(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .
A taxon-restricted duplicate of Iroquois3 is required for patterning the spider waist [1]
['Emily V. W. Setton', 'Department Of Integrative Biology', 'University Of Wisconsin-Madison', 'Madison', 'Wisconsin', 'United States Of America', 'Jesús A. Ballesteros', 'Department Of Biology', 'Kean University', 'Union']
Date: 2024-09
The chelicerate body plan is distinguished from other arthropod groups by its division of segments into 2 tagmata: the anterior prosoma (“cephalothorax”) and the posterior opisthosoma (“abdomen”). Little is understood about the genetic mechanisms that establish the prosomal-opisthosomal (PO) boundary. To discover these mechanisms, we created high-quality genomic resources for the large-bodied spider Aphonopelma hentzi. We sequenced specific territories along the antero-posterior axis of developing embryos and applied differential gene expression analyses to identify putative regulators of regional identity. After bioinformatic screening for candidate genes that were consistently highly expressed in only 1 tagma (either the prosoma or the opisthosoma), we validated the function of highly ranked candidates in the tractable spider model Parasteatoda tepidariorum. Here, we show that an arthropod homolog of the Iroquois complex of homeobox genes is required for proper formation of the boundary between arachnid tagmata. The function of this homolog had not been previously characterized, because it was lost in the common ancestor of Pancrustacea, precluding its investigation in well-studied insect model organisms. Knockdown of the spider copy of this gene, which we designate as waist-less, in P. tepidariorum resulted in embryos with defects in the PO boundary, incurring discontinuous spider germ bands. We show that waist-less is required for proper specification of the segments that span the prosoma-opisthosoma boundary, which in adult spiders corresponds to the narrowed pedicel. Our results demonstrate the requirement of an ancient, taxon-restricted paralog for the establishment of the tagmatic boundary that defines Chelicerata.
Funding: This work was supported by the National Science Foundation (IOS-1552610 and IOS-2016141 to PPS) ( nsf.gov ). Additional support to EVWS came from The National Science Foundation Graduate Research Fellowship (DGE-1747503 to EVWS) ( nsf.gov ). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Data Availability: All image data are all contained within the paper and/or Supporting Information files. Outputs of analyses and all reagents used are available in the supporting information files. Raw sequence data for the tarantula transcriptomes used in this work are available via NCBI SRA (under PRJNA1105064). All other resources used have been previously published (i.e. genomes, transcriptomes, protocols).
Copyright: © 2024 Setton et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
To advance the understanding of chelicerate body plan patterning and address possible roles for retained paralogs in chelicerate tagmosis, we generated transcriptional profiles of prosomal and opisthosomal tissues of a large-bodied spider (a tarantula), across developmental stages pertinent to posterior patterning. We applied differential gene expression (DGE) analyses to identify taxon-specific gene duplicates that were differentially expressed across the prosomal-opisthosomal (PO) boundary and screened candidates using an RNA interference (RNAi) gene silencing approach. Through this approach, we were able to identify one of the 5 spider homologs of Iroquois (Iroquois4 sensu [ 25 ]; Iroquois3-2, sensu [ 26 ]) as playing a role in patterning the segments spanning the PO boundary. Our results provide a functional link between an unexplored gene copy restricted to non-pancrustacean arthropods and the boundary between the tagmata of chelicerates.
Comparatively little is known about how these functional groups of segments are established in chelicerates, by comparison to their insect counterparts. Due to the phylogenetic distance between hexapods and chelicerates, homologs of insect candidate genes that play a role in tagmosis can exhibit dissimilar expression patterns or incomparable phenotypic spectra in gene silencing experiments in spiders, a group that includes the leading models for study of chelicerate development [ 13 , 14 , 16 – 18 ]. A further complication is the incidence of waves of whole genome duplications (WGDs) in certain subsets of chelicerate orders, such as Arachnopulmonata, a group of 6 chelicerate orders that includes spiders [ 19 – 22 ]. The retention of numerous paralogous copies that diverged prior to the Silurian represents fertile ground for understanding evolution after gene duplication but also presents the potential barrier of functional redundancy or replacement between gene copies. Accordingly, there are few functional datasets supporting a role for lineage-specific gene duplicates in the patterning of arachnid body plans [ 23 , 24 ].
These limits are accentuated in Chelicerata (e.g., spiders, scorpions, mites, horseshoe crabs), the sister group to the remaining arthropods. The bauplan of most chelicerates consists of 2 tagmata, the anterior prosoma (which bears the eyes, mouthparts, and walking legs) and the posterior opisthosoma (the analog of the insect abdomen). Even at this basic level of body plan organization, differences in architectures are markedly evident between chelicerates and the better-studied hexapods. The chelicerate prosoma typically has 7 segments and includes all mouthparts and walking legs, whereas the insect head has 6 segments and bears only the sensory (antenna) and gnathal appendages (mandible, maxilla, labium); locomotory appendages of insects occur on a separate tagma, the thorax [ 15 ].
Functional understanding of the evolution of animal body plans is frequently constrained by 2 bottlenecks. First, developmental genetic datasets and functional toolkits are often asymmetrically weighted in favor of lineages that harbor model organisms, to the detriment of phylogenetically significant non-model groups. Second, models of ontogenetic processes that are grounded in model systems vary in their explanatory power across diverse taxa, both as a function of phylogenetic distance, as well as the evolutionary lability of different gene regulatory networks (GRNs) [ 1 – 4 ]. In Arthropoda, understanding of morphogenesis, as well as the evolutionary dynamics of underlying GRNs, is largely grounded in hexapod models and, particularly, holometabolous insects. Candidate gene approaches derived from studies of insect developmental genetics have thus played an outsized role in understanding the mechanisms of arthropod evolution, with emphasis on processes like segmentation, limb axis patterning, and neurogenesis [ 5 – 10 ]. However, the candidate gene framework has its limits in investigations of taxon-specific structures (e.g., spider spinnerets, sea spider ovigers) [ 11 – 13 ], or when homologous genes or processes do not occur in non-model taxa (e.g., bicoid in head segmentation) [ 7 , 14 ].
Results
Differential gene expression, RNAi screen, and identification of waist-less To understand the genetic basis of posterior patterning in spiders, we aimed to generate tissue-specific transcriptomes of spider embryos. The leading model system for spider development, Parasteatoda tepidariorum, proved challenging in this regard, due to the small size of its embryos (500 μm) and the high internal pressure of the egg. We therefore generated DGE datasets for the tarantula Aphonopelma hentzi, which features large and synchronous broods, and embryos with large diameter (2.4 mm) and low internal pressure [27]. We dissected clutches of synchronously developing tarantula embryos and generated RNA-seq libraries for the labrum, chelicera, pedipalp, walking leg, book lung, anterior spinneret, and posterior spinneret. Dissected tissues comprised the whole appendage and attached section of body wall. This protocol was performed for 3 developmental stages, encompassing establishment and differentiation of posterior appendages (e.g., book lungs and spinnerets) [27]. DGE analysis identified 5,429 to 14,094 genes (stage 9: 7,609; stage 10: 5,429; stage 11: 14,094) as consistently differentially expressed across segments in an all-versus-all comparison (p ≤ 0.05; LFC ≥ 1) (Fig 1A). To identify genes that may play an important role in posterior patterning, we assessed the top 100 most differentially expressed genes in all-by-all comparisons for each developmental stage and screened candidates that were consistently highly expressed in only the prosomal or opisthosomal segments in at least 2 stages (stage 9: 67; stage 10: 53; stage 11: 92) (S1 Fig). We prioritized 12 genes for functional screening based on their expression profiles (S1 Table and S2 and S3 Figs). Candidate selection emphasized transcription factors (n = 8/12), with the remaining candidates comprising a secreted protein (spaetzle), a nucleotidyltransferase (Mab21-1), and 2 genes of unknown function (Ahen-TRINITY_DN6222_c0_g1 and Ahen-TRINITY_DN3695_c1_g1). PPT PowerPoint slide
PNG larger image
TIFF original image Download: Fig 1. Overview of RNA-seq design, candidate gene identification, and the ortholog identification within the Iroquois gene family. (A) Tissues from regions representing major morphological characters along the antero-posterior (AP) axis were dissected from developing Aphonopelma hentzi embryos for mRNA sequencing. DGE analysis of RNA-seq libraries generated region-specific profiles to enable the identification of genes both lowly expressed in the prosoma (blue box) and highly expressed in the opisthosoma (purple box). Arrowhead indicates the ortholog of spider waist-less. Heatmap is based upon stage 10 embryos; all transcriptional profiles are provided in S1 Fig. (B-D) Expression of waist-less in limb bud stage embryos of Parasteatoda tepidariorum, counterstained for Hoechst. Note the higher expression level in the opisthosoma compared to the prosoma. (E) Maximum likelihood gene tree of Iroquois2/3 homologs of Panarthropoda, rooted on Onychophora. Colored branches correspond to different orthologs, following F. Boldface text indicates spider waist-less orthologs. Inset: Full unrooted gene tree of Iroquois homologs. (F) Inferred evolutionary history of Iroquois gene duplications in Chelicerata. Scale bar: 100 μm. Complete dataset for heatmap in panel A is provided in S3 Data.
https://doi.org/10.1371/journal.pbio.3002771.g001 Due to the lack of gene silencing tools in the tarantula, we performed functional screening of candidate genes in the house spider Parasteatoda tepidariorum (S2 Table), following established protocols [17,28–30]. Of the 12 candidates, 10 yielded no discernable phenotype, paralleling outcomes of recent RNAi screens in this system (S1 Table) [13]. Initial identification of phenotypes from RNAi screens was performed by visual assessment of morphology at developmental stages 8 to 13, following the established P. tepidariorum staging system [31], with embryos sampled from viable egg clutches 2 to 5 (until females stopped laying). The broad range of stages was chosen to encompass establishment of the PO boundary, addition of opisthosomal segments, and limb bud development in both tagmata. Within this developmental window, we looked for morphological defects in the limb buds, the segments spanning the PO boundary, and the opisthosomal tissues. A subset of each cocoon laid during the RNAi screen was raised to at least the first instar to assess the incidence of postembryonic defects. Genes highly expressed in the prosoma were included in the initial screen toward possible identification of transcripts responsible for repressing opisthosomal identity. However, none of the prosoma-biased genes resulted in a phenotype and were not further pursued (S1 Table and S2 Fig). The 2 RNAi experiments resulting in phenotypes both targeted genes predicted to be enriched in the opisthosoma. One candidate that generated RNAi phenotypes in a small number of embryos was annotated as a GATA transcription factor. RNAi against the GATA homolog (pannier2) resulted in opisthosomal defects with low phenotypic penetrance, precluding analysis of large numbers of affected embryos (described below). The second candidate was annotated as a member of the Iroquois complex of homeobox genes (Fig 1B–1E). Previously identified as “Iroquois4” in a recent survey of homeobox family duplications [25], this transcription factor is not orthologous to the identically named vertebrate homolog Iroquois4 nor is its homology to its 2 insect homologs (mirror and arucan/caupolican) understood [32]. To forfend redundancy of nomenclature within the chelicerate Iroquois complex, we rename the differentially expressed spider copy (previously, “Iroquois4”) waist-less (wsls), reflecting the phenotypic spectrum described below. Due to the higher penetrance of RNAi against waist-less and the ensuing ability to interrogate patterning of the PO boundary, this gene became the focal point of the study.
Evolutionary history of panarthropod Iroquois homologs To better understand the evolutionary history of this gene in arthropods, we inferred a gene tree of the Iroquois family, surveying genomes and developmental transcriptomes of 4 arachnopulmonates (arachnids that share a WGD; 2 spiders, a whip spider, and a scorpion), 6 non-arachnopulmonate chelicerates (chelicerates with an unduplicated genomes; 5 sea spiders and a harvestman), 4 myriapods (sister group to chelicerates with unduplicated genomes; 2 centipedes, 2 millipedes), 3 crustaceans, and 12 hexapods. The gene tree topology (Figs 1E and S4) recovered Iroquois1, Iroquois2, and Iroquois3 homologs as 3 separate clusters, with maximal nodal support for Iroquois3. Whereas representatives of all major arthropod lineages bore Iroquois1 and Iroquois2 homologs, the cluster corresponding to Iroquois3 was comprised only of myriapod and chelicerate taxa (Fig 1E). To polarize the evolutionary history of the Iroquois complex, we examined the organization of Iroquois homologs in well-annotated genomes of Panarthropoda (Fig 1F and S3 Table). Whereas a single Iroquois homolog occurs in high-quality genomes of Tardigrada and Onychophora, chromosomal-level genomes of Myriapoda and apulmonate Chelicerata exhibited 3 Iroquois homologs arranged contiguously on single scaffolds, consistent with an origin of the arthropod Iroquois genes via 2 tandem duplications. Chromosomal-level genomes of spiders recovered 5 to 6 Iroquois copies, with homologs of Iroquois1, Iroquois2, and Iroquois3 occurring on 2 separate scaffolds, consistent with WGD in the arachnopulmonate common ancestor. The ancestral arrangement of the 3 Iroquois homologs was observed to be reordered in 1 cluster in the spider Dysdera sylvatica (see also [26]). In support of this result, non-arachnopulmonate chelicerates (e.g., the harvestman, sea spiders) bore 3 Iroquois homologs in the gene tree (1 homolog of mirror, 1 homolog of araucan/caupolican, and 1 homolog of Iroquois3), whereas spiders and scorpions bore up to 6 Iroquois homologs due to an arachnopulmonate-specific WGD. P. tepidariorum bore only 5 Iroquois homologs due to the loss of 1 mirror copy (S4 Fig). The absence of Iroquois3 in all sampled representatives of hexapods and crustaceans is consistent with a loss of this gene in the branch subtending Pancrustacea. Additionally, the duplication and subdivision of Iroquois2 into araucan and caupolican is limited to a subset of flies (e.g., Drosophila melanogaster), not all Diptera (e.g., Calliphora vicina, Anopheles gambiae, Culex pipiens quinquefasciatus) (Fig 1F).
Expression of waist-less in Parasteatoda tepidariorum Expression of spider waist-less (formerly “Irx4”, sensu [25]) was previously reported for selected stages of development, and a segmentation function had been suggested due to the segmentally reiterated stripes of expression [25,26,33]. We first surveyed waist-less expression across the embryogenesis of P. tepidariorum. The earliest expression was detected at stage 5 as a ring around the germ disc and accords with previously reported expression for stage 5 [34] (arrowhead in S5A Fig). Dynamic expression was recovered at stage 6 (dorsal field stage), corresponding to a stripe of expression in the outer margin of the embryo, in addition to a separate domain of expression in the presumptive growth zone (arrowheads in S5B and S5C Fig). Subsequent stages exhibited additional stripes generated at the posterior terminus, corresponding to presumptive segments of the prosoma (S5D–S5F Fig). In the transition from the dorsal field stage to the germ band stage (stage 7), expression of waist-less decreased in the growth zone and the strongest expression domains corresponded to the segmentally iterated stripes of the prosoma (S5G and S5H Fig). At stages 8 and 9, expression is notably stronger in opisthosomal segments compared to prosomal segments, due to the incidence of weaker domains bridging the segmentally iterated stripes of waist-less in the opisthosoma (Figs 1B–1D and S5I–S5N) and corroborating the stronger posterior expression predicted by DGE (S3 Fig). Additional expression domains include a “V” shape in the anterior head beginning at stage 8.2 (S5I–S5K Fig). At stage 9.2, expression appears in the lateral body wall, together with a distinct, distal point of expression in the prosomal appendages. At this stage, the “V” of expression in the anterior head becomes a pair of arcs, curved inward toward each other approaching the ventral midline and comprising the medial head region that lies anterior to the cheliceral limb buds (S5L–S5N Fig). At stage 10.1, the crescents of expression on each side of the developing head become more concentrated. The segmentally repeated stripes of expression are still maintained ventrally but with the stripes no longer of uniform strength across the germ band. At this stage, increased expression is seen in the opisthosomal appendages (S5O–S5Q Fig). Gene expression at stage 10.2 expression is similar to 10.1 but with increased localization to the lateral margins and appendage primordia of the opisthosoma (S5R–S5T Fig). Stage 11 embryos exhibit increased division of the stripes across the width of the germ band and continued strong expression in the lateral part of the opisthosoma (S5U–S5W Fig).
Expression of pannier2 in Parasteatoda tepidariorum The other RNAi experiment that resulted in a morphological phenotype was a GATA family transcription factor. Gene orthology was inferred using a gene tree of GATA sequences. Three spider GATA genes were identified as members of the pannier clade, with the highly expressed copy provisionally identified as Ptep-pnr2 (S15 Fig). We surveyed pnr2 in wild-type embryos at the developmental stages encompassed by the RNA-seq dataset. At all stages, surveyed Ptep-pnr2 was expressed in the lateral-most territory of the opisthosoma, which corresponds to the dorsal midline upon dorsal closure (S16 Fig). Separate expression domains were recovered in the dorso-lateral margins of the head lobes at stages 9 and 10 (S16A–S16B’ Fig). At stage 11, and the initiation of dorsal closure, expression is found throughout the lateral body wall with enrichment and expansion in the dorsally migrating opisthosomal territory (S16C and S16C’ Fig).
[END]
---
[1] Url:
https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3002771
Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.
via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/
