Reconstruction of proto-vertebrate, proto-cyclostome and proto-gnathostome genomes provides new insights into early vertebrate evolution

ARTICLE https://doi.org/10.1038/s41467-021-24573-z OPEN 1234567890():,; Reconstruction of proto-vertebrate, protocyclostome and proto-gnathostome genomes provides new insights into early vertebrate evolution Yoichiro Nakatani Aoife McLysaght 1,4,5, Prashant Shingate2,5, Vydianathan Ravi 1 ✉ & Byrappa Venkatesh 2,3 ✉ 2, Nisha E. Pillai2, Aravind Prasad 2, Ancient polyploidization events have had a lasting impact on vertebrate genome structure, organization and function. Some key questions regarding the number of ancient polyploidization events and their timing in relation to the cyclostome-gnathostome divergence have remained contentious. Here we generate de novo long-read-based chromosome-scale genome assemblies for the Japanese lamprey and elephant shark. Using these and other representative genomes and developing algorithms for the probabilistic macrosynteny model, we reconstruct high-resolution proto-vertebrate, proto-cyclostome and proto-gnathostome genomes. Our reconstructions resolve key questions regarding the early evolutionary history of vertebrates. First, cyclostomes diverged from the lineage leading to gnathostomes after a shared tetraploidization (1R) but before a gnathostome-specific tetraploidization (2R). Second, the cyclostome lineage experienced an additional hexaploidization. Third, 2R in the gnathostome lineage was an allotetraploidization event, and biased gene loss from one of the subgenomes shaped the gnathostome genome by giving rise to remarkably conserved microchromosomes. Thus, our reconstructions reveal the major evolutionary events and offer new insights into the origin and evolution of vertebrate genomes. 1 Smurfit Institute of Genetics, Trinity College Dublin, University of Dublin, Dublin 2, Ireland. 2 Comparative and Medical Genomics Laboratory, Institute of Molecular and Cell Biology, A*STAR, Biopolis, Singapore, Singapore. 3 Department of Paediatrics, Yong Loo Lin School of Medicine, National University of Singapre, Singapore, Singapore. 4Present address: Graduate School of Medicine, Osaka University, Osaka, Japan. 5These authors contributed equally: Yoichiro Nakatani, Prashant Shingate. ✉email: ; NATURE COMMUNICATIONS | (2021)12:4489 | https://doi.org/10.1038/s41467-021-24573-z | www.nature.com/naturecommunications 1 ARTICLE T NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-24573-z he emergence of morphologically complex vertebrates from invertebrate chordates is considered a major evolutionary transition that led to the emergence of more than 70,000 vertebrate species (http://vgpdb.snu.ac.kr/splist), including humans. The common ancestor of vertebrates that originated during the Lower Cambrian1 diverged to give rise to the two extant lineages of vertebrates, the cyclostomes (jawless vertebrates) and gnathostomes (jawed vertebrates). Cyclostomes are a monophyletic group2 comprising lampreys and hagfishes, while gnathostomes include cartilaginous fishes (Chondrichthyes, represented by chimaeras, sharks and rays) and bony vertebrates (Osteichthyes, represented by ray-finned fishes and lobe-finned fishes, including tetrapods). Cyclostomes are sometimes thought to be morphologically primitive as compared to gnathostomes as they lack hinged jaws, paired appendages and nostrils, mineralized tissue, and a discrete pancreas3–5. However, recent studies suggest that the lamprey and hagfish lineages independently acquired their seemingly simplified as well as their specialized morphology, and that the ancestral cyclostome already had a complex morphology and physiology distinct from the gnathostome lineage6,7. For example, although cyclostomes lack the major histocompatibility complex and immunoglobulin-based adaptive immune system (AIS) of gnathostomes, they have independently evolved somatically diversifying variable lymphocyte receptors for antigen recognition8. Evolutionary innovations at the origin of vertebrates have been proposed to be the result of ancient tetraploidization events that generated additional copies of the entire genome9,10. This view is now widely accepted because genome-wide synteny and paralogy analyses11–15 have provided convincing evidence for two rounds of tetraploidization (known as 1R and 2R, respectively) during early vertebrate evolution (see for review refs. 16,17; see also refs. 18,19). However, the timing of 1R and 2R relative to the cyclostome–gnathostome divergence has remained contentious— a significant gap in our knowledge considering the important implications for the genetic basis of the shared and derived features of these two lineages. Previous studies have produced conflicting results supporting each of the three possibilities18–26, i.e. divergence occurring prior to 1R, between 1R and 2R, and after 2R (Fig. 1). This uncertainty has been further compounded by the discovery of six Hox clusters in both lampreys and hagfish19,22,27 compared to four clusters in most gnathostome lineages, suggesting the possibility of an additional tetraploidization or chromosome-scale segmental duplications in the cyclostome ancestor18,19,22. Resolving these alternative scenarios using gene trees has proved to be challenging. This is partly due to the presence of multiple ‘ohnologues' (paralogous genes generated by polyploidy) created by successive rounds of tetraploidization; lineage-specific secondary losses of some ohnologues28,29; as well as the confounding effects of asymmetric evolutionary divergence between ohnologues30. The possibility of delayed rediploidization after a tetraploidization event28,31,32 has further complicated the interpretation of gene trees, as it uncouples gene duplication time from the divergence time of the ohnologues. In addition, the tendency of lamprey ohnologues to cluster outside gnathostome gene clades due to high GC-content and consequent codon bias22,26,33 has impeded the use of gene trees for determining the timing of 1R and 2R. An alternative and more effective strategy for the identification of ancient polyploidy is the macrosynteny-based reconstruction of ancestral genomes12–15. In particular, this strategy has the potential to reveal chromosome fusion/fission events that occurred in the interval between 1R and 2R and/or after 2R13,15. Whether such genome rearrangements are shared by cyclostomes and gnathostomes would be potentially informative for 2 determining the timing of the cyclostome–gnathostome divergence in relation to 1R and 2R. A prerequisite for such comparisons is the high-resolution reconstruction of the protocyclostome and the proto-gnathostome genomes which require high-quality, chromosome-scale genome assemblies from the most basal vertebrate lineages such as cyclostomes and cartilaginous fishes. In the present study, we generate de novo chromosome-scale genome assemblies of a cyclostome, the Japanese lamprey (Lethenteron japonicum; also known as the Arctic lamprey Lethenteron camtschaticum) and a cartilaginous fish, the elephant shark (Callorhinchus mi (...truncated)