprotein filament (Chang and Shenk, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e1990\u003c/span\u003e, Fox et al., \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e1999\u003c/span\u003e). More recently, Tr\u0026auml;ger et al. (\u003cspan citationid=\"CR130\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) found that the annealing protein P12 from bacteriophage PRD1 forms a helical filament in the same bipolar arrangement as BALF2, with cooperative binding dependent on a CTD helical motif that docks into the NTD; and homology pairing at the dimeric interface. In addition to the domain-swap, the position of Loop 2 and the results of our molecular dynamics simulations suggest that it may also play a role in oligomerisation between neighbouring subunits. From our structures, we can\u0026rsquo;t conclude as to why we see more stable filaments under reducing conditions. Two cysteine residues (C48 and C212) are within close enough proximity to each other to form an intramolecular disulfide bond, however the density present did not support this, likely due to the sample preparation containing BME which reduces disulfide bonds. \u003cem\u003eIn vivo\u003c/em\u003e, it is feasible that these residues may aid to stabilise the monomer during periods of inactivity.\u003c/p\u003e\u003cp\u003eOverall, this filament structurally resembles the ORF6 filaments observed by Ozgur and colleagues (\u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e2011\u003c/span\u003e, 2014). The ORF6 filaments presented were reported to form during the purification process; were DNA-free and could incorporate ssDNA; and required a reducing environment but not MgCl\u003csub\u003e2\u003c/sub\u003e. While we also observed longer more stable filaments under reducing conditions, these previous findings generally contradict our results. Yet, there are factors which may have contributed to this disparity: First, intracellular ssDNA may have been present within the ORF6 preparation. Likewise, the nuclease protection activity associated with this class of proteins would explain why the filaments persisted after a nuclease incubation. Second, cellular MgCl\u003csub\u003e2\u003c/sub\u003e already involved in filament formation was not accounted for when magnesium was removed and could have also resisted chelation. Third, the observation of thin and thick ORF6 filaments in the same sample was initially determined to be due to the presence and absence of ssDNA, however our results suggest that this was a mixture of thin ORF6-ssDNA complexes and thick double-helical annealing intermediates. Given that ORF6 is the closest homologue to BALF2, and that ORF6 filaments do have the ability to incorporate ssDNA, it is tempting to conclude that these are the same filament (Ozgur and Griffith, \u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e2014\u003c/span\u003e, Ozgur et al., \u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). In contrast, low-resolution NS-EM maps of ICP8 suggest it adopts a tighter conformation, though is still a domain-swapped bipolar arrangement composed of dimeric asymmetric units, which are thought to dimerize to anneal ssDNA at the same interface as we present in our structure, likely reflecting a shared mechanism (Tolun et al., \u003cspan citationid=\"CR129\" class=\"CitationRef\"\u003e2013\u003c/span\u003e, Makhov et al., \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2009\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eZinc coordination in proteins may serve either a catalytic or purely structural purpose. We were able to identify density corresponding to a zinc ion coordinated by a CCC zinc finger and the participation of a water molecule (CCCw). We feel confident defining this as a structural zinc site due to several factors: First, structural zinc atoms are often bound tetrahedrally, commonly by cysteine residues, which facilitate a higher charge transfer than other residues and thus form a stronger bond (Lee and Lim, \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2008\u003c/span\u003e, Nguyen et al., \u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e1999\u003c/span\u003e). In contrast, catalytic zinc contacts usually consist of a mixture of cysteine, histidine, glutamate, or aspartate ligands and often take on coordination numbers of five or six; these interactions contribute to a more electronegative zinc ion which more readily acts as a Lewis acid and nucleophile (Lee and Lim, \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2008\u003c/span\u003e, Wolfe et al., \u003cspan citationid=\"CR143\" class=\"CitationRef\"\u003e2000\u003c/span\u003e). For example, alcohol dehydrogenase features both a catalytic site, involving a variety of C/H/D/E/H\u003csub\u003e2\u003c/sub\u003eO contacts, and a structural site which is almost always four cysteine residues (Auld and Bergman, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Despite the participation of a water molecule, the BALF2 zinc finger is therefore more characteristic of a structural site (Lee and Lim, \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Second, the ICP8 zinc finger (CCCH) is considered structural: mutation of the cysteine ligands results in a severe reduction in viral progeny in complementation assays, bound zinc resists extensive chelation which successfully removes zinc from alcohol dehydrogenase, and when zinc is removed ICP8 retains transient ssDNA-binding activity suggesting slow unfolding of the protein (Gao et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e1988\u003c/span\u003e, Gupte et al., 1991, Auld and Bergman, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Third, the BALF2 zinc finger is positioned on the opposite side of the protein to the active site and is packed against multiple regions of the protein. Fourth, the most accurate Zn\u003csup\u003e2+\u003c/sup\u003e location predictor to date Metal3D (D\u0026uuml;rr et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), predicted the presence of a Zn\u003csup\u003e2+\u003c/sup\u003e in the experimentally identified zinc binding site. As has been noted previously for ICP8, the occupation of the binding loop by Zn\u003csup\u003e2+\u003c/sup\u003e causes it to fold and, given the non-contiguous nature of the structure, allows this loop and associated helices to stabilise the full length of the protein (Mapelli et al., \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). This is reflected in the increased flexibility of the zinc binding site region (residues 440\u0026ndash;472) upon removal of Zn\u003csup\u003e2+\u003c/sup\u003e in our MD simulations. Considering that the ICP8 zinc finger is only conserved in α-herpesviruses, it would be unsurprising if ORF6 and members of the β-herpesvirus family also coordinated zinc tetrahedrally through a CCCw site, since the residues involved in water coordination are conserved. Interestingly, to our knowledge, a structural CCCw zinc site has not been described previously. Where waters are involved, the site architecture often consists of histidines and/or acidic residues unlike in BALF2 (Karpusas et al., \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e1997\u003c/span\u003e, Bouckaert et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e1996\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThough magnesium was reported to be crucial for both filament formation and annealing activity, we were not able to identify density corresponding to Mg\u003csup\u003e2+\u003c/sup\u003e anywhere in our structure. A divalent cation binding site consisting of two universally conserved acidic residues (D1032 and D1033) identified by Bryant et al. (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2012\u003c/span\u003e) is at the border of the CTD helical region and the NTD neck cleft. In theory, Mg\u003csup\u003e2+\u003c/sup\u003e coordination here may facilitate the domain-swap and filament formation. Interestingly, though not well-resolved, this site was also at the opening of a small pocket containing a network of water density which could also contain magnesium; but we did not build magnesium into this pocket since the density was ambiguous. Alternatively, magnesium could be located at the interface between asymmetric units, which would not be resolved due to our local refinement-based approach to reconstruction. Both D1032 and D1033 were also predicted as potential Mg\u003csup\u003e2+\u003c/sup\u003e binding sites in addition to E155, E159, D190, E238, E256, E684, D800, D899, and E1035. However, during the MD simulations, all Mg\u003csup\u003e2+\u003c/sup\u003e remained bound.\u003c/p\u003e\u003cp\u003eAt the dimeric interface, we identified two major stabilisation interactions: the hydrogen bonds formed by residues of the Dimerization Helix (DH) region, and the base-pairing hydrogen bonds created by each strand of ssDNA. Previously, mutations to Q706 and F707 in ICP8 (corresponding to BALF2 residues H655 and Y656) have eliminated annealing ability (Weerasooriya et al., \u003cspan citationid=\"CR138\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). The aromaticity of F707/Y656 is universally conserved and stabilises the helix in our structure by packing closely with aromatic hydrophobic residues (F693, F703 in ICP8; Y637 and W651 in BALF2). Mutation at this position may disorder the DH enough to disrupt crucial interactions involved in dimerization, eliminating annealing activity. A similar effect has been demonstrated in gp2.5, where mutation at an equivalent position relative to the OB-fold also removed annealing activity (Rezende et al., 2003).\u003c/p\u003e\u003cp\u003eThe DH-DH and ssDNA base-pairing interactions occur within proximity to each other. Using 3D-VA, we were able to demonstrate that this creates a hinge-like interaction, resulting in flexing, rotating, and sliding between the two monomers about the central planar ssDNA-ssDNA interaction (Supp. Movie 2). This is likely what gives the filament its flexibility. This movement involves the breaking and reformation of the inter-chain hydrogen bonds of the DH region, suggesting that the driver of dimerization is instead the ssDNA itself, which may only be a weak interaction given it only occurs across three base pairs in p6-8. Microhomology-based annealing has also been observed for DSB repair in human polymerase θ helicase. Though acting as a dimer only, polymerase θ anneals homologous short 3\u0026prime; overhangs at the dimeric interface, then translocates away from the newly synthesised duplex which allows any gaps to be filled for repair of the DSB (Zerio et al., \u003cspan citationid=\"CR150\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). An analogous mechanism may occur for BALF2: after the ssDNA is coated and annealed, the current annealing model maintains that the herpesvirus annealase then dissociates from the newly synthesised ssDNA, consistent with our EMSA annealing assay (Weerasooriya et al., \u003cspan citationid=\"CR138\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Considering this, we propose the following annealing mechanism: first, ssDNA is coated by BALF2, cooperatively driven by a domain-swapping interaction between the CTD into the NTD of the neighbouring monomer. Two sufficiently homologous BALF2-coated ssDNA protofilaments are then brought together to form a bipolar filament to facilitate microhomology-based searching at the dimeric interfaces of each asymmetric unit. Once paired correctly, this causes dsDNA to form, which takes on the canonical B-DNA helical twist. In this form, the DNA is more rigid and condensed, which places steric stress on both the annealing site and non-pairing bases. This, coupled with the relatively dynamic DH-DH interactions, causes the monomers to dissociate from the complex, completing the reaction. Contrary to the description of the last step (annealed product being released), we were still able to visualise annealing intermediates in our NS-EM annealing reaction after a prolonged incubation of 24 hours, which is inconsistent with both our suggested model and literature. We interpret this as an incomplete annealing reaction that could be the result of significant homology within the same ssDNA strand, or inadequate reaction conditions to facilitate efficient homology searching on a long template, despite complete reactions being demonstrated previously for ICP8 under similar conditions (Weerasooriya et al., \u003cspan citationid=\"CR138\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eWhile the findings presented explain several key concepts of annealase-facilitated ssDNA annealing, our structure poses further questions. Why is the oligomerisation in α- and γ-herpesviruses different? What are the detailed dynamics for homology searching, and is this linked at all to the movement observed in 3D-VA? What is the basis for the high degree of structural homology between the gp2.5-like family and BALF2? Outside the scope of this work, lines of inquiry still unaddressed include how exactly BGLF5 interacts with BALF2, and how, if at all, may this system be effectively used in recombineering or as a drug target. Nonetheless, this structure sheds significant light on the mechanisms involved in the essential process of SSA in human herpesviruses.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003e3D-VA: Three-dimensional variability analysis\u003c/p\u003e\n\u003cp\u003eBALF2\u0026Delta;C: BALF2 deletion mutant, missing the C-terminal 60 residues\u003c/p\u003e\n\u003cp\u003eBME: \u0026beta;-mercaptoethanol\u003c/p\u003e\n\u003cp\u003eCTD: C-terminal domain\u003c/p\u003e\n\u003cp\u003eDH: Dimerization Helix\u003c/p\u003e\n\u003cp\u003edsDNA: double-stranded DNA\u003c/p\u003e\n\u003cp\u003eDSB: double-stranded DNA break\u003c/p\u003e\n\u003cp\u003eDSL: DNA-stabilisation loop\u003c/p\u003e\n\u003cp\u003eEBV: Epstein-Barr virus\u003c/p\u003e\n\u003cp\u003eEATR: Exonuclease-Annealase Two-component Recombinase\u003c/p\u003e\n\u003cp\u003eFRET: F\u0026ouml;rster resonance energy transfer\u003c/p\u003e\n\u003cp\u003eHHV-4: Human herpesvirus 4\u003c/p\u003e\n\u003cp\u003eHR: Homologous recombination\u003c/p\u003e\n\u003cp\u003eHSV-1: Human herpes simplex virus 1\u003c/p\u003e\n\u003cp\u003eICP8\u0026Delta;C: ICP8 deletion mutant, missing the C-terminal 60 residues\u003c/p\u003e\n\u003cp\u003eKSHV: Kaposi\u0026rsquo;s sarcoma-associated herpesvirus\u003c/p\u003e\n\u003cp\u003eNS-EM: Negative-staining electron microscopy\u003c/p\u003e\n\u003cp\u003eNTD: N-terminal domain\u003c/p\u003e\n\u003cp\u003eOB: Oligonucleotide-binding\u003c/p\u003e\n\u003cp\u003eORF6\u0026Delta;C: ORF6 deletion mutant, missing the C-terminal 60 residues\u003c/p\u003e\n\u003cp\u003eSSA: single-strand annealing\u003c/p\u003e\n\u003cp\u003eSSB: single-stranded DNA-binding protein\u003c/p\u003e\n\u003cp\u003essDNA: single-stranded DNA\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAcknowledgements\u003c/h2\u003e\u003cp\u003eParts of this research were supported by NHMRC (Ideas scheme, APP1184012 [GNT1184012]), ARC (Australian Research Council Centre of Excellence in Quantum Biotechnology project number CE230100021) and QUBIC Aspire Fellowship (awarded to N.E.S). This research was supported by resources provided by the National Computational Infrastructure (NCI) and the Pawsey Supercomputing Research Centre with funding from the Australian Government and the Government of Western Australia.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eABRAHAM, M. J., MURTOLA, T., SCHULZ, R., P\u0026Aacute;LL, S., SMITH, J. C., HESS, B. \u0026amp; LINDAHL, E. 2015. GROMACS: High performance molecular simulations through multi-level parallelism from laptops to supercomputers. \u003cem\u003eSoftwareX,\u003c/em\u003e 1-2\u003cstrong\u003e,\u003c/strong\u003e 19-25.\u003c/li\u003e\n\u003cli\u003eAMUNDSEN, S. K. \u0026amp; PARRIS, D. S. 1984. Detection of herpes simplex virus intertypic recombinant genomes in infected cell DNA. \u003cem\u003eJ Virol Methods,\u003c/em\u003e 8\u003cstrong\u003e,\u003c/strong\u003e 19-25.\u003c/li\u003e\n\u003cli\u003eANGEL, F., DECAUSSIN, G., DAILLIE, J. \u0026amp; OOKA, T. 1987. Comparison of early polypeptides induced in EBV producer and non-producer lymphoid cell lines. \u003cem\u003eAnnales de l\u0026apos;Institut Pasteur / Virologie,\u003c/em\u003e 138\u003cstrong\u003e,\u003c/strong\u003e 169-181.\u003c/li\u003e\n\u003cli\u003eAULD, D. S. \u0026amp; BERGMAN, T. 2008. Medium- and short-chain dehydrogenase/reductase gene and protein families. \u003cem\u003eCellular and Molecular Life Sciences,\u003c/em\u003e 65\u003cstrong\u003e,\u003c/strong\u003e 3961.\u003c/li\u003e\n\u003cli\u003eBAKER, M. L., JIANG, W., RIXON, F. J. \u0026amp; CHIU, W. 2005. Common ancestry of herpesviruses and tailed DNA bacteriophages. \u003cem\u003eJ Virol,\u003c/em\u003e 79\u003cstrong\u003e,\u003c/strong\u003e 14967-70.\u003c/li\u003e\n\u003cli\u003eBAKKALCI, D., JIA, Y., WINTER, J. R., LEWIS, J. E., TAYLOR, G. S. \u0026amp; STAGG, H. R. 2020. Risk factors for Epstein Barr virus-associated cancers: a systematic review, critical appraisal, and mapping of the epidemiological evidence. \u003cem\u003eJ Glob Health,\u003c/em\u003e 10\u003cstrong\u003e,\u003c/strong\u003e 010405.\u003c/li\u003e\n\u003cli\u003eBAUMFORTH, K. R., YOUNG, L. S., FLAVELL, K. J., CONSTANDINOU, C. \u0026amp; MURRAY, P. G. 1999. The Epstein-Barr virus and its association with human cancers. \u003cem\u003eMol Pathol,\u003c/em\u003e 52\u003cstrong\u003e,\u003c/strong\u003e 307-22.\u003c/li\u003e\n\u003cli\u003eBAYLIS, S. A., PURIFOY, D. J. M. \u0026amp; LITTLER, E. 1989. The characterization of the EBV alkaline deoxyribonuclease cloned and expressed in E.coli. \u003cem\u003eNucleic Acids Research,\u003c/em\u003e 17\u003cstrong\u003e,\u003c/strong\u003e 7609-7622.\u003c/li\u003e\n\u003cli\u003eBLASCHE, S., WUCHTY, S., RAJAGOPALA, S. \u0026amp; UETZ, P. 2013. The Protein Interaction Network of Bacteriophage Lambda with Its Host, Escherichia coli. \u003cem\u003eJournal of virology,\u003c/em\u003e 87.\u003c/li\u003e\n\u003cli\u003eBOEHMER, P. E. 2004. RNA binding and R-loop formation by the herpes simplex virus type-1 single-stranded DNA-binding protein (ICP8). \u003cem\u003eNucleic Acids Res,\u003c/em\u003e 32\u003cstrong\u003e,\u003c/strong\u003e 4576-84.\u003c/li\u003e\n\u003cli\u003eBOEHMER, P. E. \u0026amp; LEHMAN, I. R. 1993. Herpes simplex virus type 1 ICP8: helix-destabilizing properties. \u003cem\u003eJ Virol,\u003c/em\u003e 67\u003cstrong\u003e,\u003c/strong\u003e 711-5.\u003c/li\u003e\n\u003cli\u003eBORTNER, C., HERNANDEZ, T. R., LEHMAN, I. R. \u0026amp; GRIFFITH, J. 1993. Herpes simplex virus 1 single-strand DNA-binding protein (ICP8) will promote homologous pairing and strand transfer. \u003cem\u003eJ Mol Biol,\u003c/em\u003e 231\u003cstrong\u003e,\u003c/strong\u003e 241-50.\u003c/li\u003e\n\u003cli\u003eBOUCKAERT, J., POORTMANS, F., WYNS, L. \u0026amp; LORIS, R. 1996. Sequential structural changes upon zinc and calcium binding to metal-free concanavalin A. \u003cem\u003eJ Biol Chem,\u003c/em\u003e 271\u003cstrong\u003e,\u003c/strong\u003e 16144-50.\u003c/li\u003e\n\u003cli\u003eBREWSTER, J. L. \u0026amp; TOLUN, G. 2020. Half a century of bacteriophage lambda recombinase: In vitro studies of lambda exonuclease and Red-beta annealase. \u003cem\u003eIUBMB Life,\u003c/em\u003e 72\u003cstrong\u003e,\u003c/strong\u003e 1622-1633.\u003c/li\u003e\n\u003cli\u003eBRYANT, K. F., YAN, Z., DREYFUS, D. H. \u0026amp; KNIPE, D. M. 2012. Identification of a divalent metal cation binding site in herpes simplex virus 1 (HSV-1) ICP8 required for HSV replication. \u003cem\u003eJ Virol,\u003c/em\u003e 86\u003cstrong\u003e,\u003c/strong\u003e 6825-34.\u003c/li\u003e\n\u003cli\u003eBUSH, M., YAGER, D. R., GAO, M., WEISSHART, K., MARCY, A. I., COEN, D. M. \u0026amp; KNIPE, D. M. 1991. Correct intranuclear localization of herpes simplex virus DNA polymerase requires the viral ICP8 DNA-binding protein. \u003cem\u003eJ Virol,\u003c/em\u003e 65\u003cstrong\u003e,\u003c/strong\u003e 1082-9.\u003c/li\u003e\n\u003cli\u003eCALDERWOOD, M. A., VENKATESAN, K., XING, L., CHASE, M. R., VAZQUEZ, A., HOLTHAUS, A. M., EWENCE, A. E., LI, N., HIROZANE-KISHIKAWA, T., HILL, D. E., VIDAL, M., KIEFF, E. \u0026amp; JOHANNSEN, E. 2007. Epstein-Barr virus and virus human protein interaction maps. \u003cem\u003eProc Natl Acad Sci U S A,\u003c/em\u003e 104\u003cstrong\u003e,\u003c/strong\u003e 7606-11.\u003c/li\u003e\n\u003cli\u003eCALDWELL, B. J., NORRIS, A. S., KARBOWSKI, C. F., WIEGAND, A. M., WYSOCKI, V. H. \u0026amp; BELL, C. E. 2022. Structure of a RecT/Red\u0026beta; family recombinase in complex with a duplex intermediate of DNA annealing. \u003cem\u003eNature Communications,\u003c/em\u003e 13\u003cstrong\u003e,\u003c/strong\u003e 7855.\u003c/li\u003e\n\u003cli\u003eCERNOOKA, E., RUMNIEKS, J., TARS, K. \u0026amp; KAZAKS, A. 2017. Structural Basis for DNA Recognition of a Single-stranded DNA-binding Protein from Enterobacter Phage Enc34. \u003cem\u003eScientific Reports,\u003c/em\u003e 7\u003cstrong\u003e,\u003c/strong\u003e 15529.\u003c/li\u003e\n\u003cli\u003eCHANG, L. S. \u0026amp; SHENK, T. 1990. The adenovirus DNA-binding protein stimulates the rate of transcription directed by adenovirus and adeno-associated virus promoters. \u003cem\u003eJ Virol,\u003c/em\u003e 64\u003cstrong\u003e,\u003c/strong\u003e 2103-9.\u003c/li\u003e\n\u003cli\u003eCROLL, T. I. 2018. ISOLDE: a physically realistic environment for model building into low-resolution electron-density maps. \u003cem\u003eActa Crystallographica Section D: Biological Crystallography,\u003c/em\u003e 74\u003cstrong\u003e,\u003c/strong\u003e 519-530.\u003c/li\u003e\n\u003cli\u003eCZARNIAK, F. \u0026amp; HENSEL, M. 2015. Red-Mediated Recombineering of Salmonella enterica Genomes. \u003cem\u003eIn:\u003c/em\u003e SCHATTEN, H. \u0026amp; EISENSTARK, A. (eds.) \u003cem\u003eSalmonella: Methods and Protocols.\u003c/em\u003e New York, NY: Springer New York.\u003c/li\u003e\n\u003cli\u003eDARWISH, A. S., GRADY, L. M., BAI, P. \u0026amp; WELLER, S. K. 2016. ICP8 Filament Formation Is Essential for Replication Compartment Formation during Herpes Simplex Virus Infection. \u003cem\u003eJournal of Virology,\u003c/em\u003e 90\u003cstrong\u003e,\u003c/strong\u003e 2561-2570.\u003c/li\u003e\n\u003cli\u003eDAVISON, A. J., EBERLE, R., EHLERS, B., HAYWARD, G. S., MCGEOCH, D. J., MINSON, A. C., PELLETT, P. E., ROIZMAN, B., STUDDERT, M. J. \u0026amp; THIRY, E. 2009. The order Herpesvirales. \u003cem\u003eArch Virol,\u003c/em\u003e 154\u003cstrong\u003e,\u003c/strong\u003e 171-7.\u003c/li\u003e\n\u003cli\u003eDE BRUYN KOPS, A. \u0026amp; KNIPE, D. M. 1988. Formation of DNA replication structures in herpes virus-infected cells requires a viral DNA binding protein. \u003cem\u003eCell,\u003c/em\u003e 55\u003cstrong\u003e,\u003c/strong\u003e 857-68.\u003c/li\u003e\n\u003cli\u003eDECAUSSIN, G., LECLERC, V. \u0026amp; OOKA, T. 1995. The lytic cycle of Epstein-Barr virus in the nonproducer Raji line can be rescued by the expression of a 135-kilodalton protein encoded by the BALF2 open reading frame. \u003cem\u003eJ Virol,\u003c/em\u003e 69\u003cstrong\u003e,\u003c/strong\u003e 7309-14.\u003c/li\u003e\n\u003cli\u003eDUDAS, K. C. \u0026amp; RUYECHAN, W. T. 1998. Identification of a region of the herpes simplex virus single-stranded DNA-binding protein involved in cooperative binding. \u003cem\u003eJ Virol,\u003c/em\u003e 72\u003cstrong\u003e,\u003c/strong\u003e 257-65.\u003c/li\u003e\n\u003cli\u003eD\u0026Uuml;RR, S. L., LEVY, A. \u0026amp; ROTHLISBERGER, U. 2023. Metal3D: a general deep learning framework for accurate metal ion location prediction in proteins. \u003cem\u003eNature Communications,\u003c/em\u003e 14\u003cstrong\u003e,\u003c/strong\u003e 2713.\u003c/li\u003e\n\u003cli\u003eDUTCH, R. E., BIANCHI, V. \u0026amp; LEHMAN, I. R. 1995. Herpes simplex virus type 1 DNA replication is specifically required for high-frequency homologous recombination between repeated sequences. \u003cem\u003eJ Virol,\u003c/em\u003e 69\u003cstrong\u003e,\u003c/strong\u003e 3084-9.\u003c/li\u003e\n\u003cli\u003eDUTCH, R. E. \u0026amp; LEHMAN, I. R. 1993. Renaturation of complementary DNA strands by herpes simplex virus type 1 ICP8. \u003cem\u003eJournal of Virology,\u003c/em\u003e 67\u003cstrong\u003e,\u003c/strong\u003e 6945-6949.\u003c/li\u003e\n\u003cli\u003eFEEDERLE, R., BANNERT, H., LIPS, H., M\u0026Uuml;LLER-LANTZSCH, N. \u0026amp; DELECLUSE, H. J. 2009. The Epstein-Barr virus alkaline exonuclease BGLF5 serves pleiotropic functions in virus replication. \u003cem\u003eJ Virol,\u003c/em\u003e 83\u003cstrong\u003e,\u003c/strong\u003e 4952-62.\u003c/li\u003e\n\u003cli\u003eFELS, U., GEVAERT, K. \u0026amp; VAN DAMME, P. 2020. Bacterial Genetic Engineering by Means of Recombineering for Reverse Genetics. \u003cem\u003eFrontiers in Microbiology,\u003c/em\u003e 11.\u003c/li\u003e\n\u003cli\u003eFITSCHEN, L. J., NEWING, T. P., JOHNSTON, N. P., BELL, C. E. \u0026amp; TOLUN, G. 2024. Half a century after their discovery: Structural insights into exonuclease and annealase proteins catalyzing recombineering. \u003cem\u003eEngineering Microbiology,\u003c/em\u003e 4\u003cstrong\u003e,\u003c/strong\u003e 100120.\u003c/li\u003e\n\u003cli\u003eFIXMAN, E. D., HAYWARD, G. S. \u0026amp; HAYWARD, S. D. 1992. trans-acting requirements for replication of Epstein-Barr virus ori-Lyt. \u003cem\u003eJ Virol,\u003c/em\u003e 66\u003cstrong\u003e,\u003c/strong\u003e 5030-9.\u003c/li\u003e\n\u003cli\u003eFIXMAN, E. D., HAYWARD, G. S. \u0026amp; HAYWARD, S. D. 1995. Replication of Epstein-Barr virus oriLyt: lack of a dedicated virally encoded origin-binding protein and dependence on Zta in cotransfection assays. \u003cem\u003eJ Virol,\u003c/em\u003e 69\u003cstrong\u003e,\u003c/strong\u003e 2998-3006.\u003c/li\u003e\n\u003cli\u003eFOX, D. G., CARY, P. D. \u0026amp; KNEALE, G. G. 1999. Conformational studies of the C-terminal domain of bacteriophage Pf1 gene 5 protein. \u003cem\u003eBiochimica et Biophysica Acta (BBA) - Protein Structure and Molecular Enzymology,\u003c/em\u003e 1435\u003cstrong\u003e,\u003c/strong\u003e 138-146.\u003c/li\u003e\n\u003cli\u003eFU, X., WANG, H. \u0026amp; ZHANG, X. 2002. High-frequency intermolecular homologous recombination during herpes simplex virus-mediated plasmid DNA replication. \u003cem\u003eJ Virol,\u003c/em\u003e 76\u003cstrong\u003e,\u003c/strong\u003e 5866-74.\u003c/li\u003e\n\u003cli\u003eFUJII, K., YOKOYAMA, N., KIYONO, T., KUZUSHIMA, K., HOMMA, M., NISHIYAMA, Y., FUJITA, M. \u0026amp; TSURUMI, T. 2000. The Epstein-Barr virus pol catalytic subunit physically interacts with the BBLF4-BSLF1-BBLF2/3 complex. \u003cem\u003eJ Virol,\u003c/em\u003e 74\u003cstrong\u003e,\u003c/strong\u003e 2550-7.\u003c/li\u003e\n\u003cli\u003eGAO, M., BOUCHEY, J., CURTIN, K. \u0026amp; KNIPE, D. M. 1988. Genetic identification of a portion of the herpes simplex virus ICP8 protein required for DNA-binding. \u003cem\u003eVirology,\u003c/em\u003e 163\u003cstrong\u003e,\u003c/strong\u003e 319-29.\u003c/li\u003e\n\u003cli\u003eGAO, M. \u0026amp; KNIPE, D. M. 1989. Genetic evidence for multiple nuclear functions of the herpes simplex virus ICP8 DNA-binding protein. \u003cem\u003eJ Virol,\u003c/em\u003e 63\u003cstrong\u003e,\u003c/strong\u003e 5258-67.\u003c/li\u003e\n\u003cli\u003eGAO, Z., KRITHIVAS, A., FINAN, J. E., SEMMES, O. J., ZHOU, S., WANG, Y. \u0026amp; HAYWARD, S. D. 1998. The Epstein-Barr virus lytic transactivator Zta interacts with the helicase-primase replication proteins. \u003cem\u003eJ Virol,\u003c/em\u003e 72\u003cstrong\u003e,\u003c/strong\u003e 8559-67.\u003c/li\u003e\n\u003cli\u003eGHOSH, S., HAMDAN, S. M. \u0026amp; RICHARDSON, C. C. 2010. Two modes of interaction of the single-stranded DNA-binding protein of bacteriophage T7 with the DNA polymerase-thioredoxin complex. \u003cem\u003eJ Biol Chem,\u003c/em\u003e 285\u003cstrong\u003e,\u003c/strong\u003e 18103-12.\u003c/li\u003e\n\u003cli\u003eGOLDSTEIN, J. N. \u0026amp; WELLER, S. K. 1998. In vitro processing of herpes simplex virus type 1 DNA replication intermediates by the viral alkaline nuclease, UL12. \u003cem\u003eJ Virol,\u003c/em\u003e 72\u003cstrong\u003e,\u003c/strong\u003e 8772-81.\u003c/li\u003e\n\u003cli\u003eGOURVES, A.-S., LE GAC, N. T., VILLANI, G., BOEHMER, P. E. \u0026amp; JOHNSON, N. P. 2000. Equilibrium Binding of Single-stranded DNA with Herpes Simplex Virus Type I-coded Single-stranded DNA-binding Protein, ICP8*. \u003cem\u003eJournal of Biological Chemistry,\u003c/em\u003e 275\u003cstrong\u003e,\u003c/strong\u003e 10864-10869.\u003c/li\u003e\n\u003cli\u003eGRADY, L. M., SZCZEPANIAK, R., MURELLI, R. P., MASAOKA, T., LE GRICE, S. F. J., WRIGHT, D. L. \u0026amp; WELLER, S. K. 2017. The Exonuclease Activity of Herpes Simplex Virus 1 UL12 Is Required for Production of Viral DNA That Can Be Packaged To Produce Infectious Virus. \u003cem\u003eJ Virol,\u003c/em\u003e 91.\u003c/li\u003e\n\u003cli\u003eGUINDON, S., DUFAYARD, J.-F., LEFORT, V., ANISIMOVA, M., HORDIJK, W. \u0026amp; GASCUEL, O. 2010. New Algorithms and Methods to Estimate Maximum-Likelihood Phylogenies: Assessing the Performance of PhyML 3.0. \u003cem\u003eSystematic Biology,\u003c/em\u003e 59\u003cstrong\u003e,\u003c/strong\u003e 307-321.\u003c/li\u003e\n\u003cli\u003eGUPTE, S. S., OLSON, J. W. \u0026amp; RUYECHAN, W. T. 1991. The major herpes simplex virus type-1 DNA-binding protein is a zinc metalloprotein. \u003cem\u003eJournal of Biological Chemistry,\u003c/em\u003e 266\u003cstrong\u003e,\u003c/strong\u003e 11413-11416.\u003c/li\u003e\n\u003cli\u003eHAMMERSCHMIDT, W. \u0026amp; SUGDEN, B. 1988. Identification and characterization of oriLyt, a lytic origin of DNA replication of Epstein-Barr virus. \u003cem\u003eCell,\u003c/em\u003e 55\u003cstrong\u003e,\u003c/strong\u003e 427-33.\u003c/li\u003e\n\u003cli\u003eHAMMERSCHMIDT, W. \u0026amp; SUGDEN, B. 2013. Replication of Epstein-Barr viral DNA. \u003cem\u003eCold Spring Harb Perspect Biol,\u003c/em\u003e 5\u003cstrong\u003e,\u003c/strong\u003e a013029.\u003c/li\u003e\n\u003cli\u003eHARA, Y., WATANABE, T., YOSHIDA, M., MASUD, H. M. A. A., KATO, H., KONDO, T., SUZUKI, R., KUROSE, S., UDDIN, M. K., ARATA, M., MIYAGI, S., YANAGI, Y., SATO, Y., KIMURA, H. \u0026amp; MURATA, T. 2022. Comprehensive Analyses of Intraviral Epstein-Barr Virus Protein\u0026ndash;Protein Interactions Hint Central Role of BLRF2 in the Tegument Network. \u003cem\u003eJournal of Virology,\u003c/em\u003e 96\u003cstrong\u003e,\u003c/strong\u003e e00518-22.\u003c/li\u003e\n\u003cli\u003eHERNANDEZ, A. J. \u0026amp; RICHARDSON, C. C. 2019. Gp2.5, the multifunctional bacteriophage T7 single-stranded DNA binding protein. \u003cem\u003eSemin Cell Dev Biol,\u003c/em\u003e 86\u003cstrong\u003e,\u003c/strong\u003e 92-101.\u003c/li\u003e\n\u003cli\u003eHERNANDEZ, T. R. \u0026amp; LEHMAN, I. R. 1990. Functional interaction between the herpes simplex-1 DNA polymerase and UL42 protein. \u003cem\u003eJ Biol Chem,\u003c/em\u003e 265\u003cstrong\u003e,\u003c/strong\u003e 11227-32.\u003c/li\u003e\n\u003cli\u003eHESS, B., BEKKER, H., BERENDSEN, H. J. C. \u0026amp; FRAAIJE, J. G. E. M. 1997. LINCS: A linear constraint solver for molecular simulations. \u003cem\u003eJournal of Computational Chemistry,\u003c/em\u003e 18\u003cstrong\u003e,\u003c/strong\u003e 1463-1472.\u003c/li\u003e\n\u003cli\u003eHOLLIS, T., STATTEL, J. M., WALTHER, D. S., RICHARDSON, C. C. \u0026amp; ELLENBERGER, T. 2001. Structure of the gene 2.5 protein, a single-stranded DNA binding protein encoded by bacteriophage T7. \u003cem\u003eProc Natl Acad Sci U S A,\u003c/em\u003e 98\u003cstrong\u003e,\u003c/strong\u003e 9557-62.\u003c/li\u003e\n\u003cli\u003eHOLM, L., LAIHO, A., T\u0026Ouml;R\u0026Ouml;NEN, P. \u0026amp; SALGADO, M. 2023. DALI shines a light on remote homologs: One hundred discoveries. \u003cem\u003eProtein Science,\u003c/em\u003e 32\u003cstrong\u003e,\u003c/strong\u003e e4519.\u003c/li\u003e\n\u003cli\u003eHOOVER, W. G. 1985. Canonical dynamics: Equilibrium phase-space distributions. \u003cem\u003ePhysical review A,\u003c/em\u003e 31\u003cstrong\u003e,\u003c/strong\u003e 1695.\u003c/li\u003e\n\u003cli\u003eHUMPHREY, W., DALKE, A. \u0026amp; SCHULTEN, K. 1996. VMD: visual molecular dynamics. \u003cem\u003eJournal of molecular graphics,\u003c/em\u003e 14\u003cstrong\u003e,\u003c/strong\u003e 33-38.\u003c/li\u003e\n\u003cli\u003eHYLAND, E. M., REZENDE, L. F. \u0026amp; RICHARDSON, C. C. 2003. The DNA Binding Domain of the Gene 2.5 Single-stranded DNA-binding Protein of Bacteriophage T7*. \u003cem\u003eJournal of Biological Chemistry,\u003c/em\u003e 278\u003cstrong\u003e,\u003c/strong\u003e 7247-7256.\u003c/li\u003e\n\u003cli\u003eIVANI, I., DANS, P. D., NOY, A., P\u0026Eacute;REZ, A., FAUSTINO, I., HOSPITAL, A., WALTHER, J., ANDRIO, P., GO\u0026Ntilde;I, R., BALACEANU, A., PORTELLA, G., BATTISTINI, F., GELP\u0026Iacute;, J. L., GONZ\u0026Aacute;LEZ, C., VENDRUSCOLO, M., LAUGHTON, C. A., HARRIS, S. A., CASE, D. A. \u0026amp; OROZCO, M. 2016. Parmbsc1: a refined force field for DNA simulations. \u003cem\u003eNature Methods,\u003c/em\u003e 13\u003cstrong\u003e,\u003c/strong\u003e 55-58.\u003c/li\u003e\n\u003cli\u003eIYER, L. M., KOONIN, E. V. \u0026amp; ARAVIND, L. 2002. Classification and evolutionary history of the single-strand annealing proteins, RecT, Redbeta, ERF and RAD52. \u003cem\u003eBMC Genomics,\u003c/em\u003e 3\u003cstrong\u003e,\u003c/strong\u003e 8.\u003c/li\u003e\n\u003cli\u003eJACOB, R. J. \u0026amp; ROIZMAN, B. 1977. Anatomy of herpes simplex virus DNA VIII. Properties of the replicating DNA. \u003cem\u003eJ Virol,\u003c/em\u003e 23\u003cstrong\u003e,\u003c/strong\u003e 394-411.\u003c/li\u003e\n\u003cli\u003eJO, S., KIM, T., IYER, V. G. \u0026amp; IM, W. 2008. CHARMM-GUI: A web-based graphical user interface for CHARMM. \u003cem\u003eJournal of Computational Chemistry,\u003c/em\u003e 29\u003cstrong\u003e,\u003c/strong\u003e 1859-1865.\u003c/li\u003e\n\u003cli\u003eJORGENSEN, W. L., CHANDRASEKHAR, J., MADURA, J. D., IMPEY, R. W. \u0026amp; KLEIN, M. L. 1983. Comparison of simple potential functions for simulating liquid water. \u003cem\u003eThe Journal of chemical physics,\u003c/em\u003e 79\u003cstrong\u003e,\u003c/strong\u003e 926-935.\u003c/li\u003e\n\u003cli\u003eKARPUSAS, M., NOLTE, M., BENTON, C. B., MEIER, W., LIPSCOMB, W. N. \u0026amp; GOELZ, S. 1997. The crystal structure of human interferon \u0026beta; at 2.2-\u0026Aring;\u0026thinsp;resolution. \u003cem\u003eProceedings of the National Academy of Sciences,\u003c/em\u003e 94\u003cstrong\u003e,\u003c/strong\u003e 11813-11818.\u003c/li\u003e\n\u003cli\u003eKATOH, K. \u0026amp; STANDLEY, D. M. 2013. MAFFT Multiple Sequence Alignment Software Version 7: Improvements in Performance and Usability. \u003cem\u003eMolecular Biology and Evolution,\u003c/em\u003e 30\u003cstrong\u003e,\u003c/strong\u003e 772-780.\u003c/li\u003e\n\u003cli\u003eKAZLAUSKAS, D. \u0026amp; VENCLOVAS, C. 2012. Two distinct SSB protein families in nucleo-cytoplasmic large DNA viruses. \u003cem\u003eBioinformatics,\u003c/em\u003e 28\u003cstrong\u003e,\u003c/strong\u003e 3186-90.\u003c/li\u003e\n\u003cli\u003eKIM, Y. T. \u0026amp; RICHARDSON, C. C. 1993. Bacteriophage T7 gene 2.5 protein: an essential protein for DNA replication. \u003cem\u003eProc Natl Acad Sci U S A,\u003c/em\u003e 90\u003cstrong\u003e,\u003c/strong\u003e 10173-7.\u003c/li\u003e\n\u003cli\u003eKIM, Y. T. \u0026amp; RICHARDSON, C. C. 1994. Acidic carboxyl-terminal domain of gene 2.5 protein of bacteriophage T7 is essential for protein-protein interactions. \u003cem\u003eJ Biol Chem,\u003c/em\u003e 269\u003cstrong\u003e,\u003c/strong\u003e 5270-8.\u003c/li\u003e\n\u003cli\u003eKIM, Y. T., TABOR, S., CHURCHICH, J. E. \u0026amp; RICHARDSON, C. C. 1992. Interactions of gene 2.5 protein and DNA polymerase of bacteriophage T7. \u003cem\u003eJ Biol Chem,\u003c/em\u003e 267\u003cstrong\u003e,\u003c/strong\u003e 15032-40.\u003c/li\u003e\n\u003cli\u003eKINOSHITA, C., TAKIZAWA, Y., SAOTOME, M., OGINO, S., KURUMIZAKA, H. \u0026amp; KAGAWA, W. 2023. The cryo-EM structure of full-length RAD52 protein contains an undecameric ring. \u003cem\u003eFEBS Open Bio,\u003c/em\u003e 13\u003cstrong\u003e,\u003c/strong\u003e 408-418.\u003c/li\u003e\n\u003cli\u003eKONG, D., NOSSAL, N. G. \u0026amp; RICHARDSON, C. C. 1997. Role of the bacteriophage T7 and T4 single-stranded DNA-binding proteins in the formation of joint molecules and DNA helicase-catalyzed polar branch migration. \u003cem\u003eJournal of Biological Chemistry,\u003c/em\u003e 272\u003cstrong\u003e,\u003c/strong\u003e 8380-8387.\u003c/li\u003e\n\u003cli\u003eKOWALCZYKOWSKI, S. C., DIXON, D. A., EGGLESTON, A. K., LAUDER, S. D. \u0026amp; REHRAUER, W. M. 1994. Biochemistry of homologous recombination in Escherichia coli. \u003cem\u003eMicrobiol Rev,\u003c/em\u003e 58\u003cstrong\u003e,\u003c/strong\u003e 401-65.\u003c/li\u003e\n\u003cli\u003eLEE, C. K. \u0026amp; KNIPE, D. M. 1985. An immunoassay for the study of DNA-binding activities of herpes simplex virus protein ICP8. \u003cem\u003eJ Virol,\u003c/em\u003e 54\u003cstrong\u003e,\u003c/strong\u003e 731-8.\u003c/li\u003e\n\u003cli\u003eLEE, J., HITZENBERGER, M., RIEGER, M., KERN, N. R., ZACHARIAS, M. \u0026amp; IM, W. 2020. CHARMM-GUI supports the Amber force fields. \u003cem\u003eThe Journal of Chemical Physics,\u003c/em\u003e 153.\u003c/li\u003e\n\u003cli\u003eLEE, Y.-M. \u0026amp; LIM, C. 2008. Physical Basis of Structural and Catalytic Zn-Binding Sites in Proteins. \u003cem\u003eJournal of Molecular Biology,\u003c/em\u003e 379\u003cstrong\u003e,\u003c/strong\u003e 545-553.\u003c/li\u003e\n\u003cli\u003eLEINBACH, S. S. \u0026amp; HEATH, L. S. 1989. Characterization of the single-stranded DNA-binding domain of the herpes simplex virus protein ICP8. \u003cem\u003eBiochimica et Biophysica Acta (BBA) - Gene Structure and Expression,\u003c/em\u003e 1008\u003cstrong\u003e,\u003c/strong\u003e 281-286.\u003c/li\u003e\n\u003cli\u003eLIEBSCHNER, D., AFONINE, P. V., BAKER, M. L., BUNKOCZI, G., CHEN, V. B., CROLL, T. I., HINTZE, B., HUNG, L.-W., JAIN, S., MCCOY, A. J., MORIARTY, N. W., OEFFNER, R. D., POON, B. K., PRISANT, M. G., READ, R. J., RICHARDSON, J. S., RICHARDSON, D. C., SAMMITO, M. D., SOBOLEV, O. V., STOCKWELL, D. H., TERWILLIGER, T. C., URZHUMTSEV, A. G., VIDEAU, L. L., WILLIAMS, C. J. \u0026amp; ADAMS, P. D. 2019. Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix. \u003cem\u003eActa Crystallographica Section D,\u003c/em\u003e 75\u003cstrong\u003e,\u003c/strong\u003e 861-877.\u003c/li\u003e\n\u003cli\u003eLIN, F. L., SPERLE, K. \u0026amp; STERNBERG, N. 1984. Model for homologous recombination during transfer of DNA into mouse L cells: role for DNA ends in the recombination process. \u003cem\u003eMol Cell Biol,\u003c/em\u003e 4\u003cstrong\u003e,\u003c/strong\u003e 1020-34.\u003c/li\u003e\n\u003cli\u003eLIN, S. F., HSU, T. Y., LIU, M. Y., LIN, L. S., YANG, H. L., CHEN, J. Y. \u0026amp; YANG, C. S. 1995. Characterization of Epstein-Barr virus DNase and its interaction with the major DNA binding protein. \u003cem\u003eVirology,\u003c/em\u003e 208\u003cstrong\u003e,\u003c/strong\u003e 712-22.\u003c/li\u003e\n\u003cli\u003eLIN, S. F., LIN, S. W., HSU, T. Y., LIU, M. Y., CHEN, J. Y. \u0026amp; YANG, C. S. 1994. Functional analysis of the amino terminus of Epstein-Barr virus deoxyribonuclease. \u003cem\u003eVirology,\u003c/em\u003e 199\u003cstrong\u003e,\u003c/strong\u003e 223-7.\u003c/li\u003e\n\u003cli\u003eLIN, Y.-F., CHENG, C.-W., SHIH, C.-S., HWANG, J.-K., YU, C.-S. \u0026amp; LU, C.-H. 2016. MIB: Metal Ion-Binding Site Prediction and Docking Server. \u003cem\u003eJournal of Chemical Information and Modeling,\u003c/em\u003e 56\u003cstrong\u003e,\u003c/strong\u003e 2287-2291.\u003c/li\u003e\n\u003cli\u003eLOPES, A., AMARIR-BOUHRAM, J., FAURE, G., PETIT, M. A. \u0026amp; GUEROIS, R. 2010. Detection of novel recombinases in bacteriophage genomes unveils Rad52, Rad51 and Gp2.5 remote homologs. \u003cem\u003eNucleic Acids Res,\u003c/em\u003e 38\u003cstrong\u003e,\u003c/strong\u003e 3952-62.\u003c/li\u003e\n\u003cli\u003eMAIER, J. A., MARTINEZ, C., KASAVAJHALA, K., WICKSTROM, L., HAUSER, K. E. \u0026amp; SIMMERLING, C. 2015. ff14SB: Improving the Accuracy of Protein Side Chain and Backbone Parameters from ff99SB. \u003cem\u003eJournal of Chemical Theory and Computation,\u003c/em\u003e 11\u003cstrong\u003e,\u003c/strong\u003e 3696-3713.\u003c/li\u003e\n\u003cli\u003eMAKHOV, A. M. \u0026amp; GRIFFITH, J. D. 2006. Visualization of the annealing of complementary single-stranded DNA catalyzed by the herpes simplex virus type 1 ICP8 SSB/recombinase. \u003cem\u003eJ Mol Biol,\u003c/em\u003e 355\u003cstrong\u003e,\u003c/strong\u003e 911-22.\u003c/li\u003e\n\u003cli\u003eMAKHOV, A. M., SEN, A., YU, X., SIMON, M. N., GRIFFITH, J. D. \u0026amp; EGELMAN, E. H. 2009. The bipolar filaments formed by herpes simplex virus type 1 SSB/recombination protein (ICP8) suggest a mechanism for DNA annealing. \u003cem\u003eJ Mol Biol,\u003c/em\u003e 386\u003cstrong\u003e,\u003c/strong\u003e 273-9.\u003c/li\u003e\n\u003cli\u003eMAPELLI, M., M\u0026Uuml;HLEISEN, M., PERSICO, G., VAN DER ZANDT, H. \u0026amp; TUCKER, P. A. 2000. The 60-residue C-terminal region of the single-stranded DNA binding protein of herpes simplex virus type 1 is required for cooperative DNA binding. \u003cem\u003eJ Virol,\u003c/em\u003e 74\u003cstrong\u003e,\u003c/strong\u003e 8812-22.\u003c/li\u003e\n\u003cli\u003eMAPELLI, M., PANJIKAR, S. \u0026amp; TUCKER, P. A. 2005. The Crystal Structure of the Herpes Simplex Virus 1 ssDNA-binding Protein Suggests the Structural Basis for Flexible, Cooperative Single-stranded DNA Binding*. \u003cem\u003eJournal of Biological Chemistry,\u003c/em\u003e 280\u003cstrong\u003e,\u003c/strong\u003e 2990-2997.\u003c/li\u003e\n\u003cli\u003eMARESCA, M., ERLER, A., FU, J., FRIEDRICH, A., ZHANG, Y. \u0026amp; STEWART, A. 2010. Single-stranded heteroduplex intermediates in \u0026lambda; Red homologous recombination. \u003cem\u003eBMC Molecular Biology,\u003c/em\u003e 11\u003cstrong\u003e,\u003c/strong\u003e 54.\u003c/li\u003e\n\u003cli\u003eMARINTCHEVA, B., MARINTCHEV, A., WAGNER, G. \u0026amp; RICHARDSON, C. C. 2008. Acidic C-terminal tail of the ssDNA-binding protein of bacteriophage T7 and ssDNA compete for the same binding surface. \u003cem\u003eProc Natl Acad Sci U S A,\u003c/em\u003e 105\u003cstrong\u003e,\u003c/strong\u003e 1855-60.\u003c/li\u003e\n\u003cli\u003eMARTINEZ, R., SARISKY, R. T., WEBER, P. C. \u0026amp; WELLER, S. K. 1996. Herpes simplex virus type 1 alkaline nuclease is required for efficient processing of viral DNA replication intermediates. \u003cem\u003eJ Virol,\u003c/em\u003e 70\u003cstrong\u003e,\u003c/strong\u003e 2075-85.\u003c/li\u003e\n\u003cli\u003eMICHAUD‐AGRAWAL, N., DENNING, E. J., WOOLF, T. B. \u0026amp; BECKSTEIN, O. 2011. MDAnalysis: a toolkit for the analysis of molecular dynamics simulations. \u003cem\u003eJournal of computational chemistry,\u003c/em\u003e 32\u003cstrong\u003e,\u003c/strong\u003e 2319-2327.\u003c/li\u003e\n\u003cli\u003eMINH, B. Q., NGUYEN, M. A. T. \u0026amp; VON HAESELER, A. 2013. Ultrafast Approximation for Phylogenetic Bootstrap. \u003cem\u003eMolecular Biology and Evolution,\u003c/em\u003e 30\u003cstrong\u003e,\u003c/strong\u003e 1188-1195.\u003c/li\u003e\n\u003cli\u003eMOCARSKI, E. S. \u0026amp; ROIZMAN, B. 1982. Structure and role of the herpes simplex virus DNA termini in inversion, circularization and generation of virion DNA. \u003cem\u003eCell,\u003c/em\u003e 31\u003cstrong\u003e,\u003c/strong\u003e 89-97.\u003c/li\u003e\n\u003cli\u003eMOSBERG, J. A., LAJOIE, M. J. \u0026amp; CHURCH, G. M. 2010. Lambda Red Recombineering in Escherichia coli Occurs Through a Fully Single-Stranded Intermediate. \u003cem\u003eGenetics,\u003c/em\u003e 186\u003cstrong\u003e,\u003c/strong\u003e 791-799.\u003c/li\u003e\n\u003cli\u003eMUMTSIDU, E., MAKHOV, A. M., KONAREV, P. V., SVERGUN, D. I., GRIFFITH, J. D. \u0026amp; TUCKER, P. A. 2008. Structural features of the single-stranded DNA-binding protein of Epstein\u0026ndash;Barr virus. \u003cem\u003eJournal of Structural Biology,\u003c/em\u003e 161\u003cstrong\u003e,\u003c/strong\u003e 172-187.\u003c/li\u003e\n\u003cli\u003eNAUFER, M. N., MORSE, M., M\u0026Ouml;LLER, G. B., MCISAAC, J., ROUZINA, I., BEUNING, P. J. \u0026amp; WILLIAMS, M. C. 2021. Multiprotein E. coli SSB\u0026ndash;ssDNA complex shows both stable binding and rapid dissociation due to interprotein interactions. \u003cem\u003eNucleic Acids Research,\u003c/em\u003e 49\u003cstrong\u003e,\u003c/strong\u003e 1532-1549.\u003c/li\u003e\n\u003cli\u003eNEWING, T. P., BREWSTER, J. L., FITSCHEN, L. J., BOUWER, J. C., JOHNSTON, N. P., YU, H. \u0026amp; TOLUN, G. 2022. Red\u0026beta;177 annealase structure reveals details of oligomerization and \u0026lambda; Red-mediated homologous DNA recombination. \u003cem\u003eNature Communications,\u003c/em\u003e 13\u003cstrong\u003e,\u003c/strong\u003e 5649.\u003c/li\u003e\n\u003cli\u003eNGUYEN, L. T., LE, T. N., DE PROFT, F., CHANDRA, A. K., LANGENAEKER, W., NGUYEN, M. T. \u0026amp; GEERLINGS, P. 1999. Mechanism of [2+1] cycloadditions of hydrogen isocyanide to alkynes: Molecular orbital and density functional theory study. \u003cem\u003eJournal of the American Chemical Society,\u003c/em\u003e 121\u003cstrong\u003e,\u003c/strong\u003e 5992-6001.\u003c/li\u003e\n\u003cli\u003eNICHOLAS, J., RUVOLO, V., ZONG, J., CIUFO, D., GUO, H. G., REITZ, M. S. \u0026amp; HAYWARD, G. S. 1997. A single 13-kilobase divergent locus in the Kaposi sarcoma-associated herpesvirus (human herpesvirus 8) genome contains nine open reading frames that are homologous to or related to cellular proteins. \u003cem\u003eJ Virol,\u003c/em\u003e 71\u003cstrong\u003e,\u003c/strong\u003e 1963-74.\u003c/li\u003e\n\u003cli\u003eNOS\u0026Eacute;, S. 1984. A molecular dynamics method for simulations in the canonical ensemble. \u003cem\u003eMolecular physics,\u003c/em\u003e 52\u003cstrong\u003e,\u003c/strong\u003e 255-268.\u003c/li\u003e\n\u003cli\u003eO\u0026apos;DONNELL, M. E., ELIAS, P., FUNNELL, B. E. \u0026amp; LEHMAN, I. R. 1987. Interaction between the DNA polymerase and single-stranded DNA-binding protein (infected cell protein 8) of herpes simplex virus 1. \u003cem\u003eJ Biol Chem,\u003c/em\u003e 262\u003cstrong\u003e,\u003c/strong\u003e 4260-6.\u003c/li\u003e\n\u003cli\u003eOLSSON, M. H. M., S\u0026Oslash;NDERGAARD, C. R., ROSTKOWSKI, M. \u0026amp; JENSEN, J. H. 2011. PROPKA3: Consistent Treatment of Internal and Surface Residues in Empirical pKa Predictions. \u003cem\u003eJournal of Chemical Theory and Computation,\u003c/em\u003e 7\u003cstrong\u003e,\u003c/strong\u003e 525-537.\u003c/li\u003e\n\u003cli\u003eOZGUR, S., DAMANIA, B. \u0026amp; GRIFFITH, J. 2011. The Kaposi\u0026rsquo;s sarcoma-associated herpesvirus ORF6 DNA binding protein forms long DNA-free helical protein filaments. \u003cem\u003eJournal of Structural Biology,\u003c/em\u003e 174\u003cstrong\u003e,\u003c/strong\u003e 37-43.\u003c/li\u003e\n\u003cli\u003eOZGUR, S. \u0026amp; GRIFFITH, J. 2014. Interaction of Kaposi\u0026apos;s sarcoma-associated herpesvirus ORF6 protein with single-stranded DNA. \u003cem\u003eJ Virol,\u003c/em\u003e 88\u003cstrong\u003e,\u003c/strong\u003e 8687-95.\u003c/li\u003e\n\u003cli\u003ePARRINELLO, M. \u0026amp; RAHMAN, A. 1981. Polymorphic transitions in single crystals: A new molecular dynamics method. \u003cem\u003eJournal of Applied physics,\u003c/em\u003e 52\u003cstrong\u003e,\u003c/strong\u003e 7182-7190.\u003c/li\u003e\n\u003cli\u003ePEI, J. \u0026amp; GRISHIN, N. V. 2001. AL2CO: calculation of positional conservation in a protein sequence alignment. \u003cem\u003eBioinformatics,\u003c/em\u003e 17\u003cstrong\u003e,\u003c/strong\u003e 700-12.\u003c/li\u003e\n\u003cli\u003ePETTERSEN, E. F., GODDARD, T. D., HUANG, C. C., MENG, E. C., COUCH, G. S., CROLL, T. I., MORRIS, J. H. \u0026amp; FERRIN, T. E. 2021. UCSF ChimeraX: Structure visualization for researchers, educators, and developers. \u003cem\u003eProtein Sci,\u003c/em\u003e 30\u003cstrong\u003e,\u003c/strong\u003e 70-82.\u003c/li\u003e\n\u003cli\u003eQUINN, J. P. \u0026amp; MCGEOCH, D. J. 1985. DNA sequence of the region in the genome of herpes simplex virus type 1 containing the genes for DNA polymerase and the major DNA binding protein. \u003cem\u003eNucleic Acids Res,\u003c/em\u003e 13\u003cstrong\u003e,\u003c/strong\u003e 8143-63.\u003c/li\u003e\n\u003cli\u003eREUBEN, R. C. \u0026amp; GEFTER, M. L. 1973. A DNA-binding protein induced by bacteriophage T7. \u003cem\u003eProc Natl Acad Sci U S A,\u003c/em\u003e 70\u003cstrong\u003e,\u003c/strong\u003e 1846-50.\u003c/li\u003e\n\u003cli\u003eREUBEN, R. C. \u0026amp; GEFTER, M. L. 1974. A Deoxyribonucleic Acid-Binding Protein Induced by Bacteriophage T7: PURIFICATION AND PROPERTIES OF THE PROTEIN. \u003cem\u003eJournal of Biological Chemistry,\u003c/em\u003e 249\u003cstrong\u003e,\u003c/strong\u003e 3843-3850.\u003c/li\u003e\n\u003cli\u003eREUVEN, N. B., STAIRE, A. E., MYERS, R. S. \u0026amp; WELLER, S. K. 2003. The herpes simplex virus type 1 alkaline nuclease and single-stranded DNA binding protein mediate strand exchange in vitro. \u003cem\u003eJ Virol,\u003c/em\u003e 77\u003cstrong\u003e,\u003c/strong\u003e 7425-33.\u003c/li\u003e\n\u003cli\u003eREUVEN, N. B. \u0026amp; WELLER, S. K. 2005. Herpes simplex virus type 1 single-strand DNA binding protein ICP8 enhances the nuclease activity of the UL12 alkaline nuclease by increasing its processivity. \u003cem\u003eJ Virol,\u003c/em\u003e 79\u003cstrong\u003e,\u003c/strong\u003e 9356-8.\u003c/li\u003e\n\u003cli\u003eREUVEN, N. B., WILLCOX, S., GRIFFITH, J. D. \u0026amp; WELLER, S. K. 2004. 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. \u003cem\u003eJ Mol Biol,\u003c/em\u003e 342\u003cstrong\u003e,\u003c/strong\u003e 57-71.\u003c/li\u003e\n\u003cli\u003eREZENDE, L. F., HOLLIS, T., ELLENBERGER, T. \u0026amp; RICHARDSON, C. C. 2002. Essential amino acid residues in the single-stranded DNA-binding protein of bacteriophage T7. Identification of the dimer interface. \u003cem\u003eJ Biol Chem,\u003c/em\u003e 277\u003cstrong\u003e,\u003c/strong\u003e 50643-53.\u003c/li\u003e\n\u003cli\u003eREZENDE, L. F., WILLCOX, S., GRIFFITH, J. D. \u0026amp; RICHARDSON, C. C. 2003. A single-stranded DNA-binding protein of bacteriophage T7 defective in DNA annealing. \u003cem\u003eJ Biol Chem,\u003c/em\u003e 278\u003cstrong\u003e,\u003c/strong\u003e 29098-105.\u003c/li\u003e\n\u003cli\u003eRUYECHAN, W. T. \u0026amp; WEIR, A. C. 1984. Interaction with nucleic acids and stimulation of the viral DNA polymerase by the herpes simplex virus type 1 major DNA-binding protein. \u003cem\u003eJ Virol,\u003c/em\u003e 52\u003cstrong\u003e,\u003c/strong\u003e 727-33.\u003c/li\u003e\n\u003cli\u003eSALI, A. \u0026amp; BLUNDELL, T. L. 1993. Comparative protein modelling by satisfaction of spatial restraints. \u003cem\u003eJ Mol Biol,\u003c/em\u003e 234\u003cstrong\u003e,\u003c/strong\u003e 779-815.\u003c/li\u003e\n\u003cli\u003eS\u0026Aacute;NCHEZ-APARICIO, J.-E., TIESSLER-SALA, L., VELASCO-CARNEROS, L., ROLD\u0026Aacute;N-MART\u0026Iacute;N, L., SCIORTINO, G. \u0026amp; MAR\u0026Eacute;CHAL, J.-D. 2021. BioMetAll: Identifying Metal-Binding Sites in Proteins from Backbone Preorganization. \u003cem\u003eJournal of Chemical Information and Modeling,\u003c/em\u003e 61\u003cstrong\u003e,\u003c/strong\u003e 311-323.\u003c/li\u003e\n\u003cli\u003eSANCHEZ-GARCIA, R., GOMEZ-BLANCO, J., CUERVO, A., CARAZO, J. M., SORZANO, C. O. S. \u0026amp; VARGAS, J. 2021. DeepEMhancer: a deep learning solution for cryo-EM volume post-processing. \u003cem\u003eCommunications Biology,\u003c/em\u003e 4\u003cstrong\u003e,\u003c/strong\u003e 874.\u003c/li\u003e\n\u003cli\u003eSCHUMACHER, A. J., MOHNI, K. N., KAN, Y., HENDRICKSON, E. A., STARK, J. M. \u0026amp; WELLER, S. K. 2012. The HSV-1 Exonuclease, UL12, Stimulates Recombination by a Single Strand Annealing Mechanism. \u003cem\u003ePLOS Pathogens,\u003c/em\u003e 8\u003cstrong\u003e,\u003c/strong\u003e e1002862.\u003c/li\u003e\n\u003cli\u003eSEHRAWAT, S., KUMAR, D. \u0026amp; ROUSE, B. T. 2018. Herpesviruses: Harmonious Pathogens but Relevant Cofactors in Other Diseases? \u003cem\u003eFrontiers in Cellular and Infection Microbiology,\u003c/em\u003e 8.\u003c/li\u003e\n\u003cli\u003eSERQUI\u0026Ntilde;A, A. K. \u0026amp; ZIEGELBAUER, J. M. 2017. How herpesviruses pass on their genomes. \u003cem\u003eJ Cell Biol,\u003c/em\u003e 216\u003cstrong\u003e,\u003c/strong\u003e 2611-2613.\u003c/li\u003e\n\u003cli\u003eSEVERINI, A., SCRABA, D. G. \u0026amp; TYRRELL, D. L. 1996. Branched structures in the intracellular DNA of herpes simplex virus type 1. \u003cem\u003eJ Virol,\u003c/em\u003e 70\u003cstrong\u003e,\u003c/strong\u003e 3169-75.\u003c/li\u003e\n\u003cli\u003eSHELDRICK, P. \u0026amp; BERTHELOT, N. 1975. Inverted repetitions in the chromosome of herpes simplex virus. \u003cem\u003eCold Spring Harb Symp Quant Biol,\u003c/em\u003e 39 Pt 2\u003cstrong\u003e,\u003c/strong\u003e 667-78.\u003c/li\u003e\n\u003cli\u003eSIGNER, E. R. \u0026amp; WEIL, J. 1968. Recombination in bacteriophage lambda. I. Mutants deficient in general recombination. \u003cem\u003eJ Mol Biol,\u003c/em\u003e 34\u003cstrong\u003e,\u003c/strong\u003e 261-71.\u003c/li\u003e\n\u003cli\u003eSMITH, S., REUVEN, N., MOHNI, K. N., SCHUMACHER, A. J. \u0026amp; WELLER, S. K. 2014. Structure of the herpes simplex virus 1 genome: manipulation of nicks and gaps can abrogate infectivity and alter the cellular DNA damage response. \u003cem\u003eJ Virol,\u003c/em\u003e 88\u003cstrong\u003e,\u003c/strong\u003e 10146-56.\u003c/li\u003e\n\u003cli\u003eSTOLZENBERG, M. C. \u0026amp; OOKA, T. 1990. Purification and properties of Epstein-Barr virus DNase expressed in Escherichia coli. \u003cem\u003eJ Virol,\u003c/em\u003e 64\u003cstrong\u003e,\u003c/strong\u003e 96-104.\u003c/li\u003e\n\u003cli\u003eTAYLOR, T. J., MCNAMEE, E. E., DAY, C. \u0026amp; KNIPE, D. M. 2003. Herpes simplex virus replication compartments can form by coalescence of smaller compartments. \u003cem\u003eVirology,\u003c/em\u003e 309\u003cstrong\u003e,\u003c/strong\u003e 232-247.\u003c/li\u003e\n\u003cli\u003eTOLUN, G., MAKHOV, A. M., LUDTKE, S. J. \u0026amp; GRIFFITH, J. D. 2013. Details of ssDNA annealing revealed by an HSV-1 ICP8-ssDNA binary complex. \u003cem\u003eNucleic Acids Research,\u003c/em\u003e 41\u003cstrong\u003e,\u003c/strong\u003e 5927-37.\u003c/li\u003e\n\u003cli\u003eTR\u0026Auml;GER, L. K., DEGEN, M., PEREIRA, J., DURAIRAJ, J., TEIXEIRA, RAPHAEL D., HILLER, S. \u0026amp; HUGUENIN-DEZOT, N. 2025. Structural basis for cooperative ssDNA binding by bacteriophage protein filament P12. \u003cem\u003eNucleic Acids Research,\u003c/em\u003e 53.\u003c/li\u003e\n\u003cli\u003eTSURUMI, T. 2001. EBV Replication Enzymes. \u003cem\u003eIn:\u003c/em\u003e TAKADA, K. (ed.) \u003cem\u003eEpstein-Barr Virus and Human Cancer.\u003c/em\u003e Berlin, Heidelberg: Springer Berlin Heidelberg.\u003c/li\u003e\n\u003cli\u003eTSURUMI, T., KISHORE, J., YOKOYAMA, N., FUJITA, M., DAIKOKU, T., YAMADA, H., YAMASHITA, Y. \u0026amp; NISHIYAMA, Y. 1998. Overexpression, purification and helix-destabilizing properties of Epstein-Barr virus ssDNA-binding protein. \u003cem\u003eJ Gen Virol,\u003c/em\u003e 79 ( Pt 5)\u003cstrong\u003e,\u003c/strong\u003e 1257-64.\u003c/li\u003e\n\u003cli\u003eTSURUMI, T., KOBAYASHI, A., TAMAI, K., YAMADA, H., DAIKOKU, T., YAMASHITA, Y. \u0026amp; NISHIYAMA, Y. 1996. Epstein-Barr virus single-stranded DNA-binding protein: purification, characterization, and action on DNA synthesis by the viral DNA polymerase. \u003cem\u003eVirology,\u003c/em\u003e 222\u003cstrong\u003e,\u003c/strong\u003e 352-64.\u003c/li\u003e\n\u003cli\u003eUPRICHARD, S. \u0026amp; KNIPE, D. 2003. Conformational Changes in the Herpes Simplex Virus ICP8 DNA-Binding Protein Coincident with Assembly in Viral Replication Structures. \u003cem\u003eJournal of virology,\u003c/em\u003e 77\u003cstrong\u003e,\u003c/strong\u003e 7467-76.\u003c/li\u003e\n\u003cli\u003eVALLEDOR, M., MYERS, R. S. \u0026amp; SCHILLER, P. C. 2018. Herpes ICP8 protein stimulates homologous recombination in human cells. \u003cem\u003ePLOS ONE,\u003c/em\u003e 13\u003cstrong\u003e,\u003c/strong\u003e e0200955.\u003c/li\u003e\n\u003cli\u003eWAGNER, M. J. \u0026amp; SUMMERS, W. C. 1978. Structure of the joint region and the termini of the DNA of herpes simplex virus type 1. \u003cem\u003eJ Virol,\u003c/em\u003e 27\u003cstrong\u003e,\u003c/strong\u003e 374-87.\u003c/li\u003e\n\u003cli\u003eWANG, Y. S. \u0026amp; HALL, J. D. 1990. Characterization of a major DNA-binding domain in the herpes simplex virus type 1 DNA-binding protein (ICP8). \u003cem\u003eJ Virol,\u003c/em\u003e 64\u003cstrong\u003e,\u003c/strong\u003e 2082-9.\u003c/li\u003e\n\u003cli\u003eWEERASOORIYA, S., DISCIPIO, K. A., DARWISH, A. S., BAI, P. \u0026amp; WELLER, S. K. 2019. Herpes simplex virus 1 ICP8 mutant lacking annealing activity is deficient for viral DNA replication. \u003cem\u003eProceedings of the National Academy of Sciences,\u003c/em\u003e 116\u003cstrong\u003e,\u003c/strong\u003e 1033-1042.\u003c/li\u003e\n\u003cli\u003eWELLER, S. K. \u0026amp; SAWITZKE, J. A. 2014. Recombination promoted by DNA viruses: phage \u0026lambda; to herpes simplex virus. \u003cem\u003eAnnu Rev Microbiol,\u003c/em\u003e 68\u003cstrong\u003e,\u003c/strong\u003e 237-58.\u003c/li\u003e\n\u003cli\u003eWELLER, S. K., SEGHATOLESLAMI, M. R., SHAO, L., ROWSE, D. \u0026amp; CARMICHAEL, E. P. 1990. The herpes simplex virus type 1 alkaline nuclease is not essential for viral DNA synthesis: isolation and characterization of a lacZ insertion mutant. \u003cem\u003eJ Gen Virol,\u003c/em\u003e 71 ( Pt 12)\u003cstrong\u003e,\u003c/strong\u003e 2941-52.\u003c/li\u003e\n\u003cli\u003eWILKIE, N. M. 1973. The synthesis and substructure of herpesvirus DNA: the distribution of alkali-labile single strand interruptions in HSV-1 DNA. \u003cem\u003eJ Gen Virol,\u003c/em\u003e 21\u003cstrong\u003e,\u003c/strong\u003e 453-67.\u003c/li\u003e\n\u003cli\u003eWILKINSON, D. E. \u0026amp; WELLER, S. K. 2003. The role of DNA recombination in herpes simplex virus DNA replication. \u003cem\u003eIUBMB Life,\u003c/em\u003e 55\u003cstrong\u003e,\u003c/strong\u003e 451-8.\u003c/li\u003e\n\u003cli\u003eWOLFE, S. A., AND, L. N. \u0026amp; PABO, C. O. 2000. DNA Recognition by Cys2His2 Zinc Finger Proteins. \u003cem\u003eAnnual Review of Biophysics and Biomolecular Structure,\u003c/em\u003e 29\u003cstrong\u003e,\u003c/strong\u003e 183-212.\u003c/li\u003e\n\u003cli\u003eWONG, Y., MEEHAN, M. T., BURROWS, S. R., DOOLAN, D. L. \u0026amp; MILES, J. J. 2022. Estimating the global burden of Epstein-Barr virus-related cancers. \u003cem\u003eJ Cancer Res Clin Oncol,\u003c/em\u003e 148\u003cstrong\u003e,\u003c/strong\u003e 31-46.\u003c/li\u003e\n\u003cli\u003eWU, F. Y., AHN, J. H., ALCENDOR, D. J., JANG, W. J., XIAO, J., HAYWARD, S. D. \u0026amp; HAYWARD, G. S. 2001. Origin-independent assembly of Kaposi\u0026apos;s sarcoma-associated herpesvirus DNA replication compartments in transient cotransfection assays and association with the ORF-K8 protein and cellular PML. \u003cem\u003eJ Virol,\u003c/em\u003e 75\u003cstrong\u003e,\u003c/strong\u003e 1487-506.\u003c/li\u003e\n\u003cli\u003eXU, L., CABANAS-DAN\u0026Eacute;S, J., HALMA, M. T. J., HELLER, I., STRATMANN, SARAH A., VAN OIJEN, ANTOINE M., LEE, S.-J., PETERMAN, E. J. G. \u0026amp; WUITE, G. J. L. 2023. Regulation of T7 gp2.5 binding dynamics by its C-terminal tail, template conformation and sequence. \u003cem\u003eNucleic Acids Research\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eYU, G., SMITH, D. K., ZHU, H., GUAN, Y. \u0026amp; LAM, T. T.-Y. 2017. ggtree: an r package for visualization and annotation of phylogenetic trees with their covariates and other associated data. \u003cem\u003eMethods in Ecology and Evolution,\u003c/em\u003e 8\u003cstrong\u003e,\u003c/strong\u003e 28-36.\u003c/li\u003e\n\u003cli\u003eZAKHAROVA, K., CALDWELL, B. J., TA, S., WHEAT, C. T. \u0026amp; BELL, C. E. 2021. Mutational Analysis of Red\u0026beta; Single Strand Annealing Protein: Roles of the 14 Lysine Residues in DNA Binding and Recombination In Vivo. \u003cem\u003eInt J Mol Sci,\u003c/em\u003e 22.\u003c/li\u003e\n\u003cli\u003eZENG, Y., MIDDELDORP, J., MADJAR, J. J. \u0026amp; OOKA, T. 1997. A major DNA binding protein encoded by BALF2 open reading frame of Epstein-Barr virus (EBV) forms a complex with other EBV DNA-binding proteins: DNAase, EA-D, and DNA polymerase. \u003cem\u003eVirology,\u003c/em\u003e 239\u003cstrong\u003e,\u003c/strong\u003e 285-95.\u003c/li\u003e\n\u003cli\u003eZERIO, C. J., BAI, Y., SOSA-ALVARADO, B. A., GUZI, T. \u0026amp; LANDER, G. C. 2024. Human polymerase theta helicase positions DNA microhomologies for double-strand break repair. \u003cem\u003ebioRxiv\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eZHANG, C. X., DECAUSSIN, G., DAILLIE, J. \u0026amp; OOKA, T. 1988. Altered expression of two Epstein-Barr virus early genes localized in BamHI-A in nonproducer Raji cells. \u003cem\u003eJ Virol,\u003c/em\u003e 62\u003cstrong\u003e,\u003c/strong\u003e 1862-9.\u003c/li\u003e\n\u003cli\u003eZHANG, C. X., DECAUSSIN, G., DE TURENNE TESSIER, M., DAILLIE, J. \u0026amp; OOKA, T. 1987. Identification of an Epstein-Barr virus-specific desoxyribonuclease gene using complementary DNA. \u003cem\u003eNucleic Acids Res,\u003c/em\u003e 15\u003cstrong\u003e,\u003c/strong\u003e 2707-17.\u003c/li\u003e\n\u003cli\u003eZIMMERMANN, J. \u0026amp; HAMMERSCHMIDT, W. 1995. Structure and role of the terminal repeats of Epstein-Barr virus in processing and packaging of virion DNA. \u003cem\u003eJ Virol,\u003c/em\u003e 69\u003cstrong\u003e,\u003c/strong\u003e 3147-55.\u003c/li\u003e\n\u003cli\u003eZOU, Y., LIU, Y., WU, X. \u0026amp; SHELL, S. M. 2006. Functions of human replication protein A (RPA): from DNA replication to DNA damage and stress responses. \u003cem\u003eJ Cell Physiol,\u003c/em\u003e 208\u003cstrong\u003e,\u003c/strong\u003e 267-73.\u003c/li\u003e\n\u003cli\u003eZOU, Z., WU, S., XIONG, J., LI, H., JIANG, Y. \u0026amp; ZHANG, H. 2018. ssDNA hybridization facilitated by T7 ssDNA binding protein (gp2.5) rapidly initiates from the strand terminus or internally followed by a slow zippering step. \u003cem\u003eBiochimie,\u003c/em\u003e 147\u003cstrong\u003e,\u003c/strong\u003e 1-12.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6794668/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6794668/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eEpstein-Barr virus is an oncogenic herpesvirus present in 95% of the global population. It encodes the highly conserved BALF2 protein as an essential member of its replisome. BALF2 is a multifunctional protein which acts as a general single-stranded DNA-binding protein during replication, and as an ATP-independent recombinase involved in the single-strand annealing homologous recombination pathway. Several lines of evidence suggest that homologous recombination is an integral feature of herpesvirus DNA replication, required for the generation of concatemeric replication intermediates, genomic maintenance, and as a major driver of genetic diversity. BALF2 and its homologues are therefore promising antiviral targets. Despite over half a century of research into the herpesvirus annealase proteins, a significant roadblock persists in our understanding of their binding and annealing mechanisms. Here, we present a structure of a BALF2 DNA annealing intermediate, determined to 2.2 \u0026Aring; resolution by cryogenic electron-microscopy (cryo-EM). This structure allowed for the identification and characterisation of an oligonucleotide-binding fold, a zinc-binding loop, an active site of ssDNA-annealing, and suggests a model for cooperative binding and oligomerisation. We also investigated BALF2 through biochemical assays, bioinformatic sequence analysis and molecular dynamics simulations to further characterise regions of the protein\u0026rsquo;s structure. These findings will strongly inform future studies on herpesvirus annealases and have great potential as a starting point for structure-based drug design.\u003c/p\u003e","manuscriptTitle":"EBV BALF2 DNA annealing intermediate structure reveals the mechanism of annealing during recombination","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-22 11:51:55","doi":"10.21203/rs.3.rs-6794668/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"ea31f7b9-2d93-4c61-a943-8197f80ba1ac","owner":[],"postedDate":"July 22nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":49665308,"name":"Biological sciences/Structural biology/Electron microscopy/Cryoelectron microscopy"},{"id":49665309,"name":"Biological sciences/Molecular biology/DNA recombination"}],"tags":[],"updatedAt":"2025-07-22T11:51:55+00:00","versionOfRecord":[],"versionCreatedAt":"2025-07-22 11:51:55","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6794668","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6794668","identity":"rs-6794668","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}