The upregulation of messenger ribonucleic acids during 17α,20β ...

The upregulation of messenger ribonucleic acids during17á,20â-dihydroxy-4-pregnen-3-one-induced ovulation inthe perch ovary

D M Langenau, F W Goetz and S B RobertsUniversity of Notre Dame, Department of Biological Sciences, Notre Dame, Indiana 46556, USA

(Requests for offprints should be addressed to F W Goetz)

ABSTRACT

While progestins appear to be involved in the localovarian regulation of vertebrate ovulation, theirspecific role is unclear. In yellow perch (Percaflavescens) the progestin, 17á,20â-dihydroxy-4-pregnen-3-one (17,20-P), stimulates ovulation invitro and this induction requires gene activation.Therefore, the perch model was used to isolateprogestin-upregulated mRNAs. Perch ovaries wereincubated for 32 h with or without 17,20-P(0·1 µg/ml). Messenger ribonucleic acids wereisolated from the tissue and used for differentialdisplay PCR (DDPCR). From DDPCR, 5 bandswere eventually obtained that were verified byNorthern analysis to be consistently upregulated by17,20-P at 32 h. Using these bands, full-lengthcDNAs were obtained by library screening andcompletely sequenced. Based on similarity to knownsequences, four of the cDNAs presumably encodefor perch forms of (1) neprilysin (PNEP-1; 63%identical); (2) a lysyl oxidase-type protein (PLO-2;43·2% identical); (3) calmodulin (PCAL-1; 100%identical); and (4) a microtubule aggregate-like

protein (PMAP-1; 29·6% identical). The fifthcDNA obtained from DDPCR most likely encodesfor an egg protein and will be reported separately.Each of the cDNAs was used to probe Northernblots of ovarian mRNA taken at 0, 12, 24, 32 and42 h of incubation with 17,20-P. This temporalNorthern analysis verified that all four wereupregulated by 32 h. In addition, PNEP and PMAPtranscripts began to increase by 12 h, while PCALand PLO transcripts remained elevated through42 h. On Northern blots of RNA from other perchtissues, calmodulin was found in all tissues, PLOmRNA was ovarian specific, and PMAP mRNAwas also present in the gills and liver. Multipletranscripts were observed for PNEP, but theovarian form induced by 17,20-P was only found inhigh abundance in the heart. To our knowledge,this is the first report that specifically characterizesprogestin upregulated mRNAs in the vertebrateovary at ovulation.Journal of Molecular Endocrinology (1999) 23, 137–152

INTRODUCTION

Through the results of a number of diverse studies,progestins have been implicated in the local ovariancontrol of vertebrate ovulation. For example,spontaneous or gonadotropin-induced ovulation inrats can be blocked by steroid synthesis inhibitorssuch as aminoglutethimide or cyanoketone (Lipner& Greep 1971, Lipner & Wendelken 1971).Further, progesterone antiserum and antiprogestinsalso block human chorionic gonadotropin (hCG)-induced ovulation (Mori et al. 1977, Donath et al.1997). In some investigations, the inhibition ofovulation could be restored with exogenous proges-terone (Mori et al. 1977), although this has not

always been the case (Lipner & Greep 1971, Lipner& Wendelken 1971). In certain studies theineffectiveness of exogenous progesterone replace-ment may have been a result of the precise timing ofthe application of progesterone relative to theantagonist. For example, it was demonstrated thathypophysectomy blocked ovulation in rats andthat progesterone could reverse this inhibition(Takahashi et al. 1974). However, the timing andthe duration of exogenous progesterone treat-ment was critical in restoring ovulation in hypo-physectomized animals (Takahashi et al. 1974).

Perhaps the most convincing evidence for aprogestational role in ovulation has come fromstudies using in vitro perfused ovaries. In the

137

Journal of Molecular Endocrinology (1999) 23, 137–1520952–5041/99/023–137 ? 1999 Society for Endocrinology Printed in Great Britain

Online version via http://www.endocrinology.org

perfused rat ovary, inhibition of 3â-hydroxysteroiddehydrogenase blocked luteinizing hormone (LH)-induced ovulation, and this could be restored byprogesterone (Brannstrom & Janson 1989). Proges-terone can also stimulate ovulation by itself in theperfused fowl ovary, indicating a direct action ofprogestins in the bird ovary (Tanaka et al. 1987).However, in several studies using the perfusedrabbit ovary, inhibitors of steroidogenesis wereunable to block gonadotropin-induced ovulation(Holmes et al. 1985, Yoshimura et al. 1986, 1987),indicating that progestins may not be involved inthe control of ovulation in all vertebrates.

Finally, mice carrying a null mutation of theprogesterone receptor are anovulatory even thoughthey can undergo follicular development to pro-duce preovulatory follicles (Lydon et al. 1995).Further, exogenous gonadotropin treatment will notstimulate ovulation in these mutants.

While progestins appear to be involved in the localovarian regulation of ovulation, their specific role isunclear. In the ewe, the inhibition of steroidogenesisby isoxazol blocks ovulation and prostaglandin (PG)F production in the ovary (Murdoch et al. 1986). Ithas been hypothesized that progesterone stimulatesthe conversion of PGE to PGF in the ewe by stimu-lating the enzyme PGE2-9-ketoreductase, therebyincreasing the levels of PGF at ovulation. In supportof this, isoxazol decreases the activity of PGE2-9-ketoreductase while progesterone restores thisactivity (Murdoch & Farris 1988). Thus, one role ofprogestins in the ovary may be to regulate prosta-glandin synthesis. Progesterone inhibits the expres-sion of prostaglandin endoperoxide synthase 2(PGS-2) in the LH-stimulated rat ovary (Hedin &Eriksson 1997). Since inhibition of PGS-2 would leadto a decrease in prostaglandin synthesis, this particu-lar effect of progesterone would appear to be oppositeto the stimulatory role implied by the results of earlierinvestigations. However, it has been hypothesizedthat one function of progesterone may be to restrictthe inflammatory response at the time of ovulation(Hedin & Eriksson 1997). In addition, past inconsist-encies in the effects of progesterone on ovulation mayresult from the possibility that progesterone hasboth stimulatory and inhibitory actions at ovulationdepending on the precise time that it acts.

Besides eicosanoid synthesis, progestins have alsobeen implicated in the ovarian stimulation ofmetalloproteinase inhibitors (Morgan et al. 1994),plasminogen activator (Tsafriri et al. 1987),poly(ADP-ribose) polymerase (Murdoch 1998) andthe kallikrein/kinin system (Tanaka et al. 1992).

In yellow perch (Perca flavescens) the progestin,17á,20â-dihydroxy-4-pregnen-3-one (17,20-P),stimulates both final oocyte maturation (resumption

of meiosis) and ovulation completely in vitro (Goetz& Theofan 1979). As a result, this organism hasproven to be a very useful model in which to studythe regulation of ovulation. Ovulation induced by17,20-P can be blocked by indomethacin (Bradley &Goetz 1994), an inhibitor of prostaglandin endoper-oxide synthase, and it can be restored by primaryprostaglandins (Goetz & Theofan 1979). Thus, inperch it is hypothesized that at least one action of17,20-P on ovulation requires prostaglandins. Moreimportantly, 17,20-P-induced ovulation can beblocked by actinomycin (Theofan & Goetz 1981),indicating that mRNA transcription is necessary.In the current study, differential display PCR(DDPCR) was used to isolate mRNA transcriptsupregulated by 17,20-P in the perch ovary just priorto in vitro ovulation. During this study, fiveprogestin upregulated cDNAs were isolated. Char-acterization of four of the cDNAs is presented inthis paper. The remaining cDNA will be reportedseparately since it is likely that it represents an eggprotein of unknown function. To our knowledge,this is the first report that specifically describesprogestin-upregulated mRNAs in the vertebrateovary at ovulation. In addition, two of the cDNAsreported here encode for proteins that have neverbeen associated with the vertebrate ovary or theprocess of ovulation.

MATERIALS AND METHODS

Animals and ovarian tissue collection

Mature yellow perch females were obtained by hookand line from January to March, or were purchasedin March from commercial suppliers in Madison,WI, USA. Fish were held under natural photo-periods in 300 gallon tanks in running well-water at4 )C. All fish used for in vitro incubations wereprimed approximately 3 days prior to use with 25 IUhCG to initiate germinal vesicle (GV) migration. Tofollow GV migration accurately, eggs were sampleddaily by inserting a 100 µl glass pipette into theovipore and removing 5–10 eggs. The eggs wereexamined microscopically and when the lipiddroplets began to coalesce and the GV movedslightly off-center, fish were over-anesthetized in2-phenoxyethanol and decapitated. Ovaries wereremoved and placed in ice-cold Cortland medium(Wolf & Quimby 1969) containing HEPES (pH 7·8).

In vitro incubations for differential displayPCR and temporal Northern analysis

For large scale in vitro incubations to obtain mRNAfor DDPCR, the ovary was first dissected into large

and others · Progestin upregulated mRNAs in the perch ovary138

Journal of Molecular Endocrinology (1999) 23, 137–152

sheets (100–300 follicles) to ensure that all follicleswere exposed to the medium. These pieces werethen divided into two equal portions based onweight. Each portion was transferred to a 1·0 l flaskcontaining 200 ml Cortland medium and either17,20-P (0·1 µg/ml) or a 95% ethanol vehicle(1·0 µl/ml). The 17,20-P was initially dissolved in95% ethanol at high concentration and added to themedium to obtain the final steroid concentration.Flasks were incubated at 15 )C under intermittentagitation for 32 h.

For the temporal analysis of mRNA expressionfor specific upregulated cDNAs, smaller scaleincubations were conducted. For these incubations,fish were primed and the ovaries dissected into largesheets as described above. The sheets from eachindividual ovary were then divided randomly intonine treatment groups (2–3 mg tissue/treatment)and each group was incubated in 50 ml Cortlandmedium at 15 )C with 17,20-P (0·1 µg/ml) orethanol control (1·0 µl/ml). Control and 17,20-P-treated samples were then assayed at 12, 24, 32, and42 h of incubation. In addition, an untreated zerohour sample was also obtained.

RNA isolation

Following in vitro incubations, ovarian tissues weredeyolked by pressing them between two fine meshstainless steel screens (198 perforations/cm2). Ice-cold Cortland medium was applied to wash thesamples while the screens were gently pressedtogether. This procedure was used to separate theextrafollicular tissue and follicle layers from thechorion components and egg yolk of matureoocytes. Tissues were scraped into beakers under astream of ice-cold Cortland medium and thendistributed to 50 ml polypropylene tubes. Sampleswere lightly vortexed, spun at 3000 r.p.m. for 5 minat 4 )C, and the supernatant was decanted intonew 50 ml tubes held in ice. The remaining tissuepellet was resuspended in 30 ml new Cortlandmedium, vortexed, and spun again under the sameconditions. This procedure was repeated twice. Thesupernatants from the second and third washeswere removed and discarded, and the tissue pelletremaining after the last wash was homogenized in3 ml Tri-Reagent (Molecular Research Center,Inc., Cincinnati, OH, USA) with a TissueTearor(Biospec, Bartlesville, OK, USA). The supernatantremaining after the first wash was centrifuged at10 000 r.p.m. for 5 min at 4 )C. The supernatantwas decanted and the small tissue pellet thatremained was resuspended in 500 µl Tri-Reagent.The tissue homogenates were then combinedand briefly homogenized. All samples were then

stored at "70 )C until RNA extractions could becompleted.

RNA isolation was completed as describedpreviously (Chomczynski & Sacchi 1987,Chomczynski 1993). Briefly, tubes were thawed atroom temperature and chloroform was added to thesamples (0·2 ml chloroform/1·0 ml Tri-Reagent).The samples were shaken vigorously and then heldat room temperature for 15 min. Following a 15 mincentrifugation at 12 000 g the colorless upper phasecontaining the RNA was transferred to a new 15 mlRNAse-free tube and precipitated for 10 min withisopropanol (0·5 ml/1·0 ml of initial Tri-Reagentvolume). Samples were then centrifuged at 12 000 gfor 20 min and the supernatant was decanted. TheRNA pellet was washed twice in 6 ml 75% ethanoland dried for 10 min at room temperature. TheRNA pellet was resuspended in 500 µl RNAse-freewater and held at "70 )C until mRNA isolation wasperformed using the Poly-A-Tract mRNA IsolationSystem (Promega, Madison, WI, USA).

Differential display PCR (DDPCR)

The mRNA from large scale ovarian incubations onfive female perch was used for DDPCR analysis(RNAmap; GenHunter Corp., Nashville, TN,USA). For each sample (i.e. control and 17,20-Ptreatments/ovary) cDNA was made from purifiedmRNA (0·l µg) using Superscript II reversetranscriptase (GibcoBRL Life Technologies, GrandIsland, NY, USA) and either G or C anchorprimers (T12MG or T12MC; GenHunter Corp.).The resulting cDNA was then subjected to PCRusing the same oligo dT anchor primers and arandom 10 bp primer (AP1 through AP15; Gen-Hunter Corp.) in the presence of [á-33P]dATP(2000 Ci/mmol; DuPont NEN, Boston, MA, USA).Labeled products were separated on 6% denaturingpolyacrylamide gels (32·5 cm by 50·0 cm by0·4 mm) at 2500 V for 5 h. Gels were then blottedwithout fixing and exposed to x-ray film for 2 days.Following alignment of the x-ray film and theoriginal gel blot, differential bands were excisedusing a razor blade.

Gel pieces were soaked in 100 µl water and boiledfor 20 min in 2·0 ml microcentrifuge tubes. Thesamples were spun in a microcentrifuge and thesupernatant was transferred to a new 2·0 ml tube.The DNA was precipitated in the presence ofglycogen using 3 M sodium acetate and absoluteethanol. The samples were then centrifuged for10 min at 13 000 g to pellet the DNA. PelletedDNA was washed in 85% ethanol, dried for 10 min,and then resuspended in 10 µl water. This DNAwas reamplified by PCR using the same primer

Progestin upregulated mRNAs in the perch ovary · and others 139


pairs used to generate the original DDPCR bands.Reamplified products were separated on agarosegels, and bands of interest were excised and gelextracted (Qiaex II; Qiagen, Chatsworth, CA,USA). Gel purified DNA was ligated into pCRII(In Vitrogen, San Diego, CA, USA) and used totransform E. coli INVaF* cells (Original TACloning Kit, In Vitrogen). Positive colonies werescreened for inserts using PCR. Of the colonies thatcontained inserts of the appropriate size DNA,three clones per original DDPCR band wereselected for sequencing. Sequencing was completedon plasmid DNA using a modified dideoxy chaintermination method (Cy5 AutoRead Sequencingkit; Pharmacia Biotech, Madison, WI, USA) withCy5-labeled vector primers flanking the DNAinsertion site. The sequencing reactions wereseparated and analyzed using an ALFexpressSequencer (Pharmacia Biotech). If the sequences ofall 3 clones were identical, the DNA of one of theclones was used as a probe for Northern analysis toverify that the transcript was upregulated asobserved in the original DDPCR.

Northern analysis

Several types of Northern analysis were completed.Initial verification of 17,20-P upregulation at 32 hwas completed on ovarian mRNA from five toseven females depending on the clone. Secondly,Northern analysis was used to determine the exacttemporal expression pattern of verified 17,20-Pupregulated clones using mRNA extracted from 2–4individual fish. Finally, Northern analysis wascompleted to determine the general tissue expres-sion pattern of the four upregulated mRNAtranscripts obtained from the DDPCR screen. AllNorthern analyses were performed in the samemanner as described below.

Messenger ribonucleic acid (0·5 µg/lane) wasseparated on formaldehyde-agarose gels (1·6%agarose, 2·2 M formaldehyde, and 1#MOPS (3-(n-morpholino) propanesulfonic acid)). Samples weremixed 1:1 with a reaction/loading buffer containing4·2 M formaldehyde, bromophenol blue (0·005 mg/ml), 10% glycerol, and 2#MOPS. Samples wereheated at 55 )C for 15 min prior to loading. Gels wererun at 55 V until the dye front had migrated 8 cm.After washing, mRNA was transferred to nylonmembranes (Magna Charge Nylon Transfer Mem-branes, Micron Separations Inc., Westborough,MA, USA) by downward capillary elution using20#SSC (3 M NaCl, 0·3 M sodium citrate, pH 7·2).After washing in 2#SSC, membranes were cross-linked (Stratalinker; Stratagene, La Jolla, CA, USA)and stored under vacuum.

Nylon membranes were pre-hybridized for atleast 2 h in roller tubes at 42 )C in a buffercontaining 5#SSPE (0·75 M NaCl, 0·05 M sodiumphosphate monobasic, 5 mM EDTA, pH 7·4), 0·1%SDS, 5#Denhardt’s solution, 50% formamide, and150 µg/ml calf thymus DNA. The Northern blotswere probed with a radiolabeled, double strandedinsert obtained either from EcoRI digests of pCRIIproducts obtained from DDPCR plasmid preps orfrom Xhol/EcoRI digests of full-length clonesobtained from cDNA library screening (see below).Inserts were labeled with [á-32P]dATP (3000 Ci/mmol; ICN, Irvine, CA, USA) using Klenow andrandom primers (Prime-It II, Stratagene). Thelabeled probe was purified using a gel filtration spincolumn (CENTRI-SEP; Princeton Separations,Adelphia, NJ, USA) and heat denatured in boilingwater for 5 min. The denatured probe was addeddirectly to the pre-hybridization buffer and incu-bated with the Northern blot overnight. Followinghybridization, the blots were washed twice (15 mineach) under medium stringency (1#SSPE, 0·1%SDS, 45 )C) and twice (15 min each) underhigh stringency (0·1% SSPE, 0·5% SDS, 65 )C).Northern blots were dried, exposed to phosphor-imaging screens, and analyzed using a Storm 840phosphorimager (Molecular Dynamics, Sunnyvale,CA, USA). For the initial Northern verification of17,20-P upregulation at 32 h of incubation, RNAbands were quantified using Imagequant (MolecularDynamics) and analyzed statistically using pair-wiset-tests (Zar 1996).

Full-length cDNAs

DNA bands produced by DDPCR represent only asmall 3* region of the transcript; thus, completecDNAs were obtained by library screening. AcDNA library was constructed in Unizap (ZapExpress, Stratagene) using pooled mRNA fromperch ovaries stimulated for 32 h with 17,20-P. Thelibraries were plated at 20 000 to 50 000 plaquesper plate (12 cm diameter) and lifted to nylonmembranes (Magna Charge Nylon Transfer Mem-branes, Micron Separations Inc.). Lifts werescreened using á-32P-labeled DDPCR inserts asdescribed above (see Northern analysis section).Blots were hybridized and washed exactly asindicated above for Northern blotting and wereexposed to X-ray film for two days. Positive plaqueswere rescreened once to homogeneity. Plaques wereexcised (Rapid Excision Kit, Stratagene) and theresulting bacterial colonies were verified for insertsusing colony PCR. Plasmid DNA from clonesof interest were either (1) fully sequenced withinternal Cy-5-labeled primers (Thermo Sequenase



Fluorescent Labeled Primer Cycle Sequencing kit,Amersham Life Science, Arlington Heights, IL,USA); (2) subcloned into pCRII and sequencedusing Cy-5-labeled vector primers (Thermo Seque-nase Fluorescent Labeled Primer Cycle Sequencingkit, Amersham Life Science); and/or (3) manuallysequenced using internal primers (T7 SequenaseVersion 2·0 DNA Sequencing kit, Amersham LifeScience). The sequences for inserts obtained fromDDPCR and library screened colonies were com-pared to verify that the cDNAs obtained throughlibrary screening were the same as the DDPCRfragment. Full-length sequences were comparedusing BLAST (Basic Local Alignment Search Tool,National Center for Biotechnology Information)with known sequences in GenBank to determinesimilarity to known proteins.

Tissue expression

Tissue expression was determined for those DDPCRclones that showed upregulation by Northern analy-sis. Fish used for these studies were not primed withhCG but were obtained at the same time of the yearand contained ovaries with fully mature postvitello-genic follicles. Fish were over-anesthetized and de-capitated. Whole brain, heart, liver, spleen, headkidney, and gill filaments were dissected andweighed (approximately 0·2 mg each) on ice. In ad-dition, 0·2 mg samples of the trunk kidney (opistho-nephros), gastrointestinal tract (just posterior to thestomach) and muscle (anterior to the caudal ped-uncle) were collected. RNA/mRNA extractions andNorthern blotting were performed as describedabove and mRNA blots were probed with full-lengthinserts obtained from library screening.

RESULTS AND DISCUSSION

DDPCR generates progestin-upregulatedbands in the perch ovary

DDPCR was performed on the mRNA obtainedfrom perch ovaries following 32 h of incubationwith 17,20-P (0·1 µg/ml). This particular time pointwas chosen since it falls just prior to the time of17,20-P-induced ovulation in vitro (34–36 h) (Goetz

1. Complete nucleotide and amino acidsequence of PNEP-1. Nucleotides numbered on the left,amino acids on the right. The proposed methionine(ATG) start site is underlined and potentialglycosylation sites are underlined twice(Asn-X-Ser/Thr). The stop codon is indicated byasterisks (***). GenBank Accession No. AF077612.



et al. 1989), and is past the actinomycin-sensitiveperiod for 17,20-P-stimulated ovulation (Theofan &Goetz 1981). For DDPCR, two anchor primers(T12MG or T12MC) were used in combinationwith 15 random primers (AP-1 to AP-15). Thisgenerated approximately 60 bands on DDPCR thatinitially appeared to be upregulated by 17,20-P.While most of these bands could be reamplified andcloned, a significant number either did nothybridize with mRNA on Northern blots or did notresult in consistent upregulation with 17,20-P. Inaddition, based on the comparison of initialsequences for DDPCR bands, it was determinedthat some bands were derived from the samecDNAs. This was particularly true when multiplebands were obtained from reactions with the sameprimer pairs. From Northern analysis, 5 DDPCRbands were eventually verified to be consistentlyupregulated by 17,20-P at 32 h. Of these, one isbelieved to be an egg protein based on sequencehomology and will be reported separately.

In all in vitro incubations used in this study,17,20-P induced complete germinal vesicle break-down by 32 h, and in the temporal Northernanalysis oocytes had ovulated by the 42 h sample. Incontrast, follicles in control incubates did not

undergo germinal vesicle breakdown or ovulationthroughout the incubations.

Neprilysin-like perch cDNA

Using a 350 bp upregulated cDNA obtained fromDDPCR, a full-length cDNA clone (PNEP-1;GenBank Accession No. AF077612) of 3985nucleotides was obtained by library screening.PNEP-1 contains two in-frame methionine initia-tion sites (nucleotides 196 and 280) that wouldresult in proteins of either 740 or 770 amino acids inlength (Fig. 1). Neither of the nucleotide sequencesflanking the two methionines show similarity to theKozak consensus sequence for eukaryotic initiationsites (Kozak 1986). Assuming that the cDNAencodes for the larger 770 amino acid protein, thenPNEP-1 also contains a 195 bp untranslated 5*region and a 1477 bp untranslated 3* region.

Based on amino acid similarity to GenBanksequences, the protein encoded by PNEP-1 appearsto be a perch neprilysin (enkephalinase/neutralendopeptidase EC 3·4·24·11). The PNEP-1 proteinis 63% identical and 74% similar to mouse neprilysin(Fig. 2). In addition to sharing a high degree ofsequence similarity, the PNEP-1 protein shares

2. Amino acid alignment of the PNEP-1 protein with other known mammalian neprilysins. Amino acidsequences for perch, mouse (Chen et al. 1992), and human (Shipp et al. 1988) neprilysin were aligned using theMultiple Alignment Construction and Analysis Workbench (MACAW, National Center for BiotechnologyInformation). A consensus sequence (CONCS) in which all three sequences share identical residues is provided.Conserved domain structures (cytoplasmic domain (<<<CYTOPLASMIC DOMAIN<<<L), membrane spanningregion (T>MEMBRANE SP DOM<A), and extracellular domain (T>>>EXTRACELLULAR DOMAIN>>>) areindicated above the amino acid alignment. Conserved zinc-binding domains are underlined twice. Amino acids arenumbered on the right and correspond to individual proteins.



several characteristics of the neprilysin family ofzinc metallopeptidases (Hooper 1994). For example,all cysteine residues are conserved in the perchneprilysin as are the two zinc binding domains thatare highly characteristic of mammalian neprilysins(Fig. 2).

Neprilysin is a membrane-bound enzyme,anchored by a hydrophobic region that is flanked bya short N-terminal and long C-terminal hydro-philic stretch of amino acids (Dion et al. 1997). Inrabbit neprilysin, the short N-terminal cytoplasmicdomain consists of 27 amino acid residues; however,because the perch transcript appears to encode for aslightly larger protein, residues 1–47 may act as thecytoplasmic domain in the PNEP-1 protein. Aminoacids 47–69 presumably encode for a 23 amino acidplasma membrane spanning region that is highlyhydrophobic and the same length as the mammalianneprilysin plasma membrane spanning domain.Finally, amino acids 70 to 770 presumably encodefor the extracellular hydrophilic domain which has12 potential glycosylation sites (Fig. 1). In contrast,there are 5–6 glycosylation sites in mammalianneprilysins. While there are some differencesbetween the mammalian neprilysin and the PNEP-1protein, the high overall sequence identity andconserved neprilysin characteristics strongly suggestthat the PNEP-1 protein is a perch neprilysin.

On mRNA blots of ovarian tissue probed with theoriginal 350 bp DDPCR PNEP-1 band, a 4·0 kbtranscript was significantly (P=0·005) upregulatedby progestin treatment at 32 h in all fish examined(Fig. 3A). Given the similar size, we assume thatPNEP-1 corresponds to this transcript. The specifictemporal pattern of expression of the 4·0 kbtranscript was completed on the ovaries of threefish. In all three fish, levels of the 4·0 kb transcriptremained low in controls but were elevated from 12to 32 h in 17,20-P-treated ovaries (Fig. 3B). Levelsof the 4·0 kb transcript decreased at 42 h; howeverthey were still elevated in steroid incubates whencompared with 42 h controls. Neprilysin has alsobeen reported to be regulated by progestins in thehuman endometrium (Casey & MacDonald 1996).Specifically, in endometrial stromal cultures,neprilysin mRNA and protein levels increased inthe presence of medroxyprogesterone acetate andestradiol-17â (Casey et al. 1991).

Overall, four transcripts of 6·0 (results notshown), 4·0, 2·5, and 1·0 kb were detected whenPNEP-1 was used to probe mRNA Northern blotsof other perch tissues (Fig. 3C). The larger 6·0 kbtranscript was detected in the heart and trunkkidney (data not shown), while the 4·0 kb transcriptthat was regulated by 17,20-P in the ovary, was alsopresent in high abundance in the heart. A smaller

band encoding for a 2·5 kb transcript was detectedin the gills, gut, heart, head kidney, liver, spleen,and ovary. Finally, a small 1·0 kb transcript was

3. Northern blots of PNEP transcripts in theovary (A and B) and in other perch tissues (C). Eachlane represents 0·5 µg mRNA. (A) Expression of PNEPmRNA at 32 h under 17,20-P stimulation (0·1 µg/ml, S)or ethanol vehicle (1 µl/ml, C) in the ovaries from fivefish (F1-F5). The 4·0 kb transcript was upregulated by17,20-P in all five fish assayed (P=0·005; pair-wiset-test). (B) Representative Northern blot of the temporalexpression pattern of PNEP ovarian transcripts in onefish. Numbers above the lanes represent hours ofincubation in Cortland medium with 17,20-P (0·1 µg/ml,17,20P) or ethanol vehicle (1 µl/ml, Control). Two otherfish assayed showed a similar pattern of expression.(C) PNEP transcripts in other perch tissues. H. Kidney,head kidney; Ovary-C, ovary at 42 h in the presence ofethanol vehicle; Ovary-S, ovary at 42 h under 17,20-Pstimulation.



detected in all tissues assayed with the exception ofthe brain. Multiple neprilysin transcripts rangingfrom 2·7 to 5·7 kb have also been reported in variousmammalian cells or tissues (Shipp et al. 1988, Chenet al. 1992). In addition, localization of perchneprilysin mRNAs in various tissues throughout thebody is consistent with the previous observationthat neprilysin is located in the kidney, lung,intestine, and parts of the central nervous system inmammals (Erdos & Skidgel 1989).

Mammalian neprilysin inactivates small peptidessuch as enkephalins, substance P, atrial natriureticpeptide, endothelins, bradykinin, neurotensin, andchemotactic peptide (for review see Erdos & Skidgel1989). Some of these factors have been observed inthe vertebrate ovary (Ackland et al. 1992) and somehave also been associated with the ovulatory process(Murdoch & McCormick 1989, Hellberg et al. 1991,Furudate et al. 1994). Given that many of the smallpeptides which are cleaved by neprilysin arevasoactive, one possible role for neprilysin in theovary would be to regulate the activity of thesepeptides and, as a consequence, influence follicularblood flow. Alternatively, peptides such as endo-thelins have also been associated with the control ofovarian steroidogenesis (Tedeschi et al. 1992), andkinins have been implicated in oocyte expulsion(Hellberg et al. 1991). Thus, the actions ofneprilysin in the ovary may extend beyond vascularcontrol. Further, since neprilysin can cleavechemotactic peptides (Painter et al. 1988), it mayregulate the migration of leukocytes that is observedin the ovary at the time of ovulation (Brannstromet al. 1993).

Lysyl oxidase-like perch cDNA

Using a 257 bp upregulated cDNA obtained fromDDPCR, several larger cDNAs were eventuallyobtained by library screening and fully sequenced.The largest cDNA was 3438 bp (PLO-2; GenBankAccession No. AF143003) and contained an openreading frame that was 2685 bp, presumablyencoding a protein of 895 amino acids (Fig. 4). Aslightly smaller cDNA (3354 bp; PLO-1, GenBankAccession No. AF103901) was also completelysequenced (results not shown). PLO-1 encoded fora protein that was nearly identical with PLO-2, butdid not contain an initiation codon and lacked theN-terminal 73 amino acids of PLO-2. In addition,

4. Complete nucleotide and amino acidsequence of PLO-2. Nucleotides numbered on the left,amino acids on the right. The stop codon is indicated byasterisks (***). GenBank Accession No. AF143003.



the untranslated 5* region of PLO-2 was alsomissing. In contrast, PLO-1 contained 273 bp in the3* untranslated region that were not found inPLO-2. Even though PLO-2 appears to contain acomplete open reading frame, it is probable that thisis still not the complete cDNA since a single ovariantranscript of just less than 4·0 kb is detected onNorthern blots using these cDNAs as a probe (seebelow).

The presumed protein encoded by the PLOs issimilar to the lysyl oxidase enzyme family thatincludes lysyl oxidase (LO) (Hamalainen et al.1991), lysyl oxidase-like protein (LOL) (Kenyonet al. 1993), and lysyl oxidase-related protein (LOR)(Saito et al. 1997). These proteins show the greatestsimilarity to one another and to the PLO protein inthe carboxy region (Fig. 5). Overall, LO, LOL,LOR and PLO proteins are 41·2% identical over aC-terminal region of 206 amino acids. However, ofthe three mammalian lysyl oxidase proteins, theperch protein is most similar to the LOR protein in

the same C terminal region (68·0% identity). Overthe entire LOR sequence, PLO-2 encodes for aprotein that is 48·4% identical to LOR. A directcomparison of all of the lysyl oxidase proteins iscomplicated by the fact that LOR and the PLOproteins are significantly longer than LO or LOL,and the alignment suggests that both LOR and thePLO proteins are greatly extended between theextreme amino and carboxy regions (Fig. 5). In fact,the PLO-2 protein is the longest lysyl oxidase-likeprotein yet characterized.

The PLO cDNAs detected a single transcript onNorthern blots of perch ovaries that was slightlyless than 4·0 kb. This transcript increased signifi-cantly (P=0·01) at 32 h in all ovaries treated with17,20-P (Fig. 6A). In temporal experiments on theovaries of three fish, PLO transcripts increased inthe ovary at 32 and 42 h following 17,20-Pstimulation while PLO transcripts were notdetected in any controls or in 17,20-P-treatedovaries at 12 or 24 h (Fig. 6B). A glucocorticoid

5. Amino acid alignment of the PLO-2 protein with other human lysyl oxidase proteins. Amino acidsequences for perch (PLO2), human lysyl oxidase-related protein (LOR) (Saito et al. 1997), human lysyl oxidase (LO)(Hamalainen et al. 1991) and human lysyl oxidase-like protein (LOL) (Kenyon et al. 1993) were aligned usingMACAW. A consensus sequence (Con) in which perch and human LOR protein share identical residues is provided.Similarly, asterisks (*) below the consensus sequence denote identical (Id) residues in all four proteins. Amino acidsare numbered on the right and correspond to individual proteins.



response element is present in the 5* flanking regionof the mammalian LOL gene (Smith-Mungo &Kagan 1998), thus supporting the hypothesis thatsteroids can directly regulate the expression of theseproteins.

PLO transcripts appear to be present only in the17,20-P-stimulated ovary; however a very lowexpression of what appears to be a slightly smallertranscript may occur in the brain and gills (Fig. 6C).Human LO and LOL are expressed constitutively

in high abundance in many different tissues.Specifically, LO has been localized to human heart,placenta, lung, liver, skeletal muscle, kidney, andpancreas while LOL transcripts have been detectedin human lung, kidney, placenta, liver, brain,pancreas, heart, and skeletal muscle (Kim et al.1995). Although the tissue localization of humanLOR mRNA and protein have not been studied,LOR transcripts were first detected in fibroblastcells and are constitutively expressed in fetal lungfibroblast cell lines (Saito et al. 1997). Since the4·0 kb perch PLO transcript was detected exclu-sively in the ovary, this may indicate that it is eithernonconstitutively expressed and induced only whenneeded by various mediators, or is a novel ovarianspecific lysyl oxidase gene upregulated at ovulationand/or inflammation.

The final step in the posttranslational modifi-cation of collagen within the extracellular matrixis the crosslinking of collagen chains and fibrils(Franklin 1997). Lysyl oxidase is the enzyme thatdeaminates lysine residues causing spontaneouscrosslinking between collagen fibrils, thus makingthe collagen tough and insoluble. The same enzymealso acts on elastin, producing similar cross-linking.Based on their similar protein structures, the othermembers of the lysyl oxidase family of proteins arealso believed to be able to cross-link collagen andelastin fibers (Kim et al. 1995, Saito et al. 1997).Since ovulation is marked by an increase in proteaseactivity and follicle degradation (Espey 1980, Goetzet al. 1991), restructuring of the ovary followingovulation would be expected. Lysyl oxidase activitycould be involved in tissue repair followingovulation by crosslinking collagen or elastin fibrils.In the perfused rabbit ovary, lysyl oxidase activitypeaks just after ovulation and was hypothesized toplay an integral role in gonadal restructuring(Himeno 1986). However, in that study lysyloxidase activity was assayed indirectly, and LOprotein or mRNA characterization was not con-ducted (Himeno 1986). Given the structure of theperch lysyl oxidase protein isolated in the presentstudy and the apparent ovarian-specific expression,an intriguing hypothesis is that the lysyl oxidaseenzyme activity observed in rabbits after ovulationactually results from other lysyl oxidase enzymesinstead of LO itself. Regardless, upregulation ofPLO transcripts by 17,20-P at 42 h (postovulation),may indicate that lysyl oxidase enzyme activityregulates gonadal restructuring in perch in a similarfashion as hypothesized for rabbits.

As is the case for neprilysin, lysyl oxidase proteinshave been implicated in the stimulation of otherphysiological responses in addition to the cross-linking of collagen and elastin. For example, LO

6. Northern blots of PLO transcripts in theovary (A and B) and in other perch tissues (C). Eachlane represents 0·5 µg mRNA. (A) Expression of PLOmRNA at 32 h under 17,20-P (0·1 µg/ml, S) or ethanolvehicle (1 µl/ml, C) in the ovaries from seven fish(F1-F7). The 4·0 kb transcript was upregulated by17,20-P in all seven fish assayed (P=0·01; pair-wiset-test). Note: longer exposure time is required toobserve upregulation in fish F5. (B) RepresentativeNorthern blot of the temporal expression pattern ofPLO ovarian transcripts in one fish. Numbers above thelanes represent hours of incubation in Cortland mediumwith 17,20-P (0·1 µg/ml, 17,20P) or ethanol vehicle(1 µl/ml, Control). Two other fish assayed showed asimilar pattern of expression. (C) PLO transcripts inother perch tissues. Kidney, trunk (opisthonephros)kidney; all other abbreviations as in Fig. 3.



acts as a monocyte chemotactic substance (Lazaruset al. 1994) and thus could be involved in leukocytemigration into the ovary at ovulation. Further, theinherent enzyme activity of LO produces hydrogenperoxide as a byproduct (Green et al. 1995), and ithas been shown in fish that low levels of hydrogenperoxide can induce ovulation in vitro whencombined with certain signal transduction me-diators such as sodium orthovanadate (Hsu & Goetz1991).

Yellow perch calmodulin

Using a 190 bp upregulated cDNA obtained fromDDPCR, a 1644 bp clone (PCAL-1; GenBankAccession No. AF085250) was obtained by libraryscreening. PCAL-1 contains a 447 bp open readingframe that is initiated at a methionine encoded bynucleotides 40–42 (Fig. 7). The cDNA also containsa 39 bp 5* untranslated region and a 1104 bp 3*untranslated region.

Based on sequence comparison with GenBank,the protein encoded by PCAL-1 is definitelycalmodulin. At the amino acid level, the openreading frame of this clone is 100% identical to othervertebrate calmodulins including those from Xeno-pus laevis (Chien & Dawid 1984), medaka (Matsoet al. 1992), humans (Fischer et al. 1988), andchickens (Putkey et al. 1983). At the nucleotidelevel, PCAL-1 is approximately 86% identical toother vertebrate calmodulins.

Two transcripts (1·6 and 0·6 kb) were detected onNorthern blots of ovarian RNA when probed withPCAL-1; however, only the larger 1·6 kb transcriptwas significantly (P=0·01) upregulated by 17,20-Pafter 32 h of incubation (Fig. 8A). Temporal mRNANorthern analysis was completed on the ovaries offour fish and confirmed that only the 1·6 kbcalmodulin transcript was regulated by 17,20-P.Three of the four fish assayed showed strongupregulation of calmodulin transcripts at 32 and/or42 h (Fig. 8B). One fish exhibited a variable

7. Complete nucleotide and amino acidsequence of perch calmodulin. Nucleotides numbered onthe left, amino acids on the right. The proposedmethionine start site is underlined (Met). The stopcodon is indicated by asterisks (***). GenBankAccession No. AF085250.

8. Northern blots of calmodulin transcripts inthe ovary (A and B) and in other perch tissues (C). Eachlane represents 0·5 µg mRNA. (A) Expression of perchcalmodulin mRNA at 32 h under 17,20-P (0·1 µg/ml, S)or ethanol vehicle (1 µl/ml, C) in the ovaries from sevenfish (F1-F7). The 1·6 kb transcript was upregulated by17,20-P in all seven fish assayed (P=0·01; pair-wiset-test). (B) Representative Northern blot of the temporalexpression pattern of calmodulin ovarian transcripts inone fish. Numbers above the lanes represent hours ofincubation in Cortland medium with 17,20-P (0·1 µg/ml,17,20P) or ethanol vehicle (1 µl/ml, Control). Of threeother fish assayed, two showed a similar pattern ofexpression and one did not. (C) Calmodulin transcriptsin other perch tissues. Abbreviations as in Figs 3 and 6.



pattern of transcript expression in which nodifference between steroid treatment and controlswas observed at most time points (data not shown).However, calmodulin transcripts in this fishincreased in 17,20-P-stimulated incubates as timeprogressed, indicating that calmodulin upregulationby 17,20-P occurred in this fish as well.

The large 1·6 kb calmodulin transcript wasdetected in all other tissues assayed (Fig. 8C).

Calmodulin is found throughout the body andhas been detected in all living cells to date (Celioet al. 1996). Thus, the ubiquitous expression ofcalmodulin in perch is not surprising.

Calmodulin mediates many cellular responsesincluding cell signaling, gene regulation, DNAsynthesis, mitosis, smooth muscle contraction, andapoptotic cellular events (Celio et al. 1996). Thus,delineating a specific role for calmodulin in theovary may be difficult. However, it appears thatcalmodulin plays a role in inducing ovulation inrabbits through the activation of kinases. Calcium/calmodulin-dependent protein kinase II inducedovulation in the absence of gonadotropin stimu-lation in the perfused rabbit ovary (Kugu et al.1995). Similarly, in hCG-stimulated ovaries treatedwith calmodulin inhibitors, a significant decrease inovulation efficiency was observed compared withcontrol ovaries treated only with hCG (Kugu et al.1995). Thus, ovarian calmodulin probably interactswith kinases to alter the phosphorylation state ofovarian proteins, ultimately controlling ovulation inthe rabbit ovary. Calmodulin could also mediateovulation by regulating smooth muscle contractionin the follicle wall. The theca externa of quail andhamster follicles contains smooth muscle-like cellsand calmodulin has been localized to chicken thecalcells, smooth muscle bundles, and fibroblasts oflarge follicles (Yoshimura et al. 1990).

Finally, calmodulin has also been implicated inthe regulation of steroidogenesis (Alila et al. 1990,Johnson & Tilly 1990, van der Kraak 1991) andfinal oocyte maturation (Wasserman & Smith 1981,Bornslaeger et al. 1984, Sato 1990) in variousvertebrates. Thus, increases in calmodulin duringprogestin stimulation may be related to otherovarian processes besides ovulation.

9. Complete nucleotide and amino acidsequence of PMAP-1. Nucleotides numbered on the left,amino acids on the right. The proposed methionine startsite is underlined (Met). Potential glycosylation sites areunderlined twice (Asn-X-Ser/Thr). The stop codon isindicated by asterisks (***). GenBank Accession No.AF085251.

10. Amino acid alignment of PMAP-1 (PMAP) with human MAP-p44 (HUMAN). Deduced proteinsequences for PMAP-1 and human MAP-p44 (Takahashi et al. 1990) were aligned using MACAW. A consensussequence (Con) in which both sequences share identical residues is provided. Amino acids are numbered on the rightand correspond to individual proteins.



Perch microtubule aggregate protein-likecDNA

Using a 450 bp upregulated cDNA obtained fromDDPCR, a 1111 bp cDNA (PMAP-1; GenBankAccession No. AF085251) was obtained by libraryscreening. PMAP-1 encodes for an 882 bp openreading frame with the presumed methionineinitiation site occurring at nucleotides 19–21 andtermination occurring at nucleotides 901–903(Fig. 9). Although the sequence flanking theproposed methionine initiation site conforms wellwith Kozak sequences, it is possible that translationis initiated further upstream beyond the boundaryof this clone since the reading frame is openupstream of the proposed ATG start site. How-ever, it is likely that PMAP-1 is the full-lengthcDNA, since library screening did not yield largercDNAs and since PMAP-1 hybridizes with atranscript on Northern blots that is nearly identicalin size (see below). Assuming that the initiation siteis at nucleotide 19, PMAP-1 has an 18 bp 5*untranslated region and a 211 bp 3* untranslatedregion.

The deduced amino acid sequence of PMAP-1shows similarity to only two proteins in GenBank:human and chimpanzee microtubule aggregateprotein, MAP-p44 (Takahashi et al. 1990) (Fig. 10).The presumed protein encoded by PMAP-1 is29·6% identical and 58·5% similar to chimpanzeeMAP-p44. The PMAP-1 protein does not appear tohave a signal sequence, indicating that the protein isnot secreted.

A single 1·1 kb transcript was detected on allNorthern blots assayed. Transcript levels forPMAP significantly (P=0·012) increased in thepresence of 17,20-P after 32 h of incubation in allfish assayed (Fig. 11A). In two fish assayed fortemporal expression, PMAP levels were elevatedfollowing 32 h with 17,20-P (Fig. 11B). By 42 h,transcript levels decreased in both fish (Fig. 11B).While it appears from this figure that PMAPtranscripts increase by 12 h, it should be noted thatin the other fish assayed, no upregulation wasobserved at 12 h (data not shown). From thistemporal expression pattern, it seems likely thatPMAP mRNA levels increase by at least 32 h andthen decline following ovulation.

High transcript levels of PMAP transcriptswere detected in the gills, liver, and ovary whilevery low levels of transcript were detected in thetrunk kidney (Fig. 11C). MAP-p44 has only beenstudied in the liver (Honda et al. 1990, Takahashiet al. 1990), and no localization studies havebeen completed to determine MAP-p44 tissuedistribution in mammals.

Microtubule aggregate protein-p44 is expressedin human and chimpanzee hepatocytes infectedwith hepatitis C or D (Honda et al. 1990, Takahashiet al. 1990) and it can be induced by interferons(Kitamura et al. 1994). This protein associateswith microtubules; however, its specific biologicalfunction remains unknown. Given that the PMAP-1protein shares some similarity to these proteins,it may interact with microtubules in a mannersimilar to that described for MAP-p44 in humansand chimpanzees. However, as with MAP-p44,the ultimate function in the ovary remainsunknown.

11. Northern blots of PMAP transcripts in theovary (A and B) and in other perch tissues (C). Eachlane represents 0·5 µg mRNA. (A) Expression of PMAPmRNA at 32 h under 17,20-P (0·1 µg/ml, S) or ethanolvehicle (1 µl/ml, C) in the ovaries from five fish (F1-F5).The 1·1 kb transcript was upregulated by 17,20-P in allfive fish assayed (P=0·012; pair-wise t-test). (B)Representative Northern blot of the temporal expressionpattern of PMAP ovarian transcripts in one fish.Numbers above the lanes represent hours of incubationin Cortland medium with 17,20-P (0·1 µg/ml, 17,20P) orethanol vehicle (1 µl/ml, Control). One other fish assayedshowed a similar pattern of expression. (C) PMAPtranscripts in other perch tissues. Abbreviations as inFigs 3 and 6.



In conclusion, several progestin-upregulatedcDNAs have been isolated from the yellow perchovary just prior to ovulation. These include perchforms of neprilysin, a lysyl oxidase type protein,calmodulin and a microtubule aggregate protein. Toour knowledge, this is the first report that specifi-cally describes progestin-upregulated mRNAs inthe vertebrate ovary at ovulation. While calmodulinhas already been strongly implicated in the processof ovulation, neprilysin and microtubule aggregateprotein have never been associated with ovulation orthe vertebrate ovary. Finally, the characteristics ofthe perch lysyl oxidase expression suggest that itmay be a tissue-specific or inducible form of lysyloxidase and, therefore, different from other lysyloxidases so far characterized.

ACKNOWLEDGEMENTS

This work was supported by NSF grant no. IBN9408012. The authors gratefully acknowledge thetechnical support of Ms Priscilla Duman.

REFERENCES

Ackland J, Schwartz NB, Mayo KE & Dodson RE 1992Nonsteroidal signals originating in the gonads. PhysiologicalReviews 72 731–787.

Alila H, Davis J, Dowd J, Corradino R & Hansel W 1990Differential effects of calcium on progesterone production insmall and large bovine luteal cells. Journal of SteroidBiochemistry 36 687–694.

Bornslaeger E, Wilde M & Schultz R 1984 Regulation ofmouse oocyte maturation: involvement of cyclic AMPphosphodiesterase and calmodulin. Developmental Biology105 488–499.

Bradley J & Goetz FW 1994 The inhibitory effects ofindomethacin, nordihydroguaiaretic acid and pyrrolidine-dithiiocarbamate on ovulation and prostaglandin synthesis inyellow perch (Perca flavescens) follicle incubates.Prostaglandins 4 11–20.

Brannstrom M & Janson PO 1989 Progesterone is a mediator inthe ovulatory process of the in vitro-perfused rat ovary.Biology of Reproduction 40 1170–1178.

Brannstrom M, Mayrhofer G & Robertson SA 1993Localization of leukocyte subsets in the rat ovary during theperiovulatory period. Biology of Reproduction 48 277–286.

Casey M & MacDonald C 1996 The endothelin-parathyroidhormone-related protein vasoactive peptide system in humanendometrium: modulation by transforming growth factor-â.Human Reproduction 11 (Suppl 2) 62–82.

Casey M, Smith J, Nagai K, Hersh L & MacDonald P 1991Progesterone-regulated cyclic modulation of membranemetalloendopeptidase (enkephalinase) in humanendometrium. Journal of Biological Chemistry 26623041–23047.

Celio M, Pauls T & Schwaller B 1996 Introduction to EF-handcalcium-binding proteins. In Guidebook To The Calcium-Binding Proteins, pp 15–20. Eds M Celio, T Pauls &B Schwaller. New York: Oxford University Press.

Chen C, Salles G, Seldin M, Kister A, Reinherz E & Shipp M1992 Murine common acute lymphoblastic leukemia antigen(CD10 neutral endopeptidase 24·11). Molecularcharacterization, chromosomal localization, and modeling ofthe active site. Journal of Immunology 148 2817–2825.

Chien Y & Dawid I 1984 Isolation and characterization ofcalmodulin genes from Xenopus laevis. Molecular andCellular Biology 4 507–513.

Chomczynski P 1993 A reagent for the single-step simultaneousisolation of RNA, DNA, and proteins from cell and tissuesamples. Biotechniques 15 532–537.

Chomczynski P & Sacchi N 1987 Single step method of RNAisolation by acid guanidinium thiocyanate-phenol-chloroformextraction. Analytical Biochemistry 162 156–159.

Dion N, Cohen P, Crine P & Boileau G 1997 Characterizationof neprilysin (EC 3·4·24·11) S2* subsite. FEBS Letters 411140–144.

Donath J, Michna H & Nishino Y 1997 The antiovulatoryeffect of the antiprogestin onapristone could be related todown-regulation of intraovarian progesterone (receptors).Journal of Steroid Biochemistry and Molecular Biology 62107–118.

Erdos E & Skidgel R 1989 Neutral endopeptidase 24·11(Enkephalinase) and related regulators of peptide hormones.FASEB Journal 3 145–151.

Espey LL 1980 Ovulation as an inflammatory reaction – ahypothesis. Biology of Reproduction 22 73–106.

Fischer R, Koller M, Flura M, Mathews S, Strehler-Page M,Krebs J, Penniston J, Carafoli E & Strehler E 1988Multiple divergent mRNAs code for a single humancalmodulin. Journal of Biological Chemistry 26317055–17062.

Franklin TJ 1997 Therapeutic approaches to organ fibrosis.International Journal of Biochemistry and Cell Biology 2979–89.

Furudate S, Masaki Y & Muto T 1994 Induction of ovulationby endothelin-3 in rats. Journal of Endocrinology 141109–112.

Goetz F & Theofan G 1979 In vitro stimulation of germinalvesicle breakdown and ovulation of yellow perch (Percaflavescens) oocytes. Effects of 17á-hydroxy-20â-dihydroprogesterone and prostaglandins. General andComparative Endocrinology 37 273–285.

Goetz F, Duman P, Berndtson A & Janowsky E 1989 The roleof prostaglandins in the control of ovulation in yellow perch(Perca flavescens). Fish Physiology and Biochemistry 7163–168.

Goetz FW, Berndtson AK & Ranjan M 1991 Ovulation:mediators at the ovarian level. In Vertebrate Endocrinology:Fundamentals and Biomedical Implications, vol 4, pp 127–203.Eds KT Pang & MP Schreibman, New York: AcademicPress, Inc.

Green R, Lieb M, Weintraub A, Gacheru S, Rosenfield C,Shah S, Kagan H & Taubman M 1995 Identification of lysyloxidase and other platelet-derived growth factor-induciblegenes in vascular smooth muscle cells by differentialscreening. Laboratory Investigation 73 476–482.

Hamalainen ER, Jones TA, Sheer D, Taskinen K, PihlajaniemiT & Kivirikko KI 1991 Molecular cloning of human lysyloxidase and assignment of the gene to chromosome5q23·3-31·2. Genomics 11 508–516.

Hedin L & Eriksson A 1997 Prostaglandin synthesis issuppressed by progesterone in rat preovulatory follicles invitro. Prostaglandins 53 91–106.

Hellberg P, Larson L, Olofsson J, Hedin L & Brennstrum M1991 Stimulatory effects of bradykinin on the ovulatoryprocess in the in vitro-perfused rat ovary. Biology ofReproduction 44 269–274.



Himeno N 1986 Effect of prostaglandins on collagen synthesisin rabbit ovarian follicles during the ovulatory process. FoliaEndocrinology 62 1181–1193.

Holmes PV, Sogn J, Schillinger E & Janson P 1985 Effects ofhigh and low preovulatory concentrations of progesterone onovulation from the isolated perfused rabbit ovary. Journal ofReproduction and Fertility 75 393–399.

Honda Y, Kondo J, Maeda T, Yoshiyama Y, Yamada E,Shimizu Y, Shikata T & Ono Y 1990 Isolation andpurification of a non-A, non-B hepatitis-associatedmicrotubular aggregate protein. Journal of General Virology71 1999–2004.

Hooper N 1994 Families of zinc metalloproteases. FEBSLetters 354 1–6.

Hsu T & Goetz FW 1991 Synergistic induction of ovulationand prostaglandin synthesis in goldfish (Carassius auratus)follicles by sodium orthovanadate and hydrogen peroxide.Biology of Reproduction 45 912–917.

Johnson A & Tilly J 1990 Evidence that arachidonic acidinfluences hen granulosa cell steroidogenesis andplasminogen activator activity by a protein kinaseC-independent mechanism. Biology of Reproduction 43922–928.

Kenyon K, Modi WS, Contente S & Friedman RM 1993 Anovel human cDNA with a predicted protein similar to lysyloxidase maps to chromosome 15q24-q25. Journal ofBiological Chemistry 268 18435–18437.

Kim Y, Boyd C & Csiszar K 1995 A new gene with sequenceand structural similarity to the gene encoding human lysyloxidase. Journal of Biological Chemistry 270 7176–7182.

Kitamura A, Takahashi K, Okajima A & Kitamura N 1994Induction of the human gene for p44, a hepatitis-C-associated microtubular aggregate protein, by interferon-á/â.European Journal of Biochemistry 224 877–883.

Kozak M 1986 Point mutations define a sequence flanking theaug initiator codon that modulates translation by eukaryoticribosomes. Cell 44 283–292.

van der Kraak G 1991 Role of calcium in the control ofsteroidogenesis in preovulatory ovarian follicles of thegoldfish. General and Comparative Endocrinology 81268–275.

Kugu K, Dharmarajan A, Preutthipan S & Wallach E 1995Role of calcium/calmodulin-dependent protein kinase II ingonadotropin-induced ovulation in in vitro perfused rabbitovaries. Journal of Reproduction and Fertility 103 273–278.

Lazarus H, Cruikshank W, Narasimhan N, Kagan H & CenterD 1994 Induction of human monocyte mobility by lysyloxidase. Matrix Biology 14 727–773.

Lipner H & Greep R 1971 Inhibition of steroidogenesis atvarious sites in the biosynthetic pathway in relation toinduced ovulation. Endocrinology 88 602–607.

Lipner H & Wendelken L 1971 Inhibition of ovulation byinhibition of steroidogenesis in immature rats. Proceedings ofthe Society of Experimental Biology and Medicine 1361141–1145.

Lydon JP, DeMayo FJ, Funk CR, Mani SK, Hughes AR,Montgomery CA, Shyamala G, Conneely OM & O’MalleyBW 1995 Mice lacking progesterone receptor exhibitpleiotropic reproductive abnormalities. Genes andDevelopment 9 2266–2278.

Matso K, Sato K, Ikeshima H, Shimoda K & Takano T 1992Four synonymous genes encode calmodulin in the teleostfish, medaka (Oryzias latipes): conservation of the multigeneone-protein principle. Gene 119 279–281.

Morgan A, Keeble SC, London SN, Muse KN & Curry TE1994 Antiprogesterone (RU486) effects on metalloproteinaseinhibitor activity in human and rat granulosa cells. Fertilityand Sterility 61 949–955.

Mori T, Suzuki A, Nishimura T & Kambegawa K 1977Inhibition of ovulation in immature rats by anti-progesteroneantiserum. Endocrinology 73 185–186.

Murdoch WJ 1998 Perturbation of sheep ovarian surfaceepithelial cells by ovulation: evidence for roles ofprogesterone and poly (ADP-ribose) polymerase in therestoration of DNA integrity. Journal of Endocrinology 156503–508.

Murdoch WJ & Farris ML 1988 Prostaglandin E2-9-ketoreductase activity of preovulatory ovine follicles. Journalof Animal Sciences 66 2924–2929.

Murdoch WJ & McCormick RJ 1989 Production of lowmolecular weight chemoattractants for leukocytes byperiovulatory ovine follicles. Biology of Reproduction 4086–90.

Murdoch WJ, Petterson TA, Van Kirk EA, Vincent DL &Inskeep EK 1986 Interactive roles of progesterone,prostaglandins, and collagenase in the ovulatory mechanismof the ewe. Biology of Reproduction 35 1187–1194.

Painter R, Dukes R, Sullivan J, Carter R, Erdos E & Johnson1988 A function of neutral endopeptidase on the cellmembrane of human neutrophils. Journal of BiologicalChemistry 263 9456–9461.

Putkey J, Ts’ui K, Tanaka T, Lagace L, Stein J, Lai E &Means A 1983 Chicken calmodulin genes. A speciescomparison of cDNA sequences and isolation of a genomicclone. Journal of Biological Chemistry 258 11864–11870.

Saito H, Papaconstantinou J, Sato H & Goldstein S 1997Regulation of a novel gene encoding a lysyl oxidase-relatedprotein in cellular adhesion and senescence. Journal ofBiological Chemistry 272 8157–8160.

Sato E 1990 Suppression of spontaneous maturation of isolatedcumulus free mouse oocytes by a calmodulin antagonist.Endocrinologia Japonica 37 19–26.

Shipp MA, Richardson NE, Sayre PH, Brown NR, MastellerEL, Clayton LK, Ritz J & Reinherz EL 1988 Molecularcloning of the common acute lymphoblastic leukemia antigen(CALLA) identifies a type II integral membrane protein.Proceedings of the National Academy of Sciences of the USA85 4819–4823.

Smith-Mungo L & Kagan H 1998 Lysyl oxidase: properties,regulation and multiple functions in biology. Matrix Biology16 387–398.

Takahashi M, Ford JJ, Yoshinaga K & Greep RO 1974Induction of ovulation in hypophysectomized rats byprogesterone. Endocrinology 95 1322–1326.

Takahashi K, Kitamura N, Shibui T, Kamizono M, Matsui R,Yoshiyama Y, Maeda T, Kondo J, Honda Y, Yamada E,Shimiau Y, Teranishi Y & Nakanishi S 1990 Cloning,sequencing and expression in Esherichia coli of cDNA for anon-A, non-B hepatitis-associated microtubular aggregatesprotein. Journal of General Virology 71 2005–2011.

Tanaka K, Li ZD & Ataka Y 1987 Studies of ovulation in theperfused ovary of the fowl (Gallus domesticus). Journal ofReproduction and Fertility 80 411–416.

Tanaka N, Espey LL, Stacy S & Okamura H 1992 Epostaneand indomethacin actions on ovarian kallikrein andplasminogen activator activities during ovulation in thegonadotropin-primed immature rat. Biology of Reproduction46 665–670.

Tedeschi C, Hazum E, Kokia E, Ricciarelli E, Adashi E &Payne D 1992 Endothelin-1 as a luteinization inhibitor:inhibition of rat granulosa cell progesterone accumulation viaselective modulation of key steroidogenic steps affecting bothprogesterone formation and degradation. Endocrinology 1312476–2478.

Theofan G & Goetz FW 1981 The in vitro effects ofactinomycin D and cycloheximide on germinal vesicle



breakdown and ovulation of yellow perch (Perca flavescens)oocytes. Comparative Biochemistry and Physiology 69A557–561.

Tsafriri A, Abisogun AO & Reich R 1987 Steroids andfollicular rupture at ovulation. Journal of SteroidBiochemistry and Molecular Biology 27 359–363.

Wasserman W & Smith D 1981 Calmodulin triggers theresumption of meiosis in amphibian oocytes. Journal of CellBiology 89 389–394.

Wolf K & Quimby M 1969 Fish cell and tissue culture. In FishPhysiology, vol 3, pp 253–305. Eds W Hoar & D Randall.New York: Academic Press, Inc.

Yoshimura Y, Hosoi Y, Atlas SJ, Bongiovanni AM, Santulli R& Wallach EE 1986 The effects of ovarian steroidogenesis on

ovulation and fertilizability in the in vitro perfused rabbitovary. Biology of Reproduction 35 943–948.

Yoshimura Y, Hosoi Y, Bongiovanni AM, Santulli R, Atlas SJ& Wallach EE 1987 Are ovarian steroids required for ovummaturation and fertilization? Effects of cyanoketone on thein vitro perfused rabbit ovary. Endocrinology 1202555–2561.

Yoshimura Y, Okamoto T & Tamura T 1990 Immuno-histochemical observations on the localization ofcalmodulin in the chicken ovary. Japanese PoultryScience 27 356–362.

Zar J 1996 Biostatistical Analysis. New Jersey: Prentice Hall.

28 December 1998



The upregulation of messenger ribonucleic acids during 17α,20β ...

Documents

Transcript of The upregulation of messenger ribonucleic acids during 17α,20β ...