Loss of heterozygosity in normal tissue adjacent to breast carcinomas.

Loss of heterozygosity (LOH) was detected in morphologically normal lobules adjacent to breast cancers. The most frequent aberration was at chromosome 3p22-25; of ten cases with this LOH in the carcinoma, six displayed the same LOH in adjacent normal lobules. This suggests that in a subset of sporadic breast cancers, a tumor suppresser gene at 3p22-25 may be important in initiation or early progression of tumorigenesis. Among sixteen breast cancers with LOH at 17p13.1 and five breast cancers with LOH at 11p15.5, one case each displayed the same LOH in adjacent normal lobules. Thus the molecular heterogeneity that characterizes invasive breast cancers may occur at the earliest detectable stages of progression.

Origin recognition complex binding to a metazoan replication origin.

The initiation of DNA replication in eukaryotic cells at the onset of S phase requires the origin recognition complex (ORC) [1]. This six-subunit complex, first isolated in Saccharomyces cerevisiae [2], is evolutionarily conserved [1]. ORC participates in the formation of the prereplicative complex [3], which is necessary to establish replication competence. The ORC-DNA interaction is well established for autonomously replicating sequence (ARS) elements in yeast in which the ARS consensus sequence [4] (ACS) constitutes part of the ORC binding site [2, 5]. Little is known about the ORC-DNA interaction in metazoa. For the Drosophila chorion locus, it has been suggested that ORC binding is dispersed [6]. We have analyzed the amplification origin (ori) II/9A of the fly, Sciara coprophila. We identified a distinct 80-base pair (bp) ORC binding site and mapped the replication start site located adjacent to it. The binding of ORC to this 80-bp core region is ATP dependent and is necessary to establish further interaction with an additional 65-bp of DNA. This is the first time that both the ORC binding site and the replication start site have been identified in a metazoan amplification origin. Thus, our findings extend the paradigm from yeast ARS1 to multicellular eukaryotes, implicating ORC as a determinant of the position of replication initiation.

In vitro properties of epithelial cell lines established from human carcinomas and nonmalignant tissue.

A series of 16 epithelial cell lines derived from human carcinomatous and nonmalignant epithelial tissues were characterized for markers known to correlate with neoplasia in various model systems. The goal of these studies was to determine which in vitro transformation properties were relevant to human epithelial malignancy. The parameters tested were 1) colony formation in Methocel and on confluent monolayers of normal human epithelial cells and fibroblasts and 2) tumor induction in immunosuppressed mice. Cell lines derived from normal tissue and nonmalignant tissue peripheral to cancerous lesions were generally negative for these properties, whereas the carcinoma-derived lines varied in their expression (some lines being more abnormal than others). In most cases, the lines derived from metastatic lesions had more abnormal properties than did those derived from primary carcinomas.

Phenotypic variability in anchorage-independent growth by a human breast tumor cell line.

Possible mechanisms responsible for expression of anchorage independence were investigated with the use of the human breast carcinoma cell line Hs578T. This phenotype was not stable and most likely occurred via alterations in gene expression, causing secretion of growth factor (s). Colony-forming efficiency in methylcellulose was proportional to initial plating density at low passage number and increased with passage in culture. At high passages, there was less sensitivity to initial plating density, suggesting less dependence on surrounding cells for feeding effects. Clonal variants isolated in methylcellulose maintained a higher plating efficiency when kept in suspension, and removal of selective pressure resulted in the loss of the high level of expression upon subsequent challenge in suspension. In contrast to other systems, anchorage-dependent clones acquired the ability to grow in methylcellulose after only one passage in culture. This suggests that rapid variation at rates not typical of classical somatic mutation plays a role in expression of anchorage independence. Exposure to 5-azacytidine, which decreases methylation of DNA and may thereby activate gene transcription, increased the incidence of anchorage-independent growth 20 times; whereas treatment with 6-azacytidine had no effect. Unconcentrated medium conditioned by cells previously exposed to 5-azacytidine or by cells at high passage stimulated growth in suspension by as much as 10.9 times. The factor(s) present in this conditioned medium was stable to heat up to 63 degrees C for 30 minutes and was larger than 12,000-14,000 mol wt as determined by dialysis.

Expression of c-myc, c-Ha-ras1, and c-erbB-2 proto-oncogenes in normal and malignant human breast epithelial cells.

Short-term cultures of normal human mammary epithelial cells were used to determine the extent to which c-myc, c-Ha-ras1, and c-erbB-2 proto-oncogenes were expressed in proliferating normal cells. This level of expression was compared with that of primary tumor cells, malignant effusion cells, or permanently established breast cancer cell lines. Pure preparations of epithelial organoids from seven different reduction mammoplasty tissue samples yielded proliferating normal epithelial cells upon short-term tissue culture. In every sample, proto-oncogene transcript levels increased upon short-term culture of the epithelial cells. These levels often exceeded by 10-fold the levels measured in uncultured organoids from the same tissue. In four of the seven cultured normal breast samples, at least one of the proto-oncogenes increased its expression to a level equaling or exceeding that found in a proliferating breast cancer cell line, MCF7. One effusion metastasis sample and two primary ductal adenocarcinomas were also examined for proto-oncogene expression. The effusion metastasis sample expressed high levels of c-erbB-2 messenger RNA, in accord with its amplified gene copy number; otherwise, the levels of proto-oncogene transcripts were low in unprocessed tumor and uncultured organoids, but they increased with proliferation of the tumor cells in culture. These results indicate that the variable expression of these proto-oncogenes observed in breast biopsy specimens needs to be controlled for cellular growth rate or proliferation index. Furthermore, these findings suggest that dysregulated proto-oncogene expression, rather than overexpression per se, needs to be evaluated as a possible mechanism contributing to the development of human breast cancer.

Cytologic evaluation of breast fluid in the detection of breast disease.

The availability and cell content of aspirated breast fluid from 1,706 women were evaluated to determine the usefulness of breast fluid cytology as an indicator of breast disease. A newly developed aspirator was used to obtain cells adequate for diagnosis from approximately 50% of the women tested. Fluids were most readily available from women between the ages of 30 and 50 years. Although abnormal cytologies were observed in all age groups, the relative proportion of abnormal specimens as well as the degree of abnormality increased as a function of age. More women over 40 years of age with a high risk of breast cancer had cells classified in the more abnormal categories than did women in the normal risk group. To localize otherwise occult lesions, contrast ductography was performed on all women with very abnormal ductal cells. Women with atypical hyperplastic cytologies most commonly had benign and premalignant breast disease at subsequent biopsy. Of 27 women with fluids classified as suspected carcinomas 18 (66%) had small carcinomas. These observations strongly suggested that the presence of atypical cells in aspirated breast fluids has important clinical application for early detection of breast cancer. One limitation of cytology was that the technique rarely detected carcinomas greater than 1 cm.

Heterogeneity for allelic loss in human breast cancer.

BACKGROUND: Loss of heterozygosity (LOH) at specific chromosomal regions in the tumor cells implicates the presence of tumor suppressor genes. However, it is also possible for an LOH to be randomly acquired and irrelevant to tumor development. PURPOSE: To determine whether a particular LOH in human breast carcinomas represents a loss of tumor suppressor gene or merely a random loss, we analyzed untreated primary breast cancers for LOH. METHODS: Ninety-eight primary human breast cancers from previously untreated patients were analyzed for LOH at 12 chromosomal regions including five randomly selected regions and seven regions previously reported in other cancer types and/or breast cancers. RESULTS: The baseline incidence of LOH was five out of 124 tests (4%) using randomly selected probes on chromosomes 1p, 2p, 4p, 11q, and Xq. Incidences of LOH significantly greater than background were seen in the following chromosomal regions: 22q (10 of 26 cases, 38%); 18p (five of 24 cases, 21%); 17p (30 of 59 cases, 51%); 13q (five of 14 cases, 36%); 3p (13 of 28 cases, 46%); and 1q (18 of 70 cases, 26%). In contrast to previous reports, the incidence of LOH was not significantly different from background for 11p15. In all cases, results were the same for metastatic lymph nodes and primary tumors, suggesting that the losses occurred early in the malignant progression. In cases with LOH at more than one locus, the same DNA sample often varied in degree of signal reduction for the missing alleles. CONCLUSION: These observations indicate the presence of both intertumor and intratumor heterogeneity for LOH.

Fibronectin production by human mammary cells.

Human mammary cells were examined for the presence of the high-molecular-weight surface glycoprotein fibronectin. Early passage mammary epithelial cell and fibroblast cultures from both carcinomas and normal tissues were tested for the presence of cell-associated fibronectin by immunofluorescence microscopy and for the synthesis and secretion of fibronectin by specific immunoprecipitation of metabolically labeled protein. In vivo frozen sections of primary carcinomas and normal tissues were tested for the localization of fibronectin by immunofluorescence microscopy. In contrast to the extensive fibrillar networks of fibronectin found in the fibroblast cultures, the epithelial cell cultures from both tissue sources displayed a pattern of cell-associated fibronectin characterized by powdery, punctate staining. However, the cultured epithelial cells, as well as the fibroblasts, secreted large quantities of fibronectin into the medium. Putative myoepithelial cells also displayed extensive fibrillar networks of fibronectin. The difference in cell-associated fibronectin distribution between the epithelial cells and the fibroblasts and putative myoepithelial cells provided a simple means of quantitating stromal and myoepithelial cell contamination of the mammary epithelial cells in culture. In vivo, normal tissues showed fibronectin primarily localized in the basement membrane surrounding the epithelial cells and in the stroma. Most primary carcinomas displayed powdery, punctate staining on the epithelial cells in addition to the fibronectin present in the surrounding stroma.

Epithelial cell cultures from normal and cancerous human tissues.

Thirty epithelial cell strains were isolated from human carcinomas and normal epithelial tissues by collagenase digestion and selective removal of fibroblasts with trypsin-Versene. Most strains were obtained from metastatic carcinomas or epithelia of the urinary and intestinal tracts. The success rate for growth of both neoplastic and normal tissues (excluding skin) was 38%. Six of these strains showed gross morphologic and chromosome changes typical of malignant cells. Nine resembled normal epithelium. The other 15 exhibited some degree of morphologic change from normal.

Response to doxorubicin of cultured normal and cancerous human mammary epithelial cells.

Epithelial cells were isolated and cultured from a number of human mammary specimens of both cancerous and noncancerous origin. Doxorubicin (Dx) sensitivity was measured at second passage with the use of a highly efficient clonogenic assay. For 23 different tumor specimens derived from patients without previous chemotherapy, the drug concentrations required to kill 50% of the cells varied approximately 35-fold. In contrast, for 11 tumor specimens from patients who relapsed after regimens containing Dx, the drug concentration for 50% survival varied only fivefold and the dose-response curves for these specimens clustered at the more resistant end of the spectrum. A wide range of sensitivities was also observed among 13 noncancerous mammary specimens; however, tumor tissue and noncancerous tissue from the same donor were similar. When cultures were subjected to drug incubation periods of 1 and 4 hours, dose-response curves were superimposable when plotted as a function of drug concentration multiplied by time.

Tumor angiogenesis as a prognostic assay for invasive ductal breast carcinoma.

BACKGROUND: Tumor angiogenesis, as assayed by microvessel density, has been proposed as an independent prognostic marker for clinical outcome in breast cancer patients. PURPOSE: The present study evaluated the microvessel density assay and assessed its utility, alone and together with the evaluation of other tumor characteristics, in predicting outcome in patients with invasive ductal carcinomas. METHODS: In a blinded design, cases of invasive ductal carcinoma were selected from a registry containing the records of and tumor specimens from 386 breast cancer patients treated at the Massachusetts General Hospital from 1977 through 1982. After the exclusion of ineligible patients and inadequate specimens, 220 patients were included in the study; their median time of follow-up was 11.5 years. Half of these patients (n = 110) were positive for axillary lymph node metastases. Histologic sections of the tumors were stained immunocytochemically for factor VIII, a coagulation protein expressed by blood vessel endothelium, and for p53 protein. Independently, two analysts counted microvessels in three microscope fields selected from separate vascular regions of the tumor. Variability in microvessel scores between analysts and among different fields of the same tumor was summarized by the coefficient of variation. The kappa statistic tested for agreement between the analysts while correcting for chance agreement. The effects of tumor characteristics on metastasis-free survival and overall survival were tested univariately by the Harrington-Fleming rank test procedure. The effect of multiple factors on survival was tested under a Cox multivariate proportional hazards model. RESULTS: Microvessel count showed considerable variability between the two analysts and among regions within each tumor, with an overall concordance for tumor classification of 73%. Univariate analysis revealed no association between microvessel count and any other tumor or patient characteristic. Multivariate analysis indicated, for these patients, that nodal status and p53 staining predicted metastasis-free survival and that nodal status, estrogen receptor (ER) status and tumor grade predicted overall survival. CONCLUSIONS: Microvessel count showed much variation among different regions of each tumor. It did not predict metastasis-free survival or overall survival. Nodal status was the most powerful criterion to stratify these patients with invasive ductal carcinoma of the breast into different survival groups. Only ER status, tumor grade, and p53 staining had additional prognostic utility for these patients after they had been stratified by nodal status.

Immortalization in culture: occurrence at a late stage in the progression of breast cancer.

The properties in culture of 3 breast cancer effusion metastases, obtained over approximately 2 years from the same patient, were examined. Despite repeated attempts with cryopreserved cells, only the last specimen reproducibly exhibited immortality in culture; the first 2 specimens grew initially but failed to develop into cell lines. Each specimen was unique in morphology and growth properties, although karyotypic markers indicated a common origin. Aberrations of chromosomes 1 and 11 marked these near-diploid cells, and further structural alterations of chromosome 11 accompanied the transition of biological properties observed in the third specimen.

Two syngeneic cell lines from human breast tissue: the aneuploid mammary epithelial (Hs578T) and the diploid myoepithelial (Hs578Bst) cell lines.

We characterized two human cell lines (Hs578T and Hs578Bst), which provide several unique features that should be useful in the study of breast disease. Hs578T, derived from a carcinosarcoma, is epithelial in origin. Hs578Bst, established from normal tissue peripheral to the tumor, is myoepithelial in origin. This is the first report of companion cell lines, one malignant and one normal, established from the same organ.

Nuclear ultrastructure of epithelial cell lines derived from human carcinomas and nonmalignant tissues.

The nuclear ultrastructure of sixteen human epithelial cell lines has been characterized in detail by transmission electron microscopy. The cell lines were derived from normal tissues, nonmalignant tissues of cancerous organs, primary carcinomas, and metastatic carcinomas. Every cell section on a grid containing a clearly defined nucleus and nucleolus was scored blindly utilizing a checklist of markers. The goal of these studies was to determine whether any ultrastructural markers consistently distinguished the different stages of malignant progression represented among the lines. Nuclear bodies and perichromatin granules were found in all lines derived from cancer and were not observed in any nonmalignant lines. Nuclear envelope dilation was seen in all lines derived from cancerours organs as well as from malignant tissues but not in any lines derived from normal tissue. Margination of chromatin, irregularity of nuclear outline, redistribution of nucleolar components, and margination were expressed slightly by the normal lines, to variable degrees by the lines derived from cancerous organs, and to a much greater extent by all lines derived from malignant tissues. No differences were found between lines derived from primary carcinomas and those derived from metastatic specimens. There were no ultrastructural differences comparing subconfluent and confluent cells or cells at different passage levels. In addition, the nuclear ultrastructure of a malignant line in culture was similar to that of a tumor induced by those cells in an immunosuppressed mouse.

Production of fibronectin by human epithelial cells in culture.

Human epithelial cell lines derived from both carcinomatous and nomalignant tissues were characterized with respect to the presence and distribution of fibronectin by immunofluorescence microscopy. In cell lines derived from nonmalignant tissues or from primary carcinomas, fibronectin was found predominantly in an extracellular matrix. In contrast, cell lines derived from metastatic carcinomas displayed very little or no fibronectin. Metabolic labeling studies indicated that a positive line synthesized fibronectin de novo rather than absorbing the protein from the media. Negative lines neither synthesized fibronectin nor secreted it into the culture fluid, suggesting that they were not producing fibronectin. Evidence is presented that cells in culture change their properties after extensive subculture since a small amount of fibronectin in an extracellular matrix was observed after extensive subculture of two metastatic lines that were originally negative.

Contiguous patches of normal human mammary epithelium derived from a single stem cell: implications for breast carcinogenesis.

Tissue clonality can be assessed in females by analyzing the methylation status of polymorphic DNA markers on X-linked genes because extensive de novo methylation of one allele at the preimplantation stage is associated with its permanent inactivation. We applied X chromosome inactivation toward understanding human breast morphogenesis by examining the nonmalignant breast epithelium from two reduction mammaplasties and a mastectomy. We found that entire lobules and large ducts of normal breast tissue have the same X chromosome inactivated, suggesting that they are derived from the same stem cell. The regions of inactivation of a particular X chromosome do not extend over an entire breast, so that ducts and lobules with opposite chromosomes inactivated are present within a single breast. Potential relevance of these observations for malignant transformation is discussed.

Growth factor messenger RNA expression by human breast fibroblasts from benign and malignant lesions.

Breast tumors are a complex mix of epithelial, stromal, and vascular elements. We examined primary cultures of breast fibroblasts derived from benign and malignant lesions for expression of various growth factors. All fibroblast cultures, regardless of whether they were derived from benign or malignant lesions, expressed platelet-derived growth factor A chain, basic fibroblast growth factor, fibroblast growth factor 5, and transforming growth factor beta 1 mRNA. None expressed platelet-derived growth factor B chain or transforming growth factor alpha mRNA. However, examination of mRNA expression for the insulin-like growth factors revealed that 7 of 8 fibroblasts derived from benign lesions expressed insulin-like growth factor I (IGF-I) mRNA, while only 1 of 9 fibroblasts derived from malignancies expressed IGF-I mRNA. The opposite picture was seen for insulin-like growth factor II (IGF-II) mRNA expression, in which 1 of 9 benign-derived fibroblasts expressed IGF-II mRNA, while 5 of 9 malignant-derived fibroblasts expressed IGF-II. This correlated with previous in situ hybridization data, which showed IGF-I mRNA expression confined to the stroma of benign breast tissue. PDGF treatment of tumor fibroblasts resulted in a 3-fold increase in IGF-II mRNA. Thus there was an apparent dichotomy between IGF-I mRNA expression in the majority of fibroblasts derived from benign lesions and IGF-II mRNA expression in the majority of tumor-derived fibroblasts. Since the insulin-like growth factors are potent mitogens for breast tumor epithelial cells, this further supports the notion of a paracrine growth-promoting role for the insulin-like growth factors in breast lesions and suggests that IGF-II may be the more important growth promoter in malignant lesions.

Malignant breast epithelium selects for insulin-like growth factor II expression in breast stroma: evidence for paracrine function.

Paracrine interactions between stromal and epithelial cells are important influences on the growth and malignant behavior of breast cancers. Insulin-like growth factors I and II (IGF-I and II) are expressed by fibroblasts in benign and malignant breast lesions, and both are strong mitogens for a number of breast cancer epithelial cell lines in vitro. We have analyzed the stromal mRNA expression of IGF-I and IGF-II in matched sets of fibroblast cell lines derived from three locations in the affected breast of eight patients with breast cancer: (a) the breast tumor itself; (b) surrounding normal breast tissue; and (c) overlying breast skin. IGF-I expression was easily detected in all fibroblasts derived from normal breast tissue. In general, lesser amounts of IGF-I mRNA were detected in fibroblasts derived from breast tumors or skin. In contrast, IGF-II expression was detected at very low levels in only 3 of 8 normal breast fibroblasts, but was present in 6 of 8 tumor fibroblasts. IGF-II mRNA was expressed in all skin fibroblasts tested. IGF-II-negative stromal fibroblasts from normal breast, which were plated at low density and allowed to grow to confluence in the presence of MCF-7 breast tumor epithelial cells, demonstrated a marked increase in IGF-II mRNA expression. IGF-II in situ hybridization studies confirmed that IGF-II expression is seen at high levels in stroma of many invasive breast cancers but not normal breast. We conclude that paracrine influences, mediated by soluble factors released by breast tumor epithelium, are able to specifically increase expression of IGF-II in breast stroma, most likely by a process of clonal selection.

Selective cell culture of primary breast carcinoma.

We have used culture conditions which simulate the microenvironment of breast tumors for the isolation and propagation of primary breast tumor cells in vitro. In this monolayer setup, the mixture of cells dissociated from primary breast tumors is subjected to self-created gradients of oxygen and nutrients as well as metabolic waste and extracellular pH. The tumor populations isolated under these novel conditions have displayed phenotypic properties characteristic of breast carcinomas, including homogeneous expression of cytokeratin 19, and increased mitochondrial retention of the cationic dye rhodamine 123. Nonmalignant cultures from reduction mammoplasty were unable to survive these conditions. One tumor population which reached passage 10 was aneuploid for chromosomes 15 and 17, and displayed a p53 mutation in exon 8. These studies strongly suggest that the culture conditions described here can suppress the growth of normal breast cells, thereby allowing selective isolation of some populations of slow-growing primary tumor cells in vitro.

Partial enzymatic degradation of stroma allows enrichment and expansion of primary breast tumor cells.

The goal of this study was to isolate and expand tumor cells in culture that closely resemble invasive cells in primary breast carcinoma tissue. Based on the hypothesis that invasive tumor cells are released more readily upon digestion with connective tissue-degrading enzymes because they are not confined within a basement membrane, we have designed a novel procedure for their isolation. Using this method, we have successfully expanded in culture aneusomic tumor cells from several primary breast tumors. Twenty nine of 44 (66%) specimens processed yielded proliferative and passageable cultures of up to 2 x 10(7) cells. The original tumor tissue and cultures derived therefrom were compared for aneusomy and the abnormal expression of the erb-B2, p53, and bcl-2 gene products. Remarkable similarities were observed. However, some intratumor heterogeneity in chromosome content was found between touch preparations and cultured cells. Overexpression of erb-B2 was observed in the vast majority of cases (16 of 20), suggesting that this phenotype may be important for dysregulated proliferation in vitro. The simple and rapid method described in this report could enable routine expansion of primary breast tumors and provide adequate numbers of viable cells for studying and manipulating their functional characteristics.