Generation and initial characterization of the prolactin-inducible protein (PIP) null mouse: accompanying global changes in gene expression in the submandibular gland.

The human prolactin-inducible protein/gross cystic disease fluid protein-15 (hPIP/GCDFP-15) is a secretory glycoprotein found primarily in apocrine tissues including the breast and salivary glands. With largely unknown functions, PIP has been implicated in breast cancer and metastasis, host defense processes and T lymphocyte apoptosis. To begin to address PIP function in vivo, we generated the PIP null mouse (Pip-/-). Additionally, to determine the effect of the loss of PIP on gene expression and to gain insight into some of the molecular mechanisms underlying PIP function, microarray analysis of the submandibular gland was also undertaken. Pip-/- mice developed normally with no overt differences in behaviour or gross morphology and were fertile. However, histological examination of 3-month-old Pip-/- mice sometimes showed enlarged submandibular lymph nodes, lymphocytic aggregations within the prostate lobes, and enlarged medulla in the thymus. Functional analysis of gene expression revealed sets of multiple differentially expressed genes associated with cell death and survival, lipid metabolism, inflammation, immune disease, and cancer, as a consequence of mPIP abrogation. Taken together, these studies lend support to an immunomodulatory role for PIP in vivo and provide further insights into potentially novel signaling pathways and regulatory networks for PIP.

Elevated expression of proprotein convertases alters breast cancer cell growth in response to estrogen and tamoxifen.

Two proprotein convertase cDNAs, PC1 and furin, were stably transfected into the human breast cancer cell line MCF-7. The PC1 or furin over-expressing cells possessed an altered morphology. When grown in vitro in a serum-free medium, the population doubling time of the convertase-transfected cells was twice that of wild-type (WT) cells. High concentrations of estradiol stimulated the growth of all three cell types to a similar extent; however, at low concentrations of estradiol, the convertase-transfected cells grew more slowly than WT cells. In athymic nude mice implanted with 5 mg estradiol pellets, the growth of tumors of convertase-transfected MCF-7 cells was stimulated to a degree similar to that of WT MCF-7 tumors. However, in mice implanted with lower-dose (1.5 mg) estradiol pellets, the tumors of PC1- or furin-transfected MCF-7 cells grew approximately five times slower than those of WT MCF-7 cells. In mice implanted with tamoxifen pellets, tumors of PC1- or furin-transfected MCF-7 cells regressed approximately five times slower than the WT tumors. This study shows that the over-expression of proprotein convertases confers a greater estrogen dependency and anti-estrogen resistance on human breast cancer cells.

Estrogen receptors alpha and beta in rat decidua cells: cell-specific expression and differential regulation by steroid hormones and prolactin.

Estradiol is known to play an important role in the growth and differentiation of rat uterine stromal cells into decidual cells. In particular, this hormone with progesterone is necessary for blastocyst implantation and subsequent decidualization in the rat. Although binding experiments have demonstrated the presence of estrogen-binding sites, no evidence exists as to whether the rat decidua expresses both isoforms of the estrogen receptor (ER), alpha and beta. In this investigation, we analyzed the expression of decidual ERalpha and ERbeta, studied their regulation by PRL and steroid hormones and examined the ability of decidual ERp to transduce the estradiol signal to the progesterone receptor. Immunocytochemistry, RT-PCR, and Northern blot analysis showed that both ER species are coexpressed in the decidua during pseudopregnancy. Interestingly, these genes were preferentially found in a cell population localized in the antimesometrial site of the uterus where blastocyst implantation takes place. Using decidual cells in primary culture obtained from pseudopregnant rats and a decidua-derived cell line (GG-AD), we show a differential regulation of ERalpha and ERbeta by PRL and ovarian steroid hormones. Whereas PRL, estradiol, and progesterone all increased ERbeta messenger RNA (mRNA) expression in a dose-dependent manner, only PRL up-regulated the mRNA levels of ERa. Estradiol had no effect on ERalpha expression, whereas progesterone markedly decreased its mRNA levels. Interestingly, progesterone, which up-regulates the ability of PRL to signal to a PRL-regulated gene in mammary-gland derived cells, prevented PRL stimulation of decidual ERalpha and had no synergistic effect on ERbeta expression. The use of GG-AD cells, which express only ERbeta, allowed us to demonstrate that this receptor subtype is functional and transduces estradiol signal to the progesterone receptor. In summary, the results of this investigation have revealed that ERbeta is expressed in addition to ERalpha in the rat decidua, and that the expression of both ERs are cell specific and differentially regulated by PRL and steroids. One salient finding of this investigation is that progesterone down-regulates ERalpha, but concomitantly increases the expression of a functional ERbeta that mediates estradiol up-regulation of the decidual progesterone receptor.

Regulation of PKC delta expression by estrogen and rat placental lactogen-1 in luteinized rat ovarian granulosa cells.

Protein kinase C (PKC) delta is dramatically upregulated in the corpus luteum in the second half of pregnancy in the rat. To gain insight into the hormonal regulation of PKC delta expression, studies were undertaken to analyze the regulation of PKC delta expression in a luteinized rat granulosa cell model. PKC delta protein expression was evaluated in luteinized granulosa cells, isolated from human (h)CG-treated immature female rats 7 h after the injection of an ovulatory dose of hCG and cultured up to 12 days. Cytochrome P450 cholesterol side chain cleavage enzyme expression was observed throughout the culture period, and a majority of the cells expressed steroidogenic acute regulatory protein and responded to rat placental lactogen (rPL)-1 by exhibiting hypertrophy, consistent with maintenance of the luteal phenotype. Both PKC delta protein and mRNA expression increased 3.5-4-fold with time of culture, and PKC delta mRNA expression could be eliminated by treatment of cells with the PKC inhibitor GF109203X. E(2) caused a specific dose- and time-dependent increase in expression of PKC delta protein of twofold, whereas PKC delta mRNA was unaffected by E(2) over a 12-day culture period. Treatment of cells with 500 ng/ml rPL-1 for the final 4 days of a 12-day culture in the absence of E(2) had no effect on PKC delta protein or mRNA expression, while treatment with 500 or 3000 ng/ml rPL-1 in the presence of E(2) significantly enhanced both PKC delta protein and mRNA expression (up to threefold). These results show that two of the major regulators of luteal function in the second half of pregnancy in the rat, E(2) and rPL-1, cooperate to regulate PKC delta expression in luteinized granulosa cells.

Induction of relaxin messenger RNA expression in response to prolactin receptor activation requires protein kinase C delta signaling.

The ability of PRL or rat placental lactogen (rPL)-1 to induce relaxin mRNA expression was analyzed in a luteinized rat granulosa cell culture model. PRL receptor activation induced relaxin mRNA expression in a concentration- and time-dependent manner. High concentrations of PRL receptor agonist, equivalent to those of the second half of pregnancy in rats, were required to elicit relaxin mRNA expression. A 40-fold induction of relaxin mRNA was observed in cells treated 24 h with 1 microg/ml of rPL-1. Estrogen enhanced relaxin expression induced by PRL but did not affect relaxin expression on its own. PRL/rPL-1 induction of relaxin expression was independent of the extracellular regulated kinase (ERK) members of the mitogen-activated protein kinase (MAPK) pathway, based on the inability of the ERK kinase inhibitor PD98059 to block induction of relaxin expression. PRL/rPL-1 induction of relaxin expression required protein kinase C (PKC) delta, based on the ability of the preferential PKC delta inhibitor rottlerin to abolish induction of relaxin expression. Direct activation of PKC by phorbol myristate acetate, however, was not sufficient to promote induction of relaxin mRNA expression. Stats (signal transducers and activators of transcription) 3 and 5 DNA binding activities were induced by PRL/rPL-1 treatment of luteinized granulosa cells but only Stat 3 DNA binding was reduced by rottlerin. PRL/rPL-1 treatment of luteinized granulosa cells resulted in increased phosphorylation on tyrosine-705 and serine-727 of Stat 3, and these responses were reduced and blocked, respectively, by rottlerin. Tyrosine and serine phosphorylations of Stat 3 in the corpus luteum were also increased in the second half of pregnancy when PL levels are highest. Stat 3, but not Stat 1 or 5, coimmunoprecipitated with luteal PKC delta during pregnancy; Stat 3 transiently coimmunoprecipitated with PKC delta from luteinized granulosa cells in response to PRL receptor activation; and Stat 3/PKC delta complex formation required PKC delta kinase activity. Taken together, these results show that PKC delta is obligatory for PRL/rPL-1-dependent relaxin expression, that PKC delta complexes with Stat 3 in response to PRL receptor activation, and that PKC delta is involved in the regulation of Stat 3 phosphorylation downstream of the PRL receptor. These results demonstrate that PRL/rPL-1 promotes relaxin expression in luteal cells and that this event is mediated, at least in part, via PKC delta.

The potential role for prolactin-inducible protein (PIP) as a marker of human breast cancer micrometastasis.

The prolactin-inducible protein (PIP/GCPD15) is believed to originate from a limited set of tissues, including breast and salivary glands, and has been applied as a clinical marker for the diagnosis of metastatic tumours of unknown origin. We have investigated the potential role of PIP mRNA as a marker of human breast cancer metastasis. Using reverse transcription polymerase chain reaction and Southern or dot blot analysis, PIP mRNA was detected in 4/6 breast cell lines, independent of oestrogen receptor (ER) status. In breast primary tumours (n = 97), analysed from histologically characterized sections, PIP mRNA was detected in most cases. Higher PIP mRNA levels correlated with ER+ (P = 0.0004), progesterone receptor positive (PR+) (P = 0.0167), low-grade (P = 0.0195) tumours, and also PIP protein levels assessed by immunohistochemistry (n = 19, P = 0.0319). PIP mRNA expression was also detectable in 11/16 (69%) of axillary node metastases. PIP mRNA expression, however, was also detected in normal breast duct epithelium, skin, salivary gland and peripheral blood leucocyte samples from normal individuals. We conclude that PIP mRNA is frequently expressed in both primary human breast tumours and nodal metastases. However, the presence of PIP expression in skin creates a potential source of contamination in venepuncture samples that should be considered in its application as a marker for breast tumour micrometastases.

Selective activation of the probasin androgen-responsive region by steroid hormones.

Glucocorticoid and androgen receptors have been shown to function through the same palindromic glucocorticoid response element (GRE) and yet have differential effects on gene transcription. In this study, we examined the functional and structural relationship of the androgen and glucocorticoid receptors with the androgen responsive region (ARR) of the probasin (PB) gene containing two androgen receptor binding sites, ARBS-1 and ARBS-2. Transfection studies indicated that one copy of each cis-acting DNA element was essential for maximal androgen-induced chloramphenicol acetyltransferase (CAT) activity and that androgen selectivity was maintained when multiple copies of the minimal wild type (wt) androgen responsive region containing both ARBS-1 and ARBS-2 (-244 to -96) were subcloned in front of the thymidine kinase promoter. Furthermore, replacing the androgen response region with 1, 2 or 3 copies of either ARBS-1 or ARBS-2 restored less than 4% of the biological activity seen with the wt PB ARR. Multiple copies of either ARBS-1 or ARBS-2 did not result in glucocorticoid-induced CAT gene activity. By comparison, 1 or 2 copies of the tyrosine aminotransferase (TAT) GRE, as well as the mouse mammary tumour virus GRE, were strong inducers of CAT activity in response to both androgen and glucocorticoid treatment. In addition, band shift assays demonstrated that although the synthetic glucocorticoid receptor, GR-DNA binding domain (GR-DBD), and the synthetic androgen receptor, AR2, could interact with the TAT GRE (dissociation constants Kd of 63.9 and 14.1 respectively), only AR2 but not GR-DBD binding could be detected on ARBS-1 and ARBS-2. Our findings provide further evidence that androgen-induced regulation of gene transcription can occur through androgen-specific DNA binding sites that are distinct from the common GRE.

Influence of rat placental lactogen-I on the development of whole rat embryos in culture.

Rat placental lactogen-I (rPL-I), the first prolactin-like hormone expressed in the placenta during pregnancy in the rat, is known to influence maternal functions. In the present study, we have investigated the effects of rPL-I on the growth and development of cultured whole rat embryos. Rat embryos, with or without ectoplacental cone (EPC) attached, were explanted at day 9 of gestation. After 48 h of culture, the embryos, enclosed by the yolk sacs, were assessed by the presence of visible heart contractions ('heart beats'), crown-rump length (CRL) and yolk sac diameter (YSD). When intact embryos with EPC were cultured, the concentrations of rPL-I and rPL-II (products of EPC) in the medium were 850+/-841 and 92+/-181 ng/ml respectively (means+/-s.e.m.). In embryo cultures with the EPC removed, rPL-I levels decreased to

Analysis of tissue- and hormone-specific regulation of the human prolactin-inducible protein/gross cystic disease fluid protein-15 gene in transgenic mice.

The human prolactin-inducible protein/gross cystic disease fluid protein-15 (PIP/GCDFP-15) gene is expressed in more than 90% of human breast cancer biopsies but not in the normal mammary gland. However, it is expressed in several normal human apocrine glands such as the lacrimal and salivary glands. In human breast cancer cell lines, the gene is regulated by a number of hormones including androgen and prolactin. It is not known whether gene expression in normal tissues is under similar hormonal control. To understand the mechanisms by which hormone- and tissue-specific expression of the human PIP/GCDFP-15 gene are regulated in vivo, we generated transgenic mice using a 13.7 kb genomic DNA fragment containing the entire 7 kb human gene, together with 2.9 kilobases of 5' and 3.8 kilobases of 3' flanking sequences. The human PIP/GCDFP-15 transgene was found to be expressed in both the lacrimal and salivary glands but was not expressed in the mammary glands of transgenic mice. This tissue-specific pattern of the transgene expression in the mouse was very similar to that of the endogenous human PIP/GCDFP-15 gene, and to the endogenous mouse,gene. In the mouse salivary glands, the transgene expression was highest in the parotid, considerably less in the submaxillary (submandibular) and absent in the sublingual glands. In the mouse lacrimal gland, as in the human breast cancer cell lines, the human PIP/GCDFP-15 transgene was also up-regulated by androgen. These studies demonstrate that the human gene with its 6.3 kb flanking sequences is able to confer gene expression in vivo in a tissue-specific and hormone-responsive manner.

Pro-protein convertase gene expression in human breast cancer.

As a first step towards elucidating the role that pro-protein convertases play in the growth regulation of breast cancer, we studied the gene expression of 6 known human convertase members (PC1/PC3, PC2, furin/PACE, PACE4, PC5/PC6 and PC7/LPC) in human breast cancer tumors and cell lines. PC1, furin, PACE4 and PC7 mRNAs were detected by reverse transcriptase-polymerase chain reaction (RT-PCR) amplification in all 7 human breast cancer cell lines and 30 breast tumor tissues tested. PC5 expression was detected in 2/30 tumor tissues. PC2 mRNA, however, was not detected. In situ hybridization localized furin mRNA to the tumor cells; adjacent fibrous stroma and blood vessel elements were negative for furin gene expression. Thirty breast tumors with varying quantities of estrogen and progesterone receptors were assayed for furin, PACE4 and PC1 mRNAs by quantitative RT-PCR, and 22 tumors were assayed for PC7 mRNA. An apparent association was observed only between PACE4 and estrogen receptors. No statistically significant correlation was found between the levels of steroid receptors and the expression of human furin, PCI and PC7 genes. Convertase mRNA levels appeared similar in both the estrogen-responsive and -unresponsive breast cancer cell lines. Also, proprotein convertase mRNAs were not detected in 9 histologically normal human breast tissues. These results suggest that elevated expression of some members of the pro-protein convertase gene family is a characteristic of human breast cancer, an event which may be important for human breast tumorigenesis.

Central lactogenic regulation of maternal behavior in rats: steroid dependence, hormone specificity, and behavioral potencies of rat prolactin and rat placental lactogen I.

Adult virgin female rats display maternal behavior when continuously exposed to foster young for 5-6 days. Central infusions of PRL or placental lactogens (PLs) together with systemic treatment of progesterone (P) and estradiol (E2) stimulate maternal behavior in 1-2 days. In the present set of studies, it was asked whether the actions of lactogenic hormones are dependent upon both E2 and P and specific to lactogenic molecules. Moreover, we wanted to know whether central infusions of rat (r) PRL and PLs were equally effective in inducing maternal behavior. In the first study, adult virgin rats were ovariectomized (ovx) and stereotaxically fitted with bilateral cannulas directed at the medial preoptic area (MPOA). Rats were then assigned to one of four groups: P plus E2, blank (B) plus E2, P plus B, and B plus B. P-filled or B capsules were implanted sc on treatment day 1 and removed on day 11, whereas E2 or B capsules were implanted on day 11. All groups were infused with rPRL (40 ng/side) five times from days 11-13 and injected with bromocriptine (CB-154) sc (days 11-17) to suppress endogenous PRL release. Behavioral testing was conducted daily from days 12-17. It was found that exposure to both P and E2 was necessary to induce a fast onset of maternal behavior in PRL-infused females; priming with P or E2 alone in PRL-treated rats failed to stimulate a fast onset of behavior relative to that in nonsteroid-treated controls. In the second experiment to determine the biochemical specificity of PRL's action, adult nulliparous rats were ovx, implanted with bilateral cannulas directed at the MPOA, treated with both P and E2, injected with CB-154, and infused centrally (five times) with 40 ng (per side) of bovine GH, ovine LH, or vehicle. Central infusions of either bovine GH or ovine LH failed to stimulate maternal behavior, suggesting that the stimulatory actions of PRL are related to its lactogenic properties. In the final study, rats were ovx, fitted with bilateral cannulas directed at the MPOA; treated with P, E2, and CB-154; and given a single set of bilateral infusions of rPL-I or rPRL (40 ng/side.infusion) on day 11, three sets of infusions of rPL-I or rPRL (days 11 and 12), or vehicle infusions. Rats given three infusions of rPL-I and rPRL responded faster than controls, although the effect was not as robust as that in animals given five infusions in the initial study. rPL-I and rPRL groups did not differ from one another. Together these studies indicate that 1) both P and E2 are required for lactogenic stimulation of maternal behavior; 2) the stimulatory actions of PRL and rPLs on maternal behavior are related to their lactogenic properties; 3) extended treatment of females with lactogenic hormones is more effective in stimulating the onset of maternal behavior; and 4) the neural potencies of rPRL and rPL-I are similar. These findings provide support for the idea that the induction of maternal behavior is stimulated by the central actions of lactogenic hormones.

Rat placental lactogen-I variant (rPL-Iv), product of an unique gene, is biologically different from rPL-I.

The rat placenta expresses two rat placental lactogen-I (rPL-I) proteins: the normal rPL-I in the first half of pregnancy and a variant (rPL-Iv) in the second half of pregnancy. They are 70% identical at the amino acid level but arise from different cell types: rPL-I from giant cells and rPL-Iv from cytotrophoblasts. To assess whether rPL-Iv originates from alternative splicing of the rPL-I gene or is the product of a separate gene, genomic clones of rPL-I and rPL-Iv were isolated from a lambda EMBL-3 rat genomic library. Restriction enzyme analysis of the 14-kilobase full-length genomic clones of rPL-I and rPL-Iv indicated that the two genes are distinct. To assess the biological activity of the variant protein relative to other members of the rat PL/PRL/GH family, two expression systems were chosen to obtain the purified recombinant protein: 1) a secreted form of rPL-Iv was obtained from Chinese hamster ovary (CHO) cells transfected with rPL-Iv-complementary DNA; and 2) a rPL-Iv fusion protein (Bac-rPL-Iv) was obtained from Spodoptera frugiperda (Sf9) insect cells that had been infected with a recombinant baculovirus generated from rPL-Iv complementary DNA. An antibody was generated to the purified Bac-rPL-Iv fusion protein and used for affinity chromatography to purify recombinant rPL-Iv from the CHO cell media. The mitogenic activity of rPL-Iv was monitored in the Nb2 lymphoma cell bioassay. The relative potency of rPL-Iv compared with other members of the PL/PRL/GH family follows: ovine PRL 100, rPL-I 200, rPL-II 160, rPRL 40, CHO-rPL-Iv 0.7, and Bac rPL-Iv 0.4. Iodinated CHO-rPL-Iv showed minimal binding to Nb2 lymphoma cells, but at a 500-fold protein concentration rPL-I was able to displace [125I]rPL-I from the lymphoma cell PRL receptor. Using recombinant CHO-derived rPL-Iv as standard and antisera against the Bac-rPL-Iv fusion protein, a RIA was developed for rPL-Iv. In pregnant rats rPL-Iv appeared in the serum at day 14, rising to a peak of 2080 +/- 440 ng/ml at day 18, followed by a slight decline. These values reflect the levels of messenger RNA for rPL-Iv in rat placenta noted previously. In summary, rPL-Iv arises from a gene different from rPL-I and the rPL-I protein displays minimal binding and mitogenic activity in the Nb2 lymphoma cells.

Endocrine communication between conceptus and mother: placental lactogen stimulation of maternal behavior.

The possible role of the conceptus in stimulating the onset of maternal behavior through its secretion of placental lactogens and their passage into the brain was investigated in female rats. In the first study, significant mitogenic activity in the Nb2 lymphoma cell bioassay was detected in cerebrospinal fluid (CSF) samples collected by push-pull perfusion from rats on days 12-21 of pregnancy, coincident with the establishment of placental function. In contrast, mitogenic activity was absent from CSF in lactating and gonadectomized, virgin females. In a second study the mitogenic activity in day 12 pregnant samples was neutralized 71% with antibodies to rat placental lactogen-I (rPL-I) and > 90% with a combination of antibodies to rPL-I plus rPL-II. In contrast, activity on day 21 of pregnancy, 1 day prepartum, was reduced by antibodies to rPL-II (> 85%), but not by antibodies to rPL-I, indicating that the predominant lactogen in the CSF prepartum is rPL-II. The behavioral actions of placental secretions were assessed in the third experiment by infusing recombinant rPL-I and purified rPL-II directly into the medial preoptic area of the brain of steroid-primed, nulliparous rats. Latencies to respond maternally to foster young were significantly reduced in rPL-I- and rPL-II-treated rats (2- to 3-day latencies) when compared with latencies in control females (5- to 6-day latencies). Thus, the conceptus through its secretion of rPLs which apparently gain access to the CSF helps to prime the pregnant female's brain to respond maternally at the end of gestation. This endocrine communication between the developing conceptus and pregnant female appears to be an important part of the biological system which helps to establish successful maternal care.

A new member of the hormonally regulated rodent submaxillary gland glycoprotein gene family: cDNA cloning and tissue specific expression.

Polymerase chain reaction was used to amplify and identify two related rat submaxillary gland glycoprotein (rSMGGP and rSMGGP1) cDNAs. They were 489 bp and 594 bp long respectively. The shorter cDNA (rSMGGP) was identical to the previously published rat spot-I protein. The longer cDNA (rSMGGP1) had an additional (117 bp) unique nucleotide sequence in the 3' coding region, and the overall homology between the two cDNAs was 78%. rSMGGP also had a 68% homology to the mouse submaxillary gland glycoprotein (mSMGGP) cDNA. The predicted translated product of rSMGGP1 was 130 amino acids long, 39 amino acids longer than the rSMGGP. The region of greatest diversity between the putative peptides of the two rat cDNAs and the mouse cDNA was in the carboxy terminus. Northern blot analysis, using both rat cDNAs as probes, showed hybridization to an mRNA transcript (650 bases) in the submaxillary and lacrimal gland of the normal adult male and female rat. A larger transcript (approximately 700 bases) was induced under conditions of altered hormonal profiles: hypophysectomy, pregnancy/lactation, and castration. Dihydrotestosterone administration inhibited expression of the two transcripts in both the lacrimal and submaxillary glands of male and female rats. The labelled 117 bp DNA fragment unique to the rSMGGP1 cDNA hybridized only to the 700 base transcript in the rat lacrimal and submaxillary gland suggesting that differential exon usage produces the two variant mRNAs. The regulation of the SMGGP gene expression may provide yet another useful model for studying the mechanism of down-regulation of genes by androgen and the identification of tissue specific factors in the lacrimal and submaxillary gland.

Expression of the rat BFGF antisense RNA transcript is tissue-specific and developmentally regulated.

The basic fibroblast growth factor (bFGF) gene locus is transcribed into a number of mRNA transcripts including an antisense mRNA derived from the opposite DNA strand of the bFGF gene. Expression of this natural antisense RNA has been implicated in regulation of the bFGF sense mRNA expression and turnover. In the present study we examined the developmental pattern of expression of the bFGF antisense transcript in fetal and postnatal rat tissues. Northern hybridization with a strand-specific cRNA probe detected a 1.5-kb polyadenylated antisense RNA in all tissues examined except brain, in which two transcripts were detected as a doublet of approximately 1.3-1.5 kb in size. The level of antisense transcript expression was markedly tissue- and age-dependent. In the developing brain, both sense and antisense transcripts were detected by Northern hybridization, but the pattern of their expression was inversely related. The 6.0-kb bFGF sense transcript increased in an age-dependent manner from days 3-30 of postnatal development while the antisense transcript decreased to nearly undetectable levels over the same period. In embryonic (days 15-19) liver, kidney, heart and intestine bFGF antisense RNA expression was low but increased dramatically at parturition, rising 5-10-fold over fetal levels by 10 days of age, then declined slowly to a new steady-state level in adult tissues. The level of antisense RNA in these tissues was much higher than that of bFGF sense mRNA, which was undetectable by Northern analysis. Sense and antisense trancripts were also detected in midgestational (11.5 days) embryos by RT-PCR. Antisense expression did not increase when embryos were explanted and cultured for 48 h (9.5-11.5 days). The apparent reciprocal relationship between the abundance of sense and antisense bFGF transcripts in developing brain supports the possibility of a regulatory role for the antisense transcript in this tissue. There was no evidence for a reciprocal relationship between sense and antisense expression in the other tissues examined, indicating that the relationship between sense and antisense RNA expression may be tissue-specific.

Plasma clearance, tissue uptake and expression of pituitary peptide 23/pancreatitis-associated protein in the rat.

The secretion of peptide 23 by rat pituitary cells is stimulated by growth hormone-releasing hormone and inhibited by somatostatin. Recent cloning of the cognate cDNA for peptide 23 revealed that it is identical to pancreatitis-associated protein (PAP). In the present study, the clearance and tissue uptake of recombinant peptide 23/PAP in normal adult male rats was assessed. The plasma half-life of recombinant peptide 23/PAP was 4.8 +/- 1.4 (S.D.) min. Maximal accumulation of radio-labelled peptide 23/PAP was observed in the kidney, stomach, small intestine and pancreas whereas negligible uptake was seen in the liver, lung or heart. Peptide 23/PAP was detected in a variety of tissue extracts using a radioimmunoassay. Extracts of ileum contained the highest concentrations of peptide 23/PAP. In situ hybridization analysis showed that peptide 23/PAP mRNA was highly expressed in the columnar epithelial cells of ileum, jejunum and duodenum. These observations demonstrate that peptide 23/PAP, a protein previously thought to be of pituitary origin, is widely expressed in the gastrointestinal tract and that it is rapidly removed from the circulation by the kidney and by tissues which express peptide 23/PAP.

Age-related changes in peptide-23/pancreatitis-associated protein and pancreatic stone protein/reg gene expression in the rat and regulation by growth hormone-releasing hormone.

Peptide-23 is a 16-kilodalton protein secreted by rat pituitary cells that was first identified because it was regulated by GRF and somatostatin in a similar fashion to GH. Cloning of peptide-23 complementary DNA revealed that it is identical to pancreatitis-associated protein (PAP) and a member of the c-lectin gene family. We examined the expression of peptide-23/PAP and a structurally related protein, pancreatic stone protein (PSP/reg), in the rat gastrointestinal tract. Here we report age-related changes in the expression and GRF regulation of peptide-23. Both peptide-23/PAP messenger RNA (mRNA) and PSP/reg mRNA were virtually undetectable in the small intestine of newborn and 1- and 2-week-old rats. A dramatic increase in the expression of both genes was seen at the time of weaning in the third week postpartum. The abundance of both of these mRNA decreases after 3 and 6 months of age. Peptide-23/PAP mRNA is most abundant in the ileum, whereas PSP/reg is maximally expressed in the pancreas and duodenum. Human GRF analog pellets were implanted sc into adult male rats for 2 weeks to study the chronic effects of GRF on the expression of these genes. Both peptide-23/PAP and PSP/reg mRNA levels in duodenum and jejunum were increased in these rats compared with levels in control rats. However, no increase in peptide-23/PAP mRNA in response to GRF treatment was seen in the ileum, where the level of expression of this gene is very high, and GRF had no effect on peptide-23/PSP expression in the heart, pituitary, or hypothalamus, where expression is normally undetectable. In situ hybridization was used to localize peptide-23/PSP in the small intestine and pancreas of GRF-treated rats. An increase in peptide-23/PAP mRNA was restricted to acinar cells close to islets, whereas little expression was seen in acinar cells distant from islets, suggesting that either peptide-23/PAP may have some paracrine action on the islets, or alternatively, an islet-derived factor may function as a paracrine modulator of peptide-23/PAP expression. These data demonstrate that GRF modulates peptide-23/PAP expression in the gastrointestinal tract in a similar fashion to that previously reported for pituitary cells in primary culture.

Molecular cloning and expression of peptide 23, a growth hormone-releasing hormone-inducible pituitary protein.

Peptide 23 is a newly identified protein secreted by rat pituitary cells in primary culture. Although the secretion of this protein is stimulated by GH-releasing hormone and inhibited by somatostatin, the N-terminal amino acid sequence of peptide 23 shows no homology to rat GH. Using the polymerase chain reaction technique, we cloned and sequenced the peptide 23 complementary DNA (cDNA). By means of the mixed oligonucleotide-primed amplification of cDNA technique, primers corresponding to the NH2-amino acid sequence of peptide 23 were used to amplify, clone, and sequence a 74-basepair cDNA of peptide 23. This polymerase chain reaction product was then used as a primer to amplify the complete peptide 23 cDNA by means of the rapid amplification of cDNA ends procedure. The cDNA of peptide 23 obtained by the rapid amplification of cDNA ends procedure contained 777 nucleotides and encoded a 175-amino acid protein with a 26-amino acid putative signal peptide. The calculated mol wt of the mature protein (16,613 daltons) was in good agreement with that estimated by polyacrylamide gel electrophoresis (16 kilodaltons). Northern blot analysis revealed a major messenger RNA species of about 0.9 kilobase and a minor species of about 1.7 kilobases in cultured rat anterior pituitary cells. In rats, peptide 23 was most abundant in the pancreas and gastrointestinal tract. A GenBank sequence search revealed complete sequence identity between peptide 23 cDNA and pancreatitis-associated protein cDNA, an approximately 73% homology with human hepatocellular carcinoma cDNA from human hepatocellular carcinoma, 64% homology with bovine pancreatic thread protein cDNA, and 55% homology with rat and human reg cDNAs, which have been reported to be expressed in regenerating pancreatic islets. Therefore, peptide 23 is identical to pancreatitis-associated protein and a member of the C-type lectin supergene family.