Molecular cloning, expression, and in situ hybridization analysis of MnGPx-3 and MnGPx-4 from oriental river prawn, Macrobrachium nipponense, in response to hypoxia and reoxygenation

Glutathione peroxidase (GPx) has been the focus of increased research because of its important role as an antioxidant and in reactive oxygen species (ROS) induced damage repair. Studies on GPxs have relevance with Macrobrachium nipponense because it has poor tolerance to hypoxia in Macrobrachium nipponense. The two subunits named as MnGPx-3 and MnGPx-4 according to the glutathione peroxidase nomenclature system. Both full-length cDNAs were cloned from the hepatopancreas. In this study, we analyzed the expression of two GPxs in Macrobrachium nipponense in response to changes in environmental oxygen. Expression levels of MnGPx-3 and MnGPx-4 indicated that both have strong responses to hypoxia. In situ hybridization showed that MnGPx-3 and MnGPx-4 were located in secretory and storage cells in hepatopancreas. These results suggest that GPx gene is expressed and released by secretory cells and released response to hypoxia. In the gill tissue, however, GPxs are located in blood cells, suggesting that they perform different functions in different tissues or organs. The results of in situ hybridization were consistent with those of quantitative Real-time PCR. This study provides a basis for understanding the oxidative stress response in M. nipponense under hypoxia.


Introduction
Dissolved oxygen (DO) has been widely accepted as the main indicator of water quality in the aquaculture of prawns [1]. When organisms are adversely stimulated by the external environment, the balance between oxidation and the antioxidant system is disrupted. Once this balance is broken, it leads to increased production of reactive oxygen species (ROS) and further PLOS  concentration of the normoxic group was maintained at 6.0 ± 0.2 mg O 2 L -1 , whereas in the hypoxic group it was maintained at 2.0±0.2 mg O 2 L -1 by bubbling with N 2 gas [23]. A portable dissolved oxygen meter was used to monitore the change of dissolved oxygen in real time (Mettler Toledo, Switzerland). After the challenge with hypoxia, normal oxygen conditions were restored. The target tissues of six prawns were collected at 0, 12 and 24 h during the hypoxia process and 12 and 24 h of the reoxygenation process after 24h of hypoxia, and frozen immediately in liquid nitrogen. The samples were stored in ultra-low temperature freezer for the next steps (ThermoFisher, USA).

Cloning of two MnGPxs cDNAs
A partial fragment of opsin cDNA was obtained from transcriptomic cDNA library of the M. nipponense hepatopancreas (NCBI, SRA051767.2) created in our lab. The isolation of total RNA and the synthesis of first-strand cDNA and 3'-rapid amplification of cDNA ends (RACE) was performed using relevant kits [1]. All primers for these steps are listed in Table 1. The purification of PCR products and sequencing were completed according to instructions.

Bioinformatics analysis and nucleotide sequence
NCBI related online analysis software was used to complete the deduction of amino acid sequence, protein analysis and nucleotide sequence analysis according to the previous method of our laboratory [1]. Multi-sequence alignment and phylogenetic tree construction were completed with DNAMAN 6.0 and MEGA 5.1, respectively.

Quantitative Real-time PCR (QPCR) analysis of two MnGPx expression
The mRNA expression levels of the two MnGPx subunits following different treatments were measured by QPCR. Synthesis of the cDNAs and subsequent QPCR steps were completed as described previously [1,26]. According to previous laboratory studies, eukaryotic translation initiation factor 5A (EIF) was chosen as the reference gene of the relative quantification [27][28]. The mRNA expression levels of the two subunits were calculated by using the 2 −ΔΔCT method [29].

In situ hybridization (ISH)
Hepatopancreas and gill samples were collected according to the requirements of the chromogenic in situ hybridization (CISH) technique. Fixed paraffin-embedded sections, standard deparaffinization, sectioning and staining were prepared as previously reported by our laboratory [30]. Slides were evaluated using a light microscope. The following sequences were selected as CISH probes of MnGPx-3 and MnGPx-4, respectively, 5 0 -CTTGTAGTAGAGGG CGTTTGGATCAGCGAAAGCG-3 0 and 5 0 -CAATGGCAGGAATAGGGTTGGTCTGTGGAGAG-3 0 .

Statistical analysis
There were biological replicates for all experiments (n = 3). All data were analyzed for ANOVA by selecting the appropriate method using SPSS 20.0 software. According to the previous literature, p � 0.05 was considered to be significant [1,30].

MnGPx-3 and MnGPx-4 coding sequences
The full-length MnGPx-3 gene includes an ORF of 648 bp, encoding a 215 amino acid polypeptide with three active sites (Gln102, Trp177 and Asn178). The estimated molecular mass and theoretical pI of this protein were 24.35 kDa and 6.43, respectively. The 5'-untranslated region (UTR) and the 3'-UTR were 99 bp long and 596 bp long, respectively. A tail signal (AATAA) and polyA were also discovered ( Fig 1A). MnGPx-4 comprises an ORF of 492 bp, encoding a 164 amino acid polypeptide with three active sites (Gln75, Trp128 and Asn 129).
The deduced protein has an estimated molecular mass of 17.82 kDa and a theoretical pI of 6.30. The 5'-UTR is 259 bp long, and the 3'-UTR is 764 bp long, including a tail signal (AATAA) (Fig 1B). Three potential O-GlcNAc sites were found in MnGPx-3 (Ser23, Thr79 and Thr198) and MnGPx-4 (Ser3, Thr155 and Thr149), respectively. Three glycosylation sites in MnGPx-3 were predicted by the Net NGlyc 1.0 Server, namely, 50 NIS 52 , 107 NAT 109 , and 127 NFT 129 . Two glycosylation sites in MnGPx-4 were predicted by the Net NGlyc 1.0 Server, namely, 45 NYT 47 and 129 NFT 131 (Fig 2). The glycosylation site is where the protein chain connects to the sugar chain. Glycosylation plays an important role in protein folding, transport and half-life, as well as being involved in intercellular interaction and antigenicity. Table 1 lists all the primers we used.

Comparison and phylogenetic analysis of M. nipponense GPx
BLAST indicates that MnGPx-3 was 77.57% similar to P. monodon (GeneBank ID: AQW41378.1) and MnGPx-4 was 69.84% similar to P. vannamei (GeneBank ID: XP_027206898.1). Phylogenetic analysis of GPxs from M. nipponense confirmed that MnGPx-3 belongs to plasma GPx (GPx-3), while MnGPx-4 belongs to phospholipid hydroperoxide GPx (GPx-4). As the first GPxs reported on M. nipponense, they were named MnGPx-3 and MnGPx-4. Most of the deduced amino acid residues in the cDNAs were variable. Phylogenetic tree analysis showed that MnGPx-3 is most closely related to a GPx in P. monodon, while MnGPx-4 is the most closely related to phospholipid hydroperoxide GPx (Fig 3). GenBank accession numbers are listed in Table 2.

Tissue distribution of the two MnGPx subunits
QPCR (Quantitative real-time reverse transcription PCR) was used to quantify mRNA expression levels of two MnGPxs in different tissue, and both of these two subunits were detected in all tissues tested. MnGPx-3 mRNA levels were high in the eye, brain, gill and testis. The high-level expression of MnGPx-4 was similar to MnGPx-3 mRNA, while MnGPx-4 had less expression in the testis. The high expression of both MnGPx-3 and MnGPx-4 in the gill may indicate that these two subunits play an important role in detoxification and antioxidative damage ( Fig  4). Although MnGPx-3 and MnGPx-4 are not highly expressed in hepatopancreas, as an important antioxidant tissue, hepatopancreas is still listed as the key research tissue. Based in these expression results, hepatopancreas, gill and muscle tissue were chosen for subsequent experiments.

Expression of MnGPx subunits in response to changes in environmental oxygen
The expression patterns of MnGPx-3 and MnGPx-4 were determined by qPCR analysis. Following exposure to hypoxia in the hepatopancreas, expression of MnGPx-3 reached a maximum after 12 h but decreased thereafter. Except at 0 h, the levels of expression were significantly different between the two treatment groups (p < 0.05) (Fig 5A). This trend in the expression of MnGPx-4 reached a peak after reoxygenation for 12 h and then decreased ( Fig  5B). Under reoxygenation, MnGPx-4 expression returned to normal levels, but this was not the case for MnGPx-3 ( Fig 5). Furthermore, the situation of MnGPx-3 in gill was similar to  Expression and localization analysis of two GPxs from M. nipponense during oxygen environment changes hepatopancreas, whereas after reoxygenation for 24 h, MnGPx-3 expression returned to normal levels (p < 0.05; Fig 5C). However, MnGPx-4 expression continued to decrease after reoxygenation (p < 0.05; Fig 5D). In muscle, expression of MnGPx-3 peaked after reoxygenation for 12 h (Fig 5E). The trend in the expression of MnGPx-4 in muscle is similar to in the gill (Fig 5F).

ISH of MnGPx-3 and MnGPx-4 in hepatopancreas and gill
Hepatopancreas cells usually consist of secretory cells, basement membranes, lumen, storage cells, and transport vesicles. The volume of the transporter vesicle in the hepatopancreas first decreased and then increased during hypoxia. The morphology of the lumen also appeared to have a compression deformation and recovery with hypoxic reoxygenation (Fig 6A2-6E2). ISH analysis indicated that MnGPx-3 and MnGPx-4 were located in secretory cells and storage cells. In the negative control experiment, no signal was observed using sense-strand probes (Fig 6A3-6E3). Gill cells are mainly composed of a marginal channel, epithelial cell nuclei and hemolymph vessels. From the results of the section, it was obvious that the marginal channel of the gill cells would be significantly enlarged with an increase in the length of hypoxia, and blood cells flowed into the gill (Fig 7A2-7E2). But as the oxygen returned, the enlarged edge channel gradually returns. After reoxygenation for 24 h, some enlarged marginal channels were not restored to normal levels. ISH analysis showed that MnGPx-3 and MnGPx-4 were located in hemolymphatic vessels. In the negative control experiment, no signal was observed with the sense probe (Fig 7A3-7E3).

Discussion
The cloning and expression of two GPx subunits from M. nipponense in response to hypoxia and reoxygenation have been reported here. Previous research has shown that the catalytic Expression and localization analysis of two GPxs from M. nipponense during oxygen environment changes center of GPx was first identified as composed of Cys or Sec, Gln and Trp, and later was found to be a tetrad with an additional Asn [31][32]. The ternary catalytic center composed of Gln, Trp and Asn indicates that the GPx family has another catalytic mode, which is different from the traditional ternary catalysis and quaternary catalysis. This catalytic method participates in the fixation of selenium by using the stop codon to connect amino acid residues to form a catalytic structure [20]. Furthermore, both MnGPx-3 and MnGPx-4 include one Ser and two Thr. In addition, marker codons appeared in the amino acid sequences of both subunits, which could translate the upcoming termination codon into U [18,[33][34]. This indicates that the terminators in the two amino acid sequences are also likely to be translated as U. MnGPx-3 and MnGPx-4 are believed to belong to the plasma GPx (GPx-3) class and phospholipid hydroperoxide GPx (GPx-4) class respectively by functional domain analysis and were named according to the established GPx nomenclature system. So far, all available GPxs have been divided into eight different classes based on sequence homology and molecular characteristics, and some are species-specific [19]. MnGPx-3 was classified as the GPx3 class and MnGPx-4 was classified as the GPx4 class on the phylogenetic tree. A recent phylogenetic analysis indicated that the arthropod GPx homologues characterized to date form a clade with mammalian GPx3 [35], which is extracellular with broad substrate specificity [36]. A previous study Expression and localization analysis of two GPxs from M. nipponense during oxygen environment changes suggested that GPx3 and p53 have a synergistic effect in regulating cellular stress in P. monodon [20].
In this study, expression of MnGPx-3 increased at first and then decreased in the gill and hepatopancreas in response to hypoxia, while a slow decline occurred under reoxygenation (Fig 5A and 5C). Multiple studies have suggested that GPx mRNA levels can be an appropriate indicator of oxidative stress which is caused by various xenobiotics [37][38]. MnGPx-3 mRNA levels rose in the early stage of hypoxia, indicating that the hepatopancreas and gills of M. nipponense have a strong oxidative stress reaction in response to hypoxia. Decreased expression during hypoxia suggests that animals choose to downregulate metabolism to increase survival when the hypoxic environment does not improve. As previously reported, expression levels of mRNA for crustacean cardioactive peptide in M. nipponense have the same trend in eyestalk tissue under hypoxia [39]. However, the expression of MnGPx-3 in muscle showed a continuous upward trend, indicating that the physiological limit of muscles has not been reached after 24 hours of hypoxia. This phenomenon also suggests that muscles respond to hypoxia is different in comparison to the gills and hepatopancreas. Expression of MnGPx-3 was higher after reoxygenation compared with the control group (Fig 5A, 5C and 5E), indicating that gene expression may be stimulated by a changing level of oxygen in the environment. Similar Expression and localization analysis of two GPxs from M. nipponense during oxygen environment changes conclusions have been obtained in the study on hypoxia of glutathione S-transferase in M. nipponense [1]. The expression of MnGPx-4 in the hepatopancreas continued to increase during hypoxia and then decreased to normal levels after reoxygenation (Fig 5B). Compared with MnGPx-3, the increased expression of MnGPx-4 was relatively delayed, indicating that MnGPx-4 had a relatively late response to hypoxia in hepatopancreas. However, MnGPx-4 mRNA level has the same trend as MnGPx-3 in the gill (Fig 5D). These results suggest that MnGPx-4 plays different roles in hepatopancreas and gill. Compared with the control group, the sustained decline of MnGPx-4 indicates irreversible damage to organs in the gills following hypoxia [4,[40][41][42][43][44]. However, MnGPx-3 mRNA levels return to normal after reoxygenation. These results suggest that two subunits may be located on different organelles in the gill. This conclusion deserves further study. The expression of MnGPx-4 follow a different trend to MnGPx-3 (Fig 5), indicating that MnGPx-3 and MnGPx-4 have different functions in the same organization.
In this study, HE staining was used to observe the hepatopancreas and gill sections of M. nipponense, and ISH was used to locate the two subunits. The volume of the transporter vesicle in the hepatopancreas first decreased and then increased during hypoxia. Previous experiments have demonstrated that the volume of intracellular transport vesicles increased in Procambarus clarkii under low pH stress [45], the volume of transport vesicles increased in Portunus trituberculatus under low salt environment [46], and the volume and number of transport vesicles increased in Eriocheir sinensis nitrogen stress [47], which are consistent with our experimental results. The trend of fluorescence quantification of MnGPx-3 and MnGPx-4 first increased and then decreased under hypoxia, and then increased after reoxygenation ( Fig  5). Since MnGPx-3 and MnGPx-4 were located on secretory cells, we inferred that hypoxia and acute reoxygenation led to an increase in secretory cells. However, this inference will be confirmed by more detailed experiments. These results suggest that the GPx gene is formed in secretory cells and released into cells for a specific function. MnGPx-3 and MnGPx-4 were also located in the storage cells by ISH experiments. This suggests that in response to hypoxic stress, GPx genes are formed in secretory cells as well as in storage cells to prepare for subsequent hypoxic stress. In hypoxic conditions, the marginal channels in the gill tissue are enlarged and accompanied by inflow of blood cells. This result is consistent with the results of gill structure changes of Oncorhynchus mykiss [48] and Gymnocypris przewalskii [49] under hypoxic stress. The results also show that with the increased length of hypoxia, the marginal channels of gill tissue began to expand. However, as oxygen was restored, the expanded marginal channels were gradually restored. After reoxygenation for 24 hours, some marginal channels were still not restored. This result suggests that 24 hours of reoxygenation is not enough to recover the harm of hypoxia to the gill tissue. This prediction coincides with the results of qPCR. ISH showed that both gene subtypes were located in hemolymph vessels. This result indicates that MnGPx-3 and MnGPx-4 plays a role in gill tissue through hemolymphatic vessels in response to hypoxic stress.

Conclusion
In conclusion, we cloned the full-length cDNA sequence of MnGPx-3 and MnGPx-4 and analyzed their expression patterns in different tissues of M. nipponense. We also examined their expression profiles in the hepatopancreas, gill and muscle at five time points during hypoxia and reoxygenation. HE staining was used to observe the hepatopancreas and gill sections of M. nipponense, and ISH was used to locate the two subunits. The experimental results showed that both subtypes responded to hypoxia, but their expression patterns were different in different tissue. This study provides a basis for understanding the oxidative stress response in M. nipponensis under hypoxia.