Isolation and characterization of genes functionally involved in ovarian development of the giant tiger shrimp Penaeus monodon by suppression subtractive hybridization (SSH).

Suppression subtractive hybridization (SSH) libraries between cDNA in stages I (previtellogenic) and III (cortical rod) ovaries of the giant tiger shrimp (Penaeus monodon) were established. In all, 452 ESTs were unidirectionally sequenced. Sequence assembly generated 28 contigs and 201 singletons, 109 of which (48.0%) corresponding to known sequences previously deposited in GenBank. Several reproduction-related transcripts were identified. The full-length cDNA of anaphase promoting complex subunit 11 (PmAPC11; 600 bp with an ORF of 255 bp corresponding to a polypeptide of 84 amino acids) and selenoprotein Mprecursor (PmSePM; 904 bp with an ORF of 396 bp corresponding to a polypeptide of 131 amino acids) were characterized and reported for the first time in penaeid shrimp. Semiquantitative RT-PCR revealed that the expression levels of PmSePM and keratinocyte-associated protein 2 significantly diminished throughout ovarian development, whereas Ser/Thrcheckpoint kinase 1 (Chk1), DNA replication licensing factor mcm2 and egalitarian were down-regulated in mature ovaries of wild P. monodon (p < 0.05). Accordingly, the expression profiles of PmSePM and keratinocyte-associated protein 2 could be used as biomarkers for evaluating the degree of reproductive maturation in domesticated P. monodon.


Introduction
The giant tiger shrimp (Penaeus monodon) is one of the most economically important cultured species (Bailey-Brock and Moss, 1992;Rosenberry, 2001). Breeding P. monodon in captivity, besides being difficult (Withyachumnarnkul et al., 1998;Wongprasert et al., 2006), is very much restricted by the current dependency on wild-caught broodstock, with the consequential overexploitation of high-quality sources in the wild. As a result, aquacultural production of P. monodon has undergone a significant decline over the last several years (Limsuwan, 2004).
The low degree of reproductive maturation of captive P. monodon has also limited the ability to genetically im-prove this important species by domestication and selective breeding programs (Withyachumnarnkul et al., 1998;Kenway et al., 2006;Preechaphol et al., 2007). Eyestalk ablation is used commercially to induce ovarian maturation in penaeid shrimp but the technique leads to an eventual loss in egg quality and death of the spawners (Benzie, 1998). Therefore, predictable maturation and spawning of captive penaeid shrimp without the use of eyestalk ablation is a long-term goal for the industry (Quackenbush, 1992).
Basic information on ovarian development is somewhat limited in this shrimp. Initial steps towards an understanding of the molecular mechanisms involved in ovarian and oocyte development in this economically important species, are the identification and characterization of genes differentially expressed in the diverse stages of the process (Preechaphol et al., 2007). tide sequences of 743 EST (70.7%) significantly matched known genes previously deposited in GenBank (E-value < 10 -4 ), whereas 308 ESTs (29.3%) were regarded as newly unidentified transcripts (E-value > 10 -4 ). A total of 559 transcripts (87 contigs and 472 singletons) were obtained after sequence assembly. Several reproduction-related genes, viz., chromobox protein, ovarian lipoprotein receptor, progestin membrane receptor component 1 and ubiquitin-specific proteinase 9, X chromosome, were isolated and characterized (Preechaphol et al., 2007).
Suppression subtractive hybridization (SSH) is widely used for isolating differentially expressed genes in any two closely related samples, specimens or species (Diatchenko et al., 1996). This technique should facilitate the identification of genes involved in ovarian (and oocyte) development. The genes identified could further assist in the domestication and selective breeding programs of P. monodon.
In order to provide a further insight into the molecular mechanisms involved in the reproductive maturation processes of P. monodon, we carried out SSH of genes expressed in stages I and III ovaries of wild P. monodon. The expression profiles of five reproduction-related genes during ovarian development in wild P. monodon broodstock were further examined using semiquantitative RT-PCR. Candidate biomarkers for evaluating the degrees of reproductive maturation in captive shrimp are reported herein.

Experimental animals
Four-month-old juveniles of P. monodon, with body weights of approximately 25-30 g, were purchased from a commercial farm in Chachoengsao (eastern Thailand). These were cultured in 15 ppt seawater at ambient temperature (28-32°C) and a natural daylight cycle. Broodstock shrimp, with body weights of > 200 g, were wild-caught from Satun, located in the Andaman Sea, west of peninsular Thailand. Prior to SSH library construction, ovaries were dissected out from two broodstock and weighed. The gonadosomatic index (GSI), i.e., ovarian weight/body weight x 100, of each shrimp was calculated. In order to determine expression profiles of reproduction-related genes during P. monodon ovarian development, female juveniles and the broodstock were acclimated at normal farm conditions (28-30°C, natural daylight and 35 ppt seawater) for 2-3 days. Ovarian developmental stages of broodstock were classified according to GSI: < 1.5, 2-4, > 4-6 and > 6% for previtellogenic (I), vitellogenic (II), early cortical rod (III) and mature (IV) ovaries (N = 4 for each stage), respectively. Ovaries were dissected from each shrimp immediately after collection and kept at -80°C until use.

Isolation of total RNA and mRNA
Total RNA was extracted from various tissues of each individual with TRI-Reagent (Molecular Research Center) and mRNA was further purified using a QuickPrep Micro mRNA Purification Kit (GE Healthcare). Total RNA and mRNA were kept under absolute ethanol at -80°C, prior to reverse transcription.

Construction of suppression subtractive hybridization (SSH) cDNA libraries and EST analysis
Initially, two micrograms of mRNA from the ovaries of the P. monodon broodstock were reverse-transcribed. Suppression subtractive hybridization (SSH) between cDNA from stages III (GSI = 5.69%) and I (1.43%) and vice versa (Diatchenko et al., 1996) was carried out using a PCR-Select cDNA Subtraction Kit (BD Clontech). The subsequent products were ligated to pGEM-T Easy vector and transformed into E. coli JM109. Plasmid DNA was extracted from clones carrying > 300 bp inserts and unidirectionally sequenced using the M13 reverse primer. Sequencing data were pre-processed to remove low-quality sequences (sequence length < 100 bp, the percentage of undetermined bases > 3% and low complexity), by using SeqClean with option-A to disable the trimming of poly A tail. Repetitive sequences matching the RepBase dataset were masked by using RepeatMasker. Sequence clustering and assembly was done using TIGR Gene-Indices Clustering Tools (TGICL) (Pertea et al., 2003) with CAP3 (Huang and Madan, 1999). Nucleotide sequences of assembled and non-assembled ESTs were compared with GenBank data using BlastN and BlastX (Altschul et al., 1990). Significantly matches to nucleotides/proteins were considered when the E-value was < 1 x 10 -4 . Blast2GO was used for the additional annotation of biological activities in BlastX matched sequences, thereby enabling Gene Ontology (GO) prediction of sequence data for which no GO annotation is, as yet, available (Conesa et al., 2005).
ESTs representing P. monodon selenoprotein M precursor (PmSePM) and anaphase promoting complex subunit 11 (PmAPC11) were further sequenced from the reverse direction of the original cDNA clones by using a M13 forward primer.

Semiquantitative RT-PCR
Expression profiles of keratinocyte-associated protein 2, Ser/Thr checkpoint kinase 1, DNA replication licensing factor mcm2, PmSePM and egalitarian during ovarian development of P. monodon broodstock were analyzed by way of semiquantitative RT-PCR. EF-1a was included as the positive control. Initially, nonquantitative RT-PCR  was carried out using 100 ng of first-strand cDNA as the template, with varying concentrations of primers (0.05, 0.10, 0.15, 0.20, 0.25, 0.30 and 0.40 mM, respectively). Primer sequences are listed in Table 1. Optimal concentrations of MgCl 2 (1.0, 1.5, 2.0, and 3.0 mM) were further selected using an optimized primer concentration. Finally, RT-PCR of these genes was undertaken with an optimized primer and MgCl 2 concentrations for 20, 22, 24, 27, 30 and 35 cycles. The number of cycles before the product reached an amplification plateau was chosen.
Semiquantitative RT-PCR was undertaken with 1.5 mM of MgCl 2 and 0.2 mM of primers for the respective target genes, 0.15 mM of primers for egalitarian and 0.10 mM of those for EF 1-a, as follows: 94°C for 3 min followed by appropriate cycles (22,27,24,22 and 24 cycles for the target genes and 22 cycles for EF 1-a, respectively) of 94°C for 30 s, 53°C for 45 s and 72°C for 45 s and a final extension at 72°C for 7 min. The amplicon was electrophoretically analyzed through 1.5% agarose gels, and visualized with a UV transilluminator after ethidium bromide staining (Sambrook and Russell, 2001). The intensities of the targets and EF-1a were quantified from the gel photograph using the Quantity One software (BioRad), and relative expression levels of investigated transcripts (intensity of targets/intensity of EF-1a) in all experimental groups of P. monodon were statistically tested using analysis of variance (ANOVA), followed by the Duncan's new multiple range test. Results were considered significant when p < 0.05. The ovaries from five groups of shrimp (juveniles and stages I, II, III and IV broodstock, N = 4 for each group) were assayed for expression analysis.

Results and Discussion
An understanding of the roles of genes functionally involved in ovarian and oocyte development should ultimately lead to a plausible approach for inducing reproductive maturation in P. monodon. In this study, 220 and 232 clones, respectively, from the forward (cDNAs from stage III ovaries as the tester and those from stage I ovaries as the driver; GenBank accession no. GW775090-GW775309) and reverse (cDNAs from stage I ovaries as the tester and those from stage III ovaries as the driver; GenBank accession no. GW775310-GW775541) SSH ovarian libraries of P. monodon were unidirectionally sequenced and 136 (61.8%) and 133 (57.3%) ESTs, respectively, significantly matched known sequences in GenBank (E-value < 10 -4 , Tables 2 and 3). Homologues of thrombospondin (TSP; 39 ESTs accounting for 17.7% and 26 ESTs accounting for 11.2% of sequenced clones) and peritrophin (39 ESTs, 17.7% and 27 clones, 11.6%) were abundantly represented in both libraries similar to results from analyses of the conventional cDNA library of vitellogenic ovaries of P. monodon (79 and 87 clones accounting for 7.5 and 8.3% of clones sequenced, respectively; Preechaphol et al., 2007).
After sequence assembly, 16 contigs (from 97 ESTs) and 123 singletons were obtained for the forward and 14 contigs (from 142 ESTs) and 90 singletons for the reverse SSH libraries, respectively. In all, 229 transcripts (28 contigs from 251 transcripts and 201 singletons, i.e., 44.5%) were obtained when both libraries were analyzed simultaneously, of which 109 significantly matched known genes in GenBank (E-value < 10 -4 ). Disregarding contigs repre- 678 ESTs in ovaries of Penaeus monodon Table 1 -Nucleotide sequences of primers used for expression analysis of keratinocyte-associated protein 2, Ser/Thr checkpoint kinase 1, DNA replication licensing factor mcm2, selenoprotein M precursor and egalitarian in ovaries of wild P. monodon broodstock.

Gene Primer sequence
Keratinocyte-associated protein 2 F: 5'-CTGCTGTAAACAATCTGGAAAAC-3' DNA replication licensing factor mcm2 F: 5'-TCAAGCGAGACAACAACGAACT-3'  The percent distribution of nucleotide sequences, according to GO categories of SSH ovarian cDNA libraries of P. monodon, was analyzed (Figure 1). In the category`biological process', ESTs involved in metabolic processes were predominant (e.g. anaphase promoting complex subunit 11, S-adenosylmethionine synthetase and T-complex protein 1 subunit epsilon, i.e., 35.0% of the examined ESTs), followed by those involved in cellular processes (e.g. acidic p0 ribosomal protein, DNA replication licensing factor mcm2 and coatomer protein complex subunit beta, i.e., 25.2% of the examined ESTs). Reproductionrelated ESTs (e.g RNA binding motif protein 4, neuralized protein, dynein and egalitarian) were found in 2.4% of the examined sequences of combined SSH data. This discovery rate is higher than that of the conventional ovarian cDNA libraries of P. monodon (1.7%; Preechaphol et al., 2007).
As for the category`cellular component', ESTs functionally involved in the cell part (e.g. myosin II essential light chain, ATP synthase E chain and Ser/Thr checkpoint kinase 1, i.e., 35.5% of the examined ESTs) predominated, followed by those functionally displayed in organelles (e.g. selenoprotein M precursor, keratinocyte-associated protein 2 and interleukin enhancer binding factor 2; 25.5% of the examined ESTs).
The highly organized eukaryotic cilia and flagella contain approximately 250 proteins (Inaba, 2003). They are constructed around evolutionarily conserved microtubulebased structures called axonemes (nine outer doublet microtubules, dynein arms, a central pair of microtubules and radial spokes) (Luck, 1984;Dutcher, 1995;King, 2000). Dynein is functionally related to the transport of var- Preechaphol et al. 681 Figure 1 -The percent distribution of nucleotide sequences in the SSH ovarian cDNA library of P. monodon according to three principal GO categories: A, biological process; B, cellular components and C, molecular functions, respectively.
ious cytoplasmic organelles (Aniento et al., 1993). In Drosophila, egalitarian binds to the dynein light chain. Point mutations that specifically inhibit Egl-Dlc association disrupt microtubule-dependant trafficking both to and within the oocyte, thereby resulting in a loss of oocyte fate maintenance and polarity (Carpenter, 1994). The physiological role of carbonyl reductase was thought to be an NADPH-dependent reduction in a variety of endogenous and foreign carbonyl compounds. However, increasing evidence indicates its involvement in steroid metabolism. In ayu, its localization in ovaries, enzymatic characteristics and transcriptional increase with oocyte maturation, infer its additional function as 20b-HSD in the production of maturation inducing hormones (MIH) (Tanaka et al., 2002).
The DNA replication (or origin) licensing system is prominant in ensuring precise duplication of the genome in each cell cycle, besides being a powerful regulator of metazoan cell proliferation (Eward et al., 2004). The protein kinase Chk1 plays a role in checkpoint control. Recombinant Xenopus Chk1 phosphorylates the mitotic inducer Cdc25 in vitro at multiple sites. Nevertheless, only XChk1-catalyzed phosphorylation of Cdc25 at Ser-287 is sufficient to confer the binding of 14-3-3 proteins (Kumagai et al., 1998). Moreover, the meiotic maturation of oocytes is regulated by the maturation promoting factor (MPF), a complex of cdc2 (Cdk1), cyclin B and other Cdk/cyclin complexes (Kobayashi et al., 1991;Kishimoto, 1999Kishimoto, , 2003. Chk1-dephophorylated Cdc25 activates MPF, thereby causing meiotic resumption in oocytes (Kishimoto, 2003).
Recently, the full length cDNA of keratinocyteassociated protein 2 was isolated in the Pacific white shrimp (Litopenaeus vannamei), although the function of this protein is still unknown. Moreover, its expression was altered following infection by the White Spot Syndrome Virus, WSSV (Clavero-Salas et al., 2007).
The full length cDNAs of anaphase promoting complex subunit 11 (biological process GO:0008152; GenBank accession no. GW775392) and selenoprotein M precursor (cellular component GO:0005783; GenBank accession no. GW775333) were hereby reported and identified for the first time in penaeid shrimp.
The anaphase promoting complex subunit 11 of P. monodon (PmAPC11) was 600 bp in length, and consisted of an ORF of 255 bp corresponding to a polypeptide of 84 amino acids, with 5' and 3' UTRs of 1 and 387 bp, respectively ( Figure 2A). The closest similar sequence to PmAPC11 was the anaphase promoting complex subunit 11 homolog of Tribolium castaneum (E-value = 1 x 10 -41 ). The predicted molecular mass and pI of the deduced PmAPC11 was 9.84 kDa and 7.99, respectively. Activation of the anaphase-promoting complex (APC) by Cdc20 enables anaphase initiation and exit from mitosis (Kramer et al., 1998;Lorca et al., 1998).
The selenoprotein M precursor of P. monodon (PmSePM) was 904 bp in length, and consisted of an ORF of 396 bp, corresponding to a polypeptide of 131 amino acids, and 5' and 3' UTRs of 6 and 502 bp, respectively (Figure 2B). It significantly matched the selenoprotein M precursor of L. vannamei (E-value = 2 x 10 -58 ). The predicted molecular mass and pI of the deduced PmSePM protein was 15.10 kDa and 7.75, respectively. PmSePM contained a signal peptide located between A 21 and E 22 , as well as a Sep15_SelM domain (positions 31-107, E-value = 1.9 x 10 -34 ) that exerts the thiol-disulphide isomerase activity functionally involved in disulphide bond formation of proteins in the endoplasmic reticulum (ER) (Ferguson et al., 2006).
In addition, the EST representing selenophosphate synthetase, an enzyme involved in selenocysteine biosynthesis, was also identified. In humans, selenium deficiency leads to male infertility and susceptibility to viral infections. More than 20 selenoproteins have been identified in higher eukaryotes (Guimaraes et al., 1996;Rayman, 2000;Korotkov et al., 2002) but their functions in ovarian/oocyte 682 ESTs in ovaries of Penaeus monodon development of P. monodon remain unknown. The analysis of baseline information, acquired as part of this study addresses the paucity of data and should provide a better understanding of reproductive maturation in cultured female P. monodon.
To address the functional involvement of various genes during ovarian development of P. monodon, the expression profiles of keratinocyte-associated protein 2, Ser/Thr Chk1, DNA replication licensing factor mcm2, PmSePM and egalitarian were examined by semiquantitative RT-PCR analysis. The control gene (EF-1a) seemed to be comparably expressed in all the groups of samples examined, thereby inferring its acceptability for use in normalizing target gene expression. All transcripts were more abun-dantly expressed in the ovaries of broodstock than juveniles (p < 0.05, Figure 3). The expression level of PmSePM peaked in stage I (previtellogenic) of development (GSI < 1.5), to progressively and significantly decrease in stages II (vitellogenic), III (cortical rod) and IV (mature) (p < 0.05). Likewise, keratinocyte-associated protein 2 was initially down-regulated in stage III, and subsequently, stage IV (p < 0.05). The expression of Ser/Thr Chk1, DNA replication licensing factor mcm2 and egalitarian during stages I, II and III, was comparable (p < 0.05), although down-regulated in the final stage of ovarian development in wild P. monodon broodstock (p < 0.05, Figure 3).
In various animals, a wide variety of maternal mRNA is generally transcribed at the early oogenesis stage, to then Preechaphol et al. 683 be stored in oocytes and carried into fertilized eggs (Qiu et al., 2008;Nishimura et al., 2009). Several reproductionrelated genes that are up-regulated during the ovarian development of P. monodon, for example, Ovarian-Specific Transcript 1 (Pm-OST1) and cyclin B (PmCyB), have been previously reported Visudtiphole et al., 2009). The down-regulation of keratinocyteassociated protein 2, Ser/Thr Chk1, DNA replication licensing factor mcm2, PmSePM and egalitarian implied that lower levels of these gene products may be necessary for the development and final maturation of P. monodon oocytes. The findings facilitate the possible use of RNA interference (RNAi) for studying their functional involvement in P. monodon ovarian development. Moreover, the expression profiles of keratinocyte-associated protein 2 and selenoprotein M precursor are potentially applicable as biomarkers to indicate degrees of reproductive maturation in the domesticated shrimp. In this study, genes expressed in ovaries of P. monodon were identified by SSH analysis. The expression profiles of reproduction-related transcripts were examined. Further studies of the molecular mechanisms of those genes and proteins involved in controlling each stage of oocyte maturation should be carried out, to reach a better understanding of the reproductive maturation of P. monodon in captivity.