Breast Cancer Res Treat (2008) 110:39–49 DOI 10.1007/s10549-007-9684-7
PRECLINICAL STUDY
Cancer associated fibroblasts stimulated by transforming growth factor beta1 (TGF-b1) increase invasion rate of tumor cells: a population study Theresa M. Casey Æ Jonathan Eneman Æ Abigail Crocker Æ Jeffrey White Æ Joseph Tessitore Æ Mary Stanley Æ Seth Harlow Æ Janice Y. Bunn Æ Donald Weaver Æ Hyman Muss Æ Karen Plaut
Received: 3 July 2007 / Accepted: 9 July 2007 / Published online: 3 August 2007 Springer Science+Business Media, LLC 2007
Abstract Cancer associated fibroblasts (CAFs) are believed to promote tumor growth and progression. Our objective was to measure the effect of TGF-b1 on fibroblasts isolated from invasive breast cancer patients. Fibroblasts were isolated from tissue obtained at surgery from patients with invasive breast cancer (CAF; n = 28) or normal reduction mammoplasty patients (normal; n = 10). Myofibroblast activation was measured by counting cells immunostained for smooth muscle alpha actin (ACTA2) in cultures ± TGF-b1. Conditioned media (CM) was collected for invasion assays and RNA was isolated from cultures incubated in media ± TGF-b1 for 24 h. Q-PCR was used to measure expression of cyclin D1, fibronectin, laminin, collagen I, urokinase, stromelysin-1, and ACTA2 genes. Invasion rate was measured in chambers plated with
MDA-MB-231 cells and exposed to CM in the bottom chamber; the number of cells that invaded into the bottom chamber was counted. Wilcox Rank Sum tests were used to evaluate differences in CAFs and normal fibroblasts and the effect of TGF-b1. There was no difference in percent myofibroblasts or invasion rate between normal and CAF cultures. However, TGF-b1 significantly increased the percent of myofibroblasts (P < 0.01) and invasion rate (P = 0.02) in CAF cultures. Stromelysin-1 expression was significantly higher in normal versus CAF cultures (P < 0.01). TGF-b1 significantly increased ACTA2 expression in both normal and CAF cultures (P < 0.01). Expression of fibronectin and laminin was significantly increased by TGF-b in CAF cultures (P < 0.01). CAFs were measurably different from normal fibroblasts in
T. M. Casey (&) K. Plaut Department of Animal Science, Michigan State University, B290 Anthony Hall, East Lansing, MI 48824, USA e-mail:
[email protected]
J. Tessitore D. Weaver Department of Pathology, University of Vermont, Burlington, VT 05405, USA
K. Plaut e-mail:
[email protected] J. Eneman A. Crocker H. Muss Department of Hematology and Oncology, University of Vermont, Burlington, VT 05405, USA J. Eneman e-mail:
[email protected] A. Crocker e-mail:
[email protected] H. Muss e-mail:
[email protected] J. White Department of Animal Science, University of Vermont, Burlington, VT 05405, USA e-mail:
[email protected]
J. Tessitore e-mail:
[email protected] D. Weaver e-mail:
[email protected] M. Stanley S. Harlow Department of Surgery, University of Vermont, Burlington, VT 05405, USA M. Stanley e-mail:
[email protected] S. Harlow e-mail:
[email protected] J. Y. Bunn Department of Medical Biostatistics, University of Vermont, Burlington, VT 05405, USA e-mail:
[email protected]
123
40
response to TGF-b1, suggesting that TGF-b stimulates changes in CAFs that foster tumor invasion. Keywords Breast Cancer associated fibroblast Extracellular matrix Stroma Transforming growth factor-beta Abbreviations CAF Cancer associated fibroblast TGF-b Transforming growth factor-beta ECM Extracellular matrix FBS Fetal bovive serum BM Basal media CM Conditioned media ACTA2 Smooth muscle alpha actin CCND1 Cyclin D1 FN1 Fibronectin LAMA1 Laminin COL1A1 Collagen I MMP3 Stromelysin-1 PLAU Urokinase RQ Relative expression CT Cycle threshold
Introduction Breast cancer is the most common cancer in women and the second leading cause of death from cancer [1]. Most cancer deaths are preceded by local invasion and metastasis, with metastasis being the cause of 90% of deaths from solid tumors [2]. During the transition from DCIS to invasive breast cancer, malignant epithelial cells must infiltrate through the basement membrane extracellular matrix (ECM) barrier to gain access into the stroma. Once invasion has occurred and cancer cells have infiltrated the stroma they gain the potential to invade surrounding vasculature and to metastasize [3]. In epithelial tumors, reactive stroma, or desmoplasia, is often adjacent to carcinoma cells, and believed to form in response to the malignant transformation of epithelial cells [4]. Reactive stroma is characterized by activated fibroblast cells (myofibroblasts) that express smooth muscle alpha actin (ACTA2), modified ECM, and angiogenesis [5, 6]. Transcriptome analysis of invasive breast cancer suggested that invasion occurs in part through the acquistion of a reactive phenotype in cancer associated fibroblasts (CAFs; Casey et al., submitted data). CAFs within reactive stroma are highly proliferative and express higher levels of ECM proteases and proteins (for review, see Ref. [4, 7]). During invasion it is believed that ECM proteases secreted by CAFs hydrolyze the basement membrane barrier and newly
123
Breast Cancer Res Treat (2008) 110:39–49
synthesized ECM proteins serve as a scaffolding for motile tumor cells to track along and for structural support for angiogenesis [5, 8]. Transforming growth factor-betas (TGF-b) are multifunctional cytokines which inhibit epithelial cell growth, stimulate mesenchymal cell proliferation, regulate ECM deposition and degradation, and modulate immune function and wound repair [9–13]. TGF-bs have been shown to play a role in tumor suppression as several human cancers have been associated with reduced expression or inactivation of the type II TGF-b receptor (TGF-bR2) and/or the Smad signal transducers [14–17]. However, we and others have found that a loss of TGF-b growth inhibition in breast cancer often occurs without a loss of these signaling components [18–21]. TGF-bs have also been shown to support the progression of cancer [22] (for review). Tang et al. [23] showed using mouse models that TGF-b can transition from a tumor repressor in early stage mammary tumors to a tumor promoter in late stage tumors, and may also increase the risk of metastases. This is consistent with the higher concentrations of TGF-b1 and b3 in serum of patients with advanced tumors [24, 25], as well as increased expression of TGF-b1 in breast cancer stroma of later staged tumors [26, 27]. Dumont and Arteaga [28] have hypothesized that the switch from TGF-b’s role in tumor suppression to a role in tumor progression is caused by a dissociation of TGF-b’s antiproliferative and matrix associated effects. TGF-b1 treatment in vitro activates normal primary breast fibroblasts and CAFs into myofibroblasts [29, 30], and subcutaneous injections of TGF-b1 into mice stimulates formation of reactive stroma [31]. Further, human colon and breast cancer cell lines treated with TGF-b express higher levels of ECM related mRNAs including FN1, type IV collagenase, collagen IV, LAMA1, PLAU and PAI-1 but are not growth inhibited [32, 33]. TGF-b also induces the expression of the ECM protease stromelysin-3 (MMP1) in osteoblasts and fibroblasts [34], which has been implicated in the growth of breast cancer metastases in bone [35]. Transgenic mice that over-express CCND1 in mammary tissue develop mammary tumors associated with significant desmoplasia [36]. Cancer epithelial cells isolated from these tumors secrete TGF-b and CAFs from these tumors secrete factors that stimulate the proliferation of tumor epithelial cancer cells [36]. These data suggest that TGF-b induces a reactive phenotype in CAFs and formation of tumor associated desmoplasia. Previously, we showed that the TGF-bR2 was expressed in malignant breast epithelial cells, and that higher expression of TGF-bR2 in stroma was correlated with a poor prognosis [18]. Additionally, we showed using immunohistochemistry that there was higher expression of TGF-b1 in primary tumors from patients with recurrent breast cancer [37]. More recently we described the
Breast Cancer Res Treat (2008) 110:39–49
molecular signature of stroma associated with invasive breast cancer and reported that many ECM proteins and proteases, including collagen I (COL1A1) and fibronectin (FN1), as well as TGF-b were also more highly expressed in breast cancer stroma versus normal breast stroma (Casey et al., submitted data). We hypothesize that TGF-b facilitates breast cancer invasion by stimulating secretion of ECM proteins and proteases from CAFs, which results in remodeling the ECM and creates an environment that promotes invasion and facilitates metastasis. The objective of this study was to measure the effect of TGF-b on CAFs isolated from women with invasive breast cancer (n = 28) and fibroblasts from normal controls (reduction mammoplasty patients, n = 10). CAF and normal fibroblast cultures were treated with TGF-b to measure its effect on myofibroblast activation and expression of CCND1, FN1, LAMA1, COL1A1, PLAU, MMP3, and ACTA2 as well as the effect of conditioned media (CM) from these cultures on the rate of breast cancer cell in vitro invasion. While others have measured myofibroblast activation and expression of many of these genes in CAFs using animal models and in primary fibroblast cultures [6, 29, 30, 38–44], we are among the first to look at the effects of TGF-b on CAFs isolated from a fairly large population of women with invasive breast cancer. We report that, although there is a great deal of heterogeneity in the behavior of CAFs and normal primary fibroblasts within these populations, CAFs are measurably different from normal fibroblasts in their response to TGF-b1 and expression of genes that encode ECM proteins and proteases.
41 Table 1 Summary breast cancer patient clinical data Age Mean
56.7
Median
56
Range
40–85
Tumor size T1 (0.1–2 cm)
No. (%) 11 (39)
T2 (>2–5 cm)
15 (54)
T3 (>5 cm)
2 (7)
Node status
No. (%)
Positive
11 (39)
Negative
17 (61)
Tumor gradea Well
No. (%) 6 (22)
Moderate
13 (46)
Poor
9 (32)
Tumor type
No. (%)
Ductal
24 (86)
Lobular
3 (11)
Metaplastic Estrogen receptor statusb ER positive ER negative Progesterone receptor statusb
20 (77) 6 (23) No. (%) 16 (62)
PR negative
10 (23)
Positive Negative Stage
Study population
1 (3) No. (%)
PR positive Lymphovascular invasion
Materials and methods
Breast tissue from cancer patients and normal controls (reduction mammoplasty patients) was collected from consecutive patients that were identified through the Breast Care Center at Fletcher Allen Health Care, the University of Vermont affiliated hospital. Informed consent meeting all federal, state and institutional guidelines was obtained from all subjects. Tissue was collected from 28 patients with clinical stage I, II or III breast cancer and from 10 patients who underwent reduction mammoplasty (normal). Final pathologic tumor stage was determined with the TNM staging system (AJCC Cancer Staging Manual, 6th edition, 2002) and graded using the Nottingham system [45]. In addition tumor grade, estrogen receptor (ER) and progesterone receptor (PR) status, presence or absence of lymphovascular invasion (LVI), and lymph node status were assessed in each tumor. Table 1 summarizes the clinical characteristics of the breast cancer patient population. The
Years
No. (%) 11 (39) 17 (61) No. (%)
I
11 (39)
II
11 (39)
III
6 (22)
All tumors were obtained with consent by surgical resection for primary diagnosis of breast cancer between 2004 and 2005. The ethnic composition of patients was Caucasian. Final pathologic tumor stage was determined using the TNM staging system (AJCC Cancer Staging Manual) a
The Nottingham system was used to assess tumor grade (differentiation): grade 1 = well differentiated; grade 2 = moderately differentiated; grade 3 = poorly differentiated [46]
b
Data missing from two patients
mean, median and range of reduction patient ages were, respectively, 38.9, 39, and 21–55 years old. Isolation of fibroblast cells from human breast tissue to establish primary cell cultures Following surgical resection, a 0.5 cm3 sample of breast tissue was cut from the edge of the tumor or reduction tissue. The samples were placed in Dulbecco’s Modified
123
42
Eagle’s Medium (DMEM; Sigma Chemical Corporation, St. Louis, MO, USA), pH 7.2–7.3, supplemented with 100,000 U/l penicillin G, 100 mg/l streptomycin (Gibco Invitrogen Corporation, Grand Isle, NY, USA) and 10% fetal bovine serum (FBS, Gibco Invitrogen Corporation) referred to as basal medium (BM), and immediately transported on ice to the laboratory. The tissue was then finely minced into 1–2 mm fragments, washed twice in antibiotic PBS (PBS supplemented with 100 U/ml penicillin, 100 lg/ml streptomycin, 1.5 g/ml fungizone), and disaggregated overnight in BM with 0.1% collagenase III (Worthington Biochemical Corp., Lakewood, NJ, USA) at 37C on a rotator. After 24 h the epithelial cells were separated from stromal cells by differential centrifugation, as described by Speirs et al. [47]. Stromal cells were washed twice in antibiotic PBS and plated in 35 mm dishes with BM supplemented with 10 lg/ml insulin (Sigma Chemical Company) at 37C in a humidified chamber containing 5% CO2. Media was changed every 2 days. When cells reached confluence they were passaged to a 25 cm2 flask (Corning Plasticware Cell Culture, Corning, NY, USA) by treating with 0.25% trypsin-25 mM EDTA (Gibco Invitrogen Corporation) and agitating until cells began to detach from the surface of the flask (passage 1; p1). P2 cells were moved to a 75 cm2 flask and then passaged 1:4. All experiments were performed on fibroblasts that had been cultured for 3–10 passages. Phenotypic characterization of primary stromal fibroblasts cells To determine if isolated cultures were composed primarily of fibroblasts, cells were phenotypically characterized by immunostaining. Cells positive for vimentin and negative for multicytokeratin staining were considered fibroblasts. Cells were plated in four well chamber slides (Corning), and grown to confluence. Cells were washed with PBS and fixed with 70% ethanol and immunostained according to manufacturer’s protocols (Histostain-Plus Kit, Zymed, South San Francisco, CA, USA). Multi-cytokeratin and a vimentin mouse monoclonal antibodies (Novocastra; Novocastra Laboratories Ltd., Newcastle upon Tyne, United Kingdom; cat. # NCL-C11 and # NCL-VIM, respectively) were used at 1:10 dilutions. Negative controls for each culture were incubated for 1 h using a mouse isotype IgG (HistostainPlus kit, Zymed Laboratories Inc.). The cells were stained with amino-ethyl-carbozole (AEC; Histostain-SP kit), cover-slipped and photographed under the microscope. Assessment of fibroblast activation into myofibroblast Cells were plated and grown to confluence in 4-chamber slides in BM. Media was aspirated from the cultures and
123
Breast Cancer Res Treat (2008) 110:39–49
cells were washed two times with PBS and then incubated for 24 h in serum free media with 0 or 2.5 ng/ml TGF-b1 (R&D Systems, Minneapolis, MN, USA). After 24 h, cells were fixed, and incubated with primary sm-a-actin (ACTA2) antibody diluted 1:2,000 (Sigma Chemical Co.) and stained as described above. Percentage of myofibroblasts was assessed by counting at least 1,000 total cells and determining the proportion stained positively for ACTA2 in three fields at 200· in duplicate preparations. Collection of conditioned media and isolation of total RNA Primary fibroblasts were plated with BM in 75 cm2 flasks and grown to confluence. Media was poured off and cells were rinsed with PBS. Fresh BM was added with 0 or 2.5 ng/ml TGF-b1 and left on cultures for 24 h. CM was collected and stored at 80C until used for invasion assays. Cells were rinsed two times with PBS and 7.5 ml of TRIZOLTM Reagent (Life technologies, Rockville, MD, USA) was added. Cell lysates were passed through a pipet several times and stored at 80C until isolation of total RNA following manufacturer’s protocol. Measurement of TGF-b1 concentration in basal and conditioned media The concentration of TGF-b was determined using a human TGF-b1 Immunoassay kit (DB100B, R&D Systems) following manufacturing instructions. Both acidified and non-acidified samples of basal and CM were evaluated. PBS was substituted for the acidification treatment. Real-time quantitative polymerase chain reaction (Q-PCR) Total RNA was isolated as described above and DNase treated using the DNA-freeTM kit (Ambion, Austin, TX, USA). Quantity and quality of the RNA was assessed with the Nanodrop1 ND-1000 UV-Vis Spectrophotometer (Nanodrop Technologies, Wilmington, DE, USA) and on the Nanochip using the Bioanalyzer 2100 (Agilent Inc., Palo Alto, CA, USA), respectively. RNA was reverse transcribed into cDNA using the GeneAmp1 kit (Applied Biosystems, Foster City, CA). Q-PCR analysis was performed using the ABI Prism 7700 (Applied Biosystems) and a unique TaqMan1 Assays-on-DemandTM Gene Expression kit (Applied Biosystems) specific for human: sm-a-actin (ACTA2; Hs00426835_g1); cyclin D1 (CCND1, Hs00277039_m1), fibronectin (FN1; Hs00415006_m1), laminin (LAMA1; Hs01125436_g1), collagen I (COL1A1; Hs01076780_g1), stromelysin-1 (MMP3; Hs00968308_m1), urokinase
Breast Cancer Res Treat (2008) 110:39–49
(PLAU; Hs01547055_g1), and GAPDH (Hs99999905_m1), which was used as the housekeeping gene. Relative gene expression (RQ) was calculated according to the following equations, with cycle threshold abbreviated CT: delta CT (individual patient ± TGF-b) = CT (target gene) - CT (housekeeping gene); delta CT (individual patient ± TGFb) = CT (individual patient ± TGF-b) - CT (mean of normal cultures without TGF-b treatment); relative expression (RQ) = 2-CT.
Invasion chamber assay To assess whether CM from primary stromal fibroblast affected invasion rate of MDA-MB-231 cells (ATCC NO. HTB-26), BD BioCoatTM Growth Factor Reduced (GFR) Invasion Chambers with 8.0 lm pore inserts in 24-well plates (BD, Franklin, NJ, USA) were prepared according to manufacturer’s protocol. Inserts were removed and 750 ll of CM from each primary fibroblast cell line treated with 0 or 2.5 ng/ml TGF-b1 were added in duplicate to the bottom chambers. Inserts were replaced and 500 ll of 200,000 MDA-MB-231 cells in suspension with serum free DMEM supplemented with 0.1% bovine serum albumin (BSA; R&D Systems) were added into inserts. Chambers were incubated for 24 h at 37C. Non-invaded cells were removed from the top surface of the insert by scrubbing with cotton tip swabs. Invaded cells were fixed on the membrane with 70% ethanol for 15–30 min at 4C, washed twice with PBS, stained with hematoxylin and mounted on slides. The average number of invaded cells per chamber was determined by counting invaded cells from three fields at 200· under light microscopy.
Statistical analysis The effect of CAF and normal fibroblasts or 0 and 2.5 ng/ ml TGF-b1 were evaluated using a Wilcoxon Rank Sum test. While the effects of TGF-b1 (0 ng/ml vs. 2.5 ng/ml) within each patient group were examined using the Wilcoxon Signed Rank test. Statistical analyses for the rate of invasion assays were based on a mixed model analysis of variance to allow for control of the assay batch. All analyses were performed using SAS, Version 8.01 (30. SAS Institute Inc. 2000).
Results Phenotypic staining showed that the combination of differential centrifugation and plating in selective media resulted in the isolation of primary fibroblast cultures from
43
each patient (Fig. 1). The proportion of activated fibroblasts, myofibroblasts, in each culture treated with 0 or 2.5 ng/ml TGF-b1 was determined by counting the number of cells immunostained for ACTA2 expression (Fig. 2a). The percentage of myofibroblast varied from 0 to 70% among the normal and CAF cultures with TGF-b1 treatment causing a noticeable shift in the percentage of activated myofibroblasts in many normal and CAF cultures (Fig. 2b). TGF-b1 treatment significantly increased the mean percentage of myofibroblasts in CAF cultures (P < 0.01), but did not have a significant effect on the mean percentage of myofibroblasts in normal cultures (Fig. 2b). Similarly, there was no difference in level of ACTA2 gene expression between normal and CAF; however, TGF-b1 significantly increased the ACTA2 expression in both normal and CAF cultures (Fig. 3; P = 0.01 and P < 0.01, respectively). TGF-b1 increased the mean expression of FN1 gene expression in CAF cultures (P < 0.01) but did not have a significant effect on the mean expression among normal fibroblasts (Fig. 3b). The mean MMP3 gene expression among normal fibroblasts was significantly higher than among CAF cultures (P = 0.03). However, the addition of TGF-b1 did not affect MMP3 expression in either group. The average expression of PLAU was greater in CAF cultures (Fig. 3b). There was no difference in the expression of the other genes examined (CCDN1, LAMA1, and COL1A1) in normal and CAF cultures and TGF-b1 had no significant effect on expression of these genes in either population. When CM collected from the fibroblast cultures was used to determine its effect on in vitro invasion of MDAMB-231 cells, there was no difference in the rate of invasion stimulated by CM from CAF and normal fibroblast cultures (Fig. 4b). Following incubation with TGFb1, there was a noticeable shift in the invasion rate toward CM (Fig. 4a). In CAF cultures, this TGF-b stimulated shift resulted in a significant increase in mean invasion rate (Fig. 4b; P = 0.02). TGF-b stimulation did not significantly effect mean invasion rate toward CM from normal cultures (Fig. 4b). Since it has been reported that FBS contains TGF-b and cells in culture secrete TGF-b [36, 48–50], the levels of latent and activated TGF-b were measured in basal media (BM) and CM. The concentration of TGF-b1 was below detectable levels in non-acidified samples of BM and CM collected from normal and CAF cultures treated with 0 ng/ml TGF-b1. After acidification, the concentration of TGF-b1 in BM and CM were 1.2 ± 0.12 and 1.6 ± 0.18 ng/ml, respectively, indicating that latent TGFb1 was present in serum and secreted from primary cells, but not likely contributing to the effects attributed to TGF-b treatment.
123
44
Breast Cancer Res Treat (2008) 110:39–49
Fig. 1 Phenotypic staining demonstrates that primary cultures were selected for fibroblast cells. Breast tissue was disaggregated overnight with collagenase III and fibroblast cells were isolated by differential centrifugation followed by plating in selective media. At P3, to determine if isolated cultures were composed of fibroblasts, cells were phenotypically characterized by immunostaining. Cells positive for
vimentin and negative for multicytokeratin were considered fibroblasts. All primary cultures selected for fibroblasts immunostained positively for (A) vimentin expression and negatively for (B) multicytokeratin expression. Photograph was taken at 100· on a light microscope
Fig. 2 Effect of origin (CAF versus normal fibroblast) and TGF-b1 treatment on the percentage of activated fibroblasts-myofibroblasts. Fibroblasts were plated and grown to confluence in 4-chamber slides in BM. Media was aspirated from the cultures and cells were washed two times with PBS and then incubated for 24 h in serum free media with 0 or 2.5 ng/ml TGF-b1. After 24 h, cells were fixed, incubated with primary sm-a-actin (ACTA2) antibody and immunostained with AEC. (a) Myofibroblasts are stain positively for ACTA2 expression in
cultures treated with (A) 0 ng/ml or (B) 2.5 ng/ml TGF-b1. Photograph was taken at 200· magnification. The percentage of myofibroblasts was assessed by counting at least 1,000 total cells and determining the proportion stained positively for ACTA2 in three fields at 200· in duplicate preparations. (b) Mean percentage of myofibroblasts in cultures treated with 0 or 2.5 ng/ml TGF-b1 for 24 h. Data reported as mean percent of cells expressing ACTA2 ± SEM (*P < 0.01)
123
Breast Cancer Res Treat (2008) 110:39–49
a
45
1000
100
RQ
10
1
0.1
0.01
0.001 CCND1
ACTA2
b
COL1A1
LAMA1
FN1
PLAU
MMP3
15 14
a
13 12 11 10
RQ
9 8 7
b
b
6 5 a
3 2
a
b
4 a a
bc
a a
b b a
a a a
a a a a
a a a
1
a b
a b
0 ACTA2
CCND1
COL1A1
LAMA1
FN1
MMP3
PLAU
Fig. 3 Relative gene expression (RQ) of normal and CAF cultures treated with 0 and 2.5 ng/ml TGF-b1. Primary fibroblasts were plated and grown to confluence. Media was poured off and fresh media was added with 0 or 2.5 ng/ml TGF-b1 and left on cultures for 24 h. Total RNA was isolated from cells and reversed transcribed into cDNA. QPCR was used to measure expression of sm-a-actin (ACTA2), cyclin D1 (CCND1), fibronectin (FN1), laminin (LAMA1), collagen I (COL1A1), stromelysin-1 (MMP3), and urokinase (PLAU). RQ was calculated for each culture relative to mean expression of normal
cultures without TGF-b treatment. (a) RQ was plotted for each normal (open triangle) and CAF (open circle) primary culture and TGF-b1 treated normal (black triangle) and CAF (black circle) primary culture to show variability in expression levels among the patients. Note y-axis is log-scale. (b) Mean RQ of normal and CAF cultures treated with 0 (white and black bars, respectively) or 2.5 ng/ ml (striped and checked bars, respectively) TGF-b1 for 24 h. Values are mean ± SEM; for comparisons within each gene, means without common letters are significantly different (P 0.05)
Discussion
hour following irradiation active TGF-b1 significantly increases in murine mammary glands [53], suggesting that TGF-b1 plays a role in shortening the tumor latency period. Transgenic experiments demonstrated that when TGF-bR2 gene was knocked out in mammary fibroblasts, the knockout mice exhibited defective mammary ductal development, characterized by increased ductal epithelial cell turnover and stromal fibroblast abundance. When fibroblasts from these knockout mice were transplanted with mammary carcinoma cells, they promoted growth and invasion. In addition, treatment of tumor cells with fibroblast-conditioned medium led to increased tumor cell
The role stromal tissue and cells play in regulating the fate and morphogenesis of mammary epithelial cells has been extensively investigated through a number of elegant experiments, including: Transplantation experiments that demonstrated stroma directs organogenesis in salivary and mammary glands [51]. Irradiation experiments that showed mammary epithelial cells transplanted into irradiated mouse mammary fat pads gave rise to significantly more tumors with shorter latency periods than when they were transplanted into fat pads of non-irradiated hosts [52]. One
123
46
a
Breast Cancer Res Treat (2008) 110:39–49
500 450
num ber i nvaded cel l s
400 350 300 250 200 150 100 50 0
b
Fig. 4 Effect of conditioned media collected from primary normal fibroblast and CAF cultures on the in vitro invasion rate of MDA-MB231 cells. Primary stromal fibroblasts were plated with BM in 75 cm2 flasks and grown to confluence. Media was poured off and cells were rinsed with PBS. Fresh BM was added with 0 or 2.5 ng/ml TGF-b1 and left on cultures for 24 h. Conditioned media (CM) was collected and used for invasion assays. CM from each primary fibroblast cell line treated with 0 or 2.5 ng/ml TGF-b1 were added in duplicate to the bottom chamber of 8.0 lm pore invasion chambers. Upper chamber inserts coated with reduced growth factor matrigel were seeded with 200,000 MDA-MB-231 cells in serum free DMEM supplemented with 0.1% bovine serum albumin. Chambers were incubated for 24 h at 37C, and non-invaded cells were removed from the top surface of the insert. Invaded cells were fixed on the membrane, stained with hematoxylin and mounted on slides. The average number of invaded cells per chamber was determined by counting invaded cells from three fields at 200· under light microscopy. (a) Invasion rate of MDA-MB-231 cells toward CM from each normal culture treated with 0 ng/ml (open triangle) or 2.5 ng/ ml TGF-b1 (black triangle) and each CAF culture treated with 0 ng/ml (open circle) or 2.5 ng/ml TGF-b1 (black circle). (b) Mean invasion rate of normal and CAF cultures treated with 0 or 2.5 ng/ml (*P = 0.02)
123
proliferation and motility, which were blocked by the addition of pharmacologic inhibitors of TGF-b signaling [43]. More clinically relevant experiments suggest that genetic mutations in fibroblasts facilitate the development of carcinomas by causing abnormal stromal–epithelial interactions [54, 55]. These data demonstrate that both TGF-b and the stroma play a major role in normal mammary development and breast cancer progression. We characterized and compared primary fibroblasts from normal reduction mammoplasty patients with fibroblasts from invasive breast cancer patients. When we examined the cultures for ‘‘baseline differences’’ in the mean percent of activated fibroblasts (myofibroblasts) and the effect of CM collected from these cells on the mean in vitro invasion rate of MDA-MB-231 cells, we found no significant differences between normal and CAF cultures. However, TGF-b1 treatment caused a shift in the percent myofibroblasts and invasion rate in many of the individual cultures (Figs. 2, 4a). When we examined the populations as wholes we found that TGF-b treatment had significantly increased both the mean percent of myofibroblasts and the mean in vitro invasion rate in CAFs but did not have a significant effect on normal fibroblasts. Thus, although a shift in invasion rate and percent myofibroblast was evident in both populations, the greater shift in mean percent myofibroblasts and induction of increased invasion rate is likely due to a difference in biological response between the two populations. Our findings contrast Sieuwerts et al. [30], who report no difference in level of TGF-b induced myofibroblast activation between paired normal fibroblasts and CAFs isolated from the same breast cancer patients. However, since both normal and cancer fibroblasts were isolated from the same women in the Sieuwerts’ study, we believe the more reactive phenotype of CAFs in response to TGF-b in our study reflects innate changes in stromal cells from breast cancer patients. Moinfar et al. [55] reported that there was loss of heterozygosity (LOH) in both epithelial and mesenchymal components of 73% (8/ 11) of breast cancer samples examined (IDC and DCIS combined), and none of the control cases (women without any breast disease) revealed LOH either in the epithelial or in the stromal components. Further, although gene expression profiles revealed that normal stroma and epithelium from breast cancer patients are not statistically distinct from epithelium and stroma isolated from reduction mammoplasties [56], it is likely that TGF-b and other factors stimulate a reactive phenotype in CAFs during cancer progression, as gene expression profiles of normal stroma and cancer stroma are distinct between invasive breast cancer patients and normal reduction mammoplasty patients (Casey et al., submitted data). These data suggest that although the stromal fibroblasts from the reduction mammoplasty and cancer patients may be innately
Breast Cancer Res Treat (2008) 110:39–49
different, it is their response to exogenous factors and interaction with tumor epithelial cells that make them effectors of tumor growth and progression. Thus we show that CAFs exhibited a more ‘‘reactive’’ phenotype in response to TGF-b treatment. In our primary fibroblast cultures we also examined the expression of genes known to be regulated by TGF-b1 including ACTA2, CCND1 and several ECM proteins and proteases. ACTA2 gene expression was significantly increased in both CAFs and normal fibroblasts in response to TGF-b1. However, only CAFs exhibited a significant increase in percent myofibroblasts in response to TGF-b as measured by immunohistochemistry. It is possible that manifestation of the gene response is quicker in CAFs than in normal fibroblasts or more likely that the stability of the message is greater in CAFs, thus translating to a greater response in the cultures isolated from invasive breast cancer patients. This could be due to a number of transcriptional or post-transcriptional regulatory events that were not measured. There were no differences in CCND1 gene expression between CAF and normal cultures or due to TGF-b1 treatment. The large mean increase in CCND1 seen in CAFs treated with TGF-b1 (Fig. 3) was due in large part to one outlier. This CAF culture had high gene expression prior to treatment and increased 50-fold following TGF-b treatment (Fig. 3). This patient also overexpressed LAMA1, COL1A1, and FN1 and was hypersensitive to TGF-b. This finding demonstrates the variability inherent in patients, which may be explained by genetic variation, heterogeneity within a population or the heterogeneity of breast cancer itself. More variability in the data was observed among CAF than normal fibroblast cultures indicating that much of this variability was due to the heterogeneity of breast cancer, although a much larger dataset would be necessary to determine whether variability was due to the lack of robustness in the sample size or represents a true clinical difference among the patients. The expression of MMP3 was significantly higher in normal fibroblasts compared to CAFs regardless of whether cells were treated with TGF-b1. This is surprising since MMP3 is a matrix metalloproteinase which functions to remodel the ECM and has been associated with invasion and metastasis [57]. Transgenic mice which over-express MMP3 have a greater rate of mammary tumorigenesis [58] which is associated with reactive stroma early in development [59]. However, interpretation of function based on gene expression is dubious since MMP3’s activity is dependent on activation as well as regulation by tissue inhibitor of metalloproteinase-1 (TIMP-1). Therefore, it would have been necessary to evaluate both MMP3 protein activity and TIMP-1 expression to determine if there was differential activity in our cultures.
47
TGF-b1 treatment caused a significant increase in gene expression of FN1 in CAFs but not in normal fibroblasts. FN1 encodes the ECM protein, fibronectin, with its overexpression associated with the molecular signature of stromal invasion of IDC (Casey et al., submitted data) as well as metastasis of melanoma [60]. It was also not surprising that PLAU expression was significantly greater CAF versus normal cultures (Fig. 3b), since in breast cancer PLAU mRNA is predominantly expressed by myofibroblasts located at the invasive areas of the tumor [61]. Further, although we expected TGF-b to induce PLAU expression in primary cultures [33], others have reported that TGF-b does not stimulate PLAU expression in primary fibroblast cultures [30].
Conclusion By comparing and characterizing primary fibroblasts from a fairly large number of biological replicates we demonstrate that breast CAFs are responsive to TGF-b induced changes that may foster tumor invasion of the stroma. Further our data and recent work of others suggests that invasion and metastasis may be due in part to a combination of changes in the microenvironment, i.e., LOH in fibroblast cells and a reactive response to TGF-b. Acknowledgments This work was supported by a grant from The Breast Cancer Research Foundation, New York, NY Real-time quantitative PCR was performed in the Vermont Cancer Center DNA Analysis Facility and was supported in part by grant P30CA22435 from the NCI.
References 1. Jemal A, Siegel R, Ward E, Murray T, Xu J, Thun MJ (2007) Cancer Statistics, 2007. CA Cancer J Clin 57:43–66 2. Gupta GP, Massague J (2006) Cancer Metastasis: building a framework. Cell 127:679–695 3. Schnitt S, Guidi A (2000) Pathology of invasive breast cancer. In: Harris JR, Lippman ME, Murrow M, Osborn CK (eds) Diseases of the breast, 2nd edn. Lippincott, Williams, and Wilkins, NY, p 425 4. Radisky E, Radisky D (2007) Stromal induction of breast cancer: inflammation and invasion. Rev Endocr Metab Disord. doi: 10.1007/s11154-007-9037-1 5. Noel A, Foidart J-M (1998) The role of stroma in breast carcinoma growth in vivo. J Mammary Gland Biol Neoplasia V3:215–225 6. Tuxhorn JA, McAlhany SJ, Dang TD, Ayala GE, Rowley DR (2002) Stromal cells promote angiogenesis and growth of human prostate tumors in a differential reactive stroma (DRS) xenograft model. Cancer Res 62:3298–3307 7. Kalluri R, Zeisberg M (2006) Fibroblasts in cancer. Nat Rev Cancer 6:392–401 8. Condeelis J, Segall JE (2003) Intravital imaging of cell movement in tumours. Nat Rev Cancer 3:921–930
123
48 9. Robinson S, Silberstein G, Daniel CW (1992) Evidence supporting a role for TGF-beta isoforms in growth regulation and functional differentiation of the mouse mammary gland. In: Picciano MF, Lonnerdal B (eds) Mechanisms regulating lactation and infant nutrient utilization. Wiley-Liss, New York, pp 43–52 10. Faler B, Macsata R, Plummer D, Mishra L, Sidawy A (2006) Transforming growth factor-beta and wound healing. Perspect Vasc Surg Endovasc Ther 18:55–62 11. O’Kane S, Ferguson MWJ (1997) Transforming growth factor [beta]s and wound healing. Int J Biochem Cell Biol 29:63–78 12. Wahl SM (2007) Transforming growth factor-[beta]: innately bipolar. Curr Opin Immunol 19:55–62 13. Fleisch MC, Maxwell CA, Barcellos-Hoff M-H (2006) The pleiotropic roles of transforming growth factor beta in homeostasis and carcinogenesis of endocrine organs. Endocrinology 13:379–400 14. Markowitz S, Wang J, Myeroff L, Parsons R, Sun L, Lutterbaugh J, Fan R, Zborowska E, Kinzler K, Vogelstein B (1995) Inactivation of the type II TGF-beta receptor in colon cancer cells with microsatellite instability. Science. 268:1336–1338 15. Hahn S, Schutte M, Hoque A, Moskaluk C, da Costa L, Rozenblum E, Weinstein C, Fischer A, Yeo C, Hruban R, Kern S (1996) DPC4, a candidate tumor suppressor gene at human chromosome 18q21.1. Science 271:350–353 16. Grady W, Myeroff L, Swinler S, Rajput A, Thiagalingam S, Lutterbaugh J, Neumann A, Brattain M, Chang J, Kim S, Kinzler K, Vogelstein B, Willson J, Markowitz S (1999) Mutational inactivation of transforming growth factor beta receptor type II in microsatellite stable colon cancers. Cancer Res 50:320–324 17. Grady W, Rajput A, Myeroff L, Liu D, Kwon K, Willis J, Markowitz S (1998) Mutation of the type II transforming growth factor-beta receptor is coincident with the transformation of human colon adenomas to malignant carcinomas. Cancer Res 58:3101–3104 18. Barlow J, Yandell D, Weaver D, Casey T, Plaut K (2003) Higher stromal expression of transforming growth factor-beta type II receptors is associated with poorer prognosis breast tumors. Breast Cancer Res Treat 79:149–159 19. Tomita S, Deguchi S, Miyaguni T, Muto Y, Tamamoto T, Toda T (1999) Analyses of microsatellite instability and the transforming growth factor-beta receptor type II gene mutation in sporadic breast cancer and their correlation with clinicopathological features. Breast Cancer Res Treat 53:33–39 20. Anbazhagan R, Bornman DM, Johnston JC, Westra WH, Gabrielson E (1999) The S387Y mutations of the transforming growth factor-beta receptor type I gene is uncommon in metastases of breast cancer and other common types of adenocarcinoma. Cancer Res 59:3363–3364 21. Reiss M, Barcellos-Hoff MH (1997) Transforming growth factorbeta in breast cancer: a working hypothesis. Breast Cancer Res Treat 45:81–95 22. Roberts A, Wakefield L (2003) The two faces of transforming growth factor beta in carcinogenesis. Proc Natl Acad Sci USA 100:8621–8623 23. Tang B, Vu M, Booker T, Santner SJ, Miller FR, Anver MR, Wakefield LM (2003) TGF-beta switches from tumor suppressor to prometastatic factor in a model of breast cancer progression. J Clin Invest 112:1116–1124 24. Ghellal A, Li C, Hayes M, Byrne G, Bundred N, Kumar S (2000) Prognostic significance of TGF beta 1 and TGF beta 3 in human breast carcinoma. Anticancer Res 20:4413–4418 25. Sheen-Chen SM, Chen HS, Sheen CW, Eng HL, Chen WJ (2001) Serum levels of transforming growth factor beta1 in patients with breast cancer. Arch Surg 136:937–940 26. Gorsch SM, Memoli VA, Stukel TA, Gold LI, Arrick BA (1992) Immunohistochemical staining for transforming growth factor
123
Breast Cancer Res Treat (2008) 110:39–49
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
beta 1 associates with disease progression in human breast cancer. Cancer Res 52:6949–6952 Walker RA, Dearing SJ, Gallacher B (1994) Relationship of transforming growth factor beta 1 to extracellular matrix and stromal infiltrates in invasive breast carcinoma. Br J Cancer 69:1160–1165 Dumont N, Arteaga CL (2000) Transforming growth factor-beta and breast cancer: tumor promoting effects of transforming growth factor-beta. Breast Cancer Res 2:125–132 Ronnov-Jessen L, Petersen OW (1993) Induction of alphasmooth muscle actin by transforming growth factor-beta 1 in quiescent human breast gland fibroblasts. Implications for myofibroblast generation in breast neoplasia. Lab Invest 68:696–707 Sieuwerts A, Klijn J, Henzen-Logmans S, Foekens J (1999) Cytokine regulated urokinase type plasminogen activator (uPA) production by human breast fibroblasts in vitro. Breast Cancer Res Treat 55:9–20 Roberts AB, Sporn MB, Assoian RK, Smith JM, Roche NS, Wakefield LM, Heine UI, Liotta LA, Falanga V, Kehrl JH et al (1986) Transforming growth factor type beta: rapid induction of fibrosis and angiogenesis in vivo and stimulation of collagen formation in vitro. Proc Natl Acad Sci USA 83:4167–4171 Geiser AG, Burmester JK, Webbink R, Roberts AB, Sporn MB (1992) Inhibition of growth by transforming growth factor-beta following fusion of two nonresponsive human carcinoma cell lines. Implication of the type II receptor in growth inhibitory responses. J Biol Chem 267:2588–2593 Stampfer MR, Yaswen P, Alhadeff M, Hosoda J (1993) TGF beta induction of extracellular matrix associated proteins in normal and transformed human mammary epithelial cells in culture is independent of growth effects. J Cell Physiol 155:210–221 Delany AM, Canalis E (2001) The metastasis-associated metalloproteinase stromelysin-3 is induced by transforming growth factor-beta in osteoblasts and fibroblasts. Endocrinology 142:1561–1566 Pilkington MF, Sims SM, Dixon SJ (2001) Transforming growth factor-beta induces osteoclast ruffling and chemotaxis: potential role in osteoclast recruitment. J Bone Miner Res 16:1237–1247 Corsino P, Davis B, Law M, Chytil A, Forrester E, Norgaard P, Teoh N, Law B (2007) Tumors initiated by constitutive Cdk2 activation exhibit transforming growth factor {beta} resistance and acquire paracrine mitogenic stimulation during progression. Cancer Res 67:3135–3144 Plaut K, Bramley A, Casey T, Muss H (2004) Patients with recurrent breast cancer have higher expression of TGF-B1 in primary tumor. In: San Antonio breast cancer symposium, San Antonio, TX Singer C, Kronsteiner N, Marton E, Kubista M, Cullen K, Hirtenlehner K, Seifert M, Kubista E (2002) MMP-2 and MMP-9 expression in breast cancer-derived human fibroblasts is differentially regulated by stromal-epithelial interactions. Breast Cancer Res Treat 72:69–77 Basset P, Wolf C, Chambon P (1993) Expression of the stromelysin-3 gene in fibroblastic cells of invasive carcinomas of the breast and other human tissues: a review. Breast Cancer Res Treat 24:185–193 Tang Y, Kesavan P, Nakada MT, Yan L (2004) Tumor-stroma interaction: positive feedback regulation of extracellular matrix metalloproteinase inducer (EMMPRIN) expression and matrix metalloproteinase-dependent generation of soluble EMMPRIN. Mol Cancer Res 2:73–80 Bhowmick NA, Chytil A, Plieth D, Gorska AE, Dumont N, Shappell S, Washington MK, Neilson EG, Moses HL (2004) TGF-{beta} Signaling in fibroblasts modulates the oncogenic potential of adjacent epithelia. Science 303:848–851
Breast Cancer Res Treat (2008) 110:39–49 42. Ronnov-Jessen L, Petersen OW, Bissell MJ (1996) Cellular changes involved in conversion of normal to malignant breast: importance of the stromal reaction. Physiol Rev 76:69–125 43. Cheng N, Bhowmick NA, Chytil A, Gorksa AE, Brown KA, Muraoka R, Arteaga CL, Neilson EG, Hayward SW, Moses HL (2005) Loss of TGF-beta type II receptor in fibroblasts promotes mammary carcinoma growth and invasion through upregulation of TGF-alpha-, MSP- and HGF-mediated signaling networks. Oncogene 24:5053–5068 44. Parrott JA, Nilsson E, Mosher R, Magrane G, Albertson D, Pinkel D, Gray JW, Skinner MK (2001) Stromal-epithelial interactions in the progression of ovarian cancer: influence and source of tumor stromal cells. Mol Cell Endocrinol 175:29–39 45. Elston CW, Ellis IO (2002) Pathological prognostic factors in breast cancer. I. The value of histological grade in breast cancer: experience from a large study with long-term follow-up. C. W. Elston & I. O. Ellis. Histopathology (1991) 19:403–410. Histopathology 41:151 46. Elston C, Ellis I (1991) Pathological prognostic factors in breast cancer. I. The value of histological grade in breast cancer: experience from a large study with long-term follow-up. Histopathol 19:403–410 47. Speirs V, White MC, Green AR (1996) Collagenase III: a superior enzyme for complete disaggregation and improved viability of normal and malignant human breast tissue. In vitro Cell Dev Biol Anim 32:72–74 48. Zarzynska J, Gajewska M, Motyl T (2005) Effects of hormones and growth factors on TGF-beta1 expression in bovine mammary epithelial cells. J Dairy Res 72:39–48 49. Van Susante J, Buma P, Van Beuningen H, Van den Berg W, Veth R (2000) Responsiveness of bovine chondrocytes to growth factors in medium with different serum concentrations. J Orthoped Res 18:68–77 50. van Roozendaal C, Klijn J, van Ooijen B, Claassen C, Eggermont A, Henzen-Logmans S, Foekens J (1995) Transforming growth factor beta secretion from primary breast cancer fibroblasts. Mol Cell Endocrinol 111:1–6
49 51. Kratochwil K (1969) Organ specificity in mesenchymal induction demonstrated in the embryonic development of the mammary gland of the mouse. Dev Biol 20:46–71 52. Barcellos-Hoff MH, Ravani SA (2000) Irradiated mammary gland stroma promotes the expression of tumorigenic potential by unirradiated epithelial cells. Cancer Res 60:1254–1260 53. Barcellos-Hoff MH (1993) Radiation-induced transforming growth factor beta and subsequent extracellular matrix reorganization in murine mammary gland. Cancer Res 53:3880–3886 54. Barcellos-Hoff MH (1998) The potential influence of radiationinduced microenvironments in neoplastic progression. J Mammary Gland Biol Neoplasia 3:165–175 55. Moinfar F, Man YG, Arnould L, Bratthauer GL, Ratschek M, Tavassoli FA (2000) Concurrent and independent genetic alterations in the stromal and epithelial cells of mammary carcinoma: implications for tumorigenesis. Cancer Res 60:2562–2566 56. Finak G, Sadekova S, Pepin F, Hallett M, Meterissian S, Halwani F, Khetani K, Souleimanova M, Zabolotny B, Omeroglu A, Park M (2006) Gene expression signatures of morphologically normal breast tissue identify basal-like tumors. Breast Cancer Res 8:R58 57. Bodey B, Bodey B Jr., Siegel SE, Kaiser HE (2001) Matrix metalloproteinases in neoplasm-induced extracellular matrix remodeling in breast carcinomas. Anticancer Res 21:2021–2028 58. Sympson CJ, Bissell MJ, Werb Z (1995) Mammary gland tumor formation in transgenic mice overexpressing stromelysin-1. Semin Cancer Biol 6:159–163 59. Thomasset N, Lochter A, Sympson CJ, Lund LR, Williams DR, Behrendtsen O, Werb Z, Bissell MJ (1998) Expression of autoactivated stromelysin-1 in mammary glands of transgenic mice leads to a reactive stroma during early development. Am J Pathol 153:457–467 60. Clark EA, Goiub TR, Lander ES, Hynes RO (2000) Genomic analysis of metastasis reveals an essential role for RhoC. Nature 406:532–535 61. Nielsen B, Sehested M, Timshel S, Pyke C, Danø K (1996) Messenger RNA for urokinase plasminogen activator is expressed in myofibroblasts adjacent to cancer cells in human breast cancer. Lab Invest 74:168–177
123
