Decidual/Trophoblast Prolactin-Related Protein:Characterization of Gene Structure and Cell-SpecificExpression*

KYLE E. ORWIG†, GUOLI DAI, CHRISTINE A. RASMUSSEN‡,AND MICHAEL J. SOARES

Department of Molecular and Integrative Physiology, University of Kansas Medical Center, KansasCity, Kansas 66160

ABSTRACTDecidual/trophoblast PRL-related protein (d/tPRP) is a member of

the PRL gene family and is dually expressed in uterine and placentaltissues in a highly coordinated pattern during pregnancy. In thepresent study, we describe the isolation and characterization of thed/tPRP gene. A l DASH II Wistar-Kyoto rat genomic library wasscreened with a labeled d/tPRP complementary DNA, resulting in theisolation of two phage clones, RGLd-41 [17.7 kilobases (kb)] andRGLd-42 (15.8 kb). RGLd-41 alone was found to contain the full-length d/tPRP gene and was used for subsequent analyses. Thed/tPRP gene possesses a six-exon, five-intron organization. Relativeto other highly conserved members of the PRL gene family, d/tPRPcontains a single small additional exon (exon 3) situated betweenexons 2 and 3 of the prototypical PRL gene. The region correspondingto exon 3 of d/tPRP encodes for a unique amino acid region found ina subset of PRL family members. A reverse transcription-PCR (RT-

PCR) tissue survey for d/tPRP messenger RNA revealed that d/tPRPexpression was restricted to decidual and trophoblast tissues. A singletranscription start site 65 bp upstream of the initiation codon wasidentified in decidual tissue, whereas multiple transcription startsites ranging from 61–66 bp upstream of the initiation codon weredetected in placental tissue. Various tissue culture systems (primarycultures and cell lines) were evaluated for d/tPRP expression andactivation of a 3.96-kb d/tPRP promoter-luciferase reporter construct.Decidual, spongiotrophoblast, and trophoblast giant cell populationsexpressed d/tPRP and were capable of activating the d/tPRP promot-er-reporter construct, whereas other cell types were ineffective. Lim-ited d/tPRP promoter activation was noted in uterine stromal celllines. In summary, d/tPRP possesses a unique six-exon, five-introngene structure and exhibits cell-specific expression that is regulatedat least in part by a 3.96-kb 59-flanking region. (Endocrinology 138:2491–2500, 1997)

THE ESTABLISHMENT and maintenance of pregnancyin the female reproductive tract requires extensive re-

modeling of the uterine endometrium and development ofthe placenta. In some species, including the rat and human,uterine stromal cells differentiate into a specialized structure,referred to as the decidua (1–3). Together, the decidua andplacenta provide the conduit through which the fetus gainsaccess to nutrients and eliminates wastes. Appropriate de-velopment of these tissues is essential for reproductivesuccess.

The functions of the decidua and placenta are probablymediated in part by their production of hormones and cy-tokines. In the rat, at least nine different genes, structurallyrelated to PRL, are expressed by uterine decidual cellsand/or trophoblast cells of the chorioallantoic placenta in ahighly coordinated pattern during pregnancy (4). Decidual/trophoblast PRL-related protein (d/tPRP) is an example of a

subset of PRL family members that is dually expressed inboth decidual and trophoblast cell types (5–7). Some mem-bers of the PRL family use the PRL receptor signaling systemand mimic the actions of PRL, whereas other members of thefamily activate apparently novel signaling pathways andpossess apparently novel biological activities (4). d/tPRPfalls into this latter category. It does not activate PRL receptorsignaling pathways; however, d/tPRP does interact withheparin-containing molecules, and it increases the tumori-genicity of Chinese hamster ovary cells (8). The mechanismsthrough which the expression patterns and actions of d/tPRPfacilitate the establishment and maintenance of pregnancyare yet to be resolved.

In this report, we present information on the isolation andcharacterization of the d/tPRP gene. We show that d/tPRPpossesses a unique six-exon, five-intron gene structure andthat regulatory DNA associated with the d/tPRP gene di-rects decidua- and trophoblast-specific expression.

Materials and MethodsReagents

FBS and donor horse serum were purchased from JRH Bioscience(Lenexa, KS). Reagents for PAGE were purchased from Bio-Rad (Her-cules, CA). The chemiluminescent detection system was obtained fromAmersham Life Science (Arlington Heights, IL). The streptavidin-biotinimmunoperoxidase kit and diaminobenzidine were obtained from Vec-tor Laboratories (Burlingame, CA). Dispase was purchased from Boehr-inger Mannheim (Indianapolis, IN). All restriction enzymes, poly-merases, and DNA ligases were purchased from New England Biolabs

Received November 19, 1996.Address all correspondence and requests for reprints to: Dr. Michael

Soares, Department of Physiology, University of Kansas Medical Center,Kansas City, Kansas 66160. E-mail: msoares@kumc.edu.

* This work was supported by grants from the NICHHD (HD-29036,HD-29797, and HD-33994). Sequences reported in this manuscript havebeen deposited in the GenBank database (accession no. U44438 andL06441).

† Supported by fellowships from the Lalor and Kansas HealthFoundations.

‡ Present address: Department of Anatomy and Cell Biology, Uni-versity of Kansas Medical Center, Kansas City, Kansas 66160.

0013-7227/97/$03.00/0 Vol. 138, No. 6Endocrinology Printed in U.S.A.Copyright © 1997 by The Endocrine Society

2491

(Beverly, MA). The GH3 pituitary tumor and L929 cell lines and a Roussarcoma virus promoter-b-galactosidase (RSV-bGAL) reporter plasmidwere obtained from American Type Culture Collection (Rockville, MD).Transformation-competent Sure bacterial cells, pBluescript SK1, theFlash Nonradioactive Gene Mapping kit, and a rat genomic library wereacquired from Stratagene (La Jolla, CA). Oligonucleotide probes weresynthesized by the University of Kansas Medical Center BiotechnologySupport Facility (Kansas City, KS). DNA extraction kits were purchasedfrom Qiagen (Chatsworth, CA). Nitrocellulose and nylon membraneswere obtained from Schleicher and Schuell (Keene, NH). The pGL-2basic vector and a RSV promoter-luciferase reporter plasmid were pur-chased from Promega (Madison, WI). T7 DNA sequencing kits wereacquired from U.S. Biochemical (Cleveland, LH). The AdvantageGenomic PCR kit was obtained form Clontech (Palo Alto, CA). Radio-labeled nucleotides were purchased from DuPont-New England Nu-clear (Boston, MA). TRIzol reagent for RNA extraction, Superscriptpreamplification kits, Taq polymerase, and Lipofectamine reagent fortransfection were obtained from Life Technologies (Gaithersburg, MD).Kits for monitoring bGAL activities were acquired from Tropix (Bed-ford, MA). Unless otherwise noted, all other chemicals and reagentswere purchased from Sigma Chemical Co. (St. Louis, MO).

Animals and tissue preparation

Holtzman rats were obtained from Harlan Sprague-Dawley (India-napolis, IN). The animals were housed in an environmentally controlledfacility, with lights on from 0600–2000 h, and allowed free access to foodand water. Timed pregnancies and tissue dissections were performed aspreviously described (9). Day 0 of pregnancy was defined by the pres-ence of a copulatory plug or sperm in the vaginal smear. Protocols forthe care and use of animals were approved by the University of Kansasanimal care and use committee.

Cell cultures

A series of primary cell cultures and cell lines was evaluated for theirability to express d/tPRP and activate d/tPRP promoter/luciferase re-porter constructs. Primary decidual cell cultures were established fromdeciduomal tissue collected from rats on day 7 of pseudopregnancy.Deciduomal tissue was minced, washed three times in Hanks’ BalancedSalt Solution, and dispersed in dispase (2.4 U/ml) containing deoxyri-bonuclease I (80 U/ml) for 1.5 h at 37 C. Dispersed cells were recoveredby centrifugation and washed with Hanks’ Balanced Salt Solution toremove residual dispase. Cells were then resuspended in DMEM-MCDB302 culture medium containing 10% FBS and plated at a concentrationequivalent to one uterine horn (7–9 3 105 cells)/25-cm2 flask. After 20 h,medium and unattached cells were removed and replaced with freshmedium containing 1% FBS. Primary spongiotrophoblast cultures wereestablished according to previously published procedures (10) andmaintained in DMEM culture medium supplemented with 10% FBS. TheUI uterine stromal cell line was established (11) essentially as describedby Cohen et al. (12). CUS V2 and CUS V4 uterine stromal cells areimmortalized cells derived from rat uterine stroma by transfecting pri-mary cultures with a temperature-sensitive mutant of the simian virus40 large T antigen (13). All uterine stromal cell lines were maintained inHam’s F-10-DMEM culture medium supplemented with 10% FBS. TheRcho-1 trophoblast cell line was derived from a rat choriocarcinoma andis capable of differentiating along the trophoblast giant cell lineage (14).Rcho-1 trophoblast cells were routinely maintained in subconfluentconditions with NCTC-135 culture medium supplemented with 20%FBS, 50 mm 2-mercaptoethanol, and 1 mm sodium pyruvate (14). Rcho-1cells were induced to differentiate by growing them to near confluencein FBS-supplemented culture medium and then replacing the FBS with10% horse serum (15, 16). The HRP-1 trophoendodermal stem cell line,which exhibits both trophoblast and yolk sac attributes (17, 18), wasmaintained in RPMI 1640 medium containing 10% FBS. GH3 pituitarytumor cells (19) were maintained in DMEM culture medium supple-mented with 10% FBS, and L929 mouse fibroblast cells were maintainedin RPMI 1640 medium containing 10% FBS. All culture media weresupplemented with 100 U/ml penicillin and 100 mg/ml streptomycin.

Isolation and characterization of the d/tPRP gene

A genomic DNA library generated from liver tissue of 12-week-oldmale Wistar-Kyoto outbred rats and packaged in the l DASH II vectorwas obtained from Stratagene. The library was screened with a ratd/tPRP complementary DNA (cDNA) as previously described (5). lDNA from positive plaques was amplified and used to inoculate LE 392Escherichia coli. Phage DNA was extracted from lysates and characterizedby restriction mapping and Southern blot hybridization (20).

Two d/tPRP genomic clones were identified, RGLd-41 and RGLd-42.Oligonucleotides representing either the 59- or 39-end of d/tPRP cDNAwere end labeled with T4 polynucleotide kinase and [g-32P]ATP andused to identify clones containing the full-length d/tPRP gene. GenomicDNA was excised with NotI, and a restriction map was generated usingpartial BamHI digestion (Flash Nonradioactive Gene Mapping kit). Frag-ments containing d/tPRP exonic DNA (determined by Southern blothybridization) were subcloned into the BamHI and/or NotI sites ofpBluescript SK1, flanked by T7 and T3 promoters.

Exon-containing restriction fragments in pBluescript SK1 were se-quenced by the dideoxy chain termination method (21) using Sequenaseand [35S]deoxy-ATP. Primers corresponding to the T7 or T3 flankingsequence or internal oligonucleotide primers were used to sequence allexons and exon/intron boundaries. Reaction products were resolved in6% polyacrylamide urea gels, dried, and exposed to Kodak X-OmatX-ray film. Comparison with the published d/tPRP cDNA sequenceconfirmed the identity of the d/tPRP genomic clone (5).

Identification of the d/tPRP transcription start sites

The transcription start site of d/tPRP was determined by primerextension analysis, essentially as described by Duckworth et al. (22). Anoligonucleotide complementary to bases 131 to 110 of the d/tPRPcDNA (from ATG) was synthesized and end labeled using T4 polynu-cleotide kinase and [g-32P]ATP. The labeled primer (final concentration,10 mm) was extracted with phenol-chloroform, precipitated with etha-nol, and hybridized with 5 mg total RNA from decidua, junctional zoneplacenta, or spleen. Reverse transcription was performed using a Su-perscript Preamplification kit. The reaction was stopped by the additionof sample running buffer (98% formamide, 2.5 mm EDTA, 0.1% bro-mophenol blue, and 0.1% xylene cyanol) and electrophoretically sepa-rated on a 6% polyacrylamide-7 m urea sequencing gel. A known DNAsequence was separated on the same gel to indicate the size of theprimer-extended product.

Analysis of d/tPRP expression

Western blot analyses for d/tPRP were performed as previouslydescribed (8). Samples were separated by 12.5% PAGE under reducingconditions and transferred to nitrocellulose membranes. Immunoreac-tive bands were visualized using a chemiluminescent detection system(Arlington Heights, IL).

Tissue and cellular localization of d/tPRP was confirmed by immu-nocytochemistry using a streptavidin-biotin immunoperoxidase kit forrabbit IgG and the chromagen, diaminobenzidine (7). The immuno-stained sections were counterstained with hematoxylin. The specificityof the immunoreactions was demonstrated using preimmune serum andpreadsorbed antibodies.

Northern blots were performed as previously described (5, 23). TotalRNA was extracted from tissues and cells essentially as described byChomczynski and Sacchi (24), using TRIzol. Blots were probed with32P-labeled d/tPRP cDNA (5).

RT was performed using 0.5 mg oligo(deoxythymidine) primers and5 mg total RNA. The resulting cDNAs were amplified by PCR for 35cycles with a denaturing temperature of 94 C (1 min), an annealingtemperature of 60 C (2 min), and an extension temperature of 72 C (2min), using a Perkin-Elmer Thermocycler (model 480, Norwalk, CT).Oligonucleotide primers specifically amplified a 342-bp region of thed/tPRP cDNA: upstream primer, 59-CATGGACCTGAACATGAAAA-CATCAAA-39 (sense, 325–354; located on exon 4); and downstreamprimer, 39-GTGACGGATGCACAACTATATAAGATG-59 (antisense,637–666; located on exon 6). The PCR reaction mixture also containedprimers that amplified a 244-bp region of rat b-actin (25). b-Actin bandswere readily detectable on ethidium bromide-stained gels and used to

2492 DECIDUAL/TROPHOBLAST PRL-RELATED PROTEIN Endo • 1997Vol 138 • No 6

demonstrate equal loading and integrity of the messenger RNA (mRNA)template. Reaction products were fractionated in agarose gels and trans-ferred to nitrocellulose. Southern blots were performed using 32P-la-beled d/tPRP cDNA and visualized on Kodak X-Omat X-ray film (East-man Kodak, Rochester, NY).

d/tPRP promoter analysis

Cell-specific d/tPRP promoter activation was evaluated in variousprimary cell cultures and cell lines. PCR was used to amplify 3960 bpof d/tPRP 59-flanking DNA from the 6.9-kilobase (kb) restriction frag-ment of the d/tPRP genomic clone. The 3960-bp amplified fragmentextended to the cytosine residue located 38 bp downstream of the tran-scription start site and 28 bp upstream of the translation start site (ATG;see Fig. 2). The Advantage Genomic PCR kit was used to amplify apromoter fragment using the high fidelity Tth DNA polymerase. PCRprimers were designed so the amplified fragment contained a KpnIrestriction site at the 59-end and an XhoI site at the 39-end. This alloweddirectional cloning of the d/tPRP 59-flanking DNA into the KpnI andXhoI sites of the pGL-2 basic, luciferase reporter vector. We will refer tothis construct as d/tPRP-luc.

The d/tPRP-luc construct was transiently transfected into the pri-mary cell culture systems and cell lines using a liposome-mediateddelivery system. Cells were plated in 35-mm culture dishes (3 3 105) andtransfected with 2 mg d/tPRP-luc, RSV-luc (positive control) or pGL-2basic vector (negative control). RSV-bGAL (0.5 mg) was cotransfectedwith all constructs and used to correct for transfection efficiency. Pri-mary decidual and primary spongiotrophoblast cells were transfectedon the second day of culture; Rcho-1 trophoblast cells were transfectedon day 3 of culture, corresponding to the time that cells were exposedto differentiating conditions. Forty-eight hours after transfection, cellswere collected, and lysates were prepared via three consecutive cyclesof freezing and thawing. Luciferase activity was determined using aluminometer according to the procedure described by Brasier et al. (26).bGAL activity was evaluated using the Galacto-Light kit, and proteinconcentrations were determined using the protein-dye binding methoddescribed by Bradford (27). Luciferase activities were normalized fortransfection efficiency and protein concentration.

Rcho-1 trophoblast cells were also cotransfected with the d/tPRP-lucconstruct (3 mg/ml; 10-fold excess) and pSV2 neo (0.3 mg/ml; a plasmidproviding neomycin resistance). Cells were selected for 2 weeks in thepresence of geneticin (G418; 250 mg/ml) as we previously described (28).Stably transfected clones were isolated by limiting dilution. Cellularlysates were collected from stably transfected clones at various timesduring the transition from proliferation to differentiation states andevaluated for luciferase activity as described above.

ResultsIsolation and characterization of the d/tPRP genomicclones

Screening of the l DASH II rat genomic library with la-beled d/tPRP cDNA resulted in the isolation of two phageclones, RGLd-41 (17.7 kb) and RGLd-42 (15.8 kb). Dot blotanalysis revealed that RGLd-41 alone contained the full-length d/tPRP gene. A restriction map was generated basedon a partial digestion with BamHI (Fig. 1). Complete NotI/BamHI digestion resulted in the generation of five fragments,ranging from 2.1–6.9 kb in size (Fig. 1). Southern blots wereperformed on the resulting digestion products using randomprimer-labeled d/tPRP cDNA as a probe. Three exon-con-taining fragments were identified (open boxes in Fig. 1) andsubcloned into pBluescript SK1 for sequencing.

The complete d/tPRP cDNA sequence was identified inthe RGLd-41 clone, confirming that it was the full-lengthd/tPRP genomic DNA. All exons and exon/intron bound-aries of the d/tPRP gene were sequenced, and all intronsexhibited the consensus GT and AG splicing junctions. The

exon/intron organization of the d/tPRP gene was deter-mined by aligning d/tPRP cDNA sequence with the corre-sponding regions of RGLd-41. Sequence alignment revealedthat d/tPRP contains six exons and five introns (Figs. 1 and2). Relative to the other members of the PRL family, d/tPRPcontains one additional exon, exon 3 (Fig. 3). A homologousexon 3 is also present in PRL-like protein-C variant (PLP-Cv)(29). Exon 3 encodes a region that is also homologous withunique regions of mouse proliferin-related protein (mPRP)(30), PLP-C (31), and PLP-D (32). However, the genomicstructures of these PRL family members have not been char-acterized. This exon is part of a larger 14-amino acid regionthat has been termed the aromatic domain because it containsthree phenylalanines and one tyrosine (see Fig. 2).

Analysis of tissue- and cell-specific d/tPRP expression

We evaluated d/tPRP expression in a variety of tissues todetermine whether it was expressed in a tissue-specific man-ner. RT-PCR analysis demonstrated that d/tPRP was ex-pressed by the decidua and junctional zone of the cho-rioallantoic placenta (Fig. 4). A weaker d/tPRP signal wasalso observed in the labyrinth zone, but probably resultedfrom junctional zone contamination. In support of this con-tention, d/tPRP expression was not detectable in labyrin-thine trophoblast by in situ or immunocytochemical analyses(7).

A primary decidual cell culture system was establishedand evaluated for its ability to express d/tPRP. Northern blotanalysis revealed that d/tPRP mRNA increased until day 3of culture and was maintained until day 5 (Fig. 5, upper panel).Immunoreactive d/tPRP increased from days 1–3 of culture,as demonstrated by Western blot (Fig. 5, lower panel) and

FIG. 1. Schematic diagrams of the 17.7-kb d/tPRP genomic clone. A,Restriction mapping and Southern blot analyses demonstrated thatthree fragments contained d/tPRP exonic DNA (open boxes). B, Se-quence analysis of the d/tPRP genomic clone and comparison with thecDNA demonstrated that the d/tPRP gene is composed of six exonsand five introns. The beginning of exon 1 is defined as the transcrip-tion start site (see Fig. 9), and the 39-end of the gene is defined as thepolyadenylation signal (AATAAA). C, The 6.9-kb NotI/BamHI restric-tion fragment of the d/tPRP genomic clone contains 3960 bp of 59-flanking DNA, exon 1, and half of exon 2. Diagonal lines representregions of intron or flanking DNA that are not drawn to scale.

DECIDUAL/TROPHOBLAST PRL-RELATED PROTEIN 2493

immunocytochemical analyses (Fig. 6). In vivo, d/tPRP im-munoreactivity was localized predominantly to the antime-sometrial region of the decidua (Fig. 6d).

RT-PCR was used to identify other cell culture models thatexpress d/tPRP (Fig. 7). In addition to the primary decidualcell culture, d/tPRP expression was monitored in primaryspongiotrophoblast culture, the Rcho-1 trophoblast cell line,several uterine stromal cell lines (CUS V2, CUS V4, and UI),

L929 mouse fibroblasts, HRP-1 trophoendodermal cells, andGH3 pituitary tumor cells. Only the primary decidual cells,primary spongiotrophoblast cells, and differentiated (day 9)Rcho-1 trophoblast cells were capable of expressing detect-able levels of d/tPRP mRNA. Collectively, these results in-dicate that d/tPRP is expressed in both a cell-specific anddifferentiation-dependent manner.

Analysis of the d/tPRP transcription start site

Primer extension analysis was used to identify the d/tPRPtranscription start site in decidual tissue and junctional zoneplacenta. Decidual tissue contained one d/tPRP transcrip-tion start site located 65 bases upstream of the translationstart site (ATG; Fig. 8). In contrast, the placenta had severalapparent transcription start sites between 61–66 bases up-stream of the translation start site (Fig. 8). Spleen RNA servedas a negative control and had no primer-extended products.The sequence of the proximal promoter revealed a consensusTATA box (TATATAA) at positions 228 to 222 relative tothe transcription start site (Fig. 2, underlined and shaded).

Analysis of the d/tPRP promoter

The d/tPRP-luc construct was transiently transfected intoa variety of cell lines to determine whether the 59-flankingDNA associated with the d/tPRP gene was capable of di-recting tissue-specific expression of the luciferase reporter.The results depicted in Fig. 9 demonstrate that 3960 bp ofd/tPRP 59-flanking DNA promoted a 60-fold induction of theluciferase reporter relative to the empty pGL-2 vector inprimary decidual cells. Lower levels of d/tPRP promoteractivity were also detected in CUS V4 and UI uterine stromalcell lines (Fig. 9, upper panel). Similarly, the 3960-bp d/tPRPpromoter was able to drive reporter expression in Rcho-1trophoblast cells and the spongiotrophoblast primary culture(Fig. 9, lower panel; 700- and 400-fold induction over pGL-2,

FIG. 2. Nucleotide sequence of the d/tPRP structural gene. The lo-cations of six exons are indicated by bold capital letters; flanking andintronic regions are in lower case. The transcription start site islocated 65 bases upstream of the ATG (arrow, see Fig. 9). The TATAbox (located in the proximal promoter) and the polyadenylation signal(located at the 39-end of the gene) are underlined (black boxes). En-coded amino acids are indicated by one-letter designations beneaththe codons. The primers used for primer extension (PE) analysis andPCR are indicated by bold arrows. The site of signal peptide cleavageis indicated by the arrowhead between amino acids 21 and 11.

FIG. 3. Exon/intron organization of several members of the PRL/GHfamily. A five-exon, four-intron organization was well preserved dur-ing evolution and across species. Relative to most other members ofthe PRL/GH family, d/tPRP contains an additional exon (exon 3). Werecently demonstrated that rat PLP-Cv also contains an extra exonin a homologous region (29). sSL, Salmon somatolactin (60); cGH,chicken GH (61); hPRL, human PRL (62); mPL-II, mouse PL-II (36);rPRL, rat PRL (63, 64); mPLF, mouse proliferin-III (34).

2494 DECIDUAL/TROPHOBLAST PRL-RELATED PROTEIN Endo • 1997Vol 138 • No 6

respectively). No significant reporter activity was observedin L-929, CUS V2, GH3, or HRP-1 cells. The positive control,RSV-luc construct, was functional in all cell lines tested.

To evaluate whether the d/tPRP promoter also directsdifferentiation-dependent expression, d/tPRP-luc was sta-bly transfected into Rcho-1 trophoblast cells. Three stablytransfected clonal cell lines were established: d/tPRP-Rcho-3, d/tPRP-Rcho-4, and d/tPRP-Rcho-8. All threeclones exhibited a differentiation-dependent activation of thed/tPRP promoter (Fig. 10). d/tPRP-Rcho-3 luciferase activityincreased 95-fold between days 2–13 of culture.

Discussion

d/tPRP is one of a cluster of genes located on chromosome17 of the rat that are related to PRL. These genes are ex-pressed in a coordinated pattern by the anterior pituitary,decidua, and/or placenta during pregnancy. In this report,we have isolated and characterized the d/tPRP gene andprovided insight into the cell-specific nature of its expres-sion. The d/tPRP gene exhibits two interesting characteris-tics: 1) it possesses an atypical exon/intron organization; and2) it is dually expressed in both maternal and extraembryoniccell types.

Inclusion in the PRL gene family is largely based on struc-tural relationships among the members. Sequence analysis ofthe original d/tPRP cDNA placed it in the PRL family (5).Among the many striking similarities of members of the PRLgene family is their exon/intron organization. The PRL genehas retained a characteristic five-exon/four-intron configu-ration across all species where a structure has been reported(33). Data for other members of the PRL family, includingplacental lactogens (PLs), PRL-like proteins, and PRL-relatedproteins, are less abundant. Nonetheless, bovine PL, mousePL-II, and mouse proliferin contain a similar five-exon/four-

intron organization (34–36). d/tPRP exhibits a distinctivevariation from this common alignment and represents anexonic structure likely to be characteristic of a subset of PRLfamily members referred to as the PLP-C subfamily. ThePLP-C subfamily includes PLP-C, PLP-Cv, PLP-D, andd/tPRP (5, 29, 31, 32). Each constituent of this subfamilyshows considerable overall homology and contains a uniquesegment of amino acids that are rich in aromatic residues,which is not present in other members of the PRL family. Thegenomic structures of two members of the subfamily (PLP-Cv and d/tPRP) have now been reported (Ref. 29 and thepresent findings). Both possess a structure homologous tothose of other PRL family genes, except for an additionalsmall exon located between exons 2 and 3 of the prototypical

FIG. 4. Tissue-specific expression of d/tPRP mRNA. RT-PCR wasused to survey d/tPRP expression in a variety of tissues. d/tPRPexpression was monitored in day 10 decidual tissue, junctional zoneplacenta from d19 of pregnancy (JZ), day 19 labyrinth zone (LZ),nonpregnant (NP) uterus, ovary, liver, spleen, brain, and thymus.Upper panel, Southern blot analysis demonstrated strong d/tPRPexpression in decidual tissue and the junctional zone of the placenta,and lower levels of expression in the labyrinth zone. PCR primersspecifically amplified a 342-bp region of the d/tPRP cDNA. Lowerpanel, b-Actin was coamplified with d/tPRP, demonstrating equalloading and integrity of the mRNA template. The amplified b-actinproduct is 244 bp in length.

FIG. 5. Temporal expression of d/tPRP mRNA and protein productionby the primary decidual cell culture. Decidual cells were enzymati-cally isolated from day 7 decidual tissue. After overnight attachment,primary decidual cells were washed and placed in fresh DMEM-MCDB 302 medium containing 1% FBS. Upper panel, Total RNA wascollected from primary decidual cells at various times during culture.d/tPRP expression increased until day 3 of culture and was main-tained through day 5 of culture. Ribosomal RNA (28S) bands on theethidium bromide-stained gel are shown on the lower part of the upperpanel and demonstrate equal loading and the integrity of the RNA.Lower panel, Conditioned medium was collected from the primarydecidual cells on a daily basis and analyzed for d/tPRP content byWestern blot analysis. Primary decidual cells exhibit maximal d/tPRPproduction on day 3 of culture, coinciding with maximal expression ofd/tPRP mRNA. Recombinant d/tPRP (dPRP) was included as a pos-itive control. A signal was not detected in day 3 decidual cells whenthe antibodies were preadsorbed with d/tPRP (d3 1 dPRP), thusdemonstrating the specificity of the antibodies.

DECIDUAL/TROPHOBLAST PRL-RELATED PROTEIN 2495

PRL gene structure. This unique exon 3 encodes a region richin aromatic amino acids. We hypothesize that PLP-C andPLP-D also exhibit a six-exon/five-intron organization sim-ilar to those of PLP-Cv and d/tPRP. A member of the mousePRL family, mPRP, although showing limited overall ho-mology to the PLP-C subfamily, does contain the aromatic-rich domain characteristic of the PLP-C subfamily (30). Al-though the mPRP gene has not been cloned, we postulate thatits aromatic-rich region may also be encoded by an addi-tional exon. Previously, this similarity led us to speculate thatmembers of the PLP-C subfamily may possess biologicalactivities similar to those of proliferin-related protein, a po-tent antiangiogenic factor (8, 37). Initial experimentation hasnot supported a role of d/tPRP in the modulation of angio-genesis in the uteroplacental environment (8). Thus, inclu-sion of the aromatic domain in a protein structure does notappear to be solely responsible for instilling endothelial cell-targeted actions. The availability of the d/tPRP gene pro-vides us with a necessary prerequisite for the implementa-tion of additional molecular and transgenic strategiesdirected toward understanding the role of d/tPRP in thephysiology of pregnancy in the rat.

The d/tPRP gene is specifically expressed in maternal and

extraembryonic cells (Ref. 7 and the present study). Imme-diately, postimplantation d/tPRP expression is initiated indecidual cells located predominantly in the antimesometrialcompartment (7). The antimesometrial decidua regressesaround midgestation, and d/tPRP expression arises in tro-phoblast giant cells and in newly forming spongiotropho-blast cells of the junctional zone in the chorioallantoic pla-centa (7). Thus, there are three recognizable differentiatedcell types capable of expressing d/tPRP: 1) antimesometrialdecidual cells, 2) trophoblast giant cells, and 3) spongiotro-phoblast cells. The decidua-trophoblast shift in d/tPRP ex-pression resembles that reported for PLP-B (38–40). How-ever, in contrast to d/tPRP, PLP-B is expressed at relativelylow levels in decidua, and within the chorioallantoic placentaits expression is restricted to spongiotrophoblast cells(38–40).

Analysis of regulatory mechanisms controlling cell-spe-cific gene expression is facilitated by the availability of in vitrostrategies. In the present report, we demonstrated the use-fulness of primary culture systems for decidual and spon-giotrophoblast cells and the value of the Rcho-1 trophoblastcell line. The latter paradigm represents a valuable methodof dissecting regulatory mechanisms controlling trophoblast

FIG. 6. Immunocytochemical detection of d/tPRP in vitro and in vivo. Primary decidual cells were cultured as described in Fig. 5, and cells fromvarious stages of culture were fixed for analysis by immunocytochemistry. Primary decidual cells were stained for the presence of d/tPRP usingd/tPRP polyclonal antibodies and a streptavidin-biotin immunoperoxidase kit. d/tPRP immunoreactivity increased from day 1 (A) to day 3 (B)of culture. Day 3 primary decidual cells were also incubated with preimmune serum to demonstrate the specificity of the antibodies (C). D isa transverse section through a day 10 pregnant rat uterus. This panel demonstrates the regional expression of d/tPRP in vivo and is similarto previous observations (Rasmussen et al., 1997). d/tPRP is produced primarily by the antimesometrial decidua (AMD), with little or no d/tPRPin the mesometrial decidua (MD).

2496 DECIDUAL/TROPHOBLAST PRL-RELATED PROTEIN Endo • 1997Vol 138 • No 6

giant cell-specific gene expression (41). The three cell culturemodels exhibit characteristics that reflect the three cell typescapable of d/tPRP expression in situ. Each of these cell pop-ulations was shown in vitro to express d/tPRP and to spe-cifically activate a 3960-bp d/tPRP promoter. Modestd/tPRP promoter activation was also seen in two uterinestromal cell lines (UI and CUS V4). Although these cell linesshow minimal endogenous d/tPRP expression, they mayrepresent a stem cell population with some potential to dif-ferentiate into decidual cells. Initial attempts to induce anantimesometrial decidual cell phenotype after progesteroneand/or PGE2 treatment of the uterine stromal cell lines haveproven unsuccessful (Orwig, K. E., and M. J. Soares, unpub-lished observations). Rcho-1 trophoblast cells, which exhibitcontrolled and robust trophoblast giant cell differentiation,activated the 3960-bp d/tPRP promoter in a differentiation-dependent manner. Using a similar experimental paradigm,the PL-I and cholesterol side-chain cleavage cytochromeP450 promoters were previously shown to be transcription-ally activated during trophoblast giant cell differentiation(28, 42, 43).

Regulatory mechanisms controlling the expression of de-cidua- and trophoblast-specific genes are just beginning to be

elucidated. In the present report, we showed that regionslocated within 4 kb upstream of the d/tPRP gene are capableof specifically directing reporter gene expression to decidual,spongiotrophoblast, and trophoblast giant cells. It is likelythat at least some of the regulatory elements controllingdecidua-, spongiotrophoblast-, or trophoblast giant cell-specific expression will be differentially located within the3960-bp d/tPRP promoter.

Decidual cell-specific promoter analyses have been limitedto studies on human insulin-like growth factor binding pro-tein-1, human PRL, and rat d/tPRP genes. The insulin-likegrowth factor binding protein-1 gene is regulated in part viathe interaction of the Sp3 protein with cis-elements locatedbetween 22.8 to 22.6 kb of its 59-flanking DNA (44). De-cidual expression of PRL is directed by a 3-kb regulatoryregion located approximately 6 kb upstream of the pituitarytranscription start site (45–47). This decidual regulatory re-gion is distinct from the region controlling pituitary PRLgene expression (46, 47). Progesterone and PGE2 are activa-tors of pathways controlling decidualization (1, 47–51) andmay also activate transcription of decidual-specific genes.Distinct expression patterns for a few DNA-binding proteins,including the basic helix-loop-helix transcription factor(bHLH) Hed/Thing-2/d-HAND (52–54), Wilms’ tumor-1(55), and the retinoid X receptor-a (56), suggest the existenceof additional candidate regulators controlling decidualiza-

FIG. 7. Cell-specific expression of d/tPRP mRNA. RT-PCR was usedto evaluate d/tPRP expression in various cell lines. Upper panel,d/tPRP expression was observed in primary decidual cell cultures, butwas not detectable in uterine stromal cell lines (UI, CUS V2, and CUSV4), L929, HRP-1, or GH3 cells. Lower panel, The differentiatedRcho-1 trophoblast cells (d9) and spongiotrophoblast cells expressedd/tPRP, whereas d/tPRP expression could not be detected in prolif-erative Rcho-1 trophoblast cells (d1). The reaction mixture also con-tained primers that amplified the 244-bp region of rat b-actin todemonstrate equal loading. The bottom portion of each panel repre-sents the ethidium bromide-stained gel.

FIG. 8. Primer extension analysis of the d/tPRP transcription startsite in decidua and placenta. The transcription start site was located65 bp upstream of the initiation codon (ATG) in decidual tissue andbetween 61–66 bp upstream in the placenta. Spleen was used as anegative control and exhibited no primer-extended products. DNA ofknown sequence was used to determine the size of the primer-ex-tended product.

DECIDUAL/TROPHOBLAST PRL-RELATED PROTEIN 2497

tion and decidual cell-specific gene expression. The involve-ment of any of these factors or regulatory mechanisms in thecontrol of decidua-specific d/tPRP gene expression is cur-rently being evaluated.

d/tPRP expression in the placenta is restricted to tro-

phoblast giant cells and spongiotrophoblast cells. TheRcho-1 trophoblast cell line has proven useful in the iden-tification of at least some regulatory mechanisms control-ling trophoblast giant cell-specific gene expression. AP-1,GATA, and E-box regulatory elements and their respectiveactivators, fos, jun, GATA-2, GATA-3, and the bHLH tran-scription factor, Hxt/Thing-1/eHAND, have been impli-cated in transcriptional control of the trophoblast giant cellPL-I gene (42, 52, 57). Hxt/Thing-1/eHAND has also beenpostulated to direct differentiation along the trophoblastgiant cell lineage (52). Calzonetti and colleagues (58) useda transgenic strategy to identify a 340-bp regulatory regionthat is essential for targeting the 4311 gene to spongio-trophoblast cells. This region contains E-boxes that mayinteract with a bHLH transcription factor, Mash-2, whichwas previously shown to be essential for development ofthe spongiotrophoblast cell lineage (59). Whether similarregulatory mechanisms are involved in trophoblast giantcell- and spongiotrophoblast-specific d/tPRP gene expres-sion remain to be determined.

In summary, we have demonstrated that the d/tPRP geneis organized in a novel exon/intron arrangement and thatthe d/tPRP promoter directs cell type-specific and differ-entiation-dependent expression. Further evaluation of thed/tPRP promoter will facilitate the identification of cis-acting elements and trans-acting factors required for de-cidual- and trophoblast-specific gene expression and the

FIG. 9. Decidua and trophoblast cell-specific activation of thed/tPRP promoter. Promoter-reporter constructs were transientlytransfected into a variety of cell culture systems (see text for timeof transfection in each cell type). Forty-eight hours after transfec-tion, cells were washed and lysates were evaluated for luciferaseactivity. Luciferase activity in each cell type is reported as a per-centage of the pGL-2 promoterless vector. Luciferase activitieswere normalized for transfection efficiency using b-galactosidase.Upper panel, d/tPRP 59-flanking DNA (3960 bp) was capable ofdriving cell-specific luciferase expression in the primary decidualcells and lower levels in CUS V4 and UI uterine stromal cells. Thed/tPRP promoter showed little or no activity in L929 fibroblasts orCUS V2 cells. Luciferase activities (mean light units per 48 h/b-galunits per mg protein 6 SEM) in cells transfected with the pGL-2promoterless vector were: L-929, 158 6 52; Cus V2, 186 6 63; CusV4, 0.39 6 0.09; UI, 6.85 6 0.84; and primary decidual cells, 22.91 65.94. Luciferase activities in the RSV-Luc positive controls were:L-929, 1563 6 93; Cus V2, 6794 6 853; Cus V4, 136 6 9; UI, 2248 6710; and primary decidual cells, 10197 6 854. Lower panel. Sim-ilarly, d/tPRP-luc was expressed in a cell-specific manner by Rcho-1trophoblast cells and spongiotrophoblast primary cultures, but notin GH3 or HRP-1 cells. Luciferase activities in cells transfectedwith the pGL-2 empty vector were: GH3, 0.79 6 0.12; HRP, 47.45 68.90; Rcho-1, 20.61 6 0.52; and primary spongiotrophoblast,13.90 6 0.37. RSV-Luc-positive control luciferase activities were:GH3, 10,671 6 348; HRP, 33,314 6 599; Rcho-1, 11,449 6 694; andprimary spongiotrophoblast, 51,284 6 329. Each bar represents themean 6 SE of three replicate experiments.

FIG. 10. Differentiation-dependent d/tPRP promoter activation introphoblast cells. Rcho-1 trophoblast cells were stably transfectedwith d/tPRP-luc, and three clonal cell lines were isolated, d/tPRP-Rcho-3 (Rcho-3), d/tPRP-Rcho-4 (Rcho-4), and d/tPRP-Rcho-8(Rcho-8). The time course for activation of the d/tPRP promoter inthe Rcho-3 clonal cell line is shown. Lysates were prepared fromd/tPRP-Rcho-3 cells on days 1, 2, 5, 9, and 13 of culture and eval-uated for luciferase activity. Inset, Rcho-4 and Rcho-8 clones alsodemonstrated the ability to activate the d/tPRP promoter in adifferentiation-dependent manner. Lysates collected from cells onday 2 of culture were considered to represent a proliferative cellpopulation (P; open bars), whereas lysates collected from cells onday 9 of culture represented a differentiated population of tropho-blast giant cells (D; black bars). Each bar represents the mean 6SE of three replicate experiments.

2498 DECIDUAL/TROPHOBLAST PRL-RELATED PROTEIN Endo • 1997Vol 138 • No 6

molecular mechanisms controlling decidual and trophoblastdifferentiation.

Acknowledgments

We acknowledge the technical support of Belinda M. Chapman. CUSV2 and CUS V4 uterine stromal cells were provided by the laboratoryof Hans Zingg, McGill University (Montreal, Canada). UI uterine stro-mal cells were obtained from the laboratory of Virginia Rider, Universityof Missouri (Kansas City, MO).

References

1. DeFeo VJ 1967 Decidualization. In: Wynn RM (ed) Cellular Biology of theUterus. Appleton-Century-Crofts, New York, pp 191–290

2. Bell SC 1983 Decidualization: regional differentiation and associated function.Oxf Reprod Biol 5:220–271

3. Parr MB, Parr EL 1989 The implantation reaction. In: Wynn RM, Jollie WP (eds)Biology of the Uterus. Plenum Medical, New York, pp 233–278

4. Soares MJ, Dai G, Cohick CB, Mueller H, Orwig KE, The rodent placentalprolactin family and pregnancy. In: Bazer FW (ed) Endocrinology of Preg-nancy. Humana Press, Totowa, NJ, in press

5. Roby KF, Deb S, Gibori G, Szpirer C, Levan G, Kwok SCM, Soares MJ 1993Decidual prolactin related protein. J Biol Chem 268:3136–3142

6. Gu Y, Soares MJ, Srivastava RK, Gibori G 1994 Expression of decidualprolactin-related protein in the rat decidua. Endocrinology 135:1422–1427

7. Rasmussen CA, Orwig KE, Vellucci S, Soares MJ 1997 Dual expression ofprolactin-related protein in decidua and trophoblast tissues during pregnancy.Biol Reprod 56:647–654

8. Rasmussen CA, Hashizume K, Orwig KE, Xu L, Soares MJ 1996 Decidualprolactin-related protein: heterologous expression and characterization. En-docrinology 137:5558–5566

9. Soares MJ 1987 Developmental changes in the intraplacental distributionof placental lactogen and alkaline phosphatase in the rat. J Reprod Fertil79:93–98

10. Lu X-J, Deb S, Soares MJ 1994 Spontaneous differentiation of trophoblast cellsalong the spongiotrophoblast cell pathway: expression of members of theplacental prolactin gene family and modulation by retinoic acid. Dev Biol163:86–97

11. Piva M, Flieger O, Rider V 1996 Growth factor control of cultured rat uterinestromal cell proliferation is progesterone dependent. Biol Reprod 55:1333–1342

12. Cohen H, Pageaux J-F, Melinand C, Fayard J-M, Laugier C 1993 Normal ratuterine stromal cells in continuous culture: characterization and progestinregulation of growth. Eur J Cell Biol 61:116–125

13. Arslan A, Almazan G, Zingg HH 1995 Characterization and co-culture ofnovel nontransformed cell lines derived from rat endometrial epithelium andstroma. In Vitro Cell Dev Biol 31:140–148

14. Faria TN, Soares MJ 1991 Trophoblast cell differentiation: establishment,characterization, and modulation of a rat trophoblast cell line expressingmembers of the placental prolactin family. Endocrinology 129:2895–2906

15. Hamlin GP, Lu X-J, Roby KF, Soares MJ 1994 Recapitulation of the pathwayfor trophoblast giant cell differentiation in vitro: stage-specific expression ofmembers of the prolactin gene family. Endocrinology 134:2390–2396

16. Hamlin GP, Soares MJ 1995 Regulation of DNA synthesis in proliferating anddifferentiating trophoblast cells: involvement of transferrin, transforminggrowth factor-b, and tyrosine kinases. Endocrinology 136:322–331

17. Soares MJ, Shaberg KD, Pinal CS, De SK, Bhatia P, Andrews GK 1987Extablishment of a rat placental cell line expressing characteristics of extraem-bryonic membranes. Dev Biol 124:134–144

18. De SK, Larsen DB, Soares MJ 1995 Trophoendodermal stem cell-derivedextracellular matrices: absence of detectable entactin and presence of multiplelaminin species. Placenta 16:701–718

19. Tashjian AH, Yasumura Y, Levine L, Sato GH, Parker ML 1968 Establishmentof clonal strains of rat pituitary tumor cells that secrete growth hormone.Endocrinology 82:342–352

20. Southern E 1975 Detection of specific sequences among DNA fragments sep-arated by gel electrophoresis. J Mol Biol 98:503–517

21. Sanger F, Nicklen S, Coulson AR 1977 DNA sequencing with chain-termi-nation inhibitors. Proc Natl Acad Sci USA 74:5463–5467

22. Duckworth ML, Peden LM, Friesen HG 1988 A third prolactin-like proteinexpressed by the developing rat placenta: complementary deoxyribonucleicacid sequence and partial structure of the gene. Mol Endocrinol 2:912–920

23. Faria TN, Deb S, Kwok SCM, Talamantes F, Soares MJ 1990 Ontogeny ofplacental lactogen-I and placental lactogen-II expression in the developing ratplacenta. Dev Biol 141:279–291

24. Chomczynski P, Sacchi N 1987 Single-step method of RNA isolation by acidguanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem162:156–159

25. Koos RD, Olson CE 1989 Expression of basic fibroblast growth factor in the

rat ovary: detection of mRNA using reverse transcription-polymerase chainreaction amplification. Mol Endocrinol 3:2041–2048

26. Brasier A, Tate J, Habener J 1989 Optimized use of the firefly luciferase assayas a reporter gene in mammalian cell lines. BioTechniques 7:1116–1121

27. Bradford MM 1976 A rapid and sensitive method for the quantitation ofmicrogram quantities of protein utilizing the principle of protein-dye binding.Anal Biochem 72:249–254

28. Yamamoto T, Chapman BM, Clemens JW, Richards JS, Soares MJ 1995Analysis of cytochrome P-450 side-chain cleavage gene promoter activationduring trophoblast cell differentiation. Mol Cell Endocrinol 113:183–194

29. Dai G, Liu B, Szpirer C, Levan G, Kwok SCM, Soares MJ 1996 Prolactin-likeprotein-C variant: complementary deoxyribonucleic acid, unique six exongene structure, and trophoblast cell-specific expression. Endocrinology137:5009–5019

30. Linzer DIH, Nathans D 1985 A new member of the prolactin-growth hormonegene family expressed in mouse placenta. EMBO J 4:1419–1423

31. Deb S, Roby KF, Faria TN, Szpirer C, Levan G, Kwok SCM, Soares MJ 1991Molecular cloning and characterization of prolactin-like protein-C comple-mentary deoxyribonucleic acid. J Biol Chem 266:23027–23032

32. Iwatsuki K, Shinozaki M, Hattori N, Hirasawa K, Itagaki S-I, Shiota K,Ogawa T 1996 Molecular cloning and characterization of a new member of therat placental prolactin (PRL) family, PRL-like protein D (PLP-D). Endocrinol-ogy 137:3849–3855

33. Nicoll CS, Mayer GL, Russell SM 1986 Structural features of prolactins andgrowth hormones that can be related to their biological properties. Endocr Rev7:169–203

34. Connor AM, Waterhouse P, Khokha R, Denhardt DT 1989 Characterizationof a mouse mitogen-regulated protein/proliferin gene and its promoter: amember of the growth hormone/prolactin gene superfamily. Biochim BiophysActa 1009:75–82

35. Kessler MA, Schuler LA 1991 Structure of the bovine placental lactogen geneand alternative splicing of transcripts. DNA Cell Biol 10:93–104

36. Shida MH, Jackson-Grusby LL, Ross SR, Linzer DIH 1992 Placental-specificexpression from the mouse placental lactogen II gene promoter. Proc Natl AcadSci USA 89:3864–3868

37. Jackson D, Volpert OV, Bouck N, Linzer DIH 1994 Stimulation and inhibitionof angiogenesis by placental proliferin and proliferin-related protein. Science266:1581–1584

38. Croze F, Kennedy TG, Schroedter IC, Friesen HG 1990 Expression of ratprolactin-like protein B in deciduoma of pseudopregnant rat and in deciduaduring early pregnancy. Endocrinology 127:2665–2672

39. Duckworth ML, Schroedter IC, Friesen HG 1990 Cellular localization of ratplacental lactogen II and rat prolactin-like proteins A and B by in situ hybrid-ization. Placenta 11:143–155

40. Cohick CB, Dai G, Xu L, Deb S, Kamei T, Levan G, Szpirer C, Szpirer J, KwokSCM, Soares MJ 1996 Placental lactogen-I variant utilizes the prolactin re-ceptor signaling pathway. Mol Cell Endocrinol 116:49–58

41. Soares MJ, Chapman BM, Rasmussen CA, Dai G, Kamai T, Orwig KE 1996Differentiation of trophoblast endocrine cells. Placenta 17:277–289

42. Shida MM, Ng YK, Soares MJ, Linzer DI 1993 Trophoblast-specific tran-scription from the mouse placental lactogen-I gene promoter. Mol Endocrinol7:181–188

43. Yamamoto T, Roby KF, Kwok SCM, Soares MJ 1994 Transcriptional activa-tion of cytochrome P450 side chain cleavage enzyme expression during tro-phoblast cell differentiation. J Biol Chem 269:6517–6523

44. Gao J, Tseng L 1996 Distal Sp3 binding sites in the hIGFBP-1 gene promotersuppress trancriptional repression in decidualized human endometrial stro-mal cells: identification of a novel Sp3 form in decidual cells. Mol Endocrinol10:613–621

45. DiMattia GE, Gellersen B, Duckworth ML, Friesen HG 1990 Human pro-lactin gene expression. J Biol Chem 256:16412–16421

46. Berwaer M, Martial JA, Davis JRE 1994 Characterization of an up-streampromoter directing extrapituitary expression of the human prolactin gene. MolEndocrinol 8:635–642

47. Gellersen B, Kempf R, Telgmann R, DiMattia GE 1994 Nonpituitaryhuman prolactin gene transcription is independent of Pit-1 and differen-tially controlled in lymphocytes and in endometrial stroma. Mol Endocrinol8:356 –373

48. Kennedy TG 1983 Prostaglandin E2, adenosine 39:59-cyclic monophosphateand changes in endometrial vascular permeability in rat uteri sensitized for thedecidual cell reaction. Biol Reprod 29:1069–1076

49. Kennedy TG 1986 Intrauterine infusion of prostaglandins and decidualizationin rats with uteri differentially sensitized for the decidual cell reaction. BiolReprod 34:327–335

50. Yee GM, Kennedy TG 1993 Prostaglandin E2, cAMP and cAMP-dependentprotein kinase isozymes during decidualization of rat endometrial stromalcells in vitro. Prostaglandins 46:117–138

51. Frank GR, Brar AK, Cedars MI, Handwerger S 1994 Prostaglandin E2 en-hances human endometrial stromal cell differentiation. Endocrinology134:258–263

52. Cross JC, Flannery ML, Blanar MA, Steingrimsson E, Jenkins NA, CopelandNG, Rutter WJ, Werb Z 1995 Hxt encodes a basic helix-loop-helix transcription

DECIDUAL/TROPHOBLAST PRL-RELATED PROTEIN 2499

factor that regulates trophoblast cell development. Development121:2513–2523

53. Hollenberg SM, Sternglanz R, Cheng PF, Weintraub H 1995 Identification ofa new family of tissue-specific basic helix-loop-helix proteins with a two-hybrid system. Mol Cell Biol 15:3813–3822

54. Srivastava D, Cserjesi P, Olson EN 1995 A subclass of bHLH proteins requiredfor cardiac morphogenesis. Science 270:1995–1999

55. Zhou J, Rauscher FJ, Bondy C 1993 Wilms’ tumor (WT1) gene expression inrat decidual differentiation. Differentiation 54:109–114

56. Mangelsdorf DJ, Borgmeyer U, Heyman RA, Zhou JY, Ong ES, Oro AE,Kakizuka A, Evans RM 1992 Characterization of three RXR genes that mediatethe action of 9-cis retinoic acid. Genes Dev 6:329–344

57. Ng Y-K, George KM, Engel JD, Linzer DIH 1994 GATA factor activity isrequired for the trophoblast-specific trnascriptional regulation of the mouseplacental lactogen I gene. Development 120:3257–3266

58. Calzonetti T, Stevenson L, Rossant J 1995 A novel regulatory region is re-

quired for trophoblast-specific transcription in transgenic mice. Dev Biol171:615–626

59. Guillmot F, Nagy A, Auerbach A, Rossant J, Joyner AL 1994 Essential role ofMash-2 in extraembryonic development. Nature 371:333–336

60. Takayama Y, Rand-Weaver M, Kawauchi H, Ono M 1991 Gene structure ofchum salmon somatolactin, a presumed pituitary hormone of the growthhormone/prolactin family. Mol Endocrinol 5:778–786

61. Tanaka M, Hosokawa Y, Masanori W, Nakashima K 1992 Structure of thechicken growth hormone-encoding gene and its promoter region. Gene112:235–239

62. Truong AT, Duez C, Belayew A, Renard A, Pictet R, Bell GI, Martial JA 1984Isolation and characterization of the human prolactin gene. EMBO J 3:429–437

63. Chien Y-H, Thompson EB 1980 Genomic organization of rat prolactin andgrowth hormone genes. Proc Natl Acad Sci USA 77:4583–4587

64. Cooke NE, Baxter JD 1982 Structural analysis of the prolactin gene suggestsa separate origin for its 59 end. Nature 297:603–606

2500 DECIDUAL/TROPHOBLAST PRL-RELATED PROTEIN Endo • 1997Vol 138 • No 6