Amphioxus tails: source and fate of larval fin rays and the metamorphic transition from an ectodermal to a predominantly mesodermal tail

A median fin derived from the lateral plate mesoderm and the origin of paired fins

The development of paired appendages was a key innovation during evolution and facilitated the aquatic to terrestrial transition of vertebrates. Largely derived from the lateral plate mesoderm (LPM), one hypothesis for the evolution of paired fins invokes derivation from unpaired median fins via a pair of lateral fin folds located between pectoral and pelvic fin territories1. Whilst unpaired and paired fins exhibit similar structural and molecular characteristics, no definitive evidence exists for paired lateral fin folds in larvae or adults of any extant or extinct species. As unpaired fin core components are regarded as exclusively derived from paraxial mesoderm, any transition presumes both co-option of a fin developmental programme to the LPM and bilateral duplication2. Here, we identify that the larval zebrafish unpaired pre-anal fin fold (PAFF) is derived from the LPM and thus may represent a developmental intermediate between median and paired fins. We trace the contribution of LPM to the PAFF in both cyclostomes and gnathostomes, supporting the notion that this is an ancient trait of vertebrates. Finally, we observe that the PAFF can be bifurcated by increasing bone morphogenetic protein signalling, generating LPM-derived paired fin folds. Our work provides evidence that lateral fin folds may have existed as embryonic anlage for elaboration to paired fins.

The structure and function of centriolar rootlets

To gain a holistic understanding of cellular function, we must understand not just the role of individual organelles, but also how multiple macromolecular assemblies function collectively. Centrioles produce fundamental cellular processes through their ability to organise cytoskeletal fibres. In addition to nucleating microtubules, centrioles form lesser-known polymers, termed rootlets. Rootlets were identified over a 100 years ago and have been documented morphologically since by electron microscopy in different eukaryotic organisms. Rootlet-knockout animals have been created in various systems, providing insight into their physiological functions. However, the precise structure and function of rootlets is still enigmatic. Here, I consider common themes of rootlet function and assembly across diverse cellular systems. I suggest that the capability of rootlets to form physical links from centrioles to other cellular structures is a general principle unifying their functions in diverse cells and serves as an example of how cellular function arises from collective organellar activity.

Tail regeneration in a cephalochordate, the Bahamas lancelet, Asymmetron lucayanum

Lancelets (Phylum Chordata, subphylum Cephalochordata) readily regenerate a lost tail. Here, we use light microscopy and serial blockface scanning electron microscopy (SBSEM) to describe tail replacement in the Bahamas lancelet, Asymmetron lucayanum. One day after amputation, the monolayered epidermis has migrated over the wound surface. At 4 days, the regenerate is about 3% as long as the tail length removed. The re-growing nerve cord is a tubular outgrowth of ependymal cells, and the new part of the notochord consists of several degenerating lamellar cells anterior to numerous small vacuolated cells. The cut edges of the mesothelium project into the regenerate as tubular extensions. These tubes anastomose with each other and with midline mesodermal canals beneath the regenerating edges of the dorsal and ventral fins. SBSEM did not reveal a blastema-like aggregation of undifferentiated cells anywhere in the regenerate. At 6 days, the regenerate (10% of the amputated tail length) includes a notochord in which the small vacuolated cells mentioned above are differentiating into lamellar cells. At 10 days, the regenerate is 22% of the amputated tail length: myocytes have appeared in the walls of the myomeres, and sclerocoels have formed. By 14 days, the regenerate is 35% the length of the amputated tail, and the new tissues resemble smaller versions of those originally lost. The present results for A. lucayanum, a species regenerating quickly and with little inter-specimen variability, provide the morphological background for future cell-tracer, molecular genetic, and genomic studies of cephalochordate regeneration.

Mutation of amphioxus Pdx and Cdx demonstrates conserved roles for ParaHox genes in gut, anus and tail patterning

BACKGROUND

The homeobox genes Pdx and Cdx are widespread across the animal kingdom and part of the small ParaHox gene cluster. Gene expression patterns suggest ancient roles for Pdx and Cdx in patterning the through-gut of bilaterian animals although functional data are available for few lineages. To examine evolutionary conservation of Pdx and Cdx gene functions, we focus on amphioxus, small marine animals that occupy a pivotal position in chordate evolution and in which ParaHox gene clustering was first reported.

RESULTS

Using transcription activator-like effector nucleases (TALENs), we engineer frameshift mutations in the Pdx and Cdx genes of the amphioxus Branchiostoma floridae and establish mutant lines. Homozygous Pdx mutants have a defect in amphioxus endoderm, manifest as loss of a midgut region expressing endogenous GFP. The anus fails to open in homozygous Cdx mutants, which also have defects in posterior body extension and epidermal tail fin development. Treatment with an inverse agonist of retinoic acid (RA) signalling partially rescues the axial and tail fin phenotypes indicating they are caused by increased RA signalling. Gene expression analyses and luciferase assays suggest that posterior RA levels are kept low in wild type animals by a likely direct transcriptional regulation of a Cyp26 gene by Cdx. Transcriptome analysis reveals extensive gene expression changes in mutants, with a disproportionate effect of Pdx and Cdx on gut-enriched genes and a colinear-like effect of Cdx on Hox genes.

CONCLUSIONS

These data reveal that amphioxus Pdx and Cdx have roles in specifying middle and posterior cell fates in the endoderm of the gut, roles that likely date to the origin of Bilateria. This conclusion is consistent with these two ParaHox genes playing a role in the origin of the bilaterian through-gut with a distinct anus, morphological innovations that contributed to ecological change in the Cambrian. In addition, we find that amphioxus Cdx promotes body axis extension through a molecular mechanism conserved with vertebrates. The axial extension role for Cdx dates back at least to the origin of Chordata and may have facilitated the evolution of the post-anal tail and active locomotion in chordates.

Lineage-specific duplication of amphioxus retinoic acid degrading enzymes (CYP26) resulted in sub-functionalization of patterning and homeostatic roles

BACKGROUND

During embryogenesis, tight regulation of retinoic acid (RA) availability is fundamental for normal development. In parallel to RA synthesis, a negative feedback loop controlled by RA catabolizing enzymes of the cytochrome P450 subfamily 26 (CYP26) is crucial. In vertebrates, the functions of the three CYP26 enzymes (CYP26A1, CYP26B1, and CYP26C1) have been well characterized. By contrast, outside vertebrates, little is known about CYP26 complements and their biological roles. In an effort to characterize the evolutionary diversification of RA catabolism, we studied the CYP26 genes of the cephalochordate amphioxus (Branchiostoma lanceolatum), a basal chordate with a vertebrate-like genome that has not undergone the massive, large-scale duplications of vertebrates.

RESULTS

In the present study, we found that amphioxus also possess three CYP26 genes (CYP26-1, CYP26-2, and CYP26-3) that are clustered in the genome and originated by lineage-specific duplication. The amphioxus CYP26 cluster thus represents a useful model to assess adaptive evolutionary changes of the RA signaling system following gene duplication. The characterization of amphioxus CYP26 expression, function, and regulation by RA signaling demonstrated that, despite the independent origins of CYP26 duplicates in amphioxus and vertebrates, they convergently assume two main roles during development: RA-dependent patterning and protection against fluctuations of RA levels. Our analysis suggested that in amphioxus RA-dependent patterning is sustained by CYP26-2, while RA homeostasis is mediated by CYP26-1 and CYP26-3. Furthermore, comparisons of the regulatory regions of CYP26 genes of different bilaterian animals indicated that a CYP26-driven negative feedback system was present in the last common ancestor of deuterostomes, but not in that of bilaterians.

CONCLUSIONS

Altogether, this work reveals the evolutionary origins of the RA-dependent regulation of CYP26 genes and highlights convergent functions for CYP26 enzymes that originated by independent duplication events, hence establishing a novel selective mechanism for the genomic retention of gene duplicates.