Through genome-wide screening, a total of 35 putative A. farreri GST (AfGST) genes and 37 M. yessoensis GST (MyGST) genes were identified, which could encode proteins containing the GST_N domain (Pfam: PF02798) and GST_C domain (Pfam:PF00043) (Table 1 and Supplementary Fig. S1). Information for these GST genes, including their genome location, encoding protein length and GST domain, SwissProt annotation, and exon number, was summarized in Supplementary Table S1. The ORF (open reading frame) of AfGSTs and MyGSTs varied from 435 to 1851 bp and 312 to 1851 bp in length, encoding 144 to 616 and 103 to 616 amino acids (aa), respectively (Supplementary Table S1). To further investigate the GST structures, conserved domains were annotated in scallop GSTs. All the scallop GSTs composed of the GST domains, and most of them were ranging from 150 to 320 aa in both AfGSTs (33 out of 35) and MyGSTs (33 out of 37) (Supplementary Fig. S1). There was only one longer AfGST-ζ3 and MyGST-ζ3 consisting of 616 aa with two conserved GSTA and GST_C domains, respectively. Also, there was one shorter AfGST and three shorter MyGSTs ranging from 103-138 aa with GST_N and MAPEG domains detected. According to the SwissProt annotation, 29 AfGSTs belonged to the cytosolic subfamily, which were further divided into seven classes, including sigma (14 copies), theta (five copies), omega (four copies), zeta (three copies), pi (one copy), rho (one copy), and mu (one copy), and other one and five AfGSTs were from the mitochondrial and microsomal subfamilies, respectively (Table 1). In M. yessoensis, the 37 MyGSTs were categorized into three subfamilies, including the mitochondrial (1), microsomal (4), and cytosolic (32) GSTs. The 32 cytosolic MyGSTs were further classified as 16 in class sigma, five in theta, three in omega, three in rho, three in zeta, one in pi, and one in mu (Table 1). These classifications were further supported by the phylogenetic analysis (Fig. 1). Compared with vertebrates, mollusks and C. elegans had more cytosolic GST members, mainly due to the abundance of sigma class, which, however, was absent in humans, frog, and zebrafish. Over ten sigma GSTs were found in mollusks, including scallops, oysters, and owl limpet, which was consistent with the previous finding that sigma GSTs were expanded in invertebrate genomes (Flanagan and Smythe 2011). However, the alpha class was only absent in scallops, which was previously regarded as a GST class specific to mammals (Frova 2006), but its presence in oysters (Zou et al. 2015) implied that alpha GST gene might have been lost in the ancestor of scallops.
GST subfamily Cytosolic Mitochondrial Microsomal Alpha Mu Omega Pi Rho Sigma Theta Zeta H. sapiens 5 5 2 1 0 0 2 1 1 3 X. tropicalis 3 1 2 3 0 0 2 1 2 3 D. rerio 3 3 2 2 1 0 2 1 4 8 C. elegans 0 0 5 6 0 29 0 1 2 0 L. gigantea 1 1 4 1 1 11 2 2 1 8 C. gigas 2 4 5 1 4 10 7 3 1 6 A. farreri 0 1 4 1 1 14 5 3 1 5 M. yessoensis 0 1 3 1 3 16 5 3 1 4
Table 1. Comparison of GST gene numbers between mollusks and other selected species
Figure 1. Phylogenetic analysis of GST amino acid sequences. The ML (maximum-likelihood) tree was constructed with GST sequences from mollusk species and other selected model organisms, with tested LG + G model and bootstrapping of 1000 pseudo-replicates. Values less than 70% were not shown. The AfGSTs and MyGSTs were highlighted in red and blue, respectively (Hs: Homo sapiens; Xt: Xenopus tropicalis; Dr: Danio rerio; Ce: Caenorhabditis elegans; Lg: Lottia gigantean; Cg: Crassostrea gigas; Af: Azumapecten farreri; My: Mizuhopecten yessoensis)
The expression of GSTs in scallop embryos/larvae and adult organs/tissues was analyzed using the transcriptome dataset (Li et al. 2017; Wang et al. 2017). As shown in Fig. 2, during the development of scallops, the sigma class genes, such as AfGST-σ5, -σ7, and -σ10 in A. farreri, and MyGST-σ2, -σ4, -σ5, -σ8, -σ10, and -σ11 in M. yessoensis exhibited relatively higher expression level than other GSTs suggesting the importance of sigma GSTs in scallop development. The expansion of sigma GSTs in scallops might be related to the protection of scallop embryos/larvae from the environmental or internal oxidative stress. In addition to sigma class members, AfGST-µ1, AfGST-ω2, and AfGST-microsomal2 in A. farreri, and MyGST-μ1, MyGST-ω1, -ω2, MyGST-π1, MyGST-ρ3, and MyGST-microsomal1b in M. yessoensis showed moderate expression. Along the entire developmental process, high expression of the GST genes was mostly present at the larval stages, especially after umbo larva, and in the following juvenile stage, implying the involvement of these GSTs in antioxidation or detoxification during scallop metamorphosis and post-larval development.
Figure 2. Expression pattern of scallop GSTs during the development of Azumapecten farreri (a) and Mizuhopecten yessoensis (b). The number in each box represents the lg(RPKM) value at the corresponding developmental stage
In adult scallop tissues, higher expression of GSTs was detected mostly in hepatopancreas and kidney. For example, AfGST-σ5, -σ7 and -σ8 in A. farreri, MyGST-σ2, -σ4, -σ5, -σ8, and -ω2 in M. yessoensis (Fig. 3), most of which were sigma class members, were all highly expressed in hepatopancreas or kidney. Abundant GST transcripts in kidney and liver (or hepatopancreas) were found also in many other organisms, such as rats (Daggett et al. 1998), plaice (Pleuronectes platessa) (Leaver et al. 1993), goldfish (Carassius auratus) (Hao et al. 2008), and bivalves, such as Lamellidens marginalis (Chetty 1995) and Mytilus galloprovincialis (Fitzpatrick et al. 1995). Kidney and liver (or hepatopancreas) are the organs for eliminating toxic substances; for example, liver and kidney function in detoxifying chemicals (Wang et al. 2011) and in removing wastes (Chow et al. 2010), respectively. The scallop GSTs with abundant mRNAs in hepatopancreas and kidney, especially the sigma class members, may well participate in these processes. In addition to hepatopancreas and kidney, MyGST-σ5 was highly expressed in mantle and striated muscle, indicating its possible involvement in the protection of these organs/tissues.
Dinoflagellates of the genus Alexandrium are the major PST producers in scallop farming, but the toxicities of Alexandrium vary among species and strains due to the different PST analogs which they produce. For example, the A. minutum (AM-1 strain) and A. catenella (ACDH strain) used in this study mainly produce PST analogs of gonyautoxins (GTXs, mainly GTX1-4) and N-sulfocarbamoyl toxins (C1/2), respectively (Anderson 1990; Chang et al. 1997; Hu et al. 2019; Li et al. 2017). For scallops exposed to PST-producing algae, the toxicity varied among organs/tissues. Our previous study showed that scallop hepatopancreas and kidney could accumulate much higher concentration of PSTs than other organs, with hepatopancreas mainly absorbing the incoming PSTs from dinoflagellates directly, whereas kidneys transforming the absorbed toxins into more potent analogs, saxitoxin (STX) and neosaxitoxin (NeoSTX) (Li et al. 2017; Shimizu and Yoshioka 1981). Systematically analyzing the expression profile of GSTs in hepatopancreas and kidney of scallops exposed to different Alexandrium species could deepen our understanding of the defense mechanism against toxic algae in scallops. Therefore, in this study, 102 RNA-Seq libraries (three replicates for each test time point) from scallop hepatopancreas and kidneys of challenged and control groups were constructed, and sequenced to analyze the expression regulation of scallop GSTs after toxic dinoflagellates exposure (Supplementary Table S2).
In the hepatopancreas of A. farreri, the same five AfGSTs were differentially expressed (DE) after exposure to both A. minutum (5 with P < 0.05 and 3 with FDR < 0.05) and A. catenella (5 with both P < 0.05 and FDR < 0.05). These five genes were AfGST-σ2, -σ4, -σ10, -σ11, and -σ14, all belonging to the sigma class, but with opposite regulation patterns between the two algae exposure, except the up-regulated AfGST-σ14 (Fig. 4a and Supplementary Table S3). After exposure to A. minutum, the four DE AfGSTs were all down-regulated, with AfGST-σ4 presenting only at 15 days after exposure, whereas AfGST-σ2, -σ10, and -σ11 decreased both acutely (1-3 days) and chronically (5-15 days), with AfGST-σ10 being down-regulated at all five sampling time points. In contrast, when being exposed to A. catenella, the four AfGST genes were all significantly up-regulated, and AfGST-σ2, -σ10, and -σ11 were induced both acutely and chronically, while AfGST-σ4 was regulated chronically. Scallop hepatopancreas is the organ absorbing PSTs from the ingested algae (Li et al. 2017). These results indicated that the five AfGST-σ genes might be the key GSTs for the protection of A. farreri hepatopancreas from the harmful effects of PST-producing dinoflagellates. It was reported that sigma class GSTs showed sensitive regulation response against environmental stress inducers (Rhee et al. 2013; Won et al. 2011). The same GST expression in different individuals may be affected by various types of stimulation. In marine invertebrates, the sigma class members were found highly expressed in response to BDE-47 exposure in Venerupis philippinarum and Mytilus galloprovincialis (Li et al. 2015). The GST-mitochondrial gene in Haliotis discus discus was up-regulated after exposure to bacterial lipopolysaccharide, but down-regulated after exposure to Vibrio parahaemolyticus (Sandamalika et al. 2018). Moreover, genes involved in oxidative metabolism such as Catalase and Glutathione peroxidase were also down-regulated after A. minutum exposure in C. gigas (Mat et al. 2013). In this study, according to the PST composition of the two Alexandrium species, mainly producing PST analogs GTX1-4 and C1/2, respectively, therefore, the opposite regulating patterns of the four sigma AfGSTs between exposure to A. minutum and A. catenella may well be related to the different types of PSTs absorbed from the two algae.
Figure 4. Temporal expression of scallop GSTs in hepatopancreas and kidney of Azumapecten farreri after Alexandrium minutum and Alexandrium catenella exposure (a), and of Mizuhopecten yessoensis after Alexandrium catenella exposure (b). The heatmap was drawn based on log2(fold change) values (Supplementary Table S3). Exposure time (1, 3, 5, 10, or 15 days) was displayed below the heatmap. Significant regulation with |log2FC| > 2 and P < 0.05 was indicated with "*", and very significant regulation with |log2FC| > 2 and FDR < 0.05 was indicated with "**"
In A. farreri kidneys, where the ingested PSTs being transformed to higher toxic analogs, more AfGSTs were significantly regulated after exposure to A. minutum (11 with P < 0.05 and 8 with FDR < 0.05) and A. catenella (12 with P < 0.05 and 6 with FDR < 0.05) (Fig. 4a). After exposure to A. minutum, five GSTs were up-regulated, including AfGST-σ5 and -σ14 induced at 1 to 10 days, and AfGST-σ11, -θ3, and -θ5 induced at 15 days post-challenge. Both AfGST-σ11 and -θ5 were also down-regulated at 10 days. The other six AfGSTs were down-regulated, including AfGST-σ7, -σ12, and -θ2 regulated acutely, and AfGST-ρ1, -σ10, -σ12, and -ζ3 regulated at 10 days after exposure. It was reported that the over-expression of GST in Aedes aegypti may increase the capacity of toxin resistance (Helvecio et al. 2019). The majority of regulated AfGSTs in kidney were from the sigma class, suggesting the sigma class members being also the major GSTs involved in detoxification of PSTs in scallop kidney. Meanwhile, the up-regulation of AfGST-θ3 and -θ5 might also contribute to the protection of A. farreri kidney. In the kidney of scallops ingested A. catenella, a similar set of DE AfGSTs was detected as those with A. minutum challenge, except that AfGST-σ4 and -σ8 were regulated only for A. catenella challenge, whereas AfGST-σ11 only for A. minutum challenge (Fig. 4a). Moreover, between the challenges of the two algae, the AfGSTs showing similar up- and down-regulation patterns, whereas their regulated time points were different. For example, after A. minutum exposure, the majority of DE AfGSTs (8) were present at 10 days, and only three genes were regulated at one day. While after ingesting A. catenella, six AfGSTs were regulated at one day and only two genes were regulated at 10 days. In scallop kidney, the ingested PST analogs, GTXs from A. minutum and C1/C2 from A. catenella, could be transformed into more toxic analogs STX and NeoSTX, respectively, whereas both transformations included reductive elimination of o-sulfate groups (Li et al. 2017; Shimizu and Yoshioka 1981). The diverse expression profiles of AfGSTs indicated that, with the challenges of different PST analogs, the AfGSTs functioning in PST detoxification and antioxidation during PSTs transformation were similar in scallop kidneys, but the gene stimulation or inhibition procedure/speed might vary.
In addition, regarding the highly expressed GSTs in normal scallop hepatopancreas and kidney (Fig. 3), AfGST-σ5, -σ7, and -σ8 were only significantly regulated in kidney but not in hepatopancreas after PST-producing algae exposure, and they were constantly keeping high expression level during the challenge experiments (Supplementary Table S2). The similar pattern has also been reported in genes coding antioxidant enzymes, such as the expression of SuperOxide Dismutase (SOD), CATalase (CAT), and GST-π in mussels remaining stable after short-term exposure to Diarrhetic Shellfish Poisoning (DSP) (He et al. 2019; Prego-Faraldo et al. 2017). Moreover, the expression of SOD1 was high in A. farreri kidney, but its expression was not regulated after A. minutum and A. catenella exposure (Lian et al. 2019). The results here suggested that these highly expressed GSTs probably need to keep playing vital roles in detoxification, such as AfGST-σ5 and -σ7 in hepatopancreas and MyGST-σ5 in both kidney and hepatopancreas. On the other hand, AfGST-σ14 was hardly expressed in normal scallop tissues (Fig. 3), but presented a significant up-regulation in both kidney and hepatopancreas after PST-producing algae exposure. The up-regulation of AfGST-σ14 was mostly from 5 to 10 days, except in kidney after exposed to A. minutum with both acute (one day) and chronical (5-10 day) regulation, which may imply the reaction of AfGST-σ14 to higher toxin stimulation.
To further understand the different regulation of GSTs between scallop species, we examined MyGSTs transcripts in both hepatopancreas and kidney of M. yessoensis ingesting A. catenella (Fig. 4b). A total of five (2 with FDR < 0.05) and three (2 with FDR < 0.05) regulated sigma GSTs were detected in M. yessoensis hepatopancreas and kidney, respectively, which were less than those identified in A. farreri. For both organs, MyGST-σ2 was the only up-regulated member implying the importance of this sigma class GST in the protection of M. yessoensis from the challenge of PST-producing algae. MyGST-σ1 and -θ5 were both down-regulated in the two organs, whereas MyGST-ω2 and -σ6 were down-regulated only in the hepatopancreas. The species- and organ-specific regulation of GSTs imply the diverse function of scallop GSTs in the defense against the harmful effects of PST accumulation and metabolism, which might contribute to the adaptive evolution of scallops.