| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Center of Marine Biotechnology,3 University of Maryland Biotechnology Institute, Baltimore, Maryland 21202
Department of Zoology,4 University of Hong Kong, Hong Kong SAR, People's Republic of China
ABSTRACT
Blue crab vitellogenin (VTG) cDNA encodes a precursor that, together with two other Brachyuran VTGs, forms a distinctive cluster within a phylogenetic tree of crustacean VTGs. Using quantitative RT-PCR, we found that VTG was primarily expressed in the hepatopancreas of a vitellogenic female, with minor expression in the ovary. VTG expression in the hepatopancreas correlated with ovarian growth, with a remarkable 8000-fold increase in expression from stage 3 to 4 of ovarian development. In contrast, the VTG levels in the hepatopancreas and hemolymph decreased in stage 4. Western blot analysis and N-terminal sequencing revealed that vitellin is composed of three subunits of
78.5 kDa, 119.42 kDa, and 87.9 kDa. The processing pathway for VTG includes an initial hepatopancreatic cleavage of the primary precursor into 78.5-kDa and 207.3-kDa subunits, both of which are found in the hemolymph. A second cleavage in the ovary splits the 207.3-kDa subunit into 119.4-kDa and 87.9-kDa subunits. The hemolymph VTG profiles of mated and unmated females during ovarian development indicate that early vitellogenesis and ovarian development do not require mating, which may be essential for later stages, as VTG decreased to the basal level at stage 4 in the unmated group but remained high in the mated females. Our results encompass comprehensive overall temporal and spatial aspects of vitellogenesis, which may reflect the reproductive physiology of the female blue crab, e.g., single mating and anecdysis in adulthood.
behavior, ovary, oocyte development
Ovarian development and egg maturation are crucial processes for the success of reproduction. In crustaceans, as in other oviparous animals, ovarian development includes a growth process that consists of two consecutive phases: the primary phase is characterized by primary oocyte recruitment from oogonia [1], and the secondary phase features growth of oocytes as a result of the accumulation of yolk proteins and other cytoplasmic egg proteins [2]. Yolk proteins are the most important source of nutrients for developing embryos in oviparous animals and may constitute 60–90% of the total egg proteins [3]. Vitellogenin (VTG) is the precursor for the major yolk protein, vitellin (VT), which is a lipo-glyco-carotenoprotein. VTG in non-mammalian vertebrates and several invertebrates is produced in an extraovarian tissue and then transported as a high-density lipoprotein (HDL) into the ovary [4, 5]. VTG is serologically identical to VT and is found in the hemolymph of most crustacean species during ovarian development [6–12], including the female blue crab [13]. Thereafter, it is internalized into the oocytes through receptor-mediated endocytosis [14] and undergoes several modifications, such as specific proteolytic cleavage, to become VT [15].
It is well established that the livers of several vertebrate species [16, 17] and the fat bodies of insects [18] are the sites of VTG synthesis. However, the source of yolk proteins in crustaceans remains controversial, with evidence for hepatopancreatic [19–21], as well as intraovarian synthesis [3, 22–25].
Ovarian development, as well as reproductive physiology and behavior, have been the subjects of several studies in the female blue crab [26–28]. It has been established that the female blue crab undergoes a terminal molt at puberty and immediately mates thereafter, usually from May to October in the estuaries of Chesapeake Bay. After mating, the females migrate to high salinity waters, where egg maturation, fertilization, and spawning occur. Spawning usually takes place 1–2 mo after mating and can occur multiple times [26, 28]. A female-specific lipoprotein associated with vitellogenesis, which is composed of 48% lipid (mainly phosphatidylcholine), 50% protein, and 2% carbohydrate, has been detected in the hemolymph of blue crabs that are undergoing ovarian development [29]. Subsequently, two lipoproteins, which are cleavage products of ovarian VT, have been detected in developing blue crab embryos [30]. Furthermore, using in vitro incubation followed by Western blot or immunohistochemical analyses, it has been suggested that the ovary is the exclusive site of vitellogenesis in the female blue crab [31, 32].
During the last decade, molecular methods have been used to investigate vitellogenesis in crustaceans. The rapidly increasing list of characterized crustacean VTG genes includes those from penaeids [15, 23, 33–35], crayfish [36], prawns [19, 37], and two brachyuran species [38, 39]. From these reports, data have emerged indicating that VTG expression takes place in the hepatopancreas of Pleocyamata (including brachyurans), while in the Dendrobranchiata, both the hepatopancreas and the ovary express VTG [15, 33–35]. However, the relative contribution of each tissue during vitellogenesis, either at the mRNA or protein level, remains unclear. In addition, although VTG expression has been observed in the hepatopancreases of all the species mentioned above, the presence of the VTG protein in this tissue has yet to be demonstrated [15, 37].
The conclusion that the ovary is the sole site of vitellogenesis in the blue crab [31, 32] is not consistent with the latest suggestion of hepatopancreatic expression in Pleocyamata [37]. Therefore, we aimed to confirm and complement the original data, derived from protein analyses, with data obtained using a combined biochemical and molecular approach. In addition, because the blue crab spawning stocks in Chesapeake Bay (and along the US Atlantic coastline) have sustained a severe and persistent decline beginning in 1992 [40, 41], it is of interest to better understand vitellogenesis, ovarian development, and the control of these processes, so as to develop strategies for the recovery of this ecologically and commercially important species.
In the present study, we isolated the full-length cDNA of the blue crab vitellogenin from the female hepatopancreas and generated specific VT antibodies (anti-VT serum), to monitor the temporal and spatial profiles of vitellogenesis. More specifically, we measured VTG expression in the hepatopancreas and ovary, as well as the VTG levels in the hepatopancreas, ovary, and hemolymph during ovarian development. We also followed the cleavage pattern and transport of VTG from the hepatopancreas through the hemolymph to the ovary. In addition, a comparison of hemolymph VTG levels in mated and unmated females suggests the involvement of semen deposition in the regulation of vitellogenesis.
Final-stage juvenile female blue crabs (prepubertal) at the premolt stage were obtained from local fishermen in Chesapeake Bay or from the Aquaculture Research Center at the Center of Marine Biotechnology (Baltimore, MD). The crabs were held in 20-ppt artificial seawater at 22°C (16L:8D) and fed daily with a combination diet of frozen squid and pelleted sea bream (EWOS, Surrey, BC, Canada). Wild crabs were acclimated for 10–14 days before experimentation in a 4-m3 tank. For the monitoring of VTG hemolymph levels, females were kept individually in a 30 x 30 cm compartment in a 0.5-m3 tank. Animal collections and experiments were performed from late March to early October, 2005. Prepubertal (ripe prepubertal, approaching final molt) and pubertal females were distinguished based on rounded abdomen shape and other signs of the premolt stage [42, 43].
Ovarian developmental stage was determined based on ovarian weight and gamete size, according to criteria established by Lee and Puppione [29], in which stage 1 is previtellogenic, stages 2 to 5 are vitellogenic, and 6 to 8 are postspawning stages.
Hemolymph samples were diluted 1:1 with anticoagulant buffer (0.3 M NaCl, 30 mM sodium citrate, 26 mM citric acid, 2 mM EDTA [pH 7.4]) and stored at –20°C. Tissues for RNA extraction were dissected from ice-cold anesthetized animals, snap-frozen on dry ice, and stored at –80°C.
Amplification, Cloning, and Sequencing of VTG
Total RNA was extracted from the hepatopancreas and ovary of a vitellogenic female using Trizol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. Total RNA was quantified in a NanoDrop UV/visible spectrophotometer (NanoDrop Technologies, Wilmington, DE). First-strand cDNA was generated from 1 µg total RNA using 5'- and 3'- RACE (SMART RACE cDNA Amplification Kit; BD Biosciences, Mountain View, CA). The initial degenerate primers were designed according to the conserved amino acid domains MYKYVEA and GNMGVMTP of the VTG proteins of penaeids and C. feriatus (GenBank accession nos. AB191486, DQ288843, AY321153, AB176641, and AY724676). An amplicon of 718 bp was initially generated from hepatopancreas cDNA by PCR using the degenerate primers VitpF1 (5'-ATGTAYAARTAYGTNGARGC-3') and VitpR1 (5'-GGNGTCATNACNCCCATRTTNCC-3') and the Advantage cDNA polymerase mix (BD Biosciences), and this fragment was cloned in the TOPO TA cloning vector (Invitrogen). A 5'-RACE fragment was amplified using the reverse primer Vitfin1 (5'-CGCAGGCTTCTGGGCTCCAGCTC-3') and the adapter primer (from the kit) from a 5'-RACE hepatopancreas cDNA library, and analyzed similarly. A 6990-bp 3'-RACE fragment was amplified using primer Vitfin2 (5'-GAGCTGGAGCCCAGAAGCCTGCG-3') and the adapter primer with the GeneAmp XL PCR kit (Applied Biosystems, Branchburg, NJ). The amplicon was ligated into a vector using the TOPO TA XL cloning kit (Invitrogen). The full-length cDNA sequence of VTG was constructed from the overlapping cDNA clones.
For the phylogenetic analysis, a Neighbor Joining Tree was constructed using the MEGA phylogeny package [44] with the pair-wise deletion option and 250 bootstrap repeats. To choose the closest outgroup, a PSI Blast search was iterated three times [45] using the blue crab VTG sequence as a query, and the alignment was optimized manually according to the PSI Blast alignment results.
Purification of VT and Generation of Polyclonal Antibodies
Ovarian tissue (2 g) from a vitellogenic female blue crab (stage 4) was homogenized in 10 ml of extraction buffer (20 mM Tris [pH 7.5], 1 mM PMSF), cleared by centrifugation at 8000 x g for 10 min at 4°C, and precipitated in 50% ammonium sulfate. The pellet was resuspended in extraction buffer and the precipitation step was repeated twice.
The resulting pellet was resuspended in 1 ml of elution buffer (PBS [pH 7.5], 1 mM EDTA) and the proteins were separated on a 56 x 1.6 cm Bio-gel P-200 size exclusion column (Bio-Rad, Hercules, CA) at a flow rate of 0.1 ml/min. The protein elution profile was monitored at 280 nm and 3-ml fractions were collected. Protein fractions that were yellow in color were stored for further analysis. The purity of VT was verified on 4–15% PAGE under native and denaturing conditions and further confirmed by N-terminal sequencing of its subunits (see Supplemental Fig. 1 online at www.biolreprod.org and at http://combshare.umbi.umd.edu/zmora/zmora.html). This purified protein was then used to generate antiserum in rabbits (Sigma-Genosys Laboratories).
Total RNA (20 µg) from the ovaries and hepatopancreases of females at different stages of ovarian development were electrophoresed in a 1.2% agarose gel that contained formaldehyde, and the samples were transferred to a nylon membrane (Immobilon; Millipore Inc., Bedford, MA). For prehybridization, the membrane was incubated with hybridization buffer (50% formamide, 5x SSC, 7% SDS, 2 mM EDTA, 1% blocking reagent) for 3 h at 60°C, and then hybridized for 16 h with a digoxigenin (DIG)-labeled antisense riboprobe. The riboprobes, which were generated from the VTG cDNA clone using SP6 or T7 RNA polymerase, corresponded to nucleotides 206-1508 of VTG cDNA and were used at a final concentration of 20 ng/ml. After hybridization, the membrane was subjected to a 30-min wash in 2x SSC, 0.1% SDS at 68°C, a 30-min wash in 0.5x SSC, 0.1% SDS at 68°C, and a final wash in 0.1x SSC, 0.1% SDS at 68°C for 30 min. Hybridized probes were visualized with an AP-conjugated anti-DIG antibody (1:20 000) and the CDP-Star chemiluminescent detection system (Roche Diagnostics GmBH, Mannheim, Germany). The hybridized membrane was exposed to BioMax MS Kodak film for 5–30 min. Ribosomal RNA was visualized with ethidium bromide.
Sample preparation. Hepatopancreatic and ovarian tissues were fixed in Bouin solution for 24 h, and then dehydrated gradually through a series of increasing alcohol concentrations. Tissues were cleared and embedded in Paraplast according to conventional procedures. Sections (6-µm thickness) were prepared on 3-aminopropyl triethoxysilane (APTES)-coated slides.
Immunohistochemistry. Sections were deparaffinized, rehydrated, and rinsed in distilled water before incubation with 0.3% hydrogen peroxide in PBS (pH 7.4) for 30 min at room temperature. After washing in PBST (PBS, 0.5% Tween-20), the slides were blocked in 10% normal goat serum and incubated overnight with anti-VT serum diluted 1:3000 (or preimmune serum for the negative control), followed by incubation with HRP-conjugated goat anti-rabbit IgG (Vector Laboratories, Burlingame, CA) at 1:5000 for 1 h. Sections were then washed and incubated with 3,3'diaminobenzidine (DAB; Sigma Chemical Co., St. Louis, MO). Photographs were taken using an Olympus micro/DP70 camera mounted on a Zeiss light microscope (model Axioplan2), and converted into computer images using the Olympus DPController program.
In situ hybridization. Deparaffinized and rehydrated sections were incubated with 0.2 M HCl for 20 min, washed in PBS, treated with proteinase K (10 µg/ml in 50 mM Tris-HCl [pH 7.5], 50 mM EDTA) for 15 min, and acetylated in 0.1 M triethanolamine-HCl/0.25% (v/v) acetic anhydride. The sections were covered with 500 µl of hybridization buffer (50% formamide, 5x SSC, 50 µg/ml denatured salmon sperm DNA) and incubated for 2 h at 58°C. After prehybridization, the slides were incubated overnight at 58°C in fresh hybridization buffer that contained 40 ng/ml denatured antisense or sense DIG-labeled riboprobes, which were prepared as described above. After hybridization, the sections were washed for 30 min in 2x SSC at 25°C and for 1 h each in 2x SSC, 0.4x SSC, and 0.1x SSC at 65°C. The slides were then incubated in 1% blocking reagent [Roche] and 2% normal sheep serum in buffer I (100 mM Tris-Cl [pH 7.5], 150 mM NaCl) with AP-coupled anti-DIG antibody (Roche) (150 mU/ml in buffer I). Color was developed using NBT/BCIP plus suppressor (Pierce Biotechnology, Rockford, IL). The sections were examined and photographed as described above.
SDS-PAGE and Western Blot Analyses
Proteins from ovaries, hemolymph, and hepatopancreases of females at different stages of ovarian development were prepared as follows. Ovaries were homogenized in an extraction buffer (0.3 M sucrose, 140 mM Tris-HCl [pH 7.4]), precipitated in 50% ammonium sulfate, and resuspended in PBS. HDL was isolated from the hemolymph, to avoid interference of hemocyanin, as described by Shechter et al. [46]. Hepatopancreatic proteins were prepared as described for the ovarian preparation, layered onto 1.35 M sucrose, and centrifuged at 160 000 x g for 6 h at 4°C. The upper fraction was collected for analysis. Protein concentrations were determined using the RC Dc protein assay kit (Bio-Rad). Each protein sample (0.5 µg) was subjected to 7% SDS-PAGE and the gel was stained with Coomassie blue (GelCode, Bio-Rad) or transferred to a nitrocellulose membrane at 100 V for 1 h in transfer buffer (25 mM Tris [pH 8.3], 192 mM glycine, 20% methanol). After blocking in PBST that contained 5% skimmed milk, the membrane was incubated with anti-VT serum (1:20 000 in blocking buffer) for 1 h at room temperature. After three washes in PBST for 15 min, the membrane was incubated with HRP-goat anti-rabbit IgG (1:40 000). The signal was developed using the Super-Signal West-Pico chemiluminescence detection kit (Pierce) and detected by exposure to BioMax MS Kodak film for 30 sec to 2 min.
Preparation of VTG/VT Subunits for N-terminal Sequencing
Ovarian VT subunits were separated on 7% SDS-PAGE, as described above. In order to obtain 10 µg of the pure hemolymph and hepatopancreas VTG subunits, previously extracted HDLs were immunoprecipitated using anti-VT serum and Protein A magnetic beads (New England Biolabs, Ipswich, MA), according to the manufacturer's protocol. In brief, 30 µg hemolymph HDL (equivalent to 100 µl of hemolymph) and 1 mg hepatopancreatic HDL (equivalent to 25 mg of tissue) diluted four times with PBS that contained 1x protease inhibitor cocktail (Sigma) were preincubated with 25 µl of Protein A magnetic beads for 1 h at 4°C, in order to remove any proteins that were nonspecifically bound to Protein A. Once the beads were removed using a magnet, the sample was incubated with 20 µl anti-VT serum, followed by the addition of 20 µl Protein A magnetic beads, each for 1 h at 4°C. After incubation, the beads were retained on a magnetic rack and washed four times with 1 ml PBST. Proteins bound to Protein A were finally eluted by 3 min of heating at 70°C in 30 µl of 3x Laemmli sample loading buffer. After separation on 6% SDS-PAGE, the proteins were transferred to a PVDF membrane (Pierce) and stained with Coomassie blue. Positive bands with the expected molecular masses were excised and N-terminally sequenced on a Perkin Elmer/Applied Biosystems Edman sequencer at the synthesis and sequencing facilities of Johns Hopkins University (Baltimore, MD).
Purified VT was used to develop a competitive ELISA and to serve as a standard. The VT ELISA was based on the procedure of Lee and Watson [47], with a linear standard curve range of 50–1600 ng/ml and a detection limit of 20 ng/ml. Specifically, microplates (Costar Corp., Cambridge, MA) were coated overnight at 4°C with 100 µl of 200 ng/ml VT in 50 mM bicarbonate buffer (pH 9.6). After washing with PBST, the microplates were blocked for 1 h at 37°C with 200 µl of blocking buffer (PBST plus 1% BSA). Unknown and standard samples were incubated overnight at 4°C with anti-VT serum in blocking buffer at a final dilution of 1:20 000. Thereafter, the samples were dispensed into the wells (100 µl/well) and incubated for an additional 2 h at 37°C. After washing as described above, the plates were incubated with 100 µl of 1:5000 HRP-conjugated goat anti-rabbit IgG (Vector Laboratories, Burlingame, CA) in blocking buffer for 1 h at 37°C. Color was developed using 200 µl of 3 mg/ml 2,2'-azino-bis(3-ethylbenzo-acid-6-sulfonic acid) diammonium salt (ABTS; Sigma) in 0.1 M citrate buffer (pH 4.1) for 15–20 min at room temperature. Absorbance was measured at 405 nm using an automated microplate reader (Thermo-Max; Molecular Devices, Menlo Park, CA).
First-strand cDNA was generated from hepatopancreatic or ovarian total RNA as described above, using random hexamers as the primers and MMLV-RT (Promega, Madison, WI). Total RNA (3 µg) for each sample was subjected to DNase treatment (2 U of RQ1; Promega) to eliminate gDNA contamination. Standards were prepared from sense VTG cRNA generated from the VTG cDNA clone using T7 RNA polymerase (Roche Diagnostics) followed by purification in a ChromaSpin-100 column (BD Biosciences). The resulting sense cRNA was quantified using the RiboGreen RNA quantification kit (Molecular Probes, Eugene, OR) and converted into copy numbers based on the molecular mass of the RNA fragment. Standards that ranged from 480 to 4.8 x 106 copies were reverse-transcribed as described above.
Duplicate cDNA aliquots (20 ng of total RNA) from each sample served as templates in PCR with SYBR Green PCR core reagent (Applied Biosystems, Foster City, CA) that contained 200 nM of the gene-specific primers TaqvitF (5'-TGTACAGCTGAAAGGCGTGG-3') and TaqvitR (5'-CATGGGCCGAGAACAGTCA-3'). Amplification reactions were carried out with the ABI Prism 7700 Sequence Detection System at 50°C for 4 min, 95°C for 10 min, followed by 40 cycles of 95°C for 15 sec and 60°C for 60 sec. The copy number in each sample was determined by comparing CT (threshold cycle) values [48] to the standard (run on every plate) and normalized against the abundance of arginine kinase mRNA using the TaqAKF (5'-ACCACAAGGGTTTCAAGCAG-3') and TaqAKR (5'-GGTGGAGGAAACCTTGGACT-3') primers [49].
The data obtained from QRT-PCR and ELISA are presented as mean ± SEM. The results were examined using one-way ANOVA followed by the Tukey multiple range test. In all cases, statistical difference was accepted at P < 0.05.
The full-length cDNA sequence of VTG was determined by overlapping the 5'-RACE and 3'-RACE cDNA fragments. The complete VTG cDNA, which is composed of a 28-nt 5'-UTR, a 113-nt 3'-UTR, and a 7689-nt ORF, encodes a precursor of 2563 amino acids (aa), which includes an 18-aa signal peptide (GenBank accession no. DQ314748) with a predicted size of 282 kDa. The deduced amino acid sequence contains four putative subtilisin-like protein endopeptidase motifs (RXXR), nine putative N-glycosylation sites at amino acids 159, 658, 885, 979, 994, 1451, 1631, 1863, and 1938 (analyzed at http://www.cbs.dtu.dk/cgi-bin) but no O-glycosylation sites. Serine residues constitute 5% of the total amino acid composition of VTG, including one hepta-serine, three tri-serines, and twelve di-serines. The VTG cDNA shares 87% and 80% identity with the nucleotide and deduced amino acid sequence of the Japanese blue crab (P. trituberculatus) vitellogenin [38, 39], respectively. We also performed a phylogenetic analysis of the VTG proteins, to construct a Neighbor Joining Tree using the rainbow trout VTG as an outgroup, as described in Materials and Methods. The Brachyurans VTG formed a distinctive separate cluster from the other crustacean VTG proteins (Fig. 1).
|
Northern blot analysis (Fig. 2A) revealed that a VTG transcript of 7.8 kb in length was present only in the hepatopancreases of vitellogenic females and only at stages 2 and 3, not stage 1. No signal was detected in the ovary at any developmental stage (Fig. 2B), in female muscle, or in the male hepatopancreas (data not shown).
|
The spatial and cellular distributions of VTG expression (Fig. 3, A and B), as determined by RNA in situ hybridization, were congruent with the results of the Northern blot analysis, as shown in Figure 2. The hepatopancreas of a female at stage 3 showed positive hybridization with the antisense VTG riboprobe but not with the VTG sense probe (Fig. 3, A and B). This signal was evident in most of the epithelial cells of the hepatopancreas tubule. No signal was observed in stage 1 female or male hepatopancreases (Fig. 3, C and D).
|
Using immunohistochemistry (ICC), VTG was detected in the epithelial cells of hepatopancreatic tubules of a vitellogenic female at stage 3, as well as in the adjacent hemocytes (Fig. 3E), but not in male hepatopancreases (Fig. 3F). No positive signal was detected in previtellogenic hepatopancreas nor when preimmune serum was used (data not shown).
The results of Western blot analysis of HDLs extracted from the hepatopancreas, hemolymph, and total protein extracts of ovaries at stages 1 and 3 were stage-dependent, and differential band patterns were noted in these tissues (Fig. 4, A and B). In the hepatopancreas, three bands with molecular masses of 95 kDa, 80 kDa, and 78 kDa were detected by the anti-VT serum, while in the hemolymph, two bands were detected (Figs. 4, A and B, lanes 3 and 4). In addition, a faint 250-kDa band cross-reacted with the anti-VT serum (Fig. 4, A and B, lane 4). In the ovary, bands of 95 kDa and 78 kDa were recognized by the anti-VT serum (Fig. 4, A and B, lane 6).
|
N-terminal sequencing was performed with the resolved VT/VTG proteins obtained from all three tissues of the same female at stage 3 (a total of eight bands, numbered 1–8 in Figure 4C and a band of 250 kDa in the hemolymph). As stated earlier, hepatopancreatic and hemolymph samples were resolved in 6% SDS-PAGE, while the ovarian proteins were electrophoresed on a 7% gel. The resulting amino acid sequences are listed in Table 1. The initial amino acid sequence of band 4 of the hepatopancreas was XPYGG, which was consistent with the sequences of bands 6 and 8 of the hemolymph and ovary samples, respectively (Fig. 4C, lanes a–c). Band 5 of the hemolymph and ovarian band 7 were sequenced as SVD(X)AA and SVDYAA, respectively. The SVDYAA sequence agrees with the predicted amino acid residues at positions 732–737 that immediately follow the putative RERR cleavage site. Surprisingly, ovarian band 7 also yielded the sequence MEYTRSS, which corresponded to amino acid residues 1775–1782 of the deduced amino acid sequence of VTG (Fig. 4C, lane c).
|
Electrophoresis in 6% SDS-PAGE was chosen for the hepatopancreatic and hemolymph samples, since band 7 in the ovarian tissue samples gave two sequences (Fig. 4C, lane c). The sequence of band 1 of the hepatopancreas (Fig. 4C, lane a) started with SSSGQ, which corresponded to amino acids 1781–1785 of VTG, while band 2 sequenced as LYGPQY, which corresponded to amino acids 745–751 of VTG. Both proteins begin 6 and 15 amino acids after the expected MEYTRSS and SVDYAA motifs, respectively. The 250-kDa band did not produce a sequencing result (Fig. 4C, lane b), whereas band 3 started with the sequence DEPDGV (Fig. 4C, lane a), which is identical to the first six amino acids of cryptocyanin from the swimming crab Portunus pelagicus (GenBank accession no. ABM54471).
Late-prepubertal females approaching their final molt were divided randomly into two groups. After the final molt, the females in one group were allowed to mate and the others remained as virgins. Using a competitive ELISA, the VTG levels in the hemolymph were determined in samples collected semiweekly over 10 wk.
As shown in Figure 5, the hemolymph VTG levels in both groups were low (10–20 µg/ml) during the first 2.5 wk, then increased gradually until they peaked at 200 µg/ml by Week 5. At 9.5 wk, the VTG levels in the unmated females began to decline and returned to the basal level, while in the mated females, the extent of decrease was moderate, stabilizing at a significantly higher level of 100 µg/ml in the final 2 wk, which corresponded to ovarian stage 4 (Fig. 5).
|
The VTG levels in the hepatopancreas and ovary during vitellogenesis were measured. As shown in Figure 6, the VTG content of the hepatopancreas increased from 1.0 ± 0.3 (stage 1) to 7.0 ± 3.0 (stage 3) µg/mg of total protein, and then dropped to 3 µg/mg of total protein at stage 4.
|
As expected, the VT content of the ovary increased gradually during vitellogenesis from 128 ± 19 µg/mg at stage 1 to 478 ± 4 µg/mg at stage 4, which was equivalent to 10% to 50% of the total proteins (Fig. 6).
During the course of vitellogenesis, the VTG transcript levels in the hepatopancreas and ovary of the same animals were measured using QRT-PCR. VTG expression in the hepatopancreas, which was initially very low at stage 1 (1174 copies/20 ng total RNA), increased 900-fold at stages 2 and 3, followed by an additional dramatic increase of 73 000-fold (8.6 x 107copies/20 ng RNA) at stage 4 (Fig. 7A). The VTG expression levels in the ovary (although not detected by Northern blot analysis, Fig. 2) were generally 3000-times lower than those in the hepatopancreas, with the exception of stage 1, when the expression levels were similar. The general expression pattern in the ovary resembled that in the hepatopancreas (Fig. 7B).
|
In order to gain a better understanding of the reproductive physiology of the female blue crab, C. sapidus, we utilized a combined molecular and biochemical approach to study vitellogenesis. This complex process involves a network of tissues (hepatopancreas, hemolymph, and ovary) and includes the synthesis, transport, and processing of VTG, as well as the accumulation of VT during oocyte growth. The full-length cDNA of VTG was sequenced and a quantitative assay (QRT-PCR) to measure gene expression was developed. In addition, ovarian VT was purified, antibodies were generated, and a competitive ELISA was established.
The analysis of the VTG cDNA revealed that the ORF encodes a 2563-amino acid precursor with a predicted molecular mass of 282 kDa. This protein is most homologous (80%) to the VTG of another brachyuran species, the Japanese blue crab Portunus trituberculatus [38, 39]. Similar to other VTGs, the blue crab VTG contains multiple potential N-glycosylation sites, of which seven out of the nine are located downstream of amino acid 728 [36, 50]. Unlike vertebrates, crustacean VTGs, which included the blue crab VTG, lack the prominent motif characterized by a polyserine domain. Polyserine domains bind bivalent cations [36] and have previously been implicated in receptor binding [51, 52]. In crustaceans, the lower serine residue content (5% to 10%) is typically attributed to the general scarcity of polyserine domains (relative to most vertebrate VTGs) [36]. The serine content of blue crab VTG (8.5%) falls within this range.
A phylogenetic tree was constructed using the 11 crustacean VTGs currently available in GenBank, including the blue crab sequence, and the rainbow trout VTG as an outgroup. The tree showed two main clusters, Penaeids and Brachyurans, and clearly demonstrated that the blue crab VTG belongs to the family of crustacean VTGs. In addition, our data indicate specific lineage modifications for Brachyurans vs. Penaeids (Fig. 1). It is anticipated that additional crab VTGs will cluster with the three existing crab VTGs, allowing the construction of a more representative evolutionary distribution of crustacean VTGs.
The VTG cDNA was initially isolated from the hepatopancreas of a vitellogenic female. The results obtained in the Northern blot analysis (Fig. 2), in situ hybridization, and immunohistochemistry (Fig. 3) indicate that the hepatopancreas is the site of expression and translation of VTG. This conclusion contradicts the results of previous studies conducted at the protein level, which concluded that the ovary was the sole site of VTG production in the blue crab, while ruling out expression in the hepatopancreas [31, 32]. The ICC pattern, indicated by granular signals in the epithelial cells of the hepatopancreatic tubules (Fig. 3, E and F), is similar to those of C. quadricarinatus [46] and M. rosenbergii [53] and implies that VTG is present in a packaged form. Although the above methods suggest that the ovary is unlikely to express VTG, the highly sensitive method of QRT-PCR revealed the presence of VTG mRNA in this tissue, albeit at a level 3000-times lower than in the hepatopancreas. To ascertain that the QRT-PCR result was not an artifact, a 5'-RACE amplicon of 1.5 kb was amplified from the ovary, sequenced, and found to be identical to that prepared from the hepatopancreas (data not shown). Although the 3' portion may differ in length or sequence from the hepatopancreas form, these results suggest that vitellogenesis in the suborder Pleocyamata, as in Dendrobranchiata, originates in both the hepatopancreas and ovary. As alluded to earlier, the difference may lie in the relative VTG contribution of each tissue, which appears to be significantly lower in the ovary of Pleocyamata, at levels undetectable by most procedures.
It is widely accepted that the VT primary precursor (prepro-VTG) undergoes several proteolytic cleavages, to generate the subunits that comprise ovarian VT. Using immunoblotting and immunoprecipitation of HDL fractions, we were able to concentrate approximately 2000-fold the VTG/VT in the hepatopancreas and hemolymph (ovarian stage 3) for N-terminal sequencing. For example, 25 mg of fresh hepatopancreas yielded 10–15 µg of VTG, whereas 20 µg of VTG was obtained from 100 µl of hemolymph. While the immunoprecipitation procedure provided in pure form for N-terminal sequencing the VTG/VT subunits that are involved in the processing of VTG, it also yielded an unexpected result for the hepatopancreas sample. We anticipated fewer bands in the hepatopancreas, as the site of VTG synthesis, than in the hemolymph. To our surprise, four bands were evident in the hepatopancreas (Fig. 4C, lane a), while only two bands were noted in the hemolymph (Fig. 4C, lane b). Moreover, as listed in Table 1, the sequencing results for bands 1 and 2, compared with those of the ovarian samples, are missing 6 and 13 amino acids at the N-terminus, which may reflect aminopeptidase activity. It is possible that these differences resulted from the immunoprecipitation steps, since the hepatopancreas is the major digestive tissue with a variety of metabolic enzymes. Overall, seven out of the eight proteins detected by the anti-VT serum turned out to be subunits of VTG/VT, with the exceptions of an 80-kDa protein (Fig. 4C, band 3) in the hepatopancreas, which was identified as cryptocyanin, and an unknown 250-kDa protein in the hemolymph.
Consensus proteolytic motifs (RXXR) specifically recognized by the subtilisin-like endopeptidase or convertase are distributed throughout the VTG protein sequences. However, only one of these motifs has been proven to be utilized and is conserved among crustacean VTGs. N-terminal amino acid sequencing of the different VT subunits illustrated that this functional RXXR motif is positioned at 710–728 of other crustacean VTGs [15, 33, 34, 37, 38] Cleavage gives rise to a 78–90-kDa protein, which is homologous to invertebrate and vertebrate lipoproteins [54, 55] and fish VTG [56, 57].
The C. sapidus VTG possesses a functional RERR motif at amino acids 728–731, which results in a calculated subunit mass of 78.5 kDa for the subunit beginning with APYGGTTQ (http://bioinformatics.org/sms2/protein_mw.html) and a calculated molecular mass of 207.3 kDa for the remainder of the VTG, starting at SVDYAA (Table 1). As presented in Table 1, this result is expected and is in agreement with the bands seen for the hemolymph (Fig. 4C, band 5) and ovary (Fig. 4C, band 7). Since only two bands were observed in the hemolymph, it appears that the first cleavage of VTG occurs in the hepatopancreas immediately after synthesis, prior to its secretion into the hemolymph. From a comparison of the banding patterns of hemolymph and ovary (Fig. 4C, lanes b and c), it seems that the second cleavage takes place in the ovary, splitting band 5 to band 7, which contains two subunits. Based on our N-terminal sequencing results (Table 1), this second cleavage was expected to produce two bands with calculated molecular masses of 119.4 kDa and 87.9 kDa. However, in SDS-PAGE (Fig. 4, A and C), bands 2 (hepatopancreas) and 7 (ovary) migrated below the 100-kDa marker, indicating the molecular masses of less than 100 kDa for both subunits. This may be due to glycosylation and conjugation of lipid to this subunit, which contains many putative glycosylation sites, thereby affecting its mobility in SDS-PAGE. Interestingly, while Lee and Walker [29] observed an additional 107-kDa lipoprotein in the hemolymph, only the subunits of VTG (207.3 kDa and 78.5 kDa) and a minor band (250 kDa) of unknown sequence were found in the current study. In M. rosenbergii, the cleavage of the larger VTG subunit occurs in the hemolymph only at advanced vitellogenic stages [37]. Likewise, it is possible that cleavage of the 207.3-kDa subunit (band 5) in the blue crab hemolymph occurs after stage 4, although this motion was not tested in the present study. Based on the N-terminal sequencing results, a putative model for the processing of C. sapidus VTG is proposed in Figure 8. Since the same cleavage sites have been reported in the Penaeus semisulcatus VTG subunits [15], it can be concluded that this is a common processing pattern for crustacean VTGs.
|
The prepro-VTG, with a predicted molecular mass of 282 kDa, was not detected in the hepatopancreas (the established site of synthesis) by staining or immunoblotting (Fig. 4, A-C). This may be due to rapid processing of this protein immediately after synthesis. Alternatively, since the anti-VT serum was generated against VT, the targeted epitopes on this precursor may be unavailable for interaction with the anti-VT serum. Cleavages, glycosylations, conjugation of lipids, and consequent refolding are involved in the conversion of prepro-VTG to VT, as well as in the general formation of epitopes [58, 59]. The significance of the tertiary structure of the VTG subunits has been demonstrated in the turtle Chinemys reevesii [60], in which four different monoclonal antibodies did not recognize the denatured forms of the VTG subunits.
Similar to the hemolymph VTG (Fig. 5), the hepatopancreatic VTG levels increased significantly in stage 3 and waned in stage 4 (Fig. 6). Concomitantly, VT accumulated in the ovary (Fig. 6). However, the VTG transcripts in the hepatopancreas showed a different pattern, with a sharp increase at stage 4 (Fig. 7A) after relatively moderate increases at stages 2 and 3 of vitellogenesis. This discrepancy between the transcript level and protein content has not been observed previously in crustaceans. Instead, in P. japonicus hepatopancreas, the VTG mRNA levels decreased at the equivalent stage of vitellogenesis [34, 61].
Monitoring VTG levels in the hemolymph allowed us to determine the time required to reach each stage of ovarian development under our experimental conditions. By comparing the levels of VTG in the hemolymph samples of mated and unmated females, we observed some differences in the patterns of vitellogenesis. The effect of semen on ovarian development is of interest, since the female blue crab mates only once in her lifetime, stores the sperm in spermathechae, and uses it for several rounds of fertilization [26]. Since the regulation of vitellogenesis by peptides in semen, which simultaneously decrease female receptivity to mating and enhance vitellogenesis, has been reported in Drosophila [62], the regulation of vitellogenesis in the blue crab by factors in the semen is plausible.
In the present study, we have shown that: 1) blue crab VTG is expressed mainly in the hepatopancreas, 2) prepro-VTG undergoes two cleavages, one in the hepatopancreas and a second one primarily in the ovary, 3) VT may be serologically different from its hepatopancreatic primary precursor, 4) the VTG mRNA levels in the hepatopancreas do not correlate with VTG protein levels at stage 4 in the hepatopancreas and hemolymph, and 5) the onset and advancement of vitellogenesis in the female blue crab are not dependent upon mating. However, the fate of the ovary at stage 4 is probably dependent upon semen-borne factors.
Vitellogenesis in most crustaceans of the Pleocyamata and Dendrobranchiata suborders is co-ordinated with the molting cycles, although the pattern varies across species [63]. In some cases, the two processes are tightly linked [10] and can affect each others duration [64, 65]. Therefore, it is believed that vitellogenesis and molting are controlled by the same regulatory factors, which include neuropeptides of the XO-SG complex [31, 66, 67], ecdysteroids [68–70], and juvenoid hormones [38, 71]. Of great interest are the few species in which the molting cycle is halted upon puberty, the female blue crab being among them. On the one hand, we have found that vitellogenesis, although not directly related to molting, shares some features with other crustacean species, such as the site of vitellogenesis and processing of VTG and ovarian development. On the other hand, we have also observed a unique increase in VTG transcript levels in the hepatopancreas and ovary at stage 4 and a dependency on mating for the completion of egg maturation. These differences may be related to the halt in molting and a different mode of regulation. Nevertheless, the present study establishes the basis for studying the mechanism of regulation of vitellogenesis in this species.
ACKNOWLEDGMENTS
We thank Oded Zmora for advice and help in the execution of the project, John Stubblefield for reading the manuscript, and Sarah Bembe and Bridget Bystry for technical assistance with the animals. We also thank Professors Richard Lee (Skidaway Institute of Oceanography, Savannah, Georgia) for sharing the N-terminal sequences of VTG, Allen Place (COMB) for advice on VT purification, and Dr. Zeev Pancer for guidance with the phylogenetic analysis. This article is contribution #06–142 from the Center of Marine Biotechnology (University of Maryland Biotechnology Institute, Baltimore, MD).
FOOTNOTES
1Supported by a Program Grant (NA17FU2841) from the NOAA Chesapeake Bay Office to the Blue Crab Advanced Research Consortium. 
Correspondence: 2J. Sook Chung, Center of Marine Biotechnology, University of Maryland Biotechnology Institute, 701 E. Pratt Street, Baltimore MD 21202. FAX: 410 234 8896; e-mail: chung{at}umbi.umd.edu
Received: 10 July 2006.
First decision: 28 August 2006.
Accepted: 14 March 2007.
REFERENCES
| ||||||||||||||||||||||||
0.05; **P 
