Stat5 activation inhibits prolactin-induced AP-1 activity: distinct prolactin-initiated signals in tumorigenesis dependent on cell context.

The essential role of prolactin (PRL) in normal mammary gland growth and differentiation has implicated this hormone in the development and progression of breast cancer. Although Stat5 is the best-characterized mediator of PRL signals, PRL also activates multiple other signals, whose roles in normal and pathologic processes are not well understood. We have shown that PRL stimulates activating protein-1 (AP-1) activity in breast cancer cells, and can cooperate with estradiol in this pathway. AP-1 modulates many processes critical for carcinogenesis, including cell proliferation, survival, transformation, invasion and angiogenesis, and is elevated in many neoplasms, including breast tumors. Here, we investigated the relationship between PRL signals to AP-1 and Stat5. We found that PRL activation of Stat5a and Stat5b, but not Stat1 or Stat3, reduced PRL signals to AP-1, without altering estradiol-induced AP-1 activity. The truncation mutant, Stat5/Delta53C, but not Stat5Y699F, was an effective inhibitor, consistent with a requirement for Stat5 dimerization and nuclear accumulation, but not its C-terminal transactivation activity. The association of Stat5 with AP-1 proteins suggests that this underlies the inhibition. Predictably, the ability of PRL to activate Stat5 and AP-1 was inversely related in mammary cell lines. Further, reduction of Stat5 protein with siRNA in T47D cells, which contain elevated Stat5, increased PRL-induced AP-1 signals, transcripts for the AP-1 target, matrix metalloproteinase-2 and associated invasive behavior. This study points to the importance of cell context in determining the spectrum of PRL-induced actions, which is critical for understanding the contributions of PRL to breast cancer.

Estrogen receptor-positive mammary tumorigenesis in TGFalpha transgenic mice progresses with progesterone receptor loss.

We characterized the novel NRL-transforming growth factor alpha (NRL-TGFalpha) transgenic mouse model in which growth factor - steroid receptor interactions were explored. The NRL promoter directs transgene expression to mammary ductal and alveolar cells and is nonresponsive to estrogen manipulations in vitro and in vivo. NRL-TGFalpha mice acquire proliferative hyperplasias as well as cystic and solid tumors. Quantitative transcript analysis revealed a progressive decrease in estrogen receptor alpha (ER) and progesterone receptor (PR) mRNA levels with tumorigenesis. However, ER protein was evident in all lesion types and in surrounding stromal cells using immunohistochemistry. PR protein was identified in normal epithelial cells and in very few cells of small epithelial hyperplasias, but never in stromal or tumor cells. Prophylactic ovariectomy significantly delayed tumor development and decreased incidence. Finally, while heterozygous (+/-) p53 mice did not acquire mammary lesions, p53+/- mice carrying the NRL-TGFalpha transgene developed ER negative/PR negative undifferentiated carcinomas. These data demonstrate that unregulated TGFalpha expression in the mammary gland leads to oncogenesis that is dependent on ovarian steroids early in tumorigenesis. Resulting tumors resemble a clinical phenotype of ER+/PR-, and when combined with a heterozygous p53 genotype, ER-/PR-.

High density lipoprotein utilization by dispersed rat luteal cells.

Utilization of lipoproteins by cells prepared by collagenase dispersion of ovaries of immature gonadotropin-primed rats was studied. Human and rat HDL increased basal progestin secretion and incorporation of [14C]oleate into cellular sterol esters 2-fold during a 2 h incubation, with maximal stimulation occurring at a lipoprotein sterol concentration of 125 micrograms/ml. This concentration of HDL cholesterol also increased progestin production by cells stimulated with dibutyryl cyclic AMP. Human LDL or cholesterol-rich lipid dispersions had little effect upon either progestin secretion or sterol esterification at similar sterol concentrations. However, addition of delipidated human HDL apolipoproteins to the cholesterol-rich lipid dispersions markedly enhanced progestin production. Incubation of the dispersed cells in the presence of 25 micro M ML-236B, which blocked cellular de novo sterol synthesis by over 90%, had no effect upon progestin secretion. Specific uptake of human 125I-labeled HDL by the dispersed cells was observed. Analysis fo 125I-labeled HDL uptake as a function of lipoprotein concentration indicated that the uptake process was saturated at HDL levels of 200-400 micrograms protein/ml. The amount of HDL specifically associated with the cells at saturating levels after 1 h of incubation was sufficient to account for the increased progestin synthesis and sterol ester storage observed during this time. During the incubations cell-specific degradation of the 125I-labeled HDL apolipoprotein appeared to be minimal. We conclude that lipoprotein-carried cholesterol is an important substrate for rat luteal cells and that these cells possess a specific mechanism for the uptake of HDL.

Bovine placental prolactin-related hormones.

The bovine placenta, like that of rodents and primates, synthesizes members of the PRL/GH gene family, which may assist the pituitary hormones or perform unique functions during pregnancy. Bovine placental lactogen (bPL) potentially may act through.three receptors: as an agonist or partial antagonist at the PRL and GH receptors, and via an apparently specific receptor in the endometrium. A large distinct subfamily of diverse primary structure, including bovine PRL-related protein I (bPRP-I), evidently does not act via these receptors. Advances in our understanding of hormone-receptor interactions for this gene family have provided new tools to study the role of these hormones in the successful pregnancy.

Effects of dexamethasone treatment on IL-1 receptor mRNA levels in vivo.

Using semiquantitative reverse transcriptasepolymerase chain reaction and Northern analysis, we observed in vivo up-regulation of interleukin-1 (IL-1) RI and IL-1RII mRNA levels in peripheral blood mononuclear cells (PBMCs) and neutrophils (PMNs) from Holstein cattle injected with dexamethasone (0.04 mg/kg). Baseline levels of IL-1RI mRNA were greater than IL-1RII mRNA levels in PBMCs and PMNs before dexamethasone treatment. This is in contrast with the previously reported predominance of IL-1RII in unstimulated human PMNs. IL-1RII mRNA was strongly induced in both bovine PBMCs and PMNs at 24 h and returned to baseline levels by 72 h, after dexamethasone injection. Conversely, the greatest increase in IL-1RI mRNA in PBMCs and PMNs was not detected until 72 h after dexamethasone injection. These data provide evidence for sequential in vivo up-regulation of first IL-1RII mRNA and later IL-1RI mRNA by dexamethasone that is consistent with the anti-inflammatory activity of glucocorticoids.

In situ localization of two prolactin-related messenger ribonucleic acids to binucleate cells of bovine placentomes.

The bovine fetal placenta expresses a family of PRL-related genes, consisting of the gene encoding bovine placental lactogen (bPL), and a diverse group of related genes, exemplified by bovine PRL-related cDNA I (bPRCI). bPL and the protein encoded by bPRCI are quite distinct from one another, predicting proteins only about 36% similar in amino acid sequence. To identify the cells responsible for the expression of bPL and bPRCI, in situ hybridization experiments were performed. 35S-Labeled RNA probes were prepared from the 3' regions of bPL and bPRCI and allowed to hybridize to frozen sections of bovine placentomes. Transcripts corresponding to bPL and bPRCI colocalized to fetal binucleate cells, which is consistent with the immunocytochemical localization of bPL. Signal from radiolabeled antisense strand probe was blocked by pretreatment of the sections with a 150-fold excess of unlabeled probe, but not by an excess of unlabeled probe prepared using the other cDNA as a template. This demonstrated that the signals observed were specific for bPL and bPRCI and that the two probes did not cross-hybridize at the stringency of our conditions. Radiolabeled sense strand probes yielded no signal. We conclude that the binucleate cells of the bovine placenta transcribe at least two members of the PRL gene family that may influence fetal development and maternal adaptation to pregnancy.

Characterization of a novel prolactin-related protein from bovine fetal placenta.

The bovine fetal placenta transcribes a family of PRL-related genes that are distinct from the bovine placental lactogen gene. To demonstrate that one of these cDNAs, bovine PRL-related cDNA-I (bPRCI), is expressed at the protein level, a recombinant form of the bPRCI product was overexpressed in Escherichia coli. An antiserum raised against this recombinant product did not cross-react with the pituitary members of the PRL-GH family or with bovine placental lactogen, although cross-reactivity within the bPRC subfamily cannot be ruled out. The antiserum detected a doublet with apparent mol wt of 34,000 and 35,000 in bovine fetal placenta, but not in other fetal tissues. Treatment with several glycosidases revealed a glycoprotein with asparagine-linked carbohydrates of the biantennary complex or hybrid type. We conclude that the bovine fetal placenta expresses at least one of the novel members of this gene family at the protein level.

Expression of prolactin-related hormones in the early bovine conceptus, and potential for paracrine effect on the endometrium.

Hormones related to the pituitary hormones PRL and GH are produced by the utero-placental unit of many species. In the cow, these include bovine placental lactogen (bPL) and a distantly related subfamily including the protein encoded by bovine PRL-related complementary DNA I (bPRCI). In the present studies, we defined the onset of expression of these genes in order to begin to study the regulation of their expression and function before implantation. Messenger RNA levels of both bPL and bPRCI were assessed by dot blot analysis of total RNAs prepared from conceptuses on days 15-25. Specific transcripts were not detectable at day 15 but were readily apparent beginning at day 17. To define the portions of the placenta responsible for expressing these genes, RNA was prepared from chorion (separated into cotyledons and intercotyledonary membrane), allantois, and amnion from day 58 gestation and analyzed by Northern hybridization. Transcripts for both these hormones were confined to the chorion, with intercotyledonary RNA containing at least as much as that prepared from cotyledons. To localize the product of these genes, extraembryonic membranes were immunostained with bPL and bPRCI antisera. Mononucleated trophoblastic cells stained for bPL and bPRCI at day 18; similar staining was apparent at days 23 and 24, as well as in some but not all binucleated cells. Microsomes prepared from endometrium exhibited specific binding of bPL throughout pregnancy as well as during the luteal phase of the cycle. At midgestation, these high affinity, low capacity receptors (dissociation constant of 1.27 x 10(-10) M, maximum binding capacity of 44.2 fmol/mg protein) demonstrated a selective affinity for bPL in preference to bGH and bPRL. Using this radioreceptor assay, we demonstrated the presence of bPL at nanomolar concentrations in washings from the uterine lumen, suggesting that bPL may exert paracrine effects on the endometrium.

Regulation of ovarian cholesterol metabolism: control of 3-hydroxy-3-methylglutaryl coenzyme A reductase and acyl coenzyme A:cholesterol acyltransferase.

Activities of enzymes involved in cholesterol metabolism were measured in ovaries of PMS-hCG-primed immature rats. Microsomal 3-hydroxy-3-methyglutaryl coenzyme A (HMG-CoA) reductase, a rate-limiting enzyme in sterol synthesis, was low during the period of maximal steroid production and cholesteryl ester storage. Acyl CoA:cholesterol acyltransferase (ACAT), the microsomal enzyme catalyzing sterol esterification and cytosolic cholesteryl ester hydrolase and mitochondrial cholesterol side chain cleavage activities, paralleled storage of sterol esters and levels of plasma progesterone, being highest on day 7 post hCG. Further experiments demonstrated that HMG-CoA reductase and ACAT are regulated by different mechanisms than the sterol esterase and cholesterol side chain cleavage system. When blood sterol levels were lowered by 4-aminopyrazolo (3,4-d)pyrimidine, plasma progestin concentrations fell and ovaries failed to accumulate sterol esters. HMG-CoA reductase increased more than 10-fold as a result of this treatment, while ACAT was 15% of that measured in controls. In contrast, cholesteryl esterase and the cholesterol side chain cleavage system were not affected. Intravenous infusion of human lipoproteins reversed the effects of 4-aminopyrazolo (3,4-d)pyrimidine. Raising blood levels up to 6-fold with either cholesterol-supplemented diets or intravenous injections of rat lipoproteins had no effect upon ovarian sterol metabolism, suggesting that the process by which blood cholesterol is utilized is saturated at normal blood sterol levels. Intravenous injection of LH on day 7 post hCG increased HMG-CoA reductase and decreased ACAT activity in addition to depleting cholesteryl esters. Prior treatment of rats with aminoglutethimide prevented the effects of LH on both HMG-CoA reductase and ACAT, indicating that changes in these enzymes were a consequence of the steroidogenic stimulus imposed by LH. Treatment with aminoglutethimide alone on day 7 post hCG did not further depress the already low HMG-CoA reductase activity, but ACAT and cholesterol ester storage were stimulated while sterol esterase activity was not altered. Aminoglutethimide also produced elevated ACAT activity and sterol ester storage in animals treated on day 2 post hCG, a time when there is normally little sterol ester accumulation. We conclude that HMG-CoA reductase and ACAT are regulated, in a reciprocal fashion, by ovarian cholesterol balance. When sufficient cholesterol is available, HMG-CoA reductase is suppressed and ACAT increases to facilitate esterification of sterol in excess of cell needs. When exogenous sterol supplies cannot meet ovarian demands, HMG-CoA reductase rises, and there is a concomitant decline in ACAT. Thus, gonadotropic control over these enzymes may be exerted, in part, through modulation of the supply of cholesterol to the ovary and/or through regulation of the rate of sterol utilization for hormone synthesis.

Prolactin receptor heterogeneity in bovine fetal and maternal tissues.

Study of diverse PRL actions at a variety of fetal and maternal targets during pregnancy is complicated by receptor heterogeneity and multiple ligands circulating at this time. In the present studies, we have examined PRL receptors at a variety of potential targets by reverse transcription-PCR and Western analysis. Bovine tissues contain two different transcripts for the PRL receptor; the one that encodes a short form includes an additional 39 bases at a position identical to the deviation from the long form found in rodents and sheep. Western analyses of PRL receptors in microsomal fractions from various maternal and fetal tissues revealed considerable size heterogeneity. Collectively, the larger immunoreactive moieties (apparent Mr 100 kDa or greater) and the smaller species (47-55 kDa) correlated well with the relative abundance of the transcripts for the different forms of the receptor and varied considerably among tissues. N-Glycosylation was shown to be the major, but not the only, modification of both receptor forms when transiently transfected into COS-7 and END-6.2 cells. Much of the short form could be reduced to the mobility predicted from the complementary DNA by culture with tunicamycin; this was not true of the long form, suggesting modifications specific for its cytoplasmic domain. Differences in the pattern of immunoreactive species in the COS-7 and END-6.2 cells are consistent with cell-specific modifications. The ability of these receptor forms to mediate a transcriptional response to PRL and its placental relative, placental lactogen, was evaluated with a PRL response element inserted upstream from a thymidine kinase promoter/reporter gene construct transiently transfected into CHO-K1 cells. Both hormones were able to stimulate reporter gene expression through the long form, but not the short form, of the receptor. These studies will facilitate examination of tissue-specific actions of PRL and related hormones during pregnancy.

Molecular cloning and nucleotide sequence of a bovine alpha-lactalbumin cDNA.

A cDNA clone for the bovine milk protein, alpha-lactalbumin (alpha LA), has been identified using a rat cDNA probe. The bovine cDNA clone is 703 nucleotides (nt) long, contains 8 nt of 5'-untranslated sequence and 269 nt of 3'-untranslated sequence. When compared with previously reported sequences, the bovine alpha LA mRNA sequence has 74% similarity with rat alpha LA mRNA, 79% similarity with human mRNA and 74% similarity with guinea pig mRNA.

Long and short forms of the ovine prolactin receptor: cDNA cloning and genomic analysis reveal that the two forms arise by different alternative splicing mechanisms in ruminants and in rodents.

Prolactin is a pituitary hormone that binds to specific receptors in numerous tissues. Depending on the size of their cytoplasmic domain, long and short prolactin receptors (l-PR, s-PR) have been described. Up to now, s-PR were found in rodents only. We report here the cloning of full-length coding sequences for short and long ovine prolactin receptors (s-oPR, l-oPR). The only difference between s- and l-oPR coding sequences was, respectively, the presence or absence of a 39 base pair insert at the beginning of the cytoplasmic domain, with two contiguous inframe stop codons at its 3' end. Sequence comparison revealed that the alternative splicing producing s- and l-oPR was different from that of rodents, although the resulting proteins were very similar. PCR experiments on ovine genomic DNA showed that the 39 base pair insert was directly linked to the downstream exon, and separated from the upstream exon by an 800 base pair intron. Thus, the alternative splicing used a single intron with one 5' and two 3' sites. The same organization was found in bovine and caprine genomes, suggesting that this feature is general in ruminants and different from rodents, which use mutually exclusive exons to produce s-PR and l-PR.

SmMAK16, the Schistosoma mansoni homologue of MAK16 from yeast, targets protein transport to the nucleolus.

The SmMAK16 gene from Schistosoma mansoni was cloned by chance when an adult worm cDNA library was probed with antiserum to affinity-purified S. mansoni GSH S-transferases. SmMAK16 encodes a hydrophilic protein of 259 amino acids with a molecular mass of 31 kDa. The protein shares 43% sequence identity and 66% similarity to the nuclear protein MAK16 of Saccharomyces cerevisiae that has been implicated both in cell cycle progression and biogenesis of 60S ribosomal subunits. Both proteins display a similar degree of sequence similar to the hypothetical protein CeMAK16 from Caenorhabditis elegans. These proteins share a number of apparent protein motifs, including two nuclear localization signals (NLS), multiple sites for phosphorylation by protein kinase CK2 and four conserved cysteine residues that resemble a zinc binding domain. SmMAK16 mRNA is more highly expressed in adult female worm than males. Recombinant SmMAK16 was phosphorylated by human protein kinase CK2. When chimeric constructs containing SmMAK16 fused the green fluorescent protein (GFP) were transiently transfected into COS-7s cells, the reporter was localized not in nuclei, but exclusively in nucleoli. The yeast and nematode homologues were likewise able to direct nucleolar accumulation of the fluorescent reporter. The high degree of sequence conservation together with the ability to direct nucleolar protein transport supports the hypothesis that MAK16 proteins play a key role in the biogenesis of 60S subunits.

Effect of luteinizing hormone of the lipid composition of rat ovaries.

The lipid composition of immature rat ovaries was examined after induction of ovulation with pregnant mare serum gonadotrophin and human chorionic gonadotrophin and subsequent (7--8 days later) stimulation with 10 micgogram LH. Two hours after the administration of LH, there was a decrease of approximately 50% in the concentration of cholesteryl esters in the ovary. The percentages (by weight) of sterol esters containing stearate, linoleate, eicosatrienoate and arachidonate were reduced by LH treatment, whereas the percentage of the C24:4 acid increased. No changes were observed in either the concentrations or fatty acid composition of phospholipids and triglycerides. These observations suggest that the metabolism of cholesteryl esters is acutely affected by LH and that sterol esters bearing long-chain polyunsaturated fatty acids are preferentially mobilized. Liberation of these unsaturated fatty acyl moieties may have significant effects on metabolism in the ovarian cell.

Purification and properties of placental prolactin-related protein-I.

We used sucrose density gradient centrifugation, size exclusion chromatography, and high-pressure reversed-phase chromatography in the purification of bovine prolactin-related protein-I (bPRP-I) to homogeneity from a secretory granule-enriched fraction of fetal cotyledon. Amino terminal sequence was unambiguous, consistent with the nucleic acid sequence of the cDNA 50 codons distal to the initial AUG in the open reading frame, and began with the residues: RKSFTDRFMNAASLSHDFY. This is distinct from the signal peptide cleavage site predicted by the algorithm of von Heijne (1986) as well as that expected by comparison with other members of the growth hormone/prolactin family of hormones. The level of bPRP-I in uterine fluid was sufficient to detect by Western blot of unfractionated material and estimated as at least 0.65 microM. In contrast, bPRP-I was undetectable in the serum by this method. Interaction of [125I]-bPRP-I with high molecular weight serum components interfered with its measurement by radio-immunoassay, and could be replicated with purified alpha 2-macroglobulin with an apparent KD of about 0.41 microM. Thus, the bPRP-I gene product is processed secreted and distributed in a manner consistent with a paracrine action at the materno-fetal interface.

Molecular cloning of the bovine prolactin receptor and distribution of prolactin and growth hormone receptor transcripts in fetal and utero-placental tissues.

We have isolated a bovine prolactin (bPRL) receptor cDNA from an endometrial cDNA library, which predicts a 557 amino acid transmembrane protein similar to the long forms of other characterized prolactin receptors. The predicted cytoplasmic domain is slightly truncated primarily by a stop codon located 36 codons 5' from the stop utilized in the human hepatic transcript. When expressed in COS cells, this cDNA was shown to encode a protein which bound bovine placental lactogen (bPL) and bPRL with nearly equal affinity (KD for bPL, 2.03 x 10(-10) M; bPRL, 3.07 x 10(-10) M). Northern analysis demonstrated multiple transcripts, with maternal liver, corpus luteum, intestine, endometrium and fetal liver containing a major transcript of about 3.8 kb, and maternal corpus luteum and endometrium, a second sized transcript of apparently equal abundance of 4.4 kb. This difference did not appear to be within the coding region. Primer extension analysis of maternal hepatic and endometrial transcripts revealed considerable heterogeneity. Examination of the distribution of prolactin and growth hormone receptor transcripts at mid-pregnancy by semi-quantitative reverse transcriptase polymerase chain reaction showed that both are widespread in bovine fetal and placental tissues. This isolation of bovine prolactin receptor cDNA, and description of receptor distribution will facilitate study of the action of the placental and pituitary members of this gene family during pregnancy.

Regulation of interleukin (IL)-1alpha, IL-1beta, and IL-6 expression by growth hormone and prolactin in bovine thymic stromal cells.

Growth hormone (GH) and prolactin (PRL) have been implicated in T-cell development, but relatively little is known about the mechanism(s) of their actions on the multiple cell types in this complex tissue. Here, we investigated the effects of GH and PRL on the expression of interleukin (IL)-1alpha, IL-1beta and IL-6 in thymic stromal cells (TSC). These cytokine mRNAs were increased by GH, PRL and placental lactogen (PL) in primary cultures prepared from mid-gestational fetuses in a dose-dependent manner. IL-1 receptor antagonist (IL-1ra) abolished the hormone-induced IL-6 expression, suggesting that the induction of IL-6 was secondary to IL-1 activity. To examine the effects of these hormones on an individual cell type and develop a system in which signalling mechanisms can be studied, we generated immortalized cell lines using a strategy of conditional transformation. In the cell line, TSC-936, which displayed vimentin-positive staining and morphological characteristics of mesenchymal cells, both GH and PRL increased levels of steady-state mRNAs for IL-1alpha and IL-1beta. Nuclear run-on analysis revealed that the transcription rate of the IL-1beta gene was significantly increased by GH and PRL at 30 and 60 min, respectively, but that for IL-1alpha was not significantly changed, suggesting the possibility of an alternative mechanism mediating this response. These data suggest that modulation of cytokine gene expression is one mechanism by which GH and PRL facilitate thymic development and T-cell maturation.

Growth hormone receptor and regulation of gene expression in fetal lymphoid cells.

The role of growth hormone (GH) in modulating the adult immune response is receiving increased attention; however, its role in the development of immune competence in the fetus has not been defined. In order to begin to address the role of GH in the ontogeny of the immune response. cells from bovine fetal spleen and thymus were examined for GH receptor and responsiveness to GH. Northern analysis and ligand binding studies showed that growth hormone receptor (GHR) was readily detected in early- and mid-gestational fetal thymocytes, but it was less readily detected in thymocytes from older fetuses. In contrast, GHR was easily detected in splenocytes at all fetal ages. Thymocytes and splenocytes from mid-gestational fetuses expressed low levels of cell surface GHR by flow cytofluorometric analysis, and CD4+ and CD8 (single positive) thymocyte subsets were positive. Northern analyses were employed to determine the effects of in vitro GH treatment on expression of several proto-oncogenes, cytokines, and GHR in thymocytes from fetuses at approximately mid-gestation. GH treatment for 30 min down-regulated c-jun and c-fos mRNA approximately 2- and 2.8-fold, respectively. After 6 h treatment, GH increased transcript levels for interleukin (IL)-1alpha, IL-1beta, IL-6, and GM-CSF about 2.5-, 2.2-, 3-, and 2-fold, respectively. GH also down-regulated the expression of its own receptor about 3.2-fold after 8 h of incubation. The presence of GHR in fetal lymphoid cells and its temporal and spatial regulation suggest a potential role for GH in the development and or function of the fetal bovine immune system. Although the mechanism(s) is unclear, our results suggest that GH is intimately involved in lymphocyte function and expression of certain cytokines during a critical period of fetal immune development.

Dysregulation of mammary Stats 1,3 and 5 and PRL receptors by overexpression of TGFalpha.

Mammary TGFalpha overexpression results in delayed involution and eventually mammary cancer in transgenic mice. We hypothesized that STATs and PRL receptors (PRLR), critical regulators of mammary function, are altered in these animals and may contribute to this phenotype. We examined these factors late in the first pregnancy (d.18) and during normal involution (d.4 post-lactation) in WAP-TGFalpha transgenic mice and non-transgenic controls. Long form PRLR mRNA in WAP-TGFalpha glands at both pregnant d.18 and d.4 post-lactation was significantly reduced compared to controls, and PRLR-S3 failed to rise during involution. Total and pTyr STAT 1,3,5a and 5b also were altered. STAT 3 was higher at both times in WAP-TGFalpha glands. STAT 5a and 5b were lower at late pregnancy, but higher post-lactation; however, pTyr(694) STAT 5 was abnormally low at both times. Thus overexpression of TGFalpha has direct or indirect effects on both STATs and PRL responsiveness in vivo, which may reflect mechanisms of TGFalpha-induced mammary epithelial abnormalities.

Structure of the bovine placental lactogen gene and alternative splicing of transcripts.

Preliminary evidence for heterogeneity among bovine placental lactogen (bPL) transcripts prompted characterization of additional cDNA clones and isolation of the bPL gene. Nucleotide replacements were detected among sequenced cDNAs isolated from different animals at a total of 11 positions. Four of these predict amino acid substitutions, which are generally conservative in nature. In addition, truncated forms of bPL are predicted by the sequences of two cDNAs in which alternative splicing is evident. In one case, exon III is deleted with no effect on reading frame. However, in the other instance, a shifted reading frame resulting in a novel carboxyl terminus is generated by use of an alternative 5' splice donor site within exon IV. Nuclease protection analysis demonstrated that these variant transcripts comprise about 13% of the total bPL mRNA present in the midgestation placenta. Characterization of the bPL gene revealed that it is similar in structure to other members of this gene family. Sequence analysis demonstrated that the 5'-flanking region of the bPL gene has diverged considerably from the bovine prolactin and growth hormone genes, but shares homology with the previously characterized gene corresponding to bovine prolactin-related cDNAI (bPRCI). Primer extension as well as nuclease protection analysis indicated that a single transcription start site was utilized in the fetal placenta at midgestation. Exact matches to the consensus sequences for response elements for thyroid hormone and transcription factor AP-2 were located 50 and 70 bp, respectively, upstream from the transcription start site in cloned genomic 5'-flanking sequences. We conclude that the bovine placenta may express more than a single placental lactogen product, raising the possibility of alternative hormones with distinct biological properties, and that the bPL gene may share regulatory elements with the gene for the distinct prolactin-related product, bPRCI, based on similarities in the 5' regions of the corresponding genes.