Evolution of Eukaryotic Specific DNA Binding Sites in Asgard Archaeal RecA Recombinases

Evolution of Eukaryotic Specific DNA Binding Sites in Asgard Archaeal RecA Recombinases | bioRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (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];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-M677548'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search New Results Evolution of Eukaryotic Specific DNA Binding Sites in Asgard Archaeal RecA Recombinases Bryan Ferlez , Peter Towei Huang , Jingyi Hu , J. Brooks Crickard doi: https://doi.org/10.1101/2025.02.19.639135 Bryan Ferlez 1 Department of Molecular Biology and Genetics , Cornell University Ithaca, NY 14853, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Peter Towei Huang 1 Department of Molecular Biology and Genetics , Cornell University Ithaca, NY 14853, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jingyi Hu 1 Department of Molecular Biology and Genetics , Cornell University Ithaca, NY 14853, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site J. Brooks Crickard 1 Department of Molecular Biology and Genetics , Cornell University Ithaca, NY 14853, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site For correspondence: jbc287{at}cornell.edu Abstract Full Text Info/History Metrics Supplementary material Preview PDF Abstract RecA-type recombinases are essential for genome maintenance in all domains of life, promoting homologous recombination (HR). While the general reaction catalyzed by recombinases is the same in all domains, key differences at the amino acid level can lead to the specialization of this family of proteins. A key example of this is in the domain of Eukarya, which can utilize two different RecA homologs depending on the life cycle stage of the organism, with the activity of Rad51 utilized during mitotic growth and the activity of Dmc1 used during meiosis. Why most eukaryotes utilize two RecA homologs is an open question. A missing piece of information in this story is the identity of RecA homologs before the divergence of Rad51 and Dmc1, which occurred in early eukaryotes. We analyzed sequence data from the Archaeal homolog, RadA, in the Asgard Archaea superphyla. Our findings include evidence for the evolution of a eukaryotic-specific DNA binding site II within the Asgard lineage. We also identify sequence variants within DNA binding loops L1 and L2, part of binding site I, that imply higher mismatch tolerance within the Asgard and DPANN superphylum. These loops are also observed in the meiosis-specific recombinase Dmc1 within Eukarya. Our phylogenetic analysis suggests the last Archaeal ancestor of Rad51 and Dmc1 was more Dmc1-like in character and likely of lower fidelity. Introduction Homologous recombination (HR) is a universally conserved genome maintenance pathway that controls the flow of genetic information within populations and between species. Across the domains of life, HR protects DNA from exogenous and endogenous damage and sorts and concentrates alleles [ 1 – 4 ]. The essential function of HR is to repair DNA double-strand breaks within the genome, and this requires a template that acts as a donor to provide sequence information to repair the break. If the donor is the newly replicated sister chromatid, fidelity is maintained. However, it can benefit an organism or population to integrate pieces of foreign DNA into their genome or repair defective alleles using exogenous DNA [ 5 – 8 ]. This makes regulated integration of foreign DNA vital to both an organism’s short-term success and a species’ long-term survival. Eukaryotes have successfully compartmentalized HR into two distinct versions of the cell cycle. During mitotic growth, HR limits the exchange of genetic information by enriching for repair from the sister chromatid [ 4 , 9 , 10 ], while during meiosis, the exchange of information between homologous chromosomes is encouraged [ 8 , 11 ]. Meiosis also allows the bridging of homologous chromosomes and promotes independent assortment. This process creates new combinations of genes through the sorting of alleles [ 11 – 13 ]. Essential to regulated genetic exchange are the recombinase homologs of the RecA family [ 2 , 14 , 15 ]. The conserved feature of this group of enzymes is the ability to form filaments on double and single-strand DNA in an ATP-dependent manner, and to pair the recipient ssDNA to a homologous donor duplex DNA during the homologous recombination (HR) reaction. The DNA strand pairing and exchange process is fundamental to genetic processes and is achieved by coordinating two DNA binding sites within RecA filaments. These sites are referred to as DNA binding site I, composed of two DNA binding loops, L1 and L2 [ 16 , 17 ], and binding site II, located distally from the DNA binding loops [ 18 – 21 ]. During the reaction, the L1 and L2 from adjacent RecA protomers coordinate to separate the duplex donor DNA, which allows the faithful pairing of the incoming strand with its homologous partners. The disrupted donor strand is bound by binding site II, stabilizing the outgoing strand and making the pairing process energetically favorable [ 16 , 21 ]. In eubacteria, these enzymes are called RecA, and in Archaea, RadA. In most Eukarya, there are two recombinases, with Rad51 specializing in strand exchange during mitosis and Dmc1 facilitating strand exchange during meiosis [ 20 ]. Rad51 and Dmc1 are specialized recombinases. Rad51 has acquired novel functions to work at DNA replication forks. This includes protecting reversed replication forks and promoting lesion by-pass through template switching [ 22 – 24 ]. Some of these activities are shared with RecA but are likely missing in Dmc1. The specialization of Dmc1 for meiosis has been a more complicated question to answer. Considerable attention has been given to differences between Rad51 and Dmc1 in DNA binding site I. L1 and L2 are points of amino acid variation between Rad51 and Dmc1, with several studies highlighting that mutations in L1 and L2 can make Rad51 tolerant or Dmc1 intolerant of DNA base mismatches during strand pairing, also called homeologous recombination [ 25 – 29 ]. This led to the hypothesis that Dmc1 may be more efficient at pairing mismatched DNA, which could improve allelic recombination of variant genes on homologous chromosomes [ 25 ]. Arguments against this model include the relatively low density of polymorphisms in allelic variants within species and that the dedicated mismatch repair machinery (MMR) repairs most mismatches after strand exchange has occurred. This mechanism is the primary driver of gene conversion in S. cerevisiae [ 30 , 31 ]. This does not negate that Dmc1 is biochemically more tolerant of mismatches. Still, open questions remain about why eukaryotes utilize recombinase specialization to carry out distinctive HR pathways during mitosis and meiosis and how these recombinases diverged. Studies have focused on one of the earliest eukaryotic RecA sequences to address these questions, the protist Giardia lamblia, which has two recombinases closer in sequence to Dmc1 [ 32 ]. The exception is in L2, which contains the Rad51 sequence for both copies. This suggests there are still missing pieces in the evolutionary story of Rad51 and Dmc1. More recently, the discovery of the Asgard superphyla of Archaea has more firmly established a transition point between Archaea and Eukarya [ 33 , 34 ]. This has provided a new avenue for evaluating homologous protein sequences at this critical point in evolution. Previous work with recombinases from Archaea illustrated that RadA [ 35 ] was more closely related to recombinases from eukarya than eubacterial RecA [ 35 ]. Prior work with RadA included biochemical and structural analysis of thermophilic, methanogenic, and halophilic recombinases [ 36 – 38 ]. RadA has also been linked to genetic activities in halophilic Archaea, including strand exchange, gene conversion, and replication initiation [ 39 , 40 ]. In this study, we have evaluated the protein sequence composition of RadA recombinases from the Asgard superphyla of Archaea. By comparing these sequences with RadA sequences from other Archaea and Rad51/Dmc1 in Eukarya. Within the Asgard superphyla, we see the first evidence of fully developed Eukaryotic DNA binding sites I and II. We find evidence for variation of the DNA binding loops within the Asgard and DPANN superphyla of Archaea. With this analysis and additional genetic and biochemical experiments, we hypothesize that the last eukaryotic common ancestor was more like Dmc1 than Rad51. Dmc1 likely reflects a more ancestral recombinase, which is more tolerant of DNA mismatches. Furthermore, Rad51 has likely undergone specialization to improve the overall stringency of DNA selection during strand exchange, while the selection on Dmc1 has remained neutral. Results Analysis of recombinase diversity within the Asgard Archaea Amino acid sequences of putative Asgard recombinases were obtained using BLASTp with S. cerevisiae Rad51 (Uniprot P25454), H. sapiens Rad51 (Uniprot Q06609), and S. solfataricus RadA (Uniprot Q55075) sequences as queries to the Asgard Group database (Taxid: 1935183). These sequences were processed and used to construct a phylogenetic tree using RecA as the outgroup. From this tree, we confidently called nineteen separate groups ( Figure 1A ). One of these groups was composed of RadB, a recombination mediator protein that serves an analogous role to the Rad51 paralogs in eukarya [ 41 – 43 ]. These sequences were included in the tree to ensure the proper grouping of the RadA genes. Therefore, we identified eighteen RadA groups. These groups were occupied by recombinases from different Asgard Archaea, including Loki, Heimdall, Thor, Hela, Odin, and other Asgard groups, representing phyla. After organizing these groups, we constructed a second tree with the RadB sequences removed. This resulted in the confident identification of eighteen RadA groups ( Figure 1B ). Of the initial 18 groups, four (groups 2, 3, 14, and 15) lacked clear Walker A or B, DNA binding Loop 1 and/or DNA binding Loop 2 motifs and were not analyzed further as they are unlikely to represent RadA homologs. Nine individual sequences from the remaining 14 groups also appeared to lack canonical recombinase motifs, or were incomplete, and were removed from our analysis, resulting in 196 sequences across 14 groups ( Figure 1C ). Download figure Open in new tab Figure 1: Organization of RadA in Asgard Archaea (A) Maximum likelihood phylogenetic tree of RadA/RadB protein sequences from Asgard Archaea generated by IQ-tree (version 1.6.12). The Asgard Archaea formed 18 distinct groups of recombinases. The branch confidence was determined by the Shimodaira–Hasegawa approximate likelihood ratio test (SH-aLRT) and ultrafast bootstrap (UFBS) analyses. E. coli RecA was included as an outgroup, and RadB-like sequences are surrounded by the dashed oval. (B). A maximum likelihood phylogenetic tree of only the Asgard RadA sequences. The branch confidence was determined by bootstrap re-sampling using ultrafast bootstrapping. (C). Block alignment protein sequences from 14 or the 18 groups identified recombinases from Asgard Archaea. The length of each rectangular block reflects the overall length of the sequence conservation logo generated from a multiple sequence alignment of all unabridged sequences in each group. Each block associated with an Asgard group was aligned by the location of the Walker A box in the RecA domain. Also highlighted are the relative locations of the Walker B motif and the DNA binding loops. Also illustrated are N-terminal and C-terminal extensions unique to different groups of recombinase proteins. To understand how these fourteen groups fit in with recombinase sequences across the domains of life, we generated a large unrooted maximum likelihood phylogenetic tree composed of 2511 sequences of Archaeal recombinases from the TACK (N = 758), EURY (N = 782), DPANN (N = 621), and Asgard (N = 198) superphyla. Forty-three Rad51 sequences, twenty-eight Dmc1 sequences from eukaryotes, Escherichia coli RecA, and two eubacterial sequences from each of the forty phyla were selected for the tree (N = 81). The clustering of TACK, EURY, and DPANN sequences within groups reflecting their representative superphyla was primarily supported by SH-aLRT and UFBS statistical tests ( Figure 2 ). There were a few exceptions in these clusters, which may represent HGT signatures. Importantly, several of the Asgard groups clustered with eukaryotic Rad51 and Dmc1. Included in these groups were 5, 9, 10, and 12 ( Figure 2 ). Groups 5, 9, and 12 were composed of Ca. Heimdallarchaeota. This is consistent with previous reports that the last Archaeal ancestor of eukaryotes originated from this group [ 33 , 34 , 44 , 45 ]. Group 10 was composed of Ca. Freyaarchaeota, and Ca. Signyarchaeota ( Figure 2 ). From our tree, the last Archaeal ancestor of Rad51 and Dmc1 was likely from group 12 and a Ca. Heimdallarchaeota ( Figure 2 ). Download figure Open in new tab Figure 2: Groups of Asgard Archaea group with Eukaryotic recombinases (A) Cladogram view of maximum likelihood phylogenetic tree built from recombinase sequences spanning the domains of Archaea (TACK (Purple), EURY (Magenta), DPANN (Orange), and Asgard (Green)), Eubacteria, and Eukarya. The orange dots represent statistical significance using SH-aLRT and UFBS (SH-aLRT >80 & UFBS >95). Next, we compared the total sequence identity for each of these groups and compared them to H.s. Rad51, S.c. Rad51, H.s. Dmc1, S.c. Dmc1, Pyrococcus furiosus RadA, S.s . RadA, and E.c. RecA ( Supplemental Figure 1 ). This comparison was distinct from our tree as the total protein sequence was used for this analysis, whereas with the tree, only the core region of the recombinase was used for the final grouping. All groups had between a 17-25% sequence identity with EcRecA. Groups 1, 4, 6, 7, 8, and 12 were between 20-40% identical to P.f. RadA and S.s. RadA and showed similar sequence identity to eukaryotic recombinases. Groups 9,10,11,13, and 18 were ∼40-50% identical to archaeal and eukaryotic recombinases. Finally, groups 16 and 17 were 30-45% identical to archaeal and eukaryotic recombinases. This analysis concludes that five of our identified groups show high sequence identity to eukaryotic recombinases (>40%). Thirteen groups show significant sequence divergence from RecA, Rad51/Dmc1, and existing RadA sequences. We also performed multiple sequence alignments for these proteins ( Supplemental Figure 2 and Supplemental Figure 3 ). From this we conclude that RadA in Asgard Archaeal can sample a wide sequence space. The ATP binding site of Asgard Archaea are Eukaryotic like ATP binding is an essential feature of all RecA recombinases, and except for the T4 phage RecA recombinase UvsX [ 46 ], it promotes filament formation on DNA. The role of ATP hydrolysis in the recombination reaction is to prevent the inappropriate binding of recombinases during the reaction, permitting hydrolysis to promote the release of RecA/Rad51 from DNA [ 47 ]. This is important in the homology search and the subsequent DNA repair steps of HR [ 48 ]. Importantly, there is a significant difference in ATP hydrolysis rates observed between RecA from E. coli and Rad51/Dmc1 from H. sapiens and S. cerevisiae [ 49 ]. This is a fundamental difference between these two versions of the enzyme. The lack of high hydrolysis rates in eukaryotes has resulted in the evolution of numerous accessory factors that modulate the nucleotide binding state of Rad51/Dmc1[ 50 – 52 ]. Therefore, understanding how the ATP binding pocket has changed over evolutionary time is critical in understanding how recombinases are regulated. It has previously been shown that RadA from several Archaeal species have lower ATP hydrolysis rates than those in E. coli . Next, we evaluated the active site of Asgard Archaea to understand the variation within the superphylum. The ATP binding pocket can be divided into two regions: the Adenosine binding cap (A-cap) and the phosphate-binding loop (P-loop). The P-loop contains the Walker A motif, which includes the catalytic lysine residue required for hydrolysis. In recombinases, the adjacent protomer supplies the Walker B motif and provides the necessary aspartic acid residue to properly position H 2 O to transfer the gamma phosphate ( Figure 3A ). In eubacteria, the A-cap comprises two tyrosine (Y103, Y264) and an aspartic acid (D100). These residues are replaced by arginine (R181 Pf and R170 Hs) and proline (P334 Pf and P321 Hs) in archaeal RadA and Rad51/Dmc1, respectively. The residues interact with the adenosine by base stacking. An additional difference observed between the binding pocket in each domain is the substitution of phenylalanine (F217) in RecA for a histidine (H307 Pf, H294 Hs) in RadA and Rad51/Dmc1, respectively, and the substitution of lysine (K248) for aspartic acid (D329 Pf, D316 Hs) in RadA and Rad51/Dmc1 ( Figure 3A ). These residues aid in coordinating the phosphate residues outside of the P-loop. It should be noted that these positions showed extremely high levels of conservation within their domains. Download figure Open in new tab Figure 3. ATP binding sites of bacterial, archaeal, and eukaryotic recombinases and their relationship to Asgard RadA homologs (A) ATP binding sites of representative recombinases from the three domains of life. E. coli RecA (pdb 1XMS), P. furiosus RadA (AlphaFold model AF-O74036-F1), and human Rad51 (pdb 5H1C) chains were aligned to the cryo-EM structure of the post-synaptic filament of human Dmc1 (pdb 798C) in Pymol; human Dmc1 AMP and Ca 2+ are shown in each panel for reference. Side chains of residues proximal to either the adenosine (black dashed box, black amino acid labels) or triphosphate group (purple dashed box, purple amino acid labels) are shown as sticks. In all panels, the numbering reflects the source sequence. (B). Conservation of ATP binding site residues in Asgard RadA homologs. For reference, ATP binding site residues (pink sticks) are shown for a Group 13 RadA homolog from a Ca. Helarchaeota archaeon (AlphaFold AF-A0A850LYW3-F1);). Residues proximal to the adenosine or triphosphate group are highlighted with black or purple circles, respectively. The grey dashed circle surrounding the aspartic acid proximal to the adenosine base indicates a position that only appears to be conserved in group 7 as tryptophan (see inset). Region specific logos are derived from a full-length alignment of all 1 Asgard RadA homologs Insets surrounded by black boxes show two representative Asgard ATP binding sites (Left: Group 7, Ca. Thorarchaeota archaeon SMTZ1_45, AF-A0A135VLP8-F1; Right: Group 1, Ca. Lokiarcaheota archaeon, AF-A0A7K4CML3-F1). (C). Walker A P-loop diversity across domains of life. The first three amino acids of the Walker A P-loop show Domain-specific conservation. They are indicated by dashed black boxes on the regional logos of RecA, Rad51, Dmc1, TACK, EURY, DPANN, and Asgard recombinases. Logos for the complete Asgard dataset (n=198) (solid blue box) as well its bifurcated sub-groups, those enriched in a basic amino acid (i.e. R/K) in the third position of the P-loop (Groups 4,7-13, 18) and those that are not (Groups 1,5,6,16,17), are shown in dashed blue boxes We investigated the Asgard Archaea for variation in this region and identified that most groups were conserved with Rad51/Dmc1 and RadA from other Archaea. However, we identified a unique ATP binding site composition in group 7. In this group, the R181/170 was replaced with a bulky hydrophobic residue, and a non-conserved residue was replaced with a tryptophan (W303) ( Figure 3B ). The proline in this region was also conserved in group 7. Groups 1 and 5 had substitutions in the phosphate binding region of the pocket, with group 1 having a Lysine, Glycine, and Histidine where the canonical grouping is Aspartic Acid, Histidine, and Glutamic Acid ( Figure 3B ). Group 5 substituted has an Alanine, Histidine, Glutamic Acid organization. It should be noted that the lysine substitution in group 1 is like that used in eubacterial RecA. This data suggests that Asgard Groups 1, 5, and 7 have novel variations in their ATP binding pocket that may alter the chemistry produced during recombination. The P-loop is a conserved feature of all nucleotide-binding proteins whose motif is GXXXXGKT; it has been suggested that the ancestral form of the motif may have been GXXGXGKT[ 53 , 54 ]. Next, we evaluated the P-loop sequences from each domain. We performed a limited sequence alignment of eighty bacterial species for RecA with two representatives from each group across superphyla. This analysis generated a consensus P-loop sequence of GPESSGKT ( Figure 3C ). For both Rad51 and Dmc1, the P-loop consensus sequence is GEFRXGKT ( Figure 3C ). In the archaeal domain, we subdivided our analysis into four superphyla and expanded the number of sequences used to build the consensus. The P-loops from TACK (GEFGXGKT), EURY(GEFGSGKT), and DPANN (GXFGSGKT) were remarkably similar and allowed us to generate a consensus sequence for these groups of GXFGSGKT. In contrast, the Asgard group was highly variable in their P-loop with a consensus of GXXXXGKT ( Figure 3C ). However, within groups with higher sequence homology to the eukaryotic domain, the GEFRXGKT motif emerged ( Figure 3C ). Our analysis shows this is the only place in the Archaeal domain where the eukaryotic P-loop is observed. Notably, the arginine observed in the eukaryotic P-loop is part of the eukaryotic DNA binding site II that interacts with the non-homologous DNA strand during the strand exchange reactions. This position is different between the three domains except for the Asgard superphyla. In cooperation with three other residues, the arginine in the P-loop is essential for the completion of meiosis [ 20 , 55 ]. From this analysis, we draw two conclusions; the first is that recombinase P-loops are highly conserved and unique within each domain of life. The second is that the eukaryotic P-loop is completed in Asgard Archaea. Asgard Archaea exhibit DNA binding loops that appear like eukaryotic Dmc1 The primary DNA binding site within the RecA recombinase family comprises two DNA binding loops, L1 and L2, facilitating the strand exchange reaction. These loops represent a point of collaboration between two adjacent protomers in all domains of life [ 15 ]. While there is some variation within L1 and L2 amino acid sequences within domains, some residues appear to be fixed within and across domains. A critical variation occurs in the DNA binding loops of Dmc1 and Rad51 in eukarya. High-resolution structures demonstrate that arginine in L1 (R235, H. sapiens Rad51) forms a salt bridge with an L2 aspartic acid (D274, H. Sapiens Rad51) from an adjacent protomer ( Figure 4A ). This interaction is conserved in Rad51. However, in Dmc1, this interaction does not form because the aspartic acid has been replaced with glycine or a small amino acid such as serine or alanine ( Figure 4A ). The lone exception is Giardia intestinalis , which has two recombinases annotated as Dmc1 and containing the R-D at this position. Current models predict that this difference allows Dmc1 to allow mismatch tolerance during strand pairing [ 26 , 28 ]. Download figure Open in new tab Figure 4: Representative DNA binding loop variants of Asgard recombinases (Top left) Superposition of hRAD51 (pdb 5HC1; protomer 1 in light pink, protomer 2 in violet) and hDMC1 (pdb 7C98; protomer 1 in palegreen, protomer 2 in lime) structures highlighting the differences in amino acids proposed to regulate mismatches in DNA (also indicated by dashed boxes in amino acid logos). (Top right) Amino acid sequence logos for Rad51 (top, N = 43) and Dmc1 (bottom, N = 28) DNA binding Loop 1 and Loop 2 regions. (Bottom). Representative Alphafold2 structures for DNA binding loops observed in Asgard Archaea. Two copies of the structural model for each type (in pale cyan and cyan, respectively) are superimposed onto two adjacent protomers of hDMC1 (pdb 7C98). Below each structure is an amino acid logo for Loop 1 and Loop 2 generated from all sequences within the respective Asgard group (N = 23, 6, 12, and 21 for groups 9, 17, 13, and 7, respectively). The height of the amino acid letters represents their relative abundance. Type designation reflects the conservation of the proposed ionic interaction between R235 and D274 (hRad51 numbering); these positions are marked with orange circles in the amino acid logos. The canonical binding loop example is from Ca. Heimdallarchaeota archaeon LC3 (Group 9, AlphaFold model from UniProt:A0A1Q9NYX4), the Type I example from Ca. Heimdallarchaeota archaeon (Group 17, AlphaFold model from UniProt:A0A523TAR1), the Type II example from Ca. Helarchaeota archaeon (Group 13, AlphaFold model from UniProt:A0A850LYW3), the Type III example from Ca. Thorarchaeota archaeon (Group 7, AlphaFold model from UniProt:A0A135VV73). We evaluated the composition of DNA binding loops with the Asgard Archaea superphyla for their ability to form a predicted salt bridge during strand pairing. We structurally superimposed human Dmc1 or Rad51 with a member from each of the fourteen Asgard groups and compared these alignments to sequence logos. This was done to ensure the proper amino acid alignment in the structure. Our analysis revealed four different L1-L2 interactions. The interactions were defined as canonical (L1R/K-L2D)( Figure 4B ), which is Rad51-like, Type I (L1E/D/N/G-L2D) ( Figure 4C ), Type II (L1R-L2G/N), which is Dmc1 like ( Figure 4D ), and Type III (L1L/I-L2X(S)) ( Figure 4E ). In Asgard, six of fourteen groups had canonical interactions, and eight did not form a salt bridge between L1 and L2. This analysis suggests that Type II loops are not unique to eukarya and exist in archaea. We have also identified several types of novel loop interactions unique to archaea. We evaluated the distribution of L1 and L2 interactions in the Asgard Archaea groups. Most protein sequences were from the Heimdall, Thor, and Loki groups, with smaller contributions from Sigyna, Prometheo, Gerdar, Freya, Odin, Hela, and the undefined Asgard group. All analyzed groups except Hela had canonical types of DNA binding loops ( Supplemental Figure 4A ). Type I loops appeared in Signya, Loki, Heimdall, and the generic Asgard Group. Type II loops occurred in Heimdall, Loki, Hela, and a small number of Thor ( Supplemental Figure 4B ). Type III loops were distributed among all groups ( Supplemental Figure 4B ).) From this, we conclude that variations in DNA binding loop types are common within the Asgard superphyla. We performed the same protein sequence analysis on Archaea from the different superphyla. Not surprisingly, ∼90-92% of the DNA binding loop sequences from TACK and EURY were canonical ( Figure 5A ). Surprisingly, the remaining ∼8-10% were of the Type II variety. In the DPANN group, only ∼60% of sequences were canonical, and 40% were Type II ( Figure 5A ). This was the highest prevalence of Type II loops amongst the superphyla. The Asgard had the lowest percentage of canonical loops at ∼45%, and 10% Type I, 15% Type II, and ∼30% Type III ( Figure 5A ). This analysis indicated that Type II loops are prevalent in another superphylum of Archaea, but that Asgard contains the most diversity within their loop type. Download figure Open in new tab Figure 5: Asgard recombinases with different loop types are active (A). DNA binding loop distributions across Archaeal superphyla. DNA binding loops are scored as canonical (Black), Type I (Purple), Type II (Green), or Type III (Magenta) depending on the conservation of amino acids corresponding to R235 and D274 (hRad51 numbering). (B). Purification scheme for RadA recombines identified from Asgard Archaea (Left) and SDS-PAGE of purified product (Right) (B). Graph representing the ATP hydrolysis activity of Group 11 (Black), Group 17 (Blue), Group 13 (Green), Group 7 magenta, and S.c. Rad51 (Gold) at 30 (Left), 37 (middle), and 45° C (Right). The error bars represent the standard deviation of three independent experiments. (D). Electromobility Shift Assay for 0.3, 1.0, and 3.0 µM RadA (Groups 11, 17, 13, and 7) S.c Rad51 , and S.c.Dmc1 with and without ATP. The DPANN superphylum has small genomes, high levels of horizontal gene transfer, and a significant degree of symbiosis [ 56 – 59 ]. These conditions could favor the evolution of a low-fidelity recombinase. If this were true, we might expect high levels of low-fidelity RadA in the archaeal viruses. Even though there is a lot of sequence information on the Archaeal viruses, only a subset contains a RadA gene. Moreover, the databases had low confidence in many viral sequences we could return by blast. Despite these limitations, we aligned thirty-nine high to medium-confidence viral RadA genes ( Supplemental Figure 5A ). Of these genes, 66% had type II DNA binding loops ( Supplemental Figure 5B ). These data show that Archaeal viral RadA genes are enriched for potentially low-fidelity recombinases. Sequences from Asgard Metagenomes produce active enzymes Next, we selected protein sequences from Group 7 ( Ca. Thorarchaea , OLS31719.1), Group 11 ( Ca. Lokiarchaea, NHJ20009.1), Group 13 ( Ca. Lokiarchaea, NHI92944.1), and Group 17 ( Ca. Heimdallarchaea, TET27405.1). These recombinases were selected because the sequences were taken from marine sediment and were unlikely to be extremophiles; they represented different phyla within the Asgard group, and each had a different type of DNA binding loop. These proteins were overexpressed and purified from E. coli ( Figure 5B ) . We tested the purified proteins for ATP hydrolysis activity. As expected, the proteins exhibited ATP hydrolysis activity and increased ATP hydrolysis in the presence of ssDNA, consistent with eukaryotic recombinases ( Figure 5C ). There was no significant difference in ATP hydrolysis rate between the groups, with 1.32, 1.76, 1.44, and 1.27 µM ADP/min for groups 7, 11, 13, and 17, respectively. Activity peaked at 37°C, confirming that these proteins were not extremophilic ( Figure 5C ). These activities were comparable to S. cerevisiae Rad51 at 30°C ( Figure 5C ). We evaluated the ssDNA binding capacity of groups 7,11,13, and 17 and compared binding to S.c. Rad51 and S.c Dmc1 with and without ATP ( Figure 5D ). We found that all proteins bound to ssDNA had greater affinity in the presence of ATP ( Figure 5D ). There were apparent differences in affinity with Groups 11 and 13 binding with higher affinity than Groups 7 and 17, and S.c. Rad51 and S.c. Dmc1. Likewise, there were some differences in the minus ATP case. For example, Groups 11 and 17 displayed little binding without ATP, more like S.c . Rad51 ( Figure 5D ). In contrast, Groups 13 and 7 bound to ssDNA without ATP. This was like S.c. Dmc1( Figure 5D ). While this is a qualitative assessment and could result in other differences in the protein, it would suggest that the recombinases with Type II and Type III loops bind ssDNA more like Dmc1. We also compared preferences for binding ssDNA or dsDNA, and all proteins tested preferred ssDNA like their eukaryotic counterparts. From this, we conclude that we have expressed and purified ATPase active members of the Asgard RadA family. Canonical and Type II loops result in higher mismatch tolerance Finally, we chose to evaluate the role of DNA-binding loops in a physiological setting using an S. cerevisiae reporter assay [ 60 ]. Because the reported differences between Rad51 and Dmc1 are related to their ability to tolerate mismatches in the DNA, we chose to use a genetic reporter assay that could differentiate between effects on homologous recombination and homeologous recombination, which occur when there are mismatches in the DNA [ 60 , 61 ]. This assay is based on the orientation of the second half of the HIS3 gene, which is flanked by inverted repeats. If recombination occurs, the activity of the HIS3 gene can be restored, and the frequency of recombination can be determined by the number of HIS3+ colonies relative to the total population ( Figure 6A ). Notably, the percent identity of the inverted repeats can be modulated as wholly homologous or divergent in sequence (homeologous). In this case, we used a strain with a completely homologous inverted repeat and a strain that diverged by 9% sequence identity. By comparing the frequency of recombination in these strains, we can determine how a mutation in Rad51 affects homologous and homeologous recombination. We can then measure a ratio of homeologous to homologous (HER/HR) [ 60 ] to determine if these mutations lead to mismatch tolerance. Download figure Open in new tab Figure 6: Rad51 in Eukarya has evolved high fidelity DNA binding loops (A). Schematic diagram of reporter strain used to test the effects of DNA loop type on recombinase fidelity. These experiments used 100% and 91% homology strains in the inverted repeats. (B). Recombination frequency for WT, rad51-canonical , rad51-Type II , rad51D332G, and rad51 V331P, D332G with 100 and 91 % sequence homology within the inverted repeats. The bar represents the mean, and the error bar is the 95% confidence interval for at least eight independent experiments. (D). The bar graph represents the homeologous to homologous recombination ratio for WT, rad51-Canonical , rad51-Type II , rad51D332G, and rad51 V331P, D332G . ( D ). Cladogram view of Phylogenetic tree (From Figure 2 ) with expanded view of the split between Archaeal and Eukaryotic recombinases. To the left is a sequence logo for the DNA-binding loops within each clade. The yellow dots represent statistical confidence in the split of the branches. We measured the impact of incorporating the canonical DNA binding loop or DNA binding loop II into Rad51 ( Figure 6B ). We found in the WT strain a homologous recombination rate of 2×10 −5 +/- 3.0×10 −6 ( Figure 6B ). This was only marginally higher than the homologous recombination rate for the rad51-Canonical and rad51-Type II strains (1.2×10 −5 and 1.7×10 −5 , respectively) ( Figure 6B ). The homeologous recombination rate observed for WT at 91% homology was 6.4×10 −7 +/-3.8×10 −7 ( Figure 6C ). This was significantly lower than the recombination rates observed for rad51-canonical and rad51-TypeII (2.8×10 −6 and 3.8×10 −6 , respectively), with a difference of 4 and 4.5-fold ( Figure 6B ). These differences led to a homeologous-to-homologous ratio of 7.7 and 6.8, respectively ( Figure 6C ). This data suggests that the rad51-Canonical and rad51-TypeII are active for recombination and have a higher tolerance for mismatches. To test whether these rad51 alleles could full complement RAD51 , we used an assay to test the sensitivity of these strains to the DNA alkylating agent methyl methanesulfonate (MMS). We measured the survival of the strains on 0.0005% MMS and found that the rad51-Canonical and rad51-Type II strains only partially complemented the MMS sensitivity observed in the rad51Δ strain ( Supplemental Figure 6A ). This data suggests that while the substitution of ancestral DNA binding loops can complement some of the functions of Rad51, they are insufficient to entirely complement all DNA damage phenotypes associated with Rad51. Next, we used an assay based on allelic recombination to confirm further that these recombinases were active. In this assay, recombination between homologous chromosomes results in restoring an inactive leu2 gene ( Supplemental Figure 6B ). This allows us to measure the frequency of gene conversion resulting from homologous recombination ( Supplemental Figure 6B ). We found that there was a significant difference between the WT strains and the rad51-canonical or the rad51-Type II strains for gene conversion via HR, suggesting reduced activity ( Supplemental Figure 6C ). We observed that the rad51-canonical and rad51-TypeII strains had higher gene conversion efficiency than the rad51Δ strains ( Supplemental Figure 6C ), consistent with the MMS experiments. This indicates that these alleles were competent for some recombination levels, but the complementation was incomplete. Our analysis of the Canonical and Type II DNA binding loops suggested that both loops had similar fidelity even though the Type II loops cannot form the salt bridge that is important for fidelity [ 25 , 26 , 28 , 29 ]. However, due to the reduced complementation of the ancestral loops, we decided to make more refined mutations in the DNA binding loops. We directly targeted the salt bridge and the adjacent valine for substitution. These residues are different between Rad51 and Dmc1. Previous work has shown that swapping the VD amino acids for PG amino acids leads to increased strand exchange activity and reduced general fidelity in Rad51. These effects are most severe when both residues are swapped [ 26 ]. Furthermore, valine mutation in loop 2 to alanine has been shown to disrupt the second step in strand invasion [ 62 ]. We generated strains with rad51D332G and rad51V331P, D332G alleles and tested them for their ability to perform homologous and homeologous recombination. We found that in the rad51D332G strains, both homologous and homeologous recombination were unaffected ( Figure 6BC ). In contrast, in the rad51V331P, D332G strains, we observed a four-fold increase in homologous recombination and an eight-fold increase in homeologous recombination, leading to a HER/HR ratio of 2 ( Figure 6BC ). Notably, both strains fully complemented the MMS phenotype of the rad51Δ ( Supplemental Figure 6D ). From these experiments, we conclude that substituting the VD residues in Rad51 to PG leads to lower recombination fidelity. Discussion RecA like recombinases span the entire tree of life. Here, we have evaluated recombinase sequences near the division between Archaea and Eukarya to better understand the protein sequence information for the direct ancestors of Rad51 and Dmc1 within the Asgard Archaea. Our study found that a complete eukaryotic ATP binding site likely evolved in the Asgard Archaea. We also found a diversity of DNA-binding loops in the Asgard Archaea. Analysis of these loops suggests a connection with recombination fidelity. These data indicate that variations in the DNA binding loops may be an evolutionary mechanism for the specialization of recombinases in the Archaea/Eukarya lineage. Finally, we find that the last common Archaeal ancestor of eukaryotes likely had DNA-binding loops that had low fidelity and were more Dmc1-like. This suggests that the DNA binding loops in Rad51 may have been selected for higher fidelity, allowing low and high-fidelity recombinases to exist within a single cell. A general limitation A limitation of our study is that we have used metagenomic assemblies to analyze potential protein sequences from Asgard Archaea. While the assembly of these genomes has undergone quality control, the lack of a genetic model limits our ability to understand how this or if these recombinases work in an organism. For example, we can’t rule out the possibility that some of the sequences we identified may be viral in origin or could be errors in sequencing. Still, we think the prevalence of the sequence variants we identified limits this possibility. We have also expressed and purified examples of recombinases from Asgard Archaea. These are active and have biochemical functions. In conjunction with our stringent selection criteria for using sequences in our analysis, we believe these data represent the analysis of active recombinase enzymes. Recombinase Binding Site II A second DNA binding site is conserved in all RecA recombinases. However, there is a variation in the amino acids that make up binding site II. In E. coli RecA, these residues are R243, K245, and R227 [ 20 ]. Between eubacteria and archaea/eukarya, only R243 and K245 are physically conserved on the recombinase structure. In archaea and eukarya, the K245 is replaced by an Aspartic Acid, which does not form part of DNA binding site II. In eukarya, the third residue is part of the P-loop, and the full binding site residues are ( S.c .Rad51 R188, K361, and K371). Most archaea we analyzed lacked the third positive residue associated with the binding site II in eubacteria and eukaryotic recombinases. The exception to this occurred within the Asgard superphyla. We observed groups with the equivalent of R188 residue in their P-loop. We further found that this site is generally fixed within domains, with eubacteria having a serine, archaea a glycine, and eukarya an arginine. The exception is the Asgard archaea, which appears to be in a transition state between archaea and eukarya. Significantly, this mutation predates the split between Rad51 and Dmc1, which occurred in early eukaryotes [ 42 ]. This means that the last Archaeal ancestor of Rad51 and Dmc1 had an arginine in this position. The specific role of this substitution is unclear. However, the secondary DNA binding site is required for recombination in eukarya and is essential for meiosis [ 20 , 55 ]. Therefore, identifying this residue in another domain of life represents an important step in the evolution of eukaryotic recombinases. Fidelity of recombinases After identifying variant DNA binding loops (L1 and L2) in our initial analysis of Asgard Archaea, we further analyzed DNA binding loops within the domain of Archaea. We focused on a pair of residues, Arginine in L1 and Aspartic Acid in L2, that form a salt bridge between adjacent protomers. Biochemical and structural evidence suggests this interaction regulates fidelity during nucleotide selection. In current models, Rad51 has higher fidelity, and Dmc1 has lower fidelity [ 25 , 26 , 28 , 63 ]. The exact consequence of this difference is unclear, and how this collaborates with the mismatch repair machinery remains unknown. In most archaea, we observed that canonical loops that, by definition, should have higher fidelity. These loops were the most common type found in Archaea and the Eury and Tack superphyla, where greater than 90% of the sequences were canonical. Canonical loops comprised ∼60% and 45% of recombinases in the DPANN and Asgard superphyla, respectively. Both groups also have a high prevalence of Type II loops. Only 34% of sequences were canonical in Archaeal viral genomes, and Type II was the most common loop type. The overall variance of loop types suggests there may be functional differences in these organisms. However, when testing the canonical and Type II DNA binding loops in Saccharomyces cerevisiae, we detected no differences in their mismatch tolerance based on the homeologous to homologous recombination ratio. The caveat to this analysis is that these substituted loops do not fully complement the yeast Rad51 functions, and differences observed in fidelity reflect their activity in the yeast system. Still, this suggests that the ancestral configuration of these loops may be more tolerant of mismatches in general. This could allow them to be more promiscuous in their DNA selection. Traditionally, barriers to homologous pairing between genes and genomes can occur due to mismatch intolerance, controlled by the MutL and MutS family of proteins [ 64 , 65 ]. Archaea also possess a second family called the NucS family, which is thought to function similarly [ 66 ]. MutL/S can promote rejection of incompletely paired DNA substrates [ 67 ] or facilitate post recombination repair of persistent mismatches. However, it is unclear how recombinase proteins may communicate with the MutL/S system to prevent the incorporation of homeologous DNA. One possibility is that it is through interactions with L1 and L2. The Asgard Archaea have been shown to have the MutL/S system [ 66 ]. Future work should be directed to understand better the evolution of communication between recombinases and the mismatch repair system and whether stringency in this interaction has improved with more complex organisms. This does not answer the differences between canonical and Type II DNA binding loops. However, based on our observations, Type II loops may still have a lower fidelity than the canonical loop. Further work would be needed to fully characterize the differences between these loops in Archaea. However, during infection of Haloferax with the Haloarcula virus (HCTV1), a canonical RadA provided by the host and a Type II RadA supplied by the virus are present in the same cell [ 68 , 69 ]. In cases of bacteriophage, when the RecA homolog, UvwX, is present, it is used to initiate viral genome replication [ 46 , 49 ]. This could also be the case with RadA within the viral genome, and Type II loops with lower fidelity may be more efficient at initiating replication. This is highly speculative but would be supported by the incidence of Type II loops in the genomes of viruses and symbiotes. When more precise substitutions were made in the DNA binding loop two, there was no apparent effect on mismatch tolerance when only the salt bridge between loop one (R) and loop two (D) was disrupted. However, when both the R-D and the adjacent L-V interactions were disrupted, there was an increase in homologous recombination and mismatch tolerance. This is consistent with previous biochemical experiments performed with Dmc1 and Rad51, which swapped mismatch tolerance in these recombinases by swapping the VD in Rad51 for the PG in Dmc1[ 26 , 28 ]. Interestingly, in most of the Archaea/Eukarya lineage, the RecA homolog has a Proline instead of a Valine. The exceptions are Rad51 in Eukarya, Members of Group 5 in Asgard Archaea, and Type I and Type III DNA binding loops. The salt bridge is maintained in the cases of Rad51 and Group 5, and proline is replaced with hydrophobic amino acids such as valine, isoleucine, or methionine, which have a greater surface area for base stacking. In the case of the Type I and Type III loops, the salt bridge is also lost. The abundance of the proline within the entire lineage suggests that groups lacking a proline are deviant amongst recombinases and may reflect specialization. In the case of Rad51, this may improve the overall fidelity of the recombinase. Recently, it has been suggested that the valine in loop 2 is essential for recognizing abasic sites by Rad51, making this residue necessary for DNA repair function [ 70 ]. This supports the idea of improved fidelity. The last common Archaeal ancestor The conclusion that the last archaeal ancestor was Dmc1-like ( Figure 6D ) is likely to be controversial as strong evidence supports the transmission of a more Rad51-like recombinase in the earliest eukaryotes, as Paramecium caudatus has the Dmc1 L1 but also has the full Rad51 L2. An example of an early eukaryote with two recombinases is Giardia intestinalis, which has two Dmc1 genes, Dmc1-A and Dmc1-B. These proteins also have a Dmc1 like L1 and a Rad51 - like L2. These organisms suggest that the Type II loops were not directly inherited but may have evolved independently later in eukaryotic evolution or been acquired through horizontal gene transfer events between early eukaryotes and archaea. However, it is also possible that proto-eukaryotes that inherited a Dmc1-like recombinase have gone extinct or have not been discovered. Our model of the direct inheritance of a low-fidelity recombinase suggests that Rad51 may have undergone specialization within the eukaryotic lineage to improve overall fidelity in recombination. This may have been necessary as genomes expanded due to the inclusion of repeated elements or in response to the formation of permanent diploid states. Higher fidelity would be needed to prevent ectopic recombination or recombination between homologs under normal growth conditions. This means that Dmc1 has likely remained a low-fidelity recombinase through neutral selection, as opposed to the positive selection that Rad51 has likely experienced. Despite our efforts to categorize RecA recombinases within the Archaea/Eukarya lineage, there are still two missing pieces of information. The first is the apparent gap between eubacterial RecA and RadA in Archaea, as most recombinases in Archaea had very high sequence similarity to eukaryotes. Divergent sequences observed in the Asgard Archaea lineage are likely deep branching. Therefore, information about the last common ancestor of all RecA recombinases is missing. Likewise, sequence information for the earliest eukaryotes appears missing, as discussed above. Hopefully, these gaps will be filled as additional sequence information is discovered and added to existing databases. Author contributions BF performed all bioinformatic analysis and designed constructs for bacterial expression, PH performed biochemical experiments, analyzed data, and helped write the manuscript. JH performed experiments and analyzed data. JBC provided reagents, guided experiments, and wrote the manuscript with input from BF, PH, and JH. NIGMS R35142457 supported this work to JBC and Cornell Startup funds to JBC. Methods and Materials Retrieval and Phylogenetic Analysis of Asgard Recombinase Sequences Analysis of recombinase diversity within the Asgard Archaea Amino acid sequences of putative Asgard recombinases were obtained using BLASTp with S. cerevisiae Rad51 (Uniprot P25454), H. sapiens Rad51 (Uniprot Q06609), and S. solfataricus RadA (Uniprot Q55075) sequences as queries to the Asgard Group database (TaxId: 1935183). The expect value was 0.05 for all searches, and the maximum number of sequences was set to 5,000; for each search, less than 600 hits were returned, all were combined, and duplicate hits were removed. From this set of 653 unique sequences, those shorter than 150 (n=20) or longer than 500 (n=5) amino acids were removed, along with 38 sequences that appeared to be partial/incomplete, contained large insertions, or had non-amino acid characters; the final number of sequences used for analysis was 590. Next, we added RecA ( E. coli ) as an outgroup, as well as six RadB reference sequences to aid in the identification of potential RadB homologs in the Asgard group. Sequences were aligned using MUSCLE and manually trimmed to span only the core RecA domain (I41 to G268, E. coli numbering). After manual trimming, the alignment was automatically trimmed with BMGE with default settings and a minimum block size of 1. The trimmed alignment was then used to generate a maximum likelihood phylogenetic tree to reduce the total sequences in our analysis with Treemmer. For this tree, IQ-tree’s Model Finder selected the LG+R8 substitution model based on the best fit of the Bayesian information criterion. Next, Treemmer was used to reduce the sequences in our dataset while maintaining 95% of the overall tree length. After adding back RadB reference sequences (n = 6) and BLAST hits from Asgard members manually identified as having putative species/strain designations (n = 25), the final number of sequences was 373. The full-length sequences in this dataset were then realigned with muscle, manually trimmed to the core RecA domain using the E. coli RecA outgroup as a reference and automatically trimmed with BMGE following the same processes and parameters described above. This dataset’s maximum likelihood phylogenetic tree ( Figure 1A ) was constructed in IQ-Tree with 1000 SH-aLRT [ 71 ] and 1000 UFBS [ 72 ] replicates and the LG + R5 substitution model selected by ModelFinder. We then analyzed and annotated the tree ( Figure 1A ). To create the RadA-like subtree ( Figure 1B ), the 20 RadB-like hits and 6 RadB reference sequences were first removed from the data set. The remaining 347 sequences were processed following the same steps as described above for sequences that were aligned in MUSCLE, manually trimmed to RecA-core domain, trimmed with BMGE, and used to generate a maximum-likelihood tree with IQtree with E. coli RecA as the outgroup. The substitution model used was LG + R6 as selected by IQ-Tree’s Model Finder. Eighteen groups of putative RadA homologs were assigned based on the criteria that their clade contained four or more sequences and had SH-aLRT/UFBS branch support of >80 and >90, respectively. Of these initial 18 groups, four (groups 2, 3, 14, and 15) lacked clear Walker A, DNA binding Loop 1 and/or DNA binding Loop 2 motifs and were not analyzed further as they are unlikely to represent RadA homologs; similarly, nine individual sequences from the remaining 14 groups also appeared to lack canonical recombinase motifs, or were incomplete, and were removed from our analysis, resulting in 198 Asgard recombinase sequences across 14 groups. Collection and analysis of recombinase reference sequences from Archaea, Bacteria, and Eukarya Non-Asgard Archaea (RadA) Sequences for RadA homologs from non-Asgard Archaea were obtained following the same procedure used to recover homologs from the Asgard superphylum. Briefly, S. solfataricus RadA, S. cerevisiae Rad51, and H. sapiens Rad51 were used as queries in BLASTp searches against each archaeal superphyla (TACK Taxid: 1783275; Euryarchaeota (EURY) Taxid: 28890; DPANN Taxid: 1783276). The E-value for each search was 0.05, and the top 500 hits were collected, combined, and duplicate sequences removed. Outlier sequences >500 amino acids long, missing Walker A/B motifs, or containing X characters were removed from each data set. Five outliers were removed for TACK, and the final number of sequences used for all analyses was 758. For EURY, 13 outliers were removed, and the final number of sequences used for all analyses was 782. For DPANN, two outliers were removed, and the final number of sequences used for all analyses was 621. For each superphylum, sequences were aligned with MUSCLE using the default settings in Jalview; logos were generated with Skylign ( http://skylign.org ). Bacteria (RecA) Eighty bacterial RecA sequences spanning 40 bacterial phyla ( https://lpsn.dsmz.de/text/names-of-phyla )[ 73 ] were obtained from Uniprot and aligned with MUSCLE using the default settings in Jalview; logos were generated with Skylign ( http://skylign.org ). Eukarya (Rad51 and DMC1) The 43 Rad51 and 28 DMC1 sequences [ 42 ] used in phylogenetic analyses were obtained from Uniprot and aligned with MUSCLE using the default settings in Jalview; logos were generated with Skylign ( http://skylign.org ). Combined alignment and phylogenetic analysis of bacterial, archaeal, and eukaryotic recombinases A combined dataset (N total = 2511) of Asgard (N = 198), TACK (N = 758), EURY (N = 782), eukaryotic Rad51 (N = 43), eukaryotic Dmc1 (N = 28), and bacterial RecA (N = 80) were aligned with E. coli RecA using MAFFT [ 74 , 75 ] ( https://mafft.cbrc.jp/alignment/server/index.html ) with the G-INS-1 progressive method strategy and default settings. The alignment was trimmed manually to the core RecA domain and with BMGE as previously described for all other alignments. IQ-tree was used to generate a maximum likelihood phylogenetic tree with 1,000 SH-aLRT and 1,000 UFBS replicates using the LG + R10 substitution model selected by ModelFinder. The tree was visualized and annotated with FigTree and Adobe Illustrator. Structural modeling and comparison Alphafold models of Asgard recombinases were obtained from UniProt when available. For MCD648379 (Group 8) and MDO8063530 (Group 10), no models were associated with their respective UniProt entries and were therefore modeled using AlphaFold2 with ColabFold in ChimeraX (version 1.8dev202404120015). Superpostion of recombinase experimental structures and models was carried out with the super command in PyMol (Open source, version 2.5.0); the cryo-EM structure of the post-synaptic filament of human Dmc1 was used as a common reference for superpositions unless otherwise noted. For AlphaFold predicted structures of Asgard recombinases, arginine rotamers at sites equivalent to R235 in human Rad51 and R236 in human Dmc1 were manually adjusted to match the orientation of the corresponding sidechains in the empirical structures of these human recombinase filaments. Collection and analysis of recombinase sequences from archaeal viruses To identify recombinases associated with archaeal viruses, we used BLASTp to search IMG [ 76 ] ( https://img.jgi.doe.gov ) and NCBI databases with hDmc1 as the query. For the IMG search, we queried the viral protein database [ 77 ] with an E-value of 1E-5, limiting the maximum number of hits to 500. Next, we extracted all hits (N = 56) with “high” or “medium” confidence as defined by IMG [ 77 ]. Sequences that were greater than 500 amino acids in length (N = 4) or shared identical vOTU’s and were 100% identical (N = 17) were removed, resulting in a final IMG-derived dataset of 35 viral RadA sequences. For our NCBI searches, we queried four different taxonomic IDs with the default E-value cutoff of 0.05 and limited our results to a maximum of 500 hits: TaxID 2136008 (Archaeal Viruses), TaxID 1960247 (Uncultured Archaeal viruses), TaxID 451344 (Uncultured Archaeal Viruses), and TaxID 2731619 (bacterial and archaeal viruses with head-tail morphology). Only the last database (TaxID: 451344) returned any hits (N = 23); the top four were annotated as RadA-type recombinases associated with the Euryarchaea Haloarcua. The remaining 19 sequences were annotated as RecA-like and not included in our subsequent analyses as they likely represent bacterial viruses. Finally, we combined the 35 archaeal virus sequences from IMG with the four Haloarchaea virus sequences from NCBI. We generated a multiple-sequence alignment using Muscle with default settings in JalView that included reference sequences for eukaryotic Rad51 and Dmc1, archaeal RadA and RadB, and bacterial RecA. Expression and Purification of Asgard Archaea Recombinases Genes sequences predicted to encode recombinases Group 7 ( Ca. Thorarchaea , OLS31719.1), Group 11 ( Ca. Lokiarchaea, NHJ20009.1), Group 13 ( Ca. Lokiarchaea, NHI92944.1), and Group 17 ( Ca. Heimdallarchaea, TET27405.1) were codon optimized for E. coli and cloned into pET28a with an N-terminal 6xHis-SUMO tag. For expression, 1-L of LB was grown to an OD of 0.4-0.6, and expression was induced with 0.5 mM IPTG for 3 hours at 37 °C. Cells were then harvested and washed with 25 ml of Buffer A(30 mM Tris-Cl pH 8.0, 250 mM NaCl, 0.05% Igepal-680, 10 mM Imidazole 10% Glycerol, and protease inhibitor cocktail). Cells were stored as pellets at −80°C. Cell pellets were resuspended in 30 ml/L of resuspension buffer and thawed. The lysed cells were then sonicated with 15 sec on 45 sec at 65% duty cycle for 10 cycles. After sonication, the slurry was centrifuged at 15,000 rpm (10,000 g) for 45 min. The whole cell extract (WCE) was filtered through a 0.22 µ PDVF filter. The filtered WCE was then adjusted to 20% Ammonium Sulfate and centrifuged at 10,000 g for 10 min. The Ammonium Sulfate concentration was then adjusted to 50% and centrifuged at 10,000 g for 15 min. The pellet was resuspended in 40 ml of buffer A and incubated for 1 hour with 1 ml of settled Ni-NTA resin. The resin was washed with 10 CV of Buffer A. The resin was loaded onto a disposable column, and the protein was eluted with 4 ml of Buffer A + 200 mM Imidazole. The peak fractions were mixed with 6xHis-Ulp1 protease and dialyzed overnight in buffer A. The dialysate was bound to Ni-NTA resin to remove the Ulp1 protease and the SUMO tag. The flowthrough was collected, and the protein concentrated and stored in single use aliquots in the −80 C°. ATP hydrolysis assay A commercially available ADP-GLO kit was used to measure ATP hydrolysis rates. The ATP hydrolysis reaction was performed in HR buffer (20 mM Tris-Cl, 50 mM NaCl, 10 mM MgOAc 2 , 200 ng/µl BSA, 5 mM BME, 100 mM ATP, and 10% Glycerol) and contained 0.1 mg/ml of Phi X circular ssDNA and one µM recombinase. Reactions were conducted at 30°C, 37°C, and 45°C for 20, 40, and 60 min. Chemiluminescence was measured using an Azure 400 Fluorescent Imager. Electromobility shift assay Phi X circular ssDNA was diluted to 15 nM and incubated with 0.3, 1, and 3 µM of recombinase protein in HR buffer (20 mM Tris-Cl, 100 mM NaCl, 4 mM MgOAc 2 , 25 µg/µl BSA, 1 mM DTT, and 10% Glycerol) with or without 2.5 mM ATP at 37°C for 15 min. The samples were then electrophoresed on a 0.8% Agarose gel for 50 min at 120 V in 0.5 x TAE buffer (10 mM Tris, 5 mM acetic acid, 0.25 mM EDTA) and stained with SYBR TM Gold for 30 min. The gel was then imaged with an Azure 400 Fluorescent Imager. Yeast strains All yeast strains used in this study are present in Supplemental Table 1 and Supplemental Table 2 Serial Dilution Spot Assay For complementation spot assays, overnight cultures were diluted back to an OD 600 of 0.2 and then allowed to grow to an OD 600 of 1.0. Cells were then serially diluted and manually spotted on YPD plates containing no drug and 0.0005 % MMS. The plates were incubated at 30°C for 2-3 days and imaged at 24, 48, and 72 hours. Inverted repeat recombination Individual colonies were picked from a YP+2% Dextrose plate and grown for 2 days in YPGG (4% Galactose, 2% Glycerol). After 2 days, appropriate dilutions of cells were plated on Synthetic Complete media (4% Galactose) or Synthetic media (-His, 4% Galactose) and allowed to grow for three days. After three days, the colonies were counted. The recombination frequency and the homeologous to homologous ratio were calculated as previously described [ 61 ]. Acknowledgments We would like to acknowledge Thorsten Allers and John Logsdon Jr. for their comments on the study, Eric Alani for carefully reading the manuscript, and members of the Crickard laboratory for reading it. We would also like to thank members of the Cornell R3 groups for their comments during the development of this project. References 1. ↵ González-Torres P , Rodríguez-Mateos F , Antón J , Gabaldón T . Impact of Homologous Recombination on the Evolution of Prokaryotic Core Genomes . mBio . 2019 ; 10 ( 1 ) : doi: 10.1128/mbio.02494-18 . doi: doi:10.1128/mbio.02494-18. OpenUrl CrossRef 2. ↵ Kowalczykowski SC . An Overview of the Molecular Mechanisms of Recombinational DNA Repair . Cold Spring Harb Perspect Biol . 2015 ; 7 ( 11 ). Epub 2015/11/04. doi: 10.1101/cshperspect.a016410 . PubMed PMID: 26525148 ; PubMed Central PMCID: PMCPMC4632670 . OpenUrl Abstract / FREE Full Text 3. Symington LS , Rothstein R , Lisby M . Mechanisms and Regulation of Mitotic Recombination in Saccharomyces Cerevisiae . Genetics . 2014 ; 198 ( 3 ): 795 – 835 . Epub 2014/11/09. doi: 10.1534/genetics.114.166140 . PubMed PMID: 25381364 ; PubMed Central PMCID: PMCPMC4224172 . OpenUrl Abstract / FREE Full Text 4. ↵ Filippo JS , Sung P , Klein H . Mechanism of Eukaryotic Homologous Recombination . Annual Review of Biochemistry . 2008 ; 77 ( 1 ): 229 – 57 . doi: 10.1146/annurev.biochem.77.061306.125255 . PubMed PMID: 18275380 . OpenUrl CrossRef PubMed Web of Science 5. ↵ Andersson DI , Hughes D . Muller’s Ratchet Decreases Fitness of a DNA-Based Microbe . Proceedings of the National Academy of Sciences . 1996 ; 93 ( 2 ): 906 – 7 . doi : doi: 10.1073/pnas.93.2.906 . OpenUrl Abstract / FREE Full Text 6. Felsenstein J . The Evolutionary Advantage of Recombination . Genetics . 1974 ; 78 ( 2 ): 737 – 56 . Epub 1974/10/01. doi: 10.1093/genetics/78.2.737 . PubMed PMID: 4448362 ; PubMed Central PMCID: PMCPMC1213231 . OpenUrl Abstract / FREE Full Text 7. Colnaghi M , Lane N , Pomiankowski A . Genome Expansion in Early Eukaryotes Drove the Transition from Lateral Gene Transfer to Meiotic Sex . eLife . 2020 ; 9 : e58873 . doi: 10.7554/eLife.58873 . OpenUrl CrossRef PubMed 8. ↵ Mirzaghaderi G , Hörandl E . The Evolution of Meiotic Sex and Its Alternatives . Proceedings of the Royal Society B: Biological Sciences . 2016 ; 283 ( 1838 ): 20161221 . doi : doi: 10.1098/rspb.2016.1221 . OpenUrl CrossRef PubMed 9. ↵ Spell RM , Jinks-Robertson S . Examination of the Roles of Sgs1 and Srs2 Helicases in the Enforcement of Recombination Fidelity in Saccharomyces Cerevisiae . Genetics . 2004 ; 168 ( 4 ): 1855 – 65 . doi: 10.1534/genetics.104.032771 . OpenUrl Abstract / FREE Full Text 10. ↵ Szostak JW , Orr-Weaver TL , Rothstein RJ , Stahl FW . The Double-Strand-Break Repair Model for Recombination . Cell . 1983 ; 33 ( 1 ): 25 – 35 . doi: 10.1016/0092-8674(83)90331-8 . OpenUrl CrossRef PubMed Web of Science 11. ↵ Jones G , Kleckner N , Zickler D . Meiosis through Three Centuries . Chromosoma . 2024 ; 133 ( 2 ): 93 – 115 . Epub 2024/05/11. doi: 10.1007/s00412-024-00822-0 . PubMed PMID: 38730132 ; PubMed Central PMCID: PMCPMC11180163 . OpenUrl CrossRef PubMed 12. Hunter N. Meiotic Recombination: The Essence of Heredity . Cold Spring Harb Perspect Biol . 2015 ; 7 ( 12 ). Epub 2015/10/30. doi: 10.1101/cshperspect.a016618 . PubMed PMID: 26511629 ; PubMed Central PMCID: PMCPMC4665078 . OpenUrl Abstract / FREE Full Text 13. ↵ Murakami H , Keeney S . Regulating the Formation of DNA Double-Strand Breaks in Meiosis . Genes Dev . 2008 ; 22 ( 3 ): 286 – 92 . Epub 2008/02/05. doi: 10.1101/gad.1642308 . PubMed PMID: 18245442 ; PubMed Central PMCID: PMCPMC2731648 . OpenUrl FREE Full Text 14. ↵ Heyer WD . Biochemistry of Eukaryotic Homologous Recombination . Top Curr Genet . 2007 ; 17 : 95 – 133 . Epub 2007/03/01. doi: 10.1007/978-3-540-71021-9 . PubMed PMID: 21552479 ; PubMed Central PMCID: PMCPMC3087616 . OpenUrl CrossRef PubMed 15. ↵ Bell JC , Kowalczykowski SC . Reca: Regulation and Mechanism of a Molecular Search Engine . Trends Biochem Sci . 2016 ; 41 ( 6 ): 491 – 507 . Epub 2016/05/09. doi: 10.1016/j.tibs.2016.04.002 . PubMed PMID: 27156117 ; PubMed Central PMCID: PMCPMC4892382 . OpenUrl CrossRef PubMed 16. ↵ Yang H , Zhou C , Dhar A , Pavletich NP . Mechanism of Strand Exchange from Reca-DNA Synaptic and D-Loop Structures . Nature . 2020 ; 586 (7831): 801 -6. Epub 2020/10/16. doi: 10.1038/s41586-020-2820-9 . PubMed PMID: 33057191 ; PubMed Central PMCID: PMCPMC8366275 . OpenUrl CrossRef PubMed 17. ↵ Yang H , Pavletich NP . Insights into Homology Search from Cryo-Em Structures of Reca-DNA Recombination Intermediates . Curr Opin Genet Dev . 2021 ; 71 : 188 – 94 . Epub 2021/10/01. doi: 10.1016/j.gde.2021.09.002 . PubMed PMID: 34592688 ; PubMed Central PMCID: PMCPMC8671187 . OpenUrl CrossRef PubMed 18. ↵ Ito K , Murayama Y , Kurokawa Y , Kanamaru S , Kokabu Y , Maki T , et al. Real-Time Tracking Reveals Catalytic Roles for the Two DNA Binding Sites of Rad51 . Nat Commun . 2020 ; 11 ( 1 ): 2950 . Epub 2020/06/13. doi: 10.1038/s41467-020-16750-3 . PubMed PMID: 32528002 ; PubMed Central PMCID: PMCPMC7289862 . OpenUrl CrossRef PubMed 19. De Vlaminck I , van Loenhout MT , Zweifel L , den Blanken J , Hooning K , Hage S , et al. Mechanism of Homology Recognition in DNA Recombination from Dual-Molecule Experiments . Mol Cell . 2012 ; 46 ( 5 ): 616 – 24 . Epub 2012/05/09. doi: 10.1016/j.molcel.2012.03.029 . PubMed PMID: 22560720 . OpenUrl CrossRef PubMed Web of Science 20. ↵ Cloud V , Chan YL , Grubb J , Budke B , Bishop DK . Rad51 Is an Accessory Factor for Dmc1-Mediated Joint Molecule Formation During Meiosis . Science . 2012 ; 337 (6099): 1222 -5. Epub 2012/09/08. doi: 10.1126/science.1219379 . PubMed PMID: 22955832 ; PubMed Central PMCID: PMCPMC4056682 . OpenUrl Abstract / FREE Full Text 21. ↵ Peacock-Villada A , Yang D , Danilowicz C , Feinstein E , Pollock N , McShan S , et al. Complementary Strand Relocation May Play Vital Roles in Reca-Based Homology Recognition . Nucleic Acids Res . 2012 ; 40 ( 20 ): 10441 – 51 . Epub 2012/09/04. doi: 10.1093/nar/gks769 . PubMed PMID: 22941658 ; PubMed Central PMCID: PMCPMC3488227 . OpenUrl CrossRef PubMed Web of Science 22. ↵ Mason JM , Chan Y-L , Weichselbaum RW , Bishop DK . Non-Enzymatic Roles of Human Rad51 at Stalled Replication Forks . Nature Communications . 2019 ; 10 ( 1 ): 4410 . doi: 10.1038/s41467-019-12297-0 . OpenUrl CrossRef PubMed 23. Liu W , Saito Y , Jackson J , Bhowmick R , Kanemaki MT , Vindigni A , et al. Rad51 Bypasses the Cmg Helicase to Promote Replication Fork Reversal . Science . 2023 ; 380 ( 6643 ): 382 - 7 . doi : doi: 10.1126/science.add7328 . OpenUrl CrossRef PubMed 24. ↵ Wassing IE , Esashi F . Rad51: Beyond the Break . Seminars in Cell & Developmental Biology . 2021 ; 113 : 38 – 46 . doi: 10.1016/j.semcdb.2020.08.010 . OpenUrl CrossRef PubMed 25. ↵ Lee JY , Terakawa T , Qi Z , Steinfeld JB , Redding S , Kwon Y , et al. DNA Recombination. Base Triplet Stepping by the Rad51/Reca Family of Recombinases . Science . 2015 ; 349 ( 6251 ): 977 - 81 . Epub 2015/09/01. doi: 10.1126/science.aab2666 . PubMed PMID: 26315438 ; PubMed Central PMCID: PMCPMC4580133 . OpenUrl Abstract / FREE Full Text 26. ↵ Xu J , Zhao L , Peng S , Chu H , Liang R , Tian M , et al. Mechanisms of Distinctive Mismatch Tolerance between Rad51 and Dmc1 in Homologous Recombination . Nucleic Acids Res . 2021 ; 49 ( 22 ): 13135 – 49 . Epub 2021/12/07. doi: 10.1093/nar/gkab1141 . PubMed PMID: 34871438 ; PubMed Central PMCID: PMCPMC8682777 . OpenUrl CrossRef PubMed 27. Li WC , Lee CY , Lan WH , Woo TT , Liu HC , Yeh HY , et al. Trichoderma Reesei Rad51 Tolerates Mismatches in Hybrid Meiosis with Diverse Genome Sequences . Proc Natl Acad Sci U S A . 2021 ; 118 ( 8 ). Epub 2021/02/18. doi: 10.1073/pnas.2007192118 . PubMed PMID: 33593897 ; PubMed Central PMCID: PMCPMC7923544 . OpenUrl Abstract / FREE Full Text 28. ↵ Luo SC , Yeh HY , Lan WH , Wu YM , Yang CH , Chang HY , et al. Identification of Fidelity-Governing Factors in Human Recombinases Dmc1 and Rad51 from Cryo-Em Structures . Nat Commun . 2021 ; 12 ( 1 ): 115 . Epub 2021/01/16. doi: 10.1038/s41467-020-20258-1 . PubMed PMID: 33446654 ; PubMed Central PMCID: PMCPMC7809202 . OpenUrl CrossRef PubMed 29. ↵ Borgogno MV , Monti MR , Zhao W , Sung P , Argaraña CE , Pezza RJ . Tolerance of DNA Mismatches in Dmc1 Recombinase-Mediated DNA Strand Exchange . J Biol Chem . 2016 ; 291 ( 10 ): 4928 – 38 . Epub 2015/12/29. doi: 10.1074/jbc.M115.704718 . PubMed PMID: 26709229 ; PubMed Central PMCID: PMCPMC4777831 . OpenUrl Abstract / FREE Full Text 30. ↵ Bishop DK , Williamson MS , Fogel S , Kolodner RD . The Role of Heteroduplex Correction in Gene Conversion in Saccharomyces Cerevisiae . Nature . 1987 ; 328 ( 6128 ): 362 - 4 . Epub 1987/07/23. doi: 10.1038/328362a0 . PubMed PMID: 3299108 . OpenUrl CrossRef PubMed 31. ↵ Alani E , Reenan RA , Kolodner RD . Interaction between Mismatch Repair and Genetic Recombination in Saccharomyces Cerevisiae . Genetics . 1994 ; 137 ( 1 ): 19 – 39 . Epub 1994/05/01. doi: 10.1093/genetics/137.1.19 . PubMed PMID: 8056309 ; PubMed Central PMCID: PMCPMC1205935 . OpenUrl Abstract / FREE Full Text 32. ↵ Ramesh MA , Malik S-B , Logsdon JM , Jr . . A Phylogenomic Inventory of Meiotic Genes: Evidence for Sex in Giardia and an Early Eukaryotic Origin of Meiosis . Current Biology . 2005 ; 15 ( 2 ): 185 – 91 . doi: 10.1016/j.cub.2005.01.003 . OpenUrl CrossRef PubMed Web of Science 33. ↵ Zaremba-Niedzwiedzka K , Caceres EF , Saw JH , Bäckström D , Juzokaite L , Vancaester E , et al. Asgard Archaea Illuminate the Origin of Eukaryotic Cellular Complexity . Nature . 2017 ; 541 ( 7637 ): 353 - 8 . doi: 10.1038/nature21031 . OpenUrl CrossRef PubMed 34. ↵ Liu Y , Makarova KS , Huang W-C , Wolf YI , Nikolskaya AN , Zhang X , et al. Expanded Diversity of Asgard Archaea and Their Relationships with Eukaryotes . Nature . 2021 ; 593 ( 7860 ): 553 - 7 . doi: 10.1038/s41586-021-03494-3 . OpenUrl CrossRef PubMed 35. ↵ Haldenby S , White MF , Allers T . Reca Family Proteins in Archaea: Rada and Its Cousins . Biochem Soc Trans . 2009 ; 37 (Pt 1 ): 102 – 7 . Epub 2009/01/16. doi: 10.1042/bst0370102 . PubMed PMID: 19143611 . OpenUrl Abstract / FREE Full Text 36. ↵ Wu Y , Qian X , He Y , Moya IA , Luo Y . Crystal Structure of an Atpase-Active Form of Rad51 Homolog from Methanococcus Voltae: Insights into Potassium Dependence* . Journal of Biological Chemistry . 2005 ; 280 ( 1 ): 722 – 8 . doi: 10.1074/jbc.M411093200 . OpenUrl Abstract / FREE Full Text 37. Seitz EM , Brockman JP , Sandler SJ , Clark AJ , Kowalczykowski SC . Rada Protein Is an Archaeal Reca Protein Homolog That Catalyzes DNA Strand Exchange . Genes Dev . 1998 ; 12 ( 9 ): 1248 – 53 . Epub 1998/06/06. doi: 10.1101/gad.12.9.1248 . PubMed PMID: 9573041 ; PubMed Central PMCID: PMCPMC316774 . OpenUrl Abstract / FREE Full Text 38. ↵ McIlwraith MJ , Hall DR , Stasiak AZ , Stasiak A , Wigley DB , West SC . Rada Protein from Archaeoglobus Fulgidus Forms Rings, Nucleoprotein Filaments and Catalyses Homologous Recombination . Nucleic Acids Res . 2001 ; 29 ( 22 ): 4509 – 17 . Epub 2001/11/20. doi: 10.1093/nar/29.22.4509 . PubMed PMID: 11713300 ; PubMed Central PMCID: PMCPMC92570 . OpenUrl CrossRef PubMed Web of Science 39. ↵ Hawkins M , Malla S , Blythe MJ , Nieduszynski CA , Allers T . Accelerated Growth in the Absence of DNA Replication Origins . Nature . 2013 ; 503 ( 7477 ): 544 - 7 . Epub 2013/11/05. doi: 10.1038/nature12650 . PubMed PMID: 24185008 ; PubMed Central PMCID: PMCPMC3843117 . OpenUrl CrossRef PubMed Web of Science 40. ↵ Allers T , Ngo H-P . Genetic Analysis of Homologous Recombination in Archaea: Haloferax Volcanii as a Model Organism . Biochemical Society Transactions . 2003 ; 31 ( 3 ): 706 – 10 . doi: 10.1042/bst0310706 . OpenUrl Abstract / FREE Full Text 41. ↵ Rein HL , Bernstein KA , Baldock RA . Rad51 Paralog Function in Replicative DNA Damage and Tolerance . Curr Opin Genet Dev . 2021 ; 71 : 86 – 91 . Epub 2021/07/27. doi: 10.1016/j.gde.2021.06.010 . PubMed PMID: 34311385 ; PubMed Central PMCID: PMCPMC8671181 . OpenUrl CrossRef PubMed 42. ↵ Lin Z , Kong H , Nei M , Ma H . Origins and Evolution of the Reca/Rad51 Gene Family: Evidence for Ancient Gene Duplication and Endosymbiotic Gene Transfer . Proc Natl Acad Sci U S A . 2006 ; 103 ( 27 ): 10328 – 33 . Epub 2006/06/27. doi: 10.1073/pnas.0604232103 . PubMed PMID: 16798872 ; PubMed Central PMCID: PMCPMC1502457 . OpenUrl Abstract / FREE Full Text 43. ↵ Wardell K , Haldenby S , Jones N , Liddell S , Ngo GHP , Allers T . Radb Acts in Homologous Recombination in the Archaeon Haloferax Volcanii , Consistent with a Role as Recombination Mediator. DNA Repair (Amst ). 2017 ; 55 : 7 – 16 . Epub 2017/05/16. doi: 10.1016/j.dnarep.2017.04.005 . PubMed PMID: 28501701 ; PubMed Central PMCID: PMCPMC5480776 . OpenUrl CrossRef PubMed 44. ↵ Spang A , Eme L , Saw JH , Caceres EF , Zaremba-Niedzwiedzka K , Lombard J , et al. Asgard Archaea Are the Closest Prokaryotic Relatives of Eukaryotes . PLoS Genet . 2018 ; 14 ( 3 ): e1007080 . Epub 2018/03/30. doi: 10.1371/journal.pgen.1007080 . PubMed PMID: 29596421 ; PubMed Central PMCID: PMCPMC5875740 . OpenUrl CrossRef PubMed 45. ↵ Eme L , Tamarit D , Caceres EF , Stairs CW , De Anda V , Schön ME , et al. Inference and Reconstruction of the Heimdallarchaeial Ancestry of Eukaryotes . Nature . 2023 ; 618 (7967): 992 -9. doi: 10.1038/s41586-023-06186-2 . OpenUrl CrossRef PubMed 46. ↵ Liu J , Berger CL , Morrical SW . Kinetics of Presynaptic Filament Assembly in the Presence of Single-Stranded DNA Binding Protein and Recombination Mediator Protein . Biochemistry . 2013 ; 52 ( 45 ): 7878 – 89 . Epub 2013/10/16. doi: 10.1021/bi401060p . PubMed PMID: 24124995 ; PubMed Central PMCID: PMCPMC3864638 . OpenUrl CrossRef PubMed 47. ↵ Gataulin DV , Carey JN , Li J , Shah P , Grubb JT , Bishop DK . The Atpase Activity of E. Coli Reca Prevents Accumulation of Toxic Complexes Formed by Erroneous Binding to Undamaged Double Stranded DNA . Nucleic Acids Res . 2018 ; 46 ( 18 ): 9510 – 23 . Epub 2018/08/24. doi: 10.1093/nar/gky748 . PubMed PMID: 30137528 ; PubMed Central PMCID: PMCPMC6182174 . OpenUrl CrossRef PubMed 48. ↵ Reitz D , Chan YL , Bishop DK . How Strand Exchange Protein Function Benefits from Atp Hydrolysis . Curr Opin Genet Dev . 2021 ; 71 : 120 – 8 . Epub 2021/08/04. doi: 10.1016/j.gde.2021.06.016 . PubMed PMID: 34343922 ; PubMed Central PMCID: PMCPMC8671154 . OpenUrl CrossRef PubMed 49. ↵ Liu J , Ehmsen KT , Heyer W-D , Morrical SW . Presynaptic Filament Dynamics in Homologous Recombination and DNA Repair . Critical Reviews in Biochemistry and Molecular Biology . 2011 ; 46 ( 3 ): 240 – 70 . doi: 10.3109/10409238.2011.576007 . OpenUrl CrossRef PubMed Web of Science 50. ↵ Tsubouchi H , Argunhan B , Ito K , Takahashi M , Iwasaki H . Two Auxiliary Factors Promote Dmc1-Driven DNA Strand Exchange Via Stepwise Mechanisms . Proceedings of the National Academy of Sciences . 2020 ; 117 ( 22 ): 12062 – 70 . doi : doi: 10.1073/pnas.1917419117 . OpenUrl Abstract / FREE Full Text 51. Sullivan MR , Bernstein KA . Rad-Ical New Insights into Rad51 Regulation . Genes (Basel ). 2018 ; 9 ( 12 ). Epub 2018/12/16. doi: 10.3390/genes9120629 . PubMed PMID: 30551670 ; PubMed Central PMCID: PMCPMC6316741 . OpenUrl CrossRef PubMed 52. ↵ Crickard JB , Greene EC . Biochemical Attributes of Mitotic and Meiotic Presynaptic Complexes . DNA Repair (Amst ). 2018 ; 71 : 148 – 57 . Epub 2018/09/10. doi: 10.1016/j.dnarep.2018.08.018 . PubMed PMID: 30195641 ; PubMed Central PMCID: PMCPMC6340751 . OpenUrl CrossRef PubMed 53. ↵ Romero Romero ML , Yang F , Lin Y-R , Toth-Petroczy A , Berezovsky IN , Goncearenco A , et al. Simple yet Functional Phosphate-Loop Proteins . Proceedings of the National Academy of Sciences . 2018 ; 115 ( 51 ): E11943 – E50 . doi : doi: 10.1073/pnas.1812400115 . OpenUrl Abstract / FREE Full Text 54. ↵ Walker JE , Saraste M , Runswick MJ , Gay NJ . Distantly Related Sequences in the Alpha- and Beta-Subunits of Atp Synthase, Myosin, Kinases and Other Atp-Requiring Enzymes and a Common Nucleotide Binding Fold . Embo j . 1982 ; 1 ( 8 ): 945 – 51 . Epub 1982/01/01. doi: 10.1002/j.1460-2075.1982.tb01276.x . PubMed PMID: 6329717 ; PubMed Central PMCID: PMCPMC553140 . OpenUrl CrossRef PubMed Web of Science 55. ↵ Petiot V , White CI , Da Ines O . DNA-Binding Site Ii Is Required for Rad51 Recombinogenic Activity in Arabidopsis Thaliana . Life Sci Alliance . 2024 ; 7 ( 8 ). Epub 2024/05/28. doi: 10.26508/lsa.202402701 . PubMed PMID: 38803223 ; PubMed Central PMCID: PMCPMC11106524 . OpenUrl Abstract / FREE Full Text 56. ↵ Castelle CJ , Banfield JF . Major New Microbial Groups Expand Diversity and Alter Our Understanding of the Tree of Life . Cell . 2018 ; 172 ( 6 ): 1181 – 97 . doi: 10.1016/j.cell.2018.02.016 . OpenUrl CrossRef PubMed 57. Dombrowski N , Lee JH , Williams TA , Otre P , Spang A. Genomic Diversity, Lifestyles and Evolutionary Origins of Dpann Archaea . FEMS Microbiol Lett . 2019 ; 366 ( 2 ). Epub 2019/01/11. doi: 10.1093/femsle/fnz008 . PubMed PMID: 30629179 ; PubMed Central PMCID: PMCPMC6349945 . OpenUrl CrossRef PubMed 58. Dombrowski N , Williams TA , Sun J , Woodcroft BJ , Lee J-H , Minh BQ , et al. Undinarchaeota Illuminate Dpann Phylogeny and the Impact of Gene Transfer on Archaeal Evolution . Nature Communications . 2020 ; 11 ( 1 ): 3939 . doi: 10.1038/s41467-020-17408-w . OpenUrl CrossRef PubMed 59. ↵ Sakai HD , Nur N , Kato S , Yuki M , Shimizu M , Itoh T , et al. Insight into the Symbiotic Lifestyle of Dpann Archaea Revealed by Cultivation and Genome Analyses . Proceedings of the National Academy of Sciences . 2022 ; 119 ( 3 ): e2115449119 . doi : doi: 10.1073/pnas.2115449119 . OpenUrl Abstract / FREE Full Text 60. ↵ Spell RM , Jinks-Robertson S . Role of Mismatch Repair in the Fidelity of Rad51- and Rad59-Dependent Recombination in Saccharomyces Cerevisiae . Genetics . 2003 ; 165 ( 4 ): 1733 – 44 . Epub 2004/01/06. doi: 10.1093/genetics/165.4.1733 . PubMed PMID: 14704162 ; PubMed Central PMCID: PMCPMC1462911 . OpenUrl Abstract / FREE Full Text 61. ↵ Lee SD , Surtees JA , Alani E . Saccharomyces Cerevisiae Msh2-Msh3 and Msh2-Msh6 Complexes Display Distinct Requirements for DNA Binding Domain I in Mismatch Recognition . J Mol Biol . 2007 ; 366 ( 1 ): 53 – 66 . Epub 2006/12/13. doi: 10.1016/j.jmb.2006.10.099 . PubMed PMID: 17157869 ; PubMed Central PMCID: PMCPMC1805781 . OpenUrl CrossRef PubMed Web of Science 62. ↵ Ito K , Murayama Y , Kurokawa Y , Kanamaru S , Kokabu Y , Maki T , et al. Real-Time Tracking Reveals Catalytic Roles for the Two DNA Binding Sites of Rad51 . Nature Communications . 2020 ; 11 ( 1 ): 2950 . doi: 10.1038/s41467-020-16750-3 . OpenUrl CrossRef PubMed 63. ↵ Steinfeld JB , Beláň O , Kwon Y , Terakawa T , Al-Zain A , Smith MJ , et al. Defining the Influence of Rad51 and Dmc1 Lineage-Specific Amino Acids on Genetic Recombination . Genes Dev . 2019 ; 33 ( 17-18 ): 1191 – 207 . Epub 2019/08/03. doi: 10.1101/gad.328062.119 . PubMed PMID: 31371435 ; PubMed Central PMCID: PMCPMC6719624 . OpenUrl Abstract / FREE Full Text 64. ↵ Hunter N , Chambers SR , Louis EJ , Borts RH . The Mismatch Repair System Contributes to Meiotic Sterility in an Interspecific Yeast Hybrid . The EMBO Journal . 1996 ; 15 ( 7 ): 1726 – 33 . doi: 10.1002/j.1460-2075.1996.tb00518.x . OpenUrl CrossRef PubMed Web of Science 65. ↵ Worth L , Clark S , Radman M , Modrich P . Mismatch Repair Proteins Muts and Mutl Inhibit Reca-Catalyzed Strand Transfer between Diverged Dnas . Proceedings of the National Academy of Sciences . 1994 ; 91 ( 8 ): 3238 – 41 . doi : doi: 10.1073/pnas.91.8.3238 . OpenUrl Abstract / FREE Full Text 66. ↵ Hofstatter PG , Lahr DJG . Complex Evolution of the Mismatch Repair System in Eukaryotes Is Illuminated by Novel Archaeal Genomes . J Mol Evol . 2021 ; 89 ( 1-2 ): 12 – 8 . Epub 2021/01/08. doi: 10.1007/s00239-020-09979-5 . PubMed PMID: 33409543 ; PubMed Central PMCID: PMCPMC7884376 . OpenUrl CrossRef PubMed 67. ↵ Chakraborty U , Alani E . Understanding How Mismatch Repair Proteins Participate in the Repair/Anti-Recombination Decision . FEMS Yeast Research . 2016 ; 16 ( 6 ). doi: 10.1093/femsyr/fow071 . OpenUrl CrossRef PubMed 68. ↵ Luk AWS , Williams TJ , Erdmann S , Papke RT , Cavicchioli R. Viruses of Haloarchaea . Life . 2014 ; 4 ( 4 ): 681 - 715 . PubMed PMID : doi: 10.3390/life4040681 . OpenUrl CrossRef PubMed 69. ↵ Pietilä MK , Laurinmäki P , Russell DA , Ko CC , Jacobs-Sera D , Butcher SJ , et al. Insights into Head-Tailed Viruses Infecting Extremely Halophilic Archaea . J Virol . 2013 ; 87 ( 6 ): 3248 – 60 . Epub 2013/01/04. doi: 10.1128/jvi.03397-12 . PubMed PMID: 23283946 ; PubMed Central PMCID: PMCPMC3592136 . OpenUrl Abstract / FREE Full Text 70. ↵ Hanthi YW , Ramirez-Otero MA , Appleby R , De Antoni A , Joudeh L , Sannino V , et al. Rad51 Protects Abasic Sites to Prevent Replication Fork Breakage . Mol Cell . 2024 ; 84 ( 16 ): 3026 – 43 .e11. Epub 2024/08/24. doi: 10.1016/j.molcel.2024.07.004 . PubMed PMID: 39178838 . OpenUrl CrossRef PubMed 71. ↵ Nguyen L-T , Schmidt HA , von Haeseler A , Minh BQ . Iq-Tree: A Fast and Etective Stochastic Algorithm for Estimating Maximum-Likelihood Phylogenies . Molecular Biology and Evolution . 2014 ; 32 ( 1 ): 268 – 74 . doi: 10.1093/molbev/msu300 . OpenUrl CrossRef PubMed 72. ↵ Hoang DT , Chernomor O , von Haeseler A , Minh BQ , Vinh LS . Ufboot2: Improving the Ultrafast Bootstrap Approximation . Molecular Biology and Evolution . 2017 ; 35 ( 2 ): 518 – 22 . doi: 10.1093/molbev/msx281 . OpenUrl CrossRef PubMed 73. ↵ Oren A , Garrity GM . Valid Publication of the Names of Forty-Two Phyla of Prokaryotes . International Journal of Systematic and Evolutionary Microbiology . 2021 ; 71 ( 10 ). doi: 10.1099/ijsem.0.005056 . OpenUrl CrossRef 74. ↵ Katoh K , Rozewicki J , Yamada KD . Matt Online Service: Multiple Sequence Alignment, Interactive Sequence Choice and Visualization . Brief Bioinform . 2019 ; 20 ( 4 ): 1160 – 6 . Epub 2017/10/03. doi: 10.1093/bib/bbx108 . PubMed PMID: 28968734 ; PubMed Central PMCID: PMCPMC6781576 . OpenUrl CrossRef PubMed 75. ↵ Kuraku S , Zmasek CM , Nishimura O , Katoh K . Aleaves Facilitates on-Demand Exploration of Metazoan Gene Family Trees on Matt Sequence Alignment Server with Enhanced Interactivity . Nucleic Acids Res . 2013 ; 41 (Web Server issue): W22 – 8 . Epub 2013/05/17. doi: 10.1093/nar/gkt389 . PubMed PMID: 23677614 ; PubMed Central PMCID: PMCPMC3692103 . OpenUrl CrossRef PubMed 76. ↵ Chen I-MA , Chu K , Palaniappan K , Ratner A , Huang J , Huntemann M , et al. The Img/M Data Management and Analysis System V.7: Content Updates and New Features . Nucleic Acids Research . 2022 ; 51 ( D1 ): D723 – D32 . doi: 10.1093/nar/gkac976 . OpenUrl CrossRef 77. ↵ Camargo AP , Nayfach S , Chen IA , Palaniappan K , Ratner A , Chu K , et al. Img/Vr V4: An Expanded Database of Uncultivated Virus Genomes within a Framework of Extensive Functional, Taxonomic, and Ecological Metadata . Nucleic Acids Res . 2023 ; 51 ( D1 ): D733 – d43 . Epub 2022/11/19. doi: 10.1093/nar/gkac1037 . PubMed PMID: 36399502 ; PubMed Central PMCID: PMCPMC9825611 . OpenUrl CrossRef PubMed View the discussion thread. Back to top Previous Next Posted February 21, 2025. Download PDF Supplementary Material Email Thank you for your interest in spreading the word about bioRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Evolution of Eukaryotic Specific DNA Binding Sites in Asgard Archaeal RecA Recombinases Message Subject (Your Name) has forwarded a page to you from bioRxiv Message Body (Your Name) thought you would like to see this page from the bioRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Evolution of Eukaryotic Specific DNA Binding Sites in Asgard Archaeal RecA Recombinases Bryan Ferlez , Peter Towei Huang , Jingyi Hu , J. Brooks Crickard bioRxiv 2025.02.19.639135; doi: https://doi.org/10.1101/2025.02.19.639135 Share This Article: Copy Citation Tools Evolution of Eukaryotic Specific DNA Binding Sites in Asgard Archaeal RecA Recombinases Bryan Ferlez , Peter Towei Huang , Jingyi Hu , J. Brooks Crickard bioRxiv 2025.02.19.639135; doi: https://doi.org/10.1101/2025.02.19.639135 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Evolutionary Biology Subject Areas All Articles Animal Behavior and Cognition (7624) Biochemistry (17651) Bioengineering (13873) Bioinformatics (41887) Biophysics (21424) Cancer Biology (18566) Cell Biology (25465) Clinical Trials (138) Developmental Biology (13365) Ecology (19871) Epidemiology (2067) Evolutionary Biology (24293) Genetics (15591) Genomics (22478) Immunology (17715) Microbiology (40331) Molecular Biology (17150) Neuroscience (88492) Paleontology (666) Pathology (2828) Pharmacology and Toxicology (4817) Physiology (7635) Plant Biology (15114) Scientific Communication and Education (2044) Synthetic Biology (4286) Systems Biology (9817) Zoology (2268)