Panarthropods are a superphylum integrated by organisms with bilateral symmetry and a segmented body, divided into Tardigrada, Onychophora, and Arthropoda. The latter has been the most studied so far and is, in turn, grouped into Chelicerata (e.g., spiders and scorpions), Myriapoda (e.g., centipedes and millipedes), and Pancrustacea (e.g., crustaceans and insects) (Giribet and Edgecombe, 2019). The embryonic development of panarthropods is based on the organization of a segmented body plan, which is a shared feature in their evolutionary history (Hannibal and Patel, 2013; Chipman and Edgecombe, 2019). This developmental modality is based on the formation of similar repetitive units called segments along the anteroposterior axis (Clark et al., 2019). The establishment of this segmented patterning has been considered a significant contribution to promoting the great diversity of shapes and sizes found within panarthropods and their high adaptive success in practically all the environments on our planet (Peel et al., 2005; Auman and Chipman, 2017).

The segmented body plan of panarthropods is orchestrated by a wide variety of cellular mechanisms and genetic regulators, which can vary depending on the organism in question.

A widely conserved signaling pathway during embryonic development is the Wnt pathway, which coordinates crucial cellular processes such as proliferation, cell polarity, and the determination of cell fate (Williams and Nagy, 2017; Steinhart and Angers, 2018). Three main different Wnt signaling pathways have been described: The canonical pathway, which is dependent on β-catenin cytoplasm accumulation and subsequent nuclear translocation to the coactivation of specific gene transcription, and the noncanonical Wnt/planar cell polarity (Wnt/PCP) and Wnt/Ca²⁺ pathways, which mediate cytoskeleton dynamics and cell movements through directional information or intracellular Ca²⁺ release, respectively (Croce and McClay, 2008; Angers and Moon, 2009). Although canonical and noncanonical pathways have different signaling mechanisms, both participate in regulating different cellular processes and genetic patterning, allowing successful embryonic development.

The first studies of the Wnt pathway in panarthropods were carried out on the vinegar fly Drosophila melanogaster (Nüsslein-Volhard and Wieschaus, 1980; Baker, 1987; Rijsewijk et al., 1987; Martinez et al., 1988), initiating a wide line of research that linked this pathway to embryonic development, revealing its relationship with various developmental processes. Over the years, more research has been conducted on flies and several vertebrates, including diverse developmental processes and diseases (Jenny and Basler, 2014; Bejsovec, 2018; Wiese et al., 2018), but neglecting the high number of existing arthropods and their different modes of embryogenesis (Murat et al., 2010).

In the last years, exciting advances have been made regarding the molecular and cellular mechanisms involved during sequential segmentation and axial elongation in a variety of arthropods, and the Wnt pathway has emerged as an essential factor in these processes. Some reviews have covered this topic while not focusing specifically on the role of this signaling pathway (Martin and Kimelman, 2009; Williams and Nagy, 2017; Clark et al., 2019). Thus, in a comparative approach, it remains to discuss how and when the Wnt pathway participates in establishing the segmented pattern and regulating processes such as cell proliferation or convergent extension during the posterior body growth in panarthropods. In this review, we will start with a description of the repertoire and expression pattern of Wnt ligands in a vast number of panarthropods, including representatives of all the phyla, addressing their possible roles during germband extension. Then, we will discuss how this signaling pathway could regulate segmentation to establish this repetitive pattern and, at the same time, probably modulate different cellular processes precisely coupled to axial elongation.

The Wnt pathway is a complex signaling pathway present in all metazoans–including placozoans and sponges–that comprises several receptors and intracellular components, as well as an established repertoire of thirteen ligand subfamilies (Holstein, 2012). In vertebrates, it has been determined that Wnt1, Wnt2, Wnt3, Wnt8, and Wnt10 ligands are activators of the canonical pathway, and that Wnt4, Wnt5, Wnt6, Wnt7, Wnt9, and Wnt11 ligands are activators attributed to the noncanonical pathways (Gajos-Michniewicz and Czyz, 2020), while Wnt16 ligand has been shown to be an activator of both, canonical and noncanonical pathways (Gori et al., 2015). WntA ligand, of an indeterminate group, has not been found in vertebrates (Prud’homme et al., 2002), although it is present in urochordates and cephalochordates (Somorjai et al., 2018), as well as in panarthropods (Janssen et al., 2010; Hogvall et al., 2014).

The analysis of the full repertoire of Wnt ligands in panarthropods published so far (Figure 1) showed that a common characteristic is the loss of the Wnt3 ligand, implying that their common ancestor most probably presented ligands from only twelve subfamilies (Hayden and Arthur, 2014; Hogvall et al., 2014; Janssen and Posnien, 2014). A significant loss also occurred in insects, where the absence of Wnt2 and Wnt4 ligands in all the representatives is observed. Interestingly, an ortholog of wnt16 is present in the apid and absent in the other eight insects included in the analysis (Dearden et al., 2006; Bolognesi et al., 2008a; Murat et al., 2010; Shigenobu et al., 2010; Yin et al., 2015; Ding et al., 2019; Holzem et al., 2019; Panfilio et al., 2019; Vosburg et al., 2020). More losses are found in other lineages. However, they were not general to the clade, as they are mixed with other species that retained at least ten Wnt ligands from the ancestral twelve subfamilies (Figure 1). While these losses may depict specific eliminations in particular species within Arthropoda, we cannot exclude that missing ligands could represent low expression transcripts rather than losses at the genomic level when the database used is a transcriptome instead of the whole sequenced genome, as in the cases of Cupiennius salei, Calanus finmarchicus and Thamnocephalus platyurus, among others.

Our updated repertoire also showed that only Wnt4, Wnt5 and Wnt16 ligands were retained in all crustacean lineages–represented here by six species that belong to three of the six extant classes (Zhang, 2011) –, with a low degree of conservation of the Wnt2, Wnt8, and Wnt11 ligands (Constantinou et al., 2016; Jaramillo et al., 2016; Kao et al., 2016; Du et al., 2018).

Regarding chelicerates, they present a low loss of ligands. They also exhibit a high degree of duplications in almost all Wnt ligand subfamilies, which coincides with the whole genome duplication (WGD) that has been proposed to have taken place in the common ancestor of spiders and scorpions (Arachnopulmonata) (Schwager et al., 2017; Panfilio et al., 2019; Harper et al., 2021). Furthermore, an extreme case is found in the horseshoe crab Limulus polyphemus, that exhibit up to six copies of Wnt5 and Wnt7 ligands, most probably due to several rounds of WGD that have been recently documented for this order - Xiphosura - of chelicerates (Nossa et al., 2014; Nong et al., 2021). This leaves out harvestmen, sea spiders, ticks, and mites, represented in our compilation by Phalangium opilio, Phoxichilidium femoratum, Ixodes scapularis, and Tetranychus urticae, respectively, which show no duplications (Figure 1). Interestingly, T. urticae and the decapod crustacean Macrobrachium olfersi both lacks one of the most conserved ligands, Wnt1—only Wnt5 is fully present among panarthropods–and only retained one canonical ligand, Wnt8, and Wnt10, respectively. Thus, despite multiple Wnt ligand losses, there is always a canonical/noncanonical set of ligands in all reported species; however this repertoire is more variable for some ligands than for others. This diversity and variability within arthropods have been explained before by non-conserved specific functional redundancies among Wnt ligands in this and other phyla that are evident after single ligand loss-of-function experiments where no strong defects are found in or related to the site of expression (e.g., Gleason et al., 2006; Bolognesi et al., 2008a; Grigoryan et al., 2008).

Additionally, there are multiple lines of evidence of Wnt ligands signaling by canonical and noncanonical pathways. For example, Drosophila wnt9 (Dwnt4) activates cell motility during ovarian morphogenesis through the noncanonical pathway and salivary gland development through the canonical pathway (Cohen et al., 2002; Harris et al., 2007). In another example, human isolated chondrocytes differentiation was triggered in vitro by wnt3a that simultaneously activated both Wnt/β-catenin and Wnt/Ca²⁺ signaling pathways (Nalesso et al., 2011). All this makes the interchangeability between different ligands or their combinatorial activity to regulate the same pathways or trigger similar functions even more feasible.

On the other hand, wnt genes duplication found almost exclusively in chelicerates has allowed this group to reach up to 14–16 ligands, keeping paralogs of the same Wnt ligand subfamily despite having several other ligands. Some authors have attributed this to an ancestral division of the function between both paralogs and acquisition of a new function by one of them (Gitelman, 2009; Cho et al., 2010). It has also been suggested that larger sets of ligands could be an indication of morphological and functional diversification (Schwager et al., 2017). However, since numerous Wnt ligand subfamilies were found in simpler and basally branching metazoans, for example twelve in the sea anemone Nematostella vectensis, ten in the demosponge Halisarca dujardini and even 21 in the calcisponge Sycon ciliatum, there is no obvious relationship between higher number of ligands and morphological complexity (Kusserow et al., 2005; Borisenko et al., 2016).

To fully understand the wide and varied range of Wnt ligands present in the different panarthropods studied so far, it is necessary to analyze their expression patterns in a comparative manner throughout development. In this review we will focus on the period comprising the processes of axial elongation and segmentation, collectively known as posterior growth.

Published expression patterns of the complete repertoire of Wnt ligands are relatively scarce in panarthropods. However, we found such data for three representative insect species–unfortunately not including hemimetabolous insects–one branchiopod crustacean, two myriapods–a centipede and a millipede,—four chelicerates–including only arachnids: three spiders and one harvestman–and one representative for onychophorans and one for tardigrades (Figures 2,3).

The spatiotemporal analysis of wnt genes expression in the beetle Tribolium castaneum showed that several ligands overlap in the posterior zone, segments/parasegments, and head lobes, suggesting functional redundancy (Bolognesi et al., 2008a). However, when we compiled wnt genes expression patterns including the other panarthropods (Janssen et al., 2010; Hogvall et al., 2014; Janssen and Posnien, 2014; Constantinou et al., 2016; Holzem et al., 2019; Chavarria et al., 2021; Janssen et al., 2021), we found that overlapping is common, although diverse patterns are still observed (Figure 2). Based on the degree of patterning conservation, we were able to classify the expression of Wnt ligands as highly (Wnt1, Wnt5, Wnt6, Wnt7, and Wnt11), moderately (Wnt8, Wnt10, Wnt16, and WntA), and poorly (Wnt2, Wnt4, and Wnt9) conserved, which might suggest that the function of some of these ligands–and not others–would be essential for posterior growth in most panarthropods. Furthermore, early-branching clades, such as onychophorans–a sister group of arthropods –, chelicerates, and myriapods, showed a higher proportion of Wnt ligands in the posterior zone compared to pancrustaceans (Figure 3). Despite all this, it is unclear whether overlapping ligands fulfill similar/redundant or different/complementary functions.

wnt1 is expressed in the posterior region–known as the growth zone or, more precisely, the segment addition zone (SAZ)—of practically all the panarthropods analyzed except for the tardigrades, which do not have a proper SAZ (Chavarria et al., 2021). This suggests a leading and ancestral role of Wnt1 ligand in the regulation of the posterior growth in Panarthropoda.

Within arachnids, two wnt1 paralogs in the tarantula Acanthoscurria geniculata show a clear example of posterior subfunctionalization, where one of them is only expressed on the SAZ while the other is expressed exclusively in the hindgut primordium. Interestingly, in arachnids which have retained only one paralog, wnt1 is expressed at later stages in the putative hindgut of Parasteatoda–without expression in the SAZ–and in the SAZ of Pholcus and Phalangium, with no expression in the hindgut primordium (Janssen et al., 2021).

Specialized functions for wnt1 and wnt5 are suggested by their expression patterns within the posterior region of the Tribolium germband (Nagy and Carroll 1994; Bolognesi et al., 2008a). wnt1 expression is restricted to a specific domain within the SAZ–resembling the wnt8 domain–in a region apparently covered by the broader expression of wnt5. Since both expression domains are not equivalents, they could be covering different functions.

An analysis of Wnt1 and Wnt5 ligands performed in the spider Cupiennius salei showed their complementary expression domains within the segment, dorsal for wnt1 and ventral for wnt5 (Damen, 2002). However, the suggested “complementarity” in Cupiennius was refuted by Janssen et al. (2021) arguing that wnt5 ventral expression is restricted to the nervous system and not related to the segment polarity establishment.

On the other hand, in the onychophoran Euperipatoides kanangrensis, Hogvall et al. (2014) observed that distinct Wnt ligands were expressed at different and specific regions throughout each trunk segment; wnt5 and wnt4 extending over the anterior and posterior domains, respectively, while wnt1, wnt6, wnt9, wnt10, wnt11, and wnt16 covering the center of the segment in adjacent or overlapping regions. However, these onychophoran wnt genes are expressed in a segment polarity-like pattern only at later stages, belatedly to segment polarity genes such as engrailed. Thus, the authors suggested that wnt genes were not involved in forming segmental borders in onychophorans but probably in their intrasegmental patterning.

In summary, a complete wnt genes repertoire is published and available in twelve panarthropods, including representatives of the main subphyla. Furthermore, many wnt genes are expressed in the forming segments and the posterior region during panarthropod elongation, with a general high overlapping among them. However, there is little detailed information on these expressions. Until a better resolution in the expression of various arthropods is available, both at the tissue and cellular level, it will not be possible to determine a reliable degree of functional conservation of the different Wnt ligands.

There is also an imbalance in the analysis of the different wnt genes, mainly concentrated in the posterior expression and function of Wnt1 and Wnt8 ligands. This analysis would be greatly benefited by incorporating the complete set of Wnt ligands in more organisms, including representatives of orders not analyzed so far (e.g., hemimetabolous insects, non-arachnid chelicerates).

One of the main characteristics of panarthropods is their segmented body, the segments of which are added during development in different ways depending on the species. For example, in Drosophila, all segments are patterned almost simultaneously during syncytial blastoderm in what is called long germ segmentation (Peel et al., 2005). On the other hand, most arthropods and onychophorans present a short germ mode of segmentation, where anterior segments are patterned during the blastoderm stage, and the rest of segments are added sequentially from the SAZ, at the posterior end of the embryo/larva (Peel et al., 2005; Clark and Peel, 2018). This territory functions as an organizing center for posterior growth, where the mechanisms that regulate segmentation and axial elongation occur (Williams and Nagy, 2017). As was mentioned previously, the SAZ is characterized by expressing several Wnt ligands dynamically during segmentation, suggesting a role of this pathway in establishing the segmented pattern and the mechanisms controlling body elongation.

Functional studies on different components of the Wnt pathway during sequential segmentation have been performed mostly in the holometabolous insect Tribolium castaneum (Ober and Jockusch, 2006; Bolognesi et al., 2008b; Bolognesi et al., 2009; Beermann et al., 2011) and the spider Parasteatoda tepidariorum (McGregor et al., 2008; Setton and Sharma 2018; Setton and Sharma 2021), although the number of studies in other arthropod species is increasing.

Mainly, it has been observed that the loss of function of Wnt receptors in Tribolium (Tc-frizzled1 and Tc-frizzled2), co-receptor Arrow in Tribolium, Oncopeltus, Gryllus and Parasteatoda (Tc-arrow, Of-arrow, Gb-arrow and Pt-arrow, respectively), or final effectors (the transcription factors Gb-armadillo/β-catenin in Gryllus and Of-pangolin in Oncopeltus) of the canonical Wnt pathway, lead to the formation of truncated embryos with no posterior segments (Figure 4) (Miyawaki et al., 2004; Angelini and Kaufman 2005; Bolognesi et al., 2009; Beermann et al., 2011; Setton and Sharma, 2018; Setton and Sharma, 2021). This makes clear the essential role of this signaling pathway on the posterior growth. The same conclusion could be reached by analyzing the pharmacological inhibition of the Wnt pathway after the incubation of elongating cockroach embryos with the Inhibitor of Wnt Production-3 (IWP-3)—that blocks palmitoylation of Wnt ligands by Porcupine (Chen et al., 2009)—and after Wnt signaling activation using LiCl–that inhibits GSK3 within the β-catenin destruction complex (Hedgepeth et al., 1997)—during the elongation and segmentation of the centipede Strigamia maritima. After the IWP-3 treatment, the absence of Wnt signaling specifically during elongation disrupted segment formation and reduced the size of the SAZ as well as the number of proliferating cells, coinciding with the results showed by Oberhofer et al. (2014) where the Wnt signaling displays a dual role in the SAZ in growth control and posterior patterning. In the latter, the activation of the pathway generated shorter embryos with an expanded SAZ, revealing its role in the control of the segmentation clock (Chesebro et al., 2013; Hayden et al., 2015).

Regarding Wnt ligands, by far the most studied is wnt1/wingless, but its functional analysis has only been successful in some insects, with no loss-of-function phenotypes found in the cricket G. bimaculatus and the spider P. tepidariorum (Miyawaki et al., 2004; Setton and Sharma, 2021). Apparently, the severity of wnt1 knockdown related to posterior growth could be linked to the presence/absence of another ligand, wnt8: In Tribolium and Oncopeltus–where Wnt8 ligand is present as one and two paralogs, respectively–wnt1 RNAi only showed segmental boundary disruption without preventing elongation or the generation of the segmented pattern, as was revealed by the absence of truncated phenotypes and the regular expression of the pair-rule gene Tc-even-skipped (Tc-eve) in the beetle and a normal-sized abdomen in the case of Oncopeltus embryos (Angelini and Kaufman, 2005; Ober and Jockusch, 2006; Bolognesi et al., 2008b). On the other hand, in the domestic silk moth Bombyx mori, which do not have a wnt8 paralog, wnt1 RNAi produced a range of phenotypes from partially truncated embryos displaying immature legs (Yamaguchi et al., 2011) to drastically shortened embryos showing no signs of abdominal nor thoracic segments (Nakao, 2018). The idea of a (partial) functional interrelationship between both ligands is also based on the results obtained in Tribolium after eliminating wnt8 alone, that generated a small proportion of truncated embryos, and the drastic increasing of the truncated phenotype frequency using the wnt1 and wnt8 double RNAi (Bolognesi et al., 2008b). There are other cases in which the same different results have been reported: In the cockroach Periplaneta americana, the loss-of-function of wnt1 led embryos to develop posteriorly truncated germbands (Chesebro et al., 2013), while in the lepidopteran Dendrolimus punctatus, wnt1 knockout using CRISPR/Cas9 showed abdominal segment fusion with no signs of posterior truncation (Liu et al., 2017). However, it is not known if Wnt8 ligand is present in these two insects. Finally, wnt8 knockdown in the spider Parasteatoda (McGregor et al., 2008) caused posterior truncation or even complete absence of the opisthosomal (abdominal) region, a phenotype that resembles the effect of the RNAi against wnt8 in Tribolium (Bolognesi et al., 2008b). However, given the functional redundancy between Wnt1 and Wnt8 ligands, we cannot interpret wnt1 or wnt8 knockdown phenotypes in any species where wnt1/wnt8 double loss-of-function has not been analyzed.

The analysis of the phenotypes mentioned above reveals the Wnt pathway’s different roles during posterior growth. On the one hand, truncated phenotypes indicate early participation in this process, although we have little evidence about the mechanisms in which it would operate. In turn, we do not have sufficient evidence to relate the Wnt signaling to maintaining gene oscillations or establishing a wavefront at the anterior SAZ, as it is well known in vertebrates (Oates et al., 2012). On the other hand, the participation of wnt1 in the definition of segment borders is relatively conserved among arthropods and is a well-understood process at the molecular and cellular level in the Drosophila fly (Nüsslein-Volhard and Wieschaus, 1980; Nagy and Carroll 1994; Ober and Jockusch, 2006).

To date, other canonical and noncanonical Wnt ligands that are expressed segmentally and/or in the posterior region have only been functionally studied in the beetle Tribolium. When Bolognesi et al. (2008b) RNAi screened all nine Wnt ligands expressed in Tribolium, they only found segmental phenotypes with Tc-wnt1 and Tc-wnt8. The same result was obtained after checking different combinations of double (all combinations of Tc-wnt1, Tc-wnt5, Tc-wnt8 and Tc-wntA) and triple (Tc-wnt5, Tc-wnt8 and Tc-wntA) RNAi injections of the wnt genes that are most clearly expressed in the SAZ (Bolognesi et al., 2008a, b). However, we cannot rule out redundant functions during posterior growth for wnt5 and wntA with wnt6, which is also expressed in the SAZ but was not included in their double and triple knockdown analyses.

It is also possible that in other panarthropods, different wnt genes fulfill the absence of specific Wnt ligands, like Wnt8 in Thamnocephalus (Constantinou et al., 2016), Strigamia (Hayden et al., 2015) or the onychophoran Euperipatoides (Hogvall et al., 2014). The evidence suggests redundant and partially redundant functions between arthropod Wnt ligands. Nevertheless, the relationship between a Wnt ligand and its precise function and which ligands share the same function is still unclear.

It has been established that segmentation in most arthropods and all vertebrates is driven by a cyclical mechanism where the temporal periodicity of a clock is translated into a repetitive spatial pattern (Cooke and Zeeman, 1976; Palmeirim et al., 1997; Sarrazin et al., 2012). In vertebrates, cell-autonomous oscillatory genes expression is maintained by a posterior Wnt + Fgf signaling gradient. As this gradient decreases, the generated posterior-to-anterior kinematic waves of expression are arrested, allowing the formation of a new segment (Oates et al., 2012). Furthermore, the progressive posterior elongation of the embryo moves the Wnt + Fgf gradient together with the determination front, thus linking posterior growth with segmentation. In addition, several genes related to the Wnt, Fgf, and Notch/Delta pathways oscillate and are coupled between them, which is crucial for segment generation (Dequéant et al., 2006; Krol et al., 2011; Sonnen et al., 2018). However, despite the evidence involving the Wnt signaling pathway in the vertebrate segmentation clock, its specific role is not yet fully understood.

In arthropods, caudal, a highly conserved gene that is expressed at the posterior half of the SAZ and is crucial for its establishment and maintenance (Dearden and Akam, 2001; Chipman et al., 2004; Copf et al., 2004; Shinmyo et al., 2005; Chesebro et al., 2013; Janssen et al., 2015; Schönauer et al., 2016; Novikova et al., 2020), is regulated by the canonical Wnt signaling (Shinmyo et al., 2005; McGregor et al., 2008; Chesebro et al., 2013; Oberhofer et al., 2014; Schönauer et al., 2016). Recently, Clark and Peel (2018) proposed that in the SAZ, Wnt signaling would be regulating an ancient regulatory network formed by caudal, dichaete, and odd-paired (opa), which are sequentially expressed–both temporally and spatially–during segmentation in Tribolium, but also in the long-germ Drosophila embryo. Furthermore, one of the same authors proposed a transition from the posterior expression of caudal/dichaete to dichaete/opa expression at the anterior region of the SAZ, resembling the wavefront limit operating in vertebrates (Clark, 2021). Then it is possible that the Wnt pathway is regulating, through Caudal, genetic oscillations in arthropods, as occurs in vertebrates.

Two main kinds of cyclically expressed genes have been observed in arthropods (no dynamically expressed genes have been found in onychophorans; Janssen and Budd, 2013; Janssen and Budd, 2016): On one hand, the orthologs of Drosophila pair-rule genes as odd-skipped (Chipman et al., 2004; Sarrazin et al., 2012), even-skipped (Chipman and Akam, 2008; El-Sherif et al., 2012; Brena and Akam, 2013), paired (Eriksson et al., 2013), hairy (Chipman and Akam, 2008; Pueyo et al., 2008) and most probably runt (Schönauer et al., 2016). On the other hand, members of the Notch signaling pathway, mainly the Delta ligand, have been shown to oscillate in cockroaches (Pueyo et al., 2008), centipedes (Chipman and Akam, 2008; Brena and Akam, 2013) and spiders (Stollewerk et al., 2003; Schönauer et al., 2016). Functional analyses in a few of these organisms appear to position Wnt signaling upstream of some cyclic genes.

In Tribolium, it was already shown that blocking Wnt signaling impairs the posterior expression of the three pair-rule genes (Tc-run, Tc-eve, and Tc-odd; Bolognesi et al., 2009; Beermann et al., 2011), which are part of the regulatory circuit involved in the segmentation clock (Choe et al., 2006). In the same insect, RNA-seq comparative analysis revealed additional pair-rule genes downregulated after knocking down the Wnt/β-catenin pathway, such as Tc-hairy and Tc-dichaete, in addition to Tc-odd and Tc-eve (Oberhofer et al., 2014). However, the same experimental approach performed in the spider Parasteatoda showed no significant effect on the expression of the pair-rule genes Pt-eve, Pt-run, and Pt-hairy, with a mild effect on Pt-dichaete (Setton and Sharma, 2021). In this study, it was even seen that caudal expression was not interrupted in Parasteatoda and Gryllus after arrow knockdown, unlike in Oncopeltus, where they obtained a substantial reduction in Of-caudal expression. The discrepancies observed between Tribolium and Parasteatoda Wnt/β-catenin pathway loss-of-function results could be associated with the weak penetrance of the RNAi injection (around 50% of effect; Setton and Sharma, 2021) or with a possible low conservation of the wnt/caudal/pair-rule gene network among arthropods. In any case, if we consider that in the spider Parasteatoda wnt8 was previously shown to be necessary for the posterior expression of Pt-eve and Pt-run (Schönauer et al., 2016) and Pt-caudal (McGregor et al., 2008; Schönauer et al., 2016), one alternative explanation could be that specifically in the spider, Wnt8 function acts through an Arrow-independent pathway, that is like a noncanonical ligand.

Regarding Wnt-Notch interactions, on the other hand, Chesebro et al. (2013) showed in the cockroach that Pa-delta expression in the posterior SAZ is completely eliminated after Pa-wnt1 knockdown. Furthermore, they obtained similar results using the pharmacological inhibitor IWP-3 (Chesebro et al., 2013). At the same time, they observed the disruption of posterior Pa-wnt1 expression after injecting RNAi against Pa-notch and its partial reduction by incubating embryos with the Notch inhibitor, DAPT.

Schönauer et al. (2016), in turn, situated this signaling pathway upstream Pt-wnt8 in Parasteatoda. Depending on the SAZ region, Notch/Delta is activating (at the posterior SAZ) or inhibiting (at the anterior SAZ) this ligand. However, it was previously found that Wnt8 would be necessary for the dynamic expression of Pt-delta in the same spider (McGregor et al., 2008). In addition, the RNA-seq study by Setton and Sharma (2021) exhibited Pt-notch ″downregulation-but not Pt delta inhibition-after″ Wnt signaling loss-of-function, although they did not check the effect of knocking down Notch signaling. So then, Notch and Wnt signaling pathways seem to display a reciprocal regulation in these two animal models, the hemimetabolous cockroach Periplaneta and the chelicerate Parasteatoda, apparently partially through their regulation of caudal. Nevertheless, our limited knowledge about the complexity of the SAZ in terms of cell types and tissue organization makes it hard to understand reciprocal interactions between different pathways.

As it has been suggested that the Notch pathway does not present a functional role in the segmentation process in the onychophoran Euperipatoides kanangrensis and the holometabolous insects Drosophila melanogaster, Apis mellifera and Tribolium castaneum (Aranda et al., 2008; Wilson et al., 2010; Kux et al., 2013; Janssen and Budd, 2016), the regulatory role of the Wnt pathway on the Notch signaling is limited, considering the existing evidence, to Parasteatoda tepidariorum, Periplaneta americana and the short-germ silkworm Bombyx mori (Liu, 2013). Thus, from the cases mentioned so far, we can conclude the following (Figure 5):

1) Studies in different arthropods are lacking to determine a possible ancestral regulatory route of posterior growth.

Sequential segment addition in most arthropods is accompanied by simultaneous germband extension along the anterior-posterior axis. This complex reorganization of the developing body must include, in a species-specific manner, a variable, mutually dependent, and tightly regulated arrangement of cellular behaviors, including oriented (or not) mitotic divisions, cell rearrangements, and changes in cellular density, cell shape, and cell size, among others.

Little is known about the regulatory network behind axial elongation and the mentioned cellular processes in arthropod posterior growth. In deuterostomes, including vertebrates as well as the hemichordates that display posterior growth without segmentation, paraxial mesoderm progenitors are maintained by a posterior positive regulatory loop between wnt (wnt3a and wnt8 in zebrafish) and brachyury, so that posterior elongation depends on this molecular feedback (Martin and Kimelman, 2008; Fritzenwanker et al., 2019). However, brachyury orthologs in insects are expressed in the SAZ but seem not necessary for posterior growth, as was shown by RNAi analysis in Tribolium and Gryllus (Shinmyo et al., 2006; Berns et al., 2008). Regardless of whether Wnt-Brachyury regulation was lost in insects or corresponds to a deuterostome evolutionary innovation, which factor is maintaining Wnt signaling at the insect SAZ is unknown.

Few direct evidence links Wnt signaling to cell proliferation during arthropod posterior growth; for example, Oberhofer et al. (2014), through genetic analyses in Tribolium SAZ, showed that the Wnt pathway targets some genes involved in mitotic cell cycle and spindle organization/elongation. Moreover, RNAi treatments against wnt1 caused a decrease in cell proliferation in the cockroach SAZ, the same occurring when the pathway was inhibited by pharmacological treatment with IWP-3 (Chesebro et al., 2013). Constantinou et al. (2020), based on the cell division pattern shown by the branchiopod Thamnocephalus, divided the SAZ into an anterior no-proliferating region that expresses WntA ligand and a posterior slightly-but-constant proliferating region that expresses Wnt4 ligand. A similar correlation was found in Oncopeltus (Auman et al., 2017), where the region of cell divisions coincides with the expression of Of-eve and Of-caudal genes, two targets of Wnt signaling. The proposed model differs, however, from what was found in Dermestes (Xiang et al., 2017) and Tribolium (Cepeda et al., 2021). The proliferation region is not temporally stable in these beetles, but variable and posteriorly concentrated at specific stages. Unfortunately, no correlation has yet been found between this proliferation pattern and the expression of any gene at the SAZ, and there is evidence that the Wnt signaling target genes Tc-caudal, and Tc-eve are not involved in cell division regulation in this insect (Copf et al., 2004; Nakamoto et al., 2015). A possible mechanism that explains this patterning type would be what happens in the Drosophila wing, where cell proliferation is regulated by differences in the amount of Wnt1 ligand; high concentrations generate a decrease while intermediate concentrations trigger proliferation (Baena-Lopez et al., 2009). However, no changes in the expression of Wnt ligands correlating with this model have been demonstrated.

Another well-conserved cellular process among arthropods during germband elongation is the convergent extension (CE). The first studies on CE were carried out in Drosophila, where the pair-rule genes Dm-eve and Dm-runt are necessary for the polarized cell movements involved (Irvine and Wieschaus, 1994; Zallen and Wieschaus, 2004). Furthermore, the necessity of Tc-eve and Tc-caudal during the intercalary cell movements seen at early stages of germband elongation in Tribolium was also reported (Benton et al., 2013; Nakamoto et al., 2015). In addition, the fact that Dm-eve regulates the expression of the Toll family receptors, which direct CE during germband elongation in Drosophila (Paré et al., 2014) and is conserved in Tribolium and Parasteatoda (Benton et al., 2016), leads us to wonder if the Wnt pathway would have a role in this process. The Wnt signaling pathway could be involved in CE through its regulation of eve expression, which we have already seen occur in various arthropods during germband elongation (Bolognesi et al., 2009; Beermann et al., 2011; Oberhofer et al., 2014; Schönauer et al., 2016).

However, we will not have an answer to our question until a more significant number of studies have been carried out with a particular focus on the cellular behaviors that take place during posterior elongation in a diverse group of arthropods.

After reviewing the Wnt pathway’s role in axial elongation and segmentation processes in arthropods, more questions than answers arise.

We found the complete Wnt ligand repertoire from 42 species of panarthropods. However, the expression pattern of all the ligands has been published only in twelve. Thus, a more diverse palette of species is lacking in the analysis, especially non/underrepresented lineages (e.g., maxillopod crustaceans, non-insect hexapods, non-spider chelicerates). Furthermore, given the apparent overlapping between some ligands, it would be precious to have a detailed cellular-level analysis of the expression patterns, hopefully including double in situ hybridizations.

The incorporation of functional studies in more organisms, including non-model panarthropods with interesting phylogenetic positions as early-branching clades (e.g., onychophorans, ametabolous and hemimetabolous insects), will provide the basis for well-supported comparative analyses. In the long term, this will allow us to identify conserved and diverged aspects of the regulatory role of the Wnt pathway on posterior body elongation and segmentation. Furthermore, adding imaging tools to this analysis, like tissue-specific transgenic lines carrying fluorescent reporters for time-lapse imaging, will be essential to understanding the contribution of Wnt signaling in the highly dynamic cellular and molecular network that organizes posterior growth in most panarthropods.