Roger J.B. King

Processing of insulin-like growth factor-II (IGF-II) by human breast cancer cells

MCF-7 cells transfected with human prepro-IGF-II cDNA secreted two large precursor forms of 22 and 15 kDa, with trace amounts of the mature 7 kDa IGF-II, suggesting that overexpression leads to saturation of processing and the secretion of precursors. The 15 kDa form was separated from 22 and 7 kDa IGF-II by cation-exchange chromatography. Intracellular IGF-II, detectable only in detergent buffers, existed in two forms of 24 and 22 kDa. Conditioned media from four other breast cancer cell lines (MDA-231, HBL-100, T47D and MCF-7 McG), all contained mature 7 kDa IGF-II with trace amounts (< 10%) of the 15 kDa IGF-II. Oestradiol induced IGF-II secretion in oestrogen-sensitive MCF-7 and T47D cells, but secretion was constitutively higher in oestrogen-unresponsive MDA-231 and HBL-100 cells. This indicates, for the first time, that oestrogen regulation of IGF-II peptide in breast cancer cells, and expression throughout all cell lines tested, would support the hypothesis that IGF-II has an autocrine regulatory function in breast cancer.

A discussion of the roles of oestrogen and progestin in human mammary carcinogenesis

Oestrogens and progestins are important for both the genesis of human breast cancer and growth of those tumours once formed. Their role at different stages of the neoplastic process are reviewed and discussed within the context of a change in sensitivity of epithelial cells during either initiation or promotion stages. Evidence favours, but does not conclusively prove, the view that progestins are the predominant mitogen for normal breast epithelium whilst oestrogen assumes that function in neoplastic epithelium. Alterations in oestrogen receptor levels could provide the key for such a change. There are insufficient data on physiological progestin concentrations to judge their effect on established cancer. Models for steroidal effects on cell proliferation and oestrogen and progestin receptor regulation that are based on endometrial data are not appropriate for breast.

Estrogen and progestin effects in human breast carcinogenesis

SummaryThe influences of estrogen and progestin on human mammary neoplasia are reviewed with a view to identifying what is known about their effects. Estrogens promote growth of established cancer and pharmacological levels of progestins induce remission.In vivo, highest proliferation of histologically normal mammary epithelium occurs in the progestogenic phase of the menstrual cycle or under the progestogenic influence of oral contraceptives. Little additional hard data exist to indicate whether progestins promote or inhibit human mammary carcinogenesis. Effects on proliferation, steroid receptor content and development are discussed together with interpretation of epidemiological data on risk factors that have hormonal components. Progestins may not be the benign or beneficial agents previously supposed, and there are virtually no data to suggest that they are antiestrogenic. It is hypothesized that carcinogenesis may be accompanied by increased sensitivity to estrogen, which provides a growth advantage to the tumor by maximizing use of the low estrogen concentrations encountered in the postmenopausal state.

Oestrogen receptor gene structure and function in breast cancer

The mechanisms underlying loss of oestrogen responsiveness in breast cancer are not well-defined. Potential mechanisms include loss of receptor expression, alterations in the oestrogen receptor (ER) gene producing proteins with abnormal function, or changes to receptor-dependent or -independent pathways controlling cell proliferation. Examination by Southern analysis of the ER gene in a series of ER-negative and -positive breast tumour biopsies failed to provide evidence of gross rearrangements and in only one of thirty seven tumour DNA samples was significant gene amplification observed. No restriction fragment length polymorphisms were detected for the restriction enzymes EcoR I, Pst I or Hind III. Methylation of the ER gene as assessed by Hpa II and Msp I restriction enzyme digests varied between tumours but the degree of methylation was not correlated with levels of expression of the receptor protein. Similar findings applied in a series of ER-negative and -positive breast cancer cell lines and clonal lines of MCF-7 cells, which were developed as an in vitro model for the acquisition of oestrogen and antioestrogen resistance. In this model there was no evidence that changes to ER receptor function and/or structure at the level of the ER gene, mRNA, ligand binding, and ability to induce progesterone receptor might account for the development of hormone resistance. However, the ability of ER to interact with a DNA sequence containing the vitellogenin promoter oestrogen response element, as assessed by gel retardation assay, was impaired in the clone showing the greatest degree of oestrogen and antioestrogen resistance.

Molecular modelling of the human estrogen receptor and ligand interactions based on site-directed mutagenesis and amino acid sequence homology

A molecular model of the human estrogen receptor is reported based on a new alignment with the alpha 1-antitrypsin sequence, a homologous protein of known crystal structure. The putative ligand binding site is situated roughly equidistant between the DNA binding and dimerization regions. This binding site contains a number of amino acid residues shown by site-directed mutagenesis to be associated with the binding of agonists and antagonists. This putative ligand binding pocket is well-defined within a loop of peptide, containing complementary amino acids for binding interactions with agonists and antagonists. A leucine-rich region, common to most steroid-binding proteins, is in an optimum position for dimerization leading to DNA interaction. It is likely that ligand binding influences dimerization and DNA interaction by a conformational change in the receptor via the transcriptional activation residues. This model suggests that ligand binding may affect the hydrogen bonding pattern such that transpeptide signalling is initiated. The model accommodates steroidal estrogens and antiestrogens as well as the non-steroidal partial antagonist, hydroxytamoxifen.

Relationship of HSP27 and oestrogen receptor in hormone sensitive and insensitive cell lines.

A 27 kDa heat shock (HSP27) has been analysed by immunoassay and immunoblotting in oestradiol sensitive and insensitive cells. Oestradiol growth responsive MCF7 and T47D human breast cancer cells and growth unresponsive variants derived therefrom have unaltered levels of HSP27 as well as retaining their oestradiol receptor phenotype. MCF7 cells induced to become doxorubicin resistant in culture lose both HSP27 and oestradiol receptor. Thus, in these three pairs of cells, HSP27 content parallels oestradiol receptor (ER). Analysis of a range of ER positive and negative human cell lines supports the positive relationship between HSP27 and ER. This included six ER positive and two ER negative breast tumour lines, one ER positive and one ER negative endometrial tumour cell line and seven ER negative human lines from other sites. One ER negative osteosarcoma line (HTB96) had appreciable levels of HSP27 that were unaffected after stable transfunction with an ER cDNA. Heat shock increases HSP27 levels in some but not all cell lines tested, the effect being inversely proportional to the basal (37 degrees C) content. In a mouse mammary tumour cell line, loss of androgen sensitivity was accompanied by loss of HSP27. Loss of HSP27 occurred in MCF7 cells made drug resistant to Novatrone, vincristine and etoposide as well as doxorubicin; no detectable change was seen in cells made resistant by 5 fluorouracil or X-irradiation. In ER positive ZR75 human breast tumour cells and in both ER negative and positive variants of the HTB96 human osteosarcoma line, the intracellular distribution of HSP27 was analysed. Over 96% of the HSP27 was in the cytosol fraction and the distribution was unaffected by incubation with oestradiol. HSP27 has been discussed in the literature under three different names p29, p24 and HSP27. The data presented in this paper are reviewed in the context of the previous data. It is concluded that there is a good but not absolute correlation between the presence of ER and high amounts of HSP27 but that low amounts of HSP27 are present in many ER negative cells. The correlations between HSP27 and drug resistance are more complex.

Measurement of oestradiol receptors by five institutions on common tissue samples

The soluble oestrogen-receptor content of common breast tumours has been measured by 5 different laboratories, each using their own assay procedure. Good agreement was achieved on whether a sample was positive or negative for oestrogen receptor. Qualitative differences between laboratories could be explained by differences in thiol-reagent content of assay medium and by the method of homogenization. Recommendations are made on some of the factors involved in the routine assay of receptors in breast tumours.

Hormonal regulation of HSP27 expression in human endometrial epithelial and stromal cells

HSP27 has been analysed immunohistochemically in epithelium and stroma of premenopausal endometria obtained at different stages of the menstrual cycle and in endometria from postmenopausal women receiving oestrogens either alone or in combination with a progestin. The data indicate that HSP27 is increased by oestrogen and inhibited by progestins in glandular epithelium but not stroma. Oestrogen does not increase HSP27 in stromal cells and HSP27 is only present in stromal decidual cells seen in the late secretory phase where it continues to be detectable until the tenth week of gestation. Hormonal regulation of HSP27 is clearly different in endometrial epithelium and stroma and additional regulatory factors may be important as oestrogen alone does not increase postmenopausal epithelial HSP27 to the levels seen in the proliferative phase of premenopausal samples.

Immunoaffinity purification and characterisation of p29—An estrogen receptor related protein

p29, a 29 kDa protein recognised by D5, a monoclonal antibody prepared against partially purified cytosolic estrogen receptor (ER), has been purified to homogeneity from ZR-75-1, a human breast cancer cell line. Ammonium sulphate fractionation followed by immunoaffinity chromatography on a three column system using Protein A-Sepharose coupled D5, produced purified p29. Silver stained SDS one-dimensional polyacrylamide gel electrophoresis (PAGE) and two-dimensional PAGE showed p29 to have been purified to homogeneity. Amino acid analysis showed no unusual characteristics. Partial N-terminal sequencing studies showed that purified p29 shared a 100% homology with the sequence of a pp89, murine cytomegaloviral protein.