P2-03-01: Genetic Reduction of Circulating Insulin-Like Growth Factor (IGF)-1 Differentially Impacts the Effects of Diet-Induced Obesity and Calorie Restriction on Mammary Tumor Progression

Background: Epidemiological evidence suggests that obesity worsens breast cancer prognosis in pre- and postmenopausal women. Previously we demonstrated that diet-induced obese (DIO) mice and rats have higher circulating levels of IGF-1 and enhanced mammary tumor burden compared to normoweight controls. In contrast, we have found in several mouse models that a 30% calorie restricted (CR) diet regimen reduced serum IGF-1 levels and decreased mammary tumor growth. The present study investigated the effects of CR and DIO on mammary tumor growth in mice with normal serum levels of IGF-1 (wild-type) or mice with reduced circulating levels of IGF-1 (LID). Our goal was to identify new targets and strategies (with a focus on the IGF-1 pathway) for breaking the obesity-breast cancer link, which is important given the high prevalence of obesity in the U.S., and the strong association between obesity and breast cancer.

Materials and Methods: Virgin female liver-specific IGF-1-deficient (LID) and wild-type (FVB/NCr) mice (10-12 weeks of age) were randomized (n=10-14 per group) to receive: control diet (modified AIN-76A), a 30% CR regimen, or a DIO regimen (60 kcal% fat diet). Mice consumed the experimental diets for 11 to 15 weeks prior to orthotopic injection (4th mammary fat pad) of Met-1fvb2murine mammary tumor cells derived from a polyoma middle-T antigen transgenic mouse mammary tumor. Transplanted tumors were allowed to grow for 4 weeks and tumor volume was measured by electronic calipers biweekly. Mice were then killed, serum was collected and stored at −80C, and tumors were excised, weighed and either formalin-fixed, paraffin-embedded or flash-frozen.

Results: Using the LID mouse model of IGF-1 deficiency, this study established that decreased circulating IGF-1 (relative to wildtype mice) dramatically reduces Met-1fvb2 mammary tumor growth regardless of diet treatment (P=0.034). We also showed in LID mice that CR (44 +/− 15.6 mm3), relative to control diet (89 +/− 33 mm3) has no significant effect on Met-1fvb2 tumor growth, suggesting nearly all of the suppressive effects of CR on tumor growth (at least in the Met-1fvb2 model), can be explained by reduced systemic IGF-1. In contrast, while most of the DIO effect on tumor growth was ablated in the LID mice (332 +/− 104 mm3) vs. wild-type mice (3130 +/− 421 mm3), tumor growth in LID/DIO mice was significantly higher than LID/control mice (89 +/− 33, P=0.006) despite no change in IGF-1 levels, suggesting that a mechanism independent of IGF-1 contributes to some of the effects of DIO on Met-1fvb2 mammary tumor growth. Within LID mice, serum insulin, resistin and adiponectin were not modified by the CR or DIO diet.

Discussion: We conclude that components of the IGF-1 pathway represent promising mechanistic targets for mimicking the anticancer effects of CR and breaking the obesity-breast cancer link. However, there are likely other mechanisms independent of IGF-1 that contribute to the breast tumor enhancing effects of DIO and that need to be defined.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P2-03-01.

Diet-induced adiposity alters the serum profile of inflammation in C57BL/6N mice as measured by antibody array

Morbid obesity is considered a systemic inflammatory state. The objective of this project was to characterize the adipokine, cytokine and chemokine protein profile in serum from control, lean and obese mice. We hypothesized that chemokines and cytokines are altered by caloric restriction and diet-induced obesity as a function of changes in body composition. Six-week-old female C57BL/6N mice (n = 12 per group) were randomized to one of three diets: control (fed ad libitum); lean (30% calorie-restricted regimen relative to control) and diet-induced obese (DIO; high calorie diet, fed ad libitum). Body weight, body composition and food intake were monitored throughout the study. After 10 weeks on the diets, blood samples were collected, and adipokine/cytokine/chemokine serum profiles were measured by antibody array. Lean mice, relative to the control group, displayed increased concentrations of insulin-like growth factor (IGF) binding protein-3, -5 and -6 and adiponectin and decreased IGF-1. These mice also showed increased concentrations of interleukin (IL)-10, IL-12 p40/p70, eotaxin, monocyte chemoattractant protein-5 and SDF-1. In contrast, DIO mice displayed increased leptin, IL-6 and LPS-induced chemokine and decreased concentrations of all chemokines/cytokines measured relative to control mice. As such, these data indicate that DIO may lead to an inflammatory state characterized as a shift towards a T helper lymphocyte type 1-skewed responsiveness. The demonstration of differential adipokine, cytokine and chemokine protein profile in control, lean and DIO mice may have implications for immune responsiveness and risk of disease.

Premenopausal breast cancer and the association between estrogen receptor status, insulin-like growth factor-I (IGF-I), insulin-like growth factor binding protein-3 (IGFBP-3) and leptin

640

Background: The IGF pathway and leptin are associated with breast cancer risk in premenopausal women. These proteins are implicated in breast carcinogenesis through their interactions with the estrogen pathway, potentially resulting in specific breast cancer histopathologic subtypes. Methods: Pretreatment serum IGF-I, IGFBP-3 and leptin levels were examined in newly diagnosed premenopausal breast cancer cases to determine whether an association with tumor estrogen receptor status exists. Odds ratios were determined by logistic regression analysis. Likelihood ratio tests for interaction with BMI were conducted. Results: One hundred and eight women were evaluated (mean age 42); 82 were ER positive and 26 were ER negative. 62 were normal weight (BMI<25) and 46 were overweight and obese (BMI≥25). Mean IGF-I, IGFBP-3 and leptin levels did not differ between the ER positive and ER negative tumors. No associations were found between IGF-I, IGFBP-3 or leptin and ER status of tumors overall or by subset analysis in normal weight versus overweight and obese individuals. A suggestion that elevated IGFBP-3 was associated with ER-negative disease did not reach statistical significance. Additionally, normal-weight women with ER-negative disease had higher leptin levels than their normal-weight ER-positive counterparts, although we lacked statistical power to detect an effect of BMI on the association between leptin levels and tumor hormone receptor status. Conclusions: This is the first report examining the IGF pathway and leptin in relation to ER status in premenopausal women with newly diagnosed breast cancer. Our small number of cases suggests a number of intriguing findings that should be evaluated in larger studies. Determining links between ER status and IGFBP-3 and leptin may help better define risk factors and potentially appropriate interventions for ER-negative versus ER-positive breast cancer.

No significant financial relationships to disclose.

Genotoxicity and carcinogenicity studies of soy isoflavones

The potential cancer preventive efficacy of soy isoflavones is being investigated in preclinical and phase 1 clinical studies sponsored by the U.S. National Cancer Institute. Although 90-day oral toxicity studies with PTI G-2535 (an investigational soy isoflavone drug product) in rats and dogs, as well as teratology studies, indicated no signs of toxicity, there remains a mechanistic concern associated with the ability of isoflavones (i.e., genistein) to inhibit topoisomerase, possibly leading to DNA strand breaks. The present report describes results from two in vitro genotoxicity studies, one in vivo genotoxicity study, and a single carcinogenicity study conducted in p53 knockout mice. Bacterial mutagenesis experiments using six tester strains without metabolic activation revealed no evidence that PTI G-2535 was mutagenic. In similar experiments with exogenous metabolic activation there were statistically significant increases in revertants, but less than twofold, in a single (Salmonella typhimurium TA100) tester strain. Mouse lymphoma cell mutagenesis experiments conducted with and without metabolic activation revealed statistically significant increases in mutation frequency at PTI G-2535 concentrations > or = 0.8 and 12 microg/ml, respectively; such increases were dose related and increases in the frequency of both small and large colonies were observed. A statistically significant increase in the frequency of micronucleated polychromatic erythrocytes was also seen 24 hours after treatment in male, but not female, mice who received 500 and 1000 mg/kg body weight PTI G-2535; however,such increases were small, were not dose related, and were not observed 48 hours after treatment. In contrast, dietary genistein had no effect on survival, weight gain, or the incidence or types of tumors that developed in cancer-prone rodents lacking the p53 tumor suppressor gene, p53 knockout mice. The apparent risk/benefit of isoflavone ingestion may ultimately depend on the dose and developmental timing of exposure.

Diet and cancer prevention studies in p53-deficient mice

Progress in mechanism-based cancer prevention research may be facilitated by the use of animal models displaying specific genetic susceptibilities for cancer such as mice deficient in the p53 tumor suppressor gene, the most frequently altered gene in human cancer. We observed in p53-knockout (p53-/-) mice that calorie restriction (CR; 60% of the control group's intake of carbohydrate energy) increased the latency of spontaneous tumor development (mostly lymphomas) approximately 75%, decreased serum insulin-like growth factor (IGF)-1 and leptin levels, significantly slowed thymocyte cell cycle traverse and induced apoptosis in immature thymocytes. In heterozygous p53-deficient (p53+/-) mice, CR and 1 d/wk of food deprivation each significantly delayed spontaneous tumor development (a mix of lymphomas, sarcomas and epithelial tumors) and decreased serum IGF-1 and leptin levels even when begun late in life. We have also developed a rapid and relevant p53+/- mouse mammary tumor model by crossing p53-deficient mice with MMTV-Wnt-1 transgenic mice, and found that CR and 1 d/wk food deprivation significantly increased mammary tumor latency (greater than twofold) and reduced the mean serum IGF-1 and leptin levels to <50% of that of control mice (P < 0.0001). In addition, fluasterone, fenretinide and soy each delayed tumor development but had little effect on IGF-1 or leptin levels. We have capitalized on the susceptibility of p53+/- mice to chronic, low dose, aromatic amine-induced bladder carcinogenesis to develop a useful model for evaluating bladder cancer prevention approaches such as cyclooxygenase-2 inhibition. As demonstrated by these examples, mice with specific (and human-like) genetic susceptibilities for cancer provide powerful new tools for testing and characterizing interventions that may inhibit the process of carcinogenesis in humans.

Cancer prevention studies in p53-deficient mice

Future progress in mechanism-based cancer prevention research may be facilitated by animal models displaying specific genetic susceptibilities for cancer, such as mice deficient in 1 (+/-) or both (-/-) alleles of the p53 tumor suppressor gene. We observed in p53-/- mice that calorie restriction (CR) increased the latency of spontaneous tumor development (mostly lymphomas) by approximately 75%, decreased serum insulin-like growth factor-1 (IGF-1) and leptin levels, slowed thymocyte cell cycle traverse, and induced apoptosis in immature thymocytes. In p53+/- mice, CR and a 1 d/wk fast each delayed spontaneous tumor development (a mix of lymphomas, sarcomas, and epithelial tumors) and decreased serum IGF-1 and leptin levels, even when begun late in life. In p53+/-Wnt-1 transgenic mice, a mammary tumor model, the same interventions increased mammary tumor latency and reduced mean serum IGF-1 and leptin levels to <50% of those of control mice. We capitalized on the susceptibility of p53+/- mice to chronic, low-dose aromatic amine-induced bladder carcinogenesis to develop a useful model for evaluating bladder cancer prevention approaches. These examples clearly indicate that mice with specific (and humanlike) genetic susceptibilities for cancer are powerful models for testing interventions that may inhibit carcinogenesis in humans.

Effects of calorie restriction on thymocyte growth, death and maturation

We previously reported that calorie restriction (CR) significantly delays the spontaneous development of thymic lymphomas and other neoplasms in p53-deficient mice and their wild-type littermates. The purpose of the present study was to further characterize the anti-lymphoma effects of CR by assessing thymocyte growth, death and maturation in response to acute (6 day) and chronic (28 day) CR regimens. Male C57BL/6J mice fed a CR diet (restricted to 60% of control ad libitum intake) for 6 days displayed a severe reduction in thymic size and cellularity, as well as a decrease in splenic size and cellularity; these declines were sustained through 28 days of CR. Mice maintained on a CR diet for 28 days also displayed a significant depletion in the cell numbers of all four major thymocyte subsets defined by CD4 and CD8 expression. Analysis within the immature CD4(-)8(-) thymocyte subset further revealed an alteration in normal CD44 and CD25 subset distribution. In particular, CR for 28 days resulted in a significant decrease in the percentage of the proliferative CD44(-)25(-) subset. In addition, a significant increase in the percentage of the early, pro-T cell CD44(+)25(-) population was detected, indicative of a CR-induced delay in thymocyte maturation. Taken together, these findings suggest that CR suppresses (through several putative mechanisms) lymphomagenesis by reducing the pool of immature thymocytes that constitute the lymphoma-susceptible subpopulation.

Splenic marginal zone B-cell and thymic T-cell lymphomas in p53-deficient mice

The molecular pathology and histogenesis of lymphomas in 56 retired breeder male and 14 12-week-old male homozygous p53-deficient (p53-/-) mice (C57BL/6TacfBR-[KO]p53 N4) were evaluated. Lymphomas were assessed by serial morphologic techniques, immunohistochemistry, flow cytometry, and analysis of T cell receptor (TCR) or immunoglobulin heavy chain (IgH) gene rearrangements. We found two common types of lymphomas. T-cell lymphomas arose in the thymus through a sequence of lymphocyte depletion, medullary hyperplasia, and unilateral lymphoma. Tumor cells were CD3+, CD90+, and usually TCRalpha/beta+ and possessed clonal TCRbeta gene rearrangements. Thymic lymphoblastic lymphomas (LL) were highly malignant and quickly metastasized to the splenic white pulp and liver, even when the thymus was only slightly increased in weight. In the spleen, a novel lymphoma was found. Marginal zone hyperplasia led to marginal zone lymphoma (MZL), a well-differentiated lymphoma that usually expressed CD45R (B220) and CD5 at low levels and that had clonal IgH gene rearrangements. IgH gene rearrangements were also seen in spleens with marginal zone hyperplasias only. Hyperplastic and neoplastic marginal zone B cells expressed IgM at low to normal levels, as seen by FACS analysis and immunohistochemistry. These tumors only metastasized to the liver at a later stage, as they became less differentiated. Several mice had both types of tumors present in the spleen. Two B-cell lymphoblastic lymphomas of uncertain origin were also found. In this article, we discuss the possible mechanisms responsible for development of the lymphomas seen in these p53-deficient mice.

Calorie restriction reduces ulcerative dermatitis and infection-related mortality in p53-deficient and wild-type mice

In rodents calorie restriction (CR) reduces cancer incidence, improves health by delaying age-related declines in physiologic measures, and extends both median and maximal life span. The mechanisms underlying the various beneficial effects of CR remain undefined. In this study, heterozygous p53-deficient (p53(+/-)) mice (in which the inactivation of one allele of the p53 tumor suppressor gene increases susceptibility to spontaneous and carcinogen-induced tumor development) and wild-type (WT) litter mates were subjected to a two-stage skin carcinogenesis protocol with 7,12-dimethylbenz[a]anthracene and 12-O-tetradecanoylphorbol-13-acetate. Instead of skin carcinomas, however, the chemical treatment protocol caused ulcerous skin lesions, and 89% of mice fed ad libitum died from infection/septicemia. When WT mice were restricted to 60% of the average calorie intake of the respective ad libitum group, however, only 33% developed such lesions, and the CR mice survived twice as long on average as the ad libitum mice. CR also extended life span in p53(+/-) mice, but 50% of p53(+/-) mice subjected to CR still developed skin ulcers and mean life span was shorter than that seen in WT mice. Differences in response to CR between WT and p53(+/-) mice may be due to the reduction in p53 gene dosage, dissimilarity in the application of the CR treatment, or both. These results suggest that some of the beneficial effects of CR may need full expression of p53 for complete realization.

p53-independent inhibition of nitric oxide generation by cancer preventive interventions in ex vivo mouse peritoneal macrophages

We have previously reported that dietary interventions with purported anti-inflammatory activity delay spontaneous tumorigenesis in p53-deficient (p53-/-) mice. In the present study, 4 weeks of calorie restriction or the dietary administration of chemopreventive steroids dehydroepiandrosterone and 16alpha-fluoro-5-androsten-17-one significantly reduced the inducible generation of nitric oxide (NO) in ex vivo peritoneal macrophages from male p53 wild-type and p53-/- mice relative to the respective ad libitum-fed controls; expression of inducible nitric oxide synthase II (NOS2) protein was also markedly decreased compared with respective controls. These findings suggest that the p53-independent suppression of inducible NO production observed in this study may contribute to the anti-cancer effects of these preventive interventions in p53-/- mice.

Calorie restriction induces a p53-independent delay of spontaneous carcinogenesis in p53-deficient and wild-type mice

We reported previously that calorie restriction (CR) delays spontaneous carcinogenesis in p53-deficient (p53-/-) mice, suggesting that CR modulates carcinogenesis by p53-independent mechanisms. To further evaluate the role of p53, we monitored tumor development in p53-/- and wild-type (p53+/+) mice fed ad libitum (AL) or a CR regimen (60% of AL calorie intake). CR delayed tumor mortality in p53-/- and p53+/+ mice (mean time to death, 169 and 648 days, respectively) relative to AL feeding (104 and 470 days). The estimated age-specific cancer death rate AL:CR ratios were 4.3 for p53-/- mice and 4.4 for p53+/+ mice. Thus, despite the accelerated onset of carcinogenesis in p53-/- mice, the tumor-delaying effect of CR was similar in the two genotypes.

Effects of dehydroepiandrosterone and calorie restriction on the Bcl-2/Bax-mediated apoptotic pathway in p53-deficient mice

Modulation of apoptosis through altered expression of Bcl-2 and/or Bax may be a mechanism by which dehydroepiandrosterone (DHEA) administration and calorie restriction (CR) exert their chemopreventive effects in p53-deficient (p53-/-) mice. Using immunohistochemical detection we found that treatment with both DHEA and CR resulted in decreased expression of the PCNA proliferation marker in the thymus. In addition, treatment with DHEA also increased the rate of apoptosis in the thymus, resulting in marked thymic atrophy. Thus, both DHEA and CR appear to shift cell number homeostasis by favoring apoptosis. To further understand the molecular mechanisms by which DHEA and CR exert their effects, we examined two components of the apoptotic pathway, Bcl-2 and Bax. We found that p53-/- mice have much higher levels of Bcl-2 mRNA in the thymus than wild-type (p53+/+) mice. Treatment of p53-/- animals with DHEA resulted in decreased Bcl-2 but not Bax mRNA levels in the thymus. In contrast, CR did not change either Bcl-2 or Bax mRNA expression. The present study provides molecular evidence that DHEA and CR may modulate tumorigenesis through alterations in the apoptotic and/or proliferative pathways.

Chemoprevention of spontaneous tumorigenesis in nullizygous p53-deficient mice by dehydroepiandrosterone and its analog 16alpha-fluoro-5-androsten-17-one

Transgenic mice with both alleles of the p53 tumor suppressor gene product 'knocked out' by gene targeting are susceptible to early development of tumors, chiefly lymphomas and sarcomas. Compared with the control group, administration of dehydroepiandrosterone (DHEA) at 0.3% of the diet to male p53-deficient mice extended their lifespan by delaying death due to neoplasms (from 105 to 166 days on study, P = 0.002), primarily by suppressing lymphoblastic lymphoma (from 45 to 6% of neoplastic deaths, P = 0.010). Treatment with a synthetic DHEA analog, 16alpha-fluoro-5-androsten-17-one (compound 8354), at 0.15% of the diet also increased lifespan, to 140 days for mice that developed tumors (P = 0.037). The effects of these steroids on lifespan and tumor development did not appear to be strongly related to inhibition of food consumption and weight gain, in that a group pair-fed with control diet to the reduced food consumption of the DHEA-treated group developed and died of the same types of neoplasms at the same rate as the controls fed ad libitum. The chemopreventive effect of these steroids has been proposed to be due to suppression of DNA synthesis by inhibition of glucose 6-phosphate dehydrogenase, the rate-limiting enzyme of the pentose phosphate pathway. Although DHEA and its analog are strong non-competitive inhibitors of this enzyme in vitro, treatment with DHEA did not deplete cellular nucleotide pools in the liver, as would have been predicted. The chemopreventive effect of DHEA in this model may be due to steroid-induced thymic atrophy and suppression of T cell lymphoma, permitting these mice to survive long enough to develop tumors with longer latency.

Chemoprevention of spontaneous tumorigenesis in p53-knockout mice

Spontaneous tumorigenesis was evaluated in male p53-knockout (p53-/-) mice treated with dehydroepiandrosterone (DHEA), quercetin, d-limonene, or all-trans retinoic acid to determine whether tumor development in these mice can be modulated by cancer-chemopreventive agents. DHEA-treated mice experienced a delay in tumorigenesis (particularly lymphomas) and subsequent mortality (P < 0.01) relative to untreated control mice. Quercetin, d-limonene, and all-trans retinoic acid each had no effect on spontaneous tumor development in p53-/- mice. These data demonstrate that tumor development in p53-/- mice can be delayed by DHEA and suggest that p53-/- mice provide a useful model for evaluating strategies to offset the increased risk of tumorigenesis resulting from loss of p53 tumor suppressor function.

Calorie restriction delays spontaneous tumorigenesis in p53-knockout transgenic mice

Transgenic mice with both alleles of the p53 tumor suppressor gene (frequently mutated in human tumors) knocked out by gene targeting provide a potentially useful tumorigenesis model because these mice rapidly develop spontaneous tumors. To determine whether tumorigenesis in p53-knockout mice is sensitive to experimental manipulation, tumor development in response to calorie restriction (CR; a potent inhibitor of rodent tumors) was evaluated. Tumor development was monitored for 48 weeks in male nullizygous p53-knockout and wild-type littermate mice (28-30 per treatment group) fed ad libitum (AL) or restricted to 60% of AL carbohydrate calorie intake. CR:p53-knockout mice (median survival = 25 weeks) experienced a delay in tumor onset and subsequent mortality (P = 0.0002) relative to AL:p53-knockout mice (median survival = 16 weeks). Tumor development and mortality in wild-type littermates on either diet treatment were < 4% through 48 weeks. Cell cycle analyses were performed on splenocytes from p53-knockout mice and wild-type littermates after 4 weeks of AL feeding or CR (5 per group). The percentage of splenocytes in S phase of the cell cycle was 3-fold higher for p53-knockout mice than wild-type mice (P < 0.001), and CR reduced the percentage of S-phase splenocytes in both p53-knockout and wild-type mice (P = 0.012). These data demonstrate that tumor development in p53-knockout mice genetically predisposed to tumors can be delayed by CR (possibly via cell cycle modulation) and suggest that these mice provide a very useful model of spontaneous tumorigenesis.

Peptidyl-alpha-hydroxyglycine alpha-amidating lyase. Purification, characterization, and expression

The production of alpha-amidated peptides from their glycine-extended precursors is a two-step process involving the sequential action of two catalytic domains encoded by the bifunctional peptidylglycine alpha-amidating monooxygenase (PAM) precursor. The NH2-terminal third of the PAM precursor contains the first enzyme, peptidylglycine alpha-hydroxylating monooxygenase (PHM), a copper, molecular oxygen, and ascorbate-dependent enzyme. The middle third of the PAM precursor contains the second enzyme, peptidyl-alpha-hydroxyglycine alpha-amidating lyase (PAL). The COOH-terminal third of the PAM precursor encodes a transmembrane domain and a hydrophilic domain that may form a cytoplasmic tail. Antisera to a peptide within the PAL domain were used to identify a 50-kDa protein as the major form of PAL in bovine neurointermediate pituitary granules. This 50-kDa PAL protein was purified and found to begin at Asp434 of bPAM, indicating that it could arise through endoproteolytic cleavage of the bPAM precursor at Lys432-Lys433. With alpha-N-acetyl-Tyr-Val-alpha-hydroxyglycine as the substrate, PAL exhibits a pH optimum of 5.0; enzymatic activity is inhibited by high concentrations of salt but is relatively resistant to thiol reagents and urea. PAL activity is inhibited by EDTA and restored by a number of divalent metals, including Cd2+, Cu2+, Zn2+, and Ca2+. Kinetic studies using alpha-N-acetyl-Tyr-Val-alpha-hydroxyglycine indicate that PAL has a Km of 38 microM and a turnover number of 220/s. Expression vectors encoding only the soluble PHM domain or the PAM precursor from which the PHM domain had been deleted were constructed. hEK293 cells transfected with the PHM vector exhibited a 10-fold increase in secretion of PHM activity with no PHM activity detectable in control or transfected cells. hEK293 cells transfected with the PAL vector exhibited a 2-fold increase in secretion of PAL activity and a 15-fold increase in cellular PAL activity. Most of the PAL activity produced by the transfected cells remained membrane-associated.

pH-dependent stimulation of peptidylglycine alpha-amidating monooxygenase activity by a granule-associated factor

Peptidylglycine alpha-amidating monooxygenase (PAM; EC 1.14.17.3) is a granule-associated enzyme that catalyzes the production of alpha-amidated peptides from their glycine-extended precursors, a posttranslational modification often required for full biological activity. PAM activity in crude homogenates of bovine neurointermediate pituitary has an acidic pH optimum for the peptide substrate alpha-N-Ac-Tyr-Val-Gly. During purification, the pH optimum shifts, so that purified bovine (b)PAM exhibits an alkaline pH optimum for this substrate with virtually no activity below pH 6.5. A factor that restores the ability of purified bPAM to produce alpha-amidated products at pH 6 was identified. In rat anterior pituitary this factor (denoted SPAM for stimulator of PAM activity) was a soluble protein with a mol wt of 44 K by gel filtration; its stimulatory activity could be reduced or eliminated by trypsin digestion or boiling. SPAM stimulated PAM activity at acidic pH by increasing the apparent Vmax and decreasing the apparent Michaelis-Menten constant (Km) for the peptide substrate. Like PAM, SPAM activity is localized to the secretory granule. Levels of SPAM activity in various rat tissues correlated closely with levels of PAM activity, with the greatest amount of SPAM activity in atrium, anterior pituitary, and neurointermediate pituitary. The distribution of PAM and SPAM between soluble and membrane fractions also correlated closely. In AtT-20 cell lines transfected with a complementary DNA (cDNA) encoding the full-length bPAM precursor, both SPAM and PAM activities were increased compared to wild type cells; both activities were decreased in a cell line expressing an antisense rat (r)PAM mRNA. In marked contrast, an AtT-20 cell line transfected with a cDNA encoding a truncated, soluble form of bPAM had elevated levels of PAM activity, but levels of SPAM activity were not increased compared to wild-type cells. These results suggest that SPAM activity is closely linked to the expression of full-length PAM. The interaction of PAM and SPAM may represent a site for regulation of the synthesis of bioactive peptides, particularly at low intragranular pH.

The 108-kDA peptidylglycine alpha-amidating monooxygenase precursor contains two separable enzymatic activities involved in peptide amidation

A 43-kDa protein factor that increases the ability of purified bovine peptidylglycine alpha-amidating monooxygenase (PAM)-A and -B to produce alpha-amidated peptides at physiological pH was purified to homogeneity from bovine neurointermediate pituitary. At each step of the purification, the amount of activity correlated with the amount of protein detected on Western blots by antibody to bovine PAM(561-579). In the bovine neurointermediate pituitary the 108-kDa PAM precursor protein is cleaved to form a peptidylglycine alpha-hydroxylating monooxygenase and a peptidyl-alpha-hydroxyglycine alpha-amidating lyase, which function sequentially in the 2-step formation of alpha-amidated peptides.

Stable expression of full-length and truncated bovine peptidylglycine alpha-amidating monooxygenase complementary DNAs in cultured cells

Peptidylglycine alpha-amidating monooxygenase (PAM; EC 1.14.17.3) catalyzes the production of alpha-amidated peptides from their glycine-extended precursors, a posttranslational modification often required for full biological activity. We have previously cloned cDNAs encoding a 108-kDa bovine PAM precursor. To confirm that this cDNA encodes a functional alpha-amidating enzyme and to begin to examine the structural requirements for the biosynthesis of an active PAM enzyme, we constructed expression vectors that placed the cDNA for either the full-sized enzyme or a form truncated at the carboxyl-terminal (and thus lacking the transmembrane domain) under the control of the mouse metallothionein-1 promoter. We used the resultant plasmids to transfect AtT-20 mouse anterior pituitary corticotrope cells and selected stable lines that expressed increased levels of PAM activity. Transfected cells in which expression from the metallothionein promoter had been induced had up to 15-fold higher levels of PAM mRNA and up to 7.5-fold higher levels of PAM activity than wild-type cells. The PAM activity in the transfected cells shared many enzymatic characteristics with PAM-B, a 38-kDa soluble form of PAM purified from bovine neurointermediate pituitary. These included copper- and ascorbate-dependent activity, an alkaline pH optimum for the peptide substrate D-Tyr-Val-Gly, similar affinities for several other synthetic substrates, and comparable apparent size during gel filtration. Compared to extracts of wild-type cells, extracts from transfected cells showed increased production of five different amino acid alpha-amides. These data indicate that a single enzyme can act on a variety of peptide substrates, and that the full structure of the PAM precursor is not necessary during biosynthesis for expression of active PAM enzyme.

Effects of elevated serum insulinlike growth factor-II on growth hormone and insulinlike growth factor-I mRNA and secretion

The insulinlike growth factors (IGF) appear to exert feedback control over their own production. In an effort to determine the physiologic mechanisms for this feedback modulation, we utilized a previously developed in vivo model in which rIGF-II secreting tumor cells are transplanted into immunodeficient rats to form IGF-II secreting tumors. The tumor-bearing rat have serum IGF-II concentrations sevenfold greater than those in controls (119 +/- 16 ng/mL [mean +/- SE] v 17 +/- 2 ng/mL, P less than .0001). Serum IGF-I concentrations were reduced among the tumor-bearing rats (438 +/- 42 ng/mL v 606 +/- 32 ng/mL, P = .002) and were negatively correlated with IGF-II concentrations (r = -.47, P = .025), suggesting that IGF-II suppressed the secretion of IGF-I. Increased serum IGF-II concentrations, however, did not affect basal growth hormone concentrations (tumor-bearing, 44 +/- 12 ng/mL; control 33 +/- 6 ng/mL, P = 0.96). The GH response to GH releasing factor was likewise similar in both groups. Moreover, pituitary GH mRNA level were not different in the two groups, suggesting that IGF-II does not have a significant effect on GH secretion in this in vivo model. There was no association between serum glucose and serum IGF-I or IGF-II concentrations. To examine the effect of IGF-II on IGF-I production from the liver, we measured IGF-I mRNA levels in a subset of animals. Despite these differences in serum IGF-I concentrations, the tumor-bearing rats did not have significantly lower liver IGF-I mRNA levels.(ABSTRACT TRUNCATED AT 250 WORDS)

Beta-adrenergic stimulation of growth hormone (GH) release in vivo, and subsequent inhibition of GH-releasing factor-induced GH secretion

In vivo and in vitro studies of beta-adrenergic influences on GH secretion have produced apparently conflicting data in which the in vivo effect seems to be inhibitory and the in vitro effect to be stimulatory. The present studies were designed to observe the in vivo effect of isoproterenol (ISO), a beta-adrenergic agonist, on 1) GH release during a brief interval after intraatrial infusion, and 2) GH release in response to GRF infused 10 min after ISO. ISO was found to stimulate GH release in both intact and hypothalamus-lesioned animals within 2 min after infusion, but GH returned to control levels within 10 min. ISO also profoundly inhibited the release of GH in response to GRF. Pretreatment of animals with somatostatin (SRIF) antiserum prevented the inhibitory action of ISO on GRF-induced GH release. No change in peripheral levels of SRIF was detected. Also, there was no suppression of GRF-induced GH release by ISO when the treatments were applied in vitro to dispersed perifused pituitary cells. These data show that beta-adrenergic systems can stimulate a rapid but brief release of GH in vivo, and that the subsequent inhibitory action on GRF-induced GH release might be by means of SRIF release.

Properties of beta-adrenergic receptors on porcine corpora lutea and granulosa cells

The nature of beta-adrenergic binding by swine corpora lutea and granulosa cells was examined with the specific beta-adrenergic radioligand, (+/-)3-[125I]iodocyanopindolol (ICYP). Saturation analyses revealed the presence of high affinity (Kd = 15.2 +/- 2.1 pM; n = 8 experiments) and low capacity (6.7 +/- 0.8 fmol/mg protein) beta-adrenergic receptors on porcine corpora lutea membranes. The properties of beta-adrenergic binding were determined by computer modeling of competition studies with a variety of compounds selective for beta-adrenergic subtypes. These studies disclosed predominantly beta 1-adrenergic receptors on pig luteal membranes. This inference from radioligand binding studies was corroborated functionally by the approximately equipotent biological effects of L-norepinephrine and L-epinephrine on cAMP production by luteal tissue (respective EC50s of 282 +/- 31 and 187 +/- 66 nM; n = 3 experiments). Physiological regulation of specific beta-adrenergic receptor content in the swine ovary was indicated by prominent (up to 9-fold) variations in receptor concentrations among corpora lutea and granulosa cells at various stages of maturity. In addition, there was differential expression of beta-adrenergic receptor subtype. Whereas the beta-adrenergic receptor subtype was predominantly beta 1 in corpora hemorrhagica and corpora lutea, granulosa cells and corpora albicantia contained principally beta 2 receptors. This difference could not be accounted for by blood cell contamination of corpora lutea, since swine blood cells contained predominantly (greater than 98%) beta 2-receptors, which were present at less than 8.6% the concentration of total beta-receptors in luteal tissue. In summary, swine corpora lutea and granulosa cells contain specific high affinity, low capacity beta-adrenergic receptors that are functionally coupled to biological responses. Moreover, total receptor content as well as beta-adrenergic subtype exhibit significant physiological variation in relation to maturational status of ovarian follicular and luteal tissue.

Beta-adrenergic binding and secretory responses of the anterior pituitary

Our studies demonstrated that beta-adrenergic agonists stimulate the release of GH from rat anterior pituitary (AP) cells in vitro. Concentration-response experiments with beta-adrenergic agonists demonstrated that beta 2-adrenergic receptors mediated this phasic GH release, while having no apparent effect on PRL or LH release. The ACTH response to beta-adrenergic agonists was equivocal. Half-maximal stimulation of GH release occurred at 14 +/- 2 (+/-SE) nM isoproterenol, 160 +/- 30 nM epinephrine, and over 1 microM l-norepinephrine (n = 4). Direct binding studies in membrane particulates of rat AP confirmed receptors of the beta 2-subtype. Iodocyanopindolol binding to beta-adrenergic receptors of rat AP yielded a dissociation constant of 4.6 +/- 0.1 pM and a maximal capacity of 1.9 +/- 0.4 fmol/mg protein (n = 3). In contrast, porcine AP contained beta 1-adrenergic receptors. These results support the hypothesis that the endogenous beta 2-adrenergic agonist l-epinephrine may be a GH-releasing factor of physiological significance in the rat.

Beta-adrenergic stimulation of anterior pituitary cyclic AMP is enhanced by tumor promoters

Beta-adrenergic stimulation of cellular cyclic AMP accumulation was characterized in normal anterior pituitary cells in vitro. In the presence of isobutylmethylxanthine, the order of potency of catecholamine agonists, as well as the antagonism by propranolol and not phentolamine, aided in classifying the receptor as beta-adrenergic. Furthermore, this agonist effect was rapidly desensitized. Tumor promoters, which directly activate protein kinase C, enhanced the cyclic AMP levels achieved with beta-adrenergic agonists (1.5-fold average at 10 min). This acute effect occurred over 10-1000 nM phorbol dibutyrate or phorbol myristate acetate. Finally, 3H-phorbol dibutyrate binding to unstimulated anterior pituitary cells was predominantly associated with the cytosol (79%) versus membrane (21%) fractions. Thus, an acute role for protein kinase C in promoting beta-adrenergic receptor activation of adenylate cyclase activity is suggested for anterior pituitary cells.

Beta-adrenergic stimulation of growth hormone release from perifused rat anterior pituitary cells

Perifusion of dispersed rat anterior pituitary cells with pulses of the beta-adrenergic agonists isoproterenol and epinephrine resulted in a prompt, but brief, rise in growth hormone release. This effect was blocked by propranolol, a beta-adrenergic antagonist, but not phentolamine, an alpha-adrenergic antagonist. These results, in conjunction with other recent reports, suggest that epinephrine may act directly at the anterior pituitary gland as a physiological growth hormone-releasing factor.

Failure of dopamine and bromocriptine to affect prolactin release and cell growth in the dopamine receptor-deficient 235-1 clone

The 235-1 clone was recently derived from the 7315a transplantable pituitary tumor and continues to secrete rat prolactin. The cells have a prominent Golgi apparatus which can be stained immunocytochemically for prolactin, but there were no 600-900 nm granules which are characteristic of normal mammotrophs. In a perfused cell-column apparatus, prolactin release from the clone was unchanged by dopaminergic agonists, thyrotropin-releasing hormone and estradiol but stimulated by dibutyryl cyclic AMP. Cellular cyclic AMP content was also not changed by dopamine but was dramatically enhanced by prostaglandin E1, indicating that at least one hormone-adenylate cyclase coupling mechanism was functional. In radioligand binding studies using the dopamine antagonist [3H]spiperone, no evidence of a dopamine receptor was obtained. The [3H]spiperone binding present was not stereoselective, and exceedingly high concentrations of other ligands were required to displace the binding. In addition, the induction of a prolactin-secreting hard tumor in rats by subcutaneous inoculation of the 235-1 cells failed to induce measurable dopamine receptors associated with the tumor cells. In order to address the possibility that there were functional dopamine receptors on these cells, but that they could not be resolved with either the cell column and cyclic AMP studies or the radioreceptor assay, the clone cells were incubated with 0.1-100 nM bromocriptine for up to 8 days. Bromocriptine had no effect on the growth rate or prolactin secretion of the 235-1 clone but inhibited prolactin release from anterior pituitary cells by over 73% in control studies. We conclude that the 235-1 clone does not express dopamine receptors and that the presence of dopamine receptors is obligatory for the typical inhibitory effects of bromocriptine on prolactin release and pituitary cell growth.