Beverly S. Schaffer

Genetic bases of estrogen-induced tumorigenesis in the rat: Mapping of loci controlling susceptibility to mammary cancer in a Brown Norway × ACI intercross

Exposure to estrogens is associated with an increased risk of breast cancer. Our laboratory has shown that the ACI rat is uniquely susceptible to 17beta-estradiol (E2)-induced mammary cancer. We previously mapped two loci, Emca1 and Emca2 (estrogen-induced mammary cancer), that act independently to determine susceptibility to E2-induced mammary cancer in crosses between the susceptible ACI rat strain and the genetically related, but resistant, Copenhagen (COP) rat strain. In this study, we evaluate susceptibility to E2-induced mammary cancer in a cross between the ACI strain and the unrelated Brown Norway (BN) rat strain. Whereas nearly 100% of the ACI rats developed mammary cancer when treated continuously with E2, BN rats did not develop palpable mammary cancer during the 196-day course of E2 treatment. Susceptibility to E2-induced mammary cancer segregated as a dominant or incompletely dominant trait in a cross between BN females and ACI males. In a population of 251 female (BN x ACI)F(2) rats, we observed evidence for a total of five genetic determinants of susceptibility. Two loci, Emca4 and Emca5, were identified when mammary cancer status at sacrifice was evaluated as the phenotype, and three additional loci, Emca6, Emca7, and Emca8, were identified when mammary cancer number was evaluated as the phenotype. A total of three genetic interactions were identified. These data indicate that susceptibility to E2-induced mammary cancer in the BN x ACI cross behaves as a complex trait controlled by at least five loci and multiple gene-gene interactions.

Inhibition of Caco-2 cell proliferation by all-trans retinoic acid: role of insulin-like growth factor binding protein-6.

The present study examined the effects of all‐trans retinoic acid (tRA) on proliferation and expression of the IGF system in Caco‐2 human colon adenocarcinoma cells. tRA inhibited Caco‐2 cell proliferation in a dose‐dependent manner, with a 40 ± 2% decrease in cell number observed 48 h after the addition of 1 μM tRA. Ligand blot analysis of IGFBPs in conditioned media revealed that Caco‐2 cells produced three IGFBPs of Mr: 34,000 (IGFBP‐2), 24,000 (IGFBP‐4), and 32,000 (IGFBP‐6). The concentrations of IGFBP‐2 and IGFBP‐4 decreased by 48 ± 6 and 70 ± 13%, respectively, whereas that of IGFBP‐6 increased by 698 ± 20% with 1 μM tRA. tRA decreased mRNA levels of IGFBP‐2 and IGFBP‐4 by 20 ± 3 and 50 ± 8%, respectively, whereas tRA increased IGFBP‐6 mRNA by 660 ± 20%. tRA did not alter levels of IGF‐II mRNA or peptide. To examine if endogenous IGFBP‐6 inhibits cell proliferation, Caco‐2 cells were transfected with an IGFBP‐6 cDNA expression construct or pcDNA3 vector only and stable clones were selected. Clones overexpressing IGFBP‐6 grew more slowly than vector controls and achieved final densities 30–55% lower than those of vector controls. Accumulation of IGFBP‐6 mRNA and concentrations of IGFBP‐6 peptide in conditioned media were increased by 200–250 and 220–250%, respectively, in the IGFBP‐6 clones compared with controls. Increased expression of IGFBP‐6, which has a high binding affinity for IGF‐II, following tRA treatment suggests that the decreased proliferation caused by tRA may result, at least in part, from IGFBP‐6‐mediated disruption of the IGF‐II autocrine loop in these colon cancer cells. J. Cell. Physiol. 190: 92–100, 2002.

Synthetic low-calcaemic vitamin D3 analogues inhibit secretion of insulin-like growth factor II and stimulate production of insulin-like growth factor-binding protein-6 in conjunction with growth suppression of HT-29 colon cancer cells

The aims of the present study were to compare the ability of various synthetic analogues of 1 alpha,25-dihydroxyvitamin D(3) [1 alpha,25-(OH)(2)D(3)] to inhibit proliferation of HT-29 cells, a human colon adenocarcinoma cell line. HT-29 cells were incubated for 144 h with various concentrations (0-100 nM) of 1 alpha,25-(OH)(2)D(3), or the analogues EB1089, CB1093 or 1 beta,25-(OH)(2)D(3). All these analogues except 1 beta,25-(OH)(2)D(3) inhibited cell proliferation, but relative potencies and efficacies of EB1089 and CB1093 were much greater than that of the native vitamin. Cells grew in serum-free medium, reaching a plateau density at day 10 of culture, and addition of 10 nM 1 alpha,25-(OH)(2)D(3) or 1 beta,25-(OH)(2)D(3) did not alter the long-term growth characteristics of HT-29 cells. However, cells treated with 10 nM EB1089 or CB1093 grew at a rate slower than control and reached final densities that were 53+/-1 and 36+/-2% lower than control, respectively. Immunoblot analysis of serum-free conditioned medium using a monoclonal anti-insulin-like growth factor-(IGF)-II antibody showed that both 10 nM EB1089 and CB1093 markedly inhibited secretion of both mature 7500 M(r) and higher M(r) forms of IGF-II. Ligand blot and immunoblot analyses of conditioned media revealed the presence of IGFBPs of M(r) 24,000 (IGFBP-4), 30,000 (glycosylated IGFBP-4), 35,000 (IGFBP-2) and 32,000-34,000 (IGFBP-6). The level of IGFBP-2 was decreased by 42+/-8 and 49+/-7% by 10 nM EB 1089 and CB1093, respectively, compared to controls. IGFBP-6 was increased approximately twofold by EB1089 and CB1093, and exogenously added IGFBP-6 inhibited HT-29 cell proliferation. These results suggest that inhibition of HT-29 cell proliferation by EB1089 and CB1093 may be attributed, at least in part, to the decreased secretion of IGF-II. The increase in IGFBP-6 concentration coupled with its high affinity for IGF-II may also contribute to decreased cellular proliferation by an indirect mechanism involving sequestration of endogenously produced IGF-II.

Growth inhibition and differentiation of the human colon carcinoma cell line, Caco-2, by constitutive expression of insulin-like growth factor binding protein-3

The human colon carcinoma cell line, Caco‐2, produces insulin‐like growth factor binding protein‐3 (IGFBP‐3), the secretion of which correlates with markers of enterocyte differentiation. To investigate whether IGFBP‐3 inhibits proliferation or induces differentiation, Caco‐2 cells were stably transfected with an IGFBP‐3 cDNA expression construct or pcDNA3 vector as a control. Accumulation of IGFBP‐3 mRNA and secretion of the protein into conditioned medium 9 days after plating were readily detected in the transfected cells, whereas these parameters were undetectable in pcDNA3‐transfected cells. Insulin‐like growth factor binding protein‐3‐expressing cells grew at a rate similar to the controls for 6 days after plating, but achieved a much lower final density between days 10 and 12. By day 9 of culture, accumulation of sucrase‐isomaltase mRNA, a marker of enterocytic differentiation of Caco‐2 cells, was evident in the IGFBP‐3‐expressing cells, but was undetectable in the controls. These results indicate that IGFBP‐3 may inhibit proliferation and induce early differentiation of Caco‐2 cells.

Nonrandom pattern of chromosome aberrations in 17β-estradiol-induced rat mammary tumors: Indications of distinct pathways for tumor development

Estrogens play an important role in breast cancer etiology and the ACI rat provides a novel animal model for defining the mechanisms through which estrogens contribute to mammary cancer development. In crossing experiments between the susceptible ACI strain and two resistant strains, COP (Copenhagen) and BN (Brown Norway), several quantitative trait loci (QTL) that affect development of 17β‐estradiol (E2)‐induced mammary tumors have been defined. Using comparative genomic hybridization (CGH), we have analyzed cytogenetic aberrations in E2‐induced mammary cancers and have found clear patterns of nonrandom chromosomal involvement. Approximately two thirds of the tumors exhibited copy number changes. Losses of rat chromosome 5 (RNO5) and RNO20 were particularly common, and it was found that these two aberrations often occurred together. A third recurrent aberration involving proximal gain and distal loss in RNO6 probably defined a distinct subgroup of tumors, since it never occurred in combination with RNO5 loss. Interestingly, QTL with powerful effects on mammary cancer development have been mapped to RNO5 and RNO6. These findings suggest that there were at least two genetic pathways to tumor formation in this rat model of E2‐induced mammary cancer. By performing CGH on mammary tumors from ACI rats, F1 rats from crosses between the ACI and COP or BN strains and ACI.BN‐Emca8 congenic rats, which carry the BN allele of the Emca8 QTL on RNO5 on the ACI genetic background, we were able to determine that the constitution of the germ line influences the pattern of chromosomal aberrations.

Genetic mapping of Eutr1 , a locus controlling E2-induced pyometritis in the Brown Norway rat, to RNO5

In certain rat strains, chronic estrogen administration can lead to pyometritis, an inflammation of the uterus accompanied by infection and the accumulation of intraluminal pus. In this article, we report that the Brown Norway (BN) rat is highly susceptible to pyometritis induced by 17β-estradiol (E2). The susceptibility of the BN rat to E2-induced pyometritis appears to segregate as a recessive trait in crosses to the resistant August × Copenhagen Irish (ACI) strain. In a (BN × ACI)F2 population, we find strong evidence for a major genetic determinant of susceptibility to E2-induced pyometritis on rat chromosome 5 (RNO5). Our data are most consistent with a model in which the BN allele of this locus, designated Eutr1 (Estrogen-induced uterine response 1), acts in an incompletely dominant manner to control E2-induced pyometritis. Furthermore, we have confirmed the contribution of Eutr1 to E2-induced uterine pyometritis using an RNO5 congenic rat strain. In addition to Eutr1, we obtained evidence suggestive of linkage for five additional loci on RNO2, 4, 11, 17, and X that control susceptibility to E2-induced pyometritis in the (BN × ACI)F2 population.

Tissue-specific actions of the Ept1, Ept2, Ept6, and Ept9 genetic determinants of responsiveness to estrogens in the female rat.

Ept1, Ept2, Ept6, and Ept9 are quantitative trait loci mapped in crosses between the ACI and Copenhagen (COP) rat strains as genetic determinants of responsiveness of the pituitary gland to estrogens. We have developed four congenic rat strains, each of which carries, on the genetic background of the ACI rat strain, alleles from the COP rat strain that span one of these quantitative trait loci. Relative to the female ACI rats, female ACI.COP-Ept1 rats exhibited reduced responsiveness to 17beta-estradiol (E2) in the pituitary gland, as evidenced by quantification of pituitary mass and circulating prolactin, and in the mammary gland, as evidenced by reduced susceptibility to E2-induced mammary cancer. The ACI.COP-Ept2 rat strain exhibited reduced responsiveness to E2 in the pituitary gland but did not differ from the ACI strain in regard to susceptibility to E2-induced mammary cancer. Interestingly, female Ept2 congenic rats exhibited increased responsiveness to E2 in the thymus, as evidenced by enhanced thymic atrophy. The ACI.COP-Ept6 rat strain exhibited increased responsiveness to E2 in the pituitary gland, which was associated with a qualitative phenotype suggestive of enhanced pituitary vascularization. The ACI.COP-Ept9 rat strain exhibited reduced responsiveness to E2 in the anterior pituitary gland, relative to the ACI rat strain. Neither Ept6 nor Ept9 impacted responsiveness to E2 in the mammary gland or thymus. These data indicate that each of these Ept genetic determinants of estrogen action is unique in regard to the tissues in which it exerts its effects and/or the direction of its effect on estrogen responsiveness.

Decreased production of insulin-like growth factor-binding protein (IGFBP)-6 by transfection of colon cancer cells with an antisense IGFBP-6 cDNA construct leads to stimulation of cell proliferation

Abstract Background : Previously, we have observed that highly unsaturated dietary (n‐3) fatty acids inhibit cell proliferation in conjunction with stimulation of insulin‐like growth factor‐binding protein (IGFBP)‐6 secretion in Caco‐2 cells, a human colon carcinoma cell line.

Genetic control of estrogen action in the rat: mapping of QTLs that impact pituitary lactotroph hyperplasia in a BN × ACI intercross

Estrogens are important regulators of growth and development and contribute to the etiology of several types of cancer. Different inbred rat strains exhibit marked, cell-type-specific differences in responsiveness to estrogens as well as differences in susceptibility to estrogen-induced tumorigenesis. Regulation of pituitary lactotroph homeostasis is one estrogen-regulated response that differs dramatically between different inbred rat strains. In this article we demonstrate that the growth response of the anterior pituitary gland of female ACI rats to 17β-estradiol (E2) markedly exceeds that of identically treated female Brown Norway (BN) rats. We further demonstrate that pituitary mass, a surrogate indicator of absolute lactotroph number, behaves as a quantitative trait in E2-treated F2 progeny generated in a genetic cross originating with BN females and ACI males. Composite interval mapping analyses of the (BN×ACI)F2 population revealed quantitative trait loci (QTLs) that exert significant effects on E2-induced pituitary growth on rat chromosome 4 (RNO4) (Ept5) and RNO7 (Ept7). Continuous treatment with E2 rapidly induces mammary cancer in female ACI rats but not BN rats, and QTLs that impact susceptibility to E2-induced mammary cancer in the (BN×ACI)F2 population described here have been mapped to RNO3 (Emca5), RNO4 (Emca6), RNO5 (Emca8), RNO6 (Emca7), and RNO7 (Emca4). Ept5 and Emca6 map to distinct regions of RNO4. However, Ept7 and Emca4 map to the same region of RNO7. No correlation between pituitary mass and mammary cancer number at necropsy was observed within the (BN×ACI)F2 population. This observation, together with the QTL mapping data, indicate that with the exception of the Ept7/Emca4 locus on RNO7, the genetic determinants of E2-induced pituitary growth differ from the genetic determinants of susceptibility to E2-induced mammary cancer.

Mapping of Three Genetic Determinants of Susceptibility to Estrogen-Induced Mammary Cancer within the Emca8 Locus on Rat Chromosome 5

The ACI rat model of 17β-estradiol (E2)-induced mammary cancer has gained wide use in the study of breast cancer etiology, prevention, and genetics. Emca8, a QTL that determines susceptibility to E2-induced mammary cancer, was previously mapped to rat chromosome 5 (RNO5) in an intercross between resistant Brown Norway (BN) and susceptible ACI rats. In this study, a panel of congenic rat strains, each of which carries BN alleles across a defined segment of RNO5 on the ACI genetic background, was generated and used to map more precisely the Emca8 determinants of mammary cancer susceptibility. Three distinct genetic determinants were localized within Emca8, and two of these were mapped to intervals of less than 15 megabases. Emca8.1 harbors Cdkn2a, Cdkn2b, and other genes and is orthologous to the 9p21 breast cancer locus identified in genome-wide and candidate gene association studies. Emca8.2 harbors Cdkn2c and other genes and is orthologous to the 1p32 locus in humans that is frequently deleted in breast cancers. Both Emca8.1 and Emca8.2 harbor copy number variants that are orthologous to copy number variant regions in humans. Gene expression profiles were defined for mammary tissues from E2-treated ACI and ACI.BN-Emca8 rats to define the impact of Emca8 on gene expression and identify differentially expressed genes residing within Emca8.1 and Emca8.2. This study further illustrates the relevance of the ACI rat model of E2-induced mammary cancer for identifying novel genetic determinants of breast cancer susceptibility and defining the mechanisms through which estrogens contribute to breast cancer development. Cancer Prev Res; 6(1); 59–69. ©2012 AACR.