Fibroblast growth factor 7, secreted by breast fibroblasts, is an interleukin-1beta-induced paracrine growth factor for human breast cells.

Keratinocyte growth factor/fibroblast growth factor 7 (KGF/FGF7) is known to be a potent growth factor for mammary cells but its origin, cellular targets and mode of action in the breast are unclear. In this study, we carried out studies to determine the localisation of FGF7 and its receptor, and the related growth factor FGF10. We also determined the factors that regulate FGF7 release from stromal cells and the effects of FGF7 on normal and neoplastic breast cells. Using an FGF7-specific antibody which does not react with the FGF7 heparan sulphate proteoglycan (HSPG)-binding site, we showed epithelial and myoepithelial immunohistochemical staining in normal breast sections, and epithelial staining in breast carcinomas. Stromal staining was also detected in some lobular carcinomas as well as a subset of invasive ductal carcinomas. FGF10 and FGF receptor (FGFR)2 immunostaining showed a similar epithelial expression pattern, whereas no stromal staining was observed. We purified normal breast stromal, epithelial and myoepithelial cells and showed that FGF7 stimulated proliferation of both epithelial cell types, but not stromal fibroblasts. We also examined the effects of FGF7 on Matrigel-embedded organoids, containing both epithelial and myoepithelial cells, and showed FGF7 induced an increase in cellular proliferation. Furthermore, conditioned medium derived from stromal cells was shown to increase the proliferation of normal and neoplastic breast epithelial cells, which could be abolished by a neutralising antibody to FGF7. Finally, we showed that interleukin-1beta, but not oestradiol or other oestrogen receptor ligands, caused a dose-related FGF7 release. Further results also indicate that the epithelial localisation of FGF7 and FGF10 in breast tissue sections is likely to be due to their binding to their cognate receptor. In summary, our findings suggest that FGF7 is a paracrine growth factor in the breast. FGF7 is produced by the breast stromal fibroblasts and has profound proliferative and morphogenic roles on both epithelial and myoepithelial cells.

Fibroblast growth factor 8 is expressed at higher levels in lactating human breast and in breast cancer.

Fibroblast growth factor 8 can transform NIH3T3 cells and its expression has been found to be associated with breast and prostate cancer. Following our finding that fibroblast growth factor 8 mRNA expression is increased in breast cancer, we have undertaken an immunohistochemistry study of fibroblast growth factor 8 expression in a series of human breast tissues and other normal tissues. Our findings confirm increased expression of fibroblast growth factor 8 in malignant breast tissue but also show significant fibroblast growth factor 8 expression in non-malignant breast epithelial cells. No significant difference in fibroblast growth factor 8 expression was found between different grades of ductal carcinoma, lobular carcinoma and ductal carcinoma in-situ or cancer of different oestrogen receptor, progesterone receptor or nodal status. The highest levels of fibroblast growth factor 8 expression were found in lactating breast tissues and fibroblast growth factor 8 was also detected in human milk. A survey of other normal tissues showed that fibroblast growth factor 8 is expressed in the proliferative cells of the dermis and epithelial cells in colon, ovary fallopian tube and uterus. Fibroblast growth factor 8 appears to be expressed in several organs in man and appears to have an importance in lactation.

Altered intracellular localization of fibroblast growth factor receptor 3 in human breast cancer.

Immunohistochemical staining of human breast tissues, using an antibody against fibroblast growth factor receptor 3 [FGFR-3], showed differences in cellular distribution. Both malignant and non-malignant epithelial cells contained FGFR-3 immunoreactivity, but myoepithelial cells and stroma were negative. The staining pattern in malignant epithelial cells was predominantly nuclear, whereas epithelial cells in normal breast tissue showed both cytoplasmic and nuclear elements. Reverse transcription-polymerase chain reaction (RT-PCR) revealed two isoforms of FGFR-3 corresponding to the FGFR-3-IIIb variant and a previously described exon-deleted nuclear form of FGFR-3, which were present in both malignant and non-malignant epithelial cells. The higher level of nuclear staining and loss of cytoplasmic staining seen in malignant epithelial cells did not correspond to an increase in expression of the exon-deleted form of FGFR-3, nor to any detectable activating point mutations. Since receptor activation can result in its movement to a perinuclear localization, an alternative explanation for the redistribution of FGFR-3-IIIb could be different degrees of activation by a ligand (FGF1 or FGF9). No FGF9 was detected by immunohistochemistry in breast tissues. FGF1, however, is present in the majority of breast cancers and a different tissue distribution of FGF1 was found in breast tissues, showing predominantly nuclear, or a mix of nuclear and cytoplasmic FGFR-3. The difference in FGFR-3 staining patterns may implicate this ligand-receptor pair in breast cancer.

Retroviral infection of the FGF2 gene into MCF-7 cells induces branching morphogenesis, retards cell growth and suppresses tumorigenicity in nude mice.

FGF2 (basic fibroblast growth factor) is a multifunctional growth factor and exhibits diverse function in different cell types. In breast, loss of FGF2 expression is associated with malignant progression. In order to understand the role of FGF2 in maintenance of normal breast structures and control of cell growth, we restored FGF2 expression in the breast cancer cell line MCF-7. The FGF2 retrovirally infected MCF-7 cells (MCF-7.F2.5) not only expressed FGF2 in cytoplasm and nuclei, but also released FGF2 into culture medium both on plastic and in Matrigel conditions. The MCF-7.F2.5 cells formed branches in Matrigel and this effect was abolished by the addition of a neutralising anti-FGF2 antibody or function blocking antibodies to alpha2, alpha3 and beta1 integrins. Furthermore, MCF-7.F2.5 cells lost their ability for anchorage-independent growth in soft agar. When MCF-7 and MCF-7.F2.5 cells were injected into nude mice, there was a 1.6- to 3.2-fold reduction of tumour volume with MCF-7.F2.5 cells in comparison with the parental MCF-7 cells. MCF-7.F2.5 cells also demonstrated a reduction in oestrogen receptor-alpha (ERalpha) both in vitro and in vivo. Our results suggest that introduction of the FGF2 gene into MCF-7 cells altered the malignant tumour cells towards a more benign phenotype in vitro and in vivo.

Selecting radial basis function network centers with recursive orthogonal least squares training.

Recursive orthogonal least squares (ROLS) is a numerically robust method for solving for the output layer weights of a radial basis function (RBF) network, and requires less computer memory than the batch alternative. In this paper, the use of ROLS is extended to selecting the centers of an RBF network. It is shown that the information available in an ROLS algorithm after network training can be used to sequentially select centers to minimize the network output error. This provides efficient methods for network reduction to achieve smaller architectures with acceptable accuracy and without retraining. Two selection methods are developed, forward and backward. The methods are illustrated in applications of RBF networks to modeling a nonlinear time series and a real multiinput-multioutput chemical process. The final network models obtained achieve acceptable accuracy with significant reductions in the number of required centers.

Increased expression of fibroblast growth factor 8 in human breast cancer.

Fibroblast growth factor 8 (FGF8) is an important developmental protein which is oncogenic and able to cooperate with wnt-1 to produce mouse mammary carcinoma. The level of expression of FGF8 mRNA was measured in 68 breast cancers and 24 non-malignant breast tissues. Elevated levels of FGF8 mRNA were found in malignant compared to non-malignant breast tissues with significantly more malignant tissues expressing FGF8 (P=0.019) at significantly higher levels (P=0.031). In situ hybridization of breast cancer tissues and analysis of purified populations of normal epithelial cells and breast cancer cell lines showed that malignant epithelial cells expressed FGF8 mRNA at high levels compared to non-malignant epithelial and myoepithelial cells and fibroblasts. Although two of the receptors which FGF8 binds to (FGFR2-IIIc, FGFR3-IIIc) are not expressed in breast cancer cells, an autocrine activation loop is possible since expression of fibroblast growth factor receptor (FGFR) 4 and FGFR1 are retained in malignant epithelial cells. This is the first member of the FGF family to have increased expression in breast cancer and a potential autocrine role in its progression.

The human mammary gland basement membrane is integral to the polarity of luminal epithelial cells.

We show that myoepithelial cell basement membrane derived E3 and E8 domains of laminin-1 are capable of polarizing luminal epithelial cells with regard to epithelial membrane antigen localization. This event is dependent on the alpha6 integrin and results in aggregation and phosphorylation of the tyrosine residues of the focal adhesion kinase complex. We also demonstrate that uncultured normal luminal epithelial cells synthesize normal levels of beta and gamma laminin chains and reduced levels of alpha chains mRNA in common with malignant epithelial cells. In contrast normal myoepithelial cells synthesize all three constituent chains of laminin-1. Therefore in breast cancer the absence of myoepithelial cells could result in a lack of laminin alpha chains which may contribute to loss of polarity of malignant epithelial cells.

A paracrine role for myoepithelial cell-derived FGF2 in the normal human breast.

We have studied separated normal human breast epithelial and myoepithelial cells for the presence of basic fibroblast growth factor (FGF2) and its receptors, both low (heparan sulfate proteoglycans) and high affinity (FGFR1), and for the effects of FGF2 on the proliferation of both cell types. Our results indicate that these cells differ markedly in their synthesis and response to FGF2. We found, using PCR of purified cell populations, mRNA for FGF2 only in the myoepithelial cells, whereas immunostaining and Western blotting results demonstrated the presence of FGF2 protein in both epithelial and myoepithelial cells. FGF2 had no effect on the proliferation of myoepithelial cells, but it did maintain the survival of the separated epithelial cells in low serum and stimulate their growth in 5% and 10% FCS. Immunostainable FGFR1 was present in epithelial cells and, to a lesser extent, in myoepithelial cells. Low-affinity binding sites for FGF2 were synthesized by epithelial and myoepithelial cells, but myoepithelial cells possessed a greater proportion of higher-affinity heparan sulfate proteoglycans. These results indicate that myoepithelial cell-derived FGF2 may be an important paracrine factor controlling epithelial cell survival and growth in the normal human breast.

The detection of micrometastases in the peripheral blood and bone marrow of patients with breast cancer using immunohistochemistry and reverse transcriptase polymerase chain reaction for keratin 19.

The aim of this study was to determine whether reverse transcriptase polymerase chain reaction (RT-PCR) for keratin 19 (K19) provides additional information when combined with immunohistochemistry when used to detect micrometastases in blood and bone marrow in patients with primary breast cancer. We studied 78 patients with breast cancer who had no evidence of distant metastases. We collected blood and bone marrow, separated the mononuclear fraction and carried out RT-PCR and immunohistochemistry for K19. RT-PCR was done by two 40-cycle rounds using nested primers. In initial experiments, RT-PCR was shown to be capable of detecting one tumour cell in one million normal bone marrow cells, which was at least 10 times more sensitive than immunohistochemistry, while retaining specificity. Five per cent of the peripheral blood and 22% of the bone marrow samples contained K19 positive cells by immunohistochemistry staining. Using RT-PCR, these proportions increased to 25% and 35%, respectively. This represents a significantly greater detection frequency (P < 0.001 and P = 0.03, respectively). RT-PCR for K19 is a more sensitive method for detecting micrometastases in patients with primary breast cancer when compared with immunohistochemistry.

Separated human breast epithelial and myoepithelial cells have different growth factor requirements in vitro but can reconstitute normal breast lobuloalveolar structure.

In order to investigate the specific factors controlling the growth of normal breast cell types, purified populations of human breast epithelial and myoepithelial cells from reduction mammoplasties were grown in primary culture in three defined media and their response to foetal calf serum (FCS), epidermal growth factor (EGF) and basic fibroblast growth factor (FGF2) measured using MTT growth assays. Epithelial and myoepithelial cells differed markedly in their growth requirements. Whereas epithelial cell survival was dependent on the presence of FCS, myoepithelial cell growth was dramatically inhibited by serum. EGF and FGF2 were mitogenic for epithelial cells but not myoepithelial cells, the addition of insulin being the only essential supplement required for myoepithelial cell growth. Heparin inhibited FGF2-stimulated epithelial cell growth but also basal myoepithelial cell proliferation and this inhibition could be overcome by the addition of EGF. Neutralizing antibodies to EGF also inhibited basal myoepithelial cell growth. This suggests the possibility of an autocrine role for a heparin-binding member of the EGF family in the growth of myoepithelial cells. Purified cells combined to form lobuloalveolar structures when incubated in a reconstituted basement membrane matrix (Matrigel) in the presence of EGF and FGF2.

Expression of keratinocyte growth factor and its receptor in human breast cancer.

The level of expression of keratinocyte growth factor (KGF) mRNA has been measured in human breast cell lines, purified populations of epithelial cells, myoepithelial cells and fibroblasts from reduction mammoplasty tissue and a panel of 42 breast cancers and 30 non-malignant human breast tissues using a semiquantitative reverse transcriptase polymerase chain reaction (RT-PCR) procedure. We found similar levels of KGF mRNA in malignant and non-malignant breast tissues. The study of the amount of KGF mRNA in breast cell lines and purified populations of cells revealed that fibroblasts are the predominant source of KGF with malignant and non-malignant epithelial cells containing very low levels of KGF mRNA. We have examined the distribution of fibroblast growth factor receptor (FGFR)-2-IIIb, which is a high-affinity receptor for KGF and find that it is present on malignant and non-malignant epithelial cells. The level of FGFR-2-IIIb present on breast cancer cell lines was sufficient for KGF stimulation of breast cancer cell proliferation. Other members of the fibroblast growth factor family have been either not expressed in the human breast (FGF3, FGF4) or have been found at much reduced levels in breast cancer (FGF1, FGF2) and this is the first member of the family to potentially influence the progression of breast cancer through stimulation of cell division.

The location of acidic fibroblast growth factor in the breast is dependent on the activity of proteases present in breast cancer tissue.

Acidic fibroblast growth factor (FGF1) and two of its receptors, FGFR1 and FGFR4, were localized in cryostat sections of normal, benign and malignant human breast tissue by immunohistochemistry. Without pretreatment, FGF1 staining was mainly seen in normal epithelial cells. However, polymerase chain reaction (PCR) analysis and immunoblotting of isolated normal epithelial and myoepithelial cells showed FGF1 mRNA and protein to be present in both cell types. Following incubation of frozen sections at 37 degrees C in phosphate-buffered saline, FGF1 staining was also revealed in myoepithelial cells and basement membrane adjacent to carcinoma cells. Treatment with protease inhibitors demonstrated that this effect was due to the activity of an endogenous protease. In contrast, FGF1 staining was found to be associated with the stroma adjacent to malignant cells only in the presence of protease inhibitors. FGFR1 and FGFR4 immunostaining was localized to both normal and malignant epithelial cells and to a lesser extent to myoepithelial cells. There was no difference in the staining intensity for the FGF receptors between normal and cancer samples. The change in location of FGF1 between normal and malignant tissues and the sensitivity of stored FGF1 to the action of endogenous proteases raises the possibility of both autocrine and paracrine roles for FGF1 in the normal and malignant human breast.

Presence of exon 5-deleted oestrogen receptor in human breast cancer: functional analysis and clinical significance.

A variant form of the human oestrogen receptor (ER) mRNA lacking sequences encoded within exon 5 has been described (Fuqua SAW, Fitzgerald SD, Chamness GC, Tandon AK, McDonnell DP, Nawaz Z, O'Malloy BW, McGuire WL 1991, Cancer Res 51: 105-109). We have examined the expression of the exon 5-deleted ER (HE delta5) mRNA variant in breast biopsies using reverse transcriptase polymerase chain reaction (RT - PCR). HE delta5 mRNA was present in only 13% of non-malignant breast tissues compared with 32% of carcinomas (95% CI, P=0.05). Presence of the HE delta5 mRNA was associated with the presence of immunohistochemically detected ER (P=0.015) and progesterone receptor (PR) (P=0.02). There was a positive correlation between the presence of HE delta5 and disease-free survival (P=0.05), suggesting that the presence of HE delta5 may be an indicator of better prognosis. We have raised a monoclonal antibody specific to the C-terminal amino acids of HE delta5. This antibody recognized the variant but not the wild-type ER protein. We show that HE delta5 protein is present in breast cancer using immunohistochemical techniques. We also analysed trans-activation by HE delta5 in mammalian cells and showed that, in MCF-7 cells, HE delta5 competes with wild-type ER to inhibit ERE-dependent trans-activation. Our results indicate that this variant is unlikely to be responsible for endocrine resistance of breast cancer, but its presence at both the mRNA and protein level suggest that it may, nevertheless, be involved in regulating the expression of oestrogen-responsive genes in breast cancer.

Fibroblast growth factor 2 in breast cancer: occurrence and prognostic significance.

This paper examines the expression of fibroblast growth factor 2 (FGF-2) in the malignant human breast. Semiquantitative reverse transcriptase-polymerase chain reaction (RT-PCR) was used to assess the level of expression of FGF-2 in a series of 51 patients clinically followed up for a median of 84 months (Luqmani et al, 1992). Immunohistochemistry and Western blotting were used to show that the level of FGF-2 in breast tissues correlated with the amount of FGF-2 mRNA. FGF-2 was present in both malignant and non-malignant breast, although less was expressed in malignant tissues as determined by all three methods. Immunohistochemistry on frozen sections of breast tissue showed expression of FGF-2 in myoepithelial and epithelial cells in non-malignant samples and generally lower or undetectable levels of staining in malignant epithelial cells. The results obtained by immunohistochemistry correlated well with RT-PCR data showing similar levels of FGF-2 and FGF-2 mRNA expression in samples. No correlation was found between FGF-2 mRNA expression and T stage, nodal status or oestrogen receptor status. However, Kaplan-Meier survival plots show that higher levels of FGF-2 are associated with improved overall and disease-free survival. We suggest that FGF-2 expression may have value as a prognostic indicator in breast cancer.

Down-regulation of a novel form of fibroblast growth factor receptor 1 in human breast cancer.

Monoclonal antibodies against two epitopes of FGFR-1 have been used to investigate FGFR-1 expression in the normal and neoplastic human breast. Different forms are detected in the different cell types constituting the normal breast. Moreover, breast cancer cells lack one form of FGFR-1. Western blot analysis showed 115-kDa and 106-kDa forms of FGFR-1 within the human breast. The 115-kDa band corresponds to the beta form of FGFR-1, whereas the 106-kDa band is truncated at the carboxyl terminus. The 106-kDa form of FGFR-1 is the major form present in breast fibroblasts and myoepithelial cells, whereas epithelial cells contain equal amounts of the 115-kDa and 106-kDa forms. Breast cancer cells, however, appear to contain only the 115-kDa form of FGFR-1. This expression pattern is reflected in malignant and non-malignant tissue samples. Using reverse transcription polymerase chain reaction (RT-PCR) analysis, we have shown that the 106-kDa FGFR-1 isoform is not the previously described alpha 2 receptor that arises from a 25-base pair insertion in the second kinase domain. It is probable that the 106-kDa FGFR-1 has different signalling properties to the full-length receptor, having lost at least one tyrosine at amino acid 766, which is required for phospholipase C activation. This form of FGFR-1 appears to be lost in all breast cancer cells analysed and its absence may have a bearing on malignancy.

Effect of basic fibroblast growth factor-saporin mitotoxin on breast cancer cell lines in vitro.

Fibroblast growth factor receptors are widely expressed on breast cancer cells and we report a preliminary study to determine whether these could be useful as potential targets for delivery of a cytotoxic fibroblast growth factor 2-saporin conjugate. We show that this mitotoxin conjugate can displace I-125-fibroblast growth factor 2 binding though with reduced affinity compared to unlabeled fibroblast growth factor 2. For 4 out of 5 cell lines it is an effective inhibitor of cell growth, and cytotoxic for at least 2 of the lines. Inhibitory effects did not depend on responsiveness of cells to fibroblast growth factor 2. This activity was not achieved with free saporin. There may be potential uses for this conjugate in both experimental systems to study receptor function and subsequent processing, and also in clinical settings to eliminate breast cancer cells.

Inhibitory effects of activin on the growth and morpholgenesis of primary and transformed mammary epithelial cells.

Activin is a member of the transforming growth factor beta superfamily, which is known to have activities involved in regulating differentiation and development. By using reverse transcription-PCR analysis on immunoaffinity-purified human breast cells, we have found that activin beta a and activin type II receptor are expressed by myoepithelial cells, whereas no expression was detected in other breast cell types. In examining 15 breast cell lines, we have found only four (HBL-100, MCF10-A, PMC-42, and BT 20) to be positive for activin beta a mRNA, whereas all expressed the activin type II receptor. Furthermore, we have found activin A to be a potent growth inhibitor of MCF- 7 cells (at 2 ng/ml), where it causes an arrest in G(1). Activin A does not appear to have an effect on the cell cycle of primary myoepithelial or luminal cells. However, we demonstrate that activin is an inhibitor of tubule formation by human mammary organoids in vitro. These are the first observations of activin and activin receptor in the normal human breast and in human breast cell lines and suggest a role for activin in mammary cell growth and morphogenesis.

Expression of fibroblast growth factor 1 is lower in breast cancer than in the normal human breast.

We have measured the amount of fibroblast growth factor 1 (FGF-1) mRNA and protein in primary breast cancers and non-malignant breast tissue and have found greatly reduced levels in breast cancer compared with non-malignant tissue. A total of 116 breast cancers and 37 biopsies taken from non-malignant breast were compared for FGF-1 mRNA expression using reverse transcriptase-polymerase chain reaction (RT-PCR) and significantly lower levels were found in the cancer tissues (P < 0.001). These findings were confirmed at the protein level where four out of five breast cancers contained no detectable FGF-1 and a fifth cancer had a low level of FGF-1 compared with three samples from reduction mammoplasties. Similar results were obtained from breast cell lines in which 80% of cancer cell lines had very low levels of FGF-1, whereas all non-malignant breast cell lines contained higher levels of FGF-1. Immunohistochemical analysis indicated that FGF-1 was present in the luminal epithelial cells of the non-malignant breast but was absent from cancer cells. The decreased levels of FGF-1 in breast cancer may indicate that stimulation of cancer cells is resulting in down-regulation of FGF-1 expression or may implicate FGF-1 as a differentiation factor rather than a growth factor at its physiological concentration in the breast.

Fibroblast growth factor receptors (FGFRs) localize in different cellular compartments. A splice variant of FGFR-3 localizes to the nucleus.

We have raised specific antibodies to the second immunoglobulin-like domain of fibroblast growth factor receptors (FGFRs) and used these to investigate the expression and subcellular localization of FGFR-1, -2, -3, and -4 in breast epithelial cells. All four receptors classes could be detected in breast cell lines; however, FGFR-4 and FGFR-2 appeared to be expressed at a higher level in breast cancer cell lines than in normal epithelial cells. Surprisingly, FGFR-3 localized in the cell nucleus by immunofluorescence. A second antibody to a separate epitope confirmed this finding and showed that the form of FGFR-3 present must contain an intact kinase domain as well as the growth factor binding domain. Western analysis of fractionated cells revealed the presence of two forms of FGFR-3 of 135 and 110 kDa. The 110-kDa form was predominantly found in the nucleus, whereas the 135 kDa form was sometimes found in the nucleus. RT-PCR analysis of FGFR-3 mRNA showed the presence of a splice variant in which exons 7 and 8 are deleted. This results in the translation of FGFR-3 missing the transmembrane domain but with an intact kinase domain, which could be a soluble, intracellular receptor. Transfection experiments showed that FGFR-3 containing this deletion and no signal peptide gave an identical nuclear staining pattern to that seen in breast epithelial cells. We conclude that two forms of FGFR-3 are present in breast epithelial cells; a full-length 135-kDa receptor, which has a conventional membrane localization, and a novel soluble form of 110 kDa.

Isolation of pure populations of epithelial and myoepithelial cells from the normal human mammary gland using immunomagnetic separation with Dynabeads.

Monoclonal antibodies to epithelial membrane antigen and common acute lymphoblastic leukemia antigen were bound to second antibody-coated magnetic microspheres. These specific antibody/bead complexes were then used to directly isolate purified epithelial and myoepithelial cells from normal breast organoid cell preparations by magnetic separation. Near homogeneous cell populations were selected with yields of 10-20 x 10(6) epithelial and myoepithelial cells per organoid preparation. Repeated purification steps allowed almost complete depletion of myoepithelial cells for RNA studies. Tissue culture of separated cell populations in appropriate defined media further ensured purity of cell type and was unimpeded by Dynabead attachment.