Cancer Research Meeting Calendar Frontiers in Basic Cancer Research
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
Cancer Research 68, 2717, April 15, 2008. doi: 10.1158/0008-5472.CAN-08-0192
© 2008 American Association for Cancer Research
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hawsawi, N. M.
Right arrow Articles by Aboussekhra, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hawsawi, N. M.
Right arrow Articles by Aboussekhra, A.

Molecular Biology, Pathobiology, and Genetics

Breast Carcinoma–Associated Fibroblasts and Their Counterparts Display Neoplastic-Specific Changes

Nahed M. Hawsawi1, Hazem Ghebeh2, Siti-Faujiah Hendrayani1, Asma Tulbah3, Maha Al-Eid1, Taher Al-Tweigeri4, Dahish Ajarim4, Ayodele Alaiya1, Said Dermime2 and Abdelilah Aboussekhra1

1 Department of Biological and Medical Research, 2 Tumor Immunology/Stem Cell Therapy Program, 3 Department of Pathology, and 4 Department of Oncology, King Faisal Specialist Hospital and Research Center, Riyadh, Saudi Arabia

Requests for reprints: Abdelilah Aboussekhra, King Faisal Specialist Hospital and Research Center, BMR, MBC 03-66, P.O. Box 3354, Riyadh 11211, Saudi Arabia. Phone: 966-1464-7272; Fax: 966-1-442-7858; E-mail: aboussekhra{at}kfshrc.edu.sa.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
It has become clear that the initiation and progression of carcinomas depend not only on alterations in epithelial cells, but also on changes in their microenvironment. To identify these changes, we have undertaken cellular and molecular characterization of carcinoma-associated fibroblasts (CAF) and their tumor counterpart fibroblasts (TCF) isolated from 12 breast cancer patients. Normal breast fibroblasts (NBF) from plastic surgery were used as normal control. We present evidence that both CAFs and TCFs are myofibroblasts and show tumor-associated features. Indeed, the p53/p21 response pathway to {gamma}-rays was defective in 70% CAFs, whereas it was normal in all the TCF and NBF cells. In addition, the basal levels of the p53 and p21 proteins were significantly low in 83% of CAFs and modulated in the majority of TCFs compared with NBFs. Interestingly, both TCFs and CAFs expressed high levels of the cancer marker survivin and consequently exhibited high resistance to cisplatin and UV light. Moreover, most CAFs were positive for the proliferation marker Ki-67 and exhibited high proliferation rate compared with NBFs and TCFs. However, proliferating cell nuclear antigen was highly expressed in both CAFs and TCFs. Using the two-dimensional gel electrophoresis technique, we have also shown that CAF, TCF, and NBF cells present different proteome profiles, with many proteins differentially expressed between these cells. Taken together these results indicate that different genetic alterations can occur in breast CAFs and their corresponding adjacent counterparts, showing the important role that stroma could play in breast carcinogenesis and treatment. [Cancer Res 2008;68(8):2717–25]


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Breast cancer is the most commonly occurring malignancy in women and is responsible for ~500,000 deaths yearly worldwide (1). Carcinomas—the most frequent form of human cancer—result mainly from the accumulation of somatic mutations in epithelial cells. Carcinoma cells, like normal epithelial cells, live in a complex microenvironment (stroma) that includes the extracellular matrix and cellular components, such as immune and inflammatory cells, blood vessel cells, and fibroblasts. All these cell types may critically influence the multistep process of tumorigenesis (2–5). Fibroblasts, the predominant cells of the stroma, are responsible for the elaboration of most of the components of connective tissue (6, 7). In addition, stromal-epithelial interactions have a fundamental role in the development of normal mammary gland. Therefore, modifications in the stromal fibroblasts can play a significant role in overall cancer development (8). Nevertheless, a key question remains: which comes first, the dysfunction of epithelial cells or the dysfunction of their microenvironment? In fact, several findings from different laboratories have suggested that cancer cells themselves can alter their adjacent stroma to form a permissive and supportive environment for tumor progression (3). On the other hand, there is compelling evidence that fibroblasts play major role in the initiation and progression of carcinomas (3, 7–9). The knowledge that stromal cells have the ability to stimulate oncogenenesis has been taken a step further by recent data showing that stromal fibroblasts present in invasive human breast carcinomas promote tumor growth and angiogenesis through elevated Sdf1 secretion (10). In another study, it has been shown that extensive gene expression changes occur in all breast cell types, including fibroblasts (11). Furthermore, Kiaris et al. have shown an association between TP53 mutations in the stromal component of epithelial tumors and carcinogenesis (12).

Together, these findings indicate that stromal fibroblasts have an active role in tumorigenesis and therefore should constitute a crucial target for cancer therapy, as well as for preventive strategies (13, 14). Thereby, a deep understanding of the epithelial-stromal biochemical interactions and molecular signaling is mandatory. To this end, we undertook here molecular and cellular characterization of breast carcinoma-associated fibroblasts (CAF) and their corresponding adjacent fibroblasts. We have found that stromal fibroblasts, as well as their adjacent counterparts, show molecular and cellular modifications that are specific for tumor cells.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Tissue collection. Breast cancer specimens were collected from primary tumors of 12 patients who underwent surgery at King Faisal Specialist Hospital. Signed informed consent was obtained from all the patients. After tumor resection, an anatomic pathologist grossly examined and obtained a representative piece of the tumor tissue and an adjacent "histologically normal" breast tissue from the same affected breast. For comparison, we have also obtained normal tissue from healthy women after plastic surgery.

Generation of primary fibroblast cells from breast cancer tissues. The obtained tissues were further processed to generate the primary cultures, as previously described (15). Fibroblast cells were cultured in Medium 199 and Ham's F12 mixed 1:1 and supplemented with 10% to 20% FCS. All supplements were obtained from Sigma, except for antibiotics and antimycotics solutions, which were obtained from Life Technologies.

Immunohistochemistry of frozen sections and cell cytospin. Fresh tissues were snap frozen in liquid nitrogen and preserved at –80°C, sectioned, and adhered to slides as described (15). Cultured cells were cytospined as described (15).

Single staining. Anti-CD90 (clone 5E10, 1:50; BD Biosciences), anti–{alpha}-smooth muscle actin ({alpha}-SMA) (clone 1A4, 1:50; R&D Systems), anti-vimentin antibody (V9 clone, 1:3,000; Dako), anti–Ki-67 (Ki-67 clone, 1:50; Dako), and anti–pan-cytokeratin antibody (AE1/AE3 clone, 1:500; Dako) were used as primary antibodies. After washing, the following secondary antibodies were used: Envision+ polymer (ready to use; Dako) was used for Ki-67, CD-90, and {alpha}-SMA staining, whereas goat anti-mouse IgG1 (Southern Biotech) was used for pan-cytokeratin and vimentin staining.

Double staining. Both anti-Ki-67 (Polyclonal, Vector Laboratories) and anti-vimentin antibodies were added together. Envision + polymer with swine anti-rabbit IgG (Dako) was used as secondary antibodies. Color was developed with 3,3'-diaminobenzidine (Novocastra), Fuschin Red (DAKO), or AEC (Sigma), and instant hematoxylin (Shandon) was used for counterstaining.

UV light and {gamma}-ray treatments. For UV irradiation, the medium was removed and the cell culture monolayers in dishes were covered with PBS and exposed to a germicidal UV lamp (254 nm) at fixed distance. The UV dosimetry was done using a UV meter (Spectronics Corporation). {gamma}-Radiation was done using Cobalt (Co) source at a dose rate of 0.60 Gy/min.

Cellular lysate preparation and immunobloting. This has been done as previously described (16). Antibodies directed against p53 (DO-I), p21 (187), survivin (C-19), glyceraldehyde-3-phosphate dehydrogenase (GAPDH; FL-335), proliferating cell nuclear antigen (PCNA; PC-10), and β-actin (C-11) were purchased from Santa Cruz.

Quantification of protein expression level. The expression levels of the immunoblotted proteins were measured using the densitometer (BIO-RAD GS-800 Calibrated Densitometer) as previously described (16).

Cell death analysis by flow cytometry. Cells were challenged either with UVC (30 J m–2) or with cisplatin, whereupon the monolayers were incubated in DMEM with supplements. Detached and adherent cells were then harvested after 72 h and processed as previously described (17).

Cell proliferation analysis. Complete medium (100 µL) containing 2 to 4 x 103 cells was loaded in each well of the 96-well plate. The plate was incubated for at least 30 min in a humidified (37°C) 5% CO2 incubator and then was inserted into the real-time cell electronic sensing (RT-CES) system (ACEA Biosciences, Inc.). Cell proliferation was monitored for 70 h.

Two-dimensional gel electrophoresis, scanning, and image analysis. Proteins (100 µg) were loaded on each strip via rehydration using linear (pH 4-7 ready IPG) strips (Bio-Rad). Two-dimensional gel electrophoresis (2-DE) was done, and gels were stained with silver nitrate and scanned using a laser densitometer. Data were analyzed using PDQUEST software (Bio-Rad) as previously described (18, 19).

Data preprocessing and statistical analysis. A difference of ≥2-fold change was used as a threshold for marked quantitative difference between pairs of samples. In addition to qualitative differences (on/off protein spots), significantly differentially expressed protein spots were first selected using two different statistical methods (Student's t test and partial least square analysis) available in PDQuest 2-DE image analysis software. The data from the match set were exported from PDQUEST in the form of data table, with rows representing gels and columns representing spots. Datasets were normalized before analysis (19, 20). The resulting data were subjected to hierarchical clustering analysis using the J Express software.5


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Isolation of CAFs and their adjacent counterparts from invasive human breast cancers. Fibroblast cells were extracted from 12 human invasive mammary ductal carcinomas obtained from mastectomies. The tissues were dissociated, and various cell types were separated to obtain CAFs. We also isolated from each of the same 12 tissues a second fibroblast populations taken from a histologically noncancerous region of the breast at least 2 cm away from the outer tumor margin (tumor counterpart fibroblasts, TCF). We then verified the purity of the fibroblastic populations by cell morphology under microscope and cytospin/immunostaining. Figure 1A shows the typical morphology of fibroblasts. Furthermore, these cells strongly expressed the fibroblastic marker vimentin, whereas they were negative for cytokeratin (a marker for epithelial cells; Fig. 1A). This indicates that fibroblastic cells were well separated from the other type of cells and, therefore, are considered to be highly homogeneous with minimal contamination.


Figure 1
View larger version (71K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Breast CAFs and their adjacent counterparts are myofibroblasts. A, in vitro cultured fibroblasts were first visualized under a microscope and then the cytospin attached cells were immunostained with the indicated antibodies. Hematoxylin was used for counterstaining. Photomicrographs are at magnification of 260x. B, immunohistochemistry of frozen sections and cell cytospin were done using the indicated antibodies and cells. The circles indicate a region rich in fibroblastic cells.

 
CAFs and TCFs express the myofibroblast {alpha}-SMA marker. During the tumorigenesis process, stromal fibroblasts acquire some of the characteristics of smooth muscle cells that specifically express {alpha}-SMA. Using anti–{alpha}-SMA antibody, we have found that all the CAF and TCF cells examined were positive for this protein. However, {alpha}-SMA was almost undetectable in the fibroblastic cells either in culture or in tissue from plastic surgery (normal breast fibroblasts 6, NBF6; Fig. 1B). The proportion of positive cells varied from 5% to 100% for CAFs, with an average of 42% and from 5 to 90% for TCFs with an average of 53%. In recent studies, a link between the expression of {alpha}-SMA and the cell surface glycoprotein CD90 (Thy1) was suggested (21). CD90 is expressed at different levels by fibroblasts from different organs (22). Figure 1B shows that all CAF, TCF, and NBF cells and tissue were positive for CD-90 protein. The detection of this protein in normal NBF cells indicates that most breast fibroblasts are also expressing this protein and that there is no direct link between the expression of CD-90 and the transformation into myofibroblasts during breast carcinogenesis.

The p53/p21 DNA damage signaling pathway is defective in most CAFs but not in their adjacent counterparts. Next, we studied the effect of UV light and {gamma}-rays on the up-regulation of p53 and p21 proteins to determine their functional status and also the status of their upstream effectors. Ten CAFs/TCFs pairs and two NBFs were irradiated with UV light (5 J m–2) and {gamma}-rays (5 Gy). Irradiated cells were reincubated for different periods of time, and then whole-cell extracts were prepared and used in immunoblot analysis using specific p53 and p21 antibodies and GAPDH as internal control. The used doses were previously shown to induce both p53 and p21 proteins in breast fibroblast cells (23). The increase in the level of these proteins was considered as induction only when the protein level after the treatment became at least twice higher than in the control nonirradiated cells. Figure 2A shows that p53 and p21 protein levels augmented in response to both UV-light and {gamma}-rays in NBF6 cells. The maximum inductions of p53 (~4.5-fold) and p21 (~3-fold) were reached 14 h post–UV irradiation. In response to {gamma}-rays, the maximum levels of induction for p53 (~3-fold) and p21 (4.5-fold) were reached 3 hours after the treatment (Fig. 2A). Similar results were obtained with the NBF2 cells (data not shown).


Figure 2
View larger version (28K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. p53 and p21 expression and response to UV light and {gamma}-rays in CAFs and TCFs. A and B, confluent cells were either mock treated or challenged with UV light (5 J m–2) or {gamma}-rays (5 Gy) and then reincubated for different periods of time, as indicated. After cell lysate preparation, 30 µg of proteins were used for immunoblot analysis using the indicated antibodies. C, whole-cell lysates were prepared from the indicated cells and used for immunoblot analysis using the indicated antibodies. D, histograms showing p53 and p21 expression levels. The values were determined by densitometry and normalized against GAPDH. N1 and N6 correspond to NBF1 and NBF6. Error bars represent SDs of at least three different experiments.

 
Figure 2B shows also that, like in the NBF6 cells, p53 and p21 protein levels were up-regulated in response to both UV light and {gamma}-rays in the TCF142 cells. Comparable results were obtained in all the other TCF cells (data not shown). On the other hand, 7 of 10 CAFs (70%) exhibited a defect in the induction of p53 and its downstream effector p21 in response to {gamma}-rays. Indeed, p53 and p21 were up-regulated in CAF87 (Fig. 2B), CAF169, and CAF180 (data not shown) in response to both UV light and {gamma}-rays. However, p53 and p21 were induced only after UV light in the seven other CAFs (64, 76, 84, 114, 118, 142, 153). By contrast, following {gamma}-rays, these cells showed neither p53 nor p21 induction (Fig. 2B and data not shown).

These results indicate that the p53/p21 response pathway to {gamma}-rays is defective in most CAFs but normal in all their corresponding counterparts.

Basal expression of p21 and p53 proteins in CAFs and TCFs. Next, the basal expression levels of p21 and p53 proteins were assessed in 12 pairs CAFs/TCFs and two NBFs. Interestingly, in 10 of 12 cases (83%), the level of p53 was significantly lower in CAFs than in the normal NBFs, and eight CAFs (66%) showed low p53 level compared with their corresponding TCFs. However, the level was similarly low in both CAF/TCF169 and 84. On the other hand, CAF180 exhibited higher p53 level than in its corresponding TCF180 and the NBF cells (Fig. 2C and D). However, only four TCFs (69, 84, 118, and 169) showed significantly low p53 levels. On the other hand, p53 level was higher in five TCFs compared with the NBF cells (Fig. 2C and D). These results indicate that in most cases the level of p53 is lower in CAFs compared with their adjacent TCFs and the normal controls, and that the p53 expression is also modulated in the majority of the TCFs.

For p21, 10 of 12 CAFs (83%) exhibited low p21 level than in the NBF cells and 5 of 12 CAFs (41%) showed lower expression than in their corresponding TCFs. However, p21 level was similarly low in four cases of CAFs and TCFs and was higher in the CAF142 and CAF180 than in their corresponding TCFs (Fig. 2C and D). Interestingly, the level of p21 was significantly lower in 6 of 12 (50%) TCFs and higher in TCF153 than in the control NBF cells (Fig. 2C and D). Therefore, p21 expression is low in most CAFs and also in a significant proportion of TCFs.

Interestingly, a strong correlation exists between the expression levels of p53 and p21 proteins in these cells (Fig. 2C), which parallels the well known control of p21 expression by p53.

Survivin is highly expressed in both CAFs and TCFs. To see whether these fibroblasts display cancer-associated features, we sought to explore the expression level of survivin, the most important cancer-associated protein (24, 25). To this end, whole-cell extracts were prepared from 12 pairs CAFs/TCFs and three NBFs were used as normal control. Specific anti-survivin and anti–β-actin (used as internal control) were used for immunoblotting analysis. As expected, the level of survivin was very low to undetectable in the normal NBF fibroblasts (Fig. 3A ). However, survivin levels were higher in 100% CAFs and TCFs compared with the level detected in the three NBF cells. The increase in survivin expression was 2-fold in the CAF/TCF69 and reached 17-fold in the CAF/TCF114 (Fig. 3B). Interestingly, in 9 of 11 cases, the levels of survivin were comparable in CAFs and their corresponding adjacent fibroblasts (Fig. 3A and B). This shows that like for cancer cells, CAFs and their adjacent counterpart fibroblasts also express high levels of the cancer marker survivin.


Figure 3
View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Survivin is highly expressed in both CAFs and their corresponding TCFs. Whole-cell extracts were prepared from confluent cells and used for immunoblotting analysis using the indicated antibodies. A, immunoblots. B, histogram showing survivin expression levels. The values were determined by densitometry and normalized against β-actin. A and C correspond respectively to CAF and TCF, whereas N1, N2, and N3 correspond respectively to NBF1, NBF2, and NBF3. Error bars represent SDs of at least three different experiments.

 
Response of CAFs and TCFs to cisplatin, UV light, and {gamma}-rays. Because survivin is expressed in CAF/TCF cells and because it is a potent inhibitor of apoptosis, we sought to investigate the response of these fibroblasts to cisplatin, UV light, and {gamma}-rays. To this end, 10 CAFs and their corresponding TCFs as well as two NBF cells were treated with cisplatin (30 µg/mL), UV light (30 Jm–2), or {gamma}-rays (30 Gy) and then reincubated for 72 hours. Subsequently, cell death was assessed by flow cytometry (Fig. 4A ). Figure 4A and B show that the normal control cells NBF are sensitive to cisplatin, which triggered cell death in ~40% of these cells. However, 9 of 10 CAFs (90%) and 8 of 10 TCFs (80%) showed very high resistance to cisplatin. The proportions of cell death induced in these fibroblasts were between 0% and 5% (Fig. 4B). The TCF76 cells were more sensitive with 38% killed with the cisplatin treatment (Fig. 4B). On the other hand, the CAF/TCF114 pair exhibited very high sensitivity, with more that 70% cell death (Fig. 4B).


Figure 4
View larger version (30K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. Response of CAFs and TCFs to the killing effect of cisplatin, UV light, and {gamma}-rays. Subconfluent cells were either mock-treated or challenged with cisplatin (30 µg/mL), UV light (30 J m–2), or {gamma}-rays (30 Gy) and then reincubated for 72 h. Cell death was measured using flow cytometry. A, FACScan analysis. B-D, histograms showing the proportions of induced cell death. N2 and N6 correspond to NBF2 and NBF6.

 
In response to UV light, ~35% of the normal NBF cells underwent cell death (Fig. 4A and C). However, 9 of 10 CAF cells (90%) exhibited high resistance to the killing effect of UV light (Fig. 4C). Importantly, in 7 of 10 pairs, the CAF cells were significantly more resistant than their counterparts TCF cells. Indeed, CAFs 180, 84, 118, and 148 showed very high resistance to UV light with 1% to 2% cell death (Fig. 4C). On the other hand, only 50% of the TCFs were significantly resistant to UV light compared with the control NBF cells, whereas the other five TCFs were only slightly sensitive (Fig. 4C). This shows that in most cases the TCF cells showed an intermediate response between the normal fibroblasts and the CAFs.

Concerning the response to {gamma}-rays, all the cells including the normal NBFs exhibited very high resistance (0–5% cell death), with the exception of the CAF/TCF87 pair, wherein the proportion of cell death reached 18% (Fig. 4D). Nevertheless, CAF and TCF cells, which showed similar responses, were slightly more sensitive than the NBF cells to the killing effect of {gamma}-rays.

Proliferation rate and expression of PCNA and Ki-67 in CAFs and TCFs. It was very obvious that CAFs grow faster than their corresponding TCFs and the control normal NBF cells during in vitro culturing. Thereby, we used the real-time cell electronic sensing system to study cell proliferation and show the growing difference. We have found that the CAF114 and CAF118 grow four to five times faster than their corresponding counterparts TCFs (Fig. 5A ). Similar results were obtained with other CAF/TCF pairs (data not shown), showing that the CAF cells acquired faster proliferation rates compared with their adjacent counterpart fibroblasts. To further elucidate this, the level of the proliferation marker Ki-67 was assessed in different CAF/TCF pairs. Figure 5B shows that the Ki-67 protein was expressed at very low levels in the TCF fibroblasts. By contrast, most of the CAF cells were positive for the proliferation antigen, although to different extents (Fig. 5B). Similar results were found in breast cancer tissues, where Ki-67 was found expressed in six of seven CAFs, but was undetectable in their corresponding TCFs (data not shown). Figure 5C shows the expression of Ki-67 in epithelial cells and their corresponding stromal fibroblasts from the same tissue. The expression of Ki-67 in both cell types shows that the increase in the proliferation rate is not restricted to breast carcinomas, but it is also present in their stromal fibroblasts.


Figure 5
View larger version (65K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5. Cell proliferation and the expression of Ki-67 and PCNA in NBF, TCFs, and CAFs. A, 2,000 cells were seeded on 96-well plates and were incubated for the indicated periods of time. Cell proliferation rate was determined using the real-time cell electronic sensing system. B, in vitro cultured fibroblast cells were cytospin attached to slides and stained by Ki-67 antibody. Hematoxylin was used for counterstaining. Representative immunohistochemical staining of Ki-67 (brown, nuclear) of the indicated cell cultures is shown. Photomicrographs are at x320. C, immunohistochemistry using anti–Ki-67 antibody on breast cancer tissue. The arrows show the stained epithelial and fibroblast cells. D, 30 µg of cell lysates were immunoblotted using the indicated antibodies. A and C correspond to CAF and TCF, respectively. PCNA corresponding levels are indicated below the corresponding bands. Representative of two experiments.

 
Next, we assessed the level of the other proliferation marker PCNA and showed that PCNA levels were higher in both CAFs and TCFs compared with the control NBF1 and NBF6 cells (Fig. 5D). When comparing the PCNA level in each CAF/TCF pair, we have found that in most cases PCNA expression is only slightly higher in CAF than in TCF cells (Fig. 5D). Interestingly, the PCNA levels were found to be considerably higher in CAF87 and CAF142 compared with their corresponding TCF cells (Fig. 5D). On the other hand, the CAF180 exhibited significantly lower level of PCNA than the TCF180 (Fig. 5D).

Protein expression patterns of CAFs, TCFs, and NBF. Whole-cell lysates were prepared from one normal (NBF6) and two CAF/TCF pairs (114 and 76) and were analyzed by 2-DE. Representative 2-DE maps are shown in Fig. 6A . An average total number of 969 spots were resolved, and minimum of 84% of the spots were matched between all the gels. Interestingly, marked quantitative and qualitative changes were observed in the protein expression pattern between NBF, TCF, and CAF samples. Using the correlation analyses between pairs of samples, we observed high degree of similarities between the two CAFs (r = 0.71) and the two TCFs (r = 0.81). On the other hand, an average correlation coefficient of 0.61 was observed between pairs of TCF and CAF cells. Similar correlation was observed between NBFs and TCFs (0.61). However, we observed higher degree of heterogeneity in the protein expression pattern between the normal control NBF and CAFs, with a correlation of only 0.49.


Figure 6
View larger version (45K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6. Protein expression patterns of CAF, TCF, and NBF cells. Whole-cell lysates were prepared and subjected to 2-DE using readymade IPG strips (pH 4–7) in the first and 12.5% homogeneous SDS PAGE in the second dimension. A, two-dimensional maps showing representative examples of 2-DE gels. B, left, cluster analysis of the indicated samples using expression dataset from 15 polypeptides derived from Boolean analyses of t test and PLS; right, correspondence analysis plot of the same dataset. C, gel segments showing six protein spots that are differentially expressed between the different indicated samples.

 
Hierarchical cluster analysis of differentially expressed proteins in the NBF, TCF, and CAF samples. The differences in protein expression patterns between the matched spots in the different gels were studied. A total of 483 spots were matched to all the five samples. In an effort to reduce the dataset, multivariate data analysis of qualitative and quantitative differences were carried out to generate sets of variables for more accurate use in the clustering analysis. Using Student's t test, a total of 95 polypeptides were significantly (P < 0.05 with 98% confidence interval) differentially expressed between CAFs and TCFs. A similar analysis using PLS resulted in 25 protein spots that differed significantly.

Subsequently, we used these 95 and 25 separate datasets for possible classification of the samples into their respective groups using the hierarchical cluster analysis, and all the samples were correctly classified (data not shown). A total of 15 protein spots fall in the intersection of the above two datasets and were used in the cluster analysis of all the samples (Fig. 6B). The same dataset was subjected to correspondence analysis (Fig. 6B), and all samples were distinctly separated. These analysis showed differential protein expression pattern between the normal cells (NBF), TCFs, and CAFs, with the NBF and TCFs sharing more similarities in their protein expression patterns (Fig. 6B).

The differential expressions of some of these protein spots are shown in Fig. 6C. This figure shows the expression levels of six different proteins. Panels 1, 2, 3, and 4 show that these proteins completely disappeared in the CAF cells, whereas they were present in the corresponding TCF cells and also in the normal NBF cells (Fig. 6C). On the other hand, panels 5 and 6 showed the overexpression of proteins that were undetectable in the NBF cells and appeared in the TCF cells (Fig. 6C). Together these figures show that simultaneous analysis of multiple polypeptides by 2-DE in combination with hierarchical cluster analysis enabled clear discrimination between CAFs, TCFs, and the NBFs.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
In the present report, we provide strong evidence that CAFs, as well as their adjacent counterparts, show several tumor-specific features. Indeed, both are active myofibroblasts that express high levels of PCNA and survivin, they are highly resistant to the killing effect of cisplatin, most of them express low levels of the tumor suppressor p53 and p21 proteins, and their protein expression profiles are different from that of NBF cells. Interestingly, most CAFs but not TCF cells were defective in the {gamma}-ray–dependent induction of p53 and p21 proteins. Furthermore, like tumor cells, CAFs expressed high levels of Ki-67 and exhibited high proliferation rates. These results provide the first clear indication that TCFs, although histologically normal and are not in contact with cancer cells, bear genetic changes that distinct them from NBFs and also from their neighbor CAF cells.

We have shown here that 70% of CAF cells are defective in the {gamma}-ray–dependent up-regulation of p53 and its downstream effector p21, which are normally induced after UV light (Fig. 2B). This may suggest a defect in the p53-upstream effectors that govern the {gamma}-ray signaling pathway, such as Atm and Chk2, which play important roles in breast cancer (26–28). Interestingly, it has been recently shown that the presence of an association between the existence of p53 mutations and loss of heterozygosity/allelic imbalance at the ATM locus in the stromal compartment of breast cancer and tumor grade (29). Furthermore, Kiaris et al. have shown an association between p53 mutations in the stromal component of epithelial tumors and carcinogenesis (12). Together, these findings show the importance of the ATM/p53 tumor suppressor pathway in the stromal compartment as well. The fact that this pathway is normal in the adjacent counterpart fibroblasts suggests that the defect in the {gamma}-ray–dependent induction of p53 that has been observed in CAFs resulted from a selective pressure from the tumor cells. In addition, the levels of p53 and p21 decreased in 80% CAFs compared with NBF cells. However, only 33% and 50% of TCF cells showed lower levels of p53 and p21 proteins, respectively. These findings present the first indication that p53 and p21 expression levels decrease in CAFs and their adjacent counterparts, which could have a link with the development/progression of carcinomas. Indeed, in a mouse model of prostate cancer, it has been recently shown that initiating tumorigenesis in the epithelium by pRB inactivation led to the loss of p53 expression in stromal tumor fibroblasts and increased mesenchymal cell proliferation and tumor progression. Subsequently, some epithelium regions have also lost p53 expression (30). Furthermore, it has been recently shown that p53 suppresses the production of the chemokine Sdf1 in cultured fibroblasts (31). Because the secretion of this protein by stromal fibroblasts contributes to tumor promotion (10), a decrease in p53 level in CAFs will lead to an increase in the production and secretion of Sdf1 and, therefore, enhances tumor formation. This explains the decrease in p53 levels in most CAFs examined in this study, which further prove the nonautonomous effect of p53 tumor suppressor in carcinogenesis. A decrease in the level of the universal cyclin-dependent kinase inhibitor p21 may also have a tremendous effect on tumorigenesis because it is a modulator of both cell cycle and apoptosis and also controls the expression of various cancer-related genes (32–34).

Furthermore, we have shown that the level of the survivin protein is similarly higher in CAFs and TCFs compared with the normal NBF cells (Fig. 3). The increase in survivin expression is known to occur in most cancer cells and tissues as a consequence of activation of oncogenes or loss of tumor suppressor genes (24, 25). This suggests that CAFs, as well as their corresponding TCFs, acquired tumor-like changes that are most likely necessary for tumor growth. Because the survivin level is higher in both CAFs and their counterparts than in the normal controls, it is possible that survivin up-regulation occurs early and has a promoting role during cancer development. Therefore, the increase in the survivin level could constitute an important cancer predisposing factor, especially with its dual role in inhibiting apoptosis and activating cell proliferation. Indeed, Temme et al. have reported that overexpression of survivin increased human fibroblast proliferation (35).

Interestingly, like for cancer cells, the increase in the level of this antiapoptosis protein was accompanied with an increase in the resistance of these cells to the killing effect of UV light and cisplatin. Whereas CAFs and TCFs were similarly highly resistant to cisplatin, most CAFs showed higher resistance to UV light than their counterparts TCFs. To further elucidate the tumor-like phenotype of these stromal fibroblasts, we have shown that CAF cells and tissues express high levels of the proliferation marker Ki-67 and showed enhanced proliferation than their adjacent counterparts TCFs. However, PCNA was found to be highly expressed in both CAFs and TCFs. This explains the higher ability of these cells in proliferating and may be their predisposition to receive signals from cancer cells.

Finally, using 2-DE and hierarchical cluster analysis, we classified CAFs, TCFs, and NBF cells into three distinct groups. This shows that CAFs and TCFs are different from each other and also different from NBFs. These results also indicated that TCFs are closer to the NBF than the CAFs. Indeed, the analysis of individual spots showed various proteins in the TCFs that exhibited a level of intermediate expression between the NBF and the CAFs (Fig. 6C). These results provide the first clear indication that the proteome of the TCFs is not normal but also different from that of the CAFs, which corroborates our findings regarding the basal and induced level of p53/p21, the response to UV light, and cell proliferation.

These findings support the importance of analysis of stromal and tumor-adjacent tissue proteome in the discovery of potential biomarkers related to breast carcinomas that may have prognostic and/or predictive values. Indeed, the presence of significant associations between loss of heterozygosity/allelic imbalance in the breast cancer stroma and tumor grade and regional lymph node metastasis has been recently shown. In fact there were more correlation between clinicopathologic features and loss of heterozygosity/allelic imbalance in the stroma than in the epithelium (29). Therefore, the identification of some of the differentially expressed proteins in the stroma may shed light on the initiation and progression of breast tumor and further highlights their potential usefulness as marker of malignancy.


   Acknowledgments
 
Grant support: King Abdulaziz City for Science and Technology. This work was performed under the RAC proposal 2031091.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Dr. K. Abu-Khabar and L. Al-Haj for their help with the RT-CES system and Dr. K. Al-Hussein and P.S. Manogaran for their help with the flow cytometry.


   Footnotes
 
Note: Present address for N.M. Hawsawi: Department of Pathology, University of Newcastle upon Tyne, Royal Victoria Infirmary, Queen Victoria Road, Newcastle upon Tyne NE1 4LP, United Kingdom.

5 http://java.sun.com Back

Received 1/16/08. Revised 1/28/08. Accepted 1/28/08.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Smigal C, Jemal A, Ward E, et al. Trends in breast cancer by race and ethnicity: update 2006. CA Cancer J Clin 2006;56:168–83.[Abstract/Free Full Text]
  2. van den Hooff A. Connective tissue changes in cancer. Int Rev Connect Tissue Res 1983;10:395–432.[Medline]
  3. Mueller MM, Fusenig NE. Friends or foes-bipolar effects of the tumour stroma in cancer. Nat Rev Cancer 2004;4:839–49.[CrossRef][Medline]
  4. Bhowmick NA, Moses HL. Tumor-stroma interactions. Curr Opin Genet Dev 2005;15:97–101.[CrossRef][Medline]
  5. Kunz-Schughart LA, Knuechel R. Tumor-associated fibroblasts (part I): active stromal participants in tumor development and progression? Histol Histopathol 2002;17:599–621.[Medline]
  6. Sappino AP, Schurch W, Gabbiani G. Differentiation repertoire of fibroblastic cells: expression of cytoskeletal proteins as marker of phenotypic modulations. Lab Invest 1990;63:144–61.[Medline]
  7. Beacham DA, Cukierman E. Stromagenesis: the changing face of fibroblastic microenvironments during tumor progression. Semin Cancer Biol 2005;15:329–41.[CrossRef][Medline]
  8. Kalluri R, Zeisberg M. Fibroblasts in cancer. Nat Rev Cancer 2006;6:392–401.[CrossRef][Medline]
  9. Elenbaas B, Weinberg RA. Heterotypic signaling between epithelial tumor cells and fibroblasts in carcinoma formation. Exp Cell Res 2001;264:169–84.[CrossRef][Medline]
  10. Orimo A, Gupta PB, Sgroi DC, et al. Stromal fibroblasts present in invasive human breast carcinomas promote tumor growth and angiogenesis through elevated SDF-1/CXCL12 secretion. Cell 2005;121:335–48.[CrossRef][Medline]
  11. Allinen M, Beroukhim R, Cai L, et al. Molecular characterization of the tumor microenvironment in breast cancer. Cancer Cell 2004;6:17–32.[CrossRef][Medline]
  12. Kiaris H, Chatzistamou I, Trimis G, et al. Evidence for nonautonomous effect of p53 tumor suppressor in carcinogenesis. Cancer Res 2005;65:1627–30.[Abstract/Free Full Text]
  13. Albini A, Sporn MB. The tumour microenvironment as a target for chemoprevention. Nat Rev Cancer 2007;7:139–47.[Medline]
  14. Joyce JA. Therapeutic targeting of the tumor microenvironment. Cancer Cell 2005;7:513–20.[CrossRef][Medline]
  15. Ghebeh H, Tulbah A, Mohammed S, et al. Expression of B7-1 in breast cancer patients is strongly associated with high proliferative Ki-67-expressing tumor cells. Int J Cancer 2007;121:751–8.[CrossRef][Medline]
  16. Al-Mohanna MA, Al-Khalaf HH, Al-Yousef N, Aboussekhra A. The p16INK4a tumor suppressor controls p21WAF1 induction in response to ultraviolet light. Nucleic Acids Res 2007;35:223–33.[Abstract/Free Full Text]
  17. Al-Mohanna MA, Manogaran PS, Al-Mukhalafi Z, K AA-H, Aboussekhra A. The tumor suppressor p16(INK4a) gene is a regulator of apoptosis induced by ultraviolet light and cisplatin. Oncogene 2004;23:201–12.[CrossRef][Medline]
  18. Alaiya AA, Franzen B, Fujioka K, et al. Phenotypic analysis of ovarian carcinoma: polypeptide expression in benign, borderline and malignant tumors. Int J Cancer 1997;73:678–83.[CrossRef][Medline]
  19. Alaiya AA, Franzen B, Hagman A, et al. Classification of human ovarian tumors using multivariate data analysis of polypeptide expression patterns. Int J Cancer 2000;86:731–6.[CrossRef][Medline]
  20. Alaiya AA, Franzen B, Hagman A, et al. Molecular classification of borderline ovarian tumors using hierarchical cluster analysis of protein expression profiles. Int J Cancer 2002;98:895–9.[CrossRef][Medline]
  21. Koumas L, Smith TJ, Feldon S, Blumberg N, Phipps RP. Thy-1 expression in human fibroblast subsets defines myofibroblastic or lipofibroblastic phenotypes. Am J Pathol 2003;163:1291–300.[Abstract/Free Full Text]
  22. Baglole CJ, Ray DM, Bernstein SH, et al. More than structural cells, fibroblasts create and orchestrate the tumor microenvironment. Immunol Invest 2006;35:297–325.[CrossRef][Medline]
  23. Meyer KM, Hess SM, Tlsty TD, Leadon SA. Human mammary epithelial cells exhibit a differential p53-mediated response following exposure to ionizing radiation or UV light. Oncogene 1999;18:5795–805.[CrossRef][Medline]
  24. Ambrosini G, Adida C, Altieri DC. A novel anti-apoptosis gene, survivin, expressed in cancer and lymphoma. Nat Med 1997;3:917–21.[CrossRef][Medline]
  25. Altieri DC. Survivin, versatile modulation of cell division and apoptosis in cancer. Oncogene 2003;22:8581–9.[CrossRef][Medline]
  26. Ahmed M, Rahman N. ATM and breast cancer susceptibility. Oncogene 2006;25:5906–11.[CrossRef][Medline]
  27. Renwick A, Thompson D, Seal S, et al. ATM mutations that cause ataxia-telangiectasia are breast cancer susceptibility alleles. Nat Genet 2006;38:873–5.[CrossRef][Medline]
  28. Sullivan A, Yuille M, Repellin C, et al. Concomitant inactivation of p53 and Chk2 in breast cancer. Oncogene 2002;21:1316–24.[CrossRef][Medline]
  29. Fukino K, Shen L, Patocs A, Mutter GL, Eng C. Genomic instability within tumor stroma and clinicopathological characteristics of sporadic primary invasive breast carcinoma. JAMA 2007;297:2103–11.[Abstract/Free Full Text]
  30. Hill R, Song Y, Cardiff RD, Van Dyke T. Selective evolution of stromal mesenchyme with p53 loss in response to epithelial tumorigenesis. Cell 2005;123:1001–11.[CrossRef][Medline]
  31. Moskovits N, Kalinkovich A, Bar J, Lapidot T, Oren M. p53 attenuates cancer cell migration and invasion through repression of SDF-1/CXCL12 expression in stromal fibroblasts. Cancer Res 2006;66:10671–6.[Abstract/Free Full Text]
  32. Chang BD, Watanabe K, Broude EV, et al. Effects of p21Waf1/Cip1/Sdi1 on cellular gene expression: implications for carcinogenesis, senescence, and age-related diseases. Proc Natl Acad Sci U S A 2000;97:4291–6.[Abstract/Free Full Text]
  33. Dotto GP. p21(WAF1/Cip1): more than a break to the cell cycle? Biochim Biophys Acta 2000;1471:M43–56.[Medline]
  34. Gartel AL, Tyner AL. The role of the cyclin-dependent kinase inhibitor p21 in apoptosis. Mol Cancer Ther 2002;1:639–49.[Abstract/Free Full Text]
  35. Temme A, Diestelkoetter-Bachert P, Schmitz M, et al. Increased p21(ras) activity in human fibroblasts transduced with survivin enhances cell proliferation. Biochem Biophys Res Commun 2005;327:765–73.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 Cancer Res.Home page
Y.-L. Zheng, N. Hu, Q. Sun, C. Wang, and P. R. Taylor
Telomere Attrition in Cancer Cells and Telomere Length in Tumor Stroma Cells Predict Chromosome Instability in Esophageal Squamous Cell Carcinoma: A Genome-Wide Analysis
Cancer Res., February 15, 2009; 69(4): 1604 - 1614.
[Abstract] [Full Text] [PDF]

Home page
 Cancer Res.Home page
P. J. Mishra, P. J. Mishra, J. W. Glod, and D. Banerjee
Mesenchymal Stem Cells: Flip Side of the Coin
Cancer Res., February 15, 2009; 69(4): 1255 - 1258.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hawsawi, N. M.
Right arrow Articles by Aboussekhra, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hawsawi, N. M.
Right arrow Articles by Aboussekhra, A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online