| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Gamete Biology |
Department of Zoology,4 College of Agriculture and Life Sciences, North Carolina State University, Raleigh,North Carolina 27695-7617
South Carolina Department of Natural Resources,5 Hollings Marine Laboratory and Marine Resources Research Institute, 331 Fort Johnson Road, Charleston, South Carolina 29412
Department of Biology,6 University of South Florida, Tampa, Florida 33620-5150
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
4.1 kilobase (kb) hybridization signal in an mRNA preparation obtained from previtellogenic perch ovaries. The deduced amino acid sequence of the perch VgR was 89% and 82% identical, respectively, to that of the tilapia and rainbow trout. Because it possessed an eight-repeat ligand-binding domain (LR8) but lacked an O-linked sugar domain (–), the perch VgR was identified as a non-O-linked form of VgR (LR8–). Unlike the case in other vertebrates investigated, including tilapia and trout, no species of mRNA encoding an O-linked form of VgR (LR8+) could be detected when perch ovarian or liver mRNA reverse transcripts or cDNA libraries were screened by PCR using primer sets flanking the putative O-linked sugar domain. These novel findings call into question the assumptions that an LR8+ splice variant of the VgR always is dominantly present in somatic tissues and exists at lower levels in ovarian tissues to sequester lipoproteins distinct from Vg. A SYBR-green-based real-time reverse transcription-polymerase chain reaction assay was developed and used to quantitatively measure VgR expression in gonadal and somatic tissues, for the first time in any vertebrate. The main site of perch VgR mRNA expression was the ovary and the highest level of VgR mRNA expression was in ovaries whose largest follicles contained previtellogenic oocytes. Expression of VgR mRNA decreased with oocyte growth during vitellogenesis and was very limited in ovulated eggs. These quantitative results verify the concept that growing oocytes must extensively recycle LR8– forms of the VgR.
estradiol, female reproductive tract, gamete biology, oocyte development, ovary
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The gene encoding the chicken VgR, which is a multifunctional receptor involved in transport of Vg, very low- density lipoprotein (VLDL), 
2-macroglobulin, and lactoferrin into oocytes, has been characterized as cDNA [17, 18]. The deduced cDNA product was homologous to mammalian VLDL receptors, members of a branch of the low- density lipoprotein receptor (LDLR) superfamily, and was structurally classified as an LDLR relative with eight repeated motifs making up its ligand-binding domain (LR8). The LDLRs have only seven such repeats (LR7). The chicken expresses two forms of LR8 mRNA that appear to be generated by differential splicing of an exon called the O-linked sugar domain because it encodes a serine- and threonine-rich glycopeptide motif characteristic of some lipoprotein receptors [18]. The shorter, non-O-linked form (LR8–) of the chicken VgR is dominantly expressed in female germ cells, while the major sites of expression of the longer form (LR8+) are in somatic tissues, but no functional differences were found between translated products of the two LR8 mRNAs in terms of ligand binding. Subsequently, a full-length cDNA of the gene encoding the Xenopus VgR (chicken LR8 homologue) was cloned and characterized [19]. As was the case with the chicken, results of reverse transcription-polymerase chain reaction (RT- PCR) suggested the existence of dual Xenopus VgR mRNAs, one of which appears to be a splice variant missing the sequence encoding the O-linked sugar domain (LR8–).
More recently, two forms of cDNA encoding a rainbow trout lipoprotein receptor (LPR) were cloned and the deduced amino acid sequence revealed a type-I membrane protein of size and sequence similar to, and bearing the same characteristic suite of functional domains as, the VgR of the chicken and Xenopus [20–22]. Based on preferential expression of the corresponding mRNA in ovary, changes in mRNA expression level during oogenesis, and structure (LR8–), the shorter form of cDNA appeared to encode the rainbow trout VgR. The second cDNA encoded an LR8 product with an O-linked sugar domain (LR8+) that appears to be derived from a splice variant of VgR mRNA, as described above for the chicken and Xenopus. It was suggested that one function of the LR8+ type LPR in ovaries and somatic tissues of trout is to bind VLDL or VLDL- like proteins [20]. Molecular characterization of the VgR in more phylogenetically evolved teleost species has been limited to tilapia, Oreochromis aureus [23]. In this species, two distinct splice variants of VgR mRNA were detected by RT-PCR in various tissues, although the LR8– type of VgR message seemed to be dominant in ovaries.
Although teleost fishes represent the largest and most diverse class of vertebrates, among teleosts, molecular characterization of the VgR has been achieved for only rainbow trout and tilapia. White perch (Morone americana) are an established model for basic research on reproduction of advanced teleosts and functional characteristics of their ovarian VgR have been described [8, 10]. Unlike trout and tilapia, white perch produce fatty pelagic eggs, which is typical of most vertebrate species. However, the molecular biology of VgRs or other LPRs involved in massive transport of lipids and yolk proteins into growing oocytes of such species has not previously been investigated. The objectives of the present study were to 1) identify and characterize cDNAs encoding the VgR and other ovarian LPRs in white perch, 2) develop a real-time quantitative RT-PCR (rtqRT- PCR) assay for VgR mRNA, and 3) use the assay to verify tissue localization of VgR transcripts and changes in mRNA expression levels during oogenesis.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Experimental animals were adult white perch held in tanks at the North Carolina State University (NCSU) Aquatic Research Laboratory or the NCSU Biological Resources Facility under artificial photothermal regimes [24, 25].
Ovarian and liver tissues used to construct cDNA libraries were pooled from samples taken from three females whose most advanced clutch of oocytes had completed vitellogenic growth (maximum diameter 650 mm). The ovaries of these females also contained several (3–4) additional clutches of oocytes at earlier previtellogenic and vitellogenic growth stages. To observe maturational changes in expression of ovarian VgR mRNA, ovaries or ovulated eggs were sampled from females at various reproductive stages (stages III–VI; see Results). Ovarian and liver tissues were sampled from previtellogenic females (maximum oocyte diameter 200 mm; stage III–IV) to screen tissues for the presence of VgR mRNA splice variants via RT-PCR or to detect VgR mRNA by Northern blotting. To investigate the tissue distribution of VgR gene expression, gonadal and somatic tissues were sampled from vitellogenic females (maximum oocyte diameter 440 µm; stage V) and mature (spermiating) males.
To obtain ovulated eggs, females were injected intramuscularly with human chorionic gonadotropin (CHORULON; Intervet Inc., Millsboro, DE) at a dose of 500 IU/kg body weight [24]. Eggs were manually stripped from the females. Tissues (ovaries, eggs, testis, muscle, or liver) were excised from male and female fish, immediately frozen in liquid nitrogen (LN), and then kept at –80°C before extraction of total RNA. White perch were deeply anesthetized before handling [10] and all experiments involving live fish were carried out in accordance with the 1996 Guide for Care and Use of Laboratory Animals published by the National Research Council.
RNA Extraction
To produce a probe for screening an ovarian cDNA library, a portion of the cDNA encoding the perch VgR was cloned using RNA extracted from samples of ovary pooled from three vitellogenic females as the starting material. The pooled sample (1.0 g) of perch ovaries was finely ground under LN with a mortar and pestle and 50-mg aliquots of the homogenate were extracted using a Pharmacia Quick Prep Total RNA Extraction Kit (Pharmacia Biotech, Piscataway, NJ) following the manufacturer's instructions. Following extraction, the RNA was used immediately as the template for first strand VgR cDNA synthesis.
For analysis of VgR gene expression in different tissues by rtqRT-PCR, total RNA was extracted from 150-mg samples of white perch ovaries (or ovulated eggs), testis, liver, and muscle using Trizol Reagent (Invitrogen, San Diego, CA). The RNA was used immediately or stored in 75% ethanol at –80°C up to 4 days. Aliquots of total RNA extracted from several ovaries (n = 5) containing oocytes at different growth stages (primary growth, early secondary growth, and vitellogenic) were pooled and used as an interassay standard sample for rtqRT-PCR.
For PCR-screening of VgR splice variants or Northern blotting analysis, poly (A)+ mRNA was isolated from 1 g of perch ovarian or liver tissues using an Invitrogen FastTrak 2.0 mRNA isolation kit.
Reverse Transcription
For cloning of the VgR cDNA probe, single-stranded cDNA was synthesized from 5 mg of total ovarian RNA in a 33-µl reaction using a Pharmacia First-Strand cDNA Synthesis Kit and the Not I-d(T)18 primer.
Single-stranded cDNA also was synthesized for rtqRT-PCR of VgR mRNA or partial cloning of white perch 18S ribosomal RNA (rRNA) from 4 µg of total RNA in 20-µl reactions using the Invitrogen Superscript First Strand Synthesis System for RT-PCR. The same reaction mixture and total RNA sample was used as before, but reverse transcriptase was omitted from the first-strand synthesis reaction to produce no-RT control templates (NRT) for rtqRT-PCR. Reverse transcriptions conducted at different times invariably included a sample of the pooled total ovarian RNA as an interassay standard to normalize the efficiency of rtqRT-PCR.
As a template for PCR screening of VgR splice variants, single-stranded cDNA was synthesized from 250 ng of poly (A)+ mRNA in 20-µl reactions using the Superscript First-Strand Synthesis System as described above.
Generation of Radioisotopic VgR cDNA Probe
For cloning of the VgR cDNA probe, aliquots (2 µl) of the first-strand ovarian cDNA synthesis reaction product were amplified by PCR using degenerate oligonucleotide primers. Primers corresponded to highly conserved nucleotide sequences in chicken and Xenopus VgR (LR8) cDNAs and were located in domains encoding the epidermal growth factor (EGF)- precursor homology motif. The forward primer (VgR F1) and reverse primer (VgR R1) terminated at position 1799 (19mer) and began at position 2221 (17mer), respectively, of the chicken VgR cDNA sequence (Table 1). Amplification was carried out in a PTC-100 Programmable Thermal Controller model 60 (MJ Research, Watertown, MA) in a 50-µl reaction using 2.5 U of Taq DNA polymerase (Promega, Madison, WI). The PCR conditions were denaturation at 94°C for 40 sec, primer annealing at 43°C for 45 sec, and polymerization at 72°C for 90 sec for a total of 34 cycles.
|
The PCR products were cloned into the pCR2.1 vector using an Invitrogen original TA Cloning Kit following the manufacturer's protocol (version B). Ten clones were harvested and subcultured for extraction of plasmid DNA using the Wizard Minipreps DNA Purification System (Promega). After electrophoresis of the amplicon on a 1.2% agarose gel to verify insert size (460 base pairs [bp]), positive clones were isolated, subcultured, and extracted for plasmid DNA, which was sequenced in both directions in the Nucleic Acid Analysis Facility in the Biotechnology Resource Laboratory, Department of Biochemistry and Molecular Biology, College of Medicine, Medical University of South Carolina (MUSC). The 457-bp insert carried in the plasmid DNA of these clones was found to encode a deduced amino acid sequence with an extremely high degree (70%) of identity to the EGF precursor homology domain of the chicken VgR (LR8–). Identity at the nucleotide level was 71%.
The perch partial VgR cDNA was amplified from a representative clone (plasmid T1-1) by PCR using homologous primers (VgR F2 and VgR R2, Table 1) with annealing at 50°C. Amplicons were electrophoresed in a 0.8% agarose gel, excised from the gel, and extracted using the Promega Wizard PCR Preps DNA Purification System. This DNA was then used as a template for random primer labeling with [-32P] dATP (New England Nuclear Life Sciences, Boston, MA) to produce a high specific activity (1.4 x 109 dpm/µg DNA) probe for screening the ovarian cDNA library. The probe was labeled using a Prime-It II Random Primer Labeling Kit (Stratagene, La Jolla, CA) and purified using Stratagene NucTrap Probe Purification Columns.
Generation of Digoxygenin-Labeled VgR cDNA Probe
To produce the digoxygenin (DIG)-labeled VgR cDNA probe used for Northern blotting, aliquots (10 pg) of a full-length cDNA encoding perch VgR (clone C3; see below) were amplified by PCR using 200 nM of gene- specific primers. The forward primer (NB F1) and reverse primer (NB R1) terminated at position 3164 (19mer) and began at position 3736 (20mer), respectively, of the perch VgR cDNA sequence (Table 1). Amplification was carried out in a Hybaid PCR Express thermal cycler (Continental Lab Products, San Diego, CA) using a PCR DIG Probe Synthesis kit (Roche Diagnostics GmbH, Mannheim, Germany) according to the manufacturer's instructions. After electrophoresis on a 1.2% agarose gel, the DIG-labeled PCR product was revealed as a single band with a size of 620 bp, a slightly higher mass than the expected amplicon size (610 bp), presumably due to DIG labeling. The DIG-VgR probe was stored at –20°C until used for Northern blotting.
Construction and Screening of Ovarian and Liver cDNA Libraries
The unidirectional, size-restricted (>0.4 kilobases [kb]) ovarian cDNA library was constructed by Stratagene in the Uni-Zap XR vector (cloning site EcoRI and XhoI). The ZAP-cDNA synthesis method [26] was used with ovarian poly (A)+ mRNA from vitellogenic white perch as the starting material. General library characteristics included background of <1% nonrecombinants, average insert size of 1.9 kb, and an insert size range of 0.6–4.3 kb (calculated from 12 clones picked at random).
To clone the full-length perch VgR cDNA, approximately 1.8 x 105 recombinant phage were screened by plaque hybridization using the 32P- labeled T1-1 probe under high stringency conditions. Plaques yielding strong hybridization signals on developed films were selected for secondary and tertiary screening followed by in vivo excision of the pBluescript SK(–) phagemid from the Uni-Zap XR vector using the ExAssist/SOLR system and protocol (Stratagene). Secondary screening involved double digestion of the phagemid DNA with EcoRI and XhoI in REACT 2 buffer (Gibco BRL, Invitrogen) followed by gel electrophoresis to verify the presence of an insert of the size expected (3.5 kb) for a complete VgR cDNA. On this basis, one phagemid clone (C3) was selected for full-length nucleotide sequencing, which was performed as described above in the Nucleic Acid Analysis Facility at MUSC. Resulting sequences were compared with sequences obtained from primary publications or from databases at the National Center for Biotechnology Information using the BLAST network service [27]. Alignments of the sequences also were manually conducted using MacDNASIS (Hitachi Software Engineering America, San Francisco, CA) or MacVector (Oxford Molecular Ltd., Madison, WI) software programs.
The vitellogenic perch liver cDNA library was constructed as described above for the ovarian cDNA library except that liver samples from the same three females were pooled as the source of poly (A)+ mRNA starting material. The liver cDNA library was employed as a template for the PCR screening of VgR splice variants. General library characteristics included a background of <1% nonrecombinants, average insert size of 1.6 kb, and an insert size range of 0.6–3.5 kb (calculated from 12 clones picked at random).
Northern Blotting
Poly (A)+ mRNA (20 µg) from perch ovarian tissues or Invitrogen RNA markers (0.24–9.5 kb RNA marker) were denatured in NorthernMax Formaldehyde Load dye (Ambion Inc., Austin, TX) at 60°C for 15 min, separated by electrophoresis on a formaldehyde/MOPS denaturing 1% agarose gel, and blotted onto a Nytran membrane (Schleicher and Schuell, Keene, NH) using the Turboblotter system (Schleicher and Schuell). The nucleic acids transferred to the membrane were then ultraviolet crosslinked, prehybridized in DIG Easy Hyb solution (Roche Diagnostics GmbH) for 4 h at 50°C, and hybridized in the same buffer containing DIG-VgR probe (0.1 ml = 100 ng probe/ml) at 50°C overnight. Membranes were washed and blocked using the DIG Wash and Block Buffer set (Roche Diagnostics GmbH) and the hybridized DIG-VgR probe was visualized using antidigoxygenin antibody conjugated to alkaline phosphatase (Roche Diagnostics GmbH) and CDP-star (Roche Diagnostics GmbH) substrate as described in the manufacturer's manual for DIG-probing (DIG Application Manual for Filter Hybridization; Roche Diagnostics GmBH).
PCR Conditions for Screening of VgR Splice Variants
Three synthetic oligonucleotide primer sets (50 pmol primer/50 µl PCR reaction) targeting cDNA flanking the putative O-linked sugar domain (Table 1, OLF and OLR primers; see legend to Fig. 4 for specific primer combinations and target VgR cDNA sequences) were used to amplify templates of first-strand cDNA (1 µl/50 µl PCR reaction) and cDNA libraries (1 µl/50 µl PCR reaction) prepared from liver or ovarian mRNA of white perch. Amplification was carried out in the Hybaid thermal cycler using Hotstar Taq DNA polymerase (Qiagen Inc., Valencia, CA) with the following thermal cycling protocol: stage 1, 95°C for 15 min for 1 cycle; stage 2, 94°C for 1 min, 60°C for 1 min, and 72°C for 2 min for 30 cycles; and stage 3, 72°C for 10 min for 1 cycle.
|
Molecular Cloning of Internal Control Genes
Molecular cloning of a cDNA encoding a portion of 18S rRNA from white perch was conducted to generate an internal control for the rtqRT- PCR assay. Synthetic oligonucleotide primers (18S F1 and 18S R1; see Table 1) were designed based on the cDNA sequence encoding striped bass (Morone saxatilis) 18S rRNA (GenBank Accession AF147741). These primers were used at a concentration of 50 pmol in a 50-µl PCR reaction to amplify 1 µl of the first-strand cDNA product prepared from ovarian total RNA of white perch. Amplification was carried out using Hotstar Taq DNA polymerase in the Hybaid thermal cycler using the following thermal cycling protocol: stage 1, 95°C for 15 min for 1 cycle; stage 2, 94°C for 1 min, 60°C for 1 min, and 72°C for 1 min for 30 cycles; and stage 3, 72°C for 10 min for 1 cycle. Amplified products were separated by 1.2% agarose gel electrophoresis and a 360-bp product was excised and purified using the Gene Clean II kit (Bio 101 Inc., Carlsbad, CA). Products were subcloned into the pCRII vector using the TA Cloning kit and transformed into INVF cells using a One Shot competent cell kit (Invitrogen). Clones with a 360-bp insert were identified by restriction digestion and gel electrophoresis, subcultured, extracted for plasmid DNA, and sequenced in both directions at the NCSU DNA Sequencing and Mapping Facility or at the DNA Sequencing Facility of the Cancer Research Center at the University of Chicago. Comparison of these sequences with other known 18S rRNA sequences was conducted as described above.
Real-Time Quantitative RT-PCR
Transcript abundance of perch VgR and 18S rRNA (internal control) was quantified in gonads, muscle, and liver from both sexes by SYBR- green-based rtqRT-PCR. Gene-specific primers for rtqRT-PCR amplification were designed according to the requirements set forth by Primer Express software (Applied Biosystems, Foster City, CA). The cDNA sequences obtained in the present study were used as templates to design gene-specific primers (Table 1) for amplification of VgR mRNA (VgR F3 and VgR R3) and 18S rRNA (18S F1 and 18S R2). Primer concentrations used in rtqRT-PCR were 50 nM for VgR amplification and 20 nM for 18S rRNA amplification, based on the results of a preliminary primer optimization experiment (data not shown). The PCR amplifications and fluorescence detection were performed using the GeneAmp 5700 Sequence Detection System (Applied Biosystems) and the manufacturer's universal thermal cycling conditions. Typically, first-strand cDNA that was reverse- transcribed from, and therefore equivalent to, 1 ng of total RNA preparation (see Reverse Transcription) served as a template for duplicate samples in 25-µl PCR reactions using SYBR Green Core Reagents (Applied Biosystems). In addition, duplicate standard curves were generated using a serial dilution of first-strand cDNA template that was reverse transcribed from, and equivalent to, 0.02–2.00 ng of the interassay standard RNA. Accordingly, transcript abundance of the perch VgR gene, which was normalized to that of 18S rRNA, was reported as a fold change in abundance relative to the values obtained for cDNA reverse transcribed from the interassay standard RNA (for more details, see Relative Quantification of Gene Expression; ABI Prism 7700 Sequence Detection System, User Bulletin 2, P/N 4303859 Rev. A, Stock No. 777802-001).
Statistics
The occurrence of differences among means was verified by one-way analysis of variance (ANOVA). Where significant differences existed (P < 0.05), the differing means were identified using a Tukey-Kramer honestly significant difference (HSD) test. The JMP software program (SAS Institute Inc., Cary, NC) was used to conduct these statistical analyses.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Clone C3 contained a 230-bp 5'-untranslated region followed by a 2532-bp open reading frame and a 1259-bp 3'- untranslated region including an 18-repeat ATT microsatellite just upstream from a sequence containing short (12mer) and long (23mer) poly (A)+ signals (GenBank Accession AY173045). The open reading frame encoded an 844 AA polypeptide with a predicted mass of 93 kDa and several other characteristic features of an LR8 (Fig. 1). These characteristics include the presence of a ligand-binding domain at the N-terminus containing eight repeats with six cysteines each, an epidermal growth factor (EGF) precursor homology domain with three cysteine-rich repeats (A, B, and C), five F/YWXD motifs flanked by repeats B and C of the EGF precursor homology domain, a short transmembrane domain, and a cytoplasmic domain containing the peptide sequence (NPVY) required for receptor internalization via coated pits [28, 29]. The deduced AA sequence of the polypeptide encoded in clone C3 showed high identity with VgRs from tilapia (89%), rainbow trout (82%), chicken (67%), and Xenopus (64%), as well as high identity with several mammalian LR8s (e.g., 68% with human VLDLR) (Table 2). Identities less than 50% were revealed when the deduced AA sequence of the polypeptide encoded in clone C3 was compared with that of LR7s (LDLRs). For example, it was 46% identical to Xenopus LDLR1 (GenBank Accession Q99087) and 47% identical to human LDLR (GenBank Accession P01130). Northern blotting analysis using the specific DIG-VgR probe revealed a single intense band corresponding to a 4.14-kb mRNA (Fig. 2), which is consistent with the size of the VgR cDNA sequence inserted in clone C3 (4021 bp). These results, combined with the observations of gene expression patterns described below, confirmed that clone C3 contained a full- length cDNA encoding the white perch VgR (see Discussion).
|
|
|
PCR Screening of Perch VgR Splice Variants
Using three primer sets targeting cDNA flanking the sequence encoding the putative O-linked sugar domain, PCR screening of ovarian and liver mRNA reverse-transcript preparations (Fig. 3A) and cDNA libraries (Fig. 3B) was conducted to detect sequences encoding an O-linked (LR8+) form and/or a non-O-linked form (LR8–) of the VgR. The three primer sets (O1–O3) were chosen to amplify products 410, 450, and 190 bp in length, respectively, representing the LR8– form of VgR encoded in clone C3. For each set of primers, the size expected of an amplicon representing an LR8+ splice variant of VgR was 100 bp larger. As revealed by agarose gel electrophoresis, PCR reactions involving ovarian templates each yielded a single amplicon corresponding to the size expected for an LR8– form of VgR. The PCR preparations involving liver templates yielded only weak traces of bands of the same size as those generated from ovarian templates, bands that were too faint to document by routine gel photography (Fig. 3). No evidence of the (100-bp) larger bands expected to be amplified from a template encoding an LR8+ splice variant of the VgR was observed during these PCR screening procedures.
|
Molecular Characterization of Internal Control Gene
The TA cloning following RT-PCR amplification of ovarian RNA from perch using the 18S F1 and 18S R1 primers resulted in a cDNA clone containing a 360-bp insert. The nucleotide sequence of the insert (GenBank Accession AY370532) was 99% identical to that of Atlantic salmon (Salmo salar) 18S rRNA at positions 913 to 1273 of the salmon sequence (GenBank Accession AJ427629). In addition, the perch 18S RNA sequence was 99% identical to 18S rRNA sequences from some other teleosts, including striped bass (nucleotide positions 74–434), tilapia (Oreochromis massambicus; positions 530–890; GenBank Accession AF497908), and mummichog (Fundulus heteroclitus; positions 900–1260; GenBank Accession M91180). The perch sequence also was nearly identical (
97%) to corresponding sequences from higher vertebrates including Xenopus (positions 907–1267; GenBank Accession K01373), chicken (positions 862–1222; GenBank Accession AF173612), and human (positions 1046–1406; GenBank Accession M10098).
Validation of Real-Time Quantitative RT-PCR
Specificity of the rtqRT-PCR assay using SYBR-green reagents was confirmed by gel electrophoresis of the PCR products. Each gene-specific primer set, designed for amplification of perch VgR mRNA (VgR 3F and VgR 3R) and 18S rRNA (18S F1 and 18S R2), produced a PCR product corresponding to the expected length (69 bp and 75 bp, respectively) when reverse transcripts of ovarian total RNA were used as templates (data not shown). There was no signal produced in the no reverse transcriptase control (NRT) or the no template control (NTC) reactions. For each of the target templates (VgR mRNA or 18S rRNA), specificity of the amplicon produced by rtqRT-PCR was further verified by the following procedures: 1) A dissociation curve was run for each reaction to detect nonspecific amplification. It revealed a single fluorescence peak for each amplicon, indicating production of a single and specific PCR product. And 2) the rtqRT-PCR products were subjected to agarose gel electrophoresis, excised from the gel, extracted, and sequenced in both directions. The nucleotide sequence of each amplicon was 100% identical to the corresponding cDNA sequence of each gene in the region targeted for amplification by the respective primer set.
The standard curve of rtqRT-PCR for each target template (perch VgR mRNA and 18S rRNA) had correlation coefficients (r2) > 0.99. Intraassay variability (coefficient of variation) of the rtqRT-PCR (n = 10) was 1.32% and 0.60% for 18S rRNA and VgR mRNA, respectively, when a 200-pg template was used in 25-µl reactions. The corresponding interassay variability (n = 8) was 1.68% and 1.47% for 18S rRNA and VgR mRNA, respectively.
Tissue Distribution of VgR mRNA Expression
Transcript abundance of the VgR gene in three different tissues from both genders of white perch was normalized to that of 18S rRNA and reported as a fold change in abundance relative to the values obtained using the interassay standard. Expression of VgR mRNA was observed predominantly in ovaries, although weak expression was also found in other tissues, including liver, muscle, and testis (Fig. 4). The average level of VgR transcripts in the ovaries was 720-fold higher than in somatic tissues.
Maturational Changes in Ovarian VgR mRNA Expression
Changes in gonadosomatic index and the average diameter of the largest oocytes in female white perch sampled throughout oogenesis are shown in Fig. 5. Based on the diameter of the most advanced group of oocytes in their ovaries, females sampled during different months were categorized into four maturational stages according to Jackson and Sullivan [25]: lipidic stage (stage III), cortical granule stage (stage IV), early vitellogenic stage (stage V), and late vitellogenic stage (stage VI). Transcript abundance of the VgR gene was normalized to that of 18S rRNA and reported as a fold change in abundance (relative to the values obtained for the interassay standard) during the reproductive cycle of female white perch (Fig. 6). The highest levels of VgR mRNA expression were found in ovaries from females at stage III, when uptake of Vg into growing oocytes has not yet been initiated. A significant decrease of VgR mRNA expression occurred from stage III to stage IV, and expression levels gradually but significantly decreased from stage IV to stage VI. Ovulated eggs expressed the lowest average level of VgR mRNA, although expression levels were not significantly different from those in ovaries from females at stages V and VI. There was an 190-fold difference in average expression levels between stage III ovaries and ovulated eggs.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Molecular characterization of the full-length VgR has been accomplished for only four vertebrate species, two tetrapods and two teleost fishes. These species include the chicken, Xenopus, rainbow trout, and tilapia [17–23]. Because the chicken VgR is a multifunctional receptor [17, 30, 31], an unambiguous nomenclature based on structural features of the receptor, but not its ligand specificity, has been used to classify this protein. The two forms of chicken VgR are identical with respect to the eight ligand-binding repeats (LR8) at their N-terminus and several other structural features (see Introduction). However, they are encoded by splice variants of a single mRNA transcript and are defined as being LR8+ or LR8– forms of the receptor based on the presence (+) or absence (–) of an O-linked sugar domain immediately proximal to their transmembrane segment [18]. As this definition is convenient for representing structural features of the VgR and other lipoprotein receptors from several classes of vertebrates, we use it along with the original (functional) names of these receptors to discuss findings of the present study. The white perch VgR gene, cloned as cDNA in this study, encodes a protein that lacks an O-linked sugar domain and, with regard to this characteristic and its other structural features, is a typical vertebrate LR8– lipoprotein receptor belonging to the LDLR superfamily of receptors.
As we cloned a full-length cDNA encoding an LR8– VgR from perch ovary and an LR8+ splice variant of the VgR has been detected in the ovary and in somatic tissues of several other species (see Introduction), we attempted to detect an LR8+ version of the VgR in white perch by RT- PCR. Although we used several sets of PCR primers flanking the location of the putative O-linked sugar domain, PCR screening of ovarian and liver mRNA preparations and cDNA libraries failed to detect any sequences encoding an O-linked (LR8+) form of the VgR in white perch. These novel results suggest that the white perch may express only a single, LR8–, type of VgR. The potential physiological significance of the absence of an LR8+ VgR variant in white perch is discussed in more detail below.
We conducted Northern blotting of a poly (A)+ mRNA preparation from perch ovaries whose most advanced clutch of oocytes was at the stage of development (lipid stage, or stage III) characterized by extensive accumulation of lipid droplets, which occurs just before initiation of vitellogenesis in white perch [25]. We observed a single hybridization signal with an apparent size of 4.1 kb, which agrees well with the length of the perch VgR cDNA clone (4021 bp), is only slightly larger than estimates of transcript size obtained by Northern blotting of VgR mRNA from rainbow trout (3.5 kb) [20], tilapia (3.5 kb) [23], chicken (3.5 kb) [18], and Xenopus (3.6 kb) [19], and is similar to results from another Northern blotting study of rainbow trout VgR mRNA (3.9 kb) [22].
In a separate experiment, we had difficulty detecting VgR mRNA by Northern blotting using perch ovaries whose dominant clutch of oocytes was at a later stage of development (stage IV), one characterized by high levels of binding of Vg to its receptor and extensive deposition of Vg-derived yolk proteins in ooplasm granules [8, 25]. This difficulty in detecting VgR mRNA by Northern blotting may have been due to diminished expression of VgR mRNA by oocytes undergoing vitellogenic growth. A more sensitive, conventional RT-PCR method was used by other investigators to confirm the tissue distribution of dual VgR mRNAs in Xenopus and tilapia and successfully detected their expression in the ovary and almost all somatic tissues, while Northern blotting conducted in the same studies failed to detect the messages in somatic tissues [e.g., 19, 23]. However, conventional RT-PCR is not a strictly quantitative method; therefore, these results were not sufficient to confirm differences in the degree of VgR mRNA expression among different tissue types or stages of ovarian development.
Realizing the limitations of conventional RT-PCR procedures, we developed a more sensitive assay to evaluate levels of VgR mRNA expression among different tissues of white perch and among perch ovaries at different stages of growth and maturation. An rtqRT-PCR assay, which is a fluorescence-based kinetic RT-PCR procedure, was developed and fully validated for quantifying perch VgR mRNA in the present study.
The pattern of change in expression of the VgR gene during oocyte development in white perch is similar to that observed in the chicken [32] and rainbow trout [22], and quantitatively confirms the hypothesis that most receptors must be synthesized at early stages of oocyte development, being stored in the ooplasm for later mobilization to support oocyte growth during vitellogenesis. However, there appear to be subtle differences in ovarian VgR expression between white perch and rainbow trout, differences that are probably due to the disparate patterns of oogenesis between the two species. In trout, VgR transcripts were not detected in ovarian follicles greater than 1.0 mm in diameter (mid–late vitellogenic stage) by in situ hybridization [22] nor were they detected in ovarian follicles 3.6 mm in diameter (late vitellogenic stage) by Northern blotting or RT-PCR [20, 22]. In white perch, ovarian expression of VgR mRNA was observed at all maturational stages, although, as in trout, the relative abundance of VgR mRNA transcripts was maximal in previtellogenic ovaries and decreased continuously during vitellogenesis. Perazzolo et al. [22] suggested that apparent VgR expression by trout follicles undergoing mid- to late-vitellogenic growth (1.0–2.5 mm diameter), which was observed by Northern blotting but not in situ hybridization, may have been an artifact arising from the presence of some smaller, previtellogenic oocytes in the ovary samples. Likewise, we believe that the continuous expression of VgR mRNA in maturing perch ovaries is an artifact arising from their multiple-clutch, group-synchronous type of ovarian maturation [25]. Unlike trout ovaries, which are of the single-clutch and group-synchronous type, perch ovaries contain oocytes at all the maturation stages, even at the time of ovulation. Thus, VgR transcripts detected in ovaries whose most advanced oocytes were in late stages of vitellogenic growth (stages V and VI) were probably derived from numerous previtellogenic oocytes (stages III and IV) present in the same tissue samples. Verification of this hypothesis will require evaluation of VgR mRNA expression in situ, as has been accomplished for rainbow trout [22].
Among samples of ovary and eggs from white perch, the lowest level of VgR mRNA expression was found in ovulated eggs, which could be evaluated in the absence of less developed clutches of ovarian follicles, and in which VgR expression is unnecessary. The low levels of VgR transcripts in ovulated perch eggs (190-fold lower than in stage-III ovaries) may represent residual mRNA destined to be degraded, which was detectable in this study due to the higher sensitivity of the rtqRT-PCR assay as compared with methods used in prior studies of trout (Northern blotting in situ hybridization and RT-PCR followed by gel electrophoresis). It also is possible that the presence of VgR mRNA in samples of ovulated eggs is an artifact arising from the process of manually stripping eggs from the white perch by abdominal palpitation. This procedure may have damaged pre- or early-vitellogenic oocytes in the ovary, extruding their ooplasm and VgR mRNA therein into the samples of ovulated eggs.
Based on its high degree of similarity to VgRs of other vertebrate species with regard to genetic (cDNA) and deduced primary structure, tissue distribution of transcript expression, and pattern of gene expression during oogenesis, the white perch LR8– receptor likely functions to deliver Vg into oocytes via receptor-mediated endocytosis [21]. The very low but detectable levels of LR8– transcripts in perch somatic tissues (muscle and liver) and testis are not surprising because similar patterns of localization of LR8– transcripts were observed in the tilapia, Xenopus, and chicken by RT-PCR (see Introduction). The physiological significance of this low level of seemingly ectopic expression of the LR8– VgR in somatic tissues and testis is not known. It is possible that the LR8– receptor may play some role in recycling, metabolism, and/or storage of Vg or other lipoproteins by liver or skeletal muscle of white perch. It is becoming increasingly apparent that the best characterized vertebrate LR8 receptor, the mammalian VLDLR (LR8+ and LR8– forms), is a multifunctional receptor with broad distribution among fatty acid-active tissues that binds numerous different ligands, similar to the chicken VgR [review, 33]. We previously reported on specific binding of Vg to membrane preparations from liver and muscle of white perch [8].
The absence of transcripts for an LR8+ version of the VgR in white perch somatic tissue (and ovary) is surprising because, as noted, VgRs and VLDLRs are expressed in other animals as both LR8+ and LR8– variants, which arise from alternative splicing of a single pre-mRNA. The dual LR8 variant system has been verified to exist in at least five mammalian species (humans [34], cattle [35], mice [36, 37], rats [38], and rabbits [39]) as well as in chickens [17, 18], Xenopus [19], and two teleost fishes (rainbow trout [20] and tilapia [23]).
In the chicken, there was no functional difference between expressed LR8+ and LR8– forms of the receptor with respect to their specificity of ligand binding in vitro [18]. In oviparous vertebrates, aside from the presence or absence of an O-linked sugar domain, the only biological difference between LR8+ and LR8– receptors that has been noted involves their different degree of expression in ovary versus somatic tissues. The highest degree of expression of LR8 receptors (presumably both splice variants) was observed in ovaries of various oviparous species by Northern blotting, while LR8s were either not expressed at all in somatic tissues or they were expressed at very low levels (<1% of ovarian values) [e.g., 17, 18, 20]. These results confirm that the degree of expression of LR8+ receptors in somatic tissues is generally much lower than that of LR8– receptors in the ovary, even though the LR8+ variant is the dominant form of LR8 receptor in somatic tissues (see Introduction). Accordingly, the LR8+ splice variant may be a nonessential or supplemental receptor component in terms of lipoprotein metabolism, one that may even be omitted, as in white perch. In somatic tissues of white perch, the LR8– receptor may act as a substitute for the LR8+ variant and could be involved in lipoprotein (e.g., VLDL) metabolism or some as yet undefined aspects of Vg recycling or storage.
In mammals, the physiological significance of the O- linked sugar domain of the VLDLR is becoming better understood. It appears that the presence of the O-linked sugar domain may promote stable expression of VLDL receptors on the cell surface and its absence may favor release of the receptors into extracellular spaces [39, 40]. However, it is difficult to adapt this scenario to an explanation for selective ovarian expression of LR8– receptors in oviparous vertebrates. Exclusive or preferential expression of the LR8– VLDLR variant has been demonstrated for certain cell types, including bovine aortic endothelial cells [41], human breast carcinoma epithelial cells [42], and human intestinal and gastric cancer cell lines [43]. An LR8– VLDLR is expressed in several cell types during early development of the human brain, whereas the LR8+ receptor variant is preferentially expressed in senile plaques and some neurons and satellite glia from brains of aged donors or patients with Alzheimer disease [44]. Collectively, these findings suggest that dominant expression of an LR8– type of LPR may generally occur in very actively developing or proliferating cells, a hypothesis consistent with selective ovarian expression of the LR8– form of VgR during oogenesis of white perch and other oviparous vertebrates.
Unlike the multifunctional chicken and mammalian LR8s [33] and the perch VgR, functional VgR protein(s) in rainbow trout appear to bind only Vg with high affinity [45]. This explains the fact that uptake of Vg, but not VLDL or other lipoproteins, accounts for more than 80% of the growth of trout oocytes [46]. In contrast, ovulated eggs of white perch and many other teleosts contain a large wax ester-filled oil globule accounting for 80% of total egg lipids, despite the fact that the native Vg carries mainly phospholipids [47]. Mature oocytes of most euryhaline and marine teleosts that spawn pelagic eggs typically contain large droplets of neutral lipids (triglycerides and wax or stearyl esters) that can account for half or more of ooplasm volume [48]. In contrast, teleost Vgs, although they are generally 20% lipid by weight, contain mainly polar lipids such as phosphatidyl-choline or phosphatidyl-ethanolamine. In the white perch and its relatives (genus Morone), the time course of changes in circulating levels of Vg and triacylglycerides relative to ovarian deposition of lipids and yolk proteins suggests that triglycerides, presumably associated with VLDL, may deliver fatty acids (FA) to the ovary, allowing de novo synthesis of wax esters by growing oocytes [47]. Collectively, these observations suggest that Vg is not a major source of ovarian lipids in temperate basses and other teleosts that spawn pelagic eggs, and they imply that there could be major differences with respect to lipoprotein metabolism between these species and fishes, like trout and tilapia, that spawn demersal eggs lacking prominent oil globules [review; 49].
The mechanism by which lipid precursors other than Vg are transported in the bloodstream or taken up by oocytes has not been explored in teleosts. The hypothesis raised for rainbow trout [20], that an ovarian LR8+ variant of the VgR is involved in uptake of these lipid precursors into oocytes, is not applicable in the case of white perch because of the absence of the LR8+ type of VgR. It is possible that ovarian lipoprotein lipase (LPL) may be involved in hydrolysis of circulating lipids derived from sources other than Vg, generating FA for uptake and esterification by growing oocytes. Current models of VLDLR (LR8) function in mammals include several types of interaction of the receptor with LPL in the metabolism of triacylglycerides [33]. High levels of LPL activity and mRNA expression have been observed in ovaries of rainbow trout during vitellogenesis [50], and it has been suggested that the endothelial LPL may release FA from VLDL-associated triacylglycerides, with the free FA then entering oocytes as a source for biosynthesis of neutral lipids [48]. Relevant to this hypothesis is our observation that, while the perch VgR does not bind human lipoproteins [10], it does specifically bind chicken VLDL, suggesting that the LR8– VgR may bind native VLDL or other lipoproteins involved in delivery of FA to growing oocytes. As was the case for the mammalian VLDLR, confirmation of this hypothesis is unlikely to be accomplished via in vitro ligand-binding studies and must await development of appropriate gene knockout technology for fishes [33].
In conclusion, a full-length cDNA encoding an ovarian lipoprotein receptor was cloned from white perch and classified as an LR8– type of VgR based on its molecular structure and its dominant expression in the ovaries. The perch expresses a non-O-linked form of VgR (LR8–) but does not express an O-linked (LR8+) VgR variant, making it the first vertebrate species that has been verified to possess a single LR8 type of receptor. A simple and highly sensitive SYBR-green-based rtqRT-PCR assay was developed and used to quantitatively measure VgR expression in gonadal and somatic tissues for the first time in any vertebrate. Using the assay, we confirmed that the highest level of VgR mRNA expression was found in perch ovaries in which previtellogenic oocytes represented the most advanced stage of gametogenesis. This pattern of VgR or LR8 gene expression appears to be typical of that seen in other oviparous vertebrates and confirms that there must be little or no recruitment of VgR mRNA during vitellogensis, validating the concept that extensive recycling of the VgR must occur during oocyte growth. Although most species of oviparous vertebrates spawn fatty pelagic eggs, the white perch is the only such species for which functional characterization of the VgR [8] and molecular characterization of the VgR gene and its expression (this study) have been accomplished. These developments set the stage for exploitation of the white perch as a model for investigating the physiological mechanisms underlying massive lipidation of teleost oocytes, including the potential involvement of LR7 (e.g., LDLR) and LR8 receptors and/or LPL-based pathways in this process.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Professor Craig V. Sullivan, Room 2115 North Gardner Hall, Department of Zoology, North Carolina State University, Raleigh, NC 27695-7617. FAX: 919 515 5327; craig_sullivan{at}ncsu.edu
3 Current address: Department of Natural Sciences, Clayton College and State University, 5900 North Lee Street, Morrow, GA 30260
Received: 29 September 2003.
First decision: 16 October 2003.
Accepted: 2 February 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2-macroglobulin. J Biol Chem 1995 270:6468-6475
