Functional Characterization of the Intermediate Isoform of the Human Prolactin Receptor*

Prolactin-dependent signaling occurs as the result of ligand-induced dimerization of the prolactin receptor (PRLr). While three PRLr isoforms have been characterized in the rat, studies have suggested the existence of several human isoforms in breast carcinoma species and normal tissues. Reverse transcription polymerase chain reaction was performed on mRNA isolated from the breast carcinoma cell line T47D, revealing two predominant receptor isoforms: the previously described long PRLr and a novel human intermediate PRLr. The nucleotide sequence of the intermediate isoform was found to be identical to the long isoform except for a 573-base pair deletion occurring at a consensus splice site, resulting in a frameshift and truncated intracytoplasmic domain. Scatchard analysis of the intermediate PRLr revealed an affinity for PRL comparable with the long PRLr. While Ba/F3 transfectants expressing the long PRLr proliferated in response to PRL, intermediate PRLr transfectants exhibited modest incorporation of [3H]thymidine. Significantly, however, both the long and intermediate PRLr were equivalent in their inhibition of apoptosis of the Ba/F3 transfectants after PRL treatment. The activation of proximal signaling molecules also differed between isoforms. Upon ligand binding, Jak2 and Fyn were activated in CHO-K1 cells transiently transfected with the long PRLr. In contrast, the intermediate PRLr transfectants showed equivalent levels of Jak2 activation but only minimal activation of Fyn. Last, Northern analysis revealed variable tissue expression of intermediate PRLr transcript that differed from that of the long PRLr. Taken together, differences in signaling and tissue expression suggest that the human intermediate PRLr differs from the long PRLr in physiological function.

The neuroendocrine hormone prolactin (PRL) 1 exhibits high homology to growth hormone and is also related to the peptide hormones of the interleukin family (1,2). PRL has been implicated in the proliferation and differentiation of lobular units as well as the initiation and maintenance of lactation (3,4). It has also been shown to be an essential component of the T cell immune response, serving as a cofactor for T lymphocyte activation (5,6). Regulation by PRL may also extend to the autocrine level as the synthesis and secretion of PRL by mitogenstimulated T cells (7,8) and within breast epithelium (9,10) has been identified.
PRL exerts its effects at the molecular level by inducing the homodimerization of the prolactin receptor (PRLr). A member of the cytokine receptor family, the PRLr lacks intrinsic enzymatic activity, thus requiring the activation of associated kinases and other signaling factors for ligand-driven transduction. Two protein-tyrosine kinases found in association with the PRLr are p59 fyn (11) and p120 jak2 (12)(13)(14)(15). Through Jak2, PRL stimulation activates Stat family members in lymphocytes (16) and breast tissues (17,18), resulting in the initiation of transcription for interferon regulatory factor-1 and ␤-casein gene products, respectively. PRLr dimerization also induces the GRB2/SOS/Ras/Raf/MEK/MAPK signaling cascade, ultimately activating several transcription factors necessary for cell cycle progression including Myc, Jun, and T cell factor (19 -21).
While the diversity of PRL function is in part mediated by a variety of signaling cascades, differences in function may also be attributed to the wide variety of PRLr forms observed in nature. As members of the cytokine receptor superfamily, the PRLr isoforms show significant sequence similarity in their extracellular ligand-binding domains. Within the membraneproximal region of the intracytoplasmic domain of PRL receptors and other superfamily members lie the conserved Box 1 and Box 2 motifs. Box 1 is a hydrophobic, proline-rich region that resembles an SH3 binding domain (22)(23)(24). The Box 2 domain is hydrophobic and acidic, and its signaling function is largely uncharacterized. Several isoforms of the PRLr have been identified in both mammals (25)(26)(27)(28)(29)(30)(31) and birds (32)(33)(34). The most well characterized isoforms are those found in the rat: the short form (45 kDa) (35), long form (80 -85 kDa) (36), and a mutant intermediate form found on the PRL-dependent rat T cell lymphoma line Nb2 (65 kDa) (Fig. 3B) (36,37). In humans, the only PRLr isoform characterized thus far is the long form cloned from the liver (Fig. 3A) (38). Previous studies have, however, provided evidence that other human PRLr isoforms may be expressed in human tissues (30,39).
In this study, we identify a novel isoform of the human PRLr cloned from the human breast cancer cell line T47D. The isoform is analyzed for 1) in vivo surface expression and its ability to bind ligand, 2) induction of cell proliferation and cell survival in response to ligand, 3) the ability to activate associated kinases, and 4) the relative levels of its corresponding mRNA in normal human tissues.

EXPERIMENTAL PROCEDURES
mRNA Isolation and Reverse Transcription PCR-T47D cells, an estrogen receptor/PRLr-positive human breast cancer cell line, were used for mRNA isolation. Whole RNA was purified from 10 7 washed cells using Trizol reagent (Life Technologies, Inc.) as described previously (39). Messenger RNA was then purified from the whole RNA preparation with oligo(dT)-cellulose (Invitrogen, San Diego, CA). 5 g of T47D mRNA was used for first strand synthesis of cDNA using the Superscript II RT cDNA kit (Life Technologies, Inc.). Negative controls consisted of reactions containing no T47D mRNA or no reverse transcriptase. A positive control reaction consisted of chloramphenicol acetyltransferase mRNA template. For polymerase chain reaction, 2 l of the corresponding cDNA reactions were added to a 50-l reaction containing 5-l 10ϫ PCR buffer, 3 l of 25 mM MgCl 2 , 1 l of 10 mM dNTP mix, 5 units of Taq polymerase (Life Technologies, Inc.), and primers for amplification. As the positive control reaction, primers A (5Ј-GACATGGAAGCCATCACAGAC-3Ј) and B (5Ј-CGACCGTTCAGC-TGGATATTA-3Ј) were used to amplify a fragment of the chloramphenicol acetyltransferase gene from control cDNA. The PRLr gene amplification reaction contained primers PRLR-F3 (5Ј-ATGAAGGAA-AATGTGGCA-3Ј) and PRLR-1 (5Ј-TCAGTGAAAGGAGTGTGT-3Ј), which correspond to the 5Ј-and 3Ј-ends of the human long PRLr open reading frame. The primary cycle of the reaction consisted of 94°C for 2 min, 42°C for 1 min, 72°C for 3 min, and 94°C for 2 min, which was followed by 30 cycles of 94°C for 30 s, 47°C for 30 s, and 72°C for 2 min. It was then extended at 72°C for 3 min. Isolated PCR fragments were subcloned into the TA vector pCR 2.1 (Invitrogen, San Diego, CA) and analyzed by dideoxynucleotide sequencing. For eukaryotic expression of the intermediate isoform, the gene was reamplified by PCR with primers PRLR-Kl (5Ј-CGAATTCCACCATGAAGGAAAATGTGGCA-3Ј) and PRLR-599Ј (5Ј-GCGCTCGAGTCAGTGAAAGGAGTGTGTAAA-3Ј), which contain a 5Ј EcoRI restriction site and Kozak initiation sequence and a 3Ј XhoI restriction site, respectively. Alternative 3Ј primers were also utilized to remove the tertiary stop codon from the open reading frames of the isoforms, allowing the addition of a carboxyl-terminal V5 epitope tag when ligated into vector pEF1-V5/HisA (Invitrogen). The intermediate isoform was reamplified with primers PRLR-Kl and PRLR-INTЈ (5Ј-GCGCTCGAGGGAGTCCCGGGCTTC-3Ј), while the long isoform was reamplified with PRLR-Kl and PRLR-LONGЈ (5Ј-CG-CTCGAGGTGAAAGGAGTGTGTAAA-3Ј). The DNA fragments were digested with EcoRI and XhoI and ligated into the corresponding restriction sites of pcDNA3 and pEF1-V5/HisA. The clones were subsequently checked for amplification errors by dideoxynucleotide sequencing.
Cell Culture and Transfection-T47D cells were maintained in Dulbecco's modified Eagle's medium (Life Technologies) supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin. The mouse interleukin 3 (IL-3)-dependent pro-B cell line Ba/F3 was maintained in RPMI 1640 medium (Life Technologies) supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin in the presence of 1 ng/ml IL-3 (PeproTech, Rocky Hill, NJ). Chinese hamster ovary (CHO-K1) cells were maintained in Ham's F-12 medium (Life Technologies) supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin. Ba/F3 cells (10 7 ) were transfected with 50 g of intermediate or long isoform cDNA clones in pcDNA3 by exposure to a single voltage pulse (0.6 kV, 25 microfarads for 0.1 s) in a Gene Pulser electroporator (Bio-Rad). Stable clones were obtained by limiting dilution by selection in 750 g/ml G418. CHO cells (2 ϫ 10 5 ) were transiently transfected with 2 g of intermediate or long isoform cDNA clones in pEF1-V5/HisA in conjunction with 2 g of human Jak2 (gift of Dr. Roy Duhe) or murine Fyn (gift of Dr. Paul Stein) cDNA in pEF1-V5/HisA using Fugene 6 (Roche Molecular Biochemicals) as instructed.
Northern Analysis of PRLr Isoform Transfectants and Isoform Expression in Various Tissues-Total RNA from cells was isolated from T47D cells by extraction with Trizol reagent (Life Technologies) as described previously (39). 10 g of total RNA was denatured and subjected to electrophoresis on a 1% agarose formaldehyde gel and transferred onto a nylon membrane. A cDNA probe generated from bp 73-702 of the extracellular domain of the human PRLr long isoform was labeled with [␣-32 P]dCTP in the presence of random hexamers using the Oligo Labeling Kit (Amersham Pharmacia Biotech). The probe was hybridized to the membrane at a final concentration of 100 ng/ml at 68°C for 1 h in Express-hyb solution (CLONTECH, Palo Alto, CA) as per the manufacturer's instructions. The blot was then washed three times in 2ϫ SSC, 0.05% SDS at room temperature followed by two 20 min washes in 0.1ϫ SSC, 0.1% SDS at 50°C, followed by autoradiography.
A master tissue blot of human total mRNA (CLONTECH) was probed  (38)). Hybridization conditions were performed as instructed by CLONTECH. Under these conditions, no cross-hybridization was observed between isoforms (data not shown). The blot was exposed to x-ray film for 4 days, and signal intensities were obtained using ImageQuaNT densitometry software (Molecular Dynamics, Inc., Sunnyvale, CA).
Immunoblot Analysis-CHO cell transfectants were lysed in Laemmli buffer containing SDS and 2-mercaptoethanol (5). Lysates were electrophoresed through an 8% SDS-polyacrylamide gel and transferred to nitrocellulose. Nonspecific binding was blocked with 5% milk in PBS/Tween 20. Antigen was labeled with 1 g of horseradish peroxidase-conjugated anti-V5 antibody (Invitrogen) per ml. Antigen-antibody complexes were visualized by enhanced chemiluminescence (Amersham Pharmacia Biotech).
Immunofluoresence Microscopy-10 6 Ba/F3 transfectants were harvested and washed with PBS at 4°C. Cells were then stained with a 1:100 dilution of rabbit anti-PRLr antiserum developed by our laboratory and characterized elsewhere (40) for 1 h at 4°C. After washing three times with PBS, bovine serum albumin, 0.1% sodium azide, the cells were incubated for 30 min with a 1:2000 dilution of fluorescein 5-isothiocyanate-conjugated goat anti-rabbit secondary antibody. Cells were washed three times with PBS/bovine serum albumin/sodium azide, fixed with PBS, 4% paraformaldehyde for 15 min and resuspended in PBS/bovine serum albumin/sodium azide. Cellular immunofluoresence was examined using a Zeiss Axioskop2 immunofluorescence microscope (Carl Zeiss, Inc., Thornwood, NY) and an Apogee CCD camera (Axiom Research, Tucson, AZ). This antiserum reactivity was specific for PRLr expression, since the addition of competitive immunizing peptide was previously shown to inhibit PRLr staining (40).
Scatchard Analysis-Ligand binding affinities were determined as described previously (41). Briefly, 10 6 transfected Ba/F3 cells were incubated with increasing concentrations of human [ 125 I]PRL in a total volume of 100 l of RPMI, 0.1% sodium azide. Nonspecific binding was estimated by incubating separate tubes with unlabeled competitor PRL at a concentration 100 times greater than that of labeled ligand. Cells were incubated at 4°C for 2 h with agitation and pelleted through an oil gradient (90% dibutyl phthalate, 10% olive oil). Pellets were cut from the tubes, and cell-associated [ 125 I]PRL was measured on a 1272 Clinigamma ␥-counter (EG & G Wallac, Akron, OH).
Quantitation of Ligand-induced Proliferation and Survival of PRLr Isoform Transfectants-To assess PRL-induced cellular proliferation, 5 ϫ 10 4 long and intermediate isoform Ba/F3 transfectants were aliquoted in medium consisting of RPMI 1640 medium supplemented with sodium selenide, linoleic acid, insulin, and transferrin (ITSϩ; Calbiochem, Bedford, MA) in the presence of 0 -1 g/ml human PRL or murine IL-3. After overnight culture, cells were pulsed with 0.5 Ci of [ 3 H]thymidine at 37°C for 4 h. Incorporation of radiolabel was determined by scintillography. To assess the viability of PRLr isoform transfectants, Ba/F3 cells transfected with the PRLr intermediate isoform, the long isoform construct, or control vector were plated at 10 6 cells/well in 2 ml of RPMI 1640 ITSϩ with or without 10 ng/ml PRL. Cells were harvested over a 72-h period, and the numbers of both dead and viable cells were determined by trypan blue dye exclusion. The percentage of viability was calculated by the number of live cells/(live ϩ dead cells) ϫ 100%.
Immunoprecipitation and in Vitro Kinase Assays-After PRL stimulation (250 ng/ml), 2 ϫ 10 5 CHO cells transfected with PRLr isoforms in conjunction with Jak2 or Fyn cDNAs (all expressed in vector pEF1-V5/ HisA) were lysed and immunoprecipitated overnight as described previously (11) using 3 l anti-Fyn (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) or anti-Jak2 (Upstate Biotechnology, Lake Placid, NY) antibodies. Antigen-antibody complexes were isolated by the addition of 50 l of protein A beads. After three washes with lysis buffer, immunoprecipitates were washed once with low salt buffer (10 mM Tris (pH 7.0), 100 mM NaCl, and 100 M Na 3 VO 4 ). The protein A-Sepharose beads were then suspended in 30 l of autokinase buffer (25 mM Tris (pH 7.0), 10 mM MnCl 2 , and 10 Ci of [␥-32 P]ATP). After 20 minutes at 30°C, the reactions were stopped by the addition of 2ϫ Laemmli buffer with mercaptoethanol, and the reaction products were analyzed by 10% SDS-polyacrylamide gel electrophoresis followed by autoradiography. (42), only one has been found expressed in human tissues. Investigations from our laboratory (39) and others (30) found evidence suggesting the existence of at least one additional human isoform. To identify and characterize these putative PRLr isoforms, reverse transcription PCR was performed on cDNA generated from the breast cancer cell line T47D using oligonucleotides homologous to the 5Ј-and 3Ј-ends of the human long PRLr. Gel electrophoresis of the amplified cDNA revealed DNA fragments of 1.3, 1.5, and 1.8 kilobases (Fig. 1, lane 4). Both negative control reactions failed to amplify any DNA fragments, suggesting that the bands identified were indeed generated from cDNA template and not the result of chromosomal DNA contamination ( Fig. 1, lanes 1 and 2). The three fragments were excised and cloned, and their DNA sequences were determined. The 1.8-kilobase band was found to be the long PRLr isoform previously described (38). The 1.5-kilobase fragment was an mRNA splice variant coding for a truncated extracellular domain designated ⌬S1. 2 In contrast, the 1.3-kilobase sequence revealed an mRNA splice variant coding for an isoform with a deletion in the intracellular domain ( Fig. 2A). This DNA sequence is most likely the result of an RNA splicing event, since a consensus splice site was present at the juncture between base pairs 1009 and 1582 (39). The open reading frame is homologous to the long isoform up to base pair 1009, where a deletion of 573 nucleotides occurs, juxtaposing base pair 1009 to 1583 (Fig. 2B). While this isoform demonstrated 100% homology to the long form downstream of base pair 1582, the deletion caused a shift in the reading frame, altering the carboxyl amino acid sequence and generating a stop codon 13 residues after the splice junction (Fig. 2B). Based on the proximity of the splice junction to the gene deletion in the rat Nb2 intermediate isoform (Fig. 8) (37), this PRLr variant was designated the human intermediate PRLr.

Isolation of the Human Intermediate PRLr-Although three isoforms of the PRLr have been discovered in the rat
In Vivo Expression of the Intermediate PRLr Isoform-To investigate the physiological activities of the intermediate isoform in PRLr signaling, a eukaryotic expression vector containing the intermediate PRLr cDNA was transfected transiently into CHO cells and stably into the murine pro-B cell line Ba/F3. To serve as controls, empty vector (negative control) and the long PRLr isoform (positive control) were also used to transfect cells. Using a DNA probe homologous to the extracellular domain common to both isoforms, Northern analysis of the transfectants showed mRNA transcripts of the predicted molecular mass for both long and intermediate isoform-expressing clones (data not shown). The intermediate isoform was also analyzed for its ability to be efficiently translated. The cDNAs for both forms were subcloned into the vector pEF1V5/HisA, which enabled the addition of a V5 epitope tag to the carboxyl-terminal ends of both PRLr variants. Immunoblotting of lysates from CHO cells transiently transfected with the constructs revealed proteins of the correct molecular mass previously reported for the long form (85 kDa) (43) and approximately 50 kDa for the intermediate isoform (Fig. 3). The predicted molecular mass for the epitope-tagged intermediate PRLr is 32.7 kDa. Glycosylation of the extracellular domain in a fashion similar to that found on the long PRLr could account for the difference in the predicted and observed mobilities, since treatment of both transfectants with tunicamycin resulted in a decrease in apparent molecular mass (data not shown).
To confirm the expression of the intermediate PRLr on the cell surface, stable Ba/F3 transfectants were stained with anti-PRLr antiserum and examined for surface staining of the receptor (Fig. 4). Both long and intermediate PRLr transfectants showed high levels of a speckled, cell surface staining pattern compared with vector alone (Fig. 4, columns 1 and 2 versus  column 3), while staining with preimmune serum only resulted in background labeling of the transfectants (Fig. 4, row 2).
The Intermediate PRLr Isoform Binds Ligand with an Affinity Comparable with the Long PRLr Isoform-The rat intermediate PRLr has been shown to bind ligand with a higher affinity than the rat long form (37 (44,45). To examine the effects of human intermediate PRLr homodimerization, transfectants were incubated with increasing concentrations of human PRL and analyzed for DNA replication by tritiated thymidine incorporation (Fig. 5A). As expected, Ba/F3 cells expressing the long PRLr isoform proliferated in response to ligand, while cells transfected with vector alone did not proliferate at any PRL concentration. The intermediate PRLr could induce modest proliferation, but only at the highest concentration of PRL used (1 g/ml). This lack of significant proliferation observed was not the result of the transfection or selection process, since all three transfectants stimulated with IL-3 showed comparable levels of proliferation (Fig. 5B). Additionally, these results The PRLr Intermediate Isoform Enhances Cell Viability-While the intermediate PRLr isoform was unable to induce proliferation of Ba/F3 transfectants, this receptor variant may have an alternative physiological role. For example, the long PRLr also mediates cell survival and resistance to dexamethasone-mediated apoptosis (46,47). In addition, we have previously demonstrated that in the absence of appropriate ligand, Ba/F3 cells die an apoptotic death that corresponds to the measurement of trypan blue viability (46). To determine if the human intermediate isoform may enhance cell survival, Ba/F3 transfectants were incubated with or without PRL over a 3-day period, and cell viability was determined by trypan blue dye exclusion. Upon the addition of ligand, cells transfected with either isoform showed enhanced levels of cell viability as com-pared with vector-transfected cells (Fig. 5C), while both long and intermediate transfectants showed high levels of cell death in the absence of ligand (Fig. 5D). In summary, although the intermediate isoform is unable to invoke the complete signaling pathways necessary for cell division, the intermediate isoform inhibits apoptosis in Ba/F3 transfectants, enhancing the survival of cells in the presence of ligand.
Protein-tyrosine Kinase Activation Differs between the Long and Intermediate PRLr Isoforms-Since the Ba/F3 intermediate isoform transfectant was unable to induce proliferation upon ligand binding but did enhance cell survival, we wished to examine if there were differences in the proximal signaling pathways activated by the long and intermediate prolactin receptors. The protein-tyrosine kinases p59 fyn and Jak2 are known to be activated upon ligand stimulation of the PRLr (11,12). To investigate whether the PRLr intermediate isoform was capable of activating these proximal signaling molecules, in vitro kinase assays were carried out using CHO cells transiently transfected with constructs expressing these molecules in conjunction with the PRLr isoform constructs. The interme- diate PRLr was capable of activating Jak2 in a temporal fashion identical to the long PRLr (Fig. 6A), with maximal expression occurring 15 min after the addition of ligand. In contrast to these results, the long PRLr isoform exhibited high levels of Fyn activation, while the ability of the intermediate isoform to stimulate Fyn was greatly diminished (Fig. 6B). In summary, the pattern of proximal protein-tyrosine kinase activation differed between the long and intermediate isoforms, suggesting differences in proximal cell signaling in response to ligand.
The Human PRLr Intermediate Isoform Shows Variable Tissue Expression-Previous studies have shown that the number of isoforms and levels of PRLr expression vary between tissues in several different species (28,30,31,48). To determine if the intermediate isoform shows the same variability in expression between tissues, a dot blot containing mRNA isolated from a variety of human tissues were probed with cDNA fragments specific for either the long or intermediate isoforms (Fig. 7). To avoid cross-hybridization, probes were generated that would hybridize only to the mRNA transcripts encoding either the long or intermediate PRLr isoforms. Relative levels of expression of both isoforms were compared with those of the pituitary. Tissue expression of long PRLr mRNA varied greatly between tissues, an observation previously reported in the rat (48). The highest levels of expression of both isoforms were observed in the placenta, demonstrating over 6-fold more long PRLr and 5-fold more intermediate PRLr mRNA in comparison with the pituitary. Aside from the high placental levels, mRNA expression between isoforms differed among the remaining tissues. For example, the next three greatest amounts of long PRLr mRNA were found in the adrenal gland, pituitary, and hippocampus. In contrast, the highest levels of intermediate PRLr mRNA were found in the adrenal gland, small intestine, and kidney. In summary, the intermediate form of the PRLr shows significant variability in its expression between tissues, and this pattern differs from that of the long form, suggesting a differing physiological role for the new isoform. DISCUSSION While three PRLr isoforms have been identified and characterized in the rat, only the long PRLr has been characterized in humans. However, previous studies have strongly suggested the existence of several human PRLr isoforms. For the first time, we report the characterization of a novel, full-length human PRLr isoform isolated from the human breast carcinoma cell line T47D. The human intermediate isoform derives its name from the similarity it shares with the rat intermediate PRLr. Unlike its rat counterpart, the human intermediate PRLr results from an RNA processing event occurring at a consensus splice junction. This isoform was termed the intermediate PRLr because of the similarities it shares with the rat intermediate PRLr in both the proximity of the splice junctions and sizes of the nucleotide deletions observed (Fig. 8). Unlike the rat intermediate PRLr, the splice also induces a frameshift after residue 312, causing the addition of 13 heterologous amino acids and a premature stop codon.
Immunoblotting of transiently transfected CHO cells expressing the intermediate PRLr revealed a molecular mass of approximately 50 kDa, compared with the 85-90-kDa-long PRLr observed by us and others (43). Given that the extracellular domains of both human isoforms are identical, it can be inferred the glycosylation patterns of both forms may be similar. Therefore, the deletion of 573 base pairs resulting in the loss of 30 kDa from the intracellular domain of the intermedi- Ba/F3 cells were previously shown to be a reliable background to measure cell proliferation, since the expression of the PRLr renders them PRL-responsive (45,50). The intermediate PRLr transfectants exhibited no proliferation at physiological concentrations of PRL and only modest [ 3 H]thymidine incorporation at pharmacologic concentrations of PRL. Ba/F3 cells transfected with the long PRLr, in comparison, demonstrated robust proliferaton to physiologic concentrations of PRL. In contrast to these results, the enhancement of cell survival over a 72-h period was comparable between the intermediate and the long PRLr. Taken together, these results suggested that fundamental signaling differences existed between the two human PRLr isoforms. In conjunction with the above proliferation studies, transient transfections were undertaken to delineate differences in proximal PRLr signaling between the two human PRLr isoforms. The addition of PRL to intermediate PRLrexpressing transfectants resulted in the activation of Jak2, while the activation of Fyn was severely diminished. This was in contrast to the human long PRLr, which induced the activation of both Jak2 and Fyn upon ligand binding. These differences in proximal signaling may help explain the physiological characteristics attributable to cells expressing the intermediate PRLr. Previous mutational analysis of the PRLr by others (16,18,49) and us (50) indicated that residues in the C terminus may be necessary for maximal signaling and proliferation due to their associations with Stat5 and Fyn. Mutation of C-terminal tyrosines in the rat intermediate PRLr isoform affected Stat5-associated and Stat5-mediated gene transcription (49). This is significant in that the human intermediate PRLr is truncated by a frameshift upstream of these tyrosine residues, resulting in an intracellular domain that lacks putative Stat5 binding motifs.
The lack of C-terminal tyrosines may also explain the poor activation of Fyn via the intermediate PRLr and explain the altered mitogenic response induced by the intermediate isoform. Fyn has been shown to play an essential role in the platelet-derived growth factor receptor (51), Ig (52), and T-cell receptor (53) signaling. In addition, we have shown that Fyn is activated during PRL stimulation of Nb2 cells (11,50). We previously reported that the truncation of the rat intermediate PRLr at residue 322 or the mutation of the C-terminal Tyr 382 completely abrogated the mitogenic response, and these two forms were incapable of activating Fyn (54). Taken together, the lack of the C-terminal domain in the intermediate PRLr may preclude Fyn and Stat5 activation, thereby inhibiting the mitogenic response.
While it has been suggested that the proto-oncogene Cbl may play a role in PRL-mediated Nb2 cell survival (57), several lines of evidence also suggest that the activation of Jak2 via the intermediate PRLr may be responsible for the enhanced survival of Ba/F3 transfectants. For instance, IL-3-dependent 32Dcl3 cells transfected with a dominant negative form of Jak2 exhibited decreased levels of Erk-2 kinase and accelerated apoptosis (58). A link between Jak2 inactivation and apoptosis was also shown in human blood eosinophils (59). When cells were treated with the Jak2 inhibitor tyrphostin B42 in the presence of granulocyte-macrophage colony-stimulating factor, receptor dimerization was unable to prevent eosinophil apoptosis. Last, activated Jak2 has been implicated in increasing the expression levels of Bcl-2 (60), a member of a family of genes thought to serve as central regulators of programmed cell death (61,62). A link between PRLr signaling and the induction of another antiapoptotic gene, the Bcl-2-associated protein Bag-1, has also been shown by our laboratory (46). Ba/F3 transfectants overexpressing Bag-1 were rendered IL-3-independent, while ligand stimulation and survival of Nb2 cells was associated with increased Bag-1 levels. It is quite possible that intermediate PRLr signaling, through Jak2, may increase Bcl-2 and/or Bag-1 levels, explaining the enhanced survival of FIG. 6. Jak2 but not p59 fyn is activated by the PRLr intermediate isoform. CHO cells expressing PRLr receptor isoforms in conjunction with Jak2 (A) or p59 fyn (B) were stimulated with 250 ng/ml human PRL for the indicated times. Jak2 and Fyn were immunoprecipitated with corresponding antisera and incubated with [␥-32 P]ATP without exogenous substrates. After washing, the precipitates were separated by 10% SDS-polyacrylamide gel electrophoresis and visualized by autoradiography. The same amounts of the samples used in the in vitro kinase assays were immunoprecipitated and immunoblotted to show equal protein expression between samples (lower panels).   (63) previously reported PRLr gene expression in placental trophoblast and Western analysis detected six molecular species, two of which were approximately the size of the intermediate PRLr described herein. While the physiological function of lactogenic binding proteins in the placenta is unknown, prolactin receptors in uteroplacental tissues are believed to mediate effects of lactogenic hormones on decidual function at midgestation (64,65). Given that both PRLr isoforms were found at high levels on these tissues, regulation of hormone action during pregnancy in the placental unit could result from PRLr isoform hetero-and/or homodimerization.
Of the three tissues examined from the gastrointestinal tract (stomach, small intestine, and colon), the small intestine exhibited the greatest expression of intermediate PRLr. This correlated with previously described high levels of rat short PRLr expression (30). It is well established that PRL plays an important role in the immune system including an increase in the cytotoxic activities of natural killer cells and the proliferation of lymphocytes from the spleen or lymph nodes (66). It is possible that the intermediate PRLr may play a role in the immunomodulation of the gut, since the intestinal tract is known to be the first barrier against bacterial pathogens. Another function of PRL in the small intestine is to regulate water and electrolyte transport across the epithelium. The lack of PRL-induced responsiveness in the colon (67) (68). This suggests a direct effect of PRL on adrenal cells, and in fact it was shown that PRL increases steroidogenesis, enhancing the secretion of aldosterone and cortisol in human adrenal cell cultures (68). One or both isoforms may form part of a feedback mechanism in the adrenal gland, as cortisol is known to inhibit PRL secretion.
Another organ that expressed high levels of intermediate PRLr transcript was the kidney. Prolactin is known to induce ornithine decarboxylase activity in the kidney (69), and the low amount of detectable long PRLr transcript observed suggests that the intermediate PRLr may be capable of inducing this enzyme. On the other hand, the kidney regulates systemic water electrolyte balance and is known to play a major role in osmoregulation in amphibians and fish (70,71). There is sizable evidence supporting the existence of PRL binding sites in the kidney (72)(73)(74)(75), and PRLr mRNA was found in renal tubules (76,77). Indeed, both PRLr and PRL can be localized on the parietal epithelium of Bowman's capsule (78). Taken together, this suggests an autocrine/paracrine loop within the parietal epithelium, regulated by one or both human PRLr isoforms, which may contribute to osmoregulation.
Previous studies of PRLr isoform expression have shown they are not equally represented in all tissues and may in fact regulate the effects of PRL through hetero-and homodimerization. In the mouse, for example, examination of the transcripts of both short and long isoforms showed variation of expression in a tissue-specific manner, depending on the stage of the estrous cycle, pregnancy, and lactation (48,79). This variable expression suggests differences in the physiological roles of PRLr isoforms and co-transfection experiments utilizing different PRLr isoforms further support this hypothesis. Transient transfection of BMGE cells with rat long and short forms showed an inhibition of PRL-induced ␤-casein gene transcription directly proportional to the ratio of short to long form expressed (80). In other studies, granulocyte colony-stimulating factor receptor/PRLr chimeras were used to hetero-and homodimerize the intracellular domains of rat PRLr isoforms (45)  here may represent a physiologic mechanism through which the tissue-specific actions of this receptor complex are regulated.