Introduction

GSK3 (glycogen synthase kinase 3) is a component of the genetic pathways widely related to the anteroposterior (AP) axis specification during embryonic and post-embryonic development in Bilateria, such as the Wnt/β-catenin and hedgehog signaling pathways (Croce and McClay 2006; Kim and Kimmel 2006). The Wnt/β-catenin signaling pathway has been addressed to several developmental processes like morphogenesis, organogenesis, homeostasis, and regeneration (Croce and McClay 2006; Lee et al. 2006; Lengfeld et al. 2009). Studies in acoels (Sikes and Bely 2010) and annelids (Demilly et al. 2013; Chen et al. 2020) used successful procedures to disrupt the Wnt/β-catenin signaling pathway by using, for example, pharmacological inhibitors of GSK3. 1-Azakenpaullone (AZP) is one of the most known pharmacological inhibitors used in experiments. It is a selective inhibitor of GSK3β and, by stimulating the transcription of β-catenin-related genes, can be considered a positive regulator of the Wnt/β-catenin signaling pathway (Kunick et al. 2004; Mfopou et al. 2014).

The Wnt/β-catenin signaling pathway plays important roles during embryogenesis, larval development in the annelid Platynereis dumerilii (Audouin and Milne Edwards 1833) (Schneider and Bowerman 2007; Demilly et al. 2013; Žídek et al. 2018), and during regeneration in the annelid Aeolosoma viride Stephenson 1911 (Chen et al. 2020). In A. viride, the inhibition of GSK3β showed an increase in nuclear β-catenin during regeneration, which affected cellular proliferation, wound healing, and blastema development (Chen et al. 2020). Interestingly, data on other annelids revealed that some genes associated to the Wnt/ β-catenin pathway are expressed during regeneration. For example, β-catenin is highly expressed in the blastema of Pristina leidyi Smith 1896, asvisualized by in situ hybridization (Smith 1896; Nyberg et al. 2012) and is up-regulated during the anterior regeneration of Syllis gracilis Grube 1840, as shown by transcriptomic analysis (Ribeiro et al. 2019).

Disruption of the Wnt/β-catenin pathway has resulted in heteromorphic regeneration of multiple heads or tails in planarians (Gurley et al. 2008; Petersen and Reddien 2008) and loss of polarity in acoels (Sikes and Bely 2010). Body AP heteromorphs, such as bifurcated and bipolar heads and tails, can also arise in annelids, another group of animals with extensive regeneration abilities (Andrews 1892; Hyman 1916; Myohara et al. 1999; Kawamoto et al. 2005). Reports of heteromorphies in annelids of the family Syllidae are very intriguing. There are records of bifurcated heads (Langerhans 1881) and tails (Andrews 1891, 1892), bipolar tails (Okada 1929), and even bifurcation of internal systems, such as a double proventricle (a muscular structure of the anterior foregut) (Ribeiro et al. 2020). However, little is known whether such bifurcated and bipolar forms were resulted of alterations in the genetic regulatory processes involved in annelid regeneration.

Syllis malaquini Ribeiro et al. 2020 has been recently proposed as a suitable species for developmental studies. This species is able to reproduce by asexual fission and has remarkable ability of axial regeneration (regeneration of structures of the main body axis, e.g., heads and tails (see Zattara (2020)) (Ribeiro et al. 2020). A comprehensive study of the regenerative ability of S. malaquini is available, and it describes five regeneration stages: (1) wound closure, (2) blastema development, (3) blastema differentiation, (4) resegmentation, and (5) growth and reestablishment of digestive functions (Ribeiro et al. 2021). In this study, we assessed the morphogenetic effects of the selective pharmacological inhibitor of GSK3β 1-azakenpaullone (AZP) in S. malaquini regeneration, demonstrating that this approach is useful to investigate whether the involvement of the Wnt/β-catenin signaling pathway in annelid regeneration.

Material and methods

Specimens of S. malaquini were collected from marine aquariums at the Departamento de Biología of the Universidad Autónoma de Madrid (UAM), Spain, and the Animal Evolution and Biodiversity Department of the Georg-August University, Göttingen (GAUG), Germany. Specimens with no signs of regeneration or reproduction, bearing 60-65 segments, were selected for posttraumatic regeneration experiments. The specimens were amputated between segments 21 and 22 (intestine level, see Fig. 1) and allowed to regenerate for eight days post-amputation (dpa). The experiment at UAM was made in solutions with three different drug concentrations plus control of each: (1) 1 µM AZP in 0.01% DMSO, 0.01% DMSO as control; (2) 10 µM AZP, 0.1% DMSO as control; and (3) 20 µM AZP, 0.2% DMSO as control. The regeneration medium was prepared with filtered artificial sea water (ASW) after dissolution of AZP (Sigma-Aldrich A-3734) in dimethyl sulfoxide (DMSO, Sigma-Aldrich D8418) or only DMSO in control treatments. During experiments, those specimens (n = 10 per treatment) were maintained in six-well plates with their respective media, and some sterilized grains of sand, at room temperature (25 °C). Their respective media were changed each three days. A replicate of the experiment was performed at GAUG with the following solutions and concentrations: (1) 1 µM AZP in 0.01% DMSO, 0.01% DMSO as control, and (2) 10 µM AZP, 0.1% DMSO as control, were done (n = 10 per treatment, 8 days observation). In this case, medium was changed daily to check possible inactivation of the inhibitor during the first experiment in which the medium was changed every three days (Supplementary Material 1).

Fig. 1
figure 1

Scheme of S. malaquini body and surgical procedure

Full size image

In both experiments, we assessed regeneration during the first 8 dpa, moment in which resegmentation occurs in S. malaquini (Ribeiro et al. 2021). On the last day (8 dpa), the regenerated segments were counted in all amputees. However, there was a mortality of one posterior regenerating amputee in the trials of 1 µM AZP and 0.01 DMSO, which reduced the number of specimens examined to 9. No specimens died in the other treatments (n = 10 per treatment). Specimens were observed and photographed daily using a compound microscope Nikon Eclipse E100 with a mounted mobile phone Xiaomi Redmi 8 (UAM) and with a Leica MZ 125 stereomicroscope equipped with software analysis getIT for morphological account (GAUG). Segments were counted considering the observation of septa and body constrictions in living animals, at 200 × magnification. Statistical difference on the number of regenerated segments was tested using Kruskal–Wallis non-parametric test, performed with the software R, using RStudio version 1.4.1106 (R Core Team 2020).

Results and discussion

In both anterior and posterior regeneration, no morphological differences were noticed between the trial of 1 µM AZP and all control trials (0.01% DMSO, 0.1% DMSO, 0.2% DMSO). The animals in these treatments regenerated in a comparable pace along the stages in observation. The anterior regeneration blastema was developed around 2–4 dpa (Fig. 2a–c, e) and resegmentation could be observed from 6 dpa, with about two–three segments added until 8dpa (Fig. 2g–i, k). In the posterior regeneration, blastema differentiation occurred around 2–3 days post-amputation (dpa) (Fig. 2a’–c’, e’), and resegmentation could be observed from 4–5 dpa, with about two–three segments added until 8dpa (Fig. 2g’–i’, k). In these treatments, the amputees regenerated accordingly with previous description of S. malaquini regeneration under non-treated conditions (animals maintained in ASW only); i.e., they reached stage of resegmentation within 8 days after bisection (Ribeiro et al. 2021). No differences were observed between experiments performed at GAGU (daily change of medium) and at UAM (change of medium every three days).

Fig. 2
figure 2

Results of regeneration at 4 dpa and 8 dpa (medium changed every three days). a–l Anterior regeneration. a’–l’ Posterior regeneration. a–f Stage of blastema development. g–k Stage of resegmentation. l Stage of blastema differentiation. a’–f’ All specimens are in the stage of blastema differentiation; the pygidium has already appeared with cirri. g’, h’ Stage of resegmentation; two regenerated segments are barely seen. i’, k’ Stage of resegmentation with three regenerated segments with cirri. j’, l’ Stage of blastema differentiation. Top right rectangle shows details of bifurcated cirrus regenerated in one specimen of the 20 µM AZP treatment; arrows point to branches of bifurcated cirrus. Scale bars: 200 µm (a–l, a’–l’), 50 µm (l’ top left rectangle)

Full size image

On the other hand, in the trials with 10 µM and 20 µM AZP, resegmentation was not seen at 8 dpa. The blastema of both anterior and posterior regenerating amputees of those conditions developed around 2–4 dpa, as well as what happened in the other trials (Fig. 2e, e’, f, f’). However, the stages of blastema differentiation and resegmentation were affected. As regards the treatment of 10 µM AZP, anterior regenerates showed up to two segments within 8 dpa (Figs. 2j, 3a), while posterior regenerates showed up to one regenerated segment, and most of them have not started resegmentation (6 of 10) (Figs. 1j’ and 3b). The results obtained in controls and in the treatment with 10 µM AZP were comparable in both experiments (Figures S1 and S2, Supplementary Material 1).

Fig. 3
figure 3

Jitter plots of quantification of regenerated segments on the last day of experimentation (8 dpa). a Anterior regeneration. b Posterior regeneration

Full size image

In the treatment of 20 µM AZP, anterior regenerates did not add segments within 8 dpa (Figs. 2l and 3a), except for one specimen that had regenerated two poorly developed segments (Figure S3a, Supplementary Material 1). Meanwhile, posterior regenerates exposed to 20 µM AZP showed up to one regenerated segment, and most of them have not added segments (7 of 10) (Figs. 2l’, 3b, S1, and S2).

The number of anterior segments regenerated in controls versus animals exposed to AZP was significantly different (χ2 = 24.053, p < 0.01). Likewise, the number of posterior segments regenerated in controls versus animals exposed to AZP was significantly different (χ2 = 32.596, p < 0.001), according to the Kruskal–Wallis analysis (Fig. 3a, b). No alterations were seen in the non-regenerating body of amputees. In both experiments (medium change every three days or every 24 h), the outputs were equivalent, which reveals that results are replicable and consistent and that the inhibitor was active during both treatments.

These results reveal that blastema differentiation (during anterior regeneration) and resegmentation (during posterior regeneration) are disturbed in amputees of S. malaquini exposed to high concentrations of AZP for 8 dpa. These stages imply cellular proliferation and differentiation and, in non-treated conditions, both events occur within 4 dpa in animals amputated at the level of intestine (Ribeiro et al. 2021). Within Annelida, some studies have shown the requirement of the Wnt/β-catenin pathway in various cellular processes. In A. viride, anterior regenerating specimens treated with AZP showed an inhibition of cell proliferation in the wounded tissue, which disturbed wound healing and blastema formation (Chen et al. 2020).

Notably, one specimen of the treatment with 20 µM AZP regenerated a pygidium with a bifurcated cirrus with branches of unequal length after 8 dpa (Fig. 2l’ and S3). Bifurcated cirri and antennae have been repeatedly reported in syllids in the literature, either in specimens found in nature or in the laboratory, as a result of bisection-based experiments (Malaquin 1893; Michel 1909; Durchon and Wissocq 1964; Mohammad 1981). However, the appearance of bifurcated appendages during regeneration in pharmacological essays has not been documented before in annelids. GSK3β inhibitors (including AZP) affect the anteroposterior polarity in acoels, in which whole individuals and tissues of the worms became globular when exposed to AZP, indicating a loss of axial polarity (Sikes and Bely 2010). In annelids, posterior bipolarity can occur via regeneration after specific types of surgical procedure. For instance, bisection at the most anterior segments in aquatic oligochaetes resulted in the regeneration of posterior heads in place of tails in fragments with head, forming bipolar individuals (Müller 1908; Okada 1934a, b; Myohara et al. 1999; Kawamoto et al. 2005; Myohara 2012). There is a unique record of a bipolar-tailed syllid, Myrianida edwarsi (Saint-Joseph 1886) that regenerated an anterior tail (Okada 1929), though the study described the specimen only, and no further explanation about the process was provided.

In this study, the processes of blastema differentiation and resegmentation are affected by a selective inhibitor of GSK3β during regeneration in S. malaquini based on morphological observations. Thus, GSK3β-associated pathways (e.g., Wnt/β-catenin) are candidate modules to investigate the genetic mechanisms involved in the regular course of S. malaquini regeneration. However, solid evidence based on genetic essays should be further achieved to better understand how the GSK3β inhibition generates the morphological changes observed here.