The promise and pitfalls of synteny in phylogenomics

Abstract

Reconstructing the tree of life remains a central goal in biology. Early methods, which relied on small numbers of morphological or genetic characters, often yielded conflicting evolutionary histories, undermining confidence in the results. Investigations based on phylogenomics, which use hundreds to thousands of loci for phylogenetic inquiry, have provided a clearer picture of life's history, but certain branches remain problematic. To resolve difficult nodes on the tree of life, 2 recent studies tested the utility of synteny, the conserved collinearity of orthologous genetic loci in 2 or more organisms, for phylogenetics. Synteny exhibits compelling phylogenomic potential while also raising new challenges. This Essay identifies and discusses specific opportunities and challenges that bear on the value of synteny data and other rare genomic changes for phylogenomic studies. Synteny-based analyses of highly contiguous genome assemblies mark a new chapter in the phylogenomic era and the quest to reconstruct the tree of life.

Fig 1. Depictions of incongruence and alternate hypotheses for primates, the base of the animal tree, and teleost fish phylogenies.

(A) Example of tree incongruence. The weight of evidence strongly supports a sister relationship between bonobos and chimps, to the exclusion of humans. (B, C) Phylogenies that are incongruent would suggest a sister relationship between humans and chimps (B) or humans and bonobos (C). (D-G) The debate concerning early animal evolution has largely focused on whether sponges (D) or ctenophores (E) diverged first from all other animals: the sponge-first (F) and ctenophore-first (G) hypotheses, respectively. (H-M) Among teleost fish, the debate centers on the relationships among 3 major lineages—the Elopomorpha (mostly slim-headed fish; H), Osteoglossomorpha (mostly bony-tongued fish; I), and Clupeocephala (all other teleost fish; J). The Eloposteoglossocephala (EO-sister) hypothesis (K) suggests a sister relationship between slim-headed and bony-tongued fish, whereas the Elopomorpha-first (L) and Osteoglossomorpha-first (M) hypotheses suggest that slim-headed fish or bony-tongued fish, respectively, diverged before the other lineages split from one another. Recent studies that employed synteny as a phylogenomic marker supported the ctenophore-first (G) and EO-sister (K) hypotheses [8,9]. All images were obtained from the Wikimedia Commons (https://commons.wikimedia.org) or PhyloPic (https://www.phylopic.org) and are dedicated to the public domain; all credit goes to their respective contributors.

Fig 2. Data types for sequence-based phylogenetics.

Consider the relationships among 4 taxa (represented as T1, T2, T3, and T4), wherein the pairs T1 and T2, and T3 and T4 are sister to one another. Changes in genome architecture can be examined at the scale of microsynteny (short stretches of orthologous loci; A) or macrosynteny (long stretches of orthologous loci; B). Changes in synteny can be described by different processes, such as fusion events without-mixing (C) and with-mixing (D). (A) In the case of microsynteny, evidence of an inversion may occur between the blue and orange loci (bottom), which happened in the ancestor of T3 and T4. (B) The same phenomenon can happen in the case of macrosynteny. (C) Fusion-without-mixing events between 2 chromosomes may also reflect phylogeny. In this case, a fusion event may have occurred in the ancestor between T3 and T4 (bottom). (D) Fusion-with-mixing can also be used to reconstruct phylogeny. Note, the evolutionary scenarios at the bottom of panels A-D depict only the most likely of many possible scenarios. (E) Fusion-with-mixing events may occur in 2 steps. First, there is a fusion event, then rearrangements occur, scrambling the order of genes that once were encoded on separate chromosomes. As a result, the probability of going from a “no fusion” to “fusion-without-mixing” state (and vice versa), and going from a “fusion-without-mixing” state to a “fusion-with-mixing” state, is relatively higher than going from a “fusion-with-mixing” to a “fusion-without-mixing” state. Transitioning directly from a “no fusion” to a “fusion-with-mixing” state is highly unlikely and may require an intermediate “fusion-without-mixing” state. Transition probabilities may vary depending on the underlying genome biology of the organism, the size of the syntenic region, and other parameters.

Fig 3. Summary depictions of syntenies supporting the ctenophore-first and EO-sister hypotheses.

(A) Inferred phylogeny of animal and outgroup taxa used to examine the root of the animal tree. Under the ctenophore-first hypothesis, regions 1–7 each resulted from fusion events between 2 distinct chromosomes. The syntenic block depicted in orange for region 3 underwent a fission event in the choanoflagellate lineage, resulting in 2 chromosomes. Regions 4–7 underwent subsequent mixing events. Underneath each higher-order lineage name, the names of representatives used in the study [9] are listed. For example, among Bilateria, species from the genera Pecten and Branchiostoma were included in the study. Note, only fusion and mixing events relevant to rooting the animal tree are depicted. (B) Patterns of synteny in 7 different regions most parsimoniously support the ctenophore-first hypothesis. Examination of these regions indicates that all underwent fusion events and 4 also underwent mixing events. Each region is abbreviated as “R” along the phylogeny (for example, R1 refers to region 1). The number of genes in each syntenic region is listed at the bottom of the panel. (C) Inferred phylogeny of the 3 teleost fish groups, including an outgroup taxon (the chicken). Cartoon summary drawings of chromosomes are included for representative species. Common names of these species are provided below the taxonomic names. Highly contiguous genome assemblies facilitated the detection of chromosome fusing and mixing events after a whole genome duplication event. Chr, chromosome. (D) Chromosomes observed in extant species are depicted as cartoon summaries. Duplicated chromosomes from a whole genome duplication event are darkened. Silhouette images were obtained from PhyloPic (https://www.phylopic.org) and are dedicated to the public domain; all credit goes to their respective contributors.

Fig 4. A roadmap of challenges and opportunities for synteny-based phylogenomics.

(A) A high-level summary of steps toward best practices in synteny-based phylogenomics. Limitations in resource availability (computational power and researcher time) dictate that each project begins with a selection of taxa that are most relevant to the phylogenetic question at hand. For those taxa that lack high-quality genome assemblies, it will be necessary to sequence each genome (using long-read sequencing technology) and assemble the reads. In other cases, previously sequenced and assembled genomes may be publicly available. In either case, the next step is to annotate the genes in all selected genomes using a single high-quality annotation method. Comparisons among the gene complements of each organism should then be used to identify gene orthologs (orthologous loci are depicted in green, yellow, and blue). Orthologs can then be used in whole genome alignment and synteny detection. In addition, alignments of orthologs can be trimmed, assembled into multiple sequence alignments, and used for traditional phylogenenomics. After accounting for various sources of error, synteny blocks and multiple sequence alignments can be used to infer the topology of the tree of life. Note that obstacles in one step may be overcome by backtracking in the roadmap; for example, insufficient genome assembly completeness may benefit from additional genome sequencing. (B) Synteny data and organismal histories can be used for numerous research opportunities, including a better understanding of gene cluster function and evolution, reconstructing chromosome evolution, and inferring whole genome duplication events and ancestral genomes. For functional insights into gene clusters, fly embryos are depicted alongside gene clusters indicating how gene cluster organization may influence fly development. Silhouette images were obtained from PhyloPic (https://www.phylopic.org) and are dedicated to the public domain. Additional icons were obtained from bioicons (https://bioicons.com) and are available according to the CC-BY 4.0 license. Credit for silhouette images and icons goes to their respective contributors.