Hybridization Chain Reaction for Quantitative and Multiplex Imaging of Gene Expression in Amphioxus Embryos and Adult Tissues

An amphioxus neurula stage cell atlas supports a complex scenario for the emergence of vertebrate head mesoderm

The emergence of new structures can often be linked to the evolution of novel cell types that follows the rewiring of developmental gene regulatory subnetworks. Vertebrates are characterized by a complex body plan compared to the other chordate clades and the question remains of whether and how the emergence of vertebrate morphological innovations can be related to the appearance of new embryonic cell populations. We previously proposed, by studying mesoderm development in the cephalochordate amphioxus, a scenario for the evolution of the vertebrate head mesoderm. To further test this scenario at the cell population level, we used scRNA-seq to construct a cell atlas of the amphioxus neurula, stage at which the main mesodermal compartments are specified. Our data allowed us to validate the presence of a prechordal-plate like territory in amphioxus. Additionally, the transcriptomic profile of somite cell populations supports the homology between specific territories of amphioxus somites and vertebrate cranial/pharyngeal and lateral plate mesoderm. Finally, our work provides evidence that the appearance of the specific mesodermal structures of the vertebrate head was associated to both segregation of pre-existing cell populations, and co-option of new genes for the control of myogenesis.

When Bigger Is Better: 3D RNA Profiling of the Developing Head in the Catshark Scyliorhinus canicula

We report the adaptation of RNA tomography, a technique allowing spatially resolved, genome-wide expression profiling, to a species occupying a key phylogenetic position in gnathostomes, the catshark Scyliorhinus canicula. We focused analysis on head explants at an embryonic stage, shortly following neural tube closure and of interest for a number of developmental processes, including early brain patterning, placode specification or the establishment of epithalamic asymmetry. As described in the zebrafish, we have sequenced RNAs extracted from serial sections along transverse, horizontal and sagittal planes, mapped the data onto a gene reference taking advantage of the high continuity genome recently released in the catshark, and projected read counts onto a digital model of the head obtained by confocal microscopy. This results in the generation of a genome-wide 3D atlas, containing expression data for most protein-coding genes in a digital model of the embryonic head. The digital profiles obtained for candidate forebrain regional markers along antero-posterior, dorso-ventral and left-right axes reproduce those obtained by in situ hybridization (ISH), with expected relative organizations. We also use spatial autocorrelation and correlation as measures to analyze these data and show that they provide adequate statistical tools to extract novel expression information from the model. These data and tools allow exhaustive searches of genes exhibiting any predefined expression characteristic, such a restriction to a territory of interest, thus providing a reference for comparative analyses across gnathostomes. This methodology appears best suited to species endowed with large embryo or organ sizes and opens novel perspectives to a wide range of evo-devo model organisms, traditionally counter-selected on size criterion.

The dorsoanterior brain of adult amphioxus shares similarities in expression profile and neuronal composition with the vertebrate telencephalon

BACKGROUND

The evolutionary origin of the telencephalon, the most anterior part of the vertebrate brain, remains obscure. Since no obvious counterpart to the telencephalon has yet been identified in invertebrate chordates, it is difficult to trace telencephalic origins. One way to identify homologous brain parts between distantly related animal groups is to focus on the combinatorial expression of conserved regionalisation genes that specify brain regions.

RESULTS

Here, we report the combined expression of conserved transcription factors known to specify the telencephalon in the vertebrates in the chordate amphioxus. Focusing on adult specimens, we detect specific co-expression of these factors in the dorsal part of the anterior brain vesicle, which we refer to as Pars anterodorsalis (PAD). As in vertebrates, expression of the transcription factors FoxG1, Emx and Lhx2/9 overlaps that of Pax4/6 dorsally and of Nkx2.1 ventrally, where we also detect expression of the Hedgehog ligand. This specific pattern of co-expression is not observed prior to metamorphosis. Similar to the vertebrate telencephalon, the amphioxus PAD is characterised by the presence of GABAergic neurons and dorsal accumulations of glutamatergic as well as dopaminergic neurons. We also observe sustained proliferation of neuronal progenitors at the ventricular zone of the amphioxus brain vesicle, as observed in the vertebrate brain.

CONCLUSIONS

Our findings suggest that the PAD in the adult amphioxus brain vesicle and the vertebrate telencephalon evolved from the same brain precursor region in ancestral chordates, which would imply homology of these structures. Our comparative data also indicate that this ancestral brain already contained GABA-, glutamatergic and dopaminergic neurons, as is characteristic for the olfactory bulb of the vertebrate telencephalon. We further speculate that the telencephalon might have evolved in vertebrates via a heterochronic shift in developmental timing.

A deep learning approach for staging embryonic tissue isolates with small data

Machine learning approaches are becoming increasingly widespread and are now present in most areas of research. Their recent surge can be explained in part due to our ability to generate and store enormous amounts of data with which to train these models. The requirement for large training sets is also responsible for limiting further potential applications of machine learning, particularly in fields where data tend to be scarce such as developmental biology. However, recent research seems to indicate that machine learning and Big Data can sometimes be decoupled to train models with modest amounts of data. In this work we set out to train a CNN-based classifier to stage zebrafish tail buds at four different stages of development using small information-rich data sets. Our results show that two and three dimensional convolutional neural networks can be trained to stage developing zebrafish tail buds based on both morphological and gene expression confocal microscopy images, achieving in each case up to 100% test accuracy scores. Importantly, we show that high accuracy can be achieved with data set sizes of under 100 images, much smaller than the typical training set size for a convolutional neural net. Furthermore, our classifier shows that it is possible to stage isolated embryonic structures without the need to refer to classic developmental landmarks in the whole embryo, which will be particularly useful to stage 3D culture in vitro systems such as organoids. We hope that this work will provide a proof of principle that will help dispel the myth that large data set sizes are always required to train CNNs, and encourage researchers in fields where data are scarce to also apply ML approaches.

The ADAR Family in Amphioxus: RNA Editing and Conserved Orthologous Site Predictions

RNA editing is a relatively unexplored process in which transcribed RNA is modified at specific nucleotides before translation, adding another level of regulation of gene expression. Cephalopods use it extensively to increase the regulatory complexity of their nervous systems, and mammals use it too, but less prominently. Nevertheless, little is known about the specifics of RNA editing in most of the other clades and the relevance of RNA editing from an evolutionary perspective remains unknown. Here we analyze a key element of the editing machinery, the ADAR (adenosine deaminase acting on RNA) gene family, in an animal with a key phylogenetic position at the root of chordates: the cephalochordate amphioxus. We show, that as in cephalopods, ADAR genes in amphioxus are predominantly expressed in the nervous system; we identify a number of RNA editing events in amphioxus; and we provide a newly developed method to identify RNA editing events in highly polymorphic genomes using orthology as a guide. Overall, our work lays the foundations for future comparative analysis of RNA-editing events across the metazoan tree.