Advancing genetic and genomic technologies deepen the pool for discovery in Xenopus tropicalis

Abstract

Xenopus laevis and Xenopus tropicalis have long been used to drive discovery in developmental, cell, and molecular biology. These dual frog species boast experimental strengths for embryology including large egg sizes that develop externally, well-defined fate maps, and cell-intrinsic sources of nutrients that allow explanted tissues to grow in culture. Development of the Xenopus cell extract system has been used to study cell cycle and DNA replication. Xenopus tadpole tail and limb regeneration have provided fundamental insights into the underlying mechanisms of this processes, and the loss of regenerative competency in adults adds a complexity to the system that can be more directly compared to humans. Moreover, Xenopus genetics and especially disease-causing mutations are highly conserved with humans, making them a tractable system to model human disease. In the last several years, genome editing, expanding genomic resources, and intersectional approaches leveraging the distinct characteristics of each species have generated new frontiers in cell biology. While Xenopus have enduringly represented a leading embryological model, new technologies are generating exciting diversity in the range of discoveries being made in areas from genomics and proteomics to regenerative biology, neurobiology, cell scaling, and human disease modeling.

1 |. INTRODUCTION

Xenopus tropicalis and Xenopus laevis derive their strengths as model organisms both from their individual features and from using the two species in an intersectional or complementary fashion. Historically, the extremely large embryo sizes of X. laevis (~1 mm) made this species a primary driver of cut-and-paste embryological experiments that provided groundbreaking insight into phenomena such as induction, cell autonomy, and cellular reprogramming. While classical embryology remains enduringly powerful, genomic technologies in this species have also advanced dramatically in recent years, with the publication of the X. laevis genome,¹ successful generation of CRISPR mutants, and a rapidly-increasing number of studies using deep sequencing approaches. However, the alloteraploid genome and long (7–12 months) generation time still somewhat constrain forward genetic and epigenomic studies in X. laevis. These disadvantages are nicely complemented by X. tropicalis, its faster-maturing diploid cousin, which over the past 20 years has itself achieved full status as a model organism. Rob Grainger details the history of how the model X. tropicalis came to be in his excellent review.² X. tropicalis retain many of the characteristics that make X. laevis an attractive model system, including large (0.6–0.8 mm) eggs, rapid development, and cell-autonomous yolk stores that enable explant survival. Their diploid genome, which was published in draft form in 2010 and has recently been updated with a chromosome-level assembly^3,4 introduces a less complicated background for genetic manipulations and for genome-level analysis with next-generation sequencing applications. Leveraging the strengths of both X. tropicalis and X. laevis, evo-devo questions relating to body size control, genome duplication, and conservation of tetrapod development strategies can be easily interrogated.

X. laevis and X. tropicalis embryos are highly morphologically similar, and are even staged according to the same normal table,^5,6 which was generated for X. laevis. Additional excellent resources for embryo staging are available in the form of crowd-sourced micrographs (available through the community web resource Xenbase, www.xenbase.org) and the Zahn drawings.⁷ Nevertheless, there are some variations in embryological appearance that can make it valuable to have a full staging series for X. tropicalis as well. To this end, in this Perspective we provide a full staging series of X. tropicalis embryos from the single cell stage to the free-swimming tadpole (Figure 1). We also explore the X. tropicalis embryo as a model system and highlight technological advances that have informed diverse biological problems being newly addressed in this species.

FIGURE 1.

In this figure, we present a staging series for embryonic Xenopus tropicalis. Stages were broken into four major modules: cleavage, gastrulation, neurulation, and tailbud stages. Both animal and lateral views are shown to highlight the appearance of animal blastomeres as well as the pigment distinctions between animal and vegetal blastomeres. Gastrulation is shown as a continuum from stages 9 to 12.5. Neurulation comprises stages 13 to 18, and staging of this process is primarily predicted from lateral proximity of neural folds to one another in the dorsal view and closure of neural folds in the anterior view. When the neural fold finishes zippering, the embryo begins to elongate. During early tailbud stages (19–24) the embryo develops morphologically well-defined anterior structures including the pharyngeal arches, sensory placodes, and eyecup, while elongating along the anterior-posterior axis. Elongation, development of tail fin, resecting of gut, and development of refined organs then takes place over the rest of tailbud and early tadpole development. The lateral view is best used for staging by looking at the gut morphology, eye morphology, and body elongation. The dorsal view is best used to look at development of arches and elongation

2 |. GENETIC ENGINEERING TOOLS

Genome editing is rapidly transforming genetic analysis in X. tropicalis, since the initial adaptation of TALEN and then CRISPR Cas9 technologies in this species.^8–10 This represents an outgrowth of over two decades of work adapting and expanding the molecular and embryological strengths of X. laevis for use in X. tropicalis. As in X. laevis, tracer fluorescent mRNAs can be injected into specific X. tropicalis blastomeres to act as reporters in a whole-embryo or tissue-specific manner; and morpholinos or mRNAs can be injected for knock down or overexpression of gene products. In the same way, TALEN and CRISPR reagents can be injected to generate loss-of-function mutations.¹¹ Both F₀ mosaic genome editing and stable edited lines have proliferated in recent years, and community resources, notably at the national Xenopus Resource Center,¹² have prioritized generation of hundreds of mutant X. tropicalis lines for genes valuable to the greater Xenopus and medical genetics communities.

Ongoing efforts to expand and implement the full potential of genome editing in X. tropicalis are now targeting specific remaining hurdles. The first is the poor efficiency of homologous recombination in Xenopus cells in response to DNA damage,¹³ which has made knocking in mutations and transgenes difficult. In their review, Shi et al. describe recent methods to address this limitation by micro-homology end joining,¹⁴ homology dependent strategies,¹⁰ and homology independent strategies.¹⁵ The second limitation is that targeting essential genes leads to lethality; therefore, those mutants cannot be raised and propagated to the F₂ generation where phenotypic analysis generally occurs. While inducible technologies are being developed to circumvent this problem, a uniquely Xenopus-friendly technique, “leapfrog,” takes advantage of cut and paste embryology to transfer CRISPR engineered germ cells to a wild-type host.¹⁶This way, mutations are confined to transplanted germ cells, leaving all somatic tissues genetically intact, and the resulting F₀ frogs can be intercrossed to yield mutant compound heterozygotes in the F₁. This approach bypasses a full generation of breeding time until mutant phenotypes can be analyzed. Alternatively, F₀ animals can be outbred and their offspring intercrossed for eventual stable F₂ mutants, a strategy that still allows the researcher to make use of high doses of CRISPR reagents and the highest possible mutation efficiency in the F₀ donor germ line tissue without a risk of creating mostly biallelic embryonic-lethal mutations in somatic tissues, such that the resulting frog can be preserved as an adult. Continuing to develop efficient genetic engineering strategies in X. tropicalis will open up more robust loss-of-function strategies to probe critical genes in development, introduce targeted mutations, and build a larger libraries of reporter lines suited for live imaging and fluorescence-activated cell sorting. These technologies are indispensable to grow X. tropicalis as a premier system for mechanistic understanding of both basic biology and the basis of human disease.

3 |. GENOME-WIDE ANALYSES

Transcriptomic profiling using RNA-Seq gained immediate traction in X. tropicalis as soon as a draft genome was available,³ and in the years following publication of the genome, deep sequencing has been used to define the transcriptome across developmental stages and tissues,^17–19 during processes such as regeneration,²⁰ and under a multitude of molecular perturbations too numerous to detail. The genome assembly is now available at chromosome-level (version 9.1 as of this writing)⁴ and gene models and annotations are well-suited to RNA-Seq analysis. An exciting frontier in X. tropicalis, as in other model systems, is the potential for cell type discovery and type-specific transcriptomic profiling through single-cell RNA-Seq, which has recently been lever-aged to show that many cell types acquire their identity much earlier than previously appreciated.²¹ Both bulk and single-cell RNA-Seq are expected to continue to provide critical unbiased insight into the effects of developmental, genetic or pharmacological perturbations, and into the molecular basis of cell identity and differentiation. In addition to these transcriptomic technologies, which are also quite tractable in X. laevis, the more streamlined genome of X. tropicalis is highly amenable to epigenomic interrogation by methods like ChIP-Seq^22,23 and ATAC-Seq.²⁴ Notably, the external development and large size of early Xenopus embryos has allowed several studies characterizing the temporal dynamics of early-embryonic transcription factor occupancy, the histone code at promoters and enhancers, and the transition from maternal to zygotic transcription.^25–29 Adding an additional layer to these characterizations of transcriptional regulation, the extremely large blastomeres of Xenopus provide the extraordinary opportunity to carry out not only proteomic profiling of early embryos but even of individual cells, an approach that has so far been applied in X. laevis but is likely to be tractable in X. tropicalis as well.^30–32 Finally, genomic analysis is greatly facilitated by excellent support from the community web resource Xenbase, which has created a centralized pipeline for analysis and display of next-gen sequencing data and updates to genome assembly and annotation files³³ (http://www.xenbase.org/).

4 |. MODELING HUMAN DISEASE

The advent of cost-effective patient genome sequencing has been accompanied by an explosion of publicly available patient data for genetic diseases. Exome sequencing of triads and other technological approaches have rapidly identified hundreds of previously-unstudied genes contributing to disease phenotypes ranging from cardiac defects to autism. With this new depth of patient data comes the daunting task of creating animal models to study the mechanism of action of all these genes and the range of protein-coding allelic variants they may comprise. X. tropicalis tadpoles have proven to be an outstanding high-throughput model for this purpose, with considerable advantages in time and cost over mice.^34,35 CRISPR-mediated mutagenesis of candidate genes is rapid and efficient in F₀ animals, and gene products can be independently knocked down by morpholino oligonucleotides, allowing rapid orthogonal approaches to define the requirement for a gene in developmental processes such as autism, heart development, albinism, and craniofacial development.^36–39 A uniquely powerful feature of Xenopus is that these reagents can be delivered to just one blastomere of the two-cell embryo to inhibit gene function only on one half of the embryo, providing a contralateral control and allowing investigation of cell autonomy. Another powerful genetic approach that can be carried out within days in the F₀ animal is to use CRISPR and/or morpholino knockdown to create a genetically depleted background for the factor of interest, followed by a rescue injections with either the wildtype or patient allele of the factor, allowing one to specifically investigate the functionality of patient gene versions without requiring generations of breeding or homologous recombination.

5 |. UNIQUE OPPORTUNITIES FOR COMPARATIVE CELL BIOLOGY

The intersectional use of diploid X. tropicalis and allotetraploid X. laevis provides opportunities for approaches to cell biology and evolutionary biology that are rare, if not unique, among model organisms. X. tropicalis is estimated to have diverged from the X. laevis ancestor ~48 million years ago (mya), while the two ancestral species of X. laevis (the L and S progenitor species) are estimated to have diverged from each other 34 mya.¹ The genome size of X. laevis (3.1 Gb) as well as its karyotype (4n = 36) are approximately double those of X. tropicalis (1.7 Gb and 2n = 20, respectively). At 1 to 1.2 mm X. laevis eggs and embryos are larger than their 0.6 to 0.8 mm X. tropicalis counterparts.⁴⁰ Nevertheless, despite these substantial differences, these two species are morphologically similar enough that they are staged according to the same normal table during embryonic development. The combination of highly differentiated sizes and genomic content, with very strongly similar basic cell and developmental morphology, has formed the basis for some extraordinary insights into cell and organelle scaling. X. tropicalis spindles have been found to be roughly 30% smaller than X. laevis spindles, and spindle size is regulated by the volume and composition of cytoplasm.⁴¹ This discovery set up subsequent insights into mitotic spindle scaling, which is found to be regulated by the kinesin Kif2a, the nuclear import regulator Importin α, and the microtubule polymerase tracker XMap215.^42–44 The size of the nucleus is similarly much larger in X. laevis than X. tropicalis, and is sequentially reduced further during embryonic cleavage stages, enabling the use of these species for investigations of nuclear scaling. The size of the nucleus has been found to also be regulated by Importin α, as well as NTF2,^45,46 and Lamin components.^47,48 Nuclear envelope composition is particularly interesting in Xenopus, as it contributes to regulatory events such as the mid-blastula transition⁴⁹ as well as nuclear morphological features such as nuclear branching.⁵⁰

The genetic and cytological similarities between Xenopus species allow the formation of hybrid embryos by the fertilization of X. laevis eggs with X. tropicalis sperm, presenting a powerful model for interrogating the mechanisms of hybrid speciation and evolution. The gamete contribution of each species has profound effects on embryo viability: crosses made from X. laevis eggs and X. tropicalis sperm (l_e × t_s) are viable, while the reciprocal hybrid (t_e × l_s) die early in embryogenesis prior to gastrulation.^51–53 This hybrid incompatibility in t_e × l_s embryonic development appears to arise from mis-segregation of X. laevis chromosomes in the environment of X. tropicalis cytoplasm, followed by cell death,⁵¹ opening a new field to define the cytological factors responsible, and highlighting the power of these species for exploring mechanisms of hybrid genome formation and evolution.

6 |. WHAT IS NEXT FOR XENOPUS TROPICALIS?

We are fortunate to have adapted a great deal of technical and biological expertise from X. laevis to X. tropicalis to develop this model system. Looking forward, optimization of emerging genetic engineering techniques, proteomics, and genomics techniques and repositories will further enrich the experimental toolkit available. These will facilitate further expansion of the growing community of researchers and biological questions that can be powerfully addressed using this tractable system.