Editor's evaluation: Comparative genomics reveals insight into the evolutionary origin of massively scrambled genomes
Year: 2022
Type: peer-review
Abstract: Article Figures and data Abstract Editor's evaluation Introduction Results Discussion Methods Data availability References Decision letter Author response Article and author information Metrics Abstract Ciliates are microbial eukaryotes that undergo extensive programmed genome rearrangement, a natural genome editing process that converts long germline chromosomes into smaller gene-rich somatic chromosomes. Three well-studied ciliates include Oxytricha trifallax, Tetrahymena thermophila, and Paramecium tetraurelia, but only the Oxytricha lineage has a massively scrambled genome, whose assembly during development requires hundreds of thousands of precisely programmed DNA joining events, representing the most complex genome dynamics of any known organism. Here we study the emergence of such complex genomes by examining the origin and evolution of discontinuous and scrambled genes in the Oxytricha lineage. This study compares six genomes from three species, the germline and somatic genomes for Euplotes woodruffi, Tetmemena sp., and the model ciliate O. trifallax. We sequenced, assembled, and annotated the germline and somatic genomes of E. woodruffi, which provides an outgroup, and the germline genome of Tetmemena sp. We find that the germline genome of Tetmemena is as massively scrambled and interrupted as Oxytricha’s: 13.6% of its gene loci require programmed translocations and/or inversions, with some genes requiring hundreds of precise gene editing events during development. This study revealed that the earlier diverged spirotrich, E. woodruffi, also has a scrambled genome, but only roughly half as many loci (7.3%) are scrambled. Furthermore, its scrambled genes are less complex, together supporting the position of Euplotes as a possible evolutionary intermediate in this lineage, in the process of accumulating complex evolutionary genome rearrangements, all of which require extensive repair to assemble functional coding regions. Comparative analysis also reveals that scrambled loci are often associated with local duplications, supporting a gradual model for the origin of complex, scrambled genomes via many small events of DNA duplication and decay. Editor's evaluation The study marks a significant advance in the field of evolutionary genomics of ciliates, an ancient and highly diverse eukaryotic phylum with many idiosyncrasies that teach us valuable lessons, inter alia, about sex and the plasticity of genomes. By focusing on two species from the same family, plus a more distant outgroup within the same class, this valuable study provides new and compelling information on evolutionary trends of genome rearrangement among different species of this interesting group of organisms. The work will be of interest to anyone interested in genome dynamics. https://doi.org/10.7554/eLife.82979.sa0 Decision letter eLife's review process Introduction Organisms do not always contain a single, static genome. Programmed genome editing is a naturally occurring and essential part of development in many organisms, including ciliates (Chen et al., 2014), nematodes (Mitreva et al., 2005), lampreys (Smith et al., 2012), and zebra finches (Biederman et al., 2018). Most of these events involve precise removal and rejoining of large regions of DNA during postzygotic differentiation of a somatic genome from a germline genome. Ciliates are microbial eukaryotes with two types of nuclei: a somatic macronucleus (MAC) that differentiates from a germline micronucleus (MIC). In the model ciliate Oxytricha, the MAC is entirely active chromatin (Beh et al., 2019) and the hub of transcription. The three species that we compare are all spirotrichs, which have gene-sized ‘nanochromosomes’ in the MAC, present at high copy number (Swart et al., 2013; Chen et al., 2015; Wang et al., 2016; Chen et al., 2019; Lindblad et al., 2019; Vinogradov et al., 2012). The diploid MIC participates in sexual reproduction, but its megabase-sized chromosomes are mostly transcriptionally silent. Gene loci are often arranged discontinuously in the MIC, with short genic segments called macronuclear destined sequences (MDSs), interrupted by stretches of non-coding DNA called internally eliminated sequences (IESs) (Figure 1A). During sexual development, a new MAC genome rearranges from a copy of the zygotic MIC genome. MDSs join in the correct order and orientation, whereas MIC-limited genomic regions undergo programmed deletion, including repetitive elements, intergenic regions, and IESs (Figure 1A). Though analogous to intron splicing, these events occur on DNA. The MDSs for some MAC chromosomes are scrambled if they require translocation or inversion during MAC development (Figure 1A). Pairs of short repeats, called pointers, are present at MDS-IES junctions in both scrambled and nonscrambled loci (Mitcham et al., 1992; Prescott, 1994). Pointer sequences are present twice in the MIC, at the end of MDS n and the beginning of MDS n+1. One copy of the repeat is retained at each MDS-MDS junction in a mature MAC chromosome (Figure 1A). These microhomologous regions help guide MDS recombination, but most are non-unique, and the shortest pointers are just 2 bp. Thousands of long, noncoding template RNAs collectively program MDS joining (Nowacki et al., 2008; Lindblad et al., 2017; Yerlici and Landweber, 2014). Figure 1 Download asset Open asset Genome rearrangements in representative ciliate species. (A) Diagram of genome rearrangement in Oxytricha. Each ciliate cell contains a somatic macronucleus (MAC) and a germline micronucleus (MIC). During development, the MAC genome rearranges from a copy of the MIC genome. (1) Nonscrambled genes rearrange simply by joining consecutive macronuclear destined sequences (MDSs, blue boxes) and removing internal eliminated sequences (IESs, thin lines). (2) Rearrangement of scrambled genes requires MDS translocation and/or inversion. Pointers are microhomologous sequences (colored vertical bars) present in two copies in the MIC and only one copy in the MAC where consecutive MDSs recombine. (B) Comparison of genome rearrangement features of representative ciliates and the non-ciliate Plasmodium falciparum as an outgroup (phylogenetic information is based on Parfrey et al., 2011; Bracht et al., 2013). Conclusions from this study are shown in bold. * indicates that some scrambled pointers in Euplotes woodruffi are much longer, as discussed in the results. Statistics for pointers ≤30 bp in E. woodruffi are shown. Table information derives from the following sources: 1 - Swart et al., 2013; 2 - Lindblad et al., 2019; 3 - Chen et al., 2014; 4 - Chen et al., 2015; 5 - Sheng et al., 2020; 6 - Eisen et al., 2006; 7 - Hamilton et al., 2016; 8 - Aury et al., 2006; 9 - Guérin et al., 2017; 10 - Arnaiz et al., 2012; 11 - Riley and Katz, 2001; 12 - Maurer-Alcalá et al., 2018a; 13 - Katz and Kovner, 2010; 14 - Gao et al., 2014. Numerous studies have inferred the possible scope of genome rearrangement in different ciliate species using partial genome surveys. In Paramecium, PiggyMac-depleted cells fail to remove MIC-limited regions properly, which provided a resource to annotate ~45,000 IESs prior to assembly of a draft MIC genome (Arnaiz et al., 2012). The use of single-cell sequencing has allowed pilot studies to sample partial MIC genomes of diverse species (Chen et al., 2019; Maurer-Alcalá et al., 2018a; Maurer-Alcalá et al., 2018b; Smith et al., 2020). Alignment of tentative MIC reads to either assembled MAC genomes or single-cell transcriptome data predicts over 20 candidate scrambled loci in two basal ciliates, Loxodes sp. and Blepharisma americanum (Maurer-Alcalá et al., 2018b) and hundreds of candidate loci in the tintinnid Schmidingerella arcuata (Smith et al., 2020). Nearly one-third (31%) of approximately 5000 surveyed transcripts may be scrambled in Chilodonella uncinata (Maurer-Alcalá et al., 2018a, Figure 1B), which has four confirmed cases of scrambled genes (Katz and Kovner, 2010; Gao et al., 2014). Transcriptome-based surveys offer less precise estimates and cannot distinguish RNA splicing. Several computational pipelines have been developed to facilitate the inference of genome rearrangement features by split-read mapping in the absence of complete MIC or MAC reference genomes (Denby Wilkes et al., 2016; Zheng et al., 2020; Feng et al., 2020; Seah et al., 2021). By surveying lighter genome coverage prior to full sequencing, these tools provide partial insight into germline architecture. This helps guide selection of species for full genome sequencing and subsequent construction of complete rearrangement maps between the MIC and MAC genomes. High-quality MIC genome reference assemblies are only currently available for three ciliate genera: Oxytricha (Chen et al., 2014), Tetrahymena (Hamilton et al., 2016), and Paramecium (Guérin et al., 2017; Sellis et al., 2021). Programmed genome rearrangements in Oxytricha exhibit the highest accuracy and largest scale of any known natural gene-editing system, with exquisite control over hundreds of thousands of precise DNA cleavage/joining events. Accordingly, its germline genome structure is arguably the most complex of any model organism (Chen et al., 2014), requiring programmed deletion of over 90% of the germline DNA during development and massive descrambling of the resulting fragments to construct a new MAC genome of over 18,000 chromosomes (Lindblad et al., 2019). This differs from the distantly related Tetrahymena and Paramecium that both eliminate ~30% of the germline genome (Hamilton et al., 2016; Guérin et al., 2017). Paramecium uses exclusively 2 bp pointers and lacks evidence of any scrambled loci. A small number of scrambled loci (4 confirmed out of 2711 candidates) have been reported in Tetrahymena (Sheng et al., 2020, Figure 1B). Tetrahymena and Paramecium diverged from Oxytricha over 1 billion years ago (Parfrey et al., 2011; Bracht et al., 2013), which leaves a large gap in our understanding of the emergence of complex DNA rearrangements in the Oxytricha lineage. Open questions include how did the Oxytricha germline genome acquire its high number of IES insertions and how do scrambled loci arise and evolve. Three previous studies tackled these questions at the level of single genes and orthologs, including DNA polymerase α, actin I, and Telomere end-binding protein subunit α (Hogan et al., 2001; Chang et al., 2005; Wong and Landweber, 2006; DuBois and Prescott, 1995). Here, we provide the first comparative genomic analysis of Oxytricha trifallax and two other spirotrichous ciliates, Tetmemena sp. and Euplotes woodruffi. Tetmemena sp. is a hypotrich similar to Tetmemena pustulata, formerly Stylonychia pustulata (Chen et al., 2015), in the same family as O. trifallax (Figure 1B; Chen et al., 2014; Chen et al., 2015). Hypotrichs are noted for the presence of scrambled genes, based on previous ortholog comparisons (Chen et al., 2015; Hogan et al., 2001; Chang et al., 2005; DuBois and Prescott, 1995; Figure 1B). E. woodruffi, together with the hypotrichous ciliates, belong to the class Spirotrichea (Figure 1B). Like hypotrichs, Euplotes also has gene-sized nanochromosomes in the MAC genome (Wang et al., 2016; Chen et al., 2019; Chen et al., 2021), but this outgroup uses a different genetic code (UGA is reassigned to cysteine, Meyer et al., 1991), and little is known about its MIC genome. A partial MIC genome of Euplotes vannus was previously assembled, and it contains highly conserved TA pointers (Chen et al., 2019), consistent with previous observations in Euplotes crassus (Klobutcher and Herrick, 1995). This differs from O. trifallax, which uses longer pointers of varying lengths, with scrambled pointers typically longer than nonscrambled ones (Chen et al., 2014, Figure 1B). This observation suggests that longer pointers may supply more information to facilitate MDS descrambling, sometimes over great distances. Therefore, the preponderance of 2 bp pointers in the other Euplotes species could indicate limited capacity to support scrambled genes, and a partial genome survey of E. vannus concluded that at least 97% of loci are nonscrambled (Chen et al., 2019). Early studies of Euplotes octocarinatus, on the other hand, demonstrated its use of longer pointers (that usually contain TA) (Tan et al., 1999; Wang et al., 2005), suggesting that some members of the Euplotes genus may have the capacity to support complex genome reorganization. To investigate the origin of scrambled genomes, we choose E. woodruffi as an outgroup, because it is closely related to E. octocarinatus (Syberg-Olsen et al., 2016) and feasible to culture in the lab. This study includes the de novo assemblies of the micronuclear genome of Tetmemena sp. and both genomes of E. woodruffi. The availability of MIC and MAC genomes for both species allows us to annotate and compare their genome rearrangement maps and other key features to each other and to O. trifallax. The MIC genome of Tetmemena is extremely interrupted, like Oxytricha. While the E. woodruffi MIC genome is much more IES-sparse, it contains thousands of scrambled genes, whose architecture we compare to orthologous loci in the other species. We infer that the evolutionary origin of scrambled genes is associated with local duplications, providing strong support for a previously proposed simple evolutionary model requiring only duplication and decay (Gao et al., 2015) that allows for the evolutionary expansion of extremely rearranged chromosome architectures. Results Germline genome expansion via repetitive elements Tetmemena sp. and E. woodruffi were both propagated in laboratory culture from single cells. The E. woodruffi MAC genome was sequenced and assembled from paired-end Illumina reads from whole cell DNA, which is mostly MAC-derived. For comparative analysis, the MAC genome of E. woodruffi was assembled using the same pipeline previously used for Tetmemena sp. (Chen et al., 2015). Because MIC DNA is significantly more sparse than MAC DNA in individual cells (Prescott, 1994), MIC DNA was enriched before sequencing (see Methods); however, this leads to much lower sequence coverage of the MIC than the MAC. Third-generation long reads (Pacific Biosciences and Oxford Nanopore Technologies) were combined with Illumina paired-end reads (Methods, see genome coverage in Supplementary file 1) to construct hybrid genome assemblies for Tetmemena sp. and E. woodruffi. Though the final genome assemblies are still fragmented, often due to transposon or other repetitive insertions at boundaries (Figure 2—figure supplement 1), the current draft assemblies cover most (>90%) MDSs for 89.1% of MAC nanochromosomes in Tetmemena, and for 90.0% of MAC nanochromosomes in E. woodruffi. This allowed us to establish near-complete rearrangement maps for the newly assembled genomes of Tetmemena and E. woodruffi, at a level comparable to the published reference for O. trifallax (Chen et al., 2014), which is appropriate for comparative analysis. Table 1 shows a comparison of genome features for the three species. The three MAC genomes are similar in size, with most nanochromosomes bearing only one gene. The size distributions of MAC chromosomes are similar for the three species, though slightly shorter for E. woodruffi, consistent with prior observation via gel electrophoresis (Prescott, 1994, Figure 2—figure supplement 2). Like O. trifallax (Swart et al., 2013), the maximum number of genes encoded on one chromosome is 7–8 (Table 1). Surprisingly, the MIC genome sizes differ substantially: the Tetmemena MIC genome assembly is 237 Mbp, nearly half that of Oxytricha. The E. woodruffi MIC genome assembly is even smaller, approximately 172 Mbp (Table 1). Table 1 Statistics of somatic macronucleus (MAC) and germline micronucleus (MIC) genomes in three species. Oxytricha trifallaxTetmemena sp.Euplotes woodruffiMACa,*MICbMACcMIC†MAC†MIC†Genome size (Mbp)67.149660.623772.2172N50 (bp)374527,807333914,722270244,656GC%31.3628.4437.0532.1736.5635.31Number of contigs‡22,42625,72025,20628,44635,09917,655Two-telomere contigs14,225-15,802-19,061-Telomeric contigs20,336-21,165-28,294-Single-gene telomeric contigs76.1%-75.5%-68.5%-Maximum number of genes on a telomeric contig8-7-8- a - Swart et al., 2013; b - Chen et al., 2014; c - Chen et al., 2015. * This study used the MAC genome of Oxytricha from Swart et al., 2013 instead of the long-read assembly in Lindblad et al., 2019, because the short MAC genomes in the present study were primarily assembled from Illumina reads, as in Swart et al., 2013. Lindblad et al., 2019 updated Swart et al., 2013 by including nanochromosomes captured in single long reads, which are currently not available for the other two species. The MIC genomes of Tetmemena and E. woodruffi were assembled to a similar N50 as the reference O. trifallax genome (Chen et al., 2014) for comparative analysis. † Data from this study. ‡ Telomere-bearing element (TBE) transposon contaminants in MAC contigs were removed (Methods). Therefore, 24 Oxytricha MAC contigs and 13 Tetmemena MAC contigs were removed from the published versions. The expansion of repetitive elements in the Oxytricha lineage may contribute to the difference in MIC genome sizes (Figure 2A–C). Oxytricha has a variety of tranposable elements (TEs) in the MIC, with telomere-bearing elements (TBEs) of the Tc1/mariner family the most abundant (Chen et al., 2014; Chen and Landweber, 2016, Supplementary file 2). A complete TBE transposon contains three open reading frames (ORFs). ORF1 encodes a 42kD transposase with a DDE-catalytic motif. Though present only in the germline, TBEs are so abundant in hypotrichs that some were partially recovered and assembled from whole cell DNA (Chen and Landweber, 2016). The Oxytricha MIC genome contains ~10,000 complete TBEs and ~24,000 partial TBEs, which occupy approximately 15.20% (75 Mbp) of the genome (Figure 2A, Supplementary file 3; Chen et al., 2014; Chen and Landweber, 2016). Tetmemena, on the other hand, has many fewer TBE ORFs and only 48 complete TBEs (Supplementary file 3), comprising 1.83% (4.3 Mbp) of its MIC genome (Figure 2B). E. crassus has also been reported to have an abundant transposon family called Tec elements (Transposon of Euplotes crassus). Like TBEs, each Tec consists of three ORFs, and ORF1 also encodes a transposase from the Tc1/mariner family (Baird et al., 1989; Krikau and Jahn, 1991; Jahn et al., 1993; Jahn et al., 1989; Klobutcher and Herrick, 1997). The ~57 kD ORF2 encodes a tyrosine-type recombinase (Doak et al., 2003), and the 20kD ORF3 has unknown function (Jahn et al., 1993). Using the three ORFs of Tec1 and Tec2 as queries for search, we identified 74 complete Tec elements in E. woodruffi. Collectively, Tec ORFs occupy 3.6 Mbp, corresponding to only 2.1% of the MIC genome (Figure 2C). Notably, the transposase-encoding ORF1 is more abundant than the other two TBE/Tec ORFs in all three ciliates (Supplementary file 3), consistent with its proposed role in DNA cleavage during genome rearrangement in Oxytricha (Nowacki et al., 2009). Figure 2 with 2 supplements see all Download asset Open asset The three germline micronucleus (MIC) genomes differ in repeat content, especially transposable elements. (A–C) MIC genome categories for (A) Oxytricha trifallax, (B) Tetmemena sp., and (C) Euplotes woodruffi. Oxytricha displays the greatest proportion of repetitive elements (telomere-bearing elements [TBE], other repeats, and tandem repeats) relative to the other species. Oxytricha MIC-specific genes were annotated in Chen et al., 2014; Miller et al., 2021. (D–F) Phylogenetic analysis of the three TBE open reading frames (ORFs) in Oxytricha and Tetmemena: (D) 42 kD, (E) 22 kD, and kD, that is the transposon family in Oxytricha. For each protein sequences from each species were and maximum using et al., Oxytricha contains three of to be the most ancient among hypotrichs, based on previous analysis of limited MIC genome data (Chen and Landweber, 2016). We using TBE sequences for all three ORFs from Oxytricha and Tetmemena (Figure This confirmed that only is present in the Tetmemena MIC genome, as proposed in Chen and Landweber, This also suggests that and in Oxytricha its from other hypotrichous in Figure 2—figure supplement the MIC genome of TBEs in Oxytricha and Tetmemena are with many insertions within consistent with either IESs as for or with the model (Klobutcher and Herrick, that some insertions may have as demonstrated in Paramecium et al., Feng and Landweber, 2021). sequence evolution at the of pointers and Prescott, to boundaries that longer precisely to TBE For of the of TBE see the and Tetmemena conserved rearrangement identified that Oxytricha has more MIC in the than Tetmemena or E. woodruffi (Figure of repeat in Supplementary file 2). Mbp of the Oxytricha MIC genome which is than reported in Chen et al., that used earlier of the is repetitive TBEs, tandem repeats, and other in Figure Mbp for Tetmemena and Mbp for E. woodruffi. Mbp in repeat partially the significantly MIC genome size of Oxytricha the other The E. woodruffi genome has fewer IESs We used the genome rearrangement DNA Rearrangement et al., to annotate the MIC genomes of Oxytricha, Tetmemena, and E. woodruffi (Methods). with their genetic the genomes of O. trifallax and Tetmemena have high of (Figure We annotated over MDSs in Oxytricha and over in Tetmemena with similar MDS distributions (Figure By E. woodruffi MDSs are typically longer, which indicates a less interrupted genome (Figure We the number of MDSs between for MAC chromosomes the three species and that the have similar coding sequence (Figure supplement is a strong between number of MDSs for orthologous genes in Oxytricha and Tetmemena Figure is among number of MDSs between of E. woodruffi and Oxytricha Figure E. woodruffi typically contain fewer Figure 3 with 2 supplements see all Download asset Open asset The three germline micronucleus genomes are interrupted by internally eliminated sequences (IESs) at different (A) destined sequences of Euplotes woodruffi are longer to Oxytricha or (B) between the of MDSs for orthologous genes in Tetmemena and in Oxytricha for is the function of is (C) in E. woodruffi have fewer MDSs to Oxytricha, with that many highly discontinuous genes in Oxytricha are in E. woodruffi on one are shown. (D) of pointers on somatic macronucleus (MAC) chromosomes in Oxytricha (E) E. woodruffi, with MAC chromosomes in gene Pointers significantly at the end of MAC chromosomes in E. woodruffi. Pointer on MAC chromosomes a of the The E. woodruffi genome is much less interrupted than that of Oxytricha or of MAC nanochromosomes in E. woodruffi IESs to only and in Oxytricha and Tetmemena, The sparse IES by in E. woodruffi displays a end on MAC in gene (Figure A is also present in Oxytricha (Figure and Tetmemena (Figure supplement In E. woodruffi IESs in the a short of (Figure Notably, the between the and the in E. woodruffi is just bp for approximately half that of Oxytricha (Swart et al., 2013). E. woodruffi has an intermediate level of genome genome rearrangements in all three species, which we for the first in Tetmemena and the diverged E. woodruffi. studies have scrambled genes with confirmed rearrangement maps for a limited species of hypotrichs (Chen et al., 2014; Chen et al., 2015; Hogan et al., 2001; Chang et al., 2005; Wong and Landweber, 2006; DuBois and Prescott, and Chilodonella (Katz and Kovner, 2010; Gao et al., 2014) but not in with the of Euplotes as an earlier diverged outgroup to hypotrichs 2008; Gao et al., 2016), the E. woodruffi genome is but it contains approximately half as many scrambled genes genes encoded on or of scrambled in O. trifallax genes encoded on and 13.6% in Tetmemena genes encoded on The E. woodruffi lineage may an evolutionary intermediate between genomes with only of genome and the more massively scrambled genomes of We infer that many genes were scrambled in the of Oxytricha and Tetmemena, because these two species approximately half of their scrambled genes (Supplementary file Furthermore, most scrambled genes are not new genes, they at least one ortholog in other ciliate species (Supplementary file Supplementary file genes are associated with local Notably, scrambled genes in all three species have more (Figure We identified genes from the same gene in the of the three species (Methods). For each species, with at least one scrambled gene are significantly than scrambled genes Figure This suggests a possible role of gene duplication in the origin of scrambled Figure 4 with 3 supplements see all Download asset Open asset genes have more than nonscrambled genes in the three species. at least one scrambled gene are than that scrambled genes in (A) Oxytricha, (B) Tetmemena, and (C) Euplotes woodruffi. pointers are longer than nonscrambled ones in all three species (Figure supplement consistent with prior observations (Chen et al., 2014) and the that longer pointers in more complex rearrangements, including between MDSs by et al., and nonscrambled IESs also differ in their (Figure supplement 2). scrambled in E. woodruffi be as long as 48 the more bp These long in E. woodruffi are more partial MDS (Figure supplement 1A). We also identified MDSs that to two or more regions within the same MIC (Supplementary file representing MDS and not regions could be into the rearranged MAC we find for all three species, are significantly more scrambled chromosomes than nonscrambled MAC chromosomes that contain at least one MDS Supplementary file is shown in Figure supplement 7 and The presence of MDSs contribute to the origin of scrambled rearrangements, as proposed in an model by Gao et al., 2015; in Figure supplement The model that MDS use of either MDS copy into the mature MAC in cells that the least MDS regions into the MAC gene have both a and a to the template RNA (Nowacki et al., that rearrangement, the of into the MAC The regions more decay into and scrambled pointers be to a shorter The that we identified in E. woodruffi may an intermediate in the origin of scrambled
Source: eLife
Author: Detlef Weigel
Institution: Max Planck Institute for Biology
Language: English
Subfield: Plant Science
Domain: Life Sciences
Open Access status: gold
Estimated APC paid: $2,500
OpenAlex ID: W4312225903