Insights into the Prostanoid Pathway in the Ovary Development of the Penaeid Shrimp Penaeus monodon

The prostanoid pathway converts polyunsaturated fatty acids (PUFAs) into bioactive lipid mediators, including prostaglandins, thromboxanes and prostacyclins, all of which play vital roles in the immune and reproductive systems in most animal phyla. In crustaceans, PUFAs and prostaglandins have been detected and often associated with female reproductive maturation. However, the presence of prostanoid biosynthesis genes remained in question in these species. In this study, we outlined the prostanoid pathway in the black tiger shrimp Penaeus monodon based on the amplification of nine prostanoid biosynthesis genes: cytosolic phospholipase A2, hematopoietic prostaglandin D synthase, glutathione-dependent prostaglandin D synthase, prostaglandin E synthase 1, prostaglandin E synthase 2, prostaglandin E synthase 3, prostaglandin F synthase, thromboxane A synthase and cyclooxygenase. TBLASTX analysis confirmed the identities of these genes with 51-99% sequence identities to their closest homologs. In addition, prostaglandin F2α (PGF2α), which is a product of the prostaglandin F synthase enzyme, was detected for the first time in P. monodon ovaries along with the previously identified PUFAs and prostaglandin E2 (PGE2) using RP-HPLC and mass-spectrometry. The prostaglandin synthase activity was also observed in shrimp ovary homogenates using in vitro activity assay. When prostaglandin biosynthesis was examined in different stages of shrimp ovaries, we found that the amounts of prostaglandin F synthase gene transcripts and PGF2α decreased as the ovaries matured. These findings not only indicate the presence of a functional prostanoid pathway in penaeid shrimp, but also suggest a possible role of the PGF2α biosynthesis in shrimp ovarian development.


Introduction
Prostanoids are oxygenated derivatives of C-20 polyunsaturated fatty acids (PUFAs) that play active roles in inflammation, immune response, cardiovascular control and reproduction in most animals [1][2][3]. These PUFAs, which serve as precursors of the prostanoid pathway, include arachidonic acid (AA), eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA). The prostanoid pathway begins with the enzyme phospholipase A 2 , which releases AA from the phospholipids of cellular and intracellular membranes [3]. The released AA is then cyclized and subsequently reduced by the cyclooxygenase (COX) enzyme to form prostaglandin G 2 (PGG 2 ) and prostaglandin H 2 (PGH 2 ), respectively [4,5]. Downstream enzymes, including prostaglandin and thromboxane synthases, later convert PGH 2 to prostanoids, such as prostaglandins, prostacyclins and thromboxanes, which serve as signaling molecules in various physiological responses [3,6].
In crustaceans, one of the more prominent roles of prostanoids is the regulation of female reproductive maturation. For instance, the production of PGE 2 and PGF 2α is positively correlated with ovarian maturation in P. paeninsulanus [18] and O. senex senex [20]. Furthermore, injection of PGE 2 and PGF 2α into ovaries of O. senex senex significantly increased the number and the diameter of the oocytes in a dose-dependent manner [20]. In domesticated P. monodon, the amounts of PGE 2 in ovaries and haemolymph increased along with developing ovary stages [21]. However, the correlation between prostaglandins and crustacean ovary development may be species specific, as the amounts of PGE 2 and PGF 2α were highest in ovaries stage I and continued to decrease until stage IV in M. japonicus [10]. Nevertheless, these findings suggest a possible involvement of the prostanoid biosynthesis in crustacean female reproductive system.
Although the production of prostanoids and their precursors is well-established in most crustaceans, prostanoid biosynthesis genes in these species are poorly characterized. Thus far, the only crustacean with a fully constructed prostanoid pathway is the fresh water flea Daphnia pulex, whose annotated genome sequence revealed nine prostanoid biosynthesis genes: cytosolic phospholipase A2 (cPLA2), COX, prostaglandin D2 synthase A, prostaglandin D2 synthase B, prostaglandin E2 synthase (PGES), carbonyl reductase 1, thromboxane A and thromboxane B, and two prostanoid receptors prostanoid receptor EP4 isoform A and B [22]. In marine crustaceans, COX genes have been identified in Gammarus spp. and Caprella spp. [23], while PGES genes have been characterized in L. vannamei [24], the American lobster Homarus americanus (Accession: MGID155886 from the Marine Genomic Project) [24] and the sea lice Lepeophtheirus salmonis and Caligus rogercresseyi [25].
Due to the roles of prostanoids in the reproductive system in most crustaceans, the characterization of the prostanoid pathway in economically valuable organisms, such as penaeid shrimp, is essential for both scientific gain and potential applications in aquaculture practice. In this study, we propose a scheme for the P. monodon prostanoid pathway based on the identification of eight P. monodon prostanoid biosynthesis genes, the detection of lipid precursors and prostaglandins, and the detection of prostaglandin synthase activity. The correlations observed among gene transcription, prostaglandin production, and ovarian maturation also suggest that prostaglandin biosynthesis may be involved in the regulation of the P. monodon female reproductive system.

Analysis of conserved residues and domains in PmPGES1, PmPGES2, PmPGES3 and PmPGFS genes
The putative P. monodon prostanoid biosynthesis genes were submitted to the Conserved Domain Architecture Retrieval Tool (CDART) for protein domain prediction, revealing that the P. monodon proteins contain the same domain types and positions as the prostanoid enzymes found in other species ( Figure S1). At this point, PmPGES and PmPGFS were examined in more details, as these genes are likely to be responsible for the biosynthesis of PGE 2 and PGF 2α, which has been shown to affect ovarian development in other crustaceans [10,18,20]. In P. monodon, three isoforms of PmPGES have been cloned and characterized, namely PmPGES1, PmPGES2 and PmPGES3, which have corresponding isoforms in mammals. PGES1 or membrane-associated prostaglandin E synthase 1 is a member of the Membrane-Associated Protein involved in Eicosanoid and Glutathione metabolism (MAPEG) [27,28]. PGES1 requires glutathione for its enzymatic function, which involves converting PGH 2 to PGE 2 [27]. Multiple sequence alignment of PmPGES1 revealed the conservation of most catalytic residues. For example, D47, which is highly conserved in the PGES1 subgroup of MAPEG, was found in PmPGES1 ( Figure 1, star) [29]. In addition, key catalytic residues that interact with PGH 2 (R108 and T112 - Figure 1, white arrow heads) as well as glutathione (R36, N72, E75, H111, Y115 and R122 - Figure 1, black arrows) were conserved in PmPGES1 [29][30][31]. Lastly, PmPGES1 also contains the consensus sequence 64 ERXXXAXXNXX 75 E required for oxygenation product formation (Figure 1, underlined) [31], confirming that PmPGES1 possesses all the necessary residues for its enzymatic function.
PGES2 or membrane-associated prostaglandin E synthase 2 is a Golgi membrane-associated pre-protein that requires spontaneous cleavage of the N-terminal hydrophobic domain to become a mature cytosolic enzyme [32,33]. The catalytic domain of PGES2 is a glutathione/thioredoxin-like domain that can be activated by various thiol reducing reagents [34]. Sequence alignment of PmPGES2 revealed a conserved 66 C-X-X-69 C ( Figure 2, black arrow), which corresponded to 110 C-X-X-113 C catalytic triad in the human PGES2 active site [35]. A conserved N-terminal hydrophobic domain (Figure 2, underlined) was also identified in PmPGES2, consistent with Nterminal cleavage to generate the mature enzyme.
PGES3 or cytosolic prostaglandin E synthase (cPGES) is a 23 kDa GSH-requiring enzyme that was originally termed p23 based on its initial characterization as a co-chaperone of heat shock protein 90 [28,36]. Unlike the membrane-bound PGES1 and PGES2, PGES3 is a cytosolic, glutathione-requiring enzyme that interacts with casein kinase II (CKII) and Hsp90 [36,37]. In human PGES3, two serine residues (S113 and S118 − Figure 3, arrows) are phosphorylated by CKII to increase PGES3 enzymatic activity. However, only the putative Nterminal phosphorylated serine is conserved in PmPGES3 ( Figure 3, black arrow) [37]. CDART prediction of the putative PmPGES3 enzyme revealed the presence of an alpha crystallin-Hsps-p23 like super family domain, which is the same domain found in all PGES3 homologs ( Figure S1), further confirming the identity of the PmPGES3 gene.

RP-HPLC and mass spectrometry analysis of PUFAs and prostaglandins in shrimp ovaries
Once the prostanoid biosynthesis genes had been identified, chemical analysis was performed to detect the presence of corresponding prostanoids in wild P. monodon. Stage IV shrimp ovaries from five broodstock were pooled, homogenized and subjected to solvent extraction. Subsequent RP-HPLC analysis led to the detection of PGF 2α and PGE 2 as two small peaks that eluted at 9.26 and 10.03 minutes, respectively ( Figure 5A and 5B). The identities of these prostaglandins were later confirmed by mass analysis (Figure 5C and 5D). In addition, three prostanoid precursors EPA, DHA and AA were detected with elution times of 30.08, 31.74 and 32.56 minutes, respectively ( Figure S2, A and B). Again, mass spectra confirmed the identities of these PUFAs ( Figure S2, C-E). Other prostanoids i.e. 6-keto-PGF 1α and PGD 2 were not detected in P. monodon extracts from stage I-III shrimp ovaries, hepatopancreases, lymphoid organs or hemolymph.

Prostaglandin synthase activity in P. monodon ovarian tissue
The identification of PmPGES and PmPGFS genes and their corresponding products led us to speculate that prostaglandin synthase activity could be present in shrimp ovaries. To test this hypothesis, in vitro prostaglandin synthase activity assay was performed by incubating shrimp ovary homogenates with 25 µM AA at 28 °C, 200 rpm. The samples were subsequently collected at different time points to monitor prostaglandin production. Prior to the treatment, the basal concentration of PGE 2 in shrimp ovary homogenates was 9.6 ng/g tissue. After 30 minutes of incubation with AA, the PGE 2 concentration increased to 19.6 ng/g tissue and remained at this level before declining after 120 minutes of treatment ( Figure 6A). Similarly, the concentration of PGF 2α increased from the basal level of 10.9 ng/g tissue to 23.3 ng/g tissue after 60 minutes of incubation with AA and remained at the same level from 60 to 360 minutes after the treatment ( Figure 6B). Together, these results suggest that the prostaglandin synthase activity is present in P. monodon ovary homogenates.

Role of prostaglandin biosynthesis in shrimp ovarian development
The correlation between the amounts of prostaglandin and ovarian maturation in P. paeninsulanus [43], M. japonicus [10] and O. senex senex [20] led to the hypothesis that prostaglandins affect female reproductive development in crustaceans. To assess the possible role of prostaglandin biosynthesis in P. monodon ovarian maturation, we estimated the concentrations of PGE 2 and PGF 2α in different stages of shrimp ovaries. When compared with stage I ovaries, we observed that the PGE 2 concentrations were lower in ovaries stage II and III and higher in ovaries stage IV ( Figure 7A). On the other hand, the PGF 2α concentrations were highest in stage Figure 1. Mapping of essential residues in the predicted P. monodon PGES1 enzyme. Multiple sequence alignments of P. monodon PGES1 and their homologs showed a conserved residue for the MAPEG family (white star), catalytic residues that interact with PGH 2 (white arrow head), essential residues for H-bonding to GSH (black arrows) and consensus sequence required for oxygenation product (underline). Genus and species used in this alignment are abbreviated as followed:  I ovaries and steadily decreased as the ovaries matured ( Figure 7B). To correlate the mRNA expression levels of prostaglandin biosynthesis genes with different ovarian maturation stages, quantitative real-time PCR analysis was performed on ovaries of 27 wild-caught broodstock from the Andaman Sea (Table 2). Four sets of primers for PmPGES1, PmPGES2, PmPGES3 and PmPGFS genes were used in this study (Table 3). To determine whether the gene was up-or down-regulated at a certain developmental stage, mRNA expression levels were compared with that of stage I ovaries, which was taken as a baseline. It was observed that all three PmPGES isoforms displayed different expression profiles during the development of ovaries. PmPGES1 was up-regulated 16-fold in stage II ovaries, then continued to decrease until it reached baseline in stage IV ( Figure 8A). For PmPGES2, the change in expression from stage I to stage III ovaries was not significant. However, a significant 27-fold down-regulation of this transcript was observed in stage IV ovaries ( Figure 8B). The PmPGES3 expression level did not change significantly throughout ovarian   Figure 8C). Lastly, the amount of PmPGFS gene transcripts steadily decreased as the ovary maturation progressed from stage I to stage IV ( Figure 8D).

Discussion
In this study, the characterization of the P. monodon prostanoid pathway reveals that P. monodon contains the same types and number of prostaglandin synthase isoforms as those found in mammals (Fig, 1-4; Table 1). In addition to the three PmPGES isoforms shown in the Results section, P. monodon also encodes two isoforms of prostaglandin D synthase that matched the glutathione-dependent prostaglandin D synthase and hematopoietic prostaglandin D synthase genes originally identified in mammals. Interestingly, the proposed prostanoid pathway in another crustacean Daphnia pulex contains only one isoform of each prostanoid biosynthesis gene [22], suggesting that the organization of the P. monodon prostanoid pathway is more conserved with those in mammals, or other divergent prostanoid biosynthetic genes are present in D. pulex that are not annotated. To compare if the P. monodon prostanoid protein sequence is more closely related to its homologs in D. pulex or mammals, phylogenetic analysis of PGES1 was performed, revealing that PmPGES1 is more closely related to its crustacean homologs, including D. pulex, than to its mammalian homologs ( Figure S3).
PUFAs and PGE 2 have previously been identified in domesticated P. monodon, although it was not clear whether  Table 3. Primer sequences used in quantitative real-time PCR analysis. these molecules were synthesized de novo as the animals had been fed with high-PUFA diets [13,21]. In this study, PGF 2α was detected by RP-HPLC and mass spectrometry analysis in ovaries of wild-caught P. monodon, suggesting that these molecules occurred naturally in wild population consistent with de novo synthesis and were not the result of specific diets or rearing conditions. Although PmPGDS and PmTBXAS genes are present in P. monodon (Table 1), we were unable to detect the corresponding prostanoid products PGD 2 , PGF 1α and TBX 2 which have previously been identified in other arthropods [1,20,44,45]. Therefore, these prostanoids may be present in small amounts in P. monodon.

Genes Primers Sequence
To provide further proof of a functioning prostanoid synthetic pathway, we assessed the prostaglandin synthase activity in penaeid shrimp using an in vitro activity assay to monitor PGE 2 and PGF 2α biosynthesis in P. monodon [46]. After shrimp ovary homogenates were incubated with AA, the PGE 2 and PGF 2α concentrations increased significantly when compared to the basal concentrations ( Figure 6). In particular, the rise and fall of PGE 2 concentrations observed in this study is similar to the study performed on A. americanum [47]. As the PGE 2 biosynthesis activities have already been established in the Tobacco hornworm Manduca sexta [48], the blood sucking bug Triatoma infestans [49], the firebrat Thermobia domestica [50] and the cricket Teleogryllus commodus [51], we propose that the PGE 2 biosynthesis is conserved among different arthropods.
Once the prostanoid pathway had been established in P. monodon, we examined whether there is a correlation between prostanoid biosynthesis and shrimp ovary development. In wild P. monodon broodstock, the concentrations of PGE 2 decreased from stage I to stage III, but abruptly increased in stage IV ovaries ( Figure 7A). This is inconsistent with the trend found in domesticated shrimp, in which the amounts of PGE 2 gradually increased during ovary development [21]. The discrepancy between the two studies may be the result of different shrimp genetic background and/or dietary intake. Interestingly, the PGE 2 concentrations were highest at stage IV ovaries in both wild and domesticated shrimp, but the significance of this observation has yet to be explored. When the gene expression of each PmPGES isoform was compared to the concentration of PGE 2 found in each ovary stage, we observed no correlation between the two parameters. Furthermore, the lowest amount of PmPGES transcripts was found in stage IV ovaries, making it unlikely that the PGE 2 biosynthesis is regulated at the transcriptional level in this organ.
Unlike PGE 2 , the amount of PGF 2α steadily decreased with increasing ovary stages, which followed the same trend observed in M. japonicus [10]. Similarly, the PmPGFS gene expression also decreased as the ovaries matured, making PmPGFS gene expression and PGF 2α concentration inversely correlated with shrimp ovarian development. Together, our findings suggest that lowered PGFS gene expression resulted in decreasing concentrations of PGF 2α during shrimp ovarian maturation process. Therefore, we propose that the PGF 2α biosynthesis may be involved in the P. monodon ovarian development.
In conclusion, a prostanoid pathway in P. monodon is proposed based on the identification of nine prostanoid biosynthesis genes, two prostaglandins and three prostanoid precursors (Figure 9). In addition, PGF 2α biosynthesis may play an important role in P. monodon ovarian maturation because PGF 2α concentration and PmPGFS gene expression declined as the ovarian development progressed. Collectively, our knowledge of the P. monodon prostanoid pathway may lead to future applications in the black tiger shrimp aquaculture industry. More importantly, the identification of the P. monodon prostanoid biosynthesis genes also suggests the conservation of the prostanoid pathway between marine crustaceans and mammals.

Collection of shrimp samples
Wild female P. monodon broodstock (n = 25) were collected from the Andaman Sea, Thailand. The weight and length of broodstock were measured and recorded prior to dissection. Shrimp ovaries were dissected, weighed and either subjected to metabolite extraction or flash frozen in liquid nitrogen and stored at -80 °C for RNA extraction. The gonadosomatic index (GSI) of each shrimp was calculated using the following equation: ovarian weight/body weight x 100. Ovarian developmental stages were assigned according to GSI, separating the broodstock into stage I (GSI < 1.5), stage II (GSI = 2-4), stage III (GSI > 4-6) and stage IV (GSI > 6) [21].

Extraction of prostaglandins from shrimp tissues
Tissue homogenates underwent extraction twice with an equal volume of ethyl acetate. The solvent phases were then pooled together and dried under vacuum. Afterward, the dried crude extract was dissolved in 5% methanol. The solution was filtered to remove insoluble material before being loaded onto a 6-ml C18 SPE cartridge (VertiPak TM , Vertical Chromatography, Co., Ltd., Thailand), which was previously washed with 10 ml methanol and 10 ml water. Columns were then washed with 10 ml water, 4 ml hexane, and again with 10 ml water, before being eluted with 10 ml ethyl acetate. The eluate was then evaporated and dissolved in ethanol for subsequent HPLC analysis.

Identification of PUFAs and prostaglandins by RP-HPLC and mass spectrometry
Shrimp tissue extracts were separated by RP-HPLC using an Acclaim ® 120 C18 column (3 µm

RNA extraction and cDNA synthesis
Shrimp organs were homogenized and subjected to total RNA extraction using the Trizol reagent (Invitrogen, California, USA). mRNA was purified from total RNA using the Oligotex mRNA Mini Kit (QIAGEN, Maryland, USA). First strand cDNA was synthesized using the RevertAid TM First Strand cDNA Synthesis Kit with oligo (dT) 18 primer (Fermentas, Maryland, USA) according to the manufacturer's instructions.

Gene amplification
Initial PCR fragments were obtained using primers based on short gene sequences from The Black Tiger Shrimp EST Project at http://pmonodon.biotec.or.th [26] and the Marine Genomics Project at http://www.marinegenomics.org [24]. 5′and 3′-RACE-PCR were performed using the Advantage TM 2 PCR kit (Clontech, California, USA) and the SmartTM RACE cDNA amplification kit (Clontech). All PCR products were cloned into the pTZ57R/T vector (Fermentas), transformed into DH5α E. coli, and submitted for DNA sequencing (1 st -BASE, Malaysia). Identities of the obtained cDNA sequences were verified by TBLASTX analysis [52]. Multiple sequence Figure 9. The proposed prostanoid biosynthesis pathway in P. monodon. PUFAs, prostaglandins and full-length prostanoid biosynthesis genes identified in this study (black) were used to outline the P. monodon prostanoid pathway based on the previously published pathway in mammals. Prostanoids that have yet to be identified in P. monodon are shown in gray. doi: 10.1371/journal.pone.0076934.g009 alignment of prostaglandin biosynthesis genes were performed using CLUSTALX [53]. Rooted phylogenetic tree with branch length were performed using CLUSTALW [53]. Protein domains were predicted using Conserved Domain Architecture Retrieval Tool (CDART -http://www.ncbi.nlm.nih.gov/Structure/ lexington/lexington.cgi) [54].

Quantitative real-time PCR analysis
Quantitative real-time PCR analysis was performed using the SsoFast TM EvaGreen ® Supermix (Bio-Rad, California, USA) according to the manufacturer's instructions. Amounts of prostanoid biosynthesis gene transcripts relative to that of the house-keeping gene elongation factor 1α (EF1α) were obtained using the standard curve method [55]. The specificity of the PCR product was confirmed by agarose gel electrophoresis and melting curve analysis performed from 55 °C to 95 °C with a continuous fluorescent reading at 0.5 °C increments.

Identification of shrimp PGE 2 and PGF 2α by enzyme immunoassay (EIA)
Stage IV shrimp ovaries were harvested and homogenized in Hank's Balanced Salt Solution (HBSS) with an osmolarity of 720 mmol/kg (Sigma-Aldrich Inc., Missouri, USA). For the in vitro PGE 2 synthesis assay, ovary homogenates were incubated in a rotary shaker with 25 µM AA at 28 °C, 200 rpm and collected at 0, 30, 60, 120, 240, and 360 minutes postincubation. The homogenates were centrifuged at 12,000 x g for 2 min at 4 °C and the amounts of PGE 2 and PGF 2α in the supernatant were estimated using the prostaglandin E 2 EIA kit -Monoclonal and prostaglandin F 2α EIA kit (Cayman Chemical, Michigan, USA).

Statistical analysis
Statistical significant was assessed in this study using the Ttest with two samples assuming equal variances (P<0.05). Figure S1. Schematic representation of domain types and positions on each putative P. monodon prostanoid biosynthesis enzyme. Prostanoid biosynthesis gene sequences were submitted for the CDART analysis for domain prediction. Solid lines represent the total length of each predicted protein, while ovals and squares denote the conserved domains. C2 domain was first identified in phosphokinase C. TRX is thioredoxin-like superfamily domain. GST_C is glutathione transferase family, C-terminal alpha helical domain. MAPEG is Membrane-Associated Protein involved in Eicosanoid and Glutathione metabolism domain. EGF_CA is calcium-binding, EGF-like domain. An_peroxidaselike is animal heme peroxidases and related protein.

Supporting Information
(TIF) Figure S2. HPLC analysis and mass spectra of prostanoid precursors in shrimp ovary extract. Ovaries from 5 wild broodstock were homogenized in HBSS, pooled together, and incubated at 28 °C, 200 rpm for 1 h. The homogenate was extracted as described in materials and methods. The extract was then subjected to analysis by RP-HPLC and mass spectrometry. RP-HPLC elution profiles of commercially available prostanoid standards (A) and prostanoid precursors found in ovary homogenate (B) was obtained at 200 nm wavelength. Subsequent mass spectrometry analysis of prostanoid precursors in ovary homogenate revealed the mass spectra of EPA (C), DHA (D) and AA (E). (TIF) Figure S3. Phylogenetic trees constructed from PmPGES1 and related sequences from vertebrates and invertebrates. Predicted amino acid sequences from various organisms were obtained from GenBank and the Marine Genomics Project. Sequences were aligned using CLUSTALW multiple sequence alignment program and the rooted phylogenetic tree with branch length (UPGMA) were constructed. (TIF)