The tiny germline chromosomes of Paramecium aurelia have an exceptionally high recombination rate and are capped by a new class of Helitrons

The tiny germline chromosomes of Paramecium aurelia have an exceptionally high recombination rate and are capped by a new class of Helitrons | 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 The tiny germline chromosomes of Paramecium aurelia have an exceptionally high recombination rate and are capped by a new class of Helitrons View ORCID Profile Olivier Arnaiz , View ORCID Profile Frédéric Guérin , View ORCID Profile Arnaud Couloux , Caridad Miró-Pina , View ORCID Profile Guillaume Pellerin , Irina Nekrasova , Joëlle Amselem , View ORCID Profile Jean-Marc Aury , View ORCID Profile Simran Bhullar , View ORCID Profile Andrea Frapporti , View ORCID Profile Emmanuelle Lerat , Isabelle Luyten , Sophie Malinsky , Nathalie Mathy , Alexey Potekhin , View ORCID Profile Vinciane Régnier , View ORCID Profile Natalia Sawka-Gądek , View ORCID Profile Amandine Touzeau , Augustin de Vanssay , View ORCID Profile Coralie Zangarelli , View ORCID Profile Hadi Quesneville , View ORCID Profile Mireille Bétermier , View ORCID Profile Karine Labadie , View ORCID Profile Laurent Duret , View ORCID Profile Eric Meyer , View ORCID Profile Sandra Duharcourt , View ORCID Profile Linda Sperling doi: https://doi.org/10.1101/2025.11.06.686955 Olivier Arnaiz 1 Université Paris-Saclay , CEA, CNRS - Institute for Integrative Biology of the Cell (I2BC) , Gif-sur-Yvette, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Olivier Arnaiz Frédéric Guérin 2 Université Paris-Cité, CNRS, Institut Jacques Monod , 15 Rue Hélène Brion, F-75013, Paris, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Frédéric Guérin Arnaud Couloux 3 Génomique Métabolique, Genoscope, Institut François Jacob, CEA, CNRS, Université Evry, Université Paris- Saclay , Evry, 91057, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Arnaud Couloux Caridad Miró-Pina 2 Université Paris-Cité, CNRS, Institut Jacques Monod , 15 Rue Hélène Brion, F-75013, Paris, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Guillaume Pellerin 4 Institut de Biologie de l’ENS (IBENS), Département de Biologie, Ecole Normale Supérieure, CNRS, INSERM, Université PSL , 75005 Paris, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Guillaume Pellerin Irina Nekrasova 4 Institut de Biologie de l’ENS (IBENS), Département de Biologie, Ecole Normale Supérieure, CNRS, INSERM, Université PSL , 75005 Paris, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Joëlle Amselem 5 Université Paris-Saclay , INRAE, URGI, 78026, Versailles, France 6 Université Paris-Saclay , INRAE, BioinfOmics, URGI, 78026, Versailles, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jean-Marc Aury 3 Génomique Métabolique, Genoscope, Institut François Jacob, CEA, CNRS, Université Evry, Université Paris- Saclay , Evry, 91057, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Jean-Marc Aury Simran Bhullar 4 Institut de Biologie de l’ENS (IBENS), Département de Biologie, Ecole Normale Supérieure, CNRS, INSERM, Université PSL , 75005 Paris, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Simran Bhullar Andrea Frapporti 2 Université Paris-Cité, CNRS, Institut Jacques Monod , 15 Rue Hélène Brion, F-75013, Paris, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Andrea Frapporti Emmanuelle Lerat 7 Universite Claude Bernard Lyon 1 , LBBE, UMR 5558, CNRS, VAS, Villeurbanne, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Emmanuelle Lerat Isabelle Luyten 5 Université Paris-Saclay , INRAE, URGI, 78026, Versailles, France 6 Université Paris-Saclay , INRAE, BioinfOmics, URGI, 78026, Versailles, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Sophie Malinsky 4 Institut de Biologie de l’ENS (IBENS), Département de Biologie, Ecole Normale Supérieure, CNRS, INSERM, Université PSL , 75005 Paris, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Nathalie Mathy 1 Université Paris-Saclay , CEA, CNRS - Institute for Integrative Biology of the Cell (I2BC) , Gif-sur-Yvette, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Alexey Potekhin 4 Institut de Biologie de l’ENS (IBENS), Département de Biologie, Ecole Normale Supérieure, CNRS, INSERM, Université PSL , 75005 Paris, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Vinciane Régnier 1 Université Paris-Saclay , CEA, CNRS - Institute for Integrative Biology of the Cell (I2BC) , Gif-sur-Yvette, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Vinciane Régnier Natalia Sawka-Gądek 4 Institut de Biologie de l’ENS (IBENS), Département de Biologie, Ecole Normale Supérieure, CNRS, INSERM, Université PSL , 75005 Paris, France 8 Institute of Systematics and Evolution of Animals, Polish Academy of Sciences , Sławkowska 17 31-016 Kraków, Poland Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Natalia Sawka-Gądek Amandine Touzeau 2 Université Paris-Cité, CNRS, Institut Jacques Monod , 15 Rue Hélène Brion, F-75013, Paris, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Amandine Touzeau Augustin de Vanssay 2 Université Paris-Cité, CNRS, Institut Jacques Monod , 15 Rue Hélène Brion, F-75013, Paris, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Coralie Zangarelli 1 Université Paris-Saclay , CEA, CNRS - Institute for Integrative Biology of the Cell (I2BC) , Gif-sur-Yvette, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Coralie Zangarelli Hadi Quesneville 5 Université Paris-Saclay , INRAE, URGI, 78026, Versailles, France 6 Université Paris-Saclay , INRAE, BioinfOmics, URGI, 78026, Versailles, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Hadi Quesneville Mireille Bétermier 1 Université Paris-Saclay , CEA, CNRS - Institute for Integrative Biology of the Cell (I2BC) , Gif-sur-Yvette, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Mireille Bétermier Karine Labadie 3 Génomique Métabolique, Genoscope, Institut François Jacob, CEA, CNRS, Université Evry, Université Paris- Saclay , Evry, 91057, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Karine Labadie Laurent Duret 7 Universite Claude Bernard Lyon 1 , LBBE, UMR 5558, CNRS, VAS, Villeurbanne, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Laurent Duret Eric Meyer 4 Institut de Biologie de l’ENS (IBENS), Département de Biologie, Ecole Normale Supérieure, CNRS, INSERM, Université PSL , 75005 Paris, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Eric Meyer For correspondence: emeyer{at}biologie.ens.fr sandra.duharcourt{at}ijm.fr linda.sperling{at}i2bc.paris-saclay.fr Sandra Duharcourt 2 Université Paris-Cité, CNRS, Institut Jacques Monod , 15 Rue Hélène Brion, F-75013, Paris, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Sandra Duharcourt For correspondence: emeyer{at}biologie.ens.fr sandra.duharcourt{at}ijm.fr linda.sperling{at}i2bc.paris-saclay.fr Linda Sperling 1 Université Paris-Saclay , CEA, CNRS - Institute for Integrative Biology of the Cell (I2BC) , Gif-sur-Yvette, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Linda Sperling For correspondence: emeyer{at}biologie.ens.fr sandra.duharcourt{at}ijm.fr linda.sperling{at}i2bc.paris-saclay.fr Abstract Full Text Info/History Metrics Supplementary material Preview PDF Abstract Background Paramecia belong to the ciliate phylum of unicellular eukaryotes characterized by nuclear dimorphism. A diploid germline micronucleus (MIC) transmits genetic information across sexual generations. A polyploid transcriptionally active somatic macronucleus (MAC) develops at each sexual generation from a copy of the MIC through programmed DNA elimination (PDE) of > 30% of germline DNA. PDE requires the domesticated PiggyMac (Pgm) transposase. Assembly of Paramecium germline genomes has presented an enormous challenge owing to the difficulty of MIC isolation. Results We report chromosome-scale short-read MIC assemblies for 7 species from the P. aurelia species complex. We discovered a novel clade of Helitrons, with 9-10 kb transposase ORFs under purifying selection, that have remained active in all P. aurelia lineages. A long-read assembly for P. tetraurelia together with a genetic linkage map provided a nearly telomere-to-telomere assembly. Conclusions. The genome consists of tiny (300 kb – 1.2 Mb) and numerous (∼160) germline chromosomes with the highest recombination rate ever reported for a eukaryote (420 cM/Mb). The ends of the chromosomes consist of Helitrons inserted in telomeric C 4 A 2 repeats, forming a distinct genomic compartment that is eliminated very early during MAC development in a Pgm-independent manner. Background The monophyletic Ciliophora, that emerged over a billion years ago [ 1 ], is comprised of highly diverse, very successful unicellular eukaryotes. Their most striking common feature is nuclear dimorphism, the separation of germline from somatic chromosomes in structurally and functionally distinct nuclei [ 2 , 3 ]. Diploid micronuclei (MIC) ensure transmission of the genome across sexual generations but are not expressed during vegetative growth while a polyploid macronucleus (MAC) is responsible for gene expression. At each sexual generation, a new MAC develops from a copy of the MIC through programmed DNA elimination (PDE) that removes transposable elements (TEs) and their remnants and restores functional genes [ 4 – 6 ]. The first ciliate genomes to be sequenced were the somatic MAC genomes of Paramecium, Tetrahymena and the distantly related spirotrich, Oxytricha [ 7 – 9 ], providing gene catalogues and information about MAC chromosome organization. The P. tetraurelia MAC genome assembly and annotations revealed a series of whole genome duplications (WGD) that occurred after the divergence of Paramecium and Tetrahymena (∼500 MYa). Partially resolved by gene loss over evolutionary time [ 7 , 10 , 11 ], the most recent WGD occurred concomitantly to the emergence of the P. aurelia complex of at least 16 morphologically identical but reproductively isolated sibling species [ 12 – 14 ]. While paramecia reproduce sexually by conjugation of two cells with compatible mating types, the individuals from P. aurelia species can also undergo a self-fertilization process, autogamy, that yields 100% homozygous progeny [ 15 ]. The germline MIC genomes of only a handful of model ciliates have been assembled and annotated. The degree of fragmentation of the assemblies varies depending on repeat content and the feasibility of purifying the MICs [ 16 – 20 ]. The amount of eliminated DNA, estimated by comparison of MIC and MAC assemblies, varies from ∼30% in P. aurelia and Tetrahymena to at least 95% in Oxytricha and P. caudatum [ 4 , 14 ]. So far, the only model ciliate for which we know the number and overall architecture of the germline chromosomes with confidence is Tetrahymena thermophila [ 18 ]. Centromeric regions are found at the center of 5 large, linear germline chromosomes and most TEs are restricted to the center and the ends of the chromosomes. The assembly confirmed that Tetrahymena uses a specific 15 nt Chromosome Breakage Sequence to fragment MIC chromosomes during MAC development [ 18 ]. In striking contrast, pioneer cytogenetic studies in P. aurelia species suggested karyotypes with tens of small MIC chromosomes (Dippell 1954; Jones 1956). Further, epigenetically determined fragmentation in Paramecium creates a variable set of linear chromosomes covering MAC-destined regions [ 6 , 21 , 22 ]. Analysis of the variability led to the suggestion that MAC scaffolds could represent the arms of MIC chromosomes [ 22 ]. This “metacentric chromosome” hypothesis has been difficult to test for lack of knowledge about Paramecium centromeres. TEs can affect host viability by disrupting genes or altering their expression pattern [ 23 ]. A remarkable feature of Paramecium and other ciliates is the way they control TEs. Epigenetic mechanisms, involving meiosis-specific small RNA pathways that guide histone modifications, silence TEs in Paramecium as in most eukaryotes [ 24 ]. However, in Paramecium and other ciliates, the process is taken to the extreme of their physical elimination from the somatic MAC genome. In the germline MIC genome of Paramecium , exons are interrupted by single copy elements as small as 26 bp that are TE (Tc1/mariner DNA transposon) remnants known as Internal Eliminated Sequences (IES) [ 14 , 25 ]. Tens of thousands of IESs are precisely excised during MAC development, and time- course analysis of DNA elimination showed that IESs are eliminated earlier than TE or satellite sequences [ 16 , 26 ]. Elimination of IESs and most TEs and satellite repeats requires the domesticated PiggyMac (Pgm) transposase, which together with its Pgm-like partners, introduces DNA double strand breaks at the extremities of IESs [ 16 , 25 , 27 , 28 ]. This activity is tightly coupled to canonical non-homologous end- joining repair and gap filling of the cleavage sites [ 29 – 33 ]. Given the lack of a strictly conserved motif among eliminated sequences, how the Pgm excision complex is recruited to the boundaries of eliminated sequences is not fully understood. Data obtained so far indicate that all eliminated sequences do not rely on the same set of proteins for their elimination, suggesting that distinct yet partially overlapping pathways tether the Pgm excision machinery [ 34 – 36 ]. Indeed, most TEs and young, long IESs depend for their elimination on 25 nucleotide small RNAs, the scanRNAs (scnRNAs), whereas old, short IESs do not [ 14 ]. Initially produced from the entire MIC genome during meiosis by MIC-specific transcription factors [ 37 , 38 ] and a dedicated RNAi interference pathway [ 35 , 39 – 42 ], the scnRNA population corresponding to the non-eliminated sequences is subsequently degraded [ 41 – 44 ]. The remaining scnRNAs corresponding to TEs, in association with non-coding transcripts, then guide the Polycomb Repressive Complex 2 (PRC2) to catalyze histone modifications on TEs in the developing MAC, providing specificity despite the lack of a conserved sequence motif [ 34 , 36 , 45 – 47 ]. It is then believed that these repressive histone modifications recruit or activate the Pgm excision complex [ 24 ], which, assisted by chromatin remodelers [ 48 – 50 ], triggers DNA cleavage at the boundaries of the eliminated sequences. Purification of vegetative Paramecium MICs by a fluorescence-activated cell sorting strategy [ 16 ] was a tour de force given the MAC ploidy of ∼1600n [ 51 ]: MIC DNA represents ∼ 0.25% of total nuclear DNA, compared to ∼ 5% in Tetrahymena . The first assembly using DNA from sorted MICs was fragmented (N50 ∼37 kb) but adequate for confirmation that all IESs are retained when Pgm endonuclease expression is knocked down [ 16 ]. The assembly allowed genome-wide discovery of TEs, revealing 38 non-LTR retrotransposon families and expanding knowledge of Tc1-mariner DNA transposons to 13 families. A fraction (∼3%) of the MIC assembly, enriched in satellite DNA but not in known TEs, was found to be eliminated independently of the Pgm endonuclease [ 16 ]. The present article describes the production of chromosome-scale assemblies of the MIC genomes of 7 species from the aurelia group. Our motivation was to understand chromosome organization in P. aurelia , determine the size and number of MIC chromosomes, identify and annotate repeated sequences and provide reference genomes for functional studies. Efforts to obtain a telomere-to- telomere assembly for the P. tetraurelia model, using long-read technology coupled with information from a genetic linkage map, revealed the highest recombination rate ever reported in a eukaryote and a diploid germline genome consisting of ∼160 pairs of small chromosomes. We discovered a new clade of Helitrons that have remained active in all lineages of the aurelia clade. Unlike previously annotated Paramecium TEs, most Helitrons are inserted in germline telomeres or sub-telomeric regions. Their elimination occurs earlier than that of other TEs and does not require the Pgm excision complex. RESULTS Assembly and annotation of Paramecium MIC genomes MIC genome assemblies for 7 P. aurelia species We selected species spread across the Paramecium aurelia clade ( Figure 1A ) for vegetative MIC purification by fluorescence-activated cell sorting and high-throughput sequencing using the Illumina platform, as previously described [ 14 , 16 ]. The strategy to build chromosome-scale MIC assemblies using these datasets is summarized in SupFigure1A. MIC DNA read-pairs were assembled into contigs. Then PGM -RNAi DNA isolated from cells depleted for the Pgm endonuclease by RNAi, available in greater quantities necessary to build mate-pair libraries, was used for scaffolding, as detailed in Methods. Download figure Open in new tab Figure 1. Paramecium aurelia MIC genome content. (A) Phylogenetic tree of P. aurelia species with P. caudatum as outgroup, modified from [ 14 ], showing the size of MIC and MAC assemblies. (B) Barplots representing genome occupancy of Coding and non-coding genes, Transposable Elements, and Satellites in the MAC-destined (MDS) and MIC-limited (MLS) regions of the P. aurelia MIC genome assemblies. Genome occupancy in Mb is given in SupFigure4. (C) Heatmap showing conservation of annotated minisatellites mapped to MIC genomes of each species using RepeatMasker (Methods). Each row corresponds to a minisatellite consensus and each column to a target species. The intensity of the blue rectangle is proportional to the similarity of the best hit to the target genome. Satellites are ordered by hierarchical clustering, showing the most conserved at the top. (D) TEs annotated using REPET were classified as Helitrons, TIRs, LINEs, SINEs and uncharacterized elements. Barplots show their MIC genome occupancy as percentage of total TE annotation. (E) Circular tree of Helitron PIF1 PFAM domains calculated with nucleotide alignments for 4 species. (F) PFAM DDE_3 domains were used to make a tree representing Tc1/mariner elements. Clades are labeled according to the classification described in [ 16 ]. Note recent acquisition of P. sexaurelia and P. biaurelia copies branching near the root of the DDE3, Anchois or DDE1/DDE2 clades, denoted by a red dot. (G) Circular tree of LINE RVT_1 PFAM domains, organized in 5 Groups according to the classification in [ 16 ]. Note recent acquisition of P. sexaurelia copies that branch before the separation of Group4 and Group5. (H) Phylogeny of concatenated Rep and Hel domains from across the tree of life [ 52 ] re-aligned with concatenated Paramecium Rep and Hel domains using MAFFT. The unrooted tree was constructed with phyML and drawn with ggtree. Bootstrap support is represented by the size of the circles drawn at nodes. The tree shows that a Paramecium HLE clade (blue tips) emerged very early and is distinct from the Helitron (HLE1) (red tips) and the Helentron/Helitron2 (HLE2) (yellow tips) clades. The six scaffolded assemblies have an average MIC genome size around 100 Mb ( Figure 1A ), of which around 30% is eliminated during MAC development. With an average N50 of ∼500 kb, these assemblies are of good quality (SupTable1). Indeed, >99% of genes previously annotated using MAC genome assemblies [ 11 , 14 , 53 ] and >97% of previously annotated IESs [ 14 , 25 ] could be remapped to these assemblies. Analysis of the assemblies using the K-mer Analysis Toolkit [ 54 ] provided an assembly-agnostic assessment that the genomes are complete, with the possible exception of the P. sexaurelia assembly (SupFigure2). The P. sonneborni genome is a case apart. Although scaffolds could not be built from the contigs for technical reasons (cf. Methods), the assembly (N50 ∼33 kb) seems to be complete (SupTable1; SupFigure2). The assembly size of 217 Mb is at least twice that of the other P. aurelia MIC genomes ( Figure 1A ), consistent with previously reported genome size estimates [ 14 ]. The large genome size is essentially owing to an increase of the MIC-specific compartment (the MAC genome size is ∼ 82 Mb compared to ∼70 Mb in the other aurelia species). Examination of the origins of the MIC sequences revealed that recurrent horizontal transfer of sequences from several other aurelia species can account for the increased genome size (see [ 55 ] ). MIC genome content Paramecium MIC genomes consist of MAC-destined sequences (MDS) colinear with the corresponding MAC genome and MIC-limited sequences (MLS) that are eliminated during MAC development. To discover sequence features, especially within the MLS compartment, we used dedicated software to annotate genes, tandem repeats (satellite DNA) and dispersed repeats (TEs) for each of the 7 MIC assemblies. Paramecium aurelia MAC genomes are at least 70% coding harboring ∼40,000 genes of median size ∼1.2 kb with tiny introns, separated by very short intergenic regions [ 11 , 53 , 56 , 57 ] as shown in Figure 1B . Since examples of genes that contain MIC-specific sequences have been documented [ 14 , 40 ], we looked for genes completely embedded in P. aurelia MIC-limited regions (Methods). The high MDS gene density sharply contrasts with the low MLS gene density that was found. Indeed, as shown in Figure 1B , annotation of the assemblies detected a low density of 5 – 10% potential protein-coding genes (ATG … TGA, with or without introns), and an even lower density of potential non-coding genes (GC-content expected for coding regions but with no detectable ORF) in MLS. Moreover, the size of the ORFs in the MLS is significantly smaller than the size of ORFs in annotated MAC genes (SupFigure3A; median sizes ∼470 bp compared to ∼1.1 kb), as expected for genes that have decayed (pseudogenes). The same trend is found in a comparison of non-coding genes, though the difference is smaller (SupFigure3A). Using RNA-Seq data sets available for P. tetraurelia [ 53 ], expression of annotated MLS genes was examined (SupFigure3B). Except for a handful of genes (n = 29 with RPKM > 1, top of heatmap), none of the 3257 putative protein-coding genes or 1177 putative non-coding genes appear to be expressed, neither in vegetative cells nor during MAC development. This suggests that they are mainly annotation errors or pseudogenes. However in cells depleted for Ezl1, the catalytic subunit of Polycomb Repressive Complex 2 (PRC2) [ 45 , 46 ], the fraction of MLS genes that have transcripts (RPKM > 1 at T50) is larger (30%), showing they have the potential to be expressed. To annotate satellite repeats, we combined two approaches, MREPS and TAREAN (Methods). Minisatellites, mainly localized in MLS, represent only ∼ 2% of the size of each Paramecium MIC assembly and ∼ 6% of the MLS ( Figure 1B ; SupFigure4A). Minisatellite repeat units are generally species-specific, apart from minisatellites common to P. primaurelia and P. pentaurelia, the most closely related species ( Figure 1C , SupFigure5A). The large number of identified satellite repeats in P. sonneborni is very likely related to its large genome size. In order to see if the most abundant repeats are conserved across species, we sorted the units by abundance (SupFigure5B) and found that the most abundant units are not necessarily the most conserved. Since the diversification of the aurelia clade is relatively ancient (estimated 100 – 300 Mya) ([ 10 , 11 ], this is not surprising: rapid evolution of satellite repeats has been observed in other eukaryotes [ 58 – 61 ]. A notable exception is the 126 nt repeat (P126) similar to a WD40 repeat, the first Paramecium minisatellite identified because of its association with MAC telomeres [ 62 ]. Annotation of TEs (Methods) found elements from known superfamilies (LINE non-LTR retrotransposons and Tc1/mariner copy-and-paste DNA transposons) [ 16 ], and revealed Helitrons, DNA transposons that use a peel-and-paste mechanism of single-strand transposition [ 63 , 64 ]. SINEs, MITEs and uncharacterized repeats were also annotated. Despite the critical role of a domesticated piggyBac transposase in PDE, we found no traces of piggyBac transposons in any of the P. aurelia MIC genomes, in contrast with T. thermophila [ 18 ]. Surprisingly, although widespread in eukaryotic genomes, there was no hint of any LTR retrotransposons in these assemblies. LTR retrotransposons have not been found either in Tetrahymena thermophila [ 18 ] or the early-branching karyorelict ciliate Loxedes magnus [ 65 ], suggesting that these elements did not colonize ciliates. The annotated TEs only account for ∼5 Mb of the total genome size (almost exclusively in MLS), less than 20% of the germline DNA eliminated during MAC development ( Figure 1B ; SupFigure4A). Overall, TIR (Tc1/mariner) and LINE elements were the most abundant (∼ 2 - 3 Mb each) ( Figure 1D ; SupFigure4B). Helitron occupancy varied across the aurelia clade, on average ∼1 Mb ( Figure 1D ; SupFigure4B). In the P. sexaurelia / P. sonneborni sub-clade, LINEs appear to be less abundant than Tc1/mariner TEs. This difference could result from invasion or proliferation of elements after the divergence of the two P. aurelia sub-clades. We also used a guided approach to annotate TE protein domains for the purpose of reconstituting the evolutionary dynamics of Paramecium Helitron, Tc1/mariner and LINE superfamily elements. We used Pfam domains specific for each superfamily to identify their occurrence in six-frame translations of the genome assemblies (Methods). The protein sequences of the least degenerate copies for Pif1 DNA helicase (Helitron), DDE_3 DDE endonuclease (TIR) and reverse transcriptase RVT_1 (LINE) domains were used to build phylogenies ( Figure 1 E-G ). The phylogenetic trees support ancient invasion of Helitron, Tc1/mariner, and LINE elements, pre-dating diversification of the aurelia clade. Human curation of the Helitron transposase ORFs was carried out for the 7 P. aurelia species. Remarkably, compared to curated LINE and TIR transposase ORF copies [ 16 ], a greater proportion of the Helitron ORF copies appear to be full size (SupFigure6). Conceptual translations of 112 full size Helitron ORF copies were aligned and used to build a phyML tree that identified 29 elements (SupFigure7A and B) and has the same clade topology (A1, A2 and B) as the Pif1 domain tree ( Figure 1E ). The analysis of synonymous (dS) and non-synonymous (dN) substitution rates showed clear evidence of purifying selection, with an average dN/dS ratio over all branches of the phylogeny of 0.15±0.03 (see Methods). A phylogeny of Helitron-like Elements (HLE) from across the tree of life was recently proposed using a new automated algorithm for their identification and classification [ 52 ]. Since Helitron transposases contain an HUH endonuclease domain (Rep) [ 63 , 66 ] followed by a Pif1-like helicase (Hel) domain [ 67 ], these authors made a phylogeny using Rep-Hel domains from across the eukaryotic tree and found only two clades, HLE1 and HLE2. We added the homologous domains from curated Paramecium Helitron transposase ORFs, aligned all the sequences and built an unrooted phyML tree ( Figure 1H ). The Paramecium HLEs are found in an independent clade (“HLE Paramecium”) that diverged from HLE1 and HLE2, with high bootstrap support (Methods). Intriguingly, Helitron ORFs are interrupted by up to 4 IES insertions in 9 of the 29 identified elements (8 from clade A2 and 1 from clade B), a frequency close to that observed for cellular genes. These IESs are generally conserved in all annotated copies from all species. One remarkable case is an IES inserted at the same site, in a conserved domain of the transposase, in all copies of 5 related elements forming a well-defined A2 subclade represented in all species (SupFigure7B). This IES insertion therefore likely became fixed in a common ancestor of the 5 elements before the beginning of speciation in the P. aurelia complex. In contrast to the flanking coding sequences, the IES appears to have evolved with almost no constraint, as previously noted for IESs inserted within cellular genes (SupFigure7C). Excision timing is similar to that of other IESs (SupFigure7D). Even with improved annotation tools compared to the approach used previously for P. tetraurelia [ 16 ], a large portion of Paramecium MIC-limited genomes (50% - 70%) does not correspond to any annotated feature (gene, TE or satellite) ( Figure 1B ). It remains very difficult to know whether regions with no annotation correspond to “dark matter” i.e. sequence features no longer under selection that have evolved beyond recognition [ 68 ], or sequence features we do not yet know how to annotate. Counting chromosomes It has long been suspected that Paramecium aurelia species harbor a large number of small germline chromosomes, based on very early cytogenetics [ 69 , 70 ]. We report in this section the use of long- read sequencing to achieve a distinct, more complete assembly for P. tetraurelia as well as the construction of a genetic linkage map, leading to prediction of the karyotype of P. tetraurelia . Pgm depletion does not block elimination of ∼3 Mb of MIC-specific genome sequence, enriched in satellite DNA [ 16 ]. The short-read assemblies scaffolded with PGM RNAi DNA libraries are thus likely to be sub-optimal, especially for highly repeated regions of the genome. To try to improve closure and obtain telomere-to-telomere chromosomes, we used the Oxford Nanopore Technology (ONT) to generate long reads. Since more DNA than was available from sorted vegetative MICs was required to build the sequencing libraries, we used DNA from new MACs of P. tetraurelia cells depleted in Ezl1, the catalytic subunit of Polycomb Repressive Complex 2 (PRC2) required for PDE [ 34 , 45 , 46 ], to prepare enough unrearranged DNA for sequencing. In EZL1 -RNAi DNA, 3 Mb of additional MLS are retained compared to PGM RNAi DNA [ 16 ], however ∼33% of IESs (0.73 Mb; SupTable1) are excised [ 34 ]. With long reads of this EZL1 -RNAi DNA (1M reads with a median size ∼14 kb), a high quality de novo assembly was achieved (N50 ∼615 kb, 202 contigs, assembly size ∼104 Mb; Methods, SupFigure1B, SupFigure2, SupTable1). This assembly is less fragmented than the short-read assembly, contains all genes and the expected number of IESs (67%) and provides an improved reference to study the organization of the P. tetraurelia MIC genome. Construction of a genetic linkage map To help complete the assembly we constructed a genetic linkage map. We sequenced ( Illumina platform) and analyzed total DNA (essentially MAC DNA owing to high MAC ploidy) from 39 F2 clones of a genetic cross between two P. tetraurelia strains, 51 and 32 (Methods; SupFigure8A). We retained 220,994 biallelic SNP markers for genotyping (SNP density of 0.3% in MAC-destined regions of the genome). We detected a total of 12,206 cross over events (CO) in 39 meioses (313 COs per F2) ( Figure 2A&B ; SupFigure8A&D). This corresponds to a recombination rate of 420 cM/Mb in the MAC- destined compartment, a value that is extremely high compared to that found for most of the other species studied to date (SupFigure8E), and a genetic length of 31,000 cM, by far the highest ever reported ( Figure 2C ). Download figure Open in new tab Figure 2. P. tetraurelia chromosomes and recombination. (A) Relationship between the average number of crossovers per F2 and the length of MAC-destined regions. (cf. Methods and SupFigure8). The 161 points correspond to 158 supercontigs of the MIC genome assembly, and to 3 groups of 2 supercontigs that are linked based on the MAC assembly. The horizontal dotted line at 0.5 is equivalent to 50 cM. (B) Relationship between the recombination rate (cM/Mb) and the length of MAC-destined regions (N=161). (C) Distribution of genetic map lengths (cM) across eukaryotes. (D) Correspondence of MIC supercontig length (n=187) with the length of MAC scaffolds (MAC-constitutive scaffolds, MAC v2.0; n = 162). Each dot gives the (sum of the) length(s) of the MIC supercontig(s) as a function of the (sum of the) length(s) of the corresponding MAC scaffold(s). The dotted lines show that MIC supercontigs are up to twice as large as the corresponding MAC scaffolds. In three cases, two supercontigs encompassed contiguous segments of the same MAC scaffold (category 1:n). In 5 cases, one MIC supercontig corresponded to 2 MAC scaffolds (category n:1). See SupData1_MIC_MAC for details. (E) Illumina MIC reads with telomere repeats (mix of C4A2 and C3A3 hexamers) were clustered and aligned (Methods; SupData2_TelomereJunctions). Three categories of junctions were found: MIC ‘edge’ telomeric repeat junctions, MIC ‘internal’ telomeric repeat junctions. The last category, MAC telomeric repeat junctions are heterogeneous among the sequenced reads, the expectation for de novo telomeric repeats added to DNA extremities generated during PDE [ 21 ]. (F) Distance between telomere repeat junctions and the nearest supercontig end, for each category. The last bin of the barplots represents distances of 50 kb or greater. MIC ‘edge’ junctions are mostly at supercontig ends (n = 88): 66 (75%) map at a distance of 0 from the supercontig end as expected for true MIC telomeres in the unrearranged MIC genome. The others (n= 22, 25%) may represent MIC internal telomeric junctions. The telomere repeat clusters classified as MIC internal telomeric junctions (n=53) are mostly (n=43; 81%) less than 20 kb from supercontig ends. MAC telomeric repeat junctions are further from supercontig ends (9 / 49, 18% at <20kb). Note that telomeric repeats can be found in either of two orientations: 5’-CA or 5’-GT. Only MAC telomeres can be detected in both orientations (black and white bars). (G) Lengths of supercontigs containing MLS that are capped by telomere repeats at both ends, one end or neither end. In most eukaryotes, CO interference [ 71 ] constrains spacing of otherwise randomly located CO sites along chromosomes. We measured a 50% CO deficit at a distance of 22 kb comparable to that reported in S. cerevisiae (30 kb) (SupFigure8C) [ 72 ]. We also found evidence of aneuploidy (Methods) since 8 F2s (21%) carried some scaffolds with a doubled ploidy (5 with only one scaffold, the others with respectively 3, 4 and 13 aneuploid scaffolds) (SupFigure9). Similarly, using 74 independent published DNA-Seq datasets prepared from 100% homozygous cell lines, we found 21 additional cases of doubled ploidy for 1 or more scaffolds (28%) (SupFigure9). These cases of aneuploidy most probably result from chromosome segregation errors during meiosis and they occur with a frequency comparable to that reported in human (23.6%; [ 73 ]) (SupFigure9; cf. Discussion). Analysis of linkage revealed that 37 P. tetraurelia contigs could be clustered into 17 linkage groups, and that a few contigs consisted of two genetically unlinked segments, suggestive of chimeric assembly. We therefore built a composite genome, called the “MIC EZL1 assembly”, using the linkage map to split or join the contigs into 187 supercontigs (N50 ∼620 kb, 104.3 Mb). This assembly includes 164 supercontigs (101.4 Mb in total) that encompass 166 MAC scaffolds, and 23 supercontigs (2.9 Mb in total) consisting only of MLS, for which it was not possible to analyze genetic linkage. There is an essentially 1:1 relationship between the MIC EZL1 supercontigs and the MAC scaffolds ( Figure 2D ; SupData1_Synteny_MIC_MAC.html; SupFigure10). Often, the MIC EZL1 supercontig is at least twice as large as the MAC scaffold, corresponding to significant DNA elimination. More detailed analysis of the synteny between MIC EZL1 and MAC regions (Methods) showed that the sequences present in MIC EZL1 that are absent from the MAC scaffolds are found primarily at supercontig ends but sometimes are more centrally located (cf. annotated Circos drawings: SupData1_ Synteny_MIC_MAC.html). We conclude that the fragmentation sites used for MAC development are situated at the extremities of MIC chromosomes, although additional sites are sometimes more internal. MIC telomeres Have we achieved a telomere-to-telomere MIC genome assembly? Are MIC telomere repeats the same as MAC telomere repeats? To answer these questions, we first sought to identify MIC telomere repeats using the Illumina read data. In eukaryotes, telomeres almost always consist of tandem repeats of a 6, 7 or 8 nt motif, usually GT-rich in the strand oriented 5’ to 3’ towards the end of the chromosome. The well-characterized Paramecium MAC telomere repeats that assemble de novo to heal programmed and accidental double-strand breaks consist of random mixtures of TTGGGG and TTTGGG hexamers [ 74 ], the result of misincorporation at a single templating position in the telomerase RNA [ 75 , 76 ]. Assuming that P. aurelia MIC telomeres are also tandem repeats of a short motif, we looked for candidates using a sans a priori analysis of the k -mers in all read pairs that did not map properly to the MAC assembly (Methods). We found the most abundant G-rich motif was TTGGGG (equivalent to 5’CCCCAA3’ repeats at the beginning of the scaffold), suggesting that MIC and MAC telomere repeats are the same. The repeats are located at or near scaffold or supercontig ends. For P. tetraurelia, the repeats were not only restricted to contig ends but were much more abundant in the MIC EZL1 long-read assembly than in the short-read assembly (SupFigure11A), suggesting that MIC telomere repeats were probably collapsed or on small contigs not incorporated into the short-read assembly. We used an orthogonal approach, independent of the assemblies, to gain insight on telomere branch points. We selected Illumina MIC sequencing reads containing a junction between telomeric repeats and non-telomeric sequences, clustered them and used multiple alignments to obtain consensus sequences (Methods, SupData2). Although this approach is limited by sequencing depth and cannot provide an exhaustive catalogue, it is possible to identify the junctions between MIC telomeric repeats and flanking sequences and to distinguish them from MAC junctions. Indeed, in the case of MAC telomeres, telomere addition does not always occur at the same position among all reads because chromosome fragmentation is an imprecise process that occurs in the new MAC after several rounds of endoreplication [ 21 ]. In the case of MIC telomeres present in the unrearranged germline genome, telomere branch points are expected at exactly the same position in all reads. Surprisingly, we found two categories among the sampled MIC junctions. In the first category, telomeric repeats are flanked on only one side by non-telomeric sequences (“edge” junctions), as must be the case for the telomeres of MIC chromosomes. In the second category, the blocks of telomeric repeats are flanked on both sides with non-telomeric sequences (“internal” junctions), corresponding to telomeric repeat blocks within MIC chromosomes. It is important to note that a MIC “edge” junction does not guarantee that the telomeric repeats are at the end of a chromosome, because the clusters are short local alignments (180 – 400 bp, mean 252 bp). Examples of MAC, MIC “edge”, and MIC “internal” telomeric junctions are shown in Figure 2E . We were able to map 87% of these telomere junction consensus sequences to the P. tetraurelia MIC EZL1 genome assembly (Methods; SupData2_TelomereJunctions). After removing redundant junctions that correspond to the same loci, we retained 88 MIC “edge”, 53 MIC “internal” and 49 MAC junctions for further analysis. To see whether these telomeric repeat junctions are at chromosome ends, we plotted their distance from the closest supercontig end ( Figure 2F ). As expected, MAC telomere junctions do not map at supercontig ends, a majority of 57% are at a distance of at least 50 kb (28/49) and are found in both orientations (GT or CA repeats near the 5’ end of supercontigs). The MIC junction mapping provides a totally different picture. We found that 75% (66/88) of the MIC “edge” junctions map at the very end of supercontigs (distance = 0) in the expected orientation (CA strand of telomeric repeats on the 5’-end strand) and may thus represent the telomeres of MIC chromosomes. The “internal” junctions are also close to the supercontig ends and, intriguingly, are in the same orientation. There are 8 cases (15%) where internal junctions do map to the very end of a supercontig, but this corresponds to situations where a few long-reads suggest that the supercontig could be extended to the end of the chromosome. These results indicate that MAC telomere repeat junctions are farther from supercontig ends than MIC telomere repeat junctions. Since none of the sampled MAC junctions was closer than 1.5 kb from a supercontig end, correctly oriented telomeric repeats that cap supercontigs are most probably MIC telomeres. Considering only the 161 supercontig groups that are syntenic with MAC scaffolds ( Figure 2D and SupData1), we analyzed the size distributions of supercontigs capped by telomere repeats at both ends (n = 33), one end (n = 69) or neither end (n= 59) ( Figure 2G ). The 33 supercontigs capped at both ends are likely complete MIC chromosomes. We note that the size distribution of supercontigs that are missing telomere repeats does not differ significantly from the size distribution of supercontigs that are capped at both ends (p > 0.09, t-tests, Figure 2G ). This suggests that they all could be quasi-complete and hence that P. tetraurelia has as many as 161 MIC chromosomes. Chromosome organization Chromosome ends do not require Pgm for elimination The Paramecium germline genome is organized in distinct regions according to developmental fate. Constitutively MAC-destined regions are always retained in the new MAC and variable MAC-destined regions are partially retained via alternative elimination [ 14 ]. MIC-limited regions are eliminated during MAC development. They include IES and OES (Other Eliminated Sequences) [ 77 ]. The question addressed in this section, is where these regions are located on the chromosomes. Genomic compartments were operationally defined using short read coverage of DNA extracted from sorted nuclei: MICs, MACs, and developing MACs after PGM- RNAi (SupFigure12; Methods). This is illustrated for one P. tetraurelia supercontig in Figure 3A and drawings of all supercontigs are provided in SupData3 & 4. Constitutively MAC-destined regions represent 70.2 Mb of the assembly, variably MAC-destined regions 9.2 Mb, and the MIC-limited compartment 19.9 Mb (IES 3.9 Mb, OES 16 Mb). The OES is further divided into a Pgm-dependent compartment (13.3 Mb) maintained in new MACs after Pgm depletion and a Pgm-independent compartment (2.7 Mb) first noticed by Guérin et al. 2017 [ 16 ], that is eliminated even upon Pgm depletion. Although considered to be a distinct compartment, all IESs are Pgm-dependent [ 16 , 25 ]. Download figure Open in new tab Figure 3. Definition and characteristics of genomic compartments. (A) Genomic compartments defined by sequencing read coverage. JBrowse screen shot shows DNA-Seq coverage of a region of the MIC EZL1 assembly by MIC, PGM-RNAi and MAC Illumina sequence reads (top three tracks). MAC-constitutive and MAC-variable compartments are determined by a high and low MAC read coverage, respectively. IESs were annotated using ParTIES [ 78 ]. The OES (Other Eliminated Sequence) regions were defined by no MAC coverage (IESs excluded) then divided into two compartments: PGM-dependent and PGM-independent, based on the PGM RNAi DNA coverage. See Methods and SupFigure12 for details. (B) Density of each genome compartment across assembly supercontigs showing their relative localization on folded chromosomes: 50% corresponds to the center, 100% the ends. (C) Normalized DNA coverage (RPM) for non-overlapping 500 nt windows. The windows were separated into 3 groups, as a function of their compartment (MAC-constitutive, Pgm-dependent, or Pgm- independent). The boxplots summarize the presence of DNA of each compartment, for different DNA-Seq samples: sorted vegetative micronuclei (MIC), sorted parental old MAC fragments (FRAG), sorted developing MAC time course (DEV1 - DEV4) at different ploidy levels (32C to 128C) [ 26 ], and sorted or enriched (designated by a “*”) new MACs after gene depletion by RNAi; the samples are grouped by pathway (excision complex, scnRNA biogenesis and degradation, non-coding transcription, PRC2 components, and chromatin remodelers). These DNA-seq datasets are described in SupTable2. Figure 3B shows the relative position of each compartment calculated across all the supercontigs of the assembly. As in the example ( Figure 3A ), the relative position of the MAC-constitutive compartment and of the IESs is greatest at the center of the supercontigs and vanishes towards the ends, while the PGM-independent compartment is restricted to the ends. MAC-variable and PGM- dependent compartment occupancy curves peak towards the chromosome ends but then decrease sharply. Time-course analysis of DNA elimination indicates that the Pgm-independent compartment is eliminated very early during MAC development, before IESs are excised and before the ploidy in the new MAC reaches 32C ( Figure 3C ) [ 26 ]. To gain insight into the mechanisms of elimination of this compartment, we examined the consequences of depletion of 32 known proteins involved in PDE, using DNA-Seq of new MACs (see Methods) ( Figure 3C , SupFigure12). As expected, the Pgm excision complex (Pgm, PgmL and Ku70/80c proteins) is dispensable for the elimination of this compartment. This is also the case for the chromatin remodelers ISWI1 [ 49 ] and Spt16 [ 48 ], which are required for IES excision. Given that the assembly was constructed using long reads from DNA of cells depleted for Ezl1, the catalytic subunit of PRC2, the Pgm-independent compartment is not eliminated in this condition ( Figure 3C ). Unsurprisingly, depletion of other PRC2 components or cofactors (Eed, Suz12, Caf1, Rf2) also results in failure to eliminate this compartment. We note that two PRC2 cofactors, Eap1 and Rf4 [ 46 , 47 ], appear to have a weaker effect compared to the other PRC2 components (SupFigure12). Depletion of the proteins required for scnRNA biogenesis (Spt5 [ 37 ]; Dcl2/3 [ 34 , 35 ]), degradation of MIC-limited scnRNAs (Gtsf1 [ 43 , 44 ]), and non-coding transcription (TfIIs4, [ 36 ]) showed they are important for its elimination as well ( Figure 3C , SupFigure12). Consistent with previous reports demonstrating that scnRNAs are produced from the entire MIC genome during meiosis [ 40 , 42 ], the PGM-independent compartment is covered with scnRNAs (SupFigure13). We conclude that the elimination of the Pgm-independent compartment occurs very early during MAC development and involves the scnRNA and PRC2 pathways. The next question is the localization of annotated sequence features with respect to the genomic compartments. A majority of TEs (75.3%) are in the Pgm-dependent compartment ( Figure 4A ). More specifically, this is the case for TIR (Tc1/mariner) and LINE copies, while a majority of Helitrons are found in the Pgm-independent compartment. Consistent with the position of the different genomic compartments ( Figure 3A ), the Helitrons are overwhelmingly localized towards supercontig ends while LINEs and TIRs have a widespread distribution ( Figure 4B ). Minisatellites are also widely distributed across all compartments, as illustrated by P126. However, some abundant minisatellites (CHBD0, SAT149, SAT118, SAT189 and SAT51) are exclusively present (>90%) in the Pgm- independent compartment. As shown in Figure 4C , SAT149 and SAT51 are located at the very end of supercontigs, while CHBD0, SAT118 and SAT189 occur near the extremity. Download figure Open in new tab Figure 4. Genomic compartments and annotation landscape. (A) Barplots show the proportion of each annotated sequence feature (genes, TE and satellites) occupied by the 5 P. tetraurelia genomic compartments of the MIC EZL1 assembly. The compartment occupancy proportion for each TE order (Helitron, TIR, LINE) and of the most abundant satellite families is also presented. The last bar shows the compartment occupancy of regions (n=977) that have the potential to form G-quadruplex (pG4) and do not overlap telomeric repeats known to potentially form G4 (see Methods; SupFigure12). (B) Relative supercontig position of Helitron, TIR and LINE occupancy on MIC supercontigs (as in Figure 3B , the abscissa presents normalized distance to the nearest supercontig end); Helitrons are restricted to the ends of the supercontigs, while TIRs and LINEs are found everywhere. (C) Localization of the most abundant minisatellites on MIC supercontings. Helitron insertions are restricted to chromosome ends Helitrons are mainly localized in the Pgm-independent compartment ( Figure 4A &B). We looked more precisely at excision timing, excision requirements and expression for each of the different Helitron copies with 9-10 kb ORFS (n=79). The heatmap ( Figure 5A ), showing sequencing coverage in different contexts, confirms that these elements are present in the MIC, absent from the MAC, retained upon PRC2 depletion, but for the most part are still eliminated upon Pgm excision complex depletion. As shown for the entire Pgm-independent compartment (Figure3C), the elimination of the majority of Helitron ORFs is sensitive to depletion of factors involved in non-coding transcription and the scnRNA pathway ( Figure 5A , SupFigure14). Helitron ORFs that belong to Clade A1 (Hel01, Hel03, Hel04, and Hel18) are eliminated early in new MAC development (before 32C endoreplication stage [ 26 ]). The other Helitrons are eliminated later, but earlier than LINEs and TIRs which are eliminated very late in MAC development (from 128C). Interestingly, some Clade A2 Helitrons (Hel09, Hel19), those that contain IESs and are retained after Pgm depletion, are also eliminated very late. Finally, some Helitron ORFs can be transcribed ( Figure 5A last column of heatmap) as observed when Ezl1 or other PRC2 components are depleted (SupFigure14). Download figure Open in new tab Figure 5. Paramecium Helitron classification and characterization. (A) Phylogenetic tree of Helitron transposase copies found in the P. tetraurelia MIC EZL1 reference assembly. The phyML amino acid tree was made using MACSE to align the last 5000 nt of 79 full-length Helitron transposase ORFs, containing the conserved HUH Y2 (Rep) and PIF1 (Hel) domains (see SupFigure7A). We used this tree to define 5 Helitron Groups. Clade A2 is split into 2 groups based on the presence of a conserved IES in Hel08, Hel09, and Hel19 copies (Group 2, red labels; cf SupFigure7) but not in the other Helitrons in the clade (Group 5, gray labels). The heatmap was made using DNA-Seq and RNA-Seq RPKM coverage after normalizing copies with respect to collapse (Methods). The 3 longest TIR and LINE copies were also added to the heatmap for comparison. Samples used to calculate coverage are described in SupTable2. DNA samples from enriched but not sorted nuclei are designated by “*”. (B) Structure of some Helitron elements showing the position and orientation of the transposase ORF with respect to the supercontig end (on the left). Helitron Left Terminal Sequence (LTS) (green boxes) and Right Terminal Sequence (blue boxes) were identified by manual alignment and curation. The satellites found between the LTS and the RTS are indicated by colored boxes. For the most abundant Helitrons (Hel01/15/16) the combination of LTS-satellites-RTS defines different “Types”. The presence of an IES is indicated for Hel09/19/08/17. The schema is not to scale. See SupFigure16 for more details. (C). Schematic representation of the Hel01Copy009 copy. ORF, LTS, RTS, and minisatellites are represented by colored boxes (green: LTSc, black:ORF, red: P126, orange: SAT54/81, blue: RTSb). In the zoom to the DNA sequence level before LTS and after RTS, flanking telomeric repeats (pink boxes) are found at both extremities. Structure of complete Paramecium Helitrons Helitron transposition involves nicking of the top strand at the 5’ end of a Left Terminal Sequence (LTS) which forms a covalent bond with the transposase, unwinding of the double-helix, and introduction of a second nick at the end of the Right Terminal Sequence (RTS), shortly after a stem- loop structure, to form a circular intermediate [ 64 , 79 ]. To determine the full extent of transposable units, we looked for conserved terminal sequences marking the boundary with copy-specific flanks. We used DNA dotplots to compare ∼50 kb genomic regions centered on the ORFs, for all copies of the abundant elements Hel01 ( Figure 5A clade A1) and Hel15/Hel16 ( Figure 5A clade B). Example dotplots between pairs of different copies of Hel01 are shown in SupFigure15. The ORF sequences present a continuous diagonal, surrounded by a variety of satellites that often differ in length among closely related copies (rectangles along diagonal) and are altogether missing in more distant copies (gaps in the diagonal). Alignment of the last 350 nt before the sequences diverge revealed 3 related consensus sequences, RTSa, RTSb, and RTSc, with a short stem-loop and ending with CCAA (SupFigure17A), an observation reminiscent of the CTRR found at the end of RTS in other eukaryotes [ 52 , 80 ]. Similarly, RTSd and RTSe were found, by homology to RTSa-c, to end Hel03 and Hel04 ( Figure 5A &B). Alignment of RTSa-d consensus sequences and the unique RTSe showed characteristic RTS stem- loops with compensatory base changes, in support of this structure (SupFigure17A). The left end of the 10 Hel01 copies that did not extend beyond the end of the assembled supercontig (all Hel01 copies are oriented inwards from the end of supercontigs, see below) revealed a well-conserved LTSa sequence, which could also be identified at some supercontig ends in overhanging long-reads mapped with high confidence. The complete Helitron transposable units thus defined are schematically represented in Figure 5C . A similar analysis for the 15 copies of Hel15/Hel16 ( Figure 5A &B) revealed that most copies ended with an unusual Mosaic satellite (presented in detail in the next section; cf Figure 6A ) followed by one of several alternative RTSs (RTSf-m) that could be identified for 13 of the 15 copies (SupFigure16). The latter end with a CCAA motif, like the Hel01/Hel03/Hel04 element copies. At their left end, all 15 copies start with LTSb, homologous to LTSa, or a shortened variant LTSc ( Figure 5A &B, SupFigure16A). Download figure Open in new tab Figure 6. Non-autonomous Helitrons and satellites of clade B Helitrons (Hel15/16). (A) Three examples of the structure of ’Mosaic’ satellite/RTS at the end of Hel15Copy006, Hel15Copy007 and Hel16Copy004. Each colored bar represents an occurrence of a mosaic repeat unit. The RTS are drawn as taller blue rectangles. (B) Three dotplots compare the Hel16Copy005 with itself, with a locus that contains a complete satellite array (right end of sctg002), and this locus with itself. To the right, a schematic comparison of the two loci illustrating loss of the ORF (C) Comparison of the same satellite array found at the extremities of 2 different supercontigs. Dotplots of the 20 kb extremities of supercontigs against themselves show the characteristic series of tandem repeats that appear as squares around the diagonal of identity (SupData4). A motif consisting of at least 100 successive pyrimidines (C or T) was designated “Y_tract”. Dotplots of the 20 kb left end of sctg008 (bottom left) and the 20 kb right end of sctg020 (top right, reverse complement) are shown with these 8 motifs. A dotplot of these two different supercontig ends against each other (top left) shows the same series of motifs, but with different numbers of repeats, so that the satellites now appear as rectangles with a trajectory that diverges from the diagonal of identity. A solid line connecting the rectangles indicates little divergence of the unique sequences between satellite blocks (> 70% identity threshold, 28 nt windows). The diagram (far right) schematizes the expansion or retraction of the tandem repeats. (D) UpSet plot of the association of different satellites found at supercontig ends (n=2 x 187), using the data in SupData4. Each dot indicates the presence of one of the minisatellites. The barplot over the table shows their grouping in supercontig ends. The barplot to the left of the table shows the total occurrence of the different motifs. The black histogram bar shows the supercontig ends with a complete and ordered satellite array (LTSb/c, SAT149, SAT51, SAT41, Y_tract, SAT189, SAT118 and CHBD0). The grey bars show supercontig ends with a truncated and/or deleted satellite array. The very light grey bars represent cases that are difficult to interpret. Alignment of Helitron copies around LTS and RTS not only show that sequences diverge before LTS and after RTS, but also that there are C 4 A 2 telomere repeats before LTS and after RTS, suggesting that the copies may have inserted into the telomere repeats. The example in Figure 5C (Hel01Copy009) has a short run of telomere repeats before LTSa and a longer run after RTSb. That this is the general rule was shown by alignments around all putative LTS (SupFigure17B) and RTS (SupFigure17C for clade A1 and SupFigure17D for clade B), whether or not they are associated with full-size ORFs (Methods). The LTSs are often, but not always, flanked by telomere repeats that are at the very beginning or end of the supercontig (SupFigure17B). Remarkably, the RTSs are always followed by short stretches of telomere repeats with the same orientation. This is particularly clear for RTSa-e associated with cladeA1 (SupFigure17C; see also SupData4 scale drawings of supercontig ends). In all cases where putative LTS and RTS sequences were identified, Helitrons point away from the chromosome’s end towards its central, MAC-destined region: LTS at the end, RTS towards the center. Interestingly, the orientation of full-length ORFs with respect to the orientation of the complete Helitron is variable depending on the element ( Figure 5B ). In clade A1, the full-length ORFs are oriented in the same direction as the Helitron (except for Hel06), whereas clade B Helitron ORFs have the inverse orientation, pointing towards the chromosome end (except for Hel13). One way or the other, the orientation is always the same for all copies of each element ( Figure 5A &B, SupFigure16A, SupData4). The Pgm-independent genomic compartment is restricted to chromosome ends as are most Helitrons with full-size ORFs. Chromosome ends are also enriched in certain minisatellites ( Figure 4 ). We therefore used previous minisatellite annotations for P. tetraurelia ( Figure 1C , SupFigure5) and dotplots, which can reveal imperfect tandem repeats whatever their size, for each supercontig end (SupData4) and each locus with a full-length ORF, to annotate the satellites within and around Helitrons. As shown schematically in Figure 5B , the pattern of LTS-satellites-RTS for complete Helitrons differentiates sub-clades, hereafter designated “types”. Clade A1 contains 5 types and clade B, 3 types. Satellite annotations are drawn to scale with respect to all full-size ORFs in SupFigure16A. The manually curated Mosaic satellite that precedes the putative RTS of Hel15/Hel16 copies ( Figure 6A , SupFigure16, SupData3, SupData4) is unusual in that it is composed from 11 different ∼44 nt segments with distinct sequences, assembled singly or as tandem repeats in an apparently random order. The generally low number of identical tandem repeats in this satellite made it difficult to detect with the automated pipeline (only the “Blue” repeat segment was detected). The Mosaic loci of Hel15/Hel16 copies all end with the “Red” segment and are followed at a short distance (42-168 nt) by the putative RTS (SupFigure16B). The near-uniform size of the repeat units and their amazing random assembly order raise intriguing questions about their origin, dispersal and evolutionary history. Non-autonomous Helitrons are abundant at chromosome ends A search for LTS sequences in the assembly retrieved many copies not directly followed by Helitron ORFs. In the case of LTSa, 10 of them are followed by sequences very similar to the beginning and end of complete type 1 Hel01 copies, from which they may have been produced by recombination of two related satellites, SAT26 and SAT54/81, present 5’ and 3’ of the ORF (SupFigure18). These 10 LTSa-RTSa sequences, 6.5-9.9 kb in length, appear to represent type 1 Hel01 copies that have lost the ORF, as may also be the case for other, truncated loci. In the case of LTSb/c, 7 different satellites occur between the LTS and the ORF of Hel16Copy005 ( Figure 5B , Figure 6B ). The same array of satellites was found, without the ORF, at the ends of many supercontigs (SupData3&4; Figure 6C ). Their order from the end of each supercontig is consistent with the global analysis of satellite DNA distance from supercontig ends ( Figure 4C ). Fifteen of the 62 complete copies that start with LTSb or LTSc, end with the Mosaic satellite and/or an RTSf-RTSm sequence (SupData4). This suggests that the multiple copies of this satellite array may represent defective copies of the Hel16Copy005 type in which the ORF and in most cases the Mosaic-RTS are missing. Given the number of copies of this satellite array that appear to be complete, internally deleted or truncated ( Figure 6D ), we estimate that at least 62 (complete arrays) and up to 196 (including truncated and internally deleted arrays) of these putative non-autonomous Helitrons have inserted at a chromosome end. The conservation of the order of the satellites in the array and of the unique copy sequences that connect them (cf. dotplots: SupFigure4, Figure6C) suggests that a recent burst of transposition activity spread the arrays around the genome, after the divergence of P. tetraurelia from other aurelia species. Similarly, 20 other supercontig ends sharing blocks of P126 minisatellites between LTSb and Mosaic- RTS may be ORF-less derivatives of type 1 Hel15 copies ( Figure 5B ; SupFigure19; SupData4). The annotation of full-length Hel01/03/04 and Hel15/16 elements indicated that they contain a variety of minisatellites which may have been dispersed at the ends of MIC chromosomes by transposition. The annotations further suggested that partial, non-autonomous Helitron copies lacking the ORF are much more frequent than full-length ones. If these characteristics hold for other Helitron families with fewer ORF copies, then a large fraction, if not all, of the Pgm-independent compartment would appear to result from Helitron-mediated transposition. To test this hypothesis, we made a compendium of Helitron full-size ORFs, RTS, LTS and the specific satellites which may be assumed to be part of transposable Helitrons located at supercontig ends (50 kb ends; SupData4). We found that 254 supercontig ends have Helitron annotations and 90% of these end with a Pgm- independent compartment, while 83% of the 74 extremities with no evidence of Helitrons do not end in a Pgm-independent region ( Figure 7A ). A more chromosome-centric vision shows that the 161 chromosome-size supercontigs ( Figure 2D , SupData1) usually have Helitron-associated annotations at one or both ends (both ends, n=98; one end, n=58; neither end, n=5; Figure 7B ). The hypothesis that the Pgm-independent compartment found at chromosome ends results from Helitron transposition is well-supported by current data. Download figure Open in new tab Figure 7. Chromosome ends contain Helitron-associated sequences that define the PGM-independent frontier. (A) The number of supercontig ends for the 161 supercontigs ( Figure 2G ), with and without Helitron evidence at their ends. Helitron evidence can be an LTS, an RTS, a Helitron ORF, or a characteristic Helitron-associated satellite. See SupData4 for details. In both categories, supercontig ends without terminal PGM-independent regions are grey. (B) The number of supercontigs with Helitron evidence on both sides, on one side or on neither side of the end. (C) Heatmap showing PGM RNAi DNA-seq coverage (bin = 1) within a 20 kb region centered on the most internal boundaries of Mosaic satellite loci (n = 169). The loci (rows) are ordered according to the size of the Mosaic of interest. On the left, the black lines show the position of the Mosaic and the grey parts show the flanking regions. Note that a window can contain multiple occurrences of Mosaic. The white areas show regions that cannot be displayed because they are outside the supercontig. (D) Occurrences of CHBD0 closer to each other than 1 kb were fused (n=230) and displayed as in (D), using 30 kb windows. Note that the largest CHBD0 (bottom rows of heatmap) extend beyond the window. A sharp frontier is visible for CHBD0 loci larger than 1 kb. A striking property of the Mosaic satellite, and also of the CHBD0 satellite found in the Hel16Copy005-derived satellite array, is their localization on the Pgm-independent side of the frontier between distal Pgm-independent and more proximal Pgm-dependent genomic compartments. This is illustrated in Figure 7C &D for all occurrences of Mosaic (n = 169) and CHBD0 (n=230), using heatmaps of Pgm read coverage (Methods) in large (10 or 15 kb, respectively) intervals flanking the frontier. In both instances, the innermost, proximal extremity of the satellite corresponds almost exactly to the frontier between the Pgm-independent and Pgm-dependent compartments. We also note that the larger the satellite, the sharper the frontier. In some cases (n=39) both CHBD0 and mosaic were found in the same 30 kb windows (SupFigure20). Overall, the compartment frontier appears to be determined by the largest satellite, Mosaic or CHBD0. DISCUSSION Here, we report chromosome-scale germline genome assemblies for seven members of the Paramecium aurelia species complex and a nearly telomere-to-telomere long-read assembly and genetic linkage map for P. tetraurelia , a model extensively used to study programmed DNA elimination and non-Mendelian heredity [ 81 ]. In addition to providing an important resource, the unforeseen results of our study are (i) a karyotype with ∼160 tiny chromosomes; (ii) a novel clade of Helitron-like elements that have remained active in all aurelia lineages, with 9-10 kb ORFs; (iii) full- length autonomous and numerous non-autonomous Helitron copies that have invaded chromosome ends; and (iv) chromosome ends eliminated very early in development by an unknown mechanism. Costly karyotype Narrowing down the karyotype of P. aurelia has been no small effort given the difficulty of purifying MIC DNA and the absence of genetic or physical maps. Short-read assemblies for 7 Paramecium species allowed us to establish TE landscapes and discover minisatellites although the scaffolds did not usually extend into the telomeric regions. Our long-read assembly overcomes the fragmentation issues of the short-read assemblies for the P. tetraurelia model. It uses DNA from new MACs depleted of Ezl1, a core component of the PRC2 complex required for most DNA elimination (1/3 of IESs are however eliminated). It is important to keep in mind that the MIC EZL1 supercontigs provide the best germline assembly possible given current technological bottlenecks, but they do not constitute the reference MIC genome. Several convergent lines of evidence are in favor of a karyotype for P. tetraurelia with about 160 chromosomes, from 300 kb to 1.2 Mb in size. The first line of evidence is the assembly itself, which includes 161 supercontigs containing MAC-destined sequences. Although only 33 of these supercontigs are telomere-to-telomere, the size distribution of supercontigs capped by telomeric repeats at both ends is similar to that of other supercontigs ( Figure 2 ), which suggests that they are also quasi complete. Based on the independent observation that the Pgm-independent compartment is a marker of chromosome ends ( Figure 3 ), we counted 241 supercontig extremities with Pgm-independent genomic compartments (including 85 supercontigs where they are present at both ends) ( Figure 7A &B, SupData4). Taken together, this is strong support for a karyotype consisting of at least 120, and possibly as many as 161 chromosomes. A second line of evidence comes from analysis of aneuploid clones. Indeed, we noticed that chromosomal segregation errors are frequent in P. tetraurelia . Among the 39 independent F2 clones that we sequenced, 8 (21%) present evidence of higher read coverage (generally twice more) affecting one or more supercontigs, compared to the global median read coverage (SupFigure9A&B). We also analyzed 74 previously published DNA-Seq datasets prepared from 100% homozygous cell lines. We observed 21 additional cases (i.e. in 28% of cell lines, SupFigure9C), which shows that these duplication events are not specific to the F1’s of this particular hybrid genetic cross. These events affect entire MIC supercontigs 92% of the time i.e. the high coverage is seen all along the supercontig. This indicates that there was a segregation error during meiosis leading to extra chromosome copies (i.e. aneuploidy). Under the hypothesis that a supercontig corresponds to a fragment of a chromosome, one would expect the supercontig(s) corresponding to the other fragment(s) to also be affected each time the chromosome is duplicated. Thus, when a supercontig is detected as duplicated in an aneuploid clone, but is not co-duplicated with any other supercontig, we can conclude that the corresponding MIC chromosome does not include any other MAC-destined supercontig. Using this logic, we identified 19 supercontigs for which we could validate a one-to-one correspondence to a MIC chromosome (SupFigure9C). The size distribution of these 19 validated supercontigs is similar to that of other supercontigs (SupFigure9D), and the proportion that are telomere-capped at both ends (4/19=21%) is the same as in the entire dataset (33/161=20%) ( Figure 2G ). Although the analysis does not have the power to be exhaustive, it does show that the 161 supercontigs, even if not telomere-to-telomere, are likely to correspond to individual MIC chromosomes. A last line of evidence comes from analysis of the recombination map. The number of crossovers (COs) per chromosome is generally tightly constrained by both an upper and a lower limit [ 82 ]. A survey covering a wide range of taxa showed that in 97% of cases, each chromosome tetrad receives a minimum of 1 and a maximum of 7 COs per meiosis [ 82 ]. The lower boundary reflects the requirement of having at least one CO to ensure the proper segregation of chromosomes during meiosis [ 82 – 85 ]. Because of this constraint, small chromosomes have a higher recombination rate (in cM/Mb) than larger ones. Interestingly, P. tetraurelia fits perfectly with this pattern: for each of the 161 supercontig groups that could be genotyped, we detected between 0.7 and 3.7 CO per F2 (i.e. 1.4 to 7.4 CO per chromosome tetrad) ( Figure 2A ). Moreover, the size of the supercontigs is negatively correlated with their recombination rate ( Figure 2B ). These observations strongly suggest that P. tetraurelia , like most eukaryotes, must have at least one CO per chromosome tetrad per meiosis. Under the assumption that supercontigs corresponded to fragments of chromosomes, one would have expected many of them to have a genetic length below 50 cM (i.e. below 0.5 CO per F2: Figure 2A , dotted line). The fact that even the shortest supercontig groups are above this threshold suggests they do correspond to (quasi) full-length MIC chromosomes. The reason why Paramecium has so many chromosomes is that its genome has been shaped by a series of at least 3 whole genome duplications (WGDs) [ 7 , 86 ]. Such events have occurred repeatedly in various eukaryotic lineages such as plants [ 87 ], vertebrates [ 88 – 90 ], and yeast [ 91 ]. The pathway of resolution of these dramatic events, whose immediate consequence is to double the number of chromosomes, is that one of each pair of duplicated genes accumulates mutations and eventually decays while the other copy remains under selection, leading to a reduction in genome size. In most taxa, chromosome end-to-end fusion and/or other large-scale rearrangements subsequently reduce the total number of chromosomes back towards the pre-WGD number [ 87 – 89 , 92 , 93 ]. However, Paramecium appears to be less prone to chromosome fusions/rearrangements than most other paleopolyploid eukaryotes: the 3 most recent WGDs remain clearly visible in contemporary genomes, with extensive syntenic blocks of genes and only a few translocations [ 22 ]. Thus, the last 3 WGDs in the Paramecium lineage have led to an 8-fold karyotype expansion, from 2n = 20 chromosomes to 2n = 160. A similar pattern has been reported in ferns (from 2n = 18 chromosomes in Salvinia natans to 2n = 1440 in Ophoglossum reticulatum; [ 94 ]). The high number of chromosomes is expected to increase the risk of producing aneuploid offspring. For instance, in budding yeasts, the rate of meiotic segregation error is 0.15% per chromosome per tetrad [ 95 ]. With such an error rate, the frequency of aneuploid meiosis is expected to increase from ∼2% in a species with 16 chromosomes (like budding yeasts) (1-0.9985^16), to 21% in a species with 160 chromosomes (1-0.9985^160). In P. tetraurelia , we observed that 21% to 28% of clones carry at least one aneuploid chromosome. It should be stressed that these numbers are certainly underestimates of the true chromosomal segregation error rate, because only the viable offspring can be observed. Hence, it is hard to escape the fact that such a high rate of aneuploidy will have a fitness cost. This raises the question as to why Paramecium maintains so many chromosomes. We can hypothesize that part of the answer is that end-to-end fusion of chromosomes, as found in other taxa, may be inhibited by the Helitrons that structure and cap chromosome ends. A new class of Helitrons Eukaryotic Helitrons are a class of DNA transposons proposed to transpose via a rolling-circle replication mechanism [ 96 ]. The first elements to be described lacked terminal inverted repeats but possessed conserved 5’-TC and CTRR-3’ termini as well as a short hairpin close to the 3’ end and were found to insert within 5’-AT-3’ dinucleotides with no target site duplication. Related elements, Helitron2 and Helentron [ 97 ], possess short asymmetric terminal inverted repeats and a hairpin structure at each end. Helitron2 transposons insert within 5’-T(T/C)-3’ dinucleotides while Helentrons insert within 5’-TT-3’ and contain an additional apurinic-apyrimidinic endonuclease domain within the transposase [ 98 – 101 ]. Our phylogeny ( Figure 1H ) shows that all Paramecium elements emerge in a specific clade distinct from the two previously defined clades, HLE1 and HLE2 [ 52 ], supporting ancient acquisition and diversification of Helitrons in Paramecium . We did not find copies of all 29 elements in each of the species examined, which may be due in part to the incomplete and fragmented nature of short-read assemblies at the ends of chromosomes. Nevertheless, elements from all 3 major clades (A1, A2, and B) were found in each species. Although frequent Horizontal Transposon Transfer of Helitrons is well-documented in other eukaryotic taxa [ 102 – 104 ], we found no evidence for Horizontal Transposon Transfer between Paramecium species: the transposase tree is consistent with the species tree for those elements present in several species (SupFigure7B). An additional argument for ancient Helitron acquisition and diversification in Paramecium is the remarkable presence of an IES that interrupts a conserved region of the transposase ORF at exactly the same position in all copies of 5 different elements in an A2 subclade (SupFigure7). These elements are not in the Pgm-independent compartment and at least some copies are eliminated sufficiently late to allow the prior excision of the IES by the Pgm excision complex (SupFigure7 and 14), leaving a narrow window for expression of the transposase from the developing MAC. This IES likely derives from the insertion of a Tc1/mariner element (or a mobile, non-autonomous IES; [ 14 ]) into an ancestral Helitron ORF before diversification of the elements and before subsequent aurelia speciation, which is thought to have started several hundred million years ago [ 105 ]. This indicates that the expression constraints due to the IES have not led to the loss of that subclade, and more generally that Helitron transposition activity persisted over a very large time scale in P. aurelia species. The other available ciliate germline genome assemblies do contain evidence of Helitrons although they are not abundant (< 0.5 % of the genome) and are highly fragmented [ 17 , 18 ]. We found homology to Paramecium Helitron ORFs by tblastn at only a few T. thermophila loci, the best scoring hits located in centromeric regions of some of the 5 MIC chromosomes [ 18 ]. In Oxytricha , more distantly related to Paramecium , 3 of 301 RepeatScout consensus sequences were classified as Helitrons in a genome with very high repeat content [ 17 ], although we failed to obtain any tblastn hit on that genome. It is unclear whether all ciliate Helitrons belong to this same, third clade since the available genomes and annotations are too limited. Two unusual features of Paramecium Helitrons suggest a unique integration specificity that targets them to telomeric repeats. First, the identification of LTS and RTS sequences for clade A1 elements Hel01/Hel03/Hel04 and clade B elements Hel15/Hel16 revealed an absolute orientation bias at chromosome ends, the LTS always being closest to the 5’ end of supercontigs, and the RTS more internal. This is true for both types of elements, even though the ORFs are in opposite orientations in the two cases ( Figure 5B ). While we have not identified the LTS and RTS of other elements, we note that ORF orientation is always the same for all genomic copies (pointing away from the supercontig end or towards it, depending on the element), suggesting that this absolute strand bias is a general feature of Paramecium Helitrons. Secondly, all annotated full-length copies are flanked by CCC(C/A)AA telomere repeats on both sides, though these are often reduced to just a few degenerate repeats 3’ of the RTS of Hel15/Hel16 copies (SupFigure16). These repeats vary widely in length and sequence (alternation of C 4 A 2 and C 3 A 3 repeats) among copies of the same element and thus do not appear to be part of the TEs themselves, unlike the terminal telomere repeats present in the Tc1/mariner element TBE1 in Oxytricha [ 106 , 107 ]. Nor can they be presumed to have been added by telomerase to the ends of some linear transposition intermediate before integration, as proposed for the Tel-1 element in Tetrahymena [ 108 ], since this would result in GGG(G/T)TT repeats 3’ of the RTS instead of the observed CCC(C/A)AA. A more likely hypothesis is that these Helitrons specifically insert into the C-rich strand of telomere repeats: this would constrain their orientation at chromosome ends, fully accounting for the absolute bias observed. Insertion within MIC telomeres would both lengthen chromosomes and create internal blocks of telomere repeats of the same polarity, which are indeed frequently found near supercontig ends. Another remarkable feature is the abundance and variety of satellites, duplications and higher-order repeats present in annotated full-length copies, between the ORF and the conserved ends. Figure 5 and SupFigure16 only show the most conspicuous satellites that could be detected automatically, but closer inspection further reveals old, degenerate satellites, microsatellites and non-contiguous duplications. Together these different types of repeats occupy almost all of the non-coding sequences on both sides of the ORF, which represent from ∼50% up to 75% of the total lengths (SupFigure15). HLEs are prone to capture adjacent genes or other genomic sequences during transposition, and different models have been proposed for how this may occur [ 80 , 97 , 109 , 110 ]. Paramecium Helitrons may have captured satellites – but no gene fragments or other MAC sequences – through preferential integration into the Pgm-independent compartment, which is devoid of any gene but replete with satellites. The occurrence of distinct sets of satellites among copies of the same element (defining Hel01 types 1-5 and Hel15/Hel6 types 1-3, Figure 5 ) indicates that they can change with transposition cycles. It should be noted that the presence of captured satellites between the LTS and the ORF cannot be directly explained by the 3’-bypass model for capture, in which failure to terminate the unwinding of the transposing strand at the end of the RTS results in the inclusion of downstream flanking sequences until another possible termination site is reached. However, it could be explained by insertion of one Helitron copy within another, followed by deletion of the termini and ORF of the internal copy. Nested insertions could also explain the formation of elements with a reversed ORF orientation, as has been proposed for some Fusarium elements [ 101 ]. In some animals and plants, defective copies of Helitron-like elements lacking the ORF were shown to contain various types of satellites, including some with centromeric functions, which are thought to have been dispersed across genomes through non-autonomous transposition [ 100 , 111 – 113 ]. Although ORF-containing, autonomous elements harboring the same satellites are rarely observed in P. tetraurelia , we showed that at least 2 types of potentially autonomous elements (Hel01 type 1, SupFigure18; Hel16 type 3, Figure 7B ) appear to have given rise to ORF-less derivatives which contain the same satellite sets and are present at many more supercontig ends. Their dispersal by non-autonomous transposition may be relevant to the developmental processing of germline chromosomes, since two complex satellites from Hel16 type 3 (CHBD0 and the Mosaic) precisely coincide with the boundary of the Pgm- independent compartment. Helitron-mediated non-autonomous transposition could also have contributed to the multiplication of centromeres -- whose sequences have not yet been identified – as implied by the peculiar karyotype of P. aurelia species, consisting of a large number of very short chromosomes. Our data indicated that the ORFs of Paramecium Helitrons have been under purifying selection. This may simply reflect recent transposition activity, in line with the idea that rapidly disappearing ORFs are continuously replaced by new insertions. An alternative, but not exclusive, explanation is that Helitron transposases may provide advantageous or essential functions to the host. First, they may be required for their own elimination during development of the somatic MAC (since this does not involve the Pgm endonuclease), thus limiting the rate of transposition in the germline MIC. If we assume that the MAC ensures all gene expression, then transcription of a Helitron transposase ORF would originate from an element in the new developing MAC or from an element that failed to be excised from the maternal MAC. Another possibility is that elements that recurrently insert within telomeric repeats may contribute to the protection of chromosome ends against replication- dependent erosion, much as non-LTR retrotransposons do in some Drosophila species lacking telomerase (Saint-Leandre et al. 2019). Paramecium Helitrons may also play some structural role at chromosome ends, for instance in the formation of the “bouquet” during meiotic prophase, a cluster of telomeres attached to the nuclear membrane which is thought to facilitate homolog pairing in many organisms [ 115 , 116 ]. They could even act like the meiotic pairing centers of C. elegans , specialized chromosome sites that stabilize homolog pairing [ 117 , 118 ]. Frequent transposition of such a diverse set of elements at chromosome ends would then rapidly affect homolog recognition between populations, explaining the high rates of F2 lethality frequently observed in crosses between different isolates of the same species [ 15 ], and possibly driving speciation. Another exciting possibility is that Helitron transposases may contribute to the programmed rearrangement of rDNA loci. Indeed, in Paramecium , 9 kb rDNA units encoding 18 and 28S rRNA are found as tandem arrays within dedicated MAC chromosomes, whereas only 4 single rDNA units have been found in the MIC genome. These MIC loci contain a little more than a unit, with direct repeats of ∼ 300 bp at either end, and differ from each other by the sequence of the non-transcribed spacer [ 119 ]. It was proposed that these MIC units are excised as circles via homologous recombination between the ends during MAC development, and that rolling-circle replication then generates tandem repeats. This model would explain the observation that adjacent units in the MAC tandem repeats usually have the same spacer [ 119 ]. The tempting hypothesis that a Helitron transposase is involved in this process would reveal a second example of TE domestication, alongside that of the Pgm endonuclease [ 27 ]. Models for Pgm-independent early elimination of chromosome ends The germline genome of P. tetraurelia is organized into distinct regions based on their developmental fate. The constitutive and the variable MAC regions are, respectively, always and partially retained in the new MAC. The MIC-limited region, which includes IESs, TEs and satellites, is consistently eliminated during MAC development. Previous studies have demonstrated that separate yet overlapping pathways are involved in the Pgm-dependent elimination of IESs and TEs [ 34 – 36 ] ( Figure 8A ). Our findings now reveal the existence of a completely novel pathway for the elimination of the Pgm-independent compartment, first reported by [ 16 ]. This compartment is eliminated very early, before IESs are excised ( Figures 3C and 8A ) [ 26 ]. This newly defined compartment is located at the ends of germline chromosomes and consists of most Helitrons with full-size ORFS and multiple non-autonomous Helitron derivatives, including the CHBD0-containing satellite array ( Figures 5 - 7 , SupFigure15-17). The latter is positioned directly adjacent to a Pgm- dependent compartment, prompting the question of how the boundary between these two compartments is defined. We have shown that Helitron-associated Mosaic and CHBD0 satellites define the boundary ( Figure 7C-D ). Interestingly, predicted G4s are not found in the Mosaic satellite, but are significantly enriched within the CHBD0 repeat (70% of predicted G4 are in CHBD0) located in the proximal part of the Pgm-independent compartment ( Figure 4A ). This observation suggests that G4s may play a role in defining the boundaries between the compartments. This is reminiscent of the function of the G4-binding protein Lia3, which plays a key role in definition of the boundaries of germline-limited sequences in Tetrahymena [ 120 ]. Lia3 interacts with G-rich polypurine tracts (A 5 G 5 ) located on either side of germline-limited elements, which are eliminated by the Pgm homolog, Tpb2p. In the absence of Lia3, alternative boundaries are used. Notably, no homolog of Lia3 has been identified in the Paramecium genome as Lia3 appears to be specific to Tetrahymena . Download figure Open in new tab Figure 8. Models for Pgm-independent early elimination of chromosome ends. (A) MIC chromosomes end with sequences eliminated independently of Pgm endonuclease activity (pink), and contain Helitrons and associated satellites. The elimination of these regions involves the scnRNA RNA pathway and PRC2. These regions are eliminated earlier in new MAC development (before 32C endoreplication state) than Pgm-dependent regions (purple). Pgm-dependent DNA elimination [ 26 ] occurs between 32C-128 C endoreplication state and concerns all IESs and remaining TEs. The MAC-variable regions (seagreen) are not eliminated completely and contribute to the heterogeneity of MAC chromosomes [ 6 ]. The gene-rich constitutive-MAC (blue) regions are found in all MACs. (B) Model to explain the mechanism behind the disappearance of PGM-independent/Early eliminated regions. The first hypothesis is that a factor binds a motif (e.g. the mosaic satellite) or a DNA structure (e.g. G4) and this binding promotes introduction of a double-strand break (DSB), so that the PGM-independent region would not be replicated. (C) A second, non-exclusive hypothesis is that a particular DNA conformation, such as G4, would stall the progression of the DNA polymerase resulting in under-replication of the end of the chromosome. To explain the elimination of the Pgm-independent compartment, we propose two alternative, non- exclusive models. In the first scenario ( Figure 8B ), the presence of the Mosaic satellite or G4s promotes the formation of DNA double-strand breaks at the boundary between the two compartments. One possibility would be that the Mosaic sequence and the G4 structures mediate the recruitment of an unknown endonuclease either directly or indirectly through a G4/Mosaic- binding protein, and this would eventually lead to the elimination of the G4/Mosaic-containing region. Similarly, during class switch recombination in B cells, double-strand breaks are initiated by base modification catalyzed by AID (activation-induced deaminase) that directly binds to G4s [ 121 ]. Alternatively, one can imagine that G4 structures formed within the CHBD0 are not properly resolved and stall the DNA replication machinery, which can also cause DNA double-strand breaks [ 122 , 123 ]. Indeed, highly stable G4s can form during replication, and the failure to resolve G4s results in genome instability. G4-associated genome instability appears to largely result from the stalling of the replicative polymerase at the G4 structure and subsequent uncoupling between the replicative helicase and the DNA polymerase [ 124 ]. In the second scenario ( Figure 8C ), the stalling of the replicative DNA polymerase at G4 structures results in under-replication of the Pgm-independent compartment. We imagine here that the resolution of G4 structures during DNA replication prevents mutagenic events, including double- strand break formation. However, in the absence of G4 resolution, the prolonged stalling of the DNA polymerase at G4s and a permanent uncoupling between the replicative helicase and the polymerase would lead to under-replication of the G4-containing region. This scenario is very reminiscent of common fragile sites in vertebrates [ 125 ] and raises the possibility that fragile sites, like the one responsible for some Paramecium cortical mutants [ 126 ], could be developmentally programmed for DNA elimination, via under-replication ( Figure 8C ). To explain why depletion of the scnRNA-PRC2 pathway results in normal replication of the G4- containing region, we envision that G4s are resolved under such conditions, preventing under- replication. Unfolding of G4s requires the functions of specialized DNA polymerases, helicases, and single-stranded DNA binding proteins [ 124 ]. Remarkably, Helitrons contain a putative Pif1 helicase, known in other organisms to be involved in G4 resolution [ 124 ]. The expression of at least one gene required for G4 resolution might be negatively controlled by the scnRNA-PRC2 pathway, perhaps through the deposition of repressive histone marks. Under normal conditions, the gene would be repressed, leaving the G4s unresolved, while it would no longer be repressed in scnRNA-PRC2 depleted cells, allowing the resolution of G4s and endoreplication of the region. It is also conceivable that, similarly to G4s, Mosaic can form DNA structures that become destabilized in the absence of the scnRNA-PRC2 pathway. Further investigation will be necessary to elucidate the underlying mechanisms. Our data reveal a unique mechanism that specifically eliminates the ends of germline chromosomes. Interestingly, the loss of sub-telomeric and telomeric sequences of the germline chromosomes also occurs in Tetrahymena [ 18 ] and all nematode species where PDE has been reported [ 127 – 131 ]. In the free-living nematode Oscheius tipuale , no internal break site is generated and only the sub- telomeric regions that contain Helitron-like elements are eliminated [ 128 , 129 ]. A conserved 30-nt palindromic motif is associated with the break site [ 129 ], while no sequence feature has been identified in the parasitic nematodes Ascaris and Parascaris [ 127 , 132 – 134 ]. So far, there is no report of a protein that physically interacts with the motif or at the break sites, and the underlying mechanisms remain to be discovered. Similarly to the potential role of G4s in P. tetraurelia , DNA structures may be important for the removal of germline chromosomal ends in these organisms. The recurrent elimination of the terminal regions of germline chromosomes across various phyla may indicate convergent evolution and suggests that these germline-specific regions could have a crucial role for the chromosomes, that is yet to be discovered. Conclusions The key finding of our study is that the germline genome of P. aurelia consists of numerous tiny chromosomes (n ∼160, 300 kb to 1.2 Mb in size), with an exceptionally high recombination rate (420 cM/Mb). The chromosomes terminate in a novel genomic compartment that appears to be the result of transposition of a new class of Helitrons, that insert in C 4 A 2 telomere repeats. This compartment is eliminated very early during development of the somatic nucleus, by a mechanism independent of the well-studied PiggyMac domesticated transposase complex that was thought until now to be responsible for all transposon elimination. MATERIALS AND METHODS Strains and cell culture Paramecium cells were grown in wheat grass powder (WGP) (Pines International) infusion medium bacterized the day before use with Klebsiella pneumoniae , unless otherwise stated, and supplemented with 0.8 µg/mL β-sitosterol (Merck). Cultivation and autogamy were carried out at 27°C as described [ 135 , 136 ]. The Paramecium strains used in this study are listed in SupTable1: tetraurelia (strain 51, unless stated otherwise), octaurelia 138, biaurelia V1-4, pentaurelia 87, primaurelia AZ9-3, sonneborni ATCC 30995, sexaurelia AZ8-4. Illumina sequencing and assembly of vegetative MIC DNA Previously published [ 14 ] MIC DNA small-insert libraries of vegetative MIC DNA were used. For assemblies, read pairs containing a MAC IES junction, identified using ParTIES software [ 78 ] and MAC reference genomes [ 11 , 14 ], were removed to reduce ambiguity. The 250 nt filtered read pairs were fused with Flash version 1.2.11 [ 137 ] using default parameters to obtain ∼450 nt fragments. The fragments were assembled into MIC contigs by the Newbler overlap-layout-consensus assembler (version 2.9 Roche Diagnostics), with parameters (-mi 100 -ml 99). Because of the DNA quantities required for mate-pair libraries, we used DNA from new MACs after PGM silencing by RNAi ( PGM DNA) [ 25 ]. It was not possible to obtain PGM -RNAi DNA for P. sonneborni , so only contigs are available for this species. Large-scale cultures of tetraurelia 51, octaurelia 138, biaurelia V1-4, pentaurelia 87, primaurelia AZ9- 3, and sexaurelia AZ8-4 were collected at the end of MAC development in PGM RNAi conditions. Whole genomic DNA was isolated and used to construct four Illumina mate-pair libraries with insert sizes of 3-5 kb, 5-8 kb, 9-11 kb and 14 kb (only the first three sizes were available for P tetraurelia ) to build scaffolds with the MIC contigs. The scaffolder SSpace basic v2 [ 138 ] was used, with parameters (-a 0.5 -k 7). Gap closing was accomplished with SOAPdenovo2 GapCloser software [ 139 ]. Assembly statistics are described in SupTable1 and sequencing datasets used or generated for this study are described in SupTable2. Sequencing DNA from sorted new developing MACs after RNAi-depletion of factors As in [ 140 ], large-scale, 500 ml cultures of P. tetraurelia were collected at T30 (30 hours after T0 of autogamy, time at which 50% of cells have fragmented old MACs). RNAi KD was performed using L4440 derivatives carrying the following inserts: DCL2/3 , DCL2/3/5 , DCL5 [ 35 , 141 ] or TFIIS4 [ 36 ]. As detailed in [ 26 ], nuclei of the RNAi-depleted cells were enriched by centrifugation through a 2.1 M sucrose layer prior to flow cytometry using PgmL1 antibody labeling to separate new developing MACs from old MAC fragments. Sequencing librairies were then made using total DNA from the sorted new MAC fraction with the NEBNEXT Ultra II DNA library Prep Kit (New England Biolabs) for DCL2/3 , DCL2/3/5 , DCL5, or as previously described [ 26 ] for TFIIS4 . The libraries were subjected to paired-end sequencing on the Illumina platform. Oxford Nanopore sequencing and assembly of P. tetraurelia EZL1 -RNAi DNA Unrearranged DNA was extracted from new developing MACs of Paramecium tetraurelia Ezl1- depleted cells [ 34 , 46 ]. Four libraries were generated and sequenced using Oxford Nanopore Technology (ONT). The reads were filtered with Porechop ( https://github.com/rrwick/Porechop ). Reads containing a MAC IES junction were removed to reduce ambiguity, as described above for the Illumina assemblies. Data were assembled with SMARTdenovo [ 142 ] and polished with Racon [ 143 ] using the ONT reads as in [ 144 ]. The assembly was then subjected to two successive waves of polishing with Pilon [ 145 ] (v1.23 --mindepth 10 --minqual 30) using the sorted vegetative MIC Illumina reads. We used the recombination map (see next paragraph) to break chimeras or join contigs to obtain supercontigs. The assembly "ptetraurelia_51_EZL1_SmartDeNovo_v1.0" is throughout the manuscript referred to as the “MIC EZL1 assembly” (see SupTable1). Linkage map and recombination rate for Paramecium tetraurelia To build a genetic map of P. tetraurelia , we performed genetic crosses between strain 51 and strain 32. F1 clones from several different pairs of conjugants were allowed to undergo autogamy to generate entirely homozygous F2 clones. Total cell DNA was extracted from 42 independent F2 clones as described in [ 146 ]. The DNA samples were sequenced to an average read depth of ∼50X within MAC-destined regions (given that we extracted total DNA, most of sequence reads derive from the MAC genome, and hence we have a very limited coverage in MIC-restricted regions). Reads were mapped to the P. tetraurelia strain 51 MIC genome assembly with BWA-MEM v.0.7.15 [ 147 ] and then filtered with SAMtools v.1.3.1 [ 148 ]. Duplicate reads were removed by Picard tools v.1.98 [ 149 ]. We called variants with GATK v.3.3 [ 150 ]. This variant calling procedure was replicated using the P. tetraurelia strain 32 MAC genome assembly. To build the genetic map, we considered only biallelic SNPs (excluding indels) with good quality scores that were called consistently on both genome assemblies. Given that F2’s underwent autogamy, they are expected to be entirely homozygous. We therefore excluded SNPs that were genotyped as heterozygous. We also excluded SNPs that showed strong deviations from Mendelian expectations (see details in Supplementary Methods). We retained 220,944 SNPs to build the genetic map. Two individuals (F2.11 and F2.12) were excluded because they derive from the same meiosis as F2.10. We also excluded one individual (F2.7) that appeared to result from an abnormal autogamy process (most of its scaffolds are heterozygous). In the end, our analysis of meiotic recombination events is based on 39 independent F2’s. The strain 51 MIC reference genome consists of 187 supercontigs, including 164 supercontigs that match scaffolds of the MAC assembly. We identified three cases where two supercontigs encompassed contiguous segments of a same MAC scaffold, indicating that they derive from a same MIC chromosome. These pairs were grouped, resulting in a total of 161 supercontig groups matching MAC scaffolds. We excluded the 23 supercontigs that do not encompass any scaffold of the MAC assembly. The 161 supercontig groups that we selected to build the genetic map cover 101.4 Mb (97.3 % of the MIC reference genome assembly). For each F2, we identified crossover (COs) and non-crossover (NCOs) recombination events (see details in Supplementary Methods). Because of the low number of callable SNPs in MAC-variable and MIC-limited regions (0.06 SNPs per kb), we had very limited power to detect recombination in these regions. The SNP density is also limited in IESs (0.27 SNPs per kb compared to 3.0 SNPs per kb in MAC-constitutive regions), but COs can nevertheless be detected in IESs thanks to the presence of SNPs in flanking MAC-constitutive regions. Thus, to quantify recombination rate, we restricted our analyses to MAC-constitutive regions and IESs, that cover 74 Mb (70.9% of the MIC genome) and include 99.3% of all detected COs. We quantified CO interference using the coefficient of coincidence (CoC), as described in [ 151 ] (see details in Supplementary Methods). Collapse estimation Based on the coverage distribution shown in SupFigure12, the genomic regions with coverage above the peak were extracted, then the coverage was divided by the median coverage to obtain a collapse score. The estimate of global genome collapse was calculated by multiplying the size of the regions by the collapse scores. Assembly Completeness Kmer Analysis Toolkit v. 2.4.2 [ 54 ] was used to evaluate the completeness of genome assemblies, using K-mer frequency histograms (k = 27) and vegetative MIC DNA-seq datasets for the 7 P. aurelia species. MIC Telomeres To identify MIC telomere repeats, MIC Illumina paired-end reads improperly mapped to the corresponding MAC assembly were analyzed using Jellyfish [ 152 ] to count k-mers (k-mer lengths of 18, 21 and 24 nt) corresponding to 3 exact repeats of 6, 7 or 8 nucleotide microsatellite repeat, taking into account all permutations. Aside from repeats containing only A and T, the top abundance k-mers include 18-mers with 3 repeats of 5’-CCCCAA-3’ and of 5’-CCCAAA-3’ hexamers. For further analyses, we used the Perl regular expression (CCC(C|A)AA){3,} to identify telomere repeats: at least 3 occurrences of the same mixture of C 4 A 2 and C 3 A 3 hexamers found in Paramecium MAC DNA [ 76 ], starting with validation that perfect triplets of either 5’-CCCCAA-3’ or 5’-CCCAAA-3’ are restricted to the extremities of the P. tetraurelia EZL1 ONT supercontigs, unlike the highly abundant ((A|T){6}){3} k-mers, which are distributed all along supercontigs (SupFigure11 ). To determine the junctions between MIC telomere repeats and adjacent sequence, we defined telomere reads to be P. tetraurelia MIC Illumina HiSeq 101 nt read pairs (N = 90703803) that contained, in at least one read, a match to at least 3 consecutive repeats of a mixture of GGGGTT and GGGTTT hexamers, in either orientation. Telomere read pairs (N = 29771; 0.0328% of total read pairs) were fused using FLASH v1.2.11 [ 137 ] with default parameters to yield 20861 extended fragments of mean size 163 nt. PCR duplicates (N=93) were removed to yield 20768 fragments of mean size 165 nt. Telomere Repeat Clusters were generated by clustering telomere-repeat containing fragments with cd-hit-est version 4.8.1 (-c 0.99 -n 10 -G 0 -A 75 -g 1 -gap -8 -gap-ext -3) [ 153 ], after removing the telomere repeats to avoid spurious clustering. With these options, clustering in “accurate” mode is based on a local alignment requiring 99% nucleotide identity using a 10 nt word size, with gap and gap extension penalties slightly larger than default (default is -6 and -1, respectively), the alignment covering at least 75 nt for each sequence in the cluster. The complete sequences (i.e. with the telomere repeats) of each cluster were aligned using MAFFT (v7.40) (option - adjustdirectionaccurately) [ 154 ]. Heatmap overviews of the Telomere Repeat Cluster alignments were generated using a custom R script (SupData2). Alignment consensus sequences were mapped to genomes using Bowtie2 (v2.3.4.3 --very-sensitive-local) [ 155 ]. The mapping revealed some redundancy in the set of junctions, since some loci were identified by more than one telomere junction consensus sequence. This included cases where an internal junction and an edge junction mapped to the same location as well as cases where two clusters in antisense orientation mapped to the same location. Synteny between MAC and MIC sequences Assemblies were masked using RepeatMasker (v-open-4.0.8 -no_is -nowlow -x -species paramecium). For each species, MAC and MIC sequences were grouped using RagTag (v1.0.2 ragtag.py scaffold -w -f 2000 -I 0.5 -u) [ 156 ]. All MIC/MAC sequence groups were manually curated, using Circos (v0.69-6, [ 157 ] to display the sequence alignments (nucmer v3.1 filtered with -i 0.99 -l 1000, [ 158 ]. The SupData1_synteny_MIC_MAC.html file shows the graphics used to curate the comparison between the P. tetraurelia EZL1 ONT assembly and the MAC assembly. Genomic compartments IESs from [ 14 , 25 ] were remapped on MIC genomes using a custom Perl wrapper (remap_IES.pl, see Zenodo) and Bowtie2 [ 155 ] (v2.3.4.3 --very-sensitive) alignments guided by the groups based on synteny between MAC and MIC sequences. As previously documented [ 26 ] we characterized “buried” IESs on MIC-limited regions using ParTIES (MILORD method, [ 78 ]). We determined non-overlapping genomic compartments: MAC-constitutive, MAC-variable, Other Eliminated Regions (OES, eliminated regions apart from IESs). For this purpose, we used MAC DNA and MIC DNA bigwig (bin=1) sequencing coverage files. For each dataset, the threshold was adjusted (see SupFigure12_Compartments). For each MIC genome, the MAC-constitutive compartment was defined by regions with a MAC coverage ≥ ∼100X after fusing contiguous regions (R GenomicRanges package, reduce method, min.gapwidth=100). Previously annotated IESs were excluded. Remaining regions selected for a MAC coverage ≥ 5X, were fused as previously described (min.gapwidth=100nt), then filtered for regions longer than 100nt defining the MAC-variable compartment. The OES compartment was defined by remaining regions with a MIC DNA coverage ≥ 5X, fused (min.gapwidth=100nt), then filtered for regions longer than 100nt. SupFigure12_Compartments shows the proportion of each compartment for each MIC genome. The OES compartment for the EZL1 ONT assembly was divided into two sub-compartments: PGM - dependent and PGM -independent, using the same threshold strategy (≥ 5X, then fused at min.gapwidth=100 and filtered for segments with length ≥ 100nt) on PGM DNA. Gene annotation Previously annotated genes on MAC genomes [ 11 , 53 ] were remapped to MIC genomes using a custom Perl script (remap_gene.pl, see Zenodo) that uses refined BLAT (v36x2 -fine -maxIntron=500) [ 159 ] alignments guided by MAC/MIC sequence groups (see previous section). On each MIC genome the MAC-destined compartment (MAC-constitutive and -variable; see previous section) was masked using bedtools (v2.26.0 maskfasta). The remaining regions (MIC-limited regions) were annotated for genes using EuGene (v4.1) software and the procedure previously described [ 53 ]. Only genes with a length ≥ 100nt and G+C content < 50% were kept. Transposable Element landscape The TEdenovo module of REPET version 2.5 [ 160 ] was used for identification of Transposable Elements (TE) for each Illumina assembly, after adding all available curated Paramecium TE [ 16 ] to Repbase [ 161 ] v19.06. Default TEdenovo parameters were used (except for BLR_sensitivity : 0; TRFmaxPeriod : 500) and the MIC-specific (masked constitutive MAC) assembly was used. TEannot was run with the same parameters on the unmasked MIC assembly, and the annotations were used for human curation of the reference TEs proposed by TEdenovo. This led to discovery of Helitrons in Paramecium genomes. For the subsequent round of annotation, manually curated Helitron ORFs and the reference TEs for all species were pooled and used to annotate each assembly, including the P. tetraurelia EZL1 long-read assembly, with TEannot. Coverage of each assembly by different TE superfamilies was calculated using custom R scripts and the Genomic Ranges package (code and data in Zenodo). TE protein domains and phylogeny To identify TE independently of REPET, we searched for 24 PFAM domains curated from TEs frequently found in fungi [ 162 ] in the MIC Illumina assemblies. The constitutive MAC was masked using RepeatMasker and the genomes were translated in 6 reading frames using the transeq program of the EMBOSS suite ( http://emboss.open-bio.org/ ). HMMER 3.2. ( http://hmmer.org ) hmmsearch was used to search for the TE domains in the translated genome assemblies using default parameters. Using a cutoff domain e-value of 0.0001 to obtain only some of the least decayed TE copies, the nucleotide sequences corresponding to each hit for the genomes under consideration were extracted then aligned with MACSE v. 2.07 [ 163 , 164 ]. SeaView version 5.0.4 [ 165 ] was used to build a phylogenetic tree with phyML [ 166 ] (default parameters) from the MACSE alignment for each PFAM domain. Annotation and visualization of trees was enhanced using the R ggtree package [ 167 , 168 ]. dN/dS computation The phylogenetic tree of full size Helitron transposase copies was computed with phyML [ 166 ], based on the protein sequence alignment obtained with MAFFT using the L-INS-i method [ 154 ]. We used bio++ v.3.0.0 libraries [ 169 – 171 ] to estimate the dN/dS on each branch of the phylogenetic tree. In a first step, we used a homogeneous codon model implemented in bppml to infer the most likely branch lengths, codon frequencies at the root, and parameters of the YN98 (F3X4) substitution model [ 172 ], which allows for different nucleotide content dynamics across codon positions. In a second step, we used the MapNH substitution mapping method [ 173 ] to count synonymous and non-synonymous substitutions along each branch [ 174 ]. Minisatellite identification Our strategy was to use the sensitive MREPS software to identify tandem repeats in assemblies, and TAREAN to identify tandem repeats in sequencing reads. MREPS [ 175 ] v2.6 (-minsize 300 -minperiod 30 -exp 6 -res 150) was used to find minisatellite DNA with a repeat period of at least 30 nt in genome assemblies. Custom R scripts were used to align and build a consensus from the repeats at each locus. The consensus sequences were clustered using cd-hit-est (version 4.8.1) [ 153 ], with a first set of parameters for small repeats (-c 0.8 -n 5 -A 100 -g 1 -G 0 -sc 1 -r 0 -d 0) and a second set for longer repeats (-c 0.9 -n 9 -A 30 -g 1 -G 0 -sc 1 -r 0 -d 0). To identify minisatellite using sequencing reads, TAREAN software [ 176 ] was used with the default parameters for paired-end Illumina reads. For each species, reads were trimmed with sickle software [ 177 ] at quality -q 20, and trimmed reads were sampled to obtain reads that correspond to ∼ 0.5X to ∼1X estimated coverage of the genome. Six independent samples of reads were run for each species. The consensus sequences for the six runs were compared by blastn to determine a non-redundant set of consensus sequences for each species. The non-redundant sets of consensus sequences (for both MREPS and TAREAN workflows) were mapped to assemblies using RepeatMasker version open-4.0.8 (-s -gccalc -no_is -nolow) [ 178 ]. One TAREAN limitation is the requirement of a certain abundance of a repeat in the genome for detection. G-quadruplex identification The R package pqsfinder version 2.18.0 (-deep=TRUE and a min score of 70) [ 179 ] was used to computationally identify potential quadruplex-forming sequences in an assembly. Dotplot comparisons Dotplots were used for fine-grained visualization of tandem repeats that were too degenerate or too small for detection by the automated minisatellite pipelines. For dotplots of sequences of the same length, a custom R script was used to plot a distance matrix using the DistanceMatrix method of DECIPHER (v2.28.0; window 28, Hamming distance threshold 8) and ggplot2_v3.5.1. For non-self dotplots of Helitrons, the EMBOSS v6.6.0.0 [ 180 ] dotmatcher function was used (-graph data - windowsize 28 -threshold 54 or -threshold 60) to generate the matrix, which was drawn using R version 4.3.0 graphics::plot. Software mRNA-seq data were mapped using HISAT2 (v2.1.0 --min-intronlen 20 --max-intronlen 100) [ 181 ]. DNA-seq data were mapped using BOWTIE2 (v2.2.9 --local ) [ 155 ]. Small RNA-seq data were mapped using BWA (aln v0.7.15 0 no mismatch) [ 182 ]. Read counts on sequence features (genes,TEs, minisatellite) were calculated with htseq-count (v0.11.2, -stranded=yes) [ 183 ] on mappings after quality filtering with SAMtools (v1.9, q>=30 for data mapped on Illumina assemblies and q>=0 data mapped on ONT assemblies) [ 148 ]. Data were analyzed and displayed using R (v4.3.0) statistical computing framework [ 184 ] and R packages (Biostrings-v2.70.1 [ 185 ], ComplexHeatmap-v2.18.0, DECIPHER-v2.28.0 [ 186 ], GenomicRanges-v1.54.1 [ 187 ], ggdist-v3.3.1 [ 188 ], ggtree-v3.10.0 [ 167 ], karyoploteR-v1.28.0 [ 189 ], magrittr-v2.0.3, rtracklayer-v1.62.0 [ 190 ], seqinr-v4.2-36 [ 191 ], tidyverse-v2.0.0 [ 192 ], UpSetR-v1.4.0 [ 193 ]). Data availability Sequencing data are available in the European Nucleotide Archive under Project PRJEB98202 ( https://www.ebi.ac.uk/ena/browser/view/PRJEB98202 ) and Project PRJEB98215 ( https://www.ebi.ac.uk/ena/browser/view/PRJEB98215 ) and are described in the SupTables 1 & 2. The statistical data and scripts (bash, R or perl) used to generate results and images were deposited at Zenodo ( https://doi.org/10.5281/zenodo.17205922 ). The SupData1 – SupData4 mentioned in this article can be downloaded from Zenodo ( https://doi.org/10.5281/zenodo.17206076 ). All genome and annotation files are available from the ParameciumDB [ 194 ] download section ( https://paramecium.i2bc.paris-saclay.fr/download/ ). Abbreviations CHBD0 Chessboard 0 minisatellite CO Crossover DCL Dicer-like DEV Development EZL1 Enhancer of Zest-like 1 FRAG Fragmentation HLE Helitron-like Element IES Internal Eliminated Sequence LINE Long Interspersed Nuclear Element LTR Long Terminal Repeat LTS Left Terminal Sequence MAC Macronucleus MDS MAC-destined sequence MIC Micronucleus MLS MIC-limited Sequence MITE Miniature Inverted Repeat Transposable Element NCO Non-Crossover OES Other Eliminated Sequence ONT Oxford Nanopore Technology PDE Programmed DNA Elimination PGM PiggyMac PRC2 Polycomb Repressive Complex 2 RPKM Reads per kilobase per million mapped reads RTS Right Terminal Sequence SINE Short Interspersed Nuclear Element SNP Single Nucleotide Polymorphism TE Transposable Element TFIIS4 transcription elongation factor II paralog 4 TIR Terminal inverted repeat DNA transposon WGD Whole Genome Duplication FUNDING This work was supported by the Centre National de la Recherche Scientifique ( https://cnrs.fr ), by the Agence Nationale de la Recherche ( https://anr.fr ) (ANR-18-CE12-0005 to EM, MB, LD, SD; ANR-19- CE12-0015 to SD, OA; ANR-11-INBS-0012 to HQ; ANR-23-CE12-0027 to SD; ANR-25-CE12-7757 to SD), by the Fondation pour la Recherche Medicale ( https://www.frm.org )( Equipe FRM EQU202103012766 to MB and Equipe FRM EQU202203014643 to SD). We acknowledge the French Infrastructure for Integrated Structural Biology (FRISBI, https://frisbi.eu ) for use of the flow cytometry facility of I2BC (ANR-10-INBS-0005). Work was performed using the URGI facilities ( https://doi.org/10.15454/1.5572414581735654E12 ) supported by the French government through the ANR, as part of the France 2030 program for research infrastructure (ANR- 21-ESRE-0048). The sequencing effort was funded by France Génomique ( https://www.france-genomique.org ) through involvement of the technical facilities of Genoscope (ANR-10-INBS-09-08) to SD. We also acknowledge the sequencing and bioinformatics expertise of the I2BC High- throughput sequencing facility, supported by France Génomique (funded by the French National Program “Investissement d’Avenir” ANR-10-INBS-09). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests The authors declare that they have no competing interests. Author contributions OA played a major role in data curation, formal analysis, software development, and preparation of the manuscript. GP, IN and AP performed genetic crosses and sample preparation for the genetic linkage map. AC, JMA and KL were involved in aquistion and formal analysis of the sequence data. JA, IL and HQ played a major role in annotation, formal analysis and interpretation of repeated sequences. FG, CMP, SB, AF, SM, NM, NS, VR, AT, AdV and CZ performed RNAi and flow cytometry experiments and prepared samples for sequencing. EL annotated and interpreted repeated sequence data. MB, LD, EM, SD and LS were responsible for conception, supervision and funding of the project. OA, MB, LD, EM, SD and LS wrote and revised the manuscript, and contributed to data visualisation. LD, EM and LS performed formal analysis, annotation and data curation. KL and SD coordinated the project. All living authors have approved the manuscript. Supplementary Information HTML files SupData directories contain an html file for interactive exploration of supplementary data, the Rmd script and data used to generate the html file, and in some cases, embedded images. The html file can be opened with a web browser (JavaScript should be enabled in the browser Preferences). SupData1_Synteny_MIC_MAC Synteny between P. tetraurelia supercontigs of the MIC EZL1 -RNAi assembly and scaffolds of the MAC v2 assembly In the Circos drawings, from the center to the periphery, the arcs (red, sequence and query in the same orientation; blue, sequence and query in opposite orientations) show the syntenic regions identified by nucmer (v3.1, filtered with delta-filter -m -i 0.99 -l 1000). The darker the color the greater the sequence identity. The curves give coding gene density (orange), TE density (blue), and satellite (tandem repeat) density (green). The histograms show PGM- RNAi coverage (blue), MAC coverage (purple) and MIC coverage (red). The interactive table summarizes the data and allows the user to customize table appearance, sort by column, download a csv or excel file and search for a supercontig or scaffold. It is possible to scroll through the images without displacing the table and to jump from any image back to the table. This data was used to prepare Figure 2D . SupData2_TelomereJunctions Telomere Repeat Cluster overviews The interactive table describes intrinsic characteristics of each telomere repeat cluster, its mapping to different genome assemblies and provides a link to the overview images, which can be scrolled independently of the table. This data was used to prepare Figure 2E & F. SupData3_Poster Exhaustive overview of supercontigs and their annotations The 187 supercontigs of the P. tetraurelia EZL1 -RNAi genome assembly were drawn to scale with annotations of telomere repeats, satellites, curated Helitron ORFs, curated HLE start and end (LTS and RTS) consensus sequences, REPET automated annotations of LINE and TIR superfamily elements and IESs (detailed legend at the top of the html file). The drawings also show MAC coverage and genome compartments. The interactive table summarizes information about genome compartments and provides links to the drawings. SupData4_EndGame Zoom of supercontig ends The 50 kb at the ends of each of the 187 supercontigs of the P. tetraurelia EZL1 -RNAi ONT genome assembly are drawn to show annotations of telomere repeats, satellites, curated Helitron ORFs, Helitron start and end consensus sequences (LTS, RTS), REPET automated annotations of LINE and TIR superfamily elements and IES (detailed legend in the html file). The drawings also show MAC coverage and genome compartments. The interactive table summarizes the information and provides links to the drawings. Each drawing is accompanied by a link to a dotplot of the first 20 kb of the end with annotation of minisatellites. ACKNOWLEDGEMENTS We acknowledge use of the computing facilities of the CC LBBE/PRABI and the BIOI2 platform and computing facilities of the I2BC. We thank Pascaline Tirand, Fanny Culot, Nelly Sainsard, Cristina Delawarde and Vincent Maupu-Massamba for technical assistance. We dedicate this article to our dearly missed colleagues Janine Beisson and Sophie Malinsky, who played essential roles in the international ciliate research community and in our French Paramecium laboratories. Funder Information Declared Agence Nationale de la Recherche , ANR-18-CE12-0005 , ANR-19-CE12-0015 , ANR-11-INBS-0012 , ANR-23-CE12-0027 , ANR-25-CE12-7757 Fondation pour la Recherche Médicale , EQU202103012766 , EQU202203014643 Footnotes ↵ § Deceased REFERENCES 1. ↵ Parfrey LW , Lahr DJG , Knoll AH , Katz LA . Estimating the timing of early eukaryotic diversification with multigene molecular clocks . Proc Natl Acad Sci U S A . 2011 Aug 16; 108 ( 33 ): 13624 – 9 . OpenUrl Abstract / FREE Full Text 2. ↵ Cheng CY , Orias E , Leu JY , Turkewitz AP . The evolution of germ–soma nuclear differentiation in eukaryotic unicells . Curr Biol . 2020 May; 30 ( 10 ): R502 – 10 . OpenUrl CrossRef PubMed 3. ↵ Vogt A , Goldman AD , Mochizuki K , Landweber LF . Transposon Domestication versus Mutualism in Ciliate Genome Rearrangements . PLoS Genet . 2013 Aug 1; 9 ( 8 ): e1003659 . OpenUrl CrossRef PubMed 4. ↵ Prescott DM . The DNA of ciliated protozoa . Microbiol Mol Biol Rev . 1994 June 1; 58 ( 2 ): 233 – 67 . OpenUrl Abstract / FREE Full Text 5. Betermier M , Duharcourt S . Programmed Rearrangement in Ciliates: Paramecium . Microbiol Spectr . 2014 Dec; 2 ( 6 ). 6. ↵ Bétermier M , Klobutcher LA , Orias E. Programmed chromosome fragmentation in ciliated protozoa: multiple means to chromosome ends . Lodoen MB , editor. Microbiol Mol Biol Rev . 2023 Nov 27;e00184-22. 7. ↵ Aury JM , Jaillon O , Duret L , Noel B , Jubin C , Porcel BM , et al. Global trends of whole-genome duplications revealed by the ciliate Paramecium tetraurelia . Nature . 2006 Nov 9; 444 (7116):171–8. 8. Eisen JA , Coyne RS , Wu M , Wu D , Thiagarajan M , Wortman JR , et al. Macronuclear genome sequence of the ciliate Tetrahymena thermophila, a model eukaryote . PLoS Biol . 2006 Sept; 4 ( 9 ): e286 . OpenUrl CrossRef PubMed 9. ↵ Swart EC , Bracht JR , Magrini V , Minx P , Chen X , Zhou Y , et al. The Oxytricha trifallax macronuclear genome: a complex eukaryotic genome with 16,000 tiny chromosomes . PLoS Biol . 2013 ; 11 ( 1 ): e1001473 . OpenUrl CrossRef PubMed 10. ↵ McGrath CL , Gout JF , Johri P , Doak TG , Lynch M . Differential retention and divergent resolution of duplicate genes following whole-genome duplication . Genome Res . 2014 Oct; 24 ( 10 ): 1665 – 75 . OpenUrl Abstract / FREE Full Text 11. ↵ Gout JF , Hao Y , Johri P , Arnaiz O , Doak TG , Bhullar S , et al. Dynamics of gene loss following ancient whole-genome duplication in the cryptic Paramecium complex . Rogers R , editor. Mol Biol Evol . 2023 May 8;msad107. 12. ↵ Sonneborn T . The Paramecium aurelia complex of 14 sibling species . Trans Am Microsc Soc . 1975 ; 94 : 155 – 78 . OpenUrl 13. Sawka-Gądek N , Potekhin A , Singh DP , Grevtseva I , Arnaiz O , Penel S , et al. Evolutionary Plasticity of Mating-Type Determination Mechanisms in Paramecium aurelia Sibling Species . Genome Biol Evol . 2021 Feb 1; 13 ( 2 ):evaa258. 14. ↵ Sellis D , Guérin F , Arnaiz O , Pett W , Lerat E , Boggetto N , et al. Massive colonization of protein- coding exons by selfish genetic elements in Paramecium germline genomes . Malik HS , editor. PLOS Biol . 2021 July 29;19(7):e3001309. 15. ↵ Sonneborn TM. Paramecium aurelia . In: Handbook of Genetics . R. King . New York : Plenum Press ; 1974 . p. 469 – 594 . 16. ↵ Guérin F , Arnaiz O , Boggetto N , Denby Wilkes C , Meyer E , Sperling L , et al. Flow cytometry sorting of nuclei enables the first global characterization of Paramecium germline DNA and transposable elements . BMC Genomics . 2017 Apr 26; 18 ( 1 ): 327 . OpenUrl CrossRef PubMed 17. ↵ Chen X , Bracht JR , Goldman AD , Dolzhenko E , Clay DM , Swart EC , et al. The architecture of a scrambled genome reveals massive levels of genomic rearrangement during development . Cell . 2014 Aug 28; 158 ( 5 ): 1187 – 98 . OpenUrl CrossRef PubMed 18. ↵ Hamilton EP , Kapusta A , Huvos PE , Bidwell SL , Zafar N , Tang H , et al. Structure of the germline genome of Tetrahymena thermophila and relationship to the massively rearranged somatic genome . eLife . 2016 Nov 28; 5 . 19. Seah BKB , Singh M , Emmerich C , Singh A , Woehle C , Huettel B , et al. MITE infestation accommodated by genome editing in the germline genome of the ciliate Blepharisma . Proc Natl Acad Sci . 2023 Jan 24; 120 ( 4 ): e2213985120 . OpenUrl CrossRef PubMed 20. ↵ Villano DJ , Prahlad M , Singhal A , Sanbonmatsu KY , Landweber LF. Widespread 3D genome reorganization precedes programmed DNA rearrangement in Oxytricha trifallax [Internet] . 2025 [cited 2025 Jan 9]. Available from: http://biorxiv.org/lookup/doi/10.1101/2024.12.31.630814 21. ↵ Le Mouël A , Butler A , Caron F , Meyer E . Developmentally regulated chromosome fragmentation linked to imprecise elimination of repeated sequences in paramecia . Eukaryot Cell . 2003 Oct; 2 ( 5 ): 1076 – 90 . OpenUrl Abstract / FREE Full Text 22. ↵ Duret L , Cohen J , Jubin C , Dessen P , Goût JF , Mousset S , et al. Analysis of sequence variability in the macronuclear DNA of Paramecium tetraurelia: a somatic view of the germline . Genome Res . 2008 Apr; 18 ( 4 ): 585 – 96 . OpenUrl Abstract / FREE Full Text 23. ↵ Cosby RL , Chang NC , Feschotte C . Host–transposon interactions: conflict, cooperation, and cooption . Genes Dev . 2019 Sept 1; 33 ( 17–18 ): 1098 – 116 . OpenUrl Abstract / FREE Full Text 24. ↵ Balan T , Lerner LK , Holoch D , Duharcourt S . Small-RNA-guided histone modifications and somatic genome elimination in ciliates . WIREs RNA . 2024 Mar; 15 ( 2 ): e1848 . OpenUrl CrossRef PubMed 25. ↵ Arnaiz O , Mathy N , Baudry C , Malinsky S , Aury JM , Wilkes CD , et al. The Paramecium germline genome provides a niche for intragenic parasitic DNA: evolutionary dynamics of internal eliminated sequences . PLoS Genet . 2012 ; 8 ( 10 ): e1002984 . OpenUrl CrossRef PubMed 26. ↵ Zangarelli C , Arnaiz O , Bourge M , Gorrichon K , Jaszczyszyn Y , Mathy N , et al. Developmental timing of programmed DNA elimination in Paramecium tetraurelia recapitulates germline transposon evolutionary dynamics . Genome Res . 2022 Nov; 32 ( 11–12 ): 2028 – 42 . OpenUrl Abstract / FREE Full Text 27. ↵ Baudry C , Malinsky S , Restituito M , Kapusta A , Rosa S , Meyer E , et al. PiggyMac, a domesticated piggyBac transposase involved in programmed genome rearrangements in the ciliate Paramecium tetraurelia . Genes Dev . 2009 Nov 1; 23 ( 21 ): 2478 – 83 . OpenUrl Abstract / FREE Full Text 28. ↵ Bischerour J , Bhullar S , Denby Wilkes C , Régnier V , Mathy N , Dubois E , et al. Six domesticated PiggyBac transposases together carry out programmed DNA elimination in Paramecium . eLife . 2018 18 ; 7 . 29. ↵ Kapusta A , Matsuda A , Marmignon A , Ku M , Silve A , Meyer E , et al. Highly precise and developmentally programmed genome assembly in Paramecium requires ligase IV-dependent end joining . PLoS Genet . 2011 Apr; 7 ( 4 ): e1002049 . OpenUrl CrossRef PubMed 30. Marmignon A , Bischerour J , Silve A , Fojcik C , Dubois E , Arnaiz O , et al. Ku-mediated coupling of DNA cleavage and repair during programmed genome rearrangements in the ciliate Paramecium tetraurelia . PLoS Genet . 2014 Aug; 10 ( 8 ): e1004552 . OpenUrl CrossRef PubMed 31. Abello A , Régnier V , Arnaiz O , Le Bars R , Bétermier M , Bischerour J. Functional diversification of Paramecium Ku80 paralogs safeguards genome integrity during precise programmed DNA elimination . Lichten M , editor. PLOS Genet . 2020 Apr 16;16(4):e1008723. 32. Bischerour J , Arnaiz O , Zangarelli C , Régnier V , Iehl F , Ropars V , et al. Uncoupling programmed DNA cleavage and repair scrambles the Paramecium somatic genome . Cell Rep . 2024 Apr; 43 ( 4 ): 114001 . OpenUrl CrossRef PubMed 33. ↵ Verron B , Arnaiz O , Zangarelli C , Mathy N , Bétermier M , Bischerour J . The linker region of a development-specific DNA polymerase X ensures efficient repair of programmed DNA double- strand breaks in Paramecium tetraurelia . Nucleic Acids Res . 2025 Apr 10; 53 ( 7 ):gkaf286. 34. ↵ Lhuillier-Akakpo M , Frapporti A , Denby Wilkes C , Matelot M , Vervoort M , Sperling L , et al. Local effect of enhancer of zeste-like reveals cooperation of epigenetic and cis-acting determinants for zygotic genome rearrangements . PLoS Genet . 2014 Sept; 10 ( 9 ): e1004665 . OpenUrl CrossRef PubMed 35. ↵ Sandoval PY , Swart EC , Arambasic M , Nowacki M . Functional Diversification of Dicer-like Proteins and Small RNAs Required for Genome Sculpting . Dev Cell . 2014 Jan 27; 28 ( 2 ): 174 – 88 . OpenUrl CrossRef PubMed Web of Science 36. ↵ Maliszewska-Olejniczak K , Gruchota J , Gromadka R , Denby Wilkes C , Arnaiz O , Mathy N , et al. TFIIS-Dependent Non-coding Transcription Regulates Developmental Genome Rearrangements . PLoS Genet . 2015 July; 11 ( 7 ): e1005383 . OpenUrl CrossRef PubMed 37. ↵ Gruchota J , Denby Wilkes C , Arnaiz O , Sperling L , Nowak JK . A meiosis-specific Spt5 homolog involved in non-coding transcription . Nucleic Acids Res . 2017 May 5; 45 ( 8 ): 4722 – 32 . OpenUrl PubMed 38. ↵ Owsian D , Gruchota J , Arnaiz O , Nowak JK . The transient Spt4-Spt5 complex as an upstream regulator of non-coding RNAs during development . Nucleic Acids Res . 2022 Mar 21; 50 ( 5 ): 2603 – 20 . OpenUrl PubMed 39. ↵ Lepère G , Nowacki M , Serrano V , Gout JF , Guglielmi G , Duharcourt S , et al. Silencing-associated and meiosis-specific small RNA pathways in Paramecium tetraurelia . Nucleic Acids Res . 2009 Feb; 37 ( 3 ): 903 – 15 . OpenUrl CrossRef PubMed Web of Science 40. ↵ Singh DP , Saudemont B , Guglielmi G , Arnaiz O , Goût JF , Prajer M , et al. Genome-defence small RNAs exapted for epigenetic mating-type inheritance . Nature . 2014 May 22; 509 (7501):447–52. 41. ↵ Furrer DI , Swart EC , Kraft MF , Sandoval PY , Nowacki M . Two Sets of Piwi Proteins Are Involved in Distinct sRNA Pathways Leading to Elimination of Germline-Specific DNA . Cell Rep . 2017 July; 20 ( 2 ): 505 – 20 . OpenUrl CrossRef PubMed 42. ↵ Miró-Pina C , Charmant O , Giovannetti M , De Vanssay A , Frapporti A , Humbert A , et al. A histone methyltransferase-independent function of PRC2 controls small RNA dynamics during programmed DNA elimination in Paramecium [Internet] . Genomics ; 2023 [cited 2025 Sept 26]. Available from: http://biorxiv.org/lookup/doi/10.1101/2023.07.04.547679 43. ↵ Charmant O , Gruchota J , Arnaiz O , Nowak KP , Moisan N , Zangarelli C , et al. The PIWI- interacting protein Gtsf1 controls the selective degradation of small RNAs in Paramecium . Nucleic Acids Res . 2024 Nov 22;gkae1055. 44. ↵ Wang C , Lyv L , Solberg T , Zhang H , Wen Z , Gao F . GTSF1 is required for transposon silencing in the unicellular eukaryote Paramecium tetraurelia . Nucleic Acids Res . 2024 Nov 27; 52 ( 21 ): 13206 – 23 . OpenUrl PubMed 45. ↵ Frapporti A , Miró Pina C , Arnaiz O , Holoch D , Kawaguchi T , Humbert A , et al. The Polycomb protein Ezl1 mediates H3K9 and H3K27 methylation to repress transposable elements in Paramecium . Nat Commun . 2019 June 20; 10 ( 1 ): 2710 . OpenUrl CrossRef PubMed 46. ↵ Miró-Pina C , Charmant O , Kawaguchi T , Holoch D , Michaud A , Cohen I , et al. Paramecium Polycomb repressive complex 2 physically interacts with the small RNA-binding PIWI protein to repress transposable elements . Dev Cell . 2022 Apr; 57 ( 8 ): 1037 – 1052 .e8. OpenUrl CrossRef PubMed 47. ↵ Wang C , Solberg T , Maurer-Alcalá XX , Swart EC , Gao F , Nowacki M . A small RNA-guided PRC2 complex eliminates DNA as an extreme form of transposon silencing . Cell Rep . 2022 Aug; 40 ( 8 ): 111263 . OpenUrl CrossRef PubMed 48. ↵ de Vanssay A , Touzeau A , Arnaiz O , Frapporti A , Phipps J , Duharcourt S. The Paramecium histone chaperone Spt16-1 is required for Pgm endonuclease function in programmed genome rearrangements . Malik HS , editor. PLOS Genet . 2020 July 23;16(7):e1008949. 49. ↵ Singh A , Maurer-Alcalá XX , Solberg T , Häußermann L , Gisler S , Ignarski M , et al. Chromatin remodeling is required for sRNA-guided DNA elimination in Paramecium . EMBO J . 2022 Nov 17; 41 ( 22 ): e111839 . OpenUrl CrossRef PubMed 50. ↵ Singh A , Häußermann L , Emmerich C , Nischwitz E , Seah BK , Butter F , et al. ISWI1 complex proteins facilitate developmental genome editing in Paramecium . Genome Res . 2024 Nov 14;gr.278402.123. 51. ↵ Preer Jr JR . Quantitative predictions of random segregation models of the ciliate macronucleus . Genet Res . 1976 ; 27 ( 2 ): 227 – 38 . OpenUrl CrossRef PubMed 52. ↵ Li Z , Gilbert C , Peng H , Pollet N . Discovery of numerous novel Helitron-like elements in eukaryote genomes using HELIANO . Nucleic Acids Res . 2024 Sept 23; 52 ( 17 ): e79 – e79 . OpenUrl CrossRef PubMed 53. ↵ Arnaiz O , Van Dijk E , Bétermier M , Lhuillier-Akakpo M , de Vanssay A , Duharcourt S , et al. Improved methods and resources for paramecium genomics: transcription units, gene annotation and gene expression . BMC Genomics . 2017 June 26; 18 ( 1 ): 483 . OpenUrl CrossRef PubMed 54. ↵ Mapleson D , Garcia Accinelli G , Kettleborough G , Wright J , Clavijo BJ. KAT: a K-mer analysis toolkit to quality control NGS datasets and genome assemblies . Berger B , editor. Bioinformatics . 2017 Feb 15;33(4):574–6. 55. ↵ Bénitière F , Arnaiz O , Duharcourt S , Meyer E , Sperling L , Duret L. Inter-specific promiscuity: recurrent genomic introgressions from very distant lineages in Paramecium sonneborni . Prep. 56. ↵ Jaillon O , Bouhouche K , Gout JF , Aury JM , Noel B , Saudemont B , et al. Translational control of intron splicing in eukaryotes . Nature . 2008 Jan 17; 451 (7176):359–62. 57. ↵ Saudemont B , Popa A , Parmley JL , Rocher V , Blugeon C , Necsulea A , et al. The fitness cost of mis-splicing is the main determinant of alternative splicing patterns . Genome Biol . 2017 Oct 30; 18 ( 1 ): 208 . OpenUrl CrossRef PubMed 58. ↵ Richard GF , Kerrest A , Dujon B . Comparative Genomics and Molecular Dynamics of DNA Repeats in Eukaryotes . Microbiol Mol Biol Rev . 2008 Dec 1; 72 ( 4 ): 686 – 727 . OpenUrl Abstract / FREE Full Text 59. Silva BSML , Picorelli ACR , Kuhn GCS. In Silico Identification and Characterization of Satellite DNAs in 23 Drosophila Species from the Montium Group . Genes . 2023 Jan 23;14(2):300. 60. Thakur J , Packiaraj J , Henikoff S. Sequence , Chromatin and Evolution of Satellite DNA . Int J Mol Sci . 2021 Apr 21; 22 ( 9 ): 4309 . OpenUrl CrossRef PubMed 61. ↵ Veseljak D , Despot-Slade E , Volarić M , Horvat L , Vojvoda Zeljko T , Meštrović N , et al. Dynamic evolution of satellite DNAs drastically differentiates the genomes of Tribolium sibling species . Genome Res . 2025 Sept 16;gr.280516.125. 62. ↵ Forney J , Rodkey K . A repetitive DNA sequence in Paramecium macronuclei is related to the β subunit of G proteins . Nucleic Acids Res . 1992 ; 20 ( 20 ): 5397 – 402 . OpenUrl CrossRef PubMed Web of Science 63. ↵ Chandler M , De La Cruz F , Dyda F , Hickman AB , Moncalian G , Ton-Hoang B . Breaking and joining single-stranded DNA: the HUH endonuclease superfamily . Nat Rev Microbiol . 2013 Aug; 11 ( 8 ): 525 – 38 . OpenUrl CrossRef PubMed 64. ↵ Kosek D , Grabundzija I , Lei H , Bilic I , Wang H , Jin Y , et al. The large bat Helitron DNA transposase forms a compact monomeric assembly that buries and protects its covalently bound 5ʹ-transposon end . Mol Cell . 2021 Oct; 81 ( 20 ): 4271 – 4286 .e4. OpenUrl CrossRef PubMed 65. ↵ Seah BKB , Singh A , Vetter DE , Emmerich C , Peters M , Soltys V , et al. Nuclear dualism without extensive DNA elimination in the ciliate Loxodes magnus . Proc Natl Acad Sci . 2024 Sept 24; 121 ( 39 ): e2400503121 . OpenUrl PubMed 66. ↵ Heringer P , Kuhn GCS . Exploring the Remote Ties between Helitron Transposases and Other Rolling-Circle Replication Proteins . Int J Mol Sci . 2018 Oct 9; 19 ( 10 ): 3079 . OpenUrl PubMed 67. ↵ Heringer P , Kuhn GCS. Pif1 Helicases and the Evidence for a Prokaryotic Origin of Helitrons . Arkhipova I , editor. Mol Biol Evol . 2022 Jan 7;39(1):msab334. 68. ↵ Maumus F , Quesneville H. Deep Investigation of Arabidopsis thaliana Junk DNA Reveals a Continuum between Repetitive Elements and Genomic Dark Matter . Jordan IK , editor. PLoS ONE . 2014 Apr 7;9(4):e94101. 69. ↵ Dippell R . A preliminary report on the chromosomal constitution of certain variety 4 races of Paramecium aurelia . Suppl Caryologia . 1954 ; 6 : 1109 – 11 . OpenUrl 70. ↵ Jones K. Nuclear differentiation in Paramecium [Ph. D. thesis]. [Aberystwyth, University of Wales] ; 1956 . 71. ↵ Hillers KJ . Crossover interference . Curr Biol . 2004 Dec; 14 ( 24 ): R1036 – 7 . OpenUrl CrossRef PubMed Web of Science 72. ↵ Anderson CM , Oke A , Yam P , Zhuge T , Fung JC. Reduced Crossover Interference and Increased ZMM-Independent Recombination in the Absence of Tel1/ATM . Lichten M , editor. PLOS Genet . 2015 Aug 25;11(8):e1005478. 73. ↵ Gruhn JR , Zielinska AP , Shukla V , Blanshard R , Capalbo A , Cimadomo D , et al. Chromosome errors in human eggs shape natural fertility over reproductive life span . Science . 2019 Sept 27; 365 (6460):1466–9. 74. ↵ Baroin A , Prat A , Caron F . Telomeric site position heterogeneity in macronuclear DNA of Paramecium primaurelia . Nucleic Acids Res . 1987 Feb 25; 15 ( 4 ): 1717 – 28 . OpenUrl CrossRef PubMed Web of Science 75. ↵ McCormick-Graham M , Romero DP . A single telomerase RNA is sufficient for the synthesis of variable telomeric DNA repeats in ciliates of the genus Paramecium . Mol Cell Biol . 1996 Apr; 16 ( 4 ): 1871 – 9 . OpenUrl Abstract / FREE Full Text 76. ↵ McCormick-Graham M , Haynes WJ , Romero DP . Variable telomeric repeat synthesis in Paramecium tetraurelia is consistent with misincorporation by telomerase . EMBO J . 1997 June 2; 16 ( 11 ): 3233 – 42 . OpenUrl Abstract / FREE Full Text 77. ↵ Swart EC , Denby Wilkes C , Sandoval PY , Hoehener C , Singh A , Furrer DI , et al. Identification and analysis of functional associations among natural eukaryotic genome editing components . F1000Research . 2017 Aug 8;6:1374. 78. ↵ Denby Wilkes C , Arnaiz O , Sperling L . ParTIES: a toolbox for Paramecium interspersed DNA elimination studies . Bioinformatics . 2016 Feb 15; 32 ( 4 ): 599 – 601 . OpenUrl CrossRef PubMed 79. ↵ Grabundzija I , Hickman AB , Dyda F . Helraiser intermediates provide insight into the mechanism of eukaryotic replicative transposition . Nat Commun . 2018 Mar 29; 9 ( 1 ): 1278 . OpenUrl CrossRef PubMed 80. ↵ Barro-Trastoy D , Köhler C . Helitrons: genomic parasites that generate developmental novelties . Trends Genet . 2024 May; 40 ( 5 ): 437 – 48 . OpenUrl CrossRef PubMed 81. ↵ Beisson J , Bétermier M , Bré MH , Cohen J , Duharcourt S , Duret L , et al. Paramecium tetraurelia: the renaissance of an early unicellular model . CSH Protoc . 2010 Jan;2010(1):pdb.emo140. 82. ↵ Fernandes JB , Séguéla-Arnaud M , Larchevêque C , Lloyd AH , Mercier R . Unleashing meiotic crossovers in hybrid plants . Proc Natl Acad Sci . 2018 Mar 6; 115 ( 10 ): 2431 – 6 . OpenUrl Abstract / FREE Full Text 83. Fledel-Alon A , Wilson DJ , Broman K , Wen X , Ober C , Coop G , et al. Broad-Scale Recombination Patterns Underlying Proper Disjunction in Humans . Copenhaver GP , editor. PLoS Genet . 2009 Sept 18;5(9):e1000658. 84. Hunter N . Meiotic Recombination: The Essence of Heredity . Cold Spring Harb Perspect Biol . 2015 Oct 28; a016618 . 85. ↵ Stapley J , Feulner PGD , Johnston SE , Santure AW , Smadja CM . Variation in recombination frequency and distribution across eukaryotes: patterns and processes . Philos Trans R Soc B Biol Sci . 2017 Dec 19; 372 (1736):20160455. 86. ↵ Arnaiz O , Goût JF , Bétermier M , Bouhouche K , Cohen J , Duret L , et al. Gene expression in a paleopolyploid: a transcriptome resource for the ciliate Paramecium tetraurelia . BMC Genomics . 2010 ; 11 : 547 . 87. ↵ Carta A , Bedini G , Peruzzi L . A deep dive into the ancestral chromosome number and genome size of flowering plants . New Phytol . 2020 Nov; 228 ( 3 ): 1097 – 106 . OpenUrl CrossRef PubMed 88. ↵ Damas J , Corbo M , Kim J , Turner-Maier J , Farré M , Larkin DM , et al. Evolution of the ancestral mammalian karyotype and syntenic regions . Proc Natl Acad Sci . 2022 Oct 4; 119 ( 40 ): e2209139119 . OpenUrl CrossRef PubMed 89. ↵ Huang Z , Xu Z , Bai H , Huang Y , Kang N , Ding X , et al. Evolutionary analysis of a complete chicken genome . Proc Natl Acad Sci . 2023 Feb 21; 120 ( 8 ): e2216641120 . OpenUrl CrossRef PubMed 90. ↵ Parey E , Louis A , Montfort J , Guiguen Y , Crollius HR , Berthelot C . An atlas of fish genome evolution reveals delayed rediploidization following the teleost whole-genome duplication . Genome Res . 2022 Sept; 32 ( 9 ): 1685 – 97 . OpenUrl Abstract / FREE Full Text 91. ↵ Dujon B . Yeasts illustrate the molecular mechanisms of eukaryotic genome evolution . Trends Genet . 2006 July; 22 ( 7 ): 375 – 87 . OpenUrl CrossRef PubMed Web of Science 92. ↵ Gordon JL , Byrne KP , Wolfe KH . Mechanisms of Chromosome Number Evolution in Yeast . PLoS Genet . 2011 July 21; 7 ( 7 ): e1002190 . OpenUrl CrossRef PubMed 93. ↵ Wang X , Jin D , Wang Z , Guo H , Zhang L , Wang L , et al. Telomere-centric genome repatterning determines recurring chromosome number reductions during the evolution of eukaryotes . New Phytol . 2015 Jan; 205 ( 1 ): 378 – 89 . OpenUrl CrossRef PubMed 94. ↵ Clark J , Hidalgo O , Pellicer J , Liu H , Marquardt J , Robert Y , et al. Genome evolution of ferns: evidence for relative stasis of genome size across the fern phylogeny . New Phytol . 2016 May; 210 ( 3 ): 1072 – 82 . OpenUrl CrossRef PubMed 95. ↵ Rogers DW , McConnell E , Ono J , Greig D. Spore-autonomous fluorescent protein expression identifies meiotic chromosome mis-segregation as the principal cause of hybrid sterility in yeast . Heitman J , editor. PLOS Biol . 2018 Nov 12;16(11):e2005066. 96. ↵ Kapitonov VV , Jurka J . Rolling-circle transposons in eukaryotes . Proc Natl Acad Sci . 2001 July 17; 98 ( 15 ): 8714 – 9 . OpenUrl Abstract / FREE Full Text 97. ↵ Thomas J , Pritham EJ. Helitrons, the Eukaryotic Rolling-circle Transposable Elements . Chandler M , Craig N , editors. Microbiol Spectr . 2015 July 2;3(4):3.4.03. 98. ↵ Poulter RTM , Goodwin TJD , Butler MI . Vertebrate helentrons and other novel Helitrons . Gene . 2003 Aug 14; 313 : 201 – 12 . OpenUrl CrossRef PubMed Web of Science 99. Bao W , Jurka J . Homologues of bacterial TnpB_IS605 are widespread in diverse eukaryotic transposable elements . Mob DNA . 2013 ; 4 ( 1 ): 12 . OpenUrl CrossRef PubMed 100. ↵ Thomas J , Vadnagara K , Pritham EJ . DINE-1, the highest copy number repeats in Drosophila melanogaster are non-autonomous endonuclease-encoding rolling-circle transposable elements (Helentrons) . Mob DNA . 2014 June 4; 5 ( 1 ): 18 . OpenUrl CrossRef PubMed 101. ↵ Chellapan BV , van Dam P , Rep M , Cornelissen BJC , Fokkens L . Non-canonical Helitrons in Fusarium oxysporum . Mob DNA . 2016 Dec; 7 ( 1 ): 1 – 16 . OpenUrl PubMed 102. ↵ Castanera R , Pérez G , López L , Sancho R , Santoyo F , Alfaro M , et al. Highly expressed captured genes and cross-kingdom domains present in Helitrons create novel diversity in Pleurotus ostreatus and other fungi . BMC Genomics . 2014 Dec; 15 ( 1 ): 1 – 17 . OpenUrl CrossRef PubMed 103. Li Z , Pollet N . Impact of a horizontally transferred Helitron family on genome evolution in Xenopus laevis . Mob DNA . 2025 Apr 16; 16 ( 1 ): 19 . OpenUrl PubMed 104. ↵ Paulat NS , Storer JM , Moreno-Santillán DD , Osmanski AB , Sullivan KAM , Grimshaw JR , et al. Chiropterans Are a Hotspot for Horizontal Transfer of DNA Transposons in Mammalia . Mol Biol Evol . 2023 Apr 18; 40 ( 5 ):msad092. 105. ↵ Johri P , Gout JF , Doak TG , Lynch M. A Population-Genetic Lens into the Process of Gene Loss Following Whole-Genome Duplication . Wittkopp P , editor. Mol Biol Evol [Internet] . 2022 June 2 [cited 2025 July 15];39(6). Available from: https://academic.oup.com/mbe/article/doi/10.1093/molbev/msac118/6594306 106. ↵ Herrick G , Cartinhour S , Dawson D , Ang D , Sheets R , Lee A , et al. Mobile elements bounded by C4A4 telomeric repeats in Oxytricha fallax . Cell . 1985 Dec; 43 ( 3 Pt 2):759–68. 107. ↵ Williams K , Doak TG , Herrick G . Developmental precise excision of Oxytricha trifallax telomere- bearing elements and formation of circles closed by a copy of the flanking target duplication . EMBO J . 1993 Dec; 12 ( 12 ): 4593 – 601 . OpenUrl PubMed 108. ↵ Cherry JM , Blackburn EH . The internally located telomeric sequences in the germ-line chromosomes of tetrahymena are at the ends of transposon-like elements . Cell . 1985 Dec; 43 ( 3 ): 747 – 58 . OpenUrl CrossRef PubMed Web of Science 109. ↵ Yang L , Bennetzen JL . Structure-based discovery and description of plant and animal Helitrons . Proc Natl Acad Sci . 2009 Aug 4; 106 ( 31 ): 12832 – 7 . OpenUrl Abstract / FREE Full Text 110. ↵ Grabundzija I , Messing SA , Thomas J , Cosby RL , Bilic I , Miskey C , et al. A Helitron transposon reconstructed from bats reveals a novel mechanism of genome shuffling in eukaryotes . Nat Commun . 2016 Mar 2; 7 : 10716 . 111. ↵ Subirana JA , Albà MM , Messeguer X . High evolutionary turnover of satellite families in Caenorhabditis . BMC Evol Biol . 2015 Dec; 15 ( 1 ): 1 – 13 . OpenUrl CrossRef PubMed 112. Vojvoda Zeljko T , Pavlek M , Meštrović N , Plohl M . Satellite DNA-like repeats are dispersed throughout the genome of the Pacific oyster Crassostrea gigas carried by Helentron non- autonomous mobile elements . Sci Rep . 2020 Sept 15; 10 ( 1 ): 1 – 12 . OpenUrl CrossRef PubMed 113. ↵ Hofstatter PG , Thangavel G , Lux T , Neumann P , Vondrak T , Novak P , et al. Repeat-based holocentromeres influence genome architecture and karyotype evolution . Cell . 2022 Aug; 185 ( 17 ): 3153 – 3168 .e18. OpenUrl CrossRef PubMed 114. Saint-Leandre B , Nguyen SC , Levine MT . Diversification and collapse of a telomere elongation mechanism . Genome Res . 2019 June 1; 29 ( 6 ): 920 – 31 . OpenUrl Abstract / FREE Full Text 115. ↵ Da Ines O , White CI . Centromere Associations in Meiotic Chromosome Pairing . Annu Rev Genet . 2015 Nov 23; 49 ( 1 ): 95 – 114 . OpenUrl CrossRef PubMed 116. ↵ Zickler D , Kleckner N . Meiosis: Dances Between Homologs . Annu Rev Genet . 2023 Nov 27; 57 ( 1 ): 1 – 63 . OpenUrl CrossRef PubMed 117. ↵ MacQueen AJ , Phillips CM , Bhalla N , Weiser P , Villeneuve AM , Dernburg AF . Chromosome Sites Play Dual Roles to Establish Homologous Synapsis during Meiosis in C. elegans . Cell . 2005 Dec; 123 ( 6 ): 1037 – 50 . OpenUrl CrossRef PubMed Web of Science 118. ↵ Tsai JH , McKee BD . Homologous pairing and the role of pairing centers in meiosis . J Cell Sci . 2011 June 15; 124 ( 12 ): 1955 – 63 . OpenUrl Abstract / FREE Full Text 119. ↵ Preer LB , Rudman B , Pollack S , Preer JR . Does Ribosomal DNA Get Out of the Micronuclear Chromosome in Paramecium tetraurelia by Means of a Rolling Circle? Mol Cell Biol . 1999 Nov 1; 19 ( 11 ): 7792 – 800 . OpenUrl Abstract / FREE Full Text 120. ↵ Carle CM , Zaher HS , Chalker DL . A Parallel G Quadruplex-Binding Protein Regulates the Boundaries of DNA Elimination Events of Tetrahymena thermophila . PLOS Genet . 2016 Mar 7; 12 ( 3 ): e1005842 . OpenUrl CrossRef PubMed 121. ↵ Qiao Q , Wang L , Meng FL , Hwang JK , Alt FW , Wu H . AID Recognizes Structured DNA for Class Switch Recombination . Mol Cell . 2017 Aug; 67 ( 3 ): 361 – 373 .e4. OpenUrl CrossRef PubMed 122. ↵ Cheung I , Schertzer M , Rose A , Lansdorp PM . Disruption of dog-1 in Caenorhabditis elegans triggers deletions upstream of guanine-rich DNA . Nat Genet . 2002 Aug; 31 ( 4 ): 405 – 9 . OpenUrl CrossRef PubMed Web of Science 123. ↵ London TBC , Barber LJ , Mosedale G , Kelly GP , Balasubramanian S , Hickson ID , et al. FANCJ Is a Structure-specific DNA Helicase Associated with the Maintenance of Genomic G/C Tracts . J Biol Chem . 2008 Dec; 283 ( 52 ): 36132 – 9 . OpenUrl Abstract / FREE Full Text 124. ↵ Sato K , Knipscheer P . G-quadruplex resolution: From molecular mechanisms to physiological relevance . DNA Repair . 2023 Oct; 130 : 103552 . 125. ↵ Durkin SG , Glover TW . Chromosome Fragile Sites . Annu Rev Genet . 2007 Dec 1; 41 ( 1 ): 169 – 92 . OpenUrl CrossRef PubMed Web of Science 126. ↵ Nekrasova I , Nikitashina V , Bhullar S , Arnaiz O , Singh DP , Meyer E , et al. Loss of a Fragile Chromosome Region leads to the Screwy Phenotype in Paramecium tetraurelia . Genes . 2019 July 6; 10 ( 7 ): 513 . OpenUrl 127. ↵ Wang J , Gao S , Mostovoy Y , Kang Y , Zagoskin M , Sun Y , et al. Comparative genome analysis of programmed DNA elimination in nematodes . Genome Res . 2017 Dec; 27 ( 12 ): 2001 – 14 . OpenUrl Abstract / FREE Full Text 128. ↵ Gonzalez De La Rosa PM , Thomson M , Trivedi U , Tracey A , Tandonnet S , Blaxter M. A telomere- to-telomere assembly of Oscheius tipulae and the evolution of rhabditid nematode chromosomes . Rockman M , editor. G3 GenesGenomesGenetics . 2021 Mar 10;11(1):jkaa020. 129. ↵ Dockendorff TC , Estrem B , Reed J , Simmons JR , Zadegan SB , Zagoskin MV , et al. The nematode Oscheius tipulae as a genetic model for programmed DNA elimination . Curr Biol . 2022 Dec; 32 ( 23 ): 5083 – 5098 .e6. OpenUrl CrossRef PubMed 130. Estrem B , Wang J. Programmed DNA elimination in the parasitic nematode Ascaris . Kafsack BFC , editor. PLOS Pathog . 2023 Feb 2;19(2):e1011087. 131. ↵ Rey C , Launay C , Wenger E , Delattre M . Programmed DNA elimination in Mesorhabditis nematodes . Curr Biol . 2023 Sept; 33 ( 17 ): 3711 – 3721 .e5. OpenUrl CrossRef PubMed 132. ↵ Müller F , Wicky C , Spicher A , Tobler H . New telomere formation after developmentally regulated chromosomal breakage during the process of chromatin diminution in ascaris lumbricoides . Cell . 1991 Nov; 67 ( 4 ): 815 – 22 . OpenUrl CrossRef PubMed Web of Science 133. Wang J , Veronezi GMB , Kang Y , Zagoskin M , O’Toole ET , Davis RE . Comprehensive Chromosome End Remodeling during Programmed DNA Elimination . Curr Biol . 2020 Sept; 30 ( 17 ): 3397 – 3413 .e4. OpenUrl PubMed 134. ↵ Estrem B , Davis RE , Wang J . End resection and telomere healing of DNA double-strand breaks during nematode programmed DNA elimination . Nucleic Acids Res . 2024 Aug 27; 52 ( 15 ): 8913 – 29 . OpenUrl CrossRef PubMed 135. ↵ Beisson J , Bétermier M , Bré MH , Cohen J , Duharcourt S , Duret L , et al. Maintaining clonal Paramecium tetraurelia cell lines of controlled age through daily reisolation . Cold Spring Harb Protoc . 2010 Jan;2010(1):pdb.prot5361. 136. ↵ Beisson J , Bétermier M , Bré MH , Cohen J , Duharcourt S , Duret L , et al. Mass culture of Paramecium tetraurelia . Cold Spring Harb Protoc . 2010 Jan;2010(1):pdb.prot5362. 137. ↵ Magoč T , Salzberg SL . FLASH: fast length adjustment of short reads to improve genome assemblies . Bioinforma Oxf Engl . 2011 Nov 1;27(21): 2957 – 63 . 138. ↵ Boetzer M , Henkel CV , Jansen HJ , Butler D , Pirovano W . Scaffolding pre-assembled contigs using SSPACE . Bioinformatics . 2011 Feb 15; 27 ( 4 ): 578 – 9 . OpenUrl CrossRef PubMed Web of Science 139. ↵ Luo R , Liu B , Xie Y , Li Z , Huang W , Yuan J , et al. SOAPdenovo2: an empirically improved memory-efficient short-read de novo assembler . GigaScience . 2012 Dec 27; 1 ( 1 ): 18 . OpenUrl CrossRef PubMed 140. ↵ Bazin-Gélis M , Eleftheriou E , Zangarelli C , Lelandais G , Sperling L , Arnaiz O , et al. Inter- generational nuclear crosstalk links the control of gene expression to programmed genome rearrangement during the Paramecium sexual cycle . Nucleic Acids Res . 2023 Dec 11; 51 ( 22 ): 12337 – 51 . OpenUrl CrossRef PubMed 141. ↵ Lepère G , Bétermier M , Meyer E , Duharcourt S . Maternal noncoding transcripts antagonize the targeting of DNA elimination by scanRNAs in Paramecium tetraurelia . Genes Dev . 2008 June 1; 22 ( 11 ): 1501 – 12 . OpenUrl Abstract / FREE Full Text 142. ↵ Liu H , Wu S , Li A , Ruan J . SMARTdenovo: a de novo assembler using long noisy reads . Gigabyte . 2021 Mar 8;2021: 1 – 9 . 143. ↵ Vaser R , Sović I , Nagarajan N , Šikić M . Fast and accurate de novo genome assembly from long uncorrected reads . Genome Res . 2017 May 1; 27 ( 5 ): 737 – 46 . OpenUrl Abstract / FREE Full Text 144. ↵ Belser C , Baurens FC , Noel B , Martin G , Cruaud C , Istace B , et al. Telomere-to-telomere gapless chromosomes of banana using nanopore sequencing . Commun Biol . 2021 Sept 7; 4 ( 1 ): 1 – 12 . OpenUrl PubMed 145. ↵ Walker BJ , Abeel T , Shea T , Priest M , Abouelliel A , Sakthikumar S , et al. Pilon: An Integrated Tool for Comprehensive Microbial Variant Detection and Genome Assembly Improvement . Wang J , editor. PLoS ONE . 2014 Nov 19;9(11):e112963. 146. ↵ Bouhouche K , Gout JF , Kapusta A , Bétermier M , Meyer E . Functional specialization of Piwi proteins in Paramecium tetraurelia from post-transcriptional gene silencing to genome remodelling . Nucleic Acids Res . 2011 ; 39 ( 10 ): 4249 – 64 . OpenUrl CrossRef PubMed Web of Science 147. ↵ Li H. Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM . Available: http://arxiv.org/abs/1303.3997 . arXiv:13033997. 2013; 148. ↵ Li H , Handsaker B , Wysoker A , Fennell T , Ruan J , Homer N , et al. The Sequence Alignment/Map format and SAMtools . Bioinforma Oxf Engl . 2009 Aug 15; 25 ( 16 ): 2078 – 9 . OpenUrl 149. ↵ Broad Institute . Picard Tools . [cited 17 Nov 2020]. Available: http://broadinstitute.github.io/picard/ . 150. ↵ McKenna A , Hanna M , Banks E , Sivachenko A , Cibulskis K , Kernytsky A , et al. The Genome Analysis Toolkit: A MapReduce framework for analyzing next-generation DNA sequencing data . Genome Res . 2010 Sept; 20 ( 9 ): 1297 – 303 . OpenUrl Abstract / FREE Full Text 151. ↵ Billings T , Sargent EE , Szatkiewicz JP , Leahy N , Kwak IY , Bektassova N , et al. Patterns of Recombination Activity on Mouse Chromosome 11 Revealed by High Resolution Mapping . Kirkpatrick DT , editor. PLoS ONE . 2010 Dec 8;5(12):e15340. 152. ↵ Marçais G , Kingsford C . A fast, lock-free approach for efficient parallel counting of occurrences of k-mers . Bioinformatics . 2011 Mar 15; 27 ( 6 ): 764 – 70 . OpenUrl CrossRef PubMed Web of Science 153. ↵ Fu L , Niu B , Zhu Z , Wu S , Li W . CD-HIT: accelerated for clustering the next-generation sequencing data . Bioinformatics . 2012 Dec 1; 28 ( 23 ): 3150 – 2 . OpenUrl CrossRef PubMed Web of Science 154. ↵ Katoh K , Standley DM . MAFFT Multiple Sequence Alignment Software Version 7: Improvements in Performance and Usability . Mol Biol Evol . 2013 Apr 1; 30 ( 4 ): 772 – 80 . OpenUrl CrossRef PubMed Web of Science 155. ↵ Langmead B , Salzberg SL . Fast gapped-read alignment with Bowtie 2 . Nat Methods . 2012 Mar 4; 9 ( 4 ): 357 – 9 . OpenUrl CrossRef PubMed Web of Science 156. ↵ Alonge M , Lebeigle L , Kirsche M , Jenike K , Ou S , Aganezov S , et al. Automated assembly scaffolding using RagTag elevates a new tomato system for high-throughput genome editing . Genome Biol . 2022 Dec 15; 23 ( 1 ): 258 . OpenUrl CrossRef PubMed 157. ↵ Krzywinski M , Schein J , Birol İ, Connors J, Gascoyne R, Horsman D, et al. Circos: An information aesthetic for comparative genomics . Genome Res . 2009 Sept 1; 19 ( 9 ): 1639 – 45 . OpenUrl Abstract / FREE Full Text 158. ↵ Marçais G , Delcher AL , Phillippy AM , Coston R , Salzberg SL , Zimin A . MUMmer4: A fast and versatile genome alignment system . PLOS Comput Biol . 2018 Jan 26; 14 ( 1 ): e1005944 . OpenUrl CrossRef PubMed 159. ↵ Kent WJ . BLAT--the BLAST-like alignment tool . Genome Res . 2002 Apr; 12 ( 4 ): 656 – 64 . OpenUrl Abstract / FREE Full Text 160. ↵ Flutre T , Duprat E , Feuillet C , Quesneville H . Considering Transposable Element Diversification in De Novo Annotation Approaches . PLoS ONE . 2011 Jan 31; 6 ( 1 ): e16526 . OpenUrl CrossRef PubMed 161. ↵ Bao W , Kojima KK , Kohany O . Repbase Update, a database of repetitive elements in eukaryotic genomes . Mob DNA . 2015 Dec; 6 ( 1 ): 11 . OpenUrl CrossRef PubMed 162. ↵ Min B , Choi IG. Practical Guide for Fungal Gene Prediction from Genome Assembly and RNA- Seq Reads by FunGAP . In: Kollmar M , editor. Gene Prediction [Internet] . New York , NY: Springer New York ; 2019 [cited 2024 Apr 11]. p. 53–64. (Methods in Molecular Biology; vol. 1962). Available from: http://link.springer.com/10.1007/978-1-4939-9173-0_4 163. ↵ Ranwez V , Harispe S , Delsuc F , Douzery EJP. MACSE: Multiple Alignment of Coding SEquences Accounting for Frameshifts and Stop Codons . Murphy WJ , editor. PLoS ONE . 2011 Sept 16;6(9):e22594. 164. ↵ Ranwez V , Chantret N , Delsuc F. Aligning Protein-Coding Nucleotide Sequences with MACSE . In: Katoh K , editor. Multiple Sequence Alignment [Internet] . New York , NY: Springer US ; 2021 [cited 2024 Mar 21]. p. 51–70. (Methods in Molecular Biology; vol. 2231). Available from: http://link.springer.com/10.1007/978-1-0716-1036-7_4 165. ↵ Gouy M , Guindon S , Gascuel O . SeaView version 4: A multiplatform graphical user interface for sequence alignment and phylogenetic tree building . Mol Biol Evol . 2010 Feb; 27 ( 2 ): 221 – 4 . OpenUrl CrossRef PubMed Web of Science 166. ↵ Guindon S , Dufayard JF , Lefort V , Anisimova M , Hordijk W , Gascuel O . New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of PhyML 3.0 . Syst Biol . 2010 May; 59 ( 3 ): 307 – 21 . OpenUrl CrossRef PubMed Web of Science 167. ↵ Yu G. Data Integration, Manipulation and Visualization of Phylogenetic Trees [Internet] . 1st ed. Boca Raton: Chapman and Hall/CRC; 2022 [cited 2024 Mar 21]. Available from: https://www.taylorfrancis.com/books/9781003279242 168. ↵ Yu G , Smith DK , Zhu H , Guan Y , Lam TT. GGTREE : an R package for visualization and annotation of phylogenetic trees with their covariates and other associated data . McInerny G , editor. Methods Ecol Evol . 2017 Jan;8(1):28–36. 169. ↵ Dutheil J , Boussau B . Non-homogeneous models of sequence evolution in the Bio++ suite of libraries and programs . BMC Evol Biol . 2008 ; 8 ( 1 ): 255 . OpenUrl CrossRef PubMed 170. Guéguen L , Gaillard S , Boussau B , Gouy M , Groussin M , Rochette NC , et al. Bio++: Efficient Extensible Libraries and Tools for Computational Molecular Evolution . Mol Biol Evol . 2013 Aug; 30 ( 8 ): 1745 – 50 . OpenUrl CrossRef PubMed Web of Science 171. ↵ Bolívar P , Guéguen L , Duret L , Ellegren H , Mugal CF . GC-biased gene conversion conceals the prediction of the nearly neutral theory in avian genomes . Genome Biol . 2019 Dec; 20 ( 1 ): 5 . OpenUrl CrossRef PubMed 172. ↵ Yang Z , Nielsen R . Synonymous and nonsynonymous rate variation in nuclear genes of mammals . J Mol Evol . 1998 Apr; 46 ( 4 ): 409 – 18 . OpenUrl CrossRef PubMed Web of Science 173. ↵ Guéguen L , Duret L . Unbiased Estimate of Synonymous and Nonsynonymous Substitution Rates with Nonstationary Base Composition . Mol Biol Evol . 2018 Mar 1; 35 ( 3 ): 734 – 42 . OpenUrl CrossRef PubMed 174. ↵ Dutheil JY , Galtier N , Romiguier J , Douzery EJP , Ranwez V , Boussau B . Efficient Selection of Branch-Specific Models of Sequence Evolution . Mol Biol Evol . 2012 July; 29 ( 7 ): 1861 – 74 . OpenUrl CrossRef PubMed Web of Science 175. ↵ Kolpakov R. mreps: efficient and flexible detection of tandem repeats in DNA . Nucleic Acids Res . 2003 July 1;31(13):3672–8. 176. ↵ Novák P , Ávila Robledillo L , Koblížková A , Vrbová I , Neumann P , Macas J . TAREAN: a computational tool for identification and characterization of satellite DNA from unassembled short reads . Nucleic Acids Res . 2017 July 7; 45 ( 12 ): e111 – e111 . OpenUrl CrossRef PubMed 177. ↵ Joshi N , Fass J. Sickle: A sliding-window, adaptive, quality-based trimming tool for FastQ files . Version 1.33 https://github.com/najoshi/sickle . 2011. 178. ↵ Smit A , Hubley R , Green P. RepeatMasker Open-4.0. [Internet] . 2013 . Available from: 179. ↵ Hon J , Martínek T , Zendulka J , Lexa M. pqsfinder: an exhaustive and imperfection-tolerant search tool for potential quadruplex-forming sequences in R . Hancock J , editor. Bioinformatics . 2017 Nov 1;33(21):3373–9. 180. ↵ Rice P , Longden I , Bleasby A . EMBOSS: the European Molecular Biology Open Software Suite . Trends Genet TIG . 2000 June; 16 ( 6 ): 276 – 7 . OpenUrl CrossRef PubMed 181. ↵ Kim D , Paggi JM , Park C , Bennett C , Salzberg SL . Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype . Nat Biotechnol . 2019 Aug; 37 ( 8 ): 907 – 15 . OpenUrl CrossRef PubMed 182. ↵ Li H , Durbin R . Fast and accurate long-read alignment with Burrows–Wheeler transform . Bioinformatics . 2010 Mar 1; 26 ( 5 ): 589 – 95 . OpenUrl CrossRef PubMed Web of Science 183. ↵ Anders S , Pyl PT , Huber W. HTSeq – A Python framework to work with high-throughput sequencing data [Internet] . 2014 [cited 2025 Jan 21]. Available from: http://biorxiv.org/lookup/doi/10.1101/002824 184. ↵ R Development Core Team . R: A Language and Environment for Statistical Computing [Internet] . Vienna, Austria; 2023 . (R Foundation for Statistical Computing). Available from: http://www.R-project.org 185. ↵ Pagès H , Aboyoun P , Gentleman R , DebRoy A Saikat . Biostrings: Efficient manipulation of biological strings [Internet] . 2024 . Available from: https://bioconductor.org/packages/Biostrings 186. ↵ Wright ES . DECIPHER: harnessing local sequence context to improve protein multiple sequence alignment . BMC Bioinformatics . 2015 Dec; 16 ( 1 ): 322 . OpenUrl CrossRef PubMed 187. ↵ Lawrence M , Huber W , Pagès H , Aboyoun P , Carlson M , Gentleman R , et al. Software for Computing and Annotating Genomic Ranges . PLoS Comput Biol . 2013 Aug 8; 9 ( 8 ): e1003118 . OpenUrl CrossRef PubMed 188. ↵ Kay M. ggdist: Visualizations of Distributions and Uncertainty in the Grammar of Graphics . IEEE Trans Vis Comput Graph . 2023 ;1–11. 189. ↵ Gel B , Serra E. karyoploteR: an R/Bioconductor package to plot customizable genomes displaying arbitrary data . Hancock J , editor. Bioinformatics . 2017 Oct 1;33(19):3088–90. 190. ↵ Lawrence M , Gentleman R , Carey V. rtracklayer: an R package for interfacing with genome browsers . Bioinformatics . 2009 July 15; 25 ( 14 ): 1841 – 2 . OpenUrl CrossRef PubMed Web of Science 191. ↵ Charif D , Lobry J . Seqin{R} 1.0-2: a contributed package to the {R} project for statistical computing devoted to biological sequences retrieval and analysis. In: Structural approaches to sequence evolution: Molecules, networks, populations . U. Bastolla , M. Porto , H.E. Roman and M. Vendruscolo . New York : Springer Verlag ; 2007 . p. 207–32. (Biological and Medical Physics, Biomedical Engineering). 192. ↵ Wickham H , Averick M , Bryan J , Chang W , McGowan L , François R , et al. Welcome to the Tidyverse . J Open Source Softw . 2019 Nov 21; 4 ( 43 ): 1686 . OpenUrl CrossRef 193. ↵ Gehlenborg N. UpSetR: A More Scalable Alternative to Venn and Euler Diagrams for Visualizing Intersecting Sets [Internet] . 2019 . Available from: https://CRAN.R-project.org/package=UpSetR 194. ↵ Arnaiz O , Meyer E , Sperling L. ParameciumDB 2019: integrating genomic data across the genus for functional and evolutionary biology . Nucleic Acids Res . 2020 Jan 8;48(D1):D599–605. 195. Duharcourt S , Sperling L . The Challenges of Genome-Wide Studies in a Unicellular Eukaryote With Two Nuclear Genomes . Methods Enzymol . 2018 ; 612 : 101 – 26 . OpenUrl CrossRef PubMed 196. Duret L , Galtier N . Biased Gene Conversion and the Evolution of Mammalian Genomic Landscapes . Annu Rev Genomics Hum Genet . 2009 Sept 1; 10 ( 1 ): 285 – 311 . OpenUrl CrossRef PubMed Web of Science 197. Fu L , Gu C , Mochizuki K , Xiong J , Miao W , Wang G . The genome-wide meiotic recombination landscape in ciliates and its implications for crossover regulation and genome evolution . J Genet Genomics . 2023 Oct;S1673852723002072. View the discussion thread. Back to top Previous Next Posted November 07, 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 The tiny germline chromosomes of Paramecium aurelia have an exceptionally high recombination rate and are capped by a new class of Helitrons 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 The tiny germline chromosomes of Paramecium aurelia have an exceptionally high recombination rate and are capped by a new class of Helitrons Olivier Arnaiz , Frédéric Guérin , Arnaud Couloux , Caridad Miró-Pina , Guillaume Pellerin , Irina Nekrasova , Joëlle Amselem , Jean-Marc Aury , Simran Bhullar , Andrea Frapporti , Emmanuelle Lerat , Isabelle Luyten , Sophie Malinsky , Nathalie Mathy , Alexey Potekhin , Vinciane Régnier , Natalia Sawka-Gądek , Amandine Touzeau , Augustin de Vanssay , Coralie Zangarelli , Hadi Quesneville , Mireille Bétermier , Karine Labadie , Laurent Duret , Eric Meyer , Sandra Duharcourt , Linda Sperling bioRxiv 2025.11.06.686955; doi: https://doi.org/10.1101/2025.11.06.686955 Share This Article: Copy Citation Tools The tiny germline chromosomes of Paramecium aurelia have an exceptionally high recombination rate and are capped by a new class of Helitrons Olivier Arnaiz , Frédéric Guérin , Arnaud Couloux , Caridad Miró-Pina , Guillaume Pellerin , Irina Nekrasova , Joëlle Amselem , Jean-Marc Aury , Simran Bhullar , Andrea Frapporti , Emmanuelle Lerat , Isabelle Luyten , Sophie Malinsky , Nathalie Mathy , Alexey Potekhin , Vinciane Régnier , Natalia Sawka-Gądek , Amandine Touzeau , Augustin de Vanssay , Coralie Zangarelli , Hadi Quesneville , Mireille Bétermier , Karine Labadie , Laurent Duret , Eric Meyer , Sandra Duharcourt , Linda Sperling bioRxiv 2025.11.06.686955; doi: https://doi.org/10.1101/2025.11.06.686955 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 Genomics Subject Areas All Articles Animal Behavior and Cognition (7642) Biochemistry (17715) Bioengineering (13907) Bioinformatics (42003) Biophysics (21470) Cancer Biology (18624) Cell Biology (25533) Clinical Trials (138) Developmental Biology (13390) Ecology (19935) Epidemiology (2067) Evolutionary Biology (24356) Genetics (15617) Genomics (22529) Immunology (17753) Microbiology (40432) Molecular Biology (17200) Neuroscience (88681) Paleontology (667) Pathology (2840) Pharmacology and Toxicology (4828) Physiology (7653) Plant Biology (15171) Scientific Communication and Education (2046) Synthetic Biology (4304) Systems Biology (9826) Zoology (2271)