In Crustacea, reproductive function and mechanisms regulating vitellogenesis have not been fully elucidated. This is due in great part to a lack of information concerning the biochemical nature of the vitellogenin molecule, the hemolymph precursor of yolk protein, vitellin, as well as the functional expression of the vitellogenin-encoding gene. We have therefore cloned a cDNA encoding vitellogenin in the kuruma prawn, Penaeus japonicus based on the N-terminal amino acid sequence of the 91 kDa subunit of vitellin. The open reading frame of this cDNA encoded 2,587 amino acid residues. This is the first investigation reporting a full-length cDNA and its corresponding amino acid sequence for vitellogenin in any crustacean species.
Northern blot analysis and in situ hybridization have revealed that mRNA encoding vitellogenin was expressed in both the follicle cells in the ovary and the parenchymal cells in the hepatopancreas. In nonvitellogenic females, vitellogenin mRNA levels were negligible in both the ovary and hepatopancreas, but in vitellogenic females, levels were dramatically increased in both tissues. In the ovary, highest levels were observed during the early exogenous vitellogenic stage, and thereafter rapidly decreased, whereas in the hepatopancreas, high levels were maintained until the onset of the late vitellogenic stage. Differing profiles of vitellogenin mRNA levels in the ovary and hepatopancreas suggest that the contribution of these tissues to vitellogenin synthesis harbor separate and complementary roles during vitellogenesis.
INTRODUCTION
In Crustacea as in other oviparous animals, a large amount of vitellin (Vt), the major yolk protein appearing in eggs and consumed as an important source of nutrients during embryogenesis, is firstly accumulated in oocytes as an important step during ovarian maturation. This physiological process is known as vitellogenesis, and its regulatory mechanisms have been well-defined in oviparous vertebrates and insects. However, in Crustacea, much less is known concerning the regulation of vitellogenesis and associated reproductive function, although a great many species are important targets of fisheries and aquaculture development.
The Vt molecule is derived from vitellogenin (Vg), a yolk protein precursor which is secreted into the circulation from the ovary itself or from extraovarian tissue(s). Vt and Vg are both lipoproteins of high molecular weight which are indistinguishable from one another immunologically and electrophoretically. It has been shown that Vg is produced in the liver in vertebrates, and in the fat body in insects (Raikhel and Dhadialla, 1992). In isopods and amphipods, investigations using immunological techniques have shown that Vg is present in significant quantities in the fat body (Charniaux-Cotton and Payen, 1988; Meusy and Payen, 1988). In decapod crustaceans, three tissues have been reported as possible sites of Vg synthesis: the adipose tissue in Parapenaeus longirostris (Tom et al., 1987) and Scylla serrata (Rani and Subramoniam, 1997); the hepatopancreas in Carcinus maenas and Libinia emarginata (Paulus and Laufer, 1987), Penaeus vannamei (Quackenbush, 1989a, b), Penaeus semisulcatus (Fainzilber et al., 1992), Macrobrachium nipponense (Han et al., 1994), and S. serrata (Rani and Subramoniam, 1997); and the ovary in Pachygrapsus crassipes (Lui and O'Connor, 1977), Penaeus japonicus (Yano and Chinzei, 1987), and Callinectes sapidus (Lee and Watson, 1995). However, investigations up until present have not demonstrated conclusively that the above tissues are the actual sites of Vg synthesis in crustaceans. In order to fully resolve this question, it is necessary to obtain direct evidence concerning the precise site of Vg synthesis by demonstrating the expression of Vg mRNA in those tissues. However, such studies are just now emerging in decapod crustaceans.
We have selected the kuruma prawn, Penaeus japonicus as a model animal, because of the commercial importance of species of the Penaeus genus in fisheries and the potential to apply findings concerning this animal to the elucidation of reproductive mechanisms in a wide range of saltwater prawn species. In a previous paper, we have purified and characterized Vt from mature ovary of the kuruma prawn as a first step in elucidating mechanisms of vitellogenesis in crustaceans (Kawazoe et al., 2000). It was found that Vt consisted of three subunits of molecular weights of 91, 128, and 186 kDa. We subsequently determined the N-terminal amino acid sequences of the 91 kDa subunit as well as those of 4 peptide fragments obtained by enzymatic digestion. The objective of the present study was to clone a full-length cDNA encoding Vg in the kuruma prawn based on the amino acid sequences of the N-terminus as well as internal portions of the Vg molecule, to deduce the full primary amino acid sequence of Vg, and finally, to determine the exact site of Vg synthesis by examining the expression of its mRNA.
MATERIALS AND METHODS
Animals
Mature female kuruma prawns were obtained from the Momoshima Station of the Japan Sea-Farming Association in Hiroshima Prefecture, Japan. Immature prawns were purchased from a local dealer. Mature prawns ranged from 52 to 86 g in body weight and from 1.2 to 8.9% in gonadosomatic index (GSI). Immature prawns ranged from 18 to 20 g in body weight and from 0.2 to 0.5% in GSI. The developmental stages of oocytes were classified into 9 stages (oogonium, bouquet, chromatin nucleolus, perinucleolus, fused nucleolus (PAS-positive granule), early yolk globule, late yolk globule, early cortical alveoli, and late cortical alveoli stage) mainly according to Hong (1977). In this study, prawns were grouped into the previtellogenic stage (oogonium, bouquet, chromatin nucleolus, and perinucleolus stage), early exogenous vitellogenic stage (early yolk globule stage), and late exogenous vitellogenic stage (late yolk globule stage). Prawns at the endogenous vitellogenic stage (PAS-positive granule) and maturation stage (early and late cortical alveoli stage) were not present in this study.
Isolation of a genomic DNA fragment encoding the N-terminal amino acid sequence of the 91 kDa subunit protein
Genomic DNA was isolated from the hepatopancreas of female prawns according to a standard protocol (Sambrook et al., 1989). Two degenerate oligonucleotide primers, VtF and VtR (Fig. 1), were designed based on the N-terminal amino acid sequence of the 91 kDa subunit (Kawazoe et al., 2000), being derived from the sequences of amino acid positions 1–5 and 25–29, respectively. Polymerase chain reaction (PCR) was carried out in a 50 μl solution containing 1.5 mM MgCl2, 200 μM dNTP, 1×PCR buffer (50 mM KCl, 10 mM Tris-HCl, pH 8.3), 2 μM of each primer, 10 ng of genomic DNA, and 2 U of Taq DNA polymerase (TaKaRa, Tokyo, Japan). After an initial denaturation at 94°C for 4 min, 37 cycles of amplification were carried out using the following program: denaturation at 94°C for 0.5 min, annealing at 52°C for 1 min, and extension at 72°C for 1 min. Final extension was done at 72°C for 7 min. The PCR product was subcloned into the pBluescript SK– plasmid vector (Stratagene, CA, USA). Dideoxy sequencing reactions were carried out using a Thermo Sequenase Cycle Sequencing Kit (Amersham Pharmacia Biotech, Uppsala, Sweden) and electrophoresis and signal detection were done on an automatic sequencer SQ-5500 (HITACHI, Tokyo, Japan).
Fig. 1
Positions and nucleotide sequences of oligonucleotide primers used to clone Vg cDNA. Two degenerate primers, VtF and VtR, were used for PCR, the products of which were used as a probe for Northern blot analysis. Other primers were used for 5′ RACE PCR. The N-terminal sequence of the 91 kDa subunit of vitellin was obtained by protein sequencing (Kawazoe et al., 2000).
Isolation of RNA
The abdominal muscle containing adipose tissue and fan blade containing epidermal tissue, as well as the hepatopancreas, intestine and ovary were dissected out from mature and immature prawns. Total RNA was isolated from these tissues using ISOGEN (Nippongene, Toyama, Japan) according to instructions of the manufacturer. A part of the ovary was fixed in Bouin's solution for determination of its developmental stage (see above).
Reverse transcription-polymerase chain reaction (RT-PCR)
Reverse transcription was carried out in a 10 μl solution containing 1 μg of total RNA from the hepatopancreas or ovary from prawns in the early exogenous vitellogenic stage, 100 pM dT17 primer, 250 μM dNTP, 10 mM dithiothreitol, 15 U of RNase inhibitor (GIBCO BRL, MD, USA), and 100 U of Superscript II™ reverse transcriptase (GIBCO BRL) at 50°C for 1 hr. Using 1 μl of the cDNA solution, PCR was carried out in the same manner as described above. The PCR product was subcloned and sequenced.
Rapid amplifications of 5′ cDNA end (5′ RACE)
Using Oligotex-dT30 super (Roche, Tokyo, Japan), poly(A)+ RNA was prepared from total RNA of the ovary of prawn at the early exogenous vitellogenic stage. First-strand cDNA was synthesized in a 20 μl solution containing 1 μg of poly(A)+ RNA, 100 pM of primer VtR, 500 μM dNTP, 10 mM dithiothreitol, 25 U of RNase inhibitor (GIBCO BRL), and 200 U of Superscript II™ reverse transcriptase (GIBCO BRL) at 50°C for 1 hr. The reaction was stopped by the addition of 80 μl TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0). The resulting first-strand cDNA was tailed with poly(dA) by a 3′-end labeling kit (Amersham Pharmacia Biotech). A degenerate primer, VtR/2, was designed based on the nucleotide sequence of cDNA encoding the N-terminal sequence (Fig. 1). PCR was carried out in a 50 μl solution containing 500 nM of each primer VtR/2, adaptor, T17-adaptor (Fig. 1), 200 μM dNTP, 1×PCR buffer, and 2 μl of the poly(dA)-tailed cDNA as a template. Initial denaturation at 94°C for 4 min, 5 amplification cycles of 94°C for 1 min, 50°C for 1 min and 72°C for 3 min was followed by 35 cycles of 94°C for 0.5 min, 60°C for 0.5 min and 72°C for 2 min. Final extension was done at 72°C for 7 min. The PCR product was subcloned and sequenced as described above.
Construction of the cDNA library
The poly(A)+ RNA was subjected to cDNA library construction using a TimeSaver® cDNA synthesis kit (Amersham Pharmacia Biotech) using 370 ng of Pd(N)6 primer for first-strand cDNA synthesis. The synthesized cDNA was ligated to the λ ZAP-II vector (Stratagene). The ligation product was packaged using Gigapack® III Gold (Stratagene).
Screening of the cDNA library
The cDNA fragment (corresponding to nucleotides 1–168, Fig. 3) which had been obtained by 5′ RACE was labeled with [α-32P]dCTP (Amersham Pharmacia Biotech) using the Megaprime™ DNA labeling system (Amersham Pharmacia Biotech) and used as a probe to screen the cDNA library by plaque hybridization. Positive plaques were isolated in the second round of screening and the recombinant pBluescript SK– phagemids were rescued from the bacteriophage clones according to instructions of the manufacturer (Stratagene). After sequencing analysis of a number of clones, a DNA fragment of 398 bp in length which was located at the 3′ end of the obtained DNA sequence was used as a probe to screen the same library. Positive plaques were isolated and the recombinant pBluescript SK– phagemids were rescued from the bacteriophage clones and sequenced in the same manner as above. From the newly-obtained DNA sequence, a DNA fragment was selected as a probe to screen the same library. We named such a method library walking, i.e., repeating the procedure of selecting new probes from the most recently obtained cDNA clones, screening the same library with the new probe, and sequencing the obtained clones. Library walking was carried out five times (Fig. 2).
Fig. 2
Cloning strategy and schematic view of Vg cDNA. The merged Vg cDNA, probes, and clones sequenced are indicated by their base pair positions. Selected clones from cDNA library screening which were analyzed are designated as I-9, I-11, II-3, II-5, III-7, III-9, IV-1, IV-3, V-1, and V-3. Roman numerals indicate which probe was used to obtain a particular clone. Probes I, II, III, IV, and V were derived from 5′ RACE, I-9, II-5, III-5, and IV-1, respectively.
Fig. 3
Deduced amino acid sequence and 5′ and 3′ UTRs of Vg cDNA. The putative signal sequence is shown in italics (-18 – -1). Previously reported amino acid sequences are underlined (N-terminal sequence, LE3, LE19, LE21, and LE46). Amino acid sequences obtained in this study are underlined with a dotted line (LE101–104 and fraction 2). Ser or Thr residues that are possibly phosphorylated by casein kinase II are shaded. Two consensus cleavage sequences for processing by endoproteases of the subtilisin family are shown in white lettering with black background. Positions of nucleotides and amino acids are indicated by numbers on the left and right sides the figure, respectively. The polyadenylation signal is doubly underlined. The nucleotide sequence has been submitted to the GenBank™/EMBL Data Bank with accession number AB033719.
Extraction and purification of vitellin
We obtained further amino acid sequences of vitellin in addition to those of the 91 kDa subunit in order to confirm the obtained deduced amino acid sequence. The ovary was homogenized on ice in 0.2 M Tris-HCl (pH 7.5) containing 1.5 M NaCl and 6 M urea using a Polytron homogenizer (Switzerland). After centrifugation at 10,000 g for 10 min, the supernatant was collected and subjected to a Microconcentrators 3 filter (Amicon, MA, USA) to cut off low molecular weight proteins. The retentate was then subjected to reversed-phase HPLC using a TSKgel Phenyl-5PW RP column (4.6×75 mm, Tosoh, Tokyo, Japan). Fractionation was performed with a 40-min linear gradient of 0–80% methanol containing 5% acetonitrile in 0.1% trifluoroacetic acid (TFA) at a flow rate of 1 ml/min. Elution was monitored by measuring the absorption at 225 and 280 nm. Two major peak materials (fractions 1 and 2) were collected manually.
Enzymatic digestion and separation of the digests
Fraction 1 in 30 μl of 0.05% TFA was added with 30 μl of 1 M Tris-HCl buffer (pH 9.0) containing 2 M urea and incubated with lysyl endopeptidase (Wako Pure Chemicals, Osaka, Japan) at an enzyme to substrate ratio of 1:40 (w/w) at 37°C for 6 hr. Subsequently, 40 μl of 1 M HCl were added to stop the reaction. The digests were then separated by reversed-phase HPLC using an Asahi Pak ODP-50 column (4.6×150 mm, Showa Denko, Tokyo, Japan) with a 60 min liner gradient of 0–60% acetonitrile in 0.05% TFA at a flow rate of 1 ml/min. The fragments were collected manually by monitoring the absorbance at 225 nm.
Amino acid sequence analysis
N-terminal amino acid sequences of the two proteins (fraction 1 and 2) and fragment peptides digested with lysyl endopeptidase for fraction 1 were analyzed on a LF-3000 protein sequencer (Beckman, CA, USA) in the pulsed-liquid mode.
Northern blot analysis
For Northern blot analysis, 10 μg of total RNA isolated from the abdominal muscle, fan blade, hepatopancreas, intestine, and ovary were run on a 1% formaldehyde-agarose gel. The gel was stained with ethidium bromide, and ribosomal RNA contained in total RNA was measured using the Gel Doc 1000 Gel Documentation System and Molecular Analyst software (Bio-Rad, CA, USA). Subsequently, total RNA was transferred to a Hybond-N+ membrane (Amersham Pharmacia Biotech). The radiolabeled cDNA fragment encoding the N-terminal amino acid sequence of the 91 kDa subunit (corresponding to nucleotides 88–173, Fig. 3) was generated by PCR. The reaction was carried out in a 50 μl solution containing 1.5 mM MgCl2, 200 μM of each dATP, dGTP, dTTP, 0.67 μM (100 μCi) of [α-32P]dCTP (Amersham Pharmacia Biotech), 2 μM of each primer VtF and VtR, 5 ng of template DNA, 1×PCR buffer, and 2 U of Taq DNA polymerase (TaKaRa). After an initial denaturation at 94°C for 4 min, 40 cycles of amplification were carried out using the program described above. The membrane was then hybridized with the radiolabeled probe (1×105 cpm/ml) in 50% formamide, 6×SSC, 1×Denhardt's, 0.4% SDS and 100 μg/ml calf thymus DNA at 42°C for 16 hr and was washed with 0.2×SSC and 0.1% SDS at 55°C. RNA ladder (0.24–9.4 kb, GIBCO BRL) was used as the molecular weight marker. Hybridization signals were quantified using a Fujix Bio-imaging analyzer BAS 1000 (Fuji Film, Tokyo, Japan). An arbitrary value of 100 was applied to 4.5 ng of the pBluescript SK– plasmid DNA into which was inserted the cDNA fragment encoding the N-terminal amino acid sequence of the 91 kDa subunit protein (corresponding to nucleotides 88–173, Fig. 3).
In situ hybridization
The hepatopancreas and ovary from prawns in the early exogenous vitellogenic stage were excised and fixed overnight with 4% paraformaldehyde (PFA, TAAB Laboratories, UK) and 1% picric acid in 0.1 M phosphate buffer, pH 7.4 (PB) at 4°C. Tissues were dehydrated in ethanol and embedded in paraffin. Sections of 5 μm were cut onto APS (silane)-coated slides (Matsunami, Japan).
To generate anti-sense and sense RNA probes, DIG RNA Labeling Kit (Boehringer Mannheim, Germany) was used. The pBluescript SK– plasmid DNA into which was inserted the 5′ RACE PCR product (corresponding to nucleotides 1–168, Fig. 3) was linearized with Xba I (TaKaRa), and T7 polymerase was used to generate anti-sense probe. The sense probe was obtained after digestion with Hind III (TaKaRa) and transcription with T3 polymerase.
In situ hybridization was carried out according to Naito et al. (1991) with some modifications. Sections were digested with 20 μg/ml proteinase K (GIBCO BRL) in TE buffer for 30 min, refixing with 4% PFA in PB for 10 min then were washed in PB for 5 min. Denatured probe (10 μg/ml) was mixed with hybridization buffer (25.8% formamide, 4× SSC, 1×Denhardt's, 0.5 mg/ml calf thymus DNA, 10.3% dextran sulfate). This buffer (90 μl) was applied to each slide and covered with a parafilm. Overnight hybridization was performed in a moist chamber at 37°C. On the following day, slides were dipped in 2×SSC to remove the parafilms, followed by one wash with 2×SSC for 20 min, two washes with 1×SSC for 30 min, and another wash with 2×SSC for 20 min. Slides were washed in TNE (10 mM Tris-HCl, pH 7.6, 1 mM EDTA, 100 mM NaCl) containing 20 μg/ml RNase A for 30 min, followed by one wash in TNE for 10 min, then in 2×SSC for 20 min and finally two washes in 0.2×SSC for 20 min. The hybridized probes were immunodetected with anti-digoxigenin-AP (Boehringer Mannheim) and were visualized with the colorimetric substrates NBT/BCIP (Boehringer Mannheim) according to the instructions attached to DIG DNA Labeling and Detection Kit (Boehringer Mannheim). Slides were mounted in 90% glycerol, sealed with nail polish, and observed under a light microscope.
RESULTS
Isolation of a cDNA fragment encoding the 91 kDa subunit of Vt
Two degenerate primers, VtF and VtR, were designed based on the N-terminal amino acid sequence of the 91 kDa subunit (Fig. 1). A PCR product of the expected size (86 bp) was obtained using these primers and genomic DNA as a template (data not shown). Results of sequence analysis showed that the product was derived from a genomic DNA encoding the 91 kDa subunit, as the deduced amino acid sequence was identical to the N-terminal sequence (corresponding to nucleotides 88–173, Fig. 3). We also confirmed using nucleotide sequence analysis that the two unidentified residues in the N-terminal amino acid sequence were cysteine. The residue Lys29 was threonine in the nucleotide sequence. The same PCR product was obtained by RT-PCR using cDNA from the hepatopancreas or ovary as a template (data not shown).
Isolation and characterization of cDNA clones encoding the 91 kDa subunit protein
Results of RT-PCR analysis showed that mRNA encoding the 91 kDa subunit was present in the hepatopancreas and ovary. In order to isolate the cDNA clones encoding the 91 kDa subunit, a cDNA library was prepared from the ovary of kuruma prawn at the early exogenous vitellogenic stage. A cDNA fragment obtained by 5′ RACE (corresponding to nucleotides 1– 168, Fig. 3) was used as a probe to screen the cDNA library. Among the isolated clones, two clones with long inserts were analyzed for their nucleotide sequences. Subsequently, a DNA fragment of 398 bp which was included in the 3′ downstream region of the obtained nucleotide sequence was used as a probe to screen the same library. Newly-isolated clones were analyzed, and a new probe was selected from the obtained nucleotide sequence. As a result of five rounds of this screening, referred to as library walking (see materials and methods), 10 cDNA clones were obtained (Fig. 2), whose sequences were merged into the sole cDNA sequence containing 33 bp of a 5′ UTR, 7,761 bp of an ORF, and 176 bp of 3′ UTR (Fig. 3). The cDNA sequence contained a polyadenylation signal AATAAA at 10 nucleotides upstream of the site of poly(A) addition. The 4 partial amino acid sequences of the 91 kDa subunit reported in a previous paper (LE3, 19, 21, and LE46, Kawazoe et al., 2000) were included in the deduced amino acid sequence (Fig. 3). In addition, microsequencing of two proteins obtained in this study (fraction 1 and 2) revealed that the N-terminus of fraction 1 was identical to that of the 91 kDa subunit, and N-terminal amino acid sequence of the fraction 2 was SPVVSFVPRVYFGVWSEIVLLQHQ (amino acid position 2097–2120, Fig. 3). N-terminal amino acid sequences of the 4 peptide fragments obtained by lysyl endopeptidase digestion of the fraction 1 were located at amino acid position 233–254 (LE101), 394–402 (LE102), 698–703 (LE103), and 1381–1392 (LE104, Fig. 3). Sequences for LE104 and fraction 2 were confirmed to exist in the C-terminal side of the deduced amino acid sequence. This provided conclusive proof that the obtained cDNA clones encode Vg in the kuruma prawn.
Characterization of the deduced amino acid sequence
The cDNA had an ORF that encoded 2,587 amino acid residues with a predicted molecular mass of 286,702 Da. Based on the N-terminal sequence of the 91 kDa subunit, we were able to predict that a stretch of 18 highly hydrophobic residues (amino acid position -18–-1, Fig. 3) was a signal peptide. This deduced amino acid sequence contained two consensus cleavage sites, R–X–K/R–R (Arg707 to Arg710 and Arg1094 to Arg1097, Fig. 3) capable of undergoing processing by endoproteases of the subtilisin family (Barr, 1991). In contrast to known invertebrate and vertebrate Vg's, this protein possessed no polyserine domain or no potential N-linked glycosylation sites. However, 11 serine residues and 18 threonine residues which were included in the consensus sequence for phosphorylation by casein kinase II were found in the deduced amino acid sequence. The BLAST algorithm (Altschul et al., 1997) indicated that the deduced sequence had some similarity to apolipophorin of the tobacco hornworm (Sundermeyer et al., 1996), retinoid-and fatty-acid binding glycoprotein (RFABG) of the fruit fly (Kutty et al., 1996), human apolipoprotein B-100 (Cladaras et al., 1986), and Vg in sturgeon (Bidwell and Carlson, 1995) and killifish (LaFleur et al., 1995). As shown in Fig. 4, the region of kuruma prawn Vg which showed similarity to other Vg's and lipoproteins was limited to the N-terminal half. The C-terminal regions of kuruma prawn Vg harbored some similarity to the C-terminal region of the aforementioned apolipophorin and D domains of von Willebrand factor (vWF) in mammals (Sadler, 1991; Ruggeri and Ware, 1993; data not shown), which is involved in the coagulation of blood (Voorberg et al., 1990; Mayadas and Wagner, 1992).
Distribution of mRNA encoding vitellogenin
Tissue-specific expression of Vg mRNA was analyzed by Northern hybridization. Total RNA was extracted from the abdominal muscle containing adipose tissue and fan blade containing epidermal tissue, as well as from the hepatopancreas, intestine, and ovary from female prawn. RNA blots were hybridized with the radiolabeled Vg cDNA and hybridized signals were thus detected in the hepatopancreas and ovary, but not in other tissues (Fig. 5). The probe hybridized to a single band of approximately 8 kb in each tissue. The size of the mRNA was in agreement with that of the Vg cDNA obtained in this study (7970 bp, Fig. 3).
Fig. 5
Northern blot analysis of tissue specific expression of Vg mRNA. Results from a prawn at the early exogenous vitellogenic stage is shown as an example. Total RNA (10 μg/lane) from the abdominal muscle containing adipose tissue (lane 1), fan blade containing epidermal tissue (lane 2), hepatopancreas (lane 3), intestine (lane 4), and ovary (lane 5) were blotted onto a nylon membrane and probed with Vg cDNA which corresponded to nucleotides 88–173 (Fig. 3). Positions of the five molecular weight markers (9,490, 7,460, 4,400, 2,370 and 1,350 nucleotides) are indicated.
Fig. 6 shows the results of in situ hybridization using antisense and sense cRNA probes, indicating also the expression of Vg mRNA in the hepatopancreas and ovary. Vg mRNA signals were observed in the follicle cells that surrounded the vitellogenic oocytes (Fig. 6A). As for the hepatopancreas, signals were observed in the parenchymal cells (Fig. 6C). Hybridization with a sense control probe produced no significant signal in both tissues (Fig. 6B and D).
Fig. 6
Localization of Vg mRNA in the hepatopancreas and ovary by in situ hybridization. Strong expression of Vg was observed in the follicle cells (A) in the ovary, and in the parenchymal cells (C) in the hepatopancreas. Hybridization with a sense probe produced no significant signal in the ovary (B) and hepatopancreas (D). Bars=20 μm in A and B; 10 μm in C and D.
Changes in relative levels of mRNA encoding vitellogenin
Changes in Vg mRNA levels in the hepatopancreas and ovary during the non-vitellogenic and vitellogenic stages were examined by Northern hybridization. The same cDNA fragment was used to probe a Northern blot containing total RNA from immature and mature prawns. To ensure equal loading of RNA samples, gels were stained with ethidium bromide before blotting and levels of ribosomal RNA were measured for standardization. Results of changes in relative levels of Vg mRNA in the hepatopancreas and ovary are shown in Fig. 7. Sixteen ovaries were divided equally into four groups based on GSI and developmental stages of the oocytes as follows: GSI of 0.2–0.5% (previtellogenic stage), GSI of 1.2–3.7% (early exogenous vitellogenic stage), GSI of 4.4–7.2% (late exogenous vitellogenic stage), and GSI of 7.7–8.9% (late exogenous vitellogenic stage). In the hepatopancreas, the relative level of expression increased as vitellogenesis progressed, but decreased during later stages of maturation. In the ovary, highest levels were observed during the early exogenous vitellogenic stage and thereafter rapidly decreased.
Fig. 7
Changes in relative expression levels of Vg mRNA in the hepatopancreas and ovary. Sixteen ovaries were divided into four groups based on GSI and developmental stages of the oocytes as follows: I, previtellogenic stage (GSI of 0.2–0.5%); II, early exogenous vitellogenic stage (GSI of 1.2–3.7%); III, late exogenous vitellogenic stage (GSI of 4.4–7.2%); IV, late exogenous vitellogenic stage (GSI of 7.7–8.9%). Four females were used in each group of this experiment. Results are presented as the mean±SEM. Shaded columns, GSI; open columns, Vg mRNA levels in the hepatopancreas; solid columns, Vg mRNA levels in the ovary.
DISCUSSION
In this investigation, we cloned a cDNA encoding Vg in the kuruma prawn based on the N-terminal amino acid sequence of the 91 kDa subunit of Vt determined in our previous paper (Kawazoe et al., 2000). The ORF encoded 2,587 amino acid residues with a predicted molecular mass of 286,702 Da. This is the first investigation reporting a complete sequence for a Vg cDNA in a crustacean species. Its deduced amino acid sequence possessed two dibasic proteolytic processing sites R–X–K/R–R (Barr, 1991) that have also been found at subunit cleavage sites of insect Vg's, including Anthonomus grandis (Trewitt et al., 1992), Aedes aegypti (Chen et al., 1994), Bombyx mori (Yano et al., 1994), Lymantria dispar (Hiremath et al., 1997), and Riptortus clavatus (Hirai et al., 1998). Four peptide fragments obtained by enzymatic digestion of the 91 kDa subunit of Vt were located on the N-terminal side of the first processing site, suggesting that the 91 kDa subunit would be produced as the result of cleavage at the first processing site. If this is the case, the resultant two subunits would have calculated molecular masses of 78,466 Da and 206,509 Da, corresponding to the 91 kDa and 186 kDa subunits reported previously (Kawazoe et al., 2000). The discrepancy between the apparent molecular weight based on SDS-PAGE and the calculated one is not large and is considered to be within an acceptable error range. This can be confirmed by further elucidation of the N-terminal amino acid sequence of the 186 kDa subunit. Another subunit of 128 kDa was also found in the previous paper; however, in the present investigation no appropriate processing site which could produce a 128 kDa subunit was observed in the deduced sequence. Further investigation including isolation and sequence analysis will be necessitated in order to elucidate the 128 kDa protein. N-terminal amino acid sequence of the fraction 2 obtained in this study was located at amino acid positions 2097–2120 of kuruma prawn Vg (Fig. 3). This fraction would be produced artificially during purification because there were no dibasic proteolytic processing sites around the amino acid sequence of its N-terminus.
Kuruma prawn Vg showed low homology to several lipoproteins and Vg's thus far identified, including apolipophorin from the tobacco hornworm, RFABG from the fruit fly, human apolipoprotein B-100, and sturgeon and killifish Vg (Fig. 4). Biological similarities are seen between such lipoproteins and Vg from the point of view of binding of hydrophobic molecules, cell specific uptake, and the possibility that these proteins may have a common ancestor (Baker, 1988; Babin et al., 1999). In this investigation, the region of kuruma prawn Vg which showed homology was limited to the N-terminal half of the molecule corresponding to the 91 kDa subunit region. The C-terminal half showed no significant homology to any proteins other than the C-terminal region of apolipophorin from the tobacco hornworm and vWF in mammals; this region thus appears to possess unique properties compared with Vg of other known species. Kuruma prawn Vg did not show significant homology to Vg's of six insect species, i.e., A. grandis, A. aegypti, B. mori, L. dispar, R. clavatus, and Pimpla nipponica (Nose et al., 1997), and the plasma clotting protein of Procambarus clarkii, which is homologous to insect Vg's (Hall et al., 1998).
In Northern blot analysis, Vg mRNA of the same size was detected in the hepatopancreas and ovary of vitellogenic females and not detected in the adipose tissue, indicating that Vg is synthesized in the hepatopancreas and ovary in accordance with ovarian maturation (Fig. 5), which was reinforced by the results of in situ hybridization. Vg mRNA signals were detected in the follicle cells (Fig. 6A) and parenchymal cells (Fig. 6C). In P. semisulcatus, it has similarly been suggested through incubation studies that Vg is produced in both the hepatopancreas and ovary (Fainzilber et al., 1992). However, our investigation is one of the first in any oviparous animal to demonstrate based on mRNA expression that two differing tissues are simultaneously involved in vitellogenin synthesis. The detection of Vg mRNA expression in the follicle cells of the kuruma prawn is consistent with the results of a previous report on the same animal (Yano and Chinzei, 1987), in which it was concluded based on an immunofluorescence study that the follicle cells were responsible for ovarian Vg synthesis. Furthermore, the ovaries in several decapod crustaceans were found to synthesize Vg under in vitro tissue culture (Lui and O'Connor, 1977; Eastman-Reks and Fingerman, 1985; Yano and Chinzei, 1987; Browdy et al., 1990; Lee and Watson, 1995). In most of these studies, levels of Vg synthesis were measured through the incorporation of radiolabeled amino acids into protein synthesized de novo.
As a result of detailed analysis, changes in Vg mRNA levels were determined to differ between these two tissues (Fig. 7). In the hepatopancreas, high mRNA levels were observed during the early and late exogenous vitellogenic stages (average GSI of 2.5% and 5.9%, respectively), and those levels were decreased following later stages of vitellogenesis (average GSI of 8.2%). On the other hand, highest mRNA levels were observed during the early exogenous vitellogenic stage in the ovary, and thereafter those levels were rapidly decreased and remained at low levels. In this investigation, we demonstrated changes in GSI and mRNA levels in the two tissues, hepatopancreas and ovary, and this would allow the examination of the dynamics of tissue-specific contribution to Vg synthesis.
In decapod crustaceans, vitellogenesis is negatively regulated by a neuropeptide, vitellogenesis-inhibiting hormone (VIH), produced by the X-organ sinus gland complex in the eyestalk, and vitellogenesis can be induced artificially via bilateral eyestalk ablation (Abramowitz and Abramowitz, 1940; Brown and Jones, 1948). Soyez et al. (1987) purified a VIH from the sinus glands of Homarus americanus, and this VIH inhibited the onset of vitellogenesis of Palaemonetes varians. Its amino acid sequence was subsequently determined (Soyez et al., 1991). More recently, VIH was purified from the terrestrial isopod Armadillidium vulgare (Gréve et al., 1999). Furthermore, seven crustacean hyperglycaemic hormones (CHH) were purified from kuruma prawn (Yang et al., 1996, 1997; Nagasawa et al., 1999), and were shown to be structurally similar to VIH. These P. japonicus peptides inhibited the synthesis of protein and mRNA in the ovary of P. semisulcatus in an in vitro tissue culture system (Khayat et al., 1998). In kuruma prawn, these neuropeptides are also expected to regulate the expression of mRNA encoding Vg in the same manner as VIH in the hepatopancreas and ovary. In consideration of the changes observed in Vg mRNA levels in the hepatopancreas and ovary, it is possible to assume that the expression of Vg mRNA is regulated in a tissue-specific manner. The results of this investigation will enable in further studies on the full elucidation of the hormonal regulation of vitellogenesis including the physiological functioning of VIH and putative vitellogenesis-stimulating hormone (VSH) and vitellogenesis-stimulating ovarian hormone (VSOH).
Acknowledgments
We are grateful to the Momoshima Station, Japan Sea-Farming Association for providing the experimental animals. We also thank Prof. E. Lubzens of Israel Oceanographic and Limnological Research for valuable advice, and Dr. T.-W. Shih of the Department of Aquatic Bioscience at the University of Tokyo for encouragement and discussion. This work was supported partly by a Sasakawa Scientific Research Grant from The Japan Science Society (to N. T.), and Grants-in-Aid for Scientific Research from the Ministry of Education, Science, Sports, and Culture of Japan (to K. A.). This work in addition received support from the Japan International Research Center for Agricultural Sciences (JIRCAS), Ministry of Agricultural, Forestry, and Fisheries, where W.-J. Y. was a Science and Technology Agency (STA) Fellow.
